University of Ghana  http://ugspace.ug.edu.gh
The Prevalences of Human Papillomavirus Infections and Cervical Lesions, and 
the Patterns of Risk Behaviour Characteristics of Women in a High Risk 
Community: Akuse sub-district, Ghana 
A Thesis Submitted by 
AdolfKofi Awua (BSc, MPhil.) 
(Student Identification Number: 10019029) 
Of 
The Department of Epidemiology and Disease Control, 
School of Public Health 
To 
THE SCHOOL OF GRADUATE STUDIES OF THE UNIVERSITY OF 
GHANA, LEGON, IN FULFILMENT OF THE REQUIREMENTS FOR THE 
A WARD OF A DOCTOR OF PHILOSOPHY PUBLIC HEALTH DEGREE. 
JUNE 2014 
[j] SCHOOL OF PUBLIC 
~ HEALTH LIBRARY 
::::'~ LEGON 
University of Ghana  http://ugspace.ug.edu.gh
DECLARATION 
I hereby declare that the work described in this thesis was performed by me, Adolf Kofi Awua, 
under the supervision of Professor R.M. Adanu - Dean of the School of Public Health, College of 
Health Sciences, University of Ghana, Legon; Professor E. K. Wiredu - Pro-Vice Chancellor, 
University of Health and Allied Sciences, Ho and Professor Col. (Rtd) E. A Afari - Department 
of Epidemiology and Disease Control, School of Public Health, University of Ghana, Legon. To 
the best of my knowledge, this thesis has not been submitted for any academic degree in any 
institution. 
All references cited have been duly acknowledged. 
CANDIDATE 
AdolfKofi Awua 
SUPERVISORS 
Professor R. M Adanu ____- ---=~ _ _______  
Professor E. K Wiredu- -------------
Professor E. A Afari 
ii 
University of Ghana  http://ugspace.ug.edu.gh
DEDICATION 
TO GOD AND MY FAMILY 
III 
University of Ghana  http://ugspace.ug.edu.gh
ACKNOWLEDGMENT 
I give glory to the Almighty God, Who gave me the strength and drive to complete this study and 
very grateful to all you wonderful persons in my life who have in diverse ways supported me 
during this stage of my life. My sincere thanks go to my supervisors, Professor R.M Adanu, 
Dean of the School of Public, University of Ghana; Professor E. K. Wiredu, Pro-Vice Chancellor 
of the University of Health and Allied Sciences and Professor Col. (Rtd) E.K Afari, Department 
of Epidemiology and Disease Control, School of Public Health, University of Ghana. Your 
guidance, suggestions, contributions and support have been valuable to the successful 
completion of this study and my PhD programme. I am also very grateful to all the partners 
(Professor Fred Binka, Professor Thomas Junghanss, Professor Gerd Pluschke, and Professor 
Oliver Razum) and Coordinators (Professor R.M Adanu, Dr. M Keaser and Dr. Fenna 
Veeltmann) of the German-Ghanaian Centre for Health Research Project and the German 
Academic Exchange Programme (DAAD) for the opportunities and financial support provided 
on this programme. I thank the Head (Dr. Patricia Akweongo) and Staff of the Epidemiology and 
Disease Control Department, for all their Administrative and other support during my study the 
Administrators and Staff of the Akuse Government Hospital , particularly Dr. S. Tijani (Medical 
Superintendent), Dr. Dzanmah, Miss Yabaa Essien, Miss Gifty Tekushie and the staff of the 
Public Health Unit as well as staff of the Reproductive and Child Health Unit for hosting and 
supporting the study in diverse ways. I gracefully acknowledge the contributions of the Chiefs, 
Religious and Community leaders who accepted the study in the communities and encouraged 
the participation of the women in this study. A special thank you goes to Mr. Richard H. Asmah 
of the School of Biomedical and Allied Health Sciences (BMSAHS) College of Health Sciences, 
IV 
University of Ghana  http://ugspace.ug.edu.gh
University of Ghana, for coordinating the purchase of reagents, Specimen collection kits and the 
Pap smear test at the Pathology Laboratory of the School of Biomedical and Allied Health 
Sciences. I am grateful to Mr. Abdul Rashid Adams and Ms. Dinah Nkansah of the BMSAHS 
Pathology Laboratory for their assistance with the cytology analysis. I sincerely acknowledge the 
contribution of Dr. Alterto Severini, Head of the Viral Sexually Transmitted Disease Section and 
the National Microbiology Laboratory (NML) of Canada for hosting and funding the molecular 
laboratory analysis of this study and of Mrs. Vanessa Zubuch for her assistance with the bench 
work. ] am thankful to the other Staff of Viral Sexually Transmitted Disease Section of the NML 
for their support and all they did that made me feel at home in Canada. I am also thankful to Mr. 
Meindert Dzwart, CEO of Rovers Medical Devices for supporting this study with a donation of 
the Rover® Viba-Brush vaginal sampler, the self-collection devices used in this study. I am 
thankful to Mr. K Mensah of the Malaria Control Programme for his contributions to the design 
of the study. To my colleagues, friends and persons who have provided peer review and 
supported me in any way and particularly to Edna Dzifa Doe and Stella Dzifa Monu for all their 
encouragement, support and attention to all my complaints and frustration during this study, I say 
thank you very much. 
God Bless you all for your valuable contributions. 
v 
University of Ghana  http://ugspace.ug.edu.gh
TABLE OF CONTENTS 
DECLARATION ............................................... .................... ...... ... ............ .... ....... .. ......... .. ... ......... ii 
DEDICATION ........................................ .................. ..................................................................... iii 
ACKNOWLEDGMENT ... ................................. .... .............. .. ........... ...................... ................ ....... iv 
TABLE OF CONTENTS ........ ........................................... ............ ...................... .... ............. .. ....... vi 
LIST OF FIGURES ............................................................................. .... .............. .. .................... xiii 
LIST OF TABLES ....... .. .... ............ .. .......... .......... ...... ........ .......... ........ .... ........................... ..... .. ... xv 
LIST OF ABBREVIA TIONS ............................ .. ........ ............. .......................... ... .. ................... xvii 
ABSTRACT ... ........................................ .. ...... .................................................... .................. .... .. .. xix 
CHAPTER ONE: INTRODUCTION ........ ............. .. .. ... .............................. .. ................................. 1 
1.1 BACKGROUND .... .. .... .. ............................... ... ......................... ............. ... ............. ... ....... 1 
1.2 PROBLEM STATEMENT ... ...... .. .............. ... ..... .. ... .... .............................................. .. ..... 5 
1.3 CONCEPTUAL FRAMEWORK ........ .. ........ ... ............................ .......... ...... .... .... .. .......... 8 
1.4 RESEARCH QUESTIONS .... ... .. ...... .. ... ...... ... .. .. .. .............. ...... ................................ ... .. 10 
1.5 RATIONALE .......... .............. ................... .. .. ...................................................... .. .......... I 1 
1.6 GENERAL OBJECTIVE ............................................................................. .. ....... .. ... .... 15 
1.6. I Specific Objectives ............... ........... ...... ............... ............ .. ..... ...... ......... ....... .. ....... . 15 
1.6.2 Hypotheses ...................................... .. ................. .. ... ................................. ........... ..... 16 
CHAPTER TWO: LITERATURE REVIEW ... ........... .. .............. .... .. .. .. .......... .... .. .. ........ ............ . 17 
2.1 THE CERVIX UTERI ............. .. ................................. .. ........ ....................... ................... 17 
2.2 CANCER ........ .............. .. ... .... ............................................... ...... ....................... ........ ..... 19 
2.3 BURDEN AND EPIDEMIOLOGY OF CERVICAL CANCER ................ ..... .. ... ......... 20 
2.3.1 Variation in the Incidence of Cervical Cancer ...... ...... ................. ..... ...................... 20 
2.3.2 Variation in the Mortality of Cervical Cancer .. ....... .. ................... ........... ... .... ........ 23 
VI 
University of Ghana  http://ugspace.ug.edu.gh
2.3 .3 Summary of Burden of Cervical Cancer ....................... .... .. ...... ......... ..... ...... .......... 25 
2.4 THE NATURAL HISTORY OF CERVICAL CANCER .................... ... ................... .... 25 
2.4.1 Precursor Lesions of Cervical Cancer ..... ....... ...... .............. .. .. ....... .. ..... .... .. ...... .... ... 27 
2.5 RISKS FACTORS OF CERVICAL CANCER ..... ........ ............ ...... ....... ... ..................... 28 
2.5.1 Sexual Risk (Behaviours) Factors .............. .. .............. ............ ... ... ... ... ....... ...... .. .... .. 29 
2.5.1.1 Number of Male Sexual Partners ....... .......................... ...................... ... ...... ..... 29 
2.5.1.2 Age at First Sexual Intercourse (AFSI) ................................. ..... ... ............... ... 30 
2.5.1.3 Sexually Transmitted Infections ................ ............................... ...... ...... ... ... ..... 31 
2.5.1.4 Inconsistent use of Condom .......... ............ ..... ... ............ ... ... ............................ .. ... . 31 
2.5.2 Reproductive Risk characteristics (Known Risk Factors) ...... ........... .... ....... ....... .... 32 
2.5.2.1 Multiparity ........ ................................. .......... ............. .. ........... ..... .. ...... ............ . 32 
2.5.2.2 Oral Contraceptive Use ...................... ................... ............. .................. ........... 33 
2.5.3 Other risk factors ......... .......... ......... ....................... ...................... ..... ................... ... . 34 
2.5.3 .1 Smoking Status .... ................. ....... ........................................ .................. .... ...... 34 
2.5.3.2 Immunological Status .......... ..... .... ..... .......................... .............. ... ..... ...... .. ...... 35 
2.5.3.3 Inheritable Host Genetic Factors ............................... .. .. ..... .................. ... .... .... 36 
2.5.4 Summary of Risk Factors ........... ......... ... ... .. ... ..... .. .............. ... .... .............. ... ..... .. .... . 36 
2.6 TH E HUMAN PAPILLOMA VIRUS .. .. .. .................. ...... ................ ................ ...... ... .. .... 37 
2.6.1 Biologic Class ification ofHPV Types .... .... .. .. .... ................................. ....... .... .... .. .. 37 
2.6.2 The HPV Genolne ................ ...... ..... .. ...... .. ..... ................ .. .. .. ...... .............. ........ ....... 39 
2.6.3 Sumnlary of the HPV ............... .... .... .... ........ .. .................... .......... .......... .. .. .......... ... 4 1 
2.7 THE NATURAL HISTORY OF HPV IN FECTION ............ .. ...................... ... .............. 41 
2.7. 1 Association of HPV with Cervical Cancer .............. ................ .. .... .. ........................ 42 
2.7.2 Acquisition ofHPV Infection .............. .... .. .. ............ .. .... .. .......... .. .................... .. ...... 43 
2.7.3 Persistence and clearance of HPV Infection ............ .. .............. .... .......... .. .......... ...... 44 
2.7.4 Progression ofHPV Infection ....... .... ................ ... ........ .. .................. ......... ....... ........ 45 
VII 
University of Ghana  http://ugspace.ug.edu.gh
2.7.5 Molecular Mechanisms ofHPV Mediated Carcinogenesis ..................................... 45 
2.7.6 Summary of the Natural History of HPV infection ................ .... .................. ........ ... 47 
2.8 EPIDEMIOLOGY OF HPV INFECTION ............................... ............. ...................... .. . 48 
2.8.1 Epidemiological Risk Types Classification ofHPV ................ .. ............................. 48 
2.8.2 Prevalence of HPV Infection ................................................................................... 49 
2.8.2.1 HPV prevalence variations with grade oflesion and age ...... ..................... .. ... 50 
2.8.2.2 Age-specific HPV prevalence variations ......................................................... 51 
2.8.2.3 Type specific HPV prevalence ....................................................................... 52 
2.8.2.4 Country specific HPV prevalence ...................................................... .......... ... 54 
2.8.3 Summary of Epidemiology ofHPV ........................................................................ 55 
2.9 PREVENTION OF CERVICAL CANCER ...................................... ...... ......... ...... ........ 56 
2.9.1 Primary Prevention ........................................................................ .. .............. .. ........ 56 
2.9.1.1 Preventive Sexual Behaviour. ........................................................... ............... 57 
2.9.1.2 Prophylactic Vaccination .......................................................................... .. ..... 58 
2.9.2 Secondary Prevention: Cervical Screening ....................................... ........ .............. 59 
2.9.2.1 Screening by Cytology .................................................................................... 61 
2.9.2.2 Screening by HPV Testing .... .... ......................................... ..... ............... ....... .. 62 
2.9.2 .3 Screening by Visual Inspection ofthe Cervix ............... .. ............... ..... .... ........ 66 
2.9.2.4 Self-Sampling and Participation in Screening ................................................. 67 
2.9.2. Summary of Prevention of Cervical Cancer ............................................................... 70 
CHAPTER THREE: METHODS .. ...... .... .................. ........ .... ...... .... ...... ...... ....... .. .. .... ......... ......... 72 
3.1 STUDY LOCATIONS ....................... ................................................... ........ ................. 72 
3.2 SOURCE AND STUDY POPULATION .. ....................... ............... .. ......................... ... 74 
3.3 SAMPLE SIZE DETERMINA TION ............................................................................. 74 
3.4 STUDY DESIGN (CROSS-SECTIONAL STUDY) ..................................................... 75 
3.4.1 Study Type, Selection of Communities and Distribution of Sample Size .... ...... ..... 75 
viii 
University of Ghana  http://ugspace.ug.edu.gh
3.4.2 Community Entry and Engagement ...... .... ....... ....... .......... ... ........ .... .. ..................... 76 
3.4.3 Sampling and Reporting Strategies ...................... ......... .. ............... ......................... 77 
3.4.3.1 Reporting strategy 1. .............. ............ .... ................ ... ... ..... ..... ... ... .................... 77 
3.4.3.2 Reporting Strategy 2 ................... ....... ............ .............. .. ...... ........ .... ......... .. ..... 79 
3.4.4 Inclusion and Exclusion Criteria .................................... ..................... .... ................ 81 
3.5 ETHICAL ISSUES .... ............. .. .... .. ............ .......................................... ....................... ... 81 
3.6 QUESTIONNAIRE ADMINISTRATION .. ... ..... ... ................... ..................... .. ........ ..... 82 
3.7 SAMPLE COLLECTION AND SMEAR PREPARATION ....... ............. .. ............... .. .. 83 
3.8 PAPANICOLAOU STAINING OF SMEARS ........ ........ ............................... .... ........... 84 
3.9 DNA EXTRACTION .... ... ....... ............ ....... ... ..... .... ..................... .. ....................... .. ...... .. 85 
3.10 HPV GENOTYPING .. .. ........................................... ... ... ... .. ............... .. ... ................. ... .... 85 
3.10.1 Nested-Multiplex PCR ......... .. ...... ..................... .... ..................... ................... ......... . 85 
3.10.2 Detection of PCR Products / Genotypes .... .. ......... .. ......... ..... ... ... ....... .... ....... .. ...... ... 86 
3.11 HPV 16, 18 AND 45 VARIANT ANALYSIS .. .. ...... ... ... ................................ ........ .... ... 87 
3.11.1 HPV 16 nested PCR ........ ......... .............. ... .. ... ... ...... .. ..... ..... ....... ............ ...... ...... .... .. 87 
3.11 .2 HPVI8 nested PCR ...... .................................... ......................... .............. .... ....... .. ... 88 
3.11.3 HPV 45 nested peR ........... ................ .................................................................. .... 89 
3.11.4 Visualization of second round PCR product ........ ..... .............................................. 90 
3.11.5 Sequencing of second round PCR product .................... ............................ .... .. ...... .. 90 
3.12. STATISTICAL ANALYSIS .... ..... ................... .. ..... ....................................................... 91 
3.12.1 Data Entry and Qual ity Control. ............... ...... ................................... ...... .......... .. .... 91 
3.12.2 Main Analyses ............................. ...................... ....... ............. .... ........ ........ ............. . 91 
3.12.3 Additional Analyses ..... ... .. ........ ............................. ....... .......................... ........... .. ... 93 
3.13 DEFIN ITION OF VARIABLES ................................ ............... ................ ... .................. 93 
3.13.1 Measures / Variables of Interest .............. ......... .... ................................... ..... ... .. ...... 93 
3.13.2 Independent Variables ............................................................... .... .......... .... ......... ... 94 
IX 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER FOUR: RESULTS ............... ............................................... ....................................... 97 
4.1 PARTICIPANT DEMOGRAPHIC CHARACTERISTiCS ........................................... 97 
4.2 PARTICIPANTS REPORTING RATES/STATUS ..................................................... 100 
4.2.1 Reporting strategy 1 .... .......... .......................................... ..... .. .................. ...... ..... ... 100 
4.2.2 Reporting Strategy 2 ... ................................................................ ........... ... ..... .. ...... 102 
4.2.3 Overall Reporting Rate ...... .. ............................. ....................................... ............. 102 
4.3 SEXUAL AND REPRODUCTIVE CHARACTERISTICS (KNOWN RISK 
FACTORS) ................................................................................................................... 103 
4.3 .1 Sexual Risk Characteristic (Risk Factors) ..................... .......................... .... .......... 103 
4.3 .2 Reproductive Risk Characteristics (Risk Factors) ....... ........ ............ ........ .............. 105 
4.3.3 Other Risk Characteristics (Factors) ...................................................................... 107 
4.4 PARTICIPANTCHARACTRISTICS THAT INFLUENCED REPORTING AND 
CONDOM USE ............................................................... ... ................. ... ....... ........... .... 108 
4.4.1 Participant Characteristics that may have Influenced Reporting Status ....... ......... i 08 
4.4.2. Participant Characteristics that may have influenced Condom Use .... .. ................ 111 
4.5 ADDITONAL FINDINGS - SPECIMEN COLLECTION .... .. ............ ........................ 116 
4.5.1 Pre-performance preference of Specimen Collection Method .............................. 116 
4.5.2 Post-performance Preferences for Specimen Collection Method .............. ..... ....... 116 
4.6 HUMAN PAPILLOMA VIRUS (HPV) TESTING ........ .... ...................................... ... . 119 
4.6.1 HPV Genotype Specific Prevalence ...... .. ................ .... .................................... ...... 119 
4.6.2 Overall and Risk Type HPV Prevalence ...................................... .... ..................... 122 
4.6.3 Agreement between Self and Health Personnel Collected Specimen ................... 123 
4.6.4 Test of Hypothesis on HPV Prevalence ................................................................ 126 
4.6.5 Age Specific Prevalence (ASP) ................................ .. ......................... .............. .... 126 
4.7 TEST OF ASSOCIATIONS BETWEEN SEXUAL RISK CHARACTERISTICS AND 
HPV POSITIVITY .... ..................... ....................... .. .......................... ........ ... .. ...... ... ..... 128 
x 
University of Ghana  http://ugspace.ug.edu.gh
4.7.1 Univariable Association between Sexual Risk Characteristics and HPV Infection 
Positivity Determined With Self-Collected Specimen ........... .... ............ .... ........... 129 
4.7.2 Univariable Association between Sexual Characteristics and HPV Infection 
Positivity Determined with Health Personnel-Collected Specimen ...................... 132 
4.8 HPV VARIANT ANALYSES ................................... ............... ......... .......................... 135 
4.8.1 HPV16 Variants ................. ............. ..... ................. ............. ......................... ........... 135 
4.8.2 HPV 18 Variants ..................................................................................................... 141 
4.8.3 HPV45 Variants ..................................................................................................... 145 
4.9 CERVICAL LESIONS AND ABNORMALITIES .............................. ........... ............ 148 
CHAPTER FIVE: DISCUSSION ................................................................. ......... ............ ......... 150 
5.1 PARTICIPANT CHARACTERISTICS .............. ........ ....................... .. ............. ........... 150 
5.1.1 Distributions demographic Characteristics .......... ........ .... ..... ...................... ...... ..... 150 
5.1.2 Distribution of Sexual Risk Characteristics ............................... ... ......................... 151 
5.1.3 Distribution of Reproductive Risk Characteristics ... ... .......... ......... ... ....... ......... .... 153 
5.1.2 Participant Characteristics that may have influenced Condom Use .......... ..... ....... 155 
5.2 RESPONSE TO THE REPORTING STRATEGIES ................ .............................. .. ... 157 
5.2.1 Response Rates ... ..... ....... ..... ............................................... ..... .. ............. ............... 157 
5.2.2 Participant Characteristics that may have Influenced Reporting ................. ... ... .... 160 
5.3 ADDITONAL FINDINGS - SPECIM EN COLLECTION .................................... .. .... 161 
5.3.1 Performance and Preference for Specimen Collection ........... ................ .......... ..... 161 
5.3.2 Reasons for and characteristics that may have influenced Preferences ....... .......... 163 
5.4 HPV BURDEN WITH SC AND HPC ... .................................... .. ................................ 164 
5.4.1 Differences in HPV Prevalences ofSC and HPC. .................. ............................... 164 
5.4.2 Age-Specific HPV Prevalence ............................... ... .......................... ...... ....... ...... 168 
5.4.3 Level of Agreement between SC and HPC ............ .. .. ............... ....... ..... ................ 172 
5.6 ASSOCIATION OF SEXUAL CHARACTERISTICS WITH HPV POSITIVITy .... 173 
XI 
University of Ghana  http://ugspace.ug.edu.gh
5.7 HPV VARIANT ANALySIS ....... ...... ... ......................................... .............................. 174 
5.8 CERVICAL LESIONS AND THE ASSOCIATJON WITH RISK FACTORS .......... 177 
5.9 LIMITATIONS OF STUDY ... ..... ...................... ... .. .......... .... ..... ........... ..... ....... .... ....... 178 
CHAPTER SIX: CONCLUSION AND RECOMMENDATIOS ....... ................. ....................... 180 
6.1 CONCLUSIONS .... ......... ..................................................................... .............. .......... 180 
6 .2 RECOMMENDATION ... .... .. ......... ............ .......................................... .. ............. ......... 181 
6.2.1 Policy ........................ .... .................................. ....... .... ... .... ....... .......... .......... .............. 181 
6.2.2 Further studies .......... ......... .............. ...... .... ............... ...... .... .... .... ... ..... ........ ... ...... .. 182 
REFERENCE ..... ..... .. ... ......... .... ............ ... .. .. .... .... ........ ................. .... ....... ... .. ................. .... ...... ... 184 
APPENDIX I ...... .... ....... ... ...................... ..... ................ ... .... ... ....... .... .................... .. ... ........ ........ . 207 
APPENDIX JI ............................................ ..................... .... ... ..... ................ .......... ..... .... ... ... ....... 210 
QUESTIONNAIRE A ......................... ............ ... ... .. ... .................... ................ ............... .... .... . 210 
QUESTIONNAIRE B ....................................... ....... ...... ... ........... ....................... ... .... ... ......... . 212 
QUESTIONNAIRE C .......... ............. .......................... ................. ..................... ..... .... .... ......... 213 
APPENDIX III ....................... ..................... ............ ............ ............. ........... ... .. ............ ...... ....... .. 214 
CONSENT FORM ...... .. .... .......... ... ....... ........... ... ..... .......... ................... .. ......... ......... ..... ... ... .. . 214 
VOLUNTEER AGREEMENT FORM ................. ........... ....... ............... ........................... .... .. 218 
APPENDIX IV .......... .... ..... ........ ... ... .. ................ ....... ..... .... .... ..... ...... ... .. ... ............. ......... .. .... ... .. . 220 
SEQUENCE DATA ..... ... ............. ............. ... ....... ...... .. ................................. ..................... .. .... 220 
XII 
University of Ghana  http://ugspace.ug.edu.gh
LIST OF FIGURES 
Figure 1.1: Conceptual framework of cervical cancer and related risk factors .. ........ .............. ..... . 9 
Figure 1.2: The current status of the relationships between the country-specific data needed to 
develop a cervical cancer prevention plan in Ghana ... ............................. ....................... ............ . 14 
Figure 2. 1 Diagram of the uterus of a woman within her reproductive age.Error! Bookmark 
not defined. 
Figure 2. 2: An illustration of the multifactorial and a trifocal outcome model involved in the 
development of cervical cancer .......... .... ......... ..... ....... ......... .. ........ ............. ............ ..... .... .... ...... .. 27 
Figure 2.3: An illustration of the double stranded DNA genome ofHPY . ..... .......... ... .. .... ... ...... 40 
Figure 2. 4 An illustration of the natural histories ofHPV infection and cervical cancer. ........... 43 
Figure 3. I Population distribution in the Lower Manya Krobo District.. ... .. ....... ........... ............. 73 
F igure 3. 2: An illustration of the design for reporting for specimen collection . .. .......... ..... ..... ... 80 
Figure 4. 1: Participant enrolment and response to specimen collection.Error! Bookmark not 
defined. 
Figure 4.2: Distribution of the performance of sample collection ...... .................................. ... .. . 117 
Figure 4.3: Age specific overall HPV infection prevalence obtained with self and health 
personnel collected specimen ..... .. ...... ............... .. ... ... ............................. ....... ........ .. ..... ..... ......... 127 
Xlii 
University of Ghana  http://ugspace.ug.edu.gh
Figure 4.4: Age specific high risk (HR) HPY infection prevalence obtained with self and health 
personnel collected specimen ........................ .. .... ........................ ..... ........... .................. ......... .... 128 
Figure 4.5: Age specific low risk (LR) HPY infection prevalence obtained with self and health 
personnel collected specimen .............................................................. ..................... ............. ... .. 129 
Figure 4.6 An evolutionary relationship of HPY] 6 isolates based on sequences between the 
genome positions 7469 and 7840 within the long control region (LCR.) . .. ................ .... ........... 140 
Figure 4.7 An evolutionary relationship of the taxa of HPY 18 isolates based on sequences 
between the genome positions 7464 and 7839 within the long control region (LCR.) ........ ... ... ] 44 
Figure 4.8 An evolutionary relationship of HPY45 isolates based on sequences between the 
genome positions 7074 and 7858 within the long control region (LCR.) ............... ............ ....... ]47 
XIV 
University of Ghana  http://ugspace.ug.edu.gh
LIST OF TABLES 
Table 4.1: Distribution of the demographic characteristics of the 415 recruited participants ...... 98 
Table 4.2: Distribution of the categories of the participants' sexual risk factors .... ..... ......... ..... 104 
Table 4.3: Distributions of the reproductive risk factors among the participants ... ....... ............. 106 
Table 4.4: Distribution and Association of participants ' demographic characteristics with their 
reporting status, for Community specimen collection ....... ........ ....... ......... ...... .............. ..... ..... .. . 109 
Table 4.5: Distribution and Association of participant demographic characteristics with the 
reporting status, for Short duration hospital based specimen collection .................. .... ...... ........ 110 
Table 4.6: Distribution and Association of participants' demographic characteristics with their 
reporting status, for Long duration hospital based specimen collection ............ .............. ........... 112 
Table 4.7: Association between use of condom and demographic characteristics of the 
participants ..... .. ................. ...... .......... .. .... ......... ... ................................ ..... ......... .... ................ ...... 113 
Table 4.8: Sexual risk characteristics of the participants' which are associated with the use of 
condon1 .................. ... ........ .. ... ........ ..... .................. ... .. ..................................................... .. ..... ...... 1 14 
Table 4.9: Distribution of preferences for specimen collection methods after providing the 
specilnen ................................................... ............ .. .. ......................... .... .. .. ................. .... ...... ..... . 117 
Table 4.10: Reasons for the preference of self-specimen collection ............................ .. ............ 118 
Table 4.11: Reasons for the preference of health personnel specimen collection ....... .............. . 119 
xv 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.12: HPY genotype prevalence detennined with each specimen collection method ...... 121 
Table 4.13: HPY prevalences and the extent of agreements between self and health personnel 
collected specimen ................................... .. .......................................................... ....................... 124 
Table 4.14: Concordance and discordance in the detection ofHPY infections with self and health 
personnel collected specimen .......................... ...... ... .................................. .. ....... .............. ......... 125 
Table 4.15: Crude odds ratio for the association between sexual characteristics and HPY 
infection positivity determined with self-collected specimen ............................... .......... ... ........ 13 I 
Table 4.16: Crude odds ratio for the association between sexual characteristics and HPY 
infection positivity determined with health personnel-collected specimen ... .. ... ... ....... ...... .. ...... 133 
Table 4.17 Nucleotide sequence variations between positions 7469 and 7840 (within the long 
control region (LCR» ofHPY16 variants ..... ........ ............................ ....... .................................. 138 
Table 4.18 Nucleotide sequence variations between positions 7464 and 7839 (within the long 
control region (LCR» of HPY 18 variants ....... ... ..... ............. .... ............. ...... ............................... 143 
Table 4.19 Nucleotide sequence variations between positions 7074 and 7858 (within the long 
control region (LCR» ofHPY45 variants . .. ......... .................................... .................................. 146 
Table 4.20: Distribution and association between the location of specimen collection and the 
adequacy of the smears for evaluation ....................... .. .................................. ............... .............. 148 
Table 4.21: Distribution of diagnosis based on Pap smear evaluation .......... .. .................... .... ... 149 
XVI 
University of Ghana  http://ugspace.ug.edu.gh
LIST OF ABBREVIATIONS 
ADC Adenocarcinoma 
ADSC Adenosquamous Carcinoma 
AFSI Age at First Sexual Intercourse 
AGC-US Atypical Glandular Cells of Undetermined Significance 
ASC-US Atypical Squamous Cells of Undetermined Significance 
ASR Age-Standardized Rate 
CIN Cervical Intraepithelial Neoplasia 
CIS Carcinoma-In Situ 
CNSP Current Number of Male Partners 
HPC Health Personnel Collection 
HPV Human Papillomavirus 
HR High Risk 
IARC International Agency For Research on Cancer 
IQR Interquartile range 
LBC Liquid Based Cytology 
LCR Long Control Region 
LNSP Lifetime Number of Male Sexual Partners 
LR Low Risk 
XVII 
University of Ghana  http://ugspace.ug.edu.gh
OC Oral Contraceptive 
OR Odds Ratio 
ORF Open Reading Frames 
PHR Possibly High Risk 
PPS Probability Proportional to Size 
SC Self-Collection or Self-Collected 
SCC Squamous Cells Carcinoma 
SCJ Squamocolumnar Junction 
SIL Squamous Intraepithelial Lesion 
SIR Standardized Incidence Ratio 
TZ Transformational Zone 
VIA Visual Inspection with Acetic Acid 
VILI Visual Inspection with Lugol ' s Iodine 
VLP Viral-like-Proteins 
XVIII 
University of Ghana  http://ugspace.ug.edu.gh
ABSTRACT 
Background: Ghana continues to lack population based empirical data on the burden of cervical 
pre-cancer lesions, HPV infection and the identification of communities at risk. These together 
are hampering the control of cervical cancer in Ghana. Therefore, this study was designed to 
determine the distribution of HPV infections, cervical lesions and the known risk factors for 
cervical cancer among women living in the Akuse sub-district, Ghana. 
Methods: A cross sectional study was designed with a house-based recruitment of women 
between the ages of 15 and 70 years. Although a single strategy was designed for reporting for 
specimen collection in the study, three strategies were used. The additional strategies were a 
result of an initial low response rate resulting from the women not making time off their daily 
and economic activities to attend the Hospital for specimen collection and also based on the 
information and suggestions from the subsequent interaction with the women. These strategies 
were, a) reporting to the hospital within a short time period after recruitment at home, b) 
reporting to the hospital within a long time period after recruitment at home and c) reporting at a 
location within a community after recruitment at home. Each participant had the option of 
providing a specimen by either or both self-specimen collection and health personnel specimen 
collection. Human Papillomaviruses were detected in each of the pair of the specimen by a 
multiplex-nested PCR amplification and genotyped by the xMAP technology. 
Results: Of the 38 genotypes detected with both specimen collection methods, the following 
were the commonest detected with self-specimen; HPV16 (5.9%; 95% C13.0% - 9.0%), HPV35 
xix 
University of Ghana  http://ugspace.ug.edu.gh
(4.7%; 95% C] 2.0% - 8.0%) and HPV40 (4.7%; 95% C] 2.0% - 8.0%). The commonest 
detected with the health-personnel collected specimen were HPV35 (2.8%; 95% C] 1.0%-
5.0%) and HPV58 (2.8%; 95% C] 1.0% - 5.0%). A significant difference was obtained between 
the overall HPV prevalence determined with the self-collected specimen [43 .1%  (95% C] 38.0% 
- 51.0%)] and that with the health-personnel collected specimens [23.3% (95% C] of 19.0% -
31.0%)]. Varying reporting rates were observed for the three reporting strategies as well as 
varying preferences for the specimen collection methods. All the women were negative for 
cervical cancer and showed no dyskaryosis (intraepitheliallesion). 
Conclusion: The findings of this study provide the necessary baseline data needed to contribute 
to the development of a cervical cancer prevention plan and indicate that the Akuse sub-district 
is a high risk community for HPV infection and a low risk community for cervical cancer. 
xx. 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER ONE 
INTRODUCTION 
1.1 BACKGROUND 
Cervical cancer remains a disease of global public health concern, although it is the only 
preventable cancer with a record of a very effective prevention method (Bradley et al. , 2005). 
As such, there are consistent efforts to reduce the high rates of incidence and mortality 
recorded in the World, Africa and Ghana. The WHO has consistently published its 
Comprehensive Cervical Cancer Control Guide Book with the aim of helping to reduce the 
burden of cervical cancer on women (WHO; Guideline Development Group, 2014; WHO; 
Chronic Diseases and Health Promotion Group, 2006). This section of the thesis presents a 
brief description of the background to this study by throwing light on the WHO's position 
and research conducted on cervical cancer prevention and control , and the challenges 
associated with cervical control. 
According to the WHO, the goal of any comprehensive cervical cancer prevention and 
control programme should be to reduce the burden of cervical cancer by a) preventing and 
reducing HPV infection, b) detecting and treating cervical pre-cancer lesions (screening) and 
c) providing treatment and palliative care for those who do develop the disease (WHO; 
Guideline Development Group, 2014; WHO; Chronic Diseases and Health Promotion Group, 
2006). The development of such a programme should start with the development of a national 
policy, which should include the gathering and review of country-specific empirical data to 
answer key questions on baseline characteristics (WHO; Guideline Development Group, 
2014). These country-specific empirical data should include data on; a) the burden of HPV 
University of Ghana  http://ugspace.ug.edu.gh
infections (this includes the following prevalences; overall HPV, genotype specific HPV, 
age-specific HPV, multiple HPV infection, high risk HPV type, low risk HPV type, and if 
possible HPV genotype variants), b) the prevalence of cervical pre-cancer lesions, c) the 
distributions of the sub-categories of the known and globally accepted sexual and 
reproductive risk (behaviour characteristics) factors among women, d) the extent of and the 
factors that influence the use of condom (very important for all STI prevention), e) factors 
influencing the reporting of women for cervical cancer screening and f) the level of 
knowledge on cervical cancer among women; All of these data should be collected at the 
community level through screening activities. 
In the absence of a national cervical cancer screening programme, Ghana, like other 
developing countries, has had challenges regarding her efforts at reducing the public health 
burden of cervical cancer and HPV infection through cervical cancer screening. Some of 
these challenges include the absence of a national guideline for cervical cancer screening and 
a largely opportunistic and centralised (in Accra and Kumasi) cervical screening activities 
(Adanu et aI., 2010; WHO/ICO Information Centre on Human Papillomavirus and Cervical 
Cancer, 2010). These opportunistic screening activities have mostly been as a result of 
referrals for Pap smear and VIA by Physicians (Adanu el ai. , 2010; Adanu, 2002). A further 
challenge is that the Centres that provide these opportunistic cervical cancer screening 
services in Ghana were as a result of pilot programmes (Adanu et af. , 2010; Adanu , 2002), 
which, instead of expanding to other locations, have remained in the hospitals that hosted the 
pilot projects. One of the project centres, the Ridge Hospital, which is the regional hospital of 
the Greater Accra Region, continues to provide referral based cervical cancer screening 
services. 
2 
University of Ghana  http://ugspace.ug.edu.gh
Another challenge is limited localisation of the projects; that is, the two main programme-
based cervical screening and treatment service projects were conducted in Accra. One of 
which was sponsored by the Alliance for Cervical Cancer Prevention (ACCP) and The Johns 
Hopkins Program of International Education in Gynaecology and Obstetrics (JHPIEGO) 
(Bradley et al., 2005). That project implemented visual inspection pilots screening at two 
sites in Accra (Abotchie and Shokar, 2009). Although it resulted in the drafting of a cervical 
cancer prevention and control programme, that was never implemented. The second project, 
also localised in Accra, was part of the Women ' s Health Study in Accra (WHSA), where 
cervical cancer screening by Pap smear testing was conducted (Duda et al., 2005). These 
challenges have resulted in a very low coverage of cervical cancer screening in Ghana, which 
was estimated to be between 2.2% and 8.8%. (WHO/ICO Information Centre on Human 
Papillomavirus and Cervical Cancer, 2010; Abotchie and Shokar, 2009; Adanu , 2002). These 
data show that there is much to be done in Ghana to reduce the burden of cervical cancer. 
The study reported in this thesis, in addition to Pap smear testing, used HPV DNA testing to 
screen women in the general population of the Akuse sub-district. That is, of the three 
available screening methods (VIA, Pap test and HPV DNA test) two were used in this study. 
Although these screening methods are discussed in section 2.9, the following is a brief 
discussion of the bases of the selection ofHPV DNA testing for this study. 
HPV DNA testing (one of the secondary prevention methods for the control of cervical 
cancer) is the most sensitive and reproducible of the three secondary prevention methods. It 
involves detecting and genotyping the HPV in collected cervical swab specimen, if any. The 
presence of some type of HPV (known as high risk HPV types, discussed in section 2.6) 
indicates a high risk for the development of cervical pre-cancer lesions and cancer, while the 
presence of some other HPVs (known as low risk HPV types , discussed in section 2.6) 
3 
University of Ghana  http://ugspace.ug.edu.gh
indicate a low risk for the development of cervical pre-cancer lesions and cancer. However, 
the major limitation of this method is its slightly lower specificity for the detection of high 
grade cervical lesions, compared to the earlier developed cytology based Pap testing (Zhao et 
aI. , 2012; Katki et al., 2011; Zhao et aI. , 2010; Dillner et aI., 2008). The central debate on 
improving HPV DNA testing is currently focused on the algorithm for its inclusion in 
primary cervical cancer screening programmes and for the triage of women (Saslow et aI. , 
2012; Luesley and Leeson, 2010; Arbyn and European Commission. Directorate-General 
Health & Consumer Protection, 2008). Another significant issue regarding the usefulness of 
HPV DNA testing in cervical cancer control is its current high cost, which is limiting its use 
in developing countries such as Ghana. However, the recent development of test kits such as 
the CareHPV kit holds a great potential for reducing the cost of HPV testing and increase its 
availability in developing countries. Whiles waiting for these improvements, it is necessary 
for this country to generate background data of the HPV genotypes prevalent among its 
population and determine if there are DNA sequences differences that may be useful in the 
choice of HPV testing kits when they becomes available. This informed the HPV genotyping 
and the DNA sequencing of the most prevalent genotypes in this study. 
Although not directly a part of the study reported in this thesis , it is worth noting that the 
most recent success at addressing the immense global public health burden of cervical cancer 
has been the production of three HPV vaccines for primary prevention of HPV infection and 
therefore the control of cervical cancer and the other HPV related cancers (Chatterjee, 2014; 
Wang and Roden, 2013 ; Van de Velde et aI., 2012; Giuliano et ai., 2011; Schellenbacher et 
aI. , 2009; Aires et a!. , 2006; Harper et al., 2004). As more vaccines are developed, the choice 
of a vaccine will be informed by the data on the genotype specific prevalence in a country. 
For now, a divalent vaccine, known as Cervarix® protects against HPV -16 and HPV -18, 
4 
University of Ghana  http://ugspace.ug.edu.gh
while the tetravalent vaccine, known as Gardasil® protects against HPV -16, -18 , -6 and -11 
(Dunne et al., 2008). More recently, a nonavalent vaccine against HPV -16, 18, -6, -11 , -31 , -
35, -45 , -52 and -58 has been developed (Chatterjee, 2014; Van de Velde et al. , 2012). With 
the assistance of the Global Alliance for Vaccine initiative (GAVI), these vaccines have been 
made affordable to developing countries with demonstration project in Uganda and Tanzania. 
Additionally, through the Gardasil Access Programme (key initiative), the Merck-Quigen 
efforts and PATH support, and using a school based approach, Ghana, Guyana, Haiti , 
Honduras, Lesotho, Cameroon, Uganda and Tanzania, have generated a wealth of knowledge 
and lessons on HPV vaccination for girls between the ages of9 and 13 years (Gulland, 2012; 
Quentin et af. , 2012; Watson-Jones et af. , 20 12). Rwanda (in the year 2011) and Lesotho (in 
the year 2012) initiated a National HPV vaccination programme (Gulland, 2012; Watson-
Jones et al., 2012). Although Ghana is a GA VI eligible country for support there is currently 
no national HPV vaccination programme, however, a pilot has been initiated and the vaccines 
are also available in some private health facility (Coleman et af. , 2011 ; WHO/ICO 
Information Centre on Human Papillomavirus and Cervical Cancer, 2010). 
1.2 PROBLEM STATEMENT 
Cervical cancer is the fourth commonly diagnosed and the fourth most fatal cancer among 
women worldwide, with an estimated 528,000 new cases and 266,000 deaths in the year 2012 
(GLOBOCAN , IARC, 2012). It is common among women within the economically active 
age group; between the ages of 30 years and 55 years (GLOBOCAN, IARC, 2012; Jemal e/ 
al., 20 II). Developing countries, particularly African countries and the countries of the sub-
Saharan Africa region, continue to bear the highest proportion of the global burden of 
cervical cancer (GLOBOCAN, lARC, 2012; Jemal et af. , 2011) and cervical pre-cancer 
5 
University of Ghana  http://ugspace.ug.edu.gh
lesions among women in the general population (Clifford et aI. , 2005a, 2003). The extent of 
which was shown by the high estimated incidence rate per 100,000 women of 42.7 for 
Eastern Africa, 33.37 for Melanesia, 31.5 for Southern Africa, 30.6 for WHO Middle Africa 
region (Angola, Cameroon, Central Africa Republic, Chad, Congo Brazzaville, Congo, 
Equatorial Guinea and Gabon) and 28.7 for Western Africa; it is interesting to note that these 
are the five highest in the world (GLOBOCAN, IARC, 2012). Furthermore, cervical cancer 
was the leading cause of cancer death among women in Africa, with an estimated 56,601 
(22.6%) deaths in the year 2012 (GLOBOCAN, IARC, 2012). For Ghana, the fact that 
cervical cancer was the leading contributor to the burden of cancers diagnosed in Ghana 
(GLOBOCAN, IARC, 2012; WHOIICO Information Centre on Human Papillomavirus and 
Cervical Cancer, 2010) makes it a critical public concern. Specifically, of the 6.57 million 
Ghanaian women (aged 15 years or older) estimated to have been at risk of cervical cancer, 
3052 new cases were estimated to have occurred in the year 2012; this was about 19.3% of all 
estimated cancers in Ghana (GLOBOCAN, IARC, 2012; WHO/ICO Information Centre on 
Human Papillomavirus and Cervical Cancer, 20 I 0) . Additionally, it was estimated to have 
been the leading cause of cancer death in Ghana, with an estimated 1556 deaths (14 .6% of all 
cancer deaths) in the year 2012 (WHO/ICO Information Centre on Human Papillomavirus 
and Cervical Cancer, 20 10). Although there is paucity of published data or reports on cervical 
cancer at the regional and district level in Ghana, few studies that achieved low screening 
coverage, have indicated low prevalences of cervical pre-cancer lesions (Adanu et aI. , 20 I 0). 
In one such study in rural Accra, a prevalence of between 4.40% -5.88% for cervical pre-
cancer lesions by VIA was reported (Bradley et aI. , 2005). In the Ashanti region, prevalences 
of 3.5% in the Nkawie community and 12.6% in the Agogo community, both determined by 
Pap testing have been reported (Handlogten et al. , 2014). 
6 
University of Ghana  http://ugspace.ug.edu.gh
As has been globally identified, cervical cancer is a multifactorial and multistage disease that 
develops over a long period of time between 10 and 20 years (WHO; Chronic Diseases and 
Health Promotion Group, 2006; Sankaranarayanan et al., 2003; Sellors et ai., 2003). It is 
caused by the persistent infection of genotypes of HPY, whose acquisition is mainly the 
result of the sexual and reproductive behaviour of women and to some extent, the sexual 
behaviour of their male sexual partners (Franco et ai., 1999; Walboomers et ai., 1999). These 
characteristics are the factors that predispose a woman to acquire an HPY infection and 
enhance the persistence of the HPY infection. It is this persistent HPY infection that causes 
cervical pre-cancer lesion and cervical cancer; necessarily, the persistent HPY infection of 
the cervix by some genotypes of the Human Papillomavirus (HPY) is required, but it is not 
sufficient alone to cause cervical cancer (Franco et al., 1999; Walboomers et al., 1999). In 
other words, the persistence of an HPY infection in addition to the continued involvement in 
these same sexual and reproductive characteristics leads first to the development of low grade 
cervical pre-cancer lesions, which may progress to high grade pre-cancer lesions and then to 
cervical cancer. Each of these, progressions is influenced by the continuous persistence of the 
HPY infection and continuing participation in the sexual and reproductive behaviours. Due to 
the fact that cervical cancer screening coverage is low and has largely been opportunistic in 
Ghana, data on these risk factors and cause of cervical cancer among the general population 
are limited. In other words, there are limited community data on sexual and reproductive risk 
characteristics, HPY infection and cervical pre-cancer lesions in Ghana (have not identified 
communities at high risk) and these are hampering the control and prevention of cervical 
cancer by ensuring the persistence of a high burden of incidence, prevalence and mortality 
due to cervical cancer in Ghana. 
7 
University of Ghana  http://ugspace.ug.edu.gh
The consequence of cervical cancer on the nation is evident by the fact that cervical cancer is 
common among women within the economically active age groups. As such it reduces the 
national productive work force, which in the long run will affect the development of the 
Ghana. The role of women in Ghana's informal economy is widely acknowledged, as most of 
the commercial trading activities are largely controlled by women. Additional, the role of 
women in the formal sector of the Ghanaian economy is growingly becoming stronger by the 
day and therefore losing women to cervical cancer would hamper the development of Ghana. 
Furthermore the cost of scaling-up cryotherapy treatment for women diagnosed with cervical 
precursor lesion in Ghana was estimated to cost the nation between GH<l' 11,214,728.00 and 
GH <l' 7,850,309.00 if screening is performed every five years (Quentin et al., 20 I 0). The 
implication of cervical cancer to the family and community is similarly disturbing, because 
this same age range of women are mostly mothers of young families and sometimes 
contribute to the finances of the fami ly. Therefore, losing women to cervical cancer may 
affect the development of children, and the stability of the family. For the individual woman 
who is diagnosed with cervical cancer, the cost of health care and the need for support are 
often issues that are burdensome. For example, the costs of screening ranged from GH<l' 7.30 
to 21.86 (US$ 4.93 to 14.75) for VIA, and from GH<Z.' 70.04 to 125.19 (US$47.26 and 84.48) 
per woman treated with cryotherapy in the year 20 10 (Quentin et aI. , 20 10). The purpose of 
the study reported in this thesis was to contribute the critically needed empirical data to 
inform Ghana ' s cervical cancer prevention and control planning and implementation. 
1.3 CONCEPTUAL FRAMEWORK 
The previous section described the biological factors interacting to result in the development 
of cervical cancer and this is shown in figure 1.1 . While not one of such biological risk 
8 
University of Ghana  http://ugspace.ug.edu.gh
factors for cervical cancer, the lack of screening by women (either due to limited availability 
or lack of interest or knowledge) is another factor contributing to the high burden of cervical 
cancer and cervical pre-cancer lesions in developing countries (Scarinci et af., 20 lOa; Lowy 
et ai. , 2008; Mayrand et af., 2007a). 
Characteristics 
Cancer 
Figure 1.1: Conceptual framework of cervical cancer and related risk factors. 
9 
University of Ghana  http://ugspace.ug.edu.gh
This is because screening ensures early detection of low grade and high grade lesions and 
affords timely treatment to avert the progression of these precursor lesions to cervical cancer. 
Additionally, cervical screening, particularly using HPV testing afford the opportunity to 
avoid persistence of HPV infection. Furthermore, screening ensures the availability of the 
necessary data on the prevalence of HPV infection and related pre-cancer lesions as well as 
of the distribution of sexual and reproductive risk behaviour characteristics for the planning 
of prevention and control of cervical cancer. Therefore, by providing community based 
cervical screening, information on the sexual and reproductive behaviour characteristics, 
HPV infection and cervical pre-cancer lesions among the general population, which are 
equally significant concerns to the health of women, will be made available. 
1.4 RESEARCH QUESTIONS 
1. As suggested by the prevalence of STls in the Lower Manya Krobo District, can the 
communities of one of its sub-districts, Akuse, be said to be communities at high risk 
for HPV infection based on the distribution of the sub-categories of each of the known 
sexual risk factors for HPV infection and if so, will this high risk status be confirmed 
by the prevalence of HPV infection determined in this study? Also, can the 
communities of this sub-district be said to be high risk for cervical lesion based on the 
distribution of the sub-categories of each of the known risk factors for cervical lesion 
development and if so will this be consistent with the determined prevalence of 
cervical lesion for the sub-district? 
2. Will the HPV infection burden in the sub-district be as high as has been reported in a 
similar study in Nigeria (Thomas et al., 2004), the only large cervical cancer and HPV 
study in West Africa at the time of the design of this study? 
10 
University of Ghana  http://ugspace.ug.edu.gh
3. Will the data on the burden of HPY infection determined with self-collected specimen 
be as comparable, if not higher as expected than those determined with specimen 
collected by health personnel? 
4. Will the prevalence of cervical precursor lesions in the sub-district be as high as may 
be expected based on the Globocan estimates for Ghana in the year 2012? 
5. Are the demographic, sexual and reproductive characteristics of the study population 
associated with their use of condom and their reporting status for specimen 
collection? 
6. Are the variants of HPV 16, HPY 18 and HPY 45 in the Akuse sub-district of the 
African lineage and how similar are they to those detected in other African countries? 
1.5 RATIONALE 
In view of the fact that the distribution of HPY genotype infections varies greatly across 
populations (Castle el af. , 2012; de Sanjose el af., 2007), HPY screening needs to be 
performed in every geographical region in order to implement and evaluate the effect of 
suitable vaccination strategies. Particularly in low-resource countries like Ghana, there is the 
need to determine population/community-based HPY infection prevalences , validate low-
cost, simple, and accurate HPY tests and develop strategies to encourage women participation 
in screening programmes. Although it was estimated that over 6.5 million women in Ghana 
were at risk of cervical cancer in the year 2012 (GLOBOCAN, IARC, 2012), Ghana 
continues to lack the important country-specific empirical data (at the community level) on 
HPY infection, cervical lesions and risk factors for cervical cancer needed for prevention 
control planning (as indicated in section 1.1). As such, Ghana does not have a cervical cancer 
prevention and control plan which is based on Ghanaian-specific empirical data, as 
11 
University of Ghana  http://ugspace.ug.edu.gh
encouraged by the WHO. Specifically lacking are the population based assessments of the 
burden of overall HPV infection, genotype-specific HPV infection prevalences, age-specific 
HPV prevalence and the proportion of multiple HPV infections. Furthermore, the lack of data 
on the distribution or profile of the known/accepted risk factors (sexual and reproductive 
behaviour characteristics), the extent of and the factors that influence the consistent use of 
condom and overall prevalence of cervical pre-cancer lesions are limiting the identification of 
communities at high risk. The current state of the lack of data and issues related to prevention 
of cervical cancer in Ghana is depicted in figure 1.2. 
With these data, Ghana will be able to develop a well informed and effective cervical cancer 
prevention and control plan, which will enable the achievement of the following; a) identify 
Ghanaian communities at high risk for HPV infection and cervical cancer, b) make the 
appropriate choice of screening methods to use, d) make the appropriate choice of vaccine to 
use, e) make efforts to increase the number of women participating in cervical screening 
activities, f) identify needs/resources for cervical screening, g) plan for and provide resources 
where needed, and h) prioritize based on need and risk status in relation to the selected 
screening method. These will result in a reduced incidence and increase early detection of 
cervical cancer and thereby result in a low mortality rate due to cervical cancer in Ghana, 
which is currently high (GLOBOCAN, IARC, 2012). Additionally, these HPV data are 
important in the selection of HPV vaccine, selection of HPV test kits and the future 
evaluation of the benefits of HPV vaccination as a primary prevention of cervical cancer. 
Obtaining these HPV data will be effectively facilitated by the knowledge of which specimen 
collection methods (self-collection or health personnel collection) is more acceptable to 
women, more defining of the women ' s risk of development of cervical lesions based on HPV 
12 
University of Ghana  http://ugspace.ug.edu.gh
infection, and which screening methods (HPV test, Pap test or VIA test) the women In 
Ghanaian communities may prefer. 
Since HPV is the commonest sexually transmitted infection (STI) and that infections by the 
following STIs, Chlamydia trachomatis (CT), herpes simplex virus type 2 (HSV-2) and HIV 
(Munoz et at., 2006; Sarnoff, 2005; Smith et aI., 2002a, 2002b, 2004), encourage the 
development of cervical lesion, HPV infections and persistence (Burchell et aI., 2006; 
Sarnoff, 2005), this district was regarded as a potential high risk for HPV infection because it 
was the district with the highest rates of the most common Sexually Transmitted Infection in 
Ghana in 2010 (at the time of the design of the study) . 
13 
University of Ghana  http://ugspace.ug.edu.gh
High incidence of High mortality due to Lack of early detection and therefore women report late 
cervical cancer cervical cancer to hospitals with advance stage cervical cancer 
t 
t i i i i i 
Appropriate choice Appropriate choice of Communities at Low participation of Resources allocation Key resources not 
of the HPV vaccine screening methods high risk have women In may not meet the identified, planned for 
to use have not been have not been be not been opportunistic screening needs of and provided where 
made made identified cervical screening different communities needed 
i i i t t i 
t 
Absence of an empirical country-specific data informed 
LIMITA TION I cervical cancer prevention and control plan. 
i 
f t t .,.. t 
Lack of data on the Lack of data on the Lack of information on the Lack of data on the Lack of information on the 
country-specific overall country-specific burden level of knowledge on distribution ofthe factors that influence 
and age-specific HPV cervical pre-cancer cervical cancer among sub-categories of response rate of women to 
burden lesions women the known risk cervical screening 
factors and 
t t inconsistent use of 
f t condom i 
Lack of data and choice of a Lack of information and Lack of information on the 
specimen collection method that choice of a screening response rate of women to 
will be more informative on the method preferable by cervical screening 
burden of HPV women 
Figure 1.2: The current status of the relationships between the country-specific data needed to develop a cervical cancer prevention 
plan in Ghana. 
14 
University of Ghana  http://ugspace.ug.edu.gh
1.6 GENERAL OBJECTIVE 
To determine the prevalence of cervical pre-cancer lesions and HPV infection and their 
associated risk factors among women living in Akuse sub-district in Ghana. 
1.6.1 Specific Objectives 
1. To determine the proportion of women at risk of HPV infections and cervical pre-
cancer lesions positivity based on the distributions or profiles of the accepted risk 
factors among the women. In order to ascertain the community-risk status. 
2. To determine the difference in HPV prevalences between analysis of self-collected 
specimen and health personnel collected specimen. 
3. To assess the level of agreement and concordance between self-collected and health 
personnel collected specimen for HPV genotype detection. 
4. To determine the variants of the detected oncogenic HPVI6, 18 and 45 genotypes 
present in the study population. 
5. To determine the prevalence of cervical pre-cancer lesions. 
6. To determine participant characteristics which influence the reporting of the women 
for cervical cancer screening specimen collection 
7. To determine the factors influence the use of condom, presence of cervical lesions and 
HPV positivity 
15 
University of Ghana  http://ugspace.ug.edu.gh
1.6.2 Hypotheses 
Since there was no report of a study conducted in Ghana in respect of determining the burden 
of HPY infection and cervical lesions at the community level, as at the time of this study, the 
hypotheses of this study was based on a study designed in a similar manner by Thomas et ai. , 
2004. This was the major community based study on cervical cancer and HPY screening in 
West Africa as at the time of the design of the study reported in this thesis. That study 
reported an overall HPY prevalence of 26.3% and an odds ratio of 3.0 for the association of 
HPY positivity with cervical pre-cancer lesions. Therefore, the hypotheses of this study were 
as follows ; 
I. The prevalence of HPY infection as determined with the health personnel collected 
specimen of women in the Akuse sub-district of the Lower Manya Krobo district in 
the Eastern Region of Ghana was greater than 26.3%, as reported in the community-
based study of HPY and cervical pre-cancer lesions in Nigeria (Thomas et aI. , 2004). 
2. Women, living in the Akuse sub-district of the Lower Manya Krobo district in the 
Eastern Region of Ghana, who were infected with an HPY (exposure) were three 
times more likely (odds ratio of 3.0) to test positive for cervical pre-cancer lesion 
(outcome) compared to women who were not infected with an HPY. 
16 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER TWO 
LITERATURE REVIEW 
2.1 THE CERVIX UTERI 
The principles and application of cervical screening and related methods that are discussed 
later in this thesis would be better understood with basic description/knowledge of the 
anatomy of the uteri cervix. This description is provided in this section. This part of the 
female reproductive organ is the narrower and lower part of the uterus (about 3 - 4 cm) that 
extends into the vagina (LEPPERT, 1995). Although the shape and size of the cervix is 
influenced by the age, hormonal status, and parity of the woman, it is usually cylindrical or 
conical in shape (Figure 2.1) (WHO; Chronic Diseases and Health Promotion Group, 2006). 
The outer half of the cervix extends into the vagina (2.5 - 3.5 cm in diameter) and is the part 
that is visible during speculum examination. The opening of the cervix into the vagina, the 
external os, is seen as a centrally placed small round opening in women who have not given 
birth via the vagina, but as an irregular slit after childbirth per vaginam; the area external to 
this opening and that forms the intravaginal portion of the cervix for girls before puberty and 
is exposed to the vaginal secretions is the ectocervix. The external os leads into the upper half 
of the cervix , through the cervical canal and ends at the internal os. The upper half of the 
cervix that is not seen during a speculum examination is referred to as the endocervix and is 
traversed by the cervical canal (WHO; Chronic Diseases and Health Promotion Group, 2006; 
Sankaranarayanan et aI. , 2003) . 
17 
University of Ghana  http://ugspace.ug.edu.gh
corpus 
internal os 
------l-----f------- endocervix 
cervix -f--Ic-- cervical canal 
___________ J_ _____ _ external os 
'---L=~ __ ectocervix 
transformation 
zone 
\ , 
\ , 
Figure 2. 1 Diagram of the uterus of a woman within her reproductive age. 
Adapted and modified form comprehensive cervical cancer control WHO, 2006 
The lining of the cervix is made up of two types of epithelia, each at a different location. The 
stratified squamous epithelium covers most part of the ectocervix, extending towards the 
vagina and similar to that found in the vagina. This consists of layers of flattened cells, which 
are mainly protective. The other, the columnar (glandular) epithelium, forms the lining of the 
cervical canal and some part of the ectocervix closer to the external os. This consists of a 
sing le layer of rectangular shaped mucus secreting cells . The junction between these 
epithelia, which develops from birth to puberty as the uterus and cervix increase in size, 
referred to as the original squamocolumnar junction (OSeJ), forms a sharp visible line 
(Sankaranarayanan et al. , 2003). 
After puberty and during the reproductive period, the growth of the female reproductive 
organ results in the outward extension of the endocervical canal on to the ectocervix, 
18 
University of Ghana  http://ugspace.ug.edu.gh
exposing the columnar epithelium to the acidic condition of the vagina and pushing the 
original SCJ outwardly away from the external os. The exposed columnar epithelium gets 
gradually replaced by new stratified squamous epithelium in a process referred to as 
squamous metaplasia; resulting in a new squamocolumnar junction (NSCJ); depending on the 
woman ' s age, hormonal status, birth trauma, pregnancy status and use of oral contraceptives, 
the NSCJ is at varying location from the external os (Sankaranarayanan et al., 2003). The 
area between the original and new SCJ is termed the transformational zone (TZ) 
(Sankaranarayanan el al. , 2003 ; Sellors el al., 2003). During the collection of specimen for 
cervical cancer testing, it is the ectocervix, TZ (where about 90% of cervical cancer occur 
and the mainly sampled location) and the endocervix that are sampled. 
2.2 CANCER 
In simple terms, cancer develops when cells divide/multiply without cellular control and 
progressively acquire other traits (known as the hallmark capabilities) which make them able 
to evade cell death mechanisms and eventually outgrow the surrounding cells and form 
malignant tumours. These cells may spread and invade other normal tissues of the organism, 
destroying the tissue (Hanahan and Weinberg, 2011; WHO; Chronic Diseases and Health 
Promotion Group, 2006; Hanahan and Weinberg, 2000). The resultant disease is generally 
referred to as cancer, however, based on the tissue or organ from which the uncontrolled cell 
division arose, a type of cancer is defined; there are over a 100 types of cancers identified in 
humans. Of these, the cancer of the cervix uteri is common worldwide (Ferlay el aI. , 2010) 
and is the primary interest of the study reported in this thesis. 
19 
University of Ghana  http://ugspace.ug.edu.gh
2.3 BURDEN AND EPIDEMIOLOGY OF CERVICAL CANCER 
Although cervical cancer is common among women in the active working age group of 35 
years to 59 years, there are clear regional and both inter and intra country variation in the 
incidence and mortality rates of this cancer. For instance, while cervical cancer incidence and 
mortality have been reducing since the late 1950s in developed countries, these have been 
increasing in developing countries, primarily because of the introduction of an effective and 
quality screening programmes with Pap smear testing in only developed countries 
(WHO/ICO Information Centre on Human Papillomavirus and Cervical Cancer, 2010). The 
following sub-sections describe the extent of these variations in incidence and mortality and 
thereby emphasising the need for each country or population to determine its specific burden 
and epidemiology to inform its cervical cancer prevention and control activities. 
2.3.1 Variation in the Incidence of Cervical Cancer 
The estimated global incidence of cervical cancer in the year 2012 according to the data 
publish by GLOBOCAN was 527,624 cases, with an age-standardized incidence rate (ASR) 
of 14.0 per 100,000 women per year, making it the seventh most common cancer overall and 
the fourth most common cancer among females , after breast, colorectal and liver cancers. 
This burden has continuously been excessively high among developing countries, such that in 
the years 2002, 2008 and 2012, 83%, more than 85% and 84.3% respectively of the global 
burden of cervical cancer were in developing countries (GLOBOCAN, IARC, 2012; Jemal et 
af. , 2011; Ferlay et af. , 2010; GLOBOCAN, JARC, 2008; Parkin et af. , 2005). Furthermore 
among developing countries, a variation is seen that indicates that a highly disproportionate 
burden of the incidence of cervical cancer had been in Africa (GLOBOCAN, IARC, 2012; 
Jemal et af. , 2012). This was evident by the estimated 80,400 new cases of cervical cancer in 
the year 2008 and 92,000 new cases of cervical cancer in the year 2012 (GLOBOCAN, 
20 
University of Ghana  http://ugspace.ug.edu.gh
IARC, 2012; Jemal et aI., 2012). In addition to this global regional variability in incidence, 
regional level variability in Africa has also existed. While countries in the WHO Eastern 
Africa region continued to bear the highest estimated burden of cervical cancer globally, with 
estimates of age-standardized incidence rate (ASR) of 34.5 per 100,000 women in 2008 and 
42.7 per 100,000 women 2012, countries in, the WHO Western Africa region bore the second 
highest estimated globally burden in 2008, with an estimated incidence ASR of 33.7 per 
100,000 women and the fifth highest burden in 2012, with an estimated ASR of 29.5 per 
100,000 women. Furthermore, the countries of the WHO Southern Africa region were 
estimated to have recorded the third highest burden of cervical cancer in 2008, with an 
incidence ASR of 26.8 per 100,000 women and the third in 2012, with an estimated incidence 
ASR of 31.5 per 100,000 women (GLOBOCAN, IARC, 2012; Jemal et aI., 2011; Ferlay et 
aI., 2010; GLOBOCAN, IARC, 2008; Parkin et aI., 2005). 
The estimated high burden of cervical cancer reported for the WHO Africa regions, as 
described in the previous paragraph was also distributed varyingly at the country level 
(GLOBOCAN, IARC, 2012; Jemal et aI., 2011; Ferlay et al., 2010; GLOBOCAN, IARC, 
2008; Parkin et al., 2005). For instance, country-specific estimates of cervical cancer 
incidence, reported as age-standard ised rates per 100,000 women, in 2012 for countries of the 
Eastern African Region ranged between 15.0 in Mauritius and 75.9 in Malawi (GLOBOCAN, 
fARC , 2012). Countries of the WHO Middle Africa Region had estimates ranging from 18.8 
per 100,000 women for the Chad to 35.5 per 100,000 women for Angola. The lowest and 
highest estimates in the Northern African Region were 2.3 per 100,000 women for Egypt and 
31.] per 100,000 women for the Western Sahara (GLOBOCAN, !ARC, 2012). In the 
Southern Africa Region, Namibia and Swaziland were expected to have the lowest and 
highest estimates, 14.7 and 53 .1 per 100,000 women respectively. The estimates of the ASR 
21 
University of Ghana  http://ugspace.ug.edu.gh
per 100,000 women for countries in the Western Africa Region ranged between 8.6 in Niger 
and 44.2 in Mali. The ASR estimates for Ghana (35.4 per 100,000) was higher than those of 
its close neighbours; Nigeria (29.0), Benin (27.6), Togo (21.5), Burkina Faso (23.3), and Cote 
d'Ivoire (21.7). This converts to 3052 new cases in the year 2012, accounting for 35.4% of all 
cancers in Ghana (GLOBOCAN, IARC, 2012). 
It is worth noting that these country-specific estimates for Africa were based on very few 
empirical data from mostly cancer registries in mostly one or two major cities of some of the 
countries. For example, countries such as Ghana, where there were no empirical cancer 
registry data, estimates were obtained by modelling data from the neighbouring countries; 
Mail, Nigeria, and Cote d'Ivoire. Also, the difference between the GLOBOCAN 2012 
estimates and those of its earlier versions, do not indicate a time trend change, since the few 
data and modelling methods used in obtaining these estimation had been changing 
significantly each time for most countries, including Ghana (GLOBOCAN, IARC, 2012; 
lemal et aI., 2012; WHO/ICO Information Centre on Human Papillomavirus and Cervical 
Cancer, 2010; GLOBOCAN, fARC, 2008). 
In addition to the geographical variability in the incidence of cervical cancer discussed so far, 
there is also variation in the burden of cervical cancer with the age of women. Generally, 
cervical cancer is common among women within the working age group (Gustafsson et aI., 
1997). Data that show the worldwide age-specific incidence rates indicated that cervical 
cancer generally increased with age, but there is a sharp rise between the ages of 35 years and 
54 years and a peak between 50 years and 55 years (Parkin et aI., 2005). However, country-
based age-specific incidence varied significantly and this variation has been influenced by the 
availability of quality of screening programmes in developed countries (WHO/ICO 
Information Centre on Human Papillomavirus and Cervical Cancer, 20 10). 
22 
University of Ghana  http://ugspace.ug.edu.gh
The variability in the incidence of cervical cancer within Ghana is not currently detenninable 
because the few studies on cervical cancer thus far have been centralised and hospital based. 
For instance in one such study at the Korle-Bu Teaching Hospital, where most of the women 
were referred from almost every part of the nation, it was observed that 58.3% of the reported 
cases of gynaecological cancers were cancers of the cervix, with a prevalence of 1.6% 
(Nkyekyer, 2000). on the other hand, the only community-based study, which was in Accra, 
reported a 0.6% prevalence for high grade squamous intraepithelial lesion but did not detect 
any case of cervical cancer (Grace el al. , 2006). 
2.3.2 Variation in the Mortality of Cervical Cancer 
The global burden of the mortality due to cervical cancer, just like that for incidence, was the 
fourth leading cause of cancer death among females in the years 2012, and it was estimated 
that 265,653 deaths had occurred in that year, resulting in a mortality to incidence ratio of 
50.3%. As was observed with the global incidence of cervical cancer, there were 
geographical variation in the distribution of the burden of cervical cancer mortality 
(GLOBOCAN, fARC, 2012). Specifically, while cervical cancer mortality rates in developed 
countries have been significantly low, the rates have been high in developing countries 
(GLOBOCAN, TARC, 2012; WHO/ICO Information Centre on Human Papillomavirus and 
Cervical Cancer, 2010). Furthermore, although developing countries carry the highest burden 
(about 87%) of cervical cancer deaths; there are shape variations between these developing 
regions of the world. For instance, 144,434, 60,098 and 28,565 of the global deaths due to 
cervical cancer were estimated to have occurred in Asia, Africa and Latin America/the 
Caribbean respectively in the year 2012. Once again, the high mortality burden (age-
standardized rate) of 17.5 per 100,000 women per year reported for Africa, (cervical cancer 
was the second to breast cancer death among females) was varyingly distributed among the 
23 
University of Ghana  http://ugspace.ug.edu.gh
WHO Africa regions. For example" the estimated, age-standardize mortality rates per 
100,000 women per years of27.6, 22.2 and 20.7 for the WHO Eastern Africa, WHO Middle 
Africa Regions and Melanesia accounted for the three highest mortality rates due to cervical 
cancer worldwide in 2012, while that for the Southern Africa Region was 17.9 
(GLOBOCAN, IARC, 2012). Worldwide age-specific mortality rates of cervical cancer 
showed an increase with a sharp peak at age 44 years and then a decline after 50 years. 
Country level variations In the projected estimates of mortality were also evident. 
Specifically, 7 out of the 17 countries in the Eastern Africa Region were expected to record 
high rates of between 32.4 and 49.8; notable among these were Malawi, Mozambique and 
Comoros, while 2 of the 17 were expected to record low rates of below 10.0. In the Western 
Africa Region, Mali was expected record the highest mortality per 100000 women of 31.2 
while Ghana was expected to have record the seventh highest mortality of 18.9; which was 
higher than those of its neighbouring countries; Cote d'Ivoire (14.7), Burkina Faso (18.6), 
Togo (13.8), Nigeria (17.5) and Benin (17.8) (GLOBOCAN, IARC, 2012). 
Although there are not enough data to determine the variability of cervical cancer mortality 
within Ghana over a long period of time, an earlier 10-year review of autopsies and hospital 
mortality for the period 1990 and 2000, showed cervical cancer as the fourth (mortality 
ASCAR of 8.47) leading cause of cancer deaths in Ghana (Wiredu and Annah, 2006) . 
However, it must be noted that the availability of treatment and the state of the cancers at 
reporting to the hospital are worse for most of the other cancers compared to cervical cancer 
in Ghana. Additionally, relative to other cancers reported in Ghana, cervical cancer was 
estimated to be the leading cause of death in 2012 with an annual crude mortality rate per 
100,000 females of 18.9; amounting for an estimated 1556 of the cancer related deaths in 
24 
University of Ghana  http://ugspace.ug.edu.gh
Ghana (WHOIICO Information Centre on Human Papillomavirus and Cervical Cancer, 
2010). 
2.3.3 Summary of Burden of Cervical Cancer 
The burden (incidence and mortality) of cervical cancer is high and highest among 
developing countries. Data on the epidemiology of cervical cancer varies geographical both 
between and within global regions, continental regions and countries, so much so that for 
each country to be able to plan effectively to prevent and control cervical cancer, there is a 
critical need to determine country specific empirical data, since data such as published by 
Globocan are based on statistically modelled estimates and so are not a complete definition of 
the state in a country. The need to develop a cervical cancer prevention and control plan for 
Ghana (if not a control programme) is further stressed by the high burden of cervical cancer 
in the West Africa region and Ghana in particular, the study reported in this thesis aimed to 
provide most of these baseline data needed for the development of a cervical cancer 
prevention plan for Ghana. 
2.4 THE NATURAL HISTORY OF CERVICAL CANCER 
This section on the natural history of cervical cancer presents an overview of the stages in the 
development of cervical cancer. A fair description and understanding of this makes the 
appreciation of data collected in this study much easier, since it informed the collection of 
specimen, identification of known risk factors and the interpretations of the results of cervical 
screening activities. For Instance, it is known that most neoplasia and oncogenic influence 
(between 80% and 90% of cervical cancers) occur in the transformation zone due to its high 
turnover of cells. The neoplasia arising from the squamous cells in this zone and are referred 
to as squamous cell carcinoma. The others, approximately 12.0% and 4.0% of cervical cancer 
25 
University of Ghana  http://ugspace.ug.edu.gh
are adenocarcinomas and adenosquamous carcinoma respectively. These develop in the 
glandular, mucus-secreting surface cells of the cervical canal leading into the uterus (The 
International Collaboration of Epidemiological Studies of Cervical Cancer, 2007; Wang et 
a!., 2004; Shingleton et al., 1995). These facts informed the collection of three cervical 
specimens from the TZ, external os and endo-cervix during for the Pap smear and HPV 
testing in this study. 
Within the normal transformation zone, where the columnar cells of the epithelia are being 
replaced, the cells of the epithelia go through a series of processes including a cell maturation 
process. When this process is disrupted, mostly by HPV infection, lesions develop that have 
the potential of malignancy, referred to as dysplastic lesions; these well-defined lesions are 
mostly transient (McCredie et a!., 2008). Therefore, the detection of a lesion in a cross-
sectional study only defines a potential risk for cervical cancer. However, due to HPV 
infection and persistence, as a results of participation in high risk sexual behaviour, a few of 
these lesions may progress to higher grade lesions, and only some of the higher grade lesions 
may progress to invasive cervica l cancer when they are not effectively treated (McCredie et 
aI. , 2008). These mean that sexual risk behaviours are risk factors for the acquisition and 
persistence of HPV infection; HPV infection and reproductive risk behaviours are risk factors 
for the deve lopment and progression of cervical pre-cancer lesions; cervical pre-cancer 
lesions are a risk factor for the development of cervical cancer. The roles of the sexual and 
reproductive risk behaviours and that of HPV in the development of cervical cancer are 
discussed more detail in sections 2.5 and 2.7 respectively. 
This continuous increase in the severity of the precursor lesions (progression) may take 
between 10 and 30 years to culminate in cervical cancer, providing a long window of 
opportunity for detection by screening and for treatment to prevent the development of 
cervical cancer. These precursor lesions have been characterised and described over the years 
26 
University of Ghana  http://ugspace.ug.edu.gh
by different terminologies, with some level of overlaps, and these are described in the next 
sub-section (WHO; Chronic Diseases and Health Promotion Group, 2006; Sankaranarayanan 
et af., 2003; Sellors et aI., 2003). 
Sexual risk behaviour 
Acquisition 
Reproductive risk 
behaviour HPV infection * 
Persistent 
Cervical pre-cancer lesions* 
Progression Regression 
Normal 
Figure 2.2: An illustration of the multifactorial and a trifocal outcome model involved 
in the development of cen'ical cancer. 
Bolden text indicates outcomes, however, two of the three (*) are also risk factors leading to cervical cancer. 
T he broken line arrows indicate very less likelihood of occurrence. 
2.4.1 Precursor Lesions of Cervical Cancer 
Over the years, different cytology and histology based systems have been developed and 
improved for the description of cervical lesions (Table 2.1) (WHO; Chronic Diseases and 
Health Promotion Group, 2006; Sankaranarayanan et af., 2003; Sellors et af., 2003). 
27 
University of Ghana  http://ugspace.ug.edu.gh
For instance two of the histology classification (based on the appearance of tissue section) 
systems use the following terminologies, 
I. The dysplastic classification with grading consisting of normal, atypia, mild, 
moderate, severe carcinoma-in-situ and invasive carcinoma 
2. The cervical intraepithelial neoplasia (CrN) classifications with grading consisting of 
grade I (CrN 1), grade 2 (CrN 2), grade 3 (CIN 3) and invasive carcinoma. 
for cytology (based on the appearance of cells), the first classification system was the 
Papanicolaou classification, however, its use was limited because it lacked specificity and 
was not in line with the known biology of cervical neoplasia. Following this, the British 
developed their own cytology classification system which used the term dyskaryosis as the 
cytological equivalent of the histologic term dysplasia. Other systems such as the Bethesda 
classification with grading consisting of negative, atypical squamous cells of undetermined 
significance (ASC-US) or atypical glandular cells of undetermined significance (AGC-US), 
low-grade squamous intraepithelial lesion (SIL), high-grade SIL and invasive squamous 
ladenocarcinoma carcinoma, have been developed and improved (WHO; Chronic Diseases 
and Health Promotion Group, 2006; Sankaranarayanan et al., 2003; Sellors et aI., 2003). The 
term atypia was used by all classification systems without strict criteria making studies 
difficult to compare. 
2.5 RISKS FACTORS OF CERVICAL CANCER 
As indicated in section 2.4 sexual risk behaviours and reproductive risk behaviours are the 
factors that following an HPY infection, predict the development and progression of cervical 
lesion to cervical cancer (Picot et aI., 2012; Shields et al., 2004; Bosch et al., 2002; 
Walboomers e/ al. , 1999). This section describes each of the categories of the sexual and 
28 
University of Ghana  http://ugspace.ug.edu.gh
reproductive risk behaviours, which have been shown as risk factors for cervical cancers 
globally. The sections also describe the sub-categories of each risk behaviour showing which 
are high risk and which are low risk. This is useful for this study because, small cross-
sectional studies, such as the one reported in this thesis, are not appropriate for identifying 
these characteristics as risk factors (or correctly quantify the association), primarily because 
of their small sample sizes are often not enough (not powered enough). Often large 
longitudinal and case-control studies, multi-centre studies or meta-analysis of data of a 
number of related studies (The International Collaboration of Epidemiological Studies of 
Cervical Cancer, 2007) are needed to identify each of these behaviours as risk factors within 
a population (Bosch et aI., 2002; Walboomers et aI., 1999). In such small sample-sized cross-
sectional studies, the distribution of the participants among the low and high risk sub-
categories of each of these risk behaviours becomes a good indicator of the likelihood of a 
related outcome (Erickson et aI., 2013; Castellsague and Munoz, 2003a). For example, if 
most of the participants in a relatively small cross-sectional study report an age at first sexual 
intercourse of less than 20 years (high risk sub-category), this will imply that that population 
is most likely going to be at a high risk of HPY positivity. This mode of assessment was used 
in this study to determine if the Akuse sub-district was high-at-risk for HPY infection and 
then confimled by the results of the I-IPY test. 
2.5.1 Sexual Risk (Behaviours) Factors 
2.5.1.1 Number ofM ale Sexual Partners 
The current and life-time numbers of male sexual partners of a woman, have consistently 
been shown to be risk factors for acquiring an HPV infection. Most importantly, these factors 
have shown the strongest associations as independent risk factors for the acquisition of HPV 
infection (Lenselink et aI. , 2008) and for the development and progression of precursor 
29 
University of Ghana  http://ugspace.ug.edu.gh
lesions to invasive cervical cancer in most populations and among different age groups (The 
International Collaboration of Epidemiological Studies of Cervical Cancer, 2007). 
Specifically, there two sub-categories of these risk factors; those who have had less than 5 
and those having had 5 or more; the latter is the high risk sub-category and have recoded 
significant odds ratio and relative risk greater than 1.6, This odds ratio increases with increase 
in the number of both current and life-time number of sexual partners (Erickson et aI. , 2013; 
International Collaboration of Epidemiological Studies of Cervical Cancer, 2007; Shields et 
ai., 2004). Therefore, if a majority of the participant in this study report 5 or more current or 
life-time male sexual partners, then the community will be high risk for HPY infection. 
2.5.1.2 Age at First Sexual Intercourse (AFSI) 
Early age of first sexual intercourse is also one of the strongest risk factors for both the 
acquisition of HPY infection and the progression to high grade lesion in most studies 
(Erickson et ai., 2013; Bahmanyar et ai., 2012; Berrington de Gonzalez el aI., 2004). There 
are two sub-categories of AFSI, these are ages less than 20 years (sometime less than 18 
years) and ages 20 years and older. The former « 20 years or younger AFSI), is the high risk 
sub-category and is associated with a high risk for acquisition of an HPY infection; ages less 
or equal to 17 years showing the highest association with significant odds ratio (ORs) ranging 
between 1.8 and 2.9 (Erickson el aI. , 2013; Bahmanyar et ai., 2012; Berrington de Gonzalez 
el al. , 2004). However, in some age limited cross-sectional studies and in some atypical 
populations (Johnson el ai. , 2012; Smith et ai., 2002a; Lenselink et ai., 2008), AFSI was not 
significantly associated with the risk of HPY acquisition. For one of those studies, a derived 
risk factor, defined as sexual age (the difference between the age at first sexual intercourse 
and current age) showed significant risk for the acquisition of HPY infection (Lenselink et 
al. , 2008). The sub-categories of this derived risk factors are < II years, 11 -19 years (these 
30 
University of Ghana  http://ugspace.ug.edu.gh
are the high risk groups), 20-29 years and >30 years (these are low risk sub-categories).This 
risk factor was also derived in the study reported in this thesis. 
2.5.1.3 Sexually Transmitted Infections 
Epidemiological studies of sexually transmitted infections (STIs) other than HPV, with 
respect to the risk of cervical precursor lesions and cancer have demonstrated associations 
with Chlamydia trachomatis (CT), herpes simplex virus type 2 (HSV-2) and HIV (discussed 
as a immunocompromised risk factor), after controlling for strong risk factors such as HPV 
infection. A pooled analysis of multi-centre case-control studies conducted by the JARC 
provided evidence for significant associations ofHSV-2 with SCC (odds ratio, OR= 2.2; 95% 
CI: 1.4- 3.4) and ADC or ADSC (OR = 3.4; 95% CJ: 1.5-7.7) in a group of HPV positive 
women. The association with Chlamydia trachomatis (CT) showed an increased risk for 
cervical cancer (OR = 1.8; 95% CI, 1.2-2.7). The association of cervical lesion with 
Chlamydia trachomatis has been determined based on both a history of infection (determined 
by antibody detection) and concurrent infection with HPV (Munoz et aI., 2006; Sarnoff, 
2005; Smith el aI., 2002a, 2002b, 2004). Concurrent CT infection has been shown as an 
independent risk factor (ORs of between 1.1 and 4.1) that promotes HPV infection (by 
inducing cervical microabrasions and inflammation) and HPV persistence (by immunological 
mechanisms) (Burchell el af., 2006; Sarnoff, 2005). This formed the bases for using STI 
infection in this study as a surrogate indicator for a community at high risk for HPV infection 
and therefore the selection of the Lower Manya Krobo district for this study. 
2.5.1.4 Inconsistent use of Condom 
The promotion of the use of condom is one very important primary preventive method for 
STls in general and more so for the prevention of HPV infection and cervical lesion (WHO; 
31 
University of Ghana  http://ugspace.ug.edu.gh
Guideline Development Group, 2014; WHO; Chronic Diseases and Health Promotion Group, 
2006). The consistent use of condom has been largely shown as a protective factor against 
HPY acquisition, promotes HPY clearance thereby reducing the risk of the development of 
cervical lesion and promoting regression of cervical pre-cancer lesion (WHO; Guideline 
Development Group, 2014; WHO; Chronic Diseases and Health Promotion Group, 2006; 
Shields et aI., 2004; Shepherd et aI., 2000). 
Although it is the consistent use of condom that provides protection against HPY infection, it 
use provides partial protection since it does not protect against HPY infection that may be 
due to genital skin-to-skin contact or self-inoculation (Burchell et aI., 2006; Winer et al., 
2006) . However, because it provides protection against other STI that promote the acquisition 
and persistence of HPY infection as well as the progression of pre-cancer lesions (section 
2.5.1 .3), it is very important to cervical cancer control and prevention. Therefore, it is 
important to determine the extent of the consistent use of condom during cervical screening 
as well as to determine factors that influence its consistent or inconsistent use. This will be 
useful in planning and targeting education on condom use and the need for consistent usage 
among a population. To this end, the study reported in this thesis determined the extent of 
condom use among women participating in the screening activity and analysed for the 
participant characteristics that may influence the inconsistent use of condom among the 
participants. 
2.5.2 Reproductive Risk characteristics (Known Risk Factors) 
2.5.2.1 Multiparity 
The epidemiological data that support the role of parity (defined either as the number of full 
term pregnancies or number of live births) as a risk factor for cervical pre-cancer lesion and 
32 
University of Ghana  http://ugspace.ug.edu.gh
cancer have mostly been from systematic/pooled reviews and meta-analysis of mostly case-
control studies and a few large sample cohort studies, as was the case for most of these risk 
factors. Individual relatively small-sample studies (case-control or cross-sectional) are 
usually challenged by choice of controls, number of participants in subgroups and time of 
assessment of parity (at recruitment, during the study or at the end of follow-up) 
(Castellsague and Munoz, 2003a). Although parity is of a very low and non-significant risk 
for HPV acquisition and for low grade precursor lesions development, there are inconsistent 
reports of an association with HPV persistence and a moderate risk for progression of 
cervical lesions to Carcinoma-in situ (CIS) and invasive cervical cancer (Kim e/ af., 2012; 
Vaccarella et aI., 2006). The two major sub-categories of parity are less than 3 full tenn 
pregnancies and 3 or more pregnancies. The latter (full term pregnancies of more than 3) are 
often significant high risk factors and mostly show a linear trend with the occurrence of CIS 
and invasive cancer. The 95% CI of OR of these association could range between 1.3 and 7.9 
(Castellsague and Munoz, 2003a). Therefore, for this study, since the sample size was 
relative very small, determining the odds ratio of this association will be misleading. 
However, the distribution of the participants within the sub-categories of these characteristics 
were used as an indication of the likelihood of detecting cervical lesions among the women . 
This likelihood was thereafter compared with the detection of cervical lesion among the 
women by Pap test. 
2.5.2.2 Oral Contraceptive Use 
The current and duration of use of oral contraceptive (OC) as well as the duration since the 
stop of oral contraceptive use have consistently not been a significant risk for the acquisition 
of HPV infection. However, while the current and duration of use of oral contraceptive have 
been associated with an increased risk for the progression to high grade lesions and invasive 
33 
University of Ghana  http://ugspace.ug.edu.gh
cervical cancer, the duration since the stop of oral contraceptive use has been associated with 
a reduced risk of these (Lenselink et aI., 2008; Vaccarella et af. , 2006). The two sub-
categories for these risk factors are usage for less than 5 years and usage for between 5 and 
10 years. Higher risks are observed among women who have been using oral contraceptives 
for 5-10 years (OR of between 1.4 and 5.4) and the highest risk was observed among users of 
more than 10 years (OR of between 2.0 and 7.8) (International Collaboration of 
Epidemiological Studies of Cervical Cancer, 2007; Castellsague and Munoz, 2003a; Smith et 
ai., 2003). Different study designs, and differences and difficulties in the choice of controls 
for case-control designs may influence the varying levels of risk as measured either by odds 
ratio or relative risk for cervical cancer in different studies (Shields et aI., 2004). 
2.5.3 Other risk factors 
2.5.3.1 Smoking Status 
Current smoking status is associated with an increased risk of precursor lesions and invasive 
cervical cancer and as such is thought to playa role in HPV persistence/clearance. This risk 
has been low to moderate and inconsistent in most populations (Kapeu et aI. , 2008; Lenselink 
et aI., 2008) and has been shown to be confounded by cofactors of sexual behaviour, such as 
the number of sexual partners (Vaccarella et al., 2008). This confounding may come about 
because smoking is often associated with promiscuous life styles in most populations. Based 
on pooled analyses of both prospective and retrospective cohort studies, smoking generally is 
moderately associated with the risk of developing cervical cancer; OR range between 1.09 
and 1.88 (Erickson et ai., 2013 ; Vaccarella et af., 2008; Berrington de Gonzalez et ai. , 2004). 
However, compared to never smokers, current smokers are at the highest risk followed by 
ever smokers, while past smokers (of about 10 years) are often not-significantly at risk of 
cervical cancer as compared to never smokers (Sarian et af., 2009). This is often explained on 
34 
University of Ghana  http://ugspace.ug.edu.gh
the bases that during the period of smoking the related oncogenic agents accumulates in the 
body due to a slow clearance/excretions, However, following the stop of smoking, the level 
of the accumulated oncogenic agents reduces overtime to the levels as in non-smokers 
(Sarian et af., 2009). The intensity of smoking of more than 15 or 20 cigarettes per day is 
often significantly associated with the risk for HPV prevalence and progression to cervical 
cancer, but the duration of smoking has not shown any such association (Erickson et ai. , 
2013; Vaccarella et af., 2008; Berrington de Gonzalez et aI., 2004). 
2.5.3.2 Immunological Status 
The two most important immunocompromised states, in relation to cervical cancer are HIV 
infection/AIDS and use of immunosuppressive medication, because they have been shown to 
be significantly associated with an increased risk of cervical HPV infection and progression 
of pre-cancer lesions (Erickson et aI., 2013; Firnhaber et aI., 2009; Veroux et aI., 2009). In a 
meta-analysis of the incidence of cancers among these peculiar populations, the risk of the 
incidence of HPV related cancers were increased, with HIV infected population bearing the 
highest risk for most of the cancers studied. Specifically for cervical cancer, the meta-
analysis ' standardized incidence ratio (SIR) range between 2.9 and 11.3 for HIV infected 
women and between 1.3 and 3.3 for organ transplant recipients undergoing 
immunosuppressive therapy (Grulich et ai. , 2007). Issues of the interaction of HIV and HPV 
(including type specific concordance) are thought to account for the higher risk for the 
HIV/AIDS population although this is not clearly confirmed (Firnhaber et af. , 2009; 
Mbulawa et af. , 2009; Palefsky, 2009; Grulich et af., 2007; Palefsky and Holly, 2003). 
Therefore, if a population has a large proportion of its members with such 
immunocompromised statuses, that population will be at a high risk of bearing a high burden 
of HPV infection and cervical pre-cancer lesion. The district that hosted the study (the Lower 
35 
University of Ghana  http://ugspace.ug.edu.gh
Many Krobo district) described in this thesis has been reported to be one of the high burden 
of HIV infection in Ghana. 
2.5.3.3 Inheritable Host Genetic Factors 
Although this risk factor was not considered in this study, it is worth describing its 
contributions to cervical cancer. Based on limited data form a number of population based 
cohort studies, some variants of host inheritable genes have been suggested as possible risk 
factors for cervical cancer. For example, genetic variations in the HLA DOB1 and CD83 
genes of the innate immune system have been shown to be associated with invasive cervical 
cancer in studies of families (Neuman et aI. , 2000). Furthermore, homozygous alleles of 
single nucleotide polymorphism (SNP) at codon 72 of the PT53 gene that changes the 
nucleotide C to G and a resultant replacement of arginine with proline in the p53 protein, 
have been suggested to lead to higher susceptibility to HPV -associated cervical cancer 
(Tommasino et aI., 2003; Storey et al., 1998). This has been due to enhancement of the 
persistence of HPV infection and the progression of cervical lesions. Other such inheritable 
variations in genes such as, a C309G SNP in the MDM2 promoter, a C609T SNP in exon 6 of 
the NQOl rs1800566 gene and another SNP C465T in exon 4 of the same NQOI, have been 
reported as associated with the development of cervical cancer (Hu et af. , 2010; Wang et al., 
2010; Wang and Hildesheim, 2003) . 
2.5.4 Summary of Risk Factors 
The risk factors for the development of cervical cancer are mainly HPV infections and 
cervical pre-cancer lesions; however, some sexual and reproductive behaviour have been 
shown globally to consistently be risk factors for the acquisition of HPV infections, 
persistence of HPV infections, the development of cervical pre-cancer lesions and the 
36 
University of Ghana  http://ugspace.ug.edu.gh
progression of cervical pre-cancer lesions to cervical cancer. However, it must be noted that 
of the two or more sub-categories of each of these risk factors, one or two are high risk while 
the others are low risk behaviours. For example age at first sexual intercourse has two sub-
categories, < 20 years and 2: 20 years and the former is high risk for the acquisition of HPV 
infection. The risk status of these sub-categories have been consistent globally but may be 
correctly determined or quantified with large size studies such as, multi-centre study, pooled-
analysis or meta-analysis of studies. Therefore, this relative small-sample sized cross-
sectional study will not be appropriate to quantify or confirm the risk status of these known 
risk factors. However, the distribution of the women between the high risk and low risk sub-
categories will serve as a surrogate indicator ofthe risk status of the community. 
2.6 THE HUMAN PAPILLOMAV IRUS 
As discussed in the previous sections and indicated in figure 2.2, infections of HPV are the 
major determinant of the development of cervical cancer. As such, as part of this study, data 
on HPV infection was collected. It is therefore important to present a summary of the updated 
knowledge ofHPV, in the content of this study. 
2.6.1 Biologic Classification of HPV Types 
In the study reported in this thesis , the detection of the genotypes and variants of HPV was a 
major objective that has implications for the selection of vaccines and HPV testing kit for 
cervical cancer screening in Ghana. As such, a brief description of the classification of the 
general group of viruses that HPV belongs is presented in this section. This group known as 
Papillomaviruses are a family of viruses referred to as Papillomaviridae, which is made of 29 
genera at the most recent amendments (in 2004) (Bernard et al., 2010; de Villiers and Gunst, 
37 
University of Ghana  http://ugspace.ug.edu.gh
2009; de Yilliers et af., 2004). Of the 189 known types (more have been detected but not yet 
included in the classification), 120 have humans as host (these are referred to as human 
Papillomaviruses) while the others have other mammals, birds and reptiles as host (Bernard el 
aI. , 2010; de Yilliers el al., 2004). The 120 human Papillomavirus (HPYs) have been grouped 
in 5 genera, these are Alphapapillomavirus, Betapapillomavirus, Gammapapillomavirus, 
Mupapillomavirus and Nupapillomavirus. These consist of 30 species (identified by 
numbers) that are further grouped as types (often referred to as genotypes) and subtypes 
(known as variants or lineage of a genotype). For example, the genus Alphapapillomavirus 
includes a species number-9 that consists of the following HPY genotypes 16, 31, 35, 33, 52, 
58 and 67 (Bernard el al., 2010) and the genotype HPY16 has variants HPY16Afr-l, 
HPYI6Afri-2, HPY 16AA 1, HPYI6AA2 and others. These classifications are based on the 
level of similarity/identity in the nucleotide sequences of the L1 gene of these viruses (the 
genome of the HPY is discussed in section 2.6.2 below). An HPY genotype is different from 
another if its full length genome and/or if the DNA sequence of its Ll gene defer by more 
than 10% of that of the other genotype. If this difference is between I % and 10%, then the 
Papillomavirus is a subtype of the other genotype (as referred to as variant or lineage). If the 
difference is between 0.5% and I %, then the Papillomavirus is of the same genotype as the 
other, but it is a sub-variant (also referred to as sub-lineage) of the genotype (Chen ef aI. , 
2013 ; de Yilliers and Gunst, 2009; de Yilliers ef al. , 2004). In the study reported in this 
thesis, the genotypes of the detected HPY were determined and variants of the HPY 
genotypes 16, 18 and 45 were determined. These will help understand how different the 
variants in Ghana may be from the known variants of these genotypes and understand the 
implication of these data on the selection of an HPY testing kit for HPY screening in Ghana. 
Additionally, these may contribute to studying the evolutionary trends of these genotypes in 
Ghana and the association of HPY variants and cervical lesions and cancer development. 
38 
University of Ghana  http://ugspace.ug.edu.gh
2.6.2 The HPV Genome 
As indicated in the previous section the classification or identity of an HPV depends on the 
whole genome or the L I gene sequence and in furtherance of this understanding, this section 
provides a description of the genome of the generalized HPV. The human Papillomavirus 
particle is made-up of a circular double stranded 7.9 kbp DNA genome encapsulated by a two 
protein shell of size 50 mm. The genome consists of nine open reading frames (ORP) that 
control eight genes (Figure 2.3). Six of these genes are transcribed during the early stage of 
the life cycle of the virus and therefore are referred to as the early genes; these are E1, E2, 
E3, E4, E5, E6, E7 and E8). The others are transcribed during the late stages of the life cycle 
of the HPV and so are referred to as late genes (L1 and L2). The ninth ORP is the long 
control region (LCR) also referred to as the upstream regulatory region (URR). It controls the 
expression of the whole genome (Alp AVCl, 2012; Fertey et ai., 2010; de Villiers and Gunst, 
2009) and therefore variation in it sequence has implication for oncogenicity of the detected 
HPV variant. For this purpose as well as for variant classification, the study reported in this 
thesis determined the variation in the LCR of 3 selected genotypes . 
The other genes of the HPV genome have been shown to have the following function. Three 
of the early genes E5, E6, and E7, are oncogenes that negatively influence the host cells ' 
control of transcription and cell cycle, thereby inducing the transformation to malignancy. It 
should however be noted that the Beta-papillomavirus and the Gamma-papillomavirus do not 
have the E5 gene and therefore do not encode the E5 protein shell (Venuti et aI., 2011 ; 
Munoz et al., 2006; de Villiers et aI., 2004). 
39 
University of Ghana  http://ugspace.ug.edu.gh
o 
6000 J410() 
4000 
Figure 2. 3: An illustration of the double stranded DNA genome of HPV. 
Adopted with modifications form Villa, 2006; Sol tar et aI., 2004; Lukaszuk et aI. , 2003 
The other two, El and E2, are involved with the regulation of the transcription and 
replication of the viral genome. The E2 protein (a product of the E2 gene) also suppresses the 
expression of the E6 gene when the viral genome is not integrated in the host genome. 
Additionally, the product of the E8 gene, which is expressed as a fusion protein with the C-
terminal of the E2 protein , known as E8/\E2C (a protein known only to be expressed by HPV 
genotypes 1, 11 , 16, 31, and 33), in addition to the E2 protein repress the E6/E7 promotor, 
resulting in controlling the expression of the E6/E7 oncogenes and is also involved in the 
control of viral replication (Dreer et aI., 2016; Straub et aI., 2015, 2014; Fertey et aI., 2010). 
The function of the recently identified small putative E3 gene are yet to be identified and no 
protein expressed from it has also been identified (Alp Avcl, 2012). The two late genes, Ll 
and L2, code for the structural proteins that compose the viral capsid/shell (Munoz el aI. , 
40 
University of Ghana  http://ugspace.ug.edu.gh
2006; de Yilliers et al. , 2004) . It is worth noting that the available vaccines for HPY are viral-
like-proteins (YLP) based on the products of the LI and L2. 
Although classified as an early gene, the E4 gene product has been suggested to play roles in 
the late stage of the HPY life cycle and these roles have been a subject of speculation as it has 
remained largely unclear (Doorbar, 2013). The suggested major role of the E4 protein has 
been its involvement with virus release after assembly, due to its association with cytokeratin 
network reorganization (Doorbar, 2013). Additional suggested roles for the E4 protein have 
been in HPY genome amplification, expression of capsid proteins, and possible perturbations 
in RNA processing, signal transduction, protein degradation or cell differentiation (Doorbar, 
2013; Peh et ai. , 2004). 
2.6.3 Summary of the HPV 
The HPY is one of a number of Papillomaviruses that is made up of several genera, species, 
genotypes and variants of each genotype; the classification of which is based on DNA 
sequence variations of some of the genes of its 7.9 kbp circular genome. The function of the 
products of its nine open reading frames synergistically defines its oncogenicity and the 
variation therein with different genotypes. Implying that based on the distribution of the 
different genotypes in a population, the risk of cervical lesion development will vary. 
2.7 THE NATURAL HISTORY OF HPV INFECTION 
The natural history describes the role of HPY in the development of cervical cancer, as has 
been mentioned in the previous sections. It is the understanding of the natural history that has 
informed the development of screening methods for cervical cancer and has greatly supported 
cervical cancer related research. The study reported in this thesis was largely informed and 
41 
University of Ghana  http://ugspace.ug.edu.gh
designed to assess data representing most of the stages in the natural history. Therefore, an 
understanding of the description of this section is critical to the understanding of this study. 
2.7.1 Association of HPV with Cervical Cancer 
The causal relationship between HPV and cervical cancer was indirectly established through 
the findings of epidemiological studies that initially identified some sexual behaviours as 
significant risk factors for cervical cancer. It was so until the employment of molecular 
techniques in epidemiological studies (molecular epidemiology) that it became undoubtedly 
clear that the sexual behaviour risk factors were only indicative of the risk of HPV infection 
and that it was persistent HPV infection which was a risk factor for the development of 
cervical pre-cancer lesions (Bosch el aI., 2002; Franco el aI., 1999; Walboomers et aI. , 1999). 
Furthermore, reviews and pooled multi-centre studies have shown that between 90% and 
99% of cervical cancer cases in most world regions were associated with HPV infections and 
odds ratios (ORs) of between 35 and 130 have been reported for these associations (Munoz et 
aI. , 2003; Bosch et aI., 2002). Using the causality criteria for human diseases, Bosch et aI. , 
(2002), have discussed in detail the causal relationship between specific genotypes of HPV 
and cervical cancer. That report indicated that of the over 170 HPV characterised types, only 
about 40 infect the anogenital area, leading possibly to the development of cervical pre-
cancer lesions and eventually cancer. These 40 types are often referred to as genital HPV. 
These imply that, however few, there are some cervical cancers that may not be due to 
persistent HPV infections and so studies may report cervical pre-cancer lesions among 
women who are HPV negative. 
The major processes In the natural history of HPV infection (acquisition, clearance, 
persistence and progression) are considered as a spectrum of occurrences in association with 
42 
University of Ghana  http://ugspace.ug.edu.gh
the multi-step natural history of cervical carcinogenesis. These are illustrated In figure 2.4 
below. 
Normal Tissue ) Initial transient 
infection with HPY 
~Cle arance, ... -- -~ 6 to 12 months + : Regression ,""'<ted n«ue 
9 to 15 months 
prOgreSSiO~ ,~  Viral infection ,,  
,,  
2toSmonthj ++ -- _ .:.  
Cytologic changes Persistent infection with 
(Warts. CIN I or LSIL) high risk HPV tor ,- I year 
~ __ _  _  P e rsistent intection and Cotactors 
j (.:.g. sexual adlnty) 
Carcinoma in situ INVASIVE CANCER 
Persistent infection and!or 
multiple infection 
Figure 2. 4 An illustration of the natural histories of HPV infection and cervical cancer. 
Adopted and modified based on data from (Goldstein e/ 01. , 2009; Kahn, 2009; Schiffman and Wacholder, 
2009; Flores e / 01. , 2006; Schiffman and Castle, 2005) . The relative thickness of the arrows indicates the 
proportion of cases that progress from one state to the nex1. Clearance is immunologically mediated . 
2.7.2 Acq uisition of HPV Infection 
Human Papillomaviruses are ubiquitous and the commonest of sexually transmitted 
infections (STI), particularly common among young sexually active women. Although it 
infects the cutaneous and mucosal epithelia of most parts of the body, its infection of the 
metaplastic epithelium of the cervical transformation zone is of greater clinical significance 
(Lenselink et aI. , 2008; Moscicki et aI., 2006). It is clear that almost all forms of sexual 
contacts, including penetrative intercourse, genital contacts and genital skin to skin contact 
43 
University of Ghana  http://ugspace.ug.edu.gh
with an infected partner is necessary for the transmission of the virus. In addition, self-
inoculation also aids in the spread of the virus to other body locations, most likely by skin to 
skin contact (Moscicki et ai., 2006; Schiffman and Kjaer, 2003). Other known routes of 
transmission, including, vertical and perinatal transmission, and trans-placental (intrauterine) 
transmission, have very marginal implication for the development of cervical cancer (Bosch 
et ai., 2006). 
2.7.3 Persistence and clearance of HPV Infection 
The figure 2.3 indicates that acquiring an HPY infection does not mean a woman may 
develop cervical pre-cancer lesion, since it is the persistence of the HPY infection that is 
required for the development. However, some persistent HPY infections may eventually be 
cleared even before any cervical pre-cancer lesion develops (Goldstein et ai., 2009; Kahn, 
2009; Schiffman and Wacholder, 2009; Flores et aI., 2006; Schiffman and Castle, 2005). 
Therefore, it is possible for a woman to be HPY positive and not have any cervical pre-cancer 
lesion. Persistent HPY infection is loosely described as the detection of the same HPY type 
two or more times within a period of testing (there are no standard or recommended periods), 
otherwise that infection is said to have been cleared (Koshiol et al. , 2008; Moscicki et aI., 
2006; Schiffman and Kjaer, 2003). Different time intervals between testing have been 
reported by studies resulting in variations in the definition of HPY persistence/clearance; 
while some study have reported 6 to 12 months duration, others have reported long periods of 
between 5 years and 7 years (Koshiol et ai., 2008; Moscicki et al., 2006; Schiffman and 
Kjaer, 2003) . Inherently, there are no recommended durations indicated neither for research 
nor as part of cervical cancer prevention programme guideline. It is worth noting that 
differences in HPY type-specific clearance and persistence rates do exist, such that some 
HPY genotype are easily cleared than others. One of such order of clearance for high risk 
44 
University of Ghana  http://ugspace.ug.edu.gh
HPY infection (from highest to lowest) was reported as follows, HPY- 59, 35, 45, 18, 33 , 52, 
31 , 58 and 16, while that of persistence was HPY 52, 16, 58, 33, 31, 35, 45,18 and 59 
(Insinga el aI. , 2011). 
2.7.4 Progression of HPV Infection 
The important characteristics that influence the ability of an HPY to lead to pre-cancer 
lesions and to cervical cancer are primarily, the genotype of the infecting HPY, the length of 
persistence of the HPY infection, host characteristics and exposure to some risk factors (as 
discussed in section 2.5) (Insinga el aI., 2011; Moscicki el aI., 2006; Schiffman and Kjaer, 
2003). These imply that cervical pre-cancer lesions (cytological changes) in themselves are 
risk factors for cervical cancer since some of the pre-cancer lesion may regress (figure 2.4). 
A period of between 10 years and 25 years elapses before a persistent HPY infection may 
finally result in the development of cervical cancer (Bosch el aI., 2008; McCredie et aI., 
2008; Woodman et aI., 2007; Moscicki et aI., 2006; Schiffman and Kjaer, 2003). This 
provides a long enough time for the detection, and treatment of cervical pre-cancer lesions 
and therefore, enough time for the prevention of the cancer that may develop if the pre-cancer 
lesions are not detected and treated. 
2.7.5 Molecular Mechanisms of HPV Mediated Carcinogenesis 
The roles of the genes of the HPY genome and their relationship with the natural history of 
cervical cancer (discussed earlier) together contributes to the understanding of the mechanism 
of HPY mediated carcinogenesis. Although the related genotype specific HPY biological 
mechanism of the carcinogenesis is not completely and clearly understood, this section 
presents a generalized model of the carcinogenesis of HPYs. This is particularly true of a 
group of HPY genotypes referred to as high risk HPY types (HR HPY). As will be discussed 
45 
University of Ghana  http://ugspace.ug.edu.gh
in sub-section 2.8.1, these are more effective at causing the progression of pre-cancer lesions 
to cancer (Wang and Hildesheim, 2003). 
The principal genes responsible for oncogenicity are the E6 and E7 oncogenes, as well as the 
E2 gene, if the viral genome is not integrated (Venuti et aI., 2011; Lowy et aI., 2008; Munoz 
et aI., 2006). The oncogenicity of the E6 and E7 proteins of each HPV genotype has to do 
primarily with the strength of their interaction with a number of cellular proteins, however 
only few of these are understood (Venuti et al., 2011; Lowy et al., 2008; Munoz et al., 2006). 
The most understood of these interactions are those between the oncoproteins (E7 and E6) 
and two tumour suppressive proteins, retinoblastoma protein (pRb) and p53 protein 
respectively. The interaction between the HPV E7 protein and human pRb protein results in 
an increased DNA replication in the infected cell and therefore unscheduled (uncontrolled) 
cell-cycle activities occur. Such cells are expected to be lead to cell death (be killed) through 
a process mediated by the actions of the protein p53. However, the HPV E6 protein binds 
with this p53 protein, and leads to the degradation of the p53. As a result, there is a loss of 
cell cycle control, and these cells usually are not killed. These result in an increased loss of 
cell-cycle control, and genomic instability, which may eventually lead to cell 
immortalization and malignant transformation (Venuti et aI., 2011; Lowy et aI., 2008; Munoz 
et aI., 2006). For a viral genome that is not integrated in the host genome, the E2 viral protein 
negatively influences the transcription (formation) of the E6 and E7 viral proteins, (reduces 
the amounts produced). However, when the viral genome integrates into the host genome, the 
integration occurs at the E2 gene of the virus, thereby disrupting the formation of E2 protein 
and related function. This ensures that the E6 and E7 proteins are produced at a faster rate, 
with the ultimate result of cancer development (Lowy et al., 2008; Munoz et aI., 2006). It is 
therefore understandable that HPV genotypes that are able to integrate their genome more 
often tend to be more efficient at tumour progression. The role of another gene product, the 
46 
University of Ghana  http://ugspace.ug.edu.gh
in sub-section 2.8.1, these are more effective at causing the progression of pre-cancer lesions 
to cancer (Wang and Hildesheim, 2003). 
The principal genes responsible for oncogenicity are the E6 and E7 oncogenes, as well as the 
E2 gene, if the viral genome is not integrated (Venuti et aI., 20 11; Lowy et al., 2008; Munoz 
et al., 2006). The oncogenicity of the E6 and E7 proteins of each HPV genotype has to do 
primarily with the strength of their interaction with a number of cellular proteins, however 
only few of these are understood (Venuti et aI., 2011; Lowy et al., 2008; Munoz et al., 2006). 
The most understood of these interactions are those between the oncoproteins (E7 and E6) 
and two tumour suppressive proteins, retinoblastoma protein (PRb) and p53 protein 
respectively. The interaction between the HPV E7 protein and human pRb protein results in 
an increased DNA replication in the infected cell and therefore unscheduled (uncontrolled) 
cell-cycle activities occur. Such cells are expected to be lead to cell death (be killed) through 
a process mediated by the actions of the protein p53. However, the HPV E6 protein binds 
with this p53 protein, and leads to the degradation of the p53. As a result, there is a loss of 
cell cycle control, and these cells usually are not killed. These result in an increased loss of 
cell-cycle control, and genomic instability, which may eventually lead to cell 
immortalization and malignant transformation (Venuti el at., 2011; Lowy el at., 2008; Munoz 
et aI., 2006). For a viral genome that is not integrated in the host genome, the E2 viral protein 
negatively influences the transcription (formation) of the E6 and E7 viral proteins, (reduces 
the amounts produced). However, when the viral genome integrates into the host genome, the 
integration occurs at the E2 gene of the virus, thereby disrupting the formation of E2 protein 
and related function . This ensures that the E6 and E7 proteins are produced at a faster rate, 
with the ultimate result of cancer development (Lowy el aI., 2008; Munoz et ai., 2006). It is 
therefore understandable that HPV genotypes that are able to integrate their genome more 
often tend to be more efficient at tumour progression. The role of another gene product, the 
46 
University of Ghana  http://ugspace.ug.edu.gh
E5 protein, IS still not clearly understood; however, it has been shown to be a weak 
transforming protein in vitro. It interacts with a number of cellular proteins involved with 
HPV life cycle, cell transformation, cell proliferation and evasion of the immune response 
(Greco et aI., 2011 ; Venuti et at. , 2011 ; Genther et at., 2003). For example HPV 16 E5 protein 
has been shown to interact with and down-regulate the Major Histocompatibility Complex 
(MHC) class ] and class II leading to the evasion of immunosurveillance of the immune 
system. ]t interacts and activates the epithelial growth factor receptor (EGF -R) involved with 
cell proliferation and also interacts with the HPV E6 oncoprotein to enhance the oncogenic 
activities of both proteins (Greco et at., 2011; Venuti et al. , 2011). The HPV 16 E5 has also 
been shown to be able to down-regulate the expression of the tumour suppressor proteins p21 
and p27 (Venuti et al., 2011). Therefore, a high prevalence of HR HPV infection increases 
the risk for the development of pre-cancer lesions and cervical cancer. Additionally, and 
although not part of the study reported in this thesis, sequence variation within these genes 
(E5, £6, £7) may have implication for variation of the risk of developing pre-cancer lesion 
within a population. 
2.7.6 Summary of the Natural History of HPV infection 
The natural history indicates that HPV infections are necessary for the development of 
cervical pre-cancer lesion and therefore cervical cancer, however, most HPV infection do not 
lead to cervical pre-cancer lesions because the infection are often cleared by human immune 
system. It is the few HPV infections that persist that are able to induce the development of 
cervical pre-cancer lesion. Furthermore, the cervical pre-cancer lesions that may develop may 
also progress to cervical cancer or may regress. The small proportion of women who may 
develop cervical pre-cancer lesions and cervical cancer do so as a result of the additional 
continuous exposure to risk factors, most of which are the sexual and reproductive behaviours 
47 
University of Ghana  http://ugspace.ug.edu.gh
of these women. The long duration between acquiring and maintaining an HPV infection, and 
the development of cervical cancer (10-25 years) provide a long enough window of 
opportunity to prevent the development of cervical cancer. 
2.8 EPIDEMIOLOGY OF HPV INFECTION 
Data generated by the study reported in this thesis was primarily focused on providing the 
country-specific HPV molecular epidemiological information for a community in Ghana. 
This section provides the necessary background information at the Global , African and 
Ghanaian levels so as to put in context the contribution of this study to knowledge on HPV in 
Ghana. This will include a description of the risk types of HPV, prevalences of HPV 
stratified by grade of lesion, age and country. 
2.8.1 Epidemiological Risk Types Classification of HPV 
As indicated in section 2.7.6, each HPV genotype' s (the over 40 genital HPVs) ability to 
cause cervical cancer (also referred to as oncogenic potential/risk) differs and this has formed 
the basis for a classification of HPV into risk type categories. High risk HPV types are those 
that are most prevalent, most associated with high grade and invasive cervical cancer and 
often occur as single infections in high grade lesion (Munoz et al., 2003). This molecular 
epidemiolog ic-based classification, categorises the following types as high risk HPV types 
(HR HPV); HPV 16, 18, 31 , 33, 35, 39, 45 , 51 , 52, 56,58, 59,68, 73 , and 82. The following ; 
6,11 , 40,42, 43, 44, 54, 55, 57, 61 , 70, 71 , 72, 81 , 84 and CP6108 have been categorized as 
low risk HPY types. The HPY genotypes 26, 30, 34, 53, 66, 67, 69, 85 and 97 are considered 
as probable high risk while HPY types 34, 57, and 83 were considered to be associated with 
undetermined risk (Munoz el aI. , 2003). Therefore, the prevalence or distribution ofHR HPV 
in a community will be indicative of a high risk for cervical pre-cancer development. For the 
48 
University of Ghana  http://ugspace.ug.edu.gh
study reported in this thesis these determinations were made. The following sections describe 
the different prevalences ofHPY infections that are important for a population. 
2.8.2 Prevalence of HPV Infection 
Among every general population, the overall HPY prevalence, HPY genotypes-specific 
prevalence, HPY risk type prevalences and age-specific HPY prevalences are important 
information for cervical cancer prevention. These have been shown to vary with the grade of 
cervical lesion (normal, low grade and high grade) and therefore stratified HPY infection by 
grade of lesion are importance information (Bruni et af., 2010; de Sanjose et aI. , 2007). Also 
varying are geographical distributions of HPY prevalences both within and between 
countries. Different HPY prevalences have also been reported among some peculiar 
populations, defined by sexual practices or immunocompromised state, for example 
commercial sex workers and HIY infected persons respectively (Ortiz et aI., 2006). 
Therefore, the proportion of such peculiar population among the study population is an 
important information to note. Due to these variations, it always important for every country 
to determine its county-level data to inform its planning of a cervical cancer prevention 
programme (WHO; Guideline Development Group, 2014); which was one of the major 
objectives of the study reported in this thesis. 
It must be noted that in addition to the geographical and biological factors that contribute to 
the observed variations in HPY prevalences, issues such as study design (case-control, cross 
sectional etc.), the methods of HPY detection and genotyping primers used may also 
contribute marginally and often none significantly to the observed variations (de Sanjose et 
aI. , 2007 ; Ortiz et af., 2006; Clifford et aI., 2005a). 
49 
University of Ghana  http://ugspace.ug.edu.gh
2.8.2.1 HPV prevalence variations with grade ofl esion and age 
The global distributions of HPV infection among women with normal cytology have 
indicated consistently that the African region has a high HPV prevalence. According to the 
report of a multi-centre study, Europe recorded the lowest average prevalence of 5.2%. South 
America and Asia regions recorded, 14.3% and 8.7 % respectively, while sub-Saharan Africa 
recorded the highest prevalence of 25.6%; it is worth noting that the pooled analysis 
included only one study from sub-Saharan Africa and as such that data may be biased by an 
under representation (Clifford et aI., 2005b). In a more current systematic analysis (de 
Sanjose et aI., 2007) that included more countries for each world region (78 studies in all), a 
similar trend was observed with much more detail. The highest overall HPV prevalences 
among women with normal cytology were observed in the following five world regions, three 
of which are in Africa; Eastern Africa (31.6%), Eastern Europe (29.1 %), Northern Africa 
(21.5%), Central America (20.4%) and Western Africa (17.0%). The five world regions with 
the lowest overall HPV prevalence did not include any region of Africa, these were Southeast 
Asia (6.2%), Southern Europe (6.8%), Japan/Taiwan region (7.0%), South-Central Asia 
(7.5%) and Northern Europe (7.9%). These imply that there was a high possibility of 
recording a HPV prevalence in the study reported in this thesis, since it was conducted in an 
African community, as has been observed. 
Among women with low grade cervical precursor lesions, the overall HPV prevalence, as 
reported in a meta-analysis of 55 studies (8,308 women) from different geographical regions 
(Clifford el aI., 2005b), also showed variations, but with higher HPV prevalences. North 
America recorded the highest prevalence of 80.1 %, followed by South/Central America, 
68.3%; Europe, 67.8%; Asia, 67.1 % and Africa 59.1 %. However, the prevalence of HPV 
infections in high grade lesions and invasive cervical cancer shows very little variation but 
50 
University of Ghana  http://ugspace.ug.edu.gh
were very high (70% - 99%). In a meta-analysis of studies from different world regions, an 
overall prevalence of between 84.2% and 87.6% was reported with less variation across 
geographical areas of the world (Clifford et aI., 2003). 
2.8.2.2 Age-specific HPV prevalence variations 
The age-specific distributions of overall HPV infections among women within the general 
population show an overall trend that is largely similar in most world regions. For example, 
in a meta-analysis of 78 multi-centre-study of women with normal cytology, it was observed 
that women younger than 25 years were most infected with HPV (with a peak between 20 -
22 years). This was followed by a progressive decrease in prevalence with age until a small 
increase was observed in women between 34 years and 54 years (Lenselink et aI., 2008; de 
Sanjose et aI., 2007; Burchell et aI., 2006; Schiffman and Castle, 2005). Despite, general 
trend, there were major variations among the different geographical areas and mainly with the 
age range for which a second increase in prevalence occurred; for Africa and Europe the age 
range of 45-54 years experienced a small second increase in prevalence, while for the North, 
Central and South America regions, the second increase in prevalence was among the 34-45 
years range. Data from the Asia region did not show a second increase in prevalence as 
observed for the other regions (de Sanjose el at., 2007) . 
Furthermore, age-specific HPY prevalence variations have been observed at the country 
level. For example, analysing data of women with normal cytology, it was observed that data 
from Thailand and Vietnam showed two peaks at 25-34 years and 55-64 years, data form 
Columbia, Argentina and Chile all showed one peak at 25-34 years, 35-44 years and 35-44 
years respectively (Clifford et aI. , 2005a). Data form Italy and Netherlands showed one peak 
at 55-64 years and 45-54 years while data on women in the general population in Spain 
51 
University of Ghana  http://ugspace.ug.edu.gh
showed two peaks at 20-24 years and 45-54 (Bosch et al., 2006). Data were only available 
from a Nigerian study for Africa and this showed two peaks, at 25-34 years and 45-54 years 
(Clifford et aI., 2005a), Another study among women in the general population in Canada, 
which detected only high risk HPV types, showed a single peak at 20-24 years (Ogilvie et aI. , 
2013). Therefore, it is important for every country to generate its own data to be able to 
understand the dynamic of HPV prevalence among its population in order for the effective 
planning of an HPV and cervical cancer prevention programme. The study presented in this 
thesis provides such data for a community in Ghana using two separate cervicovaginal 
specimens from each woman. 
2.8.2.3 Type specific HPV prevalence 
As indicated in section 2.8.1, high risk HPV are most commonly detected in cervical cancer 
tissues and as such the high risk genotypes HPV 16 and 18 are the most prevalent HPV types 
among women globally; although there are variations in the HPV genotype specific 
prevalence among women with normal cytology in the different world regions (de Sanjose el 
aI., 2007). Specifically, a pooled analysis reported by de Sanjose el al. , (2007) showed that 
among women with normal cytology, HPV 16 was the commonest genotype in all regions 
except in Eastern Africa and Japan/Taiwan regions where HPV 52 was the commonest. 
However, it must be noted that the data on HPV prevalence in Africa included in that pooled 
analysis was very limited and therefore the reported type specific prevalence was most likely 
not a true reflection of the situation in Africa. This limitation of data form Africa was one of 
the motivations for the conduct of the study reported in this thesis, which was aimed at 
providing this and other importance information on HPV prevalence. 
52 
University of Ghana  http://ugspace.ug.edu.gh
Furthermore, even wider variations were observed when the second commonest HPV types in 
each of the regions were considered; Eastern Africa and Japan/Taiwan region (HPV 16); 
Eastern Europe, Central America (HPV 31); Northern Africa, Northern Europe, Western 
Europe, Eastern Asia, Southeast Asia (HPV 18); and Western Africa and South America 
(HPV 58); North America (HPV 53), Southern Europe (HPV 66) and South-Central Asia 
(HPV 42). These variations confirmed the observation of an earlier review of studies among 
women with normal cytology by Clifford et al. , (2005b), which reported the following as the 
three most common HPV types in each of these regions; Sub-Saharan Africa (HPV 42, 16 
and 35), Asia (HPV 16, 42 and 33), South America (HPV 16, 58, and 18); Europe (HPV 16, 
3 I, and 18). 
These variations in HPV type specific prevalence are also observed among women with low 
grade lesions, but to lesser extent (Clifford et af., 2005a). The three most prevalent HPV 
types observed for each of the regions were as follows, North America (HPV 16, 51, 56), 
Europe (HPV 16, 31 , 51), Africa (HPV 16, 53, and 58), South/Central America (HPV 16, 33 
and 6) and Asia (HPV 16, 58, and 18). A further narrower variation in HPV prevalence was 
observed among women with high grade and invasive cervical cancer, HPV 16, 18, 33 , 31 
and 45 were the commonest with HPV 16 and 18 accounting for 66.9% of all HPV infection 
in all world regions (Clifford et al. , 2005b). 
It must be stated that these reported prevalences of the genital HPVs were both as single and 
multiple type infections; however, higher rates of single infections were often observed. 
These multiple type infections were most often observed among younger women and women 
with immunosuppressive conditions. They were often also observed among women with 
lower grade lesions than high-grade lesion and normal cytology (Trottier, 2006). Although 
different combinations of genotype multiple infections have been reported, some studies have 
53 
University of Ghana  http://ugspace.ug.edu.gh
generated data that seem to suggest some level of association between specific HPY types 
that are within the same species (Mejlhede et ai., 2010; Cones a-Zamora et aI., 2009; Awua et 
aI., 2007). Also, the limited data on whether HPY genotypes compete or facilitate another 
HPY genotype' s acquisition or persistence makes it difficult to understand clearly the 
influence of multiple infection on HPY prevalence and therefore how HPY prevalence might 
change in this era of type specific vaccination (Rositch et ai., 2011). 
2.8.2.4 Country specific HPV prevalence 
Another fact that encourages the determination of country-specific data on HPY infection 
was the well identified inter- and intra-country levels variation. HPY studies in Spain serves a 
good example of this intra-country variation. Specifically, studies in Madrid and Murcia have 
reported HPY prevalence of20.2% and 49.7% respectively among women attending cervical 
pathology clinics (Martin et ai., 2011; Conesa-Zamora et aI., 2009). These studies also 
reported HPY genotypes 16, 53, 31 , 52 and 6 as the commonest types in Madrid (Martin et 
al., 2011) and HPY 16, 45 , 5 J, 33 and 31 as the most detected in Murcia (Conesa-Zamora et 
al., 2009). Furthermore, these studies have discussed how similar and different their data 
were relative to studies in other countries in Europe such as France and Italy (Martin et ai., 
2011; Conesa-Zamora el aI., 2009) . 
The inter-country variations in HPY prevalence is exemplified by the few studies from the 
West Africa region; a community-based study in Ibadan, Nigeria determined that the 95% 
confidence interval of overall HPV prevalence was between 24.8% and 26.3% among women 
with normal cytology. HPY 42, 16, 35, 31 and 81 were the five most prevalent HPV 
genotypes reported by that study (Thomas et aI., 2004). However, a hospital based study in 
Benin, reported a 95% confidence interval for the overall prevalence as between 26.7% and 
54 
University of Ghana  http://ugspace.ug.edu.gh
33.2% among women with normal cytology; HPV 59, 35, 16, 18, and 58 were detected the 
most (Piras et ai. , 2011). Furthermore, a study in the Gambia determined an overall 
prevalence of 13.4% and HPV 16,35, 18,33 and 58 as the most prevalent HPVs (Wall et ai. , 
2005). In yet another study of women 35 years or older in Senegal, an overall prevalence of 
18.0% was reported and HPV 16,58, 18,33 and 54 were the commonest HPV types (Xi et 
al. , 2003). A hospital based study in Accra, Ghana that did not report type specific 
prevalence, reported an overall prevalence of 10.7% (Domfeh et at., 2008). In a case-control 
study in Abidjan, Cote d'lvoire, HPV prevalence among the control group (non HIV positive) 
were 31.1 %; the overall most common HPV types was confounded by the HIV positive status 
of the cases and therefore is not considered herein (Adjorlolo-lohnson et af., 2010). 
In respect of the prevalence of HPV infection among women diagnosed with advance 
cervical cancer both inter- and intra-country variations were very minimal and often 
negligible. For instance, Two studies of women diagnosed with cervical cancer in Ghana 
recorded HPV prevalences of 89.8% (95% CI 85.7-93.4) and 98.0% (95% CI 92.6-99.8) with 
the following as the most prevalent, HPV 18, 59, and 45 in both studies (Awua et at., 2016; 
Attoh et al. , 2010; Awua et al., 2007). Similar prevalences were reported by multicentre 
study in Ghana, Nigeria and South Africa (Denny el af. , 2014). 
2.8.3 Summary of Epidemiology of HPV 
The body of knowledge on the epidemiology of HPV indicates a lot of variability. Genotypes 
of HPV have shown differences in their risk potential for inducting cervical pre-cancer 
lesions and therefore cancer. The extensive variations in prevalence have been seen in 
respects of their distributions among the different grades of cervical pre-cancer lesions and 
cervical cancers, such that there is high variability among low grade lesions but lower 
55 
University of Ghana  http://ugspace.ug.edu.gh
variability among high grade lesions and cancer. This extensive variability is also seen in the 
age-specific, country level genotype specific prevalences. 
2.9 PREVENTION OF CERVICAL CANCER 
The primary aim of the study reported in this thesis was to provide information that was 
needed for the planning of an effective cervical cancer prevention programme in Ghana; as 
such, this brief description of the prevention of cervical cancer was necessary as a 
background to the study. Although there are primary, secondary and tertiary prevention of 
cervical cancer, this description is limited to only primary and secondary prevention, based 
on the context of this study. 
2.9.1 Primary Prevention 
Primary prevention refers to taking measures to ensure that a disease does not occur among 
individuals at risk by ensuing minimal or no exposure to the associated risk (Shepherd et aI., 
2000). For cervical cancer, this will include measures to prevent acquiring and maintaining 
persistent HPY infection. Until the mid-2000s, the major primary prevention of cervical 
cancer had been the use of educational intervention, aimed at promoting sexual-risk reduction 
behaviours. These included abstinence from sex before marriage (casual sex), delay of first 
sexual intercourse, reduction of the number of sexual partners and consistent use of condoms 
(Shepherd et al., 2000). However, following the association of HPY with cervical cancer and 
the subsequent development of the HPV vaccines, prophylactic vaccination against two, four 
or nine of the about 40 genital HPV have become the major promoted primary prevention 
strategy for prevention of cervical cancer. 
56 
University of Ghana  http://ugspace.ug.edu.gh
2.9.1.1 Preventive Sexual Behaviour 
As with most STI, the use of preventive education on abstinence does not seem effective and 
may not have any significant effect in the prevention of cervical cancer, both globally and in 
Ghana. For example, in spite of the numerous sexual educational campaigns in the Damgbe 
West district, in Ghana, the incidence of teenage sexual activity, which is a surrogate for 
abstinence, has been estimated to still be high; a study in that district (close to the location of 
the study reported in this thesis) reported that 55% of teenagers were sexually active by the 
age of between 13 years and 19 years (Afenyadu and Goparaju, 2003). Similarly, reports In 
other countries have indicated that educational campaigns encouraging young women to 
abstain from sexual activities till marriage orland reduce their number of sexual partners, did 
not seem to have had any noticeable effect on the incidence of HPV infection, cervical 
lesions and cervical cancer (Kahn, 2009). These sexual risk reducing behaviour interventions 
have been observed to be effective for up to 6 months , and showed different levels of 
behavioural changes. This low duration of changes in behaviour will most likely not be 
adequate to influence a reduction in the incidence of cervical cancer (Norton et aI. , 2012 ; 
Picot et af. , 2012; Shepherd et af. , 2000) . 
Consistent use of condom by male partners was reported to be associated with a reduced risk 
for HPV infection and cervical les ion. In a study by Winer et af. (2006), it was observed that 
compared to women whose male partners use condom at a frequency of 5% or less, women 
whose male partners use condom 100% of the time during penetrative sexual contact were 
70% less likely to be infected with HPV. Furthermore, women whose male partners used 
condom for more than 50% but less 100% of the time were 50% less likely to be infected 
with HPV. Also, although not significant, the use of male condom was associated with a less 
likelihood of developing cervical lesions by female partners. The fact that HPV transmission 
57 
University of Ghana  http://ugspace.ug.edu.gh
is possible by genital skin to skin contact and the male condom does not cover the peno-
scrotal region explained the less than 100% protection among women whose male partners 
used condom always (100%). The study reported in this thesis provided community 
education on sexual risk behaviours associated with cervical cancer development 
2.9.1.2 Prophylactic Vaccination 
This section describes the development and the current state of efforts at promoting HPV 
vaccination in the prevention of cervical cancer. Following the findings that persistent 
infection with high risk HPV is the causal factor of cervical cancer, one of the major 
developments in cervical cancer prevention has been the production of three prophylactic 
vaccines; these are Gardasil (which protects against infections of HPV16, HPVI8, HPV6 and 
HPV II), Cervarix (which protects against HPV 16 and HPV 18) and a nonavalent vaccine 
(which protect against HPV6, 11, 16, 18,31,33,35,45, and 58) (Chatterjee, 2014; Van de 
Velde et al., 20 12; PATH; Practical experience, 2010a; Dunne et af., 2008). 
The first WHO position paper on HPV vaccination encouraged a programmatic approach to a 
national vaccination scheme that selects the target population based on the population's data 
on the age of initiation of sexual activity. In addition, the cost effectiveness of the HPV 
vaccination strategies as well as sustainable funding should be considered before 
implementation. As such, the target group for cost-effective vaccination in most countries 
were reported to be adolescent females ranging between the ages of 911 0 years and 13 years 
(PA TH; Practical experience, 201 Ob; Kim and Goldie, 2008). On the other hand, vaccinations 
in boys have been considered by some studies as not cost-effective, if a good coverage were 
achieved for girls, and so vaccination for boys has mostly been uncommon (Kahn, 2009; 
Barnabas et af. , 2006). 
58 
University of Ghana  http://ugspace.ug.edu.gh
In order for developing countries (most affected by HPV infection and related diseases) to 
overcome the challenges of cost-effectiveness and sustainable funding, a number of 
initiatives and demonstration programmes have actively contributed to the reduction in the 
cost ofHPV vaccine for developing countries. Initiatives such as the PATH HPV Vaccination 
Pilots, Gardasil Access Programmes and Global Alliance for Vaccines and Immunisations 
(GA VI) initiatives have been undertaken in countries such as India, Peru, Uganda and 
Vietnam (Banura et aI., 2012; Gulland, 2012; Ramanathan and Varghese, 2010). African 
countries such as Tanzania have already benefited, while since February 2013, countries 
such Ghana, Kenya, Madagascar, Malawi, Niger, and Sierra Leone have been participating in 
GA VI supported demonstration programmes (GA VI Alliance Support, 2013; Quentin et aI., 
2012; Watson-lones et al., 2012). In spite of the fact that routine HPV vaccination in Ghana 
has been shown to be an appropriate option, and that an HPV vaccines pilot has started in 
Ghana, the need for secondary prevention (screening) remains relevant and must be 
supported with relevant empirical data on baseline HPV prevalence by studies such as that 
reported in this thesis (Edwin, 2010). 
2.9.2 Secondary Prevention: Cervical Screening 
The study reported in thi s thesis is primarily a secondary prevention activity that employed 
two of the three known cervical screening methods. This section provides a description of the 
development of the three available cervical screening methods (cytology, visual inspection 
and H PV testing), the advantages and Ii mitations, as well as the bases of the choice of which 
to use in this or any other study. 
As has been described thus far, once a woman is persistently infected with HPV or has 
developed a cervical pre-cancer lesion (whichever is detected first), there is a risk for cervical 
cancer development and thus the need to detect and stop the progression of the pre-cancer 
59 
University of Ghana  http://ugspace.ug.edu.gh
lesion. Despite the fact that about 90% of HPV infections are often cleared and lesions may 
regress in about I or 2 years (figure 2.4), the fact that the burden of cervical cancer is still 
high, calls for action to reduce the burden. Measures taken to achieve these (detection of 
HPV and cervical pre-cancer lesions by screening and the treatments of detected precursor 
lesions) constitute secondary prevention of cervical cancer. Consistent cervical screening and 
appropriate treatment of pre-cancers have so far been the most effective, at reducing cases 
and deaths attributable to cervical cancer (Sanghvi et al., 2008; Sankaranarayanan et aI., 
2005). 
Of the three well studied screening methods for the early detection of pre-cancer lesions, 
cytology (Papanicolaou or Pap smear) has been the most effective (Sanghvi et aI., 2008; 
Sankaranarayanan et aI., 2005). However, visual inspection of the cervix after the application 
of either acetic acid (VIA) or Lugol's Iodine (VILI) has been useful in the detection of pre-
cancer lesions in developing countries (Sanghvi et aI., 2008; Sankaranarayanan et aI., 2005). 
The third approach, which involves the detection of HPV DNA (HPV testing), has shown 
great promise, however, some few challenges remain to be addressed in order for it to be 
implemented as a primary testing method in routine clinical practice of most countries 
(Cuzick el aI., 2012). 
Screening activities remall1 relevant even with the development of vaccines against HPV 
infection, because while a small fraction of cervical cancer are not due to HPV infection 
(Clifford et aI., 2003), not all high risk HPV infection are protected against by HPV vaccines 
available (Chatterjee, 2014; Van de Velde et aI., 2012; PATH; Practical experience, 2010a; 
Dunne el aI., 2008). Additionally, for monitoring the progress and effectiveness of 
vaccination at protecting cervical cancer, there is the critical need to screen women for 
cervical pre-cancer lesion (Rositch et af., 20 II). The following sub-sections provide an 
60 
University of Ghana  http://ugspace.ug.edu.gh
overvIew of each of these screening methods although screening by cytology and HPV 
testing were the only used screening methods in the study reported in this thesis. 
2.9.2.1 Screening by Cytology 
The commonest cytological method for cervical screening (detecting precursor lesions), 
referred to as conventional cytology, involves performing a Pap smear test. Briefly, this 
involves the collection of exfoliated cells from entire transformation zone of the cervix and 
the cervical canal (figure 2.1) using a spatula and/or cervical brush (Moss et aI., 2004). The 
collected cells (cervical swab sample) are then "smeared" uniformly on to a microscope slide, 
followed immediately by the application of a fixative. The smear is subsequently stained with 
dyes (Papanicolaou stain) and examined for indications of cellular abnormalities under a 
microscope (WHO; Chronic Diseases and Health Promotion Group, 2006). 
The implementation of Pap smear, as an intervention/screening tool in well-organised 
national cervical cancer screening programmes, has had the most profound reduction in the 
incidence, prevalence and deaths related to cervical cancer; however, these have only been 
achieved in developed countries (Laara et af., 1987). For instance, between the years 1965 
and 1982, a reduction in cervical cancer deaths of between 34% and 81 % were observed 
among the Nordic countries that achieved nationwide coverage in their implementation of 
Pap smear based cervical cancer screening programmes (Laara et af., 1987). Additionally, the 
1960 to 1980 screening programme carried out in Canada resulted in the reduction of cervical 
cancer incidence from about 28.4 to 6.9 per 100,000 women per year and a reduction in 
related mortality, form 11.4 to 3.3 per 100,000 women per year (Denny, 2012). The 
effectiveness of conventional cytology at this reduction has been due to its ability to detected 
61 
University of Ghana  http://ugspace.ug.edu.gh
pre-cancer lesions (CIN2 and CIN3) with a sensitivity that ranges from 47% to 62% and a 
specificity that ranges between 60% and 95%, globally (Sankaranarayanan et aI., 2005). 
Another cytology-based screening tool known as liquid based cytology (LBC) was developed 
as an improvement of the conventional cytology method (Moss et aI., 2004). Its sensitivity 
averages as 57% and specificity of about 97%, LBC has shown relatively small 
improvements on the sensitivity and specificity of conventional cytology. Although there 
were some controversies regarding the initial studies of the LBC technology, it is being 
widely used in the USA and UK for routine screening for cervical lesions (Moss et al., 2004). 
However, like conventional cytology, LBC is currently difficult to implement in developing 
countries for screening programmes because it has similar requirements for implementation 
and so does not address the difficulties developing countries have with cytology based 
methods, such as, limited quality controls, limited instrumentations, high cost, limited trained 
personnel and inefficiencies in follow-ups (Denny, 2012). However, in studies (both in 
developed and developing countries), such as the one reported in this thesis, cytology testing 
is often used as gold standard for assessing other screening methods. 
2.9.2.2 Screel1il1g by HPV Testil1g 
The most recent addition to the options for cervical cancer screening has been the detection 
of HPY infection in exfoliated cervical cells collected for Pap smear testing. Following 
extensive evaluation in numerous studies, HPY test as a screening tool is acceptable in both 
developing and developed countries. These studies have consistently shown that HPY testing 
is more sensitive (between 84% and 90%) but less specific (between 70% and 86%) for the 
detection of pre-cancer lesion (CIN2 and CIN3 grade lesions) as compared to cytology based 
testing, which reported sensitivity of between 57% and 61 % and specificity of between 82% 
62 
University of Ghana  http://ugspace.ug.edu.gh
and 90% in the same group of studies (Patanwala et af. , 2013; Castle et af. , 2012; Cuzick et 
ai., 2006; Kulasingam et ai., 2002; Solomon et ai., 2001). Other important findings of these 
HPV studies were that HPV testing was more reproducible. 
Subsequent studies with improved study designs and testing technologies (more sensitive 
molecular techniques) have also compared HPV testing to cytology based testing. These 
provided stronger evidence for the higher sensitivity but slightly lower specificity of HPV 
testing (Ronco et al., 2010; Leinonen et aI., 2009; Mayrand et aI., 2007b; Ronco et aI., 2006). 
One of the largest comparative studies of cervical screening methods in a developing country 
was a randomised control trial conducted in India. In that study, it was shown that a single 
HPV testing was associated with a significant reduction in cervical cancer cases and related 
deaths relative to a control group; while cytology based testing on the other hand did not 
significantly reduce cervical cancer incidence or related deaths (Patanwala et ai., 2013; 
Castle et ai. , 2012; Sankaranarayanan et ai. , 2009). 
In order to overcome the observed low specificity of HPV testing for the detection of CIN2 or 
worse lesions, concurrently HPV and cytology based testing for cervical cancer screening, 
known as co-testing, was evaluated (Dillner et ai., 2008). The findings of those studies 
indicated that co-testing of women by HPV and Pap testing was a useful approach for 
cervical cancer screening (Dillner et af., 2008). One of the major studies that evaluated co-
testing for cervical cancer screening, the so-called Joint European Cohort Study, reported that 
being negative for both HPV and Pap testing predicted the lowest risk for the development of 
cervical cancer (Dillner et ai. , 2008). It was also observed that being HPV negative but Pap 
positive marginally increased the risk of the incidence of cervical cancer, which was at a 
constant rate over a period of six years. However, being HPV positive but Pap negative 
progressively increased the risk of cervical cancer incidence over the same period. Also, 
63 
University of Ghana  http://ugspace.ug.edu.gh
being positive for both HPV and Pap testing was observed as the most significant predictor of 
cervical cancer incidence (Dillner et ai. , 2008). 
In routine real-world clinical practice, co-testing had similar predictive values as compared to 
the experimental study conditions. In one of the major studies that showed this (Katki et ai., 
2011), the cumulative incidence ofCIN3 or worse lesion over a five year period was 0.17% 
and 0.36% for women who tested negative by only HPV or Pap testing (respectively). For the 
women who tested positive by only HPV or Pap testing, their cumulative incidences were 
7.6% and 4.7% respectively (Katki et af. , 2011). However, having a negative result for both 
HPV and Pap testing had a 0.16% cumulative incidence over the same period, indicating that 
performing both tests was more reassuring for women who had negative results, particularly 
compared to those with negative Pap results. The cumulative incidence for a woman with a 
negative HPV test but a positive Pap test only went up marginally to 0.86%, while having a 
positive HPV result and a negative Pap results increased the cumulative incidence to 5.9%. 
Furthermore, having positive results for both tests increased the cumulative incidence to 
12.1 % (Katki el af. , 2011). This clearly shows that having an HPV result makes a big 
different for women who are Pap positive, that is, a risk of 0.86% ifHPV negative or a risk of 
12.1 % if H PV positive. Additionally, some studies have shown that genotyping the detected 
HPV in order to identify high risk HPV (HR-HPV) type infections, particularly those of 
HPV -16, 18, 31 and 33 improves the positive predictive value of HPV testing for the 
detection of CIN 3 or worse lesions (Soderlund-Strand et af. , 2011; Castle et af., 2005 ; Khan 
et af. , 2005). However, it is worth noting that current algorithms that include genotyping 
triages are targeted at women 30 years or older, because younger women often have transient 
HPV infection. 
64 
University of Ghana  http://ugspace.ug.edu.gh
A number of co-testing implementation strategies have been evaluated globally, which has be 
summarised thus: an initial HPY testing followed by cytology triage of HPY positive women 
or an initial cytology testing followed by an HPY test triage of women with CIN2 or worse 
grade lesions (Patanwala et aI., 2013). The ATHENA HPY study, a randomized controlled 
trial , which evaluated nine strategies, including a triage with HPY genotyping, concluded that 
a screening strategy that will have an HPY testing triaged by a combination of HPY 
genotyping and cytology does well on the three major screening outcome measures; 
sensitivity, specificity, and reduction of potential harm (Cox et aI., 2013; Patanwala et aI. , 
2013 ; Castle et aI. , 2012; Cuzick el aI. , 2012; Castle et aI. , 2011 a). The FOCAL trial , another 
randomi zed controlled trial of HPY testing for cervical cancer screening using CIN3 or worse 
lesion as the end point, when completed and finally reported, is expected to provide data 
regarding the efficacy of two co-testing strategies; an initial hr-HPY DNA testing 
(genotyping) followed by liquid based cytology (LBC) triage of hr-HPY-positive women and 
an initial liquid based cytology (LBC) testing followed by hr-HPY triage (Ogilvie et aI. , 
2010). 
Although testing technologies for HPY detection has rapidly improved, from the 
conventional PCR through the signal amplification technologies (as Hybrid Capture I and II) 
to the combination of multiplex, nested and real time PCRs, only the sensitivity of HPY 
testing has been improved, without an improvement of its relatively lower specificity. This is 
due to the fact that whiles the HPY test detects current HPY infections, which also includes 
those that are transient; it is rather persistent HPY infections that account for the incidence of 
cervical cancer. Therefore, newer molecular technologies for the detection of other viral 
biomarkers such as HPY RNA, E6 and E7 HPY proteins and related expressions, and non-
viral biomarkers (such as pl6INK4a and Ki67 antigens, methylation of CADMJ and MAL 
65 
University of Ghana  http://ugspace.ug.edu.gh
genes6) are rapidly being improved in order to be able to differentiate between persistent and 
transient HPV infections, with the possibility of an improvement of the lower specificity of 
HPV testing (Patanwala el aI., 2013; Cuzick et ai., 2012). The introduction ofHPV testing in 
developing countries may face some of the problems associated with a national cytology-
based screening programme such as the high cost of setting up infrastructure for call and 
recall , which remains very poor, and inefficiencies in follow-up. 
The approach adopted in the study reported in this thesis was that the women who consented 
to participate were concurrently (co-tested) tested by Pap smear and samples form the spatula 
and brush were recovered for HPV DNA testing and genotyping. 
2.9.2.3 Screening by Visual Inspection of the Cervix 
A third major screening method for cervical cancer prevention involves the direct inspection 
of the TZ, SCJ and the external os with the unaided eye and a bright light after the application 
of either Acetic acid (VIA) or Lugol's Iodine (VILI). For VIA, the inspection is done at least 
1 minute after a 3%-5% solution of acetic acid has been applied to the cervix and the 
appearance of a distinct acetowhite area in the TZ or near the SJC is looked for. In the case of 
VILI, the inspection is performed following the application of Lugol's Iodine and the 
appearance of a distinct yellow area in the TZ close to the SJC is indicative of a positive test. 
These are very simple, affordable and easy to implement approaches in screening 
programmes, particularly for developing countries. The nature of these testing methods 
makes them amenable to what has come to be known as the 'single visit' or 'screen-and-treat' 
approach in cervical cancer prevention. With that approach, women with a positive test, 
without further triage, are directly treated by cryotherapy. The evaluations of these testing 
methods in many developing countries have shown that they reduce loss to follow-up, delays 
66 
University of Ghana  http://ugspace.ug.edu.gh
in treatment and missed cases (Cuzick et af., 2012; Sankaranarayanan et aI., 2005). Of the 
two testing methods, VIA has been the most evaluated, mostly in developing countries 
including Ghana, Thailand, India, South Africa and Zimbabwe. The sensitivity and 
specificity of these testing methods have varied greatly in most of the cross-sectional studies 
that evaluated them; they have both ranged between 14% and 95% with an average sensitivity 
of 50% and average specificity of 80% for the detection of CIN2 or CIN3 lesions (Cuzick et 
af., 2012; Sanghvi et af. , 2008; Gaffikin et af., 2007). 
The major limitations of these testing methods are that the appearance of either the 
acetowhite or yellow area (for VIA and VILI respectively) may not be due to cervical 
neoplasia, since inflammation, regeneration of the cervical epithelium and immature 
squamous metaplasia give the observed colour changes, resulting in treatment of women who 
do not need any treatment (over-treatment). The results of these tests may be difficult to 
determine for postmenopausal women. Additionally, these test are predominantly unable to 
detect glandular precursor lesions, since it is difficult to visualize the endocervical canal. 
Issues, such as, the lack of a uniform criteria for reporting results, subjectivity in the 
interpretation of results, the need for frequent re-training, all result in the widely varying 
sensitivity and specificity of these tests (Cuzick et al., 2012). Balancing the positives and 
limitations of this screening method, the VIAIVILI although may be easier to implement in 
resource limited countries, it is most likely going to lead to some appreciable level of over 
treatment and over referrals for colposcopy with accompanying increase in cost. 
2.9.2.4 Self-Sampling and Participation in Screening 
Although not a screening method, self-sample collection is a strategy developed such that 
women are able to collect their own samples which may be useful for both cytology and HPV 
67 
University of Ghana  http://ugspace.ug.edu.gh
DNA testing. This strategy was developed to support the most sensitive and most specific 
screening methods (HPY DNA and Pap testing respectively), in order to overcome the burden 
of logistics and training needs associated with the implementation in developing countries of 
Pap and HPY testing. Additionally, self-sample collection strategy was developed to address 
the challenges that limit women's participation in screening activities, these include lack of 
time, discomfort, inconvenience, cultural objections, unwillingness and inability to travel to 
health facilities for Pap and HPY tests (Arrossi et aI., 2015; Yirtanen et aI., 2011; Anhang et 
aI., 2005). Although most developing countries are evaluating and introducing screening 
based on VIA and YILI, as discussed earlier, these are not as sensitive and specific as Pap 
and DNA testing (Cuzick et aI., 2012; Petignat and Vassilakos, 2012) and therefore there is 
the need to study the usefulness and acceptability of self-sample collection. 
In order that most women get tested and are able to benefit from the advantage of HPV 
testing, self-collected cervicovaginal cell samples for HPY testing have been evaluated 
relative to both HPY testing of health-personnel collected cervical samples and Pap testing. 
Findings of cross-sectional studies indicated that there was a high rate of agreement between 
health-personnel collection (HPC) and self-collection (SC) with regards to the HPV 
positivity. In one sllch study, an overall agreement of 88.1 % with a kappa of 0.73 was 
reported ; while HPV testing of SC did not detect HPY infections in 14 women who were 
HPV positive by the corresponding HPC, HPV testing of HPC samples also missed the HPV 
infection in 18 women detected as positive with sc. The statistically none significant 
discordance and other differences observed between the two types of samples was due to the 
fact that more low risk types of HPV were detected in SC than HPC specimen (Gravitt et aI., 
2001). In more recent studies, overall agreements of between 91.8 and 93.6% were reported 
between the two sample types with kappa of between 0.67 and 0.73. The levels of 
68 
University of Ghana  http://ugspace.ug.edu.gh
discordance were also similar. In one such study, HPV types were detected in IS women 
using their HPC samples that were not detected by their SC samples. On the other hand, those 
detected in 17 women using their SC specimen were not detected in their HPC specimen. The 
rates of type-specific concordance in women with multiple infections were also fairly high 
(Petignat and Vassilakos, 2012; Zhao et af., 2012; Bhatia et af. , 2009). 
In respect of the sensitivity and specificity for the detection of CIN2 or worse lesions, 
generally SC was 10%-19% less sensitivity but of similar specificity relative to HPC. A study 
had reported even very similar sensitivity and specificity for both sample types; 93 .6% and 
82.S%, respectively for SC and 93.2%, and 87.S%, respectively for HPC. In the same study, 
the sensitivity and specificity for cytology were 77.S%, and 87.3%, respectively, indicating 
how similar SC is to HPC relative to cytology (Bhatia et ai., 2009). In a community based 
randomized controlled trial in Mexico, the so called MARCH trial , a relative sensitivity for 
the detection of CIN2 or worse lesions and invasive cancer were 3.4 and 4.2 times more 
respectively, for SC as compared to cytology although it recorded a lower positive predictive 
value and a lower specificity because it detected more low risk HPV types (Lazcano-Ponce et 
af. , 2011). In all , these reports show that self-collection is a viable alternative to HPC and 
complementary to cytology testing (Scarinci ef af. , 2013; Belinson et al., 2003). The fact that 
self-sampling does not need pelvic (speculum) examination, attendance at clinics or the 
presence of health personnel , and that it can be performed at the convince of a woman at 
home and the collected specimen may sent through the regular mail, makes it potentially a 
useful tool for increasing the coverage and participation of women in cervical cancer 
screening. It has been shown that the response to self-collection can be varying for different 
populations. For example, women who were non-attendees of regular cervical screening in 
Italy, Finland, Sweden and Netherlands responded to self-collection at the following rates, 
69 
University of Ghana  http://ugspace.ug.edu.gh
19.6%,31%,39% and 34% respectively, which was more than the use of letter to remind 
non-attendees (about 10.0%). However, for all these studies, an average of 5% increase in 
overall screening coverage due to the introduction of self-collection were reported (Virtanen 
et af., 20 11; Rossi et af. , 2011; Gok et af., 20 10; Sanner et af., 2009). 
The overall influence of self-collection on cervical cancer screenmg has led to its 
recommendation as an alternative method of sampling in national screening programmes and 
most women in most studies including those referred to above, were of the view that self-
sampling was easy to perform and that they did not experience difficulties with it. The study 
reported in this thesis applied two separate approaches without the provision of incentives; a 
hospital based approach and a community centred approach for self-collection. Detailed 
description of these approaches are provided in the method chapter of this report. The second 
approach was designed with the aim of overcoming the limitations of not meeting women at 
home and the difficulties women have when they needed to travel to the hospital for kits and 
for the return of their samples. 
2.9.2. Summary of Prevention of Cervical Cancer 
Primary prevention of cervical cancer, which was mainly by education on abstinence form 
causal sex and the promotion of the use of condom, was on a low key, however, the 
development of three vaccines against nine different genotypes has brought the primary 
prevention of cervical cancer to the fore. Research continues towards the development of new 
vaccines (' second-generation' vaccines) against the HPV types that are not protected against 
by the current available vaccines (Chatterjee, 2014). In spite of these improvements in 
primary prevention, secondary prevention which has mostly been by screening for cervical 
lesions (cytology tests) and lately screening for HPV infection (HPV DNA test) has been the 
major tool for reducing cervical cancer incidence and mortality. ]n the era of vaccination, 
70 
University of Ghana  http://ugspace.ug.edu.gh
screening remains relevant as it serves as a means of assessing the success of vaccination 
programmes in preventing the development of cervical cancer precursor lesions among other 
important issues .. 
71 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER THREE 
METHODS 
3.1 STUDY LOCATIONS 
On the basis of the following facts , that HPV is the commonest sexually transmitted infection 
(ST!), that other STJs share similar sexual behaviour risk factor profiles (Burchell et aI. , 
2006) and that Chlamydia trachomatis (CT) , herpes simplex virus type 2 (HSV-2) and HIV 
encourage the development of cervical lesions (Munoz et aI. , 2006; Sarnoff, 2005; Smith et 
af. , 2002a, 2002b, 2004) and HPV infections and persistence (Burchell et aI. , 2006; Sarnoff, 
2005), STls may serve as good surrogate indicators of the risk of HPV infections in a 
community. Therefore, a high risk community in this study was defined as any district that 
recorded one of the highest STI prevalences in Ghana in the year 2010. The Lower Manya 
Krobo District of the Eastern Region was that district with the highest STI prevalence and 
therefore was selected for this study. Subsequently, the Akuse sub-district of the Lower 
Manya Krobo District (Figure 3.1) was randomly selected, among the 5 sub-districts by 
random number selection using the Ms. Excel software, for the conduct of this study. 
The Akuse sub-district is made-up of 18 communities, 12 of which are located closely 
(centred at the coordinates 6" 06'02.43N and 0" 1' 16.87E), while the other 6 are between 3 
km and 5 km west of the 12 communities. The sub-district has a Government District 
Hospital with some of the communities having a Community based Health Planning and 
Services (CHPS) Centre. As at the time of the design of this study (December 2010), the 
available population data for the Akuse sub-district was the Ghana 2000 census's estimate of 
population size for the year 2011 (the 2011 estimates were found not to be substantially 
72 
University of Ghana  http://ugspace.ug.edu.gh
different from the 20 I 0 census data when it became available). The population of the Akuse 
sub-district was estimated to be 8887 (for 20 II), of which 29.2% (2595 women) were 
femal es between the ages of 15 and 65 years (Ghana Statistical Service, 2000). There were 
four common Ghanaian languages spoken within the 18 communities of the sub-district; 
these were Hausa, Ewe, Ga-Adangbe and Twi. The study questionnaire was pre-tested in the 
Atua community of the Odumase sub-district. 
Ghana 
Eastern Region 
POPULATION 
D LESS POPUlATE 
_ AVEARGE PU'OlATES 
_ OENSLY POPUlATED 
Figure 3. 1 Population distribution in the Lower Manya Krobo District. 
The sub-district se lected for the study is indicated by the small arrow. 
73 
University of Ghana  http://ugspace.ug.edu.gh
3.2 SOURCE AND STUDY POPULATION 
Women living within the Lower Manya Krobo District between March 2012 and March 2013 
were targeted in this study, however, those who were living in the communities of the Akuse 
sub-district within this period were recruited to participate in this study. 
3.3 SAMPLE SIZE DETERMINATION 
A minimum sample size was calculated based on the random sample size equations and a 
modification (S) since the population was less than 10000. 
2 
N Z p(l-p) 5= 
2 N 
d 1 + Population 
N= sample size. 
Z= the standard normal deviation (set at 1.96 corresponding to the 95% Confidence Interval) . 
p = the prop0l1ion In the target population estimated to have an HPY infection (HPY 
prevalence) . 
d = degree of accuracy desired or the maximum difference to detect 
(Expected prevalence - Worst acceptable prevalence). 
Since there were no data for the prevalence of HPY infection in the general population of any 
community in Ghana, the HPY prevalence (P) reported in a Nigerian community (26.3%) and 
74 
University of Ghana  http://ugspace.ug.edu.gh
neighboring West African countries with the highest and lowest reported HPV prevalences 
were used. The degree of accuracy (d) was set as 0.065, which is the closest to the average 
difference between the HPV prevalence in West African countries (Jemal et at., 2012; Piras 
et at., 2011 ; Vaccarella et aI. , 2008; Thomas et at. , 2004). Since the study was a cross-
sectional study, a design factor of 1.2 was applied to obtain a sample size of 211 participants. 
A response rate of 60% was expected, based on literature with respect to response to cervical 
screening and the sample size was adjusted by this rate to obtain the target of 352 women for 
the study. 
3.4 STUDY DESIGN (CROSS-SECTIONAL STUDY) 
3.4.1 Study Type, Selection of Communities and Distribution of Sample Size 
A cross-sectional study design was used for this study. Of the 5 sub-districts of the Lower 
Manya Krobo District (figure 3.1), one (the Akuse sub-district) was randomly selected by the 
use of the random number function in Ms. Excel software. Specifically, the names of the five 
sub-districts were listed , not in any order, and numbered 1 to 5. The Excel formula 
"=RANDBETWEEN(I,5)" was used to randomly select one of the numbers from I to 5, and 
that which corresponded to the number assigned the Akuse sub-district on the list (3) was 
returned . All its communities included in the study. Based on the population of women 
between the ages of 15 years and 64 years in each community, the number of women targeted 
for the study (352 participants) was distributed among the communities by the probability 
proportional to size (PPS) method. That is, the population of women (older than 15 years) for 
each of the community was determined and the sum calculated. The ratio of the population 
for each community to the sum of the populations was determined and the resulting fraction 
75 
University of Ghana  http://ugspace.ug.edu.gh
was used to multiple the number of women targeted for the study in order to obtain the 
number of women targeted for each community. ) 
3.4.2 Community Entry and Engagement 
For this study location, it was found that the Akuse Government Hospital was the most 
influential institution through which the community could be successfully engaged. 
Therefore, with the assistance of the hospital administration, a meeting involving the 
religious leaders of the orthodox churches, researchers and the hospital administration took 
place. At this meeting, a community education programme was planned. Presentations on 
cervical cancer and the study were made in two churches during the Christian Home Week 
Celebration, (during this celebration, members of all the orthodox churches in the sub-district 
meet in one of the member churches for specific activities). Also, announcements were made 
regularly at all the orthodox churches regarding the study and its presence in the 
communities. A second presentation was made mid-way through the field work in a different 
church during a women's group celebration. The leaders of the Muslim community were 
separately contacted and during a Friday Muslim prayer service the community members 
were informed of the study and issues relating to cervical cancer were discussed . 
Each of the communities in the sub-district was entered by contacting the traditional leaders 
of the respective community with the assistance of the Public Health Nurses. The community 
members were reached by the traditional system of announcement and the activities of the 
Community Health Volunteers within those communities. The participants were subsequently 
recruited by house visits. The Public Health Nurses and the Community Health Volunteers 
engaged the communities and participants continuously during their regular community 
outreach programmes. They explained the study and provided information about cervical 
cancer as well as taking feedbacks from the study team to the members of the community and 
76 
University of Ghana  http://ugspace.ug.edu.gh
the participants. This study completely avoided the involvement of the political leaders 
and/or members of the local government system because of the existence of some political 
disputes and legal issues within the sub-district to ensure freedom and willingness of 
participation. 
3.4.3 Sampling and Reporting Strategies 
The study was designed to employ a single reporting strategy for the participants, described 
below as reporting strategy I (section 3.4.3.1) and depicted as figure 3.2a. However, due to 
an initial low reporting rates, which resulted from women not able to make time off their 
daily and economic activities to go to the hospital during working hours for specimen 
collected, a discussions with mostly the women who were not reporting (by phone calls), the 
Nurses of the Public Health Unit and the Community Health Workers of the Akuse Hospital, 
resulted in the gathering of important information and suggestions. Based on these 
information and suggestions, a new reporting strategy was developed that improved the 
reporting rate. This is described below as reporting strategy 2 and depicted in figure 3.2b. 
Another recruitment/reporting strategy that evolved during this study was what was referred 
to as peer-recruitment (described below as additional observation). Women recruited by this 
means were allowed to participate, so that the study may be able to understand how what 
women within a community say among themselves regarding a screening programme, may 
influence the overall participation of women in cervical cancer screening. 
3.4.3.1 Reporting strategy 1 
Women were identified by a house survey of 17 of the 18 communities in the Akuse sub-
district. In each of those communities, the major roads/paths through it were identified by the 
use of Google Map computer application and grouped in order of direction, from one end of 
77 
University of Ghana  http://ugspace.ug.edu.gh
the community to the other. Based on the crude estimate of the average number of houses and 
the population of women in each of the communities, it was assessed that visiting every 5th 
house or less will ensure a wide coverage of the communities, therefore a number was 
selected from among the numbers 1 to 5, which was 3 to represent the random count of 
houses to visit. Therefore, starting from the first selected road/path, every third house to the 
right and left of the road/path was visited. The women in each house meeting the inclusion 
criteria were invited to participate in the study after explaining the study, the consent form 
and the options available to them as participants as well as giving them basic information 
about cervical cancer. Each consenting woman was assisted by a researcher to complete the 
study questionnaire. The women were given the option to set a date (a period within a week 
or more than one week) they were going to come to the hospital for a dual specimen 
collection, which were self-specimen collection performed by the women and Trained Health 
Personnel collected specimen. This was repeated for each road until the desired sample size 
was attained for each community. Women, who had not reported to the hospital for more than 
two weeks after the date they opted for, were called on the mobile phone number(s) they 
gave, until they were reached once, they were reminded and a new date arranged for them to 
visit the hospital for sample collection. These are illustrated in figure 3.2 below. 
3.4.3.1.1 Peer-recruitment 
It was observed that some women who reported for sample collection at the hospital were not 
recruited by the researchers, but were encouraged by the women who lived with them (in the 
same house), who met the researchers during the house survey. For each of the houses 
surveyed that had less than three women recruited by the researchers, the peer invited 
/women were allowed into the study and treated as was done for the other women recruited 
by the researchers. That is , the study was explained to them, those who were willing 
78 
University of Ghana  http://ugspace.ug.edu.gh
provided informed consent and the study questionnaire was administered as described above. 
Specimens were subsequently collected in the same way as described for the other women. 
3.4.3.2 Reporting Strategy 2 
This field-developed approach (figure 3.2b) involved identifying, a location within a 
community where a room was available for the women to be attended to, in privacy. This 
location was mostly central, well known and convenient to quickly reach from all other parts 
of the community. Women recruited by the house survey were invited to participate by going 
to the selected location at any time between 9:00 am and 7:00 pm. Women who reported and 
met the inclusion criteria were invited to participate in the study after administering the 
consent process as described above. Consenting women were assisted by the researcher to 
complete the study questionnaire. 
Since this approach evolved on the field, only 4 of the 17 communities that had not yet been 
involved in the study at all, were reached with this approach and as such the study was not 
sufficiently powered to determine the differences in the reporting rates between this and the 
original approach. By the nature of this second approach (particularly the fact that there was 
more time to report within a day and that the completion of questionnaire was at the 
collection point but not at home during recruitment), the level of motivation for the women in 
these 4 communities were different from those who were to report at the hospital. Therefore, 
differences in performance between this and the other approach are mainly suggestive of their 
relative potential influence of cervical cancer screening activities in Ghana and other 
developing countries with similar settings. 
79 
University of Ghana  http://ugspace.ug.edu.gh
a) b) Reporting Strategy 2 
Reporting Strategy 1 
Recruit by House Survey Recruit by House Survey 
~ I 
I Complete Questionnaire I 
... 
Choose reporting time: Report to a location within the Do not report to 
Short (within 1 week) or Long (within 3 month) Community location, Out of study 
Do no•t re port to 11 Report• to H ospital Complete Questionnaire Hospital 1 
+ ~ 
Remind by Provide HPC and/or SC 
I--
phone call specimen Provide HPC and SC specimen 
+ ~ 
Do not report, nd Complete a 2 nd 
out of study questionnaire Complete a 2 questionnaire 
Figure 3. 2: An illustration of the design for reporting for specimen collection. 
80 
University of Ghana  http://ugspace.ug.edu.gh
3.4.4 Inclusion and Exclusion Criteria 
Self-reported healthy women between the ages of 15 and 70 years who were willing to provide 
specimen by at least one of the two specimen collection methods were included in this study. 
Women who had undergone hysterectomy or cervical conisation were excluded from the study. 
Women who reported they had never had sexual intercourse were also excluded from the study. 
Self-reported pregnant women were also excluded from the study; this was because of the 
controversial nature of cervical screening during pregnancy and the occurrence of maternal 
anxiety and the potential of association of participation in screening with any future occurrence 
related to the pregnancy (Flannelly, 20 10). Additionally, there is a high potential of over- and 
underestimation of cytological changes or possible lesion which may have been due to the 
normal physiological changes associated with pregnancy (Flannelly, 2010; Hunter et at., 2008; 
Van Calsteren et ai., 2005). 
3.5 ETHICAL ISSUES 
The study was approved by the Ethics Review Board of the Ghana Health Service (10 No GHS-
ERC: 06/11/1 0). For the participants who were able to read and understand English, they read 
and signed an informed consent form (in English) that explained the main research goals, sample 
collection procedures, potential benefits and harms, and privacy and confidentiality of data 
collected for the study. For those who could not read English, the informed consent form was 
81 
University of Ghana  http://ugspace.ug.edu.gh
translated to a local language and explained to them; each of the four languages spoken in the 
communities (Hausa, Ewe, Ga-Adangbe and Twi) was translated by one particular researcher 
throughout the study. Each woman was given a card indicating her community, date of 
recruitment, a short title of the study and the place to report for sample collection. Questionnaires 
were administered by a one-on-one interview, in such a manner as to ensure that no other person 
heard the participant and the researcher. The description of the process for self-collection of 
cervical swabs (self-specimen collection) was also done one-on-one and with the aid of 
pictures/illustrations. The self-specimen collection was done by the participants alone in a 
separate room and with the illustrations of how to perform the process in hand. Subsequently, a 
vaginal examination and health-personnel sample collection was performed with only the health 
personnel (Medical officer or Nurse) and the participant behind a screen in a room privately. 
Since no woman was found to have intraepithelial lesions or Pap positive results, there were no 
referrals for subsequent management. However, women HPY positive were advised to re-test 
after 3 years. 
3.6 QUESTIONNAffiE ADMINISTRATION 
The study questionnaire was finalized after the pre-test, in the Atua Community, Odumase sub-
district. Two trained Public Health Nurses assisted the study participants, by a one-on-one 
interview, to complete a study questionnaire (appendix J). This was used to collect information 
on the socio-demographic characteristics, sexual behaviour, sexual and reproductive history, 
menstrual factors , use of oral contraception, and history of sexually transmitted infections and 
cervical cancer screening history. 
82 
University of Ghana  http://ugspace.ug.edu.gh
3.7 SAMPLE COLLECTION AND SMEAR PREPARATION 
Each participant was given the option to choose to perform both SC and HPC or anyone of the 
two as they wished. However, for the most of them who choose to perform both, they performed 
the self-collection followed by the Health-personnel collection. This order, as has been reported 
by most studies which made both methods available, was due to the advantage of identifying 
women who may have been within the period of menses and should be excluded from the Pap 
test (Lazcano-Ponce et aI. , 2011; Adler et aI. , 2012; Bosgraaf et aI., 2014; Gravitt et aI., 2011; 
Bhatia el al., 2009; Bais et aI. , 2007». To collect the sample(s), each participant, one at a time, 
was instructed on how to use the Rover® Viba-Brush vaginal sampler, (Rovers Medical Devices, 
The Netherlands), using the illustrations on the packaging of the brush. The women were 
directed to a private room and reminded to wash their hands before using the kit. The collected 
specimen, on the detachable part of the brush, was transferred into a sterile 15 mL screw-capped 
tube containing 5 mL PBS with 5 mM ethylene diamine tetraacetic acid (EDTA), pH 8.0. These 
specimens were designated as self-collected (SC) for HPV testing. Women having their menses 
were asked to come back one week after the end of their menstrual period. 
The nurse practitioner then performed a standardized pelvic examination. With the aid of a 
sterile plastic disposable speculum, the same trained Nurse, used the PAP-PAK@ Cytology kit 
(Medical Packaging Corporation, Camarillo, USA) to collect cervical specimen, according to the 
manufacturer' s instructions, for Papanicolaou (Pap) smear and HPV testing. Briefly, the head of 
a wooden Ayre 's spatula was used to collect cellular material from the cervix and the 
squamocolumnar junction, and then a cytobrush was used to collect cellular material form the 
endocervix. These were smeared uniformly on separate sections of a microscopic slide (for Pap 
83 
University of Ghana  http://ugspace.ug.edu.gh
testing) followed by the application of a fixative (95% ethanol and 5% polyethylene glycol) 
provided with the kit. The cytobrush was then rinsed in 50 ilL of DNAgard®Tissue (Biomatrica, 
Inc., San Diego) in a 2 mL screw-capped micro-centrifuge tube. Also, the head of the wooden 
spatula was broken and added to the same 2 mL screw-capped tube to complete the sample used 
for HPV testing. These specimens were designated as Health Personnel Samples (HPS). All 
specimens for HPV testing were transported at 4 °C and store at -20°C until used for HPV 
testing. 
3.8 PAPANICOLAOU STAINING OF SMEARS 
The slides were completely immersed in two changes of95% ethanol to remove the carbowax in 
the fixative, overnight. The slides were thereafter placed in racks with a continuous flow of 
running water to hydrate for 5 minutes and then stained with Mayer's Hematoxylin for 5 
minutes. The slides were washed in running water for 5 minutes, after which they were 
dehydrated by immersion in 95% ethanol for 2 minutes. The slides were passed through Orange 
G-6 for 2 minutes, followed by two rinses of 95% ethanol, each time for 10 seconds. The slides 
were then transferred into a solution of EA for 3 minutes and rinsed very well with two changes 
of 95% ethanol, each time for 10 seconds. Dehydration was completed in absolute ethanol for I 
minute. The slides were air dried on the bench, cleared in xylene and coverslip fixed with DPX 
mountant (GCC Diagnostics, Gainland Chemical Co., Sandycroft Flints, UK). The slides were 
then examined under an Olympus BX51 TF microscope with an Olympus DP20 camera 
(Olympus Corporation, Tokyo, Japan), for cellular characteristics. 
84 
University of Ghana  http://ugspace.ug.edu.gh
3.9 DNA EXTRACTION 
The volume of each specimen collected by HPS was made-up to I mL with PBS and vortex 
thoroughly. For each of both specimen (SS and HPS), 200 ilL of the cell suspension were used 
for DNA extraction on the MagNA Pure LC automated system (Roche Molecular Systems, Inc., 
Pleasanton, USA) using the MagNA Pure DNA Extraction Kit (Roche Molecular Systems, Inc., 
Pleasanton, USA) according to the manufacturer's specification with a modification of an 
addition of 15 ilL of RNase to the lysis buffer for each specimen. The negative controls included 
in each set of extraction were PBS, DNase/RNase free water, and DNAGard, while the positive 
control was a suspension of SiHa cells with an integrated HPV 16 genome. The specimen/DNA 
quality was assessed by the amplification of the human RNase H gene in 10 ilL of DNA extract 
and analysed with Light Cycler 480 and related software (Roche Molecular Systems, Inc, 
Pleasanton, USA). 
3.10 HPV GENOTYPING 
3.10.1 Nested-Multiplex PCR 
For each specimen, 15 ilL of its extracted DNA were used in amplifying the HPV Ll gene with 
the consensus PGMY09/PGMYll primer sets (sequences presented in Table A in the Appendix) 
in a first round PCR. In addition to the 15 ilL of extracted DNA, the 50 ilL final reaction volume 
contained 200 11M of each of the four dNTPs, IX PCR Buffer (200 mM Tris-HCI pH 8.4; 500 
mM KCI), 6 mM of MgCb, 0.6 11M of each of the 18 PGMY primers and 1.25 U of Taq DNA 
polymerase. The PCR conditions, for the first round of nested PCR were, an initial denaturation 
step at 95°C for 9 min, followed by 40 cycles of, denaturation at 95 °C for 30 seconds, 
85 
University of Ghana  http://ugspace.ug.edu.gh
annealing at 55 °C for 30 seconds and elongation at n °C for 30 seconds. This was followed by 
a single final extension at n °C for 5 min before holding the temperature at 4°C. 
The second round of the nested PCR was with 2 ilL of the first round PCR products as DNA 
template and the GP5+/GP6+ primer set, (5'-TTIGTTACTGTGGTAGATACTAC-3 ' ) and (3 ' -
CTfATACTAAATGTCAAATAAAAAG-5') respectively. The GP6+ primer (Invitrogen, 
Burlington, Canada) carried a 5' biotin label and phosphorothioate bonds in the first 4 
nucleotides on the 5' end. The 50 ilL final reaction volume contained, 200 11M of each dNTP, IX 
PCR Buffer (200 mM Tris-HCI pH 8.4; 500 mM KCI), 4 mM of MgCb, 1.0 11M of GP5+ 
primer, 1.01lM of GP6+ primer, 1.25U of Hotsat Taq DNA polymerase. The amplification was 
achieved with the following conditions, an initial denaturation step at 94°C for 5 minutes was 
followed by 30 cycles of denaturation at 94°C for 30 seconds, primer annealing at 40 °C for 20 
seconds and an elongation at n °C for 30 seconds. This was followed by a 7 minute final 
extension at n oc and holding the temperature at 4°C. 
3.10.2 Detection ofPCR Products I Genotypes 
The second round PCR products were used for the detection of 46 genotypes of HPV (HPV 16, 
35 , 40, 45 , 58, 18, 66, 59, 6, 31 , 74, 42, 52, 51 , 62, 81 , 83, n , 90, 67, 53, 56, 33, 70, 87, 39, 30, 
54, 44, 86, 82, 84, 68,69, 85, 91 , 73, 43 , 11 , 13, 32, 61 , 71,89, 97 and 26) by the map 
technology, as described by Zubach et al., (2011). This involved an initial coupling of 
microspheres labelled (according to the manufacturer' s instructions) with different ratios of red 
and infrared tluorophores (Luminex Corporation, Austin, TX) to HPV type-specific probes 
carrying a 5' -CI2 amino modification that reacted with the carboxyl groups on the microspheres. 
The double stranded second round PCR products were linearized in 96-well PCR microplates 
86 
LtJ SCHOOL OF :~JBLlC 
_~~ \ H: AL.lI1 L.! C.., hARY 
CK5 ~ LEGON 
L ,,--,~ 
University of Ghana  http://ugspace.ug.edu.gh
(Fisher, Ottawa, ON, Canada) with 2 flL of bacteriophage T7 gene6 exonuclease (New England 
BioLabs, Pickering, ON, Canada) that removed the non-labelled strand after 40 minutes 
Incubation at room temperature. The PCR microplates were kept at 95°C for more than I 0 
minutes while adding 33 flL of the microsphere mix (15 microsphereslflL for each set) to each of 
the wells of each PCR microplate. These were incubated for hybridization at 60°C for 10 
minutes. After this, 25 flL of a 0.04-flglflL solution of streptavidin-phycoerythrin (PE) 
(Invitrogen) in I-tetramethyl ammonium chloride (TMAC) (Sigma), were added and incubated 
for 5 minutes at 60 °C. Genotype specific hybridization was detected on a Luminex Liquid Chip 
200 flow cytometer (Qiagen) using the Luminex IS software (Luminex). The analysis was 
carried out at 60°C with a maximum volume of 50 flL of the sample and a minimum count of 
100 microspheres per type, with settings of 8,300 and 16,500 for the lower and upper gates, 
respectively, as recommended by the manufacturer. 
3.11 HPV 16,18 AND 45 VARIANT ANALYSIS 
The specimens positive for HPVI6, 18 and 45 infections were selected for variant analyses since 
these are the known most oncogenic HPV genotypes and are the genotype whose variant analysis 
is important for understanding the contribution of variants on carcinogeneisis of an HPV type. 
The long control region (LCR) of the HPV16, 18 and 45 genomes were separately amplified by a 
nested PCR and second round products sequenced. 
3.11.1 HPV16 nested peR 
Using 5 flL of DNA extracted from each of the specimens, the first round PCRs were performed 
with the following primer sets; HPV16RIF (5'CACCTACTAATTGTGTTGTGG3'), 
HPV 16R I R (5 'GTTTGCACACACCCATG3 '). The 50 flL final reaction volume also contained, 
87 
University of Ghana  http://ugspace.ug.edu.gh
400 11M of each dNTP, IX PCR buffer (200 mM Tris-HCl pH 8.4; 500 mM KCl), 2.5 mM of 
MgCb, 1.25 liM of each of the two primers, and 5U of AmpliTaq DNA polymerase. The 
amplification was achieved with the following conditions; an initial denaturation step at 95°C 
for 5 minutes was followed by 60 cycles of denaturation at 95 °C for 30 seconds, primer 
annealing at 65°C for 30 seconds and an elongation at 72 °C for 30 seconds followed by a 7 
minute final extension at 72 °C and holding the temperature at 4 0c. Extracted HPV16 DNA 
positive control and PCR water negative controls were used in each set of amplification. 
The second round PCR used 5 ilL of the first round product and the following primer sets 
HPVI6R2F (5 'GGGGTACCTCGGTTGCA TGCTTTTTGGC3') and HPVI6R2R 
(S'GGTCTAGACGGTTTGCACACACCCATGT3') for the amplification. The 100 ilL final 
reaction volume contained 200 11M of each dNTP, 1 X PCR buffer (200 mM Tris-HCI pH 8.4; 
500 mM KCI), 2.S mM of MgCb, I.2S 11M of each of the two primers, and 6.3 U of Taq DNA 
polymerase. The amplifications were achieved with the same cycling conditions as that of the 
first round PCR. 
3.11.2 HPV18 nested peR 
The nested PCR for HPY 18 positive specimen used the same concentrations and cycling 
conditions as those for HPY16 except that the following primers sets HPV18RIF 
(5'GTGTIJGTGGTATGGGTGTT3') and HPYI8RIR (S'GTATAGTATGTGCTGCCCAA3') 
were used for the first round and HPVI8R2F (S'CGGTTGCATAAACTATGTAT3') and 
HPY18R2R (S 'TCGGTTGCCnJGGCTTATG3') were used for the second round . The controls 
used were a plasmid containing an HPY 18 DNA positive control and PCR water negative 
controls. 
88 
University of Ghana  http://ugspace.ug.edu.gh
3.11.3 HPV45 nested PCR 
Using 5 JlL of DNA extracted from a specimen each time, three separate first round PCRs were 
performed with the following sets of primers, for fragment HPV4 5F 
(5'GACCTCGTAAGCGTCCTGC3') and HPV45R 5'GGATGCTGTGTAGTATG 
CAAGTTTATA3'); for Fragment 2 HPV45F2 (5'GTATGGTGTTACTGTACATA 
ATTGTGG3') and HPV45R2 (5'CATAGGGTGTGGATACAGTTGTG3') and for Fragment 3 
HPV45F3 5'CTGGCACATTTACAACCCCTAC3') HPV45R3 (5'CGCCAT 
CCTGCAAT GCAC 3 ') . The 50 JlL final reaction volume also contained, 400 JlM of each dNTP, 
IX PCR buffer (200 mM Tris-HCI pH 8.4; 500 mM KCI), 3.0 mM ofMgCh, 1.25 JlM of each of 
the forward and reverse primers, and 5U of AmpliTaq DNA polymerase. The amplification was 
achieved with the following conditions an initial denaturation step at 95°C for 10 minutes was 
followed by 40 cycles of denaturation at 95°C for 30 seconds, primer annealing at 55 °C for 30 
seconds and an elongation at 72 °C for 30 seconds. This was followed by a 7 minute final 
extension at 72 °C and holding the temperature at 4 0c. Extracted HPV45 DNA positive control 
and PCR water negative controls were used in each set of amplifications. 
For the first round amplifications that did not yield enough PCR products for sequencing, a 
second round PCR was performed that used 2 JlL of the first round product and the same primer 
sets as for the first round amplification. The 50 JlL final reaction volume contained 400 JlM of 
each dNTP, I X PCR buffer (200 mM Tris-HCI pH 8.4; 500 mM KCI), 2.5 mM of MgCh, 1.25 
JlM of each of the two primers, and 6.3 U of Taq DNA polymerase. The amplifications were 
achieved with the same cycling conditions as that of the first round PCR. 
89 
University of Ghana  http://ugspace.ug.edu.gh
3.11.4 Visualization of second round PCR product 
For all the second round PCR products of HPV 16 18 and 45, 20 JlL were resolved on a 1.5% 
agarose gel in I X TBE buffer with I Jlg/mL ethidium bromide at 120 V for an hour. The gels 
were documented with GelDoc™ XR+ System (Bio-Rad Laboratory, Mississauga, Canada). 
3.11.5 Sequencing of second round PCR product 
The second round PCR products were purified with the Amicon Ultra centrifuge filers (Millipore 
Ireland Ltd, Tullagreen, Ireland) with slight modification of the manufacturer's instruction. 
Specifically, a filter column was placed upright, into a micro-centrifuge tube for each sample, 
after which 200 ilL of PCR water and 80 ilL of a PCR product were added . The assembly was 
spun at 14000xg for 12 minutes. Additional 100 ilL of PCR water were added to the filter 
columns. The filter columns were thereafter transferred, inverted, to a new set of micro-
centrifuge tubes and spun for 2 minutes at 1000xg. The columns were discarded and the eluted 
DNA in each tube was quantified at 230 nm using NanoDrop ND- IOO spectrophotometer 
(Thermo Fisher Scientific Inc, Wilmington, USA). The eluted DNA was diluted to 50 nglllL and 
sequenced with ABI Applied Biosystems 3730xl DNA Analyzer BigDye® Terminator v 3.0 
(Applied Biosystem, Canada) . 
DNA Sequences obtained with the forward and reverse primers were assembled with SeqMan 
Pro software (DNAStar) . The sequences obtained for HPV16, HPV18 and HPV45 PCR products 
were aligned to HPVI6, HPVI8 and HPV45 reference sequences (listed in Table SI) by the 
Cluster W method, using the MEGA 5.2 software. Phylogenetic trees were constructed based on 
the Test Neighbour Joining using the MEGA 5.2 software 
90 
University of Ghana  http://ugspace.ug.edu.gh
3.12. STATISTICAL ANALYSIS 
3.12.1 Data Entry and Quality Control 
Double data entry was performed in SPSS version 20.0 (IBM Corporation, USA), after which 
data validation and cleaning were performed by merging the two data entries and checking for 
duplicates, illegal entries, non-entries, inconsistent coding and extreme/unreasonable values. 
SPSS version 20.0 and OpenEpi software were used for further data analyses. 
3.12.2 Main Analyses 
Initially, the data on the demographic characteristics, sexual and reproductive risk characteristics 
were described with frequencies, percentages and descriptive statistics. Additionally, the 
distributions of these were presented showing sub-categories for each. 
Furthermore, in order to understand the influence of these characteristics (demographic, sexual 
and reproductive) on the reporting for specimen collection, a Chi-square analysis was used to 
determine the association between these participant characteristics and reporting for specimen 
collection, with special interest for the association with the inconsistent use of condom. It must 
be noted that the analyses of the relationship with reporting status excluded data on demographic 
characteristics and rep0l1ing status of the 38 peer-recruited women. This helped identify which 
category of women should be encouraged the more to participate in cervical screening. In other 
words, to determine which characteristics were associated with non-reporting. The relationship 
with condom use was assessed to inform the identification and targeting of education on condom 
use to the appropriate category of women; depending on which demographic characteristic 
and/or risk characteristics were more associated with inconsistent condom used. It must be noted 
91 
University of Ghana  http://ugspace.ug.edu.gh
that the analysis excluded data on the women who did not report and the 38 peer-recruited 
women. 
Following the observations of the distribution of the participants' risk characteristics, which are 
indicators of potential burden of HPV and cervical lesion, and in order to determine and confirm 
the burden of HPV, the overall, genotype specific, age-specific, high risk, low risk, multiple and 
single HPV infection prevalences for each specimen collection method, self-sampling (SC) and 
health personnel sampling (HPC), were described as proportions of the total number of 
participants tested for HPV within each category, and reported with a 95% confidence intervals. 
Having used two methods for specimen collection for HPY burden determination, a comparison 
of the methods was necessary to inform future choice of method for screening. Therefore, the 
levels of agreement between the two specimen collection methods, in respect of each of the HPY 
infection prevalences, were analysed for using the Cohen's Kappa analysis and a determination 
of the proportions of concordance of HPV detection per woman, using only data of women who 
provided specimen by both SC and HPC methods was made. A McNemar's test was used test for 
the significance of the marginal proportion (concordance and discordance). Additionally, the 
age-specific prevalences for each of the specimen collection methods were obtained as a 
proportion of women, within each age-group that tested positive for HPV and these were 
compared and presented as paired line plots. Further to the comparison between the methods, an 
assessment of the strength of the association between sexual risk characteristics with HPY 
positivity for each specimen collection was performed by the use of logistic regression, to 
determine crude odds ratios (OR). 
92 
University of Ghana  http://ugspace.ug.edu.gh
On the other hand, some analyses were not possible to perform, although intended, due to non-
detection and limited power of the study. These include, all analyses related to the presence of 
cervical lesions (prevalence of and associations of cervical lesions with HPY positivity and 
known risk factors): these were not performed since no cervical lesion was detected in this study. 
Furthermore, this study was not powered to determine the magnitude of the association between 
the known sexual risk factors and HPY positivity (the sample size was smaller than needed for 
determining such associations), therefore this was not performed. 
3.12.3 Additional Analyses 
]n addition to the analyses directly related to the specific objectives of this study, as indicated in 
section 3.12.2 above, analyses of additional collected data were performed and reported as 
additional findings. 
• Chi-square analyses were performed to determine which demographic characteristics and 
known risk factors (excluding data of women who did not report for specimen collection 
and those recruited by peers), were associated with the stated post-performance 
preference for each of the two specimen collection methods used in this study. This will 
help identify what category of women who may prefer each of the specimen collection 
methods. 
3.13 DEFINITION OF VARIABLES 
3.13.1 Measures / Variables of Interest 
The main outcome variables assessed in this study were as follows: 
93 
University of Ghana  http://ugspace.ug.edu.gh
I . The detection of an HPV infection with each of the two specimen collection methods. 
This included the prevalence of HPV categorized as follows; overall, genotype specific, 
age-specific, high risk type, low risk type, single and multiple HPV infections, 
2. The extent of agreement between the two specimen collection methods in respect of HPV 
positivity. Specifically, these included the Kappa values and the proportion of 
concordances of the overall, type specific, age-specific, high risk type, low risk type, 
single and multiple HPV infections. 
3. The variant types of the detected HPV 16, 18 and 45 genotypes. 
4 . The prevalence of cervical lesions (cytological abnormalities). 
5. The distribution or proportions of the participants within the high risk and low sub-
categories of each of the known sexual and reproductive risk factors. The high and low 
risk sub-categories of each of the risk factors are described under section 3.13.2 below. 
Additional findings were also reported for this study, these were, the reporting rates for the 
different reporting strategies that evolved during this study, the levels of both post and pre-
performance preferences of the two specimen collection methods and the associations of 
demographic characteristic and risk factors with condom lise. 
3.13.2 Independent Variables 
For the analysis of the associations of sexual risk characteristics with condom use and with l-IPV 
positivity, the following were defined as the independent variables: 
• Age at first sexual intercourse, as a risk factor for the acquisition of an l-IPV infection, 
was defined as the calendar age at which a woman first experienced penetrative vaginal 
94 
University of Ghana  http://ugspace.ug.edu.gh
sexual intercourse. The AFSI sub-category, less than 20 years, is known to be high risk 
for HPV infection while the AFS] sub-category, greater 20 years, is known to be low risk 
for HPV infection (Erickson et aI., 2013; Bahmanyar et al., 2012; Berrington de 
Gonzalez et aI., 2004). 
• Sexual age, as a risk factor for the acquisition of an HPV infection, was defined as the 
difference between the current calendar age of a woman and her calendar age at first 
sexual intercourse. The sexual age sub-category, <II years is known to be high risk for 
HPV positivity while the category II to 19 years is known to be moderate risk and the 
category > 20 years is known to be protective of HPV infection (Lenselink et al., 2008). 
• Lifetime number of male sexual partners (LNSP), as a risk factor for the acquisition and 
persistence of HPV infection, was defined as the number of males a woman had had 
penetrative vaginal sexual intercourse with in her lifetime. The LNSP sub-category, 2: 3 is 
known to be high risk for HPV positivity and progression of cervical lesion, while the 
sub-categories of less than 3 are known to be protective of HPV infection (Erickson el 
al., 201 3; International Collaboration of Epidemiological Studies of Cervical Cancer, 
2007; Shields el at. , 2004). 
• Consistent use of condoms during sexual intercourse, as a risk factor for the acquisition 
of an HPV infection, was defined as the use of condom by either the male or female 
partner during penetrative vaginal sexual intercourse. The inconsistent use or not of 
condoms are known to be high risk for HPV positivity, while the consistent use of 
condom is known to be protective ofHPV infection. 
• Number of pregnancies, as a risk factor for the acquisition of an HPV infection and the 
progression of cervical lesions, were defined to include all the pregnancies the woman 
95 
University of Ghana  http://ugspace.ug.edu.gh
had carried, both full term or otherwise. The sub-category of > 4 pregnancies are known 
to be high risk for HPV positivity and progression of cervical lesions, while the sub-
categories of < 5 pregnancies are known to be of low risk. 
The general confounding factors including age, educational status and marital status were 
included in the analysis of the associations with condom use. 
• Educational status, was defined as the highest attained educational level as at the time of 
the study and each sub-category included different levels of both the older and new 
educational systems in Ghana. 
• No-formal education referred to not having a formal education. 
• The Primary Category included the primary levels of both the older and new 
educational system. 
• Junior Secondary Category included the Middle school level and the Junior 
Secondary level of the older systems, as well as the Junior High level of the current 
educational system. 
• Senior secondary category included secondary level and senior secondary levels of 
the older systems, as well as the senior high levels of the current educational 
system. 
• Post-secondary category included the Teacher Training College, Nursing Training 
College, Polytechnic and the University levels). 
96 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER FOUR 
RESULTS 
4.1 PARTICIPANT DEMOGRAPHIC CHARACTERISTICS 
Data of the 38 peer-recruited women were included in the analysis of the demographic 
characteristics based on the fact that the same criteria of selection (they were from the same 
houses selected during the study and not more than three women in all, including the peer-
recruited, from each of the selected houses were allowed to participate) and consenting processes 
were used. They were therefore expected to be similar to those recruited by the researchers. 
More importantly, the distributions of the demographic (Table 4.1) and risk characteristics of the 
38 peer-recruited women were similar to those of the women who were recruited by the 
researchers; except for the distribution of religion (because almost all Muslims were within one 
community). Therefore, the potential bias due to the inclusion of the distribution of the 38 peer-
recruited women was not evident. The table 4.1 shows the distribution of the demographic 
characteristics stratified by the reporting strategies in the study. Similar tables for the distribution 
of the risk characteristics stratified by reporting strategies were not reported. Therefore, the 
demographic characteristics of the 415 women (377 study recruited and 38 peer-recruited) who 
consented, and completed the study questionnaire are described in this section. 
97 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.1: Distribution of the demographic characteristics of the 415 recruited 
participants. 
Reporting Strategy, n (%) 
Community Peer Long Short All Women 
Sub-categories 
based Recruited duration duration 
15 - 19 6 (5.8) 2 (5.6) 12(7.7) 6 (5.1) 26 (6.3) 
20 - 24 18(17.5) 2 (5.6) 26 (16.8) 19 (16.1) 65 (15 .8) 
25 - 29 28 (27.2) 8 (22.2) 36 (23.2) 25 (21.2) 97 (23.5) 
30 - 34 6 (5 .8) 3 (8.3) 20 (12.9) 17(14.4) 46 (11.2) 
35 - 39 12 (11.7) 7(19.4) 15 (9.7) 18(15.3) 52 (12.6) 
Grouped Age 40 - 44 8 (7.8) 8 (22.2) 18(11.6) 17(14.4) 51 (12.4) 
45 - 49 6 (5.8) 2 (5.6) 15 (9.7) 7 (5 .9) 30 (7.3) 
50 - 54 8 (7.8) I (28) 10(6.5) 3 (2.5) 22 (5.3) 
55 - 59 4 (3.9) 3 (8.3) 2 (1.3) 4 (3 .4) 13 (3 .2) 
60 Or Older 7 (6.8)' 0(0.0) I (0.6) 2 (1.7) 10 (2.4) 
Total 103 <IOO .O) 36 (100.0) 155 (100.0) 118(100.0) 412 ( 100.0) 
Christian 103 (100.0) 36(100.0) 129 (82 .7)' 105 (897) 373 (90.5) 
Muslim 0(0.0) o (0.0) 25 (16.0)' 10(8.5) 35 (8 .5) 
Religion 
Other 0(0.0) o (0.0) 2 (1.3) 2 (1. 7) 4 (1.0) 
Total 103 (100.0) 36 (100.0) 156 (100.0) 117 (100.0) 412 (100.0) 
Unmarried 38 (38.0) 10 (28.6) 49 (31.8) 53 (44.9) 150 (36.9) 
Marital Status Married 62 (62.0) 25 (71.4) 105 (68.2) 65(55 .1) 257 (63.1) 
Total 100 (100.0) 35 (100.0) 154 (100.0) 118(100.0) 407 (100.0) 
No Formal 21 (20.8) 6 (16.7) 30 (19.5) II (9.3) 68 (16.6) 
Primary 26(25 .7) 5 (13.9) 30 (19.5) 16 (13.6) 77 (18.8) 
Educational Junior Secondary 39 (38.6) 14,(38.9) 69 (44.8)' 72 (61.0)' 194 (47.4) 
Status Senior Secondary 9 (8.9) 6 (16.7) 18 (11. 7) 15 (12.7) 48(11.7) 
Post-Secondary 6 (5.9) 5(13 .9) 7 (4.5) 4 (3.4) 22 (5.4) 
Total 101 (100.0) 36 (100.0) 154 (100.0) 118 (100.0) 409 (100.0) 
Unemployed 12 (12.0) o (0.0) 7 (4 .7) 6 (5.2) 25 (6.2) 
Formally employed 18 (18.0) 9 (25.0) 23 (15.3) 16(13.8) 66 (16.4) 
Skilled Worker 23 (23.0) 8 (22.2) 25 (16.7) 37(31.9) 93 (231) 
occupation 
Trader 34 (34.0) 16 (44.4) 75 (50.0) 47 (40.5) i 72 (42 .8) 
Agro-worker 13 (13 .0) 3 (8.3) 20 (13.3) 10 (8.6) 46 (11.4) 
Total 100 (100.0) 36 (1000) 150(1000) 116 (100.0) 402 (100.0) 
.. denotes for each sub-category of demographic characteristics, column proportions of a reporting strategy that differ 
significantly the other at the 95% confidence level. * The difference in religious characteristics that showed Muslim in only two 
of the four groups in the column was primary due to the fact that there was a major Muslim community that was approached at 
the early stage of the study. 
98 
University of Ghana  http://ugspace.ug.edu.gh
The ages of the 415 participants ranged between 15 years and 70 years, and were not normally 
distributed; the test of normality (both Kolmoqorov-Smirnov and Shapiro-Wilk) showed a 
significant difference (p = 0.000l) and a skewness of 0.68 ± 0.12 and Kurtosis of -0.16 ± 0.24. 
The median age of the participants was 31.0 years with an interquartile range (IQR) of 16.0 
years. The ages were subsequently grouped in a 5 year interval and the distribution presented in 
Table 4.1. The age group with the highest proportion of the participants was 25-29 years (n = 97; 
23.5%), which was followed by the age group 20-24 years (n = 65; 15.8%). On the other hand 
the age groups with the lowest proportion of the women in this study were the age groups 60 
years and older (n = 10; 2.4%) and 15-19 years (n = 26; 6.3%). 
The distribution of the religious affiliations, were as follows, 90.5% Christians (n = 373),8.5 % 
Muslims (n = 35) and 1.0% (n = 4) other religions. The Table 4.1 also shows that more than half 
(63.1 %; n = 257) of the participants were married, while 36.9% were unmarried (n = 150). 
Among the women who were married, 79.4% (197 of 248) were in monogamous marriages 
while 14.9% (n = 37) were in polygamous marriages. The distribution of the educational status 
of the participants showed that 47.4% of the participants reported junior secondary as their 
highest attained educational level , as at the time of the study. The next most common educational 
status reported was the primary level, which was by 18.8% of the participants. The others were 
as follows, no formal education, 16.6%; senior secondary 11.7%; and post-secondary 5.4%. 
Additionally, 6.2% of the women were unemployed, and the occupations of those who were 
employed were categorised into four groups. Traders (42.8%) and skilled workers (23.1 %) 
formed the majority of the participants. The others were in formal employment (16.4%) or in 
Agro-business (11.4%). 
99 
University of Ghana  http://ugspace.ug.edu.gh
4.2 PARTICIPANTS REPORTING RATES/STATUS 
4.2.1 Reporting strategy 1 
Of the 473 eligible women invited to participate in this study, 415 consented and completed the 
study questionnaire. Of these, 377 participants were invited by the researchers while 38 were 
peer-invited. Since the number of women invited by their peers, of which 38 reported, could not 
have been known, and that they did not fit in any of the reporting strategies, these women were 
excluded from the analysis of the reporting rates as it will bias the estimated (over-estimate) 
overall reporting rate if included (figure 4.1). 
Of the 377 participants invited by the researchers, 274 were invited to participate using reporting 
strategy 1 (which was the original strategy designed for the study). Among these, 118 of them 
opted to report within a week, while 156 opted to report at any time after a week but within 3 
months after questionnaire administration, for specimen collection at the hospital (figure 4.1). 
These formed two sub-groups of reporting strategy 1, hereinafter referred to as short and long 
duration for reporting respectively. Among the 156 participants who opted for the long duration 
of reporting, 60 of them (38 .5%) reported for specimen collection at the hospital within the 
stated period . Among the remaining 96 participants who did not report after the date they opted 
to, 27 of them were reached by a phone call , reminded of their participation and were ali owed to 
set a new date. Of these, 12 participants (44.4% of the 27 participants or 12.5% of the 96 
participants who did not report initially) reported for specimen collection at the hospital. On the 
other hand , among the 118 participants who opted for the short duration of reporting to the 
hospital , 55 (46.6%) reported within the duration they opted for. 
100 
University of Ghana  http://ugspace.ug.edu.gh
415 of the 473 invited women consented and completed study questionnaire 
I 
Approach One Approach Two 
I 
11 8 women opted for sample 156 women opted for sample 103 women scheduled for 38 women who were 
collectIon at hospital within a collection at hospital within 3 community based specimen 
invited by other 
week (short duration) months (long duration) collection 
participants (peer 
recruited) consented, 
completed the STUdy 
63 did not 96 did not 
questionnaire and 
report report 
provided specimen 
29 reminded by 27 reminded by 
phone call phone call 
•
55(46.6%) 3 (10.3
 %) 12 (4•4.4 %) 60 (38.5%) 98(95.1%) 
reported reported reported reported reported 
Overall, 228 of the 377 (60.4%) of the study recruited women reported 
Figure 4. 1: Participant enrolment and response to specimen collection. 
101 
University of Ghana  http://ugspace.ug.edu.gh
Among the remaining 63 participants who did not report, 29 were reached by a phone call, 
reminded of their participation and allowed to fix a new date. Of these 29 participants, only 2 
(10.3% of those who were reached on their phone or 4.8% of those who did not report initially) 
reported to the hospital for specimen collection. 
4.2.2 Reporting Strategy 2 
In respect of the strategy that was developed on the field (reporting strategy 2), which involved 
specimen collection at a location within the communities, 103 women were invited to participate 
in this study by it and 98 (95.1 %) of these women reported for specimen collection at the 
location within their respective communities (Figure 4.1). 
4.2.3 Overall Reporting Rate 
Overall, 228 of the 377 participants recruited by the study (excluding the 38 peer-recruited 
participants) reported for specimen collection; in other words, a reporting rate of 60.4% was 
achieved in this study. A comparison of the response rates for the three reporting strategies (long 
duration to report to the hospital , short duration to report to the hospital and report at a location 
within community), showed a statistically significant difference (p = 0.0001) and an LSD 
multiple comparison (post hoc analysis) indicated that all the pairwise comparisons of response 
rates were significantly different. The order therefore, from highest to lowest was, reporting at a 
location within a community (95.1%) > short duration to report (46.6%) > long duration to report 
(38.5%). 
102 
University of Ghana  http://ugspace.ug.edu.gh
4.3 SEXUAL AND REPRODUCTIVE CHARACTERISTICS (KNOWN RISK 
FACTORS) 
The sexual and reproductive characteristics of the 415 women (377 study recruited and 38 peer-
recruited) who consented, and completed the study questionnaire were described together in this 
section. The inclusion of the data collected from the 38 peer-recruited participants was based on 
the reasons as described above for demographic characteristics. That is, the analysis of the risk 
characteristics stratified by the reporting strategies showed no significant differences. 
4.3.1 Sexual Risk Characteristic (Risk Factors) 
The ages at first sexual intercourse (AFSI) reported by the participants ranged between 9 years 
and 36 years and were not normally distributed; the test of normality (both Kolmoqorov-Smimov 
and Shapiro-Wilk) showed a significant difference (p = 0.0001), a skewness of 0.75 ± 0.12 and 
Kurtosis of 4.75 ± 0.24. The median age at first sexual intercourse of the participants was 18 
years with an interquartile range of 3.0 years. The ages were grouped into two sub-categories and 
the distribution depicted in Table 4.2. This table shows that slightly more than half of the 
participants (52.4%, n = 208) had their sexual debut before they were 19 years old. The 
remaining 47.6% started having sex when they were at the ages of 19 and older. A small number 
of the women (n = 18) did not state when they had their first sexual intercourse. 
Regarding the number of male sexual partners the participants had had since they became 
sexually active (Table 4.2), a minimum of I, a maximum of 10 with a median of 2 (interquartile 
range of I) male sexual partners were recorded. Those who reported three or more male sexual 
103 
University of Ghana  http://ugspace.ug.edu.gh
partners formed about 47.6% (n = 195) of the participants, while those who reported 2 and 
male sexual partners were 27.6% and 22.2% respectively. 
Table 4.2: Distribution of the categories of the participants' sexual risk factors 
Sexual risk factors Categories Frequency Percentage 
-S18 208 52.4 
Age at first sexual intercourse (years) 2: 19 189 47.6 
Total 397 
<11 175 44.2 
11-19 97 24.5 
Sexual age (number of years since first 
20 - 29 82 20.7 
sexual intercourse) 
2: 30 42 10.6 
Total 396 
91 22.9 
Lifetime number of male sexual partners 2 111 28.0 
(LNSP) 2:3 195 49.1 
Total 397 
0 61 14.9 
347 84.6 
Current number of male sexual partners (CNSP) 2 I 0.2 
3 1 0.2 
Total 410 
No 240 58.7 
Use of condom during sexual intercourse Yes * 169 41.3 
Total 409 
No 189 46.4 
Had an STI in the past 5-10 years Yes 218 53.6 
Total 407 
Candida 213 97.7 
HIV 1 0.5 
Other STls 
Gonorrhoea 2 1.0 
Total 216 
Itafi=ed sub-categories are the high risk categories/or HPV positivity. * all were inconsistent users 
104 
University of Ghana  http://ugspace.ug.edu.gh
With respect to the current (at the time of the study) number of male sexual partners the 
participants had, most of them (84.6%) reported having one, while 14.9% reported they did not 
have a male sexual partner at the time of the study. Only one participant each reported 2 and 3 
male sexual partners, while 5 participants did not report the number of male sexual partners they 
had. The table also shows slightly more than half, 58.7% (n = 240 of 409) of the participants 
stated that they and their male sexual partners had never used condoms during sexual 
intercourse, while 41.3% stated they and their male sexual partners had been using condoms, but 
inconsistently. 
A little over half, 53.6% (n = 218) of the 407 participants who had indicated STI history, had 
contracted an STI in the past 5 to 10 years prior to this study, while 46.4% had not contracted 
any STI during that period and 8 women did not indicate their history of STI. Of these 218 
participants who had contracted an ST!, 97.7% (n = 213) reported they have been diagnosed of 
candidiasis, while only a single participant reported infection with the HIV, and 2 others reported 
a diagnosis of Gonorrhoea within the stated period . 
4.3.2 Reproductive Risk Characteristics (Risk Factors) 
The use of oral contraceptives (OC) was not common among the participants of this study (Table 
4.3); only 23.3% (n = 94) of them stated they had or have been using OC within the previous 10 
years at the time of the study. Most of them 76.7% (n = 310) reported they had not used any oral 
contraceptives within the sated period, while 11 women did not indicate their status of. 
105 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.3: Distributions of the reproductive risk factors among the participants. 
Reproductive risk factors Categories Frequency Percentage 
No 310 76.7 
Use ofOC in the past 10 years Yes 94 23.3 
Total 404 100 
< I year 47 52.2 
I - 4 years 36 40.0 
Duration of OC use 
~ 5 years 7 7.7 
Total 90 100 
No 194 48.4 
History of menstrual difficulties Yes 207 51.6 
Total 401 100 
No 56 13.6 
Ever been pregnant Yes 355 86.4 
Total 411 100 
57 16.1 
Number of pregnancies (among ever 2-4 162 45 .6 
pregnant) ~5 136 38.3 
Total 355 
No 182 51.5 
History of abortion (among ever 
Yes 172 48.7 
pregnant) 
Total 353 100 
100 58.1 
The number of induced abortion (among 
~2 72 41.8 
ever had an abortion) 
Total 172 
No 245 69.8 
History of miscarriage (among ever 
Yes 106 30.2 
pregnant) 
Total 351 100 
67 62.6 
The number of miscarriages (among ever 
~ 2 39 36.4 
had a miscarriage) 
Total 106 100 
Among the participants who had used OC, slightly more than half, 52.2% (n = 47 of 90), had 
used it for less than a year, 40.0% had used it for between I and 4 years and only 7.7%, of the 
106 
University of Ghana  http://ugspace.ug.edu.gh
women had used OC for 5 or more years. The table 4.3 also shows that while about half of the 
participants, 51.6% (n = 207) had experienced menstrual difficulties, 48.4% (n = 194) stated they 
had not and 14 women did not provide any such information. 
Also, a majority of the participants, 86.4% (n = 355 of 415) reported that they had ever been 
pregnant and the number of pregnancies per woman ranged between 1 and 18, with a median of 
3.0 and an interquartile range of 4.0. A smaller proportion, 16.1 % of these 355 participants had 
been pregnant only once, as at the time of the study. Those who had been pregnant for between 2 
and 4 times constituted 45.6% (n = 162 of 355), while 38.3% had been pregnant for at least 5 
times. One hundred and seventy two (48.7%) of the participants who had ever been pregnant, 
had ever had an abortion and \06 (30.2%) had ever experienced a miscarriage. The number of 
abortions reported ranged between 1 and 5 with a median of I.More than half (58.1 %) reported 
they had had one abortion while 41.8% reported they had had 2 or more abortions. The number 
of miscarriages reported, by the 106 participants who had experienced miscarriages, ranged 
between I and 5 with a median of 1. While 62.6% who miscarried had a single experience, 
36.4% had miscarried two or more times. 
4.3.3 Other Risk Characteristics (Factors) 
While more than half (61.3%) of the 415 participants in this study were not consumers of 
alcoholic products as at the time of the study, 39.7% were. Almost all, 97.3%, of the 415 
participants reported they had never smoked, with only 0.5% (n = 1) reported that she was a 
smoker, while 2.2% did not report their smoking status. Only a small proportion, 17.1% (n = 71) 
of the 415 participants had heard of cervical cancer; an even smaller proportion stated what, in 
their view, was the cause of cervical cancer (4.1 %, n = 17) and how it could be prevented (1.7%, 
107 
University of Ghana  http://ugspace.ug.edu.gh
n = 7). Of the 17 participants who stated a cause, 3 (17.6%) stated having sex at a tender age, 5 
(29.4%) stated having multiple male sexual partners, 2 (11.7%) stated sexual transmission from 
males, 3 (17.6%) stated bacterial and virus infections of the cervix, I (5.8%) stated the insertion 
of dough in the vagina, 2 (11.7%) stated STI that when treated did not resolve and I (5.8%) 
stated unprotected sex, as the cause of cervical cancer. Of the 7 participants who stated a 
preventative measure, 1 (14.3%) stated being faithful to one sexual partner, 3 (42.8%) stated 
going for regular check-ups at the hospital and reporting unusual conditions to the doctor, 2 
(28.6%) stated regular use of condom, 1 (14.3%) stated regular examination by cervical test, and 
I (14.3%) stated participating in screening activities, as preventative measures. 
4.4 PARTICIPANT CHARACTRISTICS THAT INFLUENCED REPORTING AND 
CONDOM USE 
4.4.1 Participant Characteristics that may have Influenced Reporting Status 
These analyses of the association of the demographic characteristics with reporting status were 
performed separately for the three reporting strategies, which are; report within the community, 
long duration to report to the hospital and short duration to report to the hospital. This 
automatically excludes the 38 peer-recruited women. Among the participants who were to report 
within the community, it was determined that there were no statistically significant associations 
between reporting status with all the demographic characteristics (Table 4.4) . Similarly, for the 
participants who opted for the short duration within which they were to report for specimen 
collection, reporting status was not significantly associated with any of the demographic 
characteristics (Table 4.5). 
108 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.4: Distribution and Association of participants' demographic characteristics with 
their reporting status, for Community specimen collection. 
Response to Community approach for 
s~ecimen collection, n {%} 
Demographic 
Categories Did not J: p value 
characteristics Reported Total 
= report (n 99) (N = 103) 
{n = 4} 
15 - 19 6 (6.1) 0(0.0) 6 (5.8) 
20 - 24 16 (16.2) 2 (50.0) 18 (17.5) 
25 - 29 27 (27.3) I (25.0) 28 (27.2) 
30 - 34 6 (6.1) 0(0.0) 6 (5.8) 
35 - 39 12 (12.1) 0(0.0) 12(11.7) 
Age (categorised) 40 - 44 7 (7.1) 1 (25.0) 8 (7.8) 
6.097 
(0.730) 
45 - 49 6 (6.1) 0(0.0) 6 (5.8) 
50 - 54 8 (8.1) 0(0.0) 8 (7.8) 
55 - 59 4 (4.0) 0(0.0) 4 (3.9) 
60 or older 7 (7.1) 0(0.0) 7 (6.8) 
Total 99 4 103 
Unmarried 36 (37.5) 2 (50.0) 38 (38.0) Fisher exact 
Marital status Married 60 (62.5) 2 (50.0) 62 (62.0) (0.633) 
Total 96 4 100 
Christian 99 (100) 4 (100) 103 (100) 
Muslim 0(0.0) 0(0.0) 0(0.0) 
Religion 0.001 
Other 0(0.0) 0(0.0) 0(0.0) 
Total 99 4 103 
No formal 
21 (21.6) 0(0.0) 21 (20.8) 
education 
Primary 25 (25 .8) I (25.0) 26 (25.7) 
Educational status Junior secondary 38 (39.2) I (25.0) 39 (38.6) 
9.205 
(0.056) 
Senior secondary 7 (7.2) 2 (50.0) 9 (8.9) 
Post-Secondary 6 (6.2) 0(0.0) 6 (5.9) 
Total 97 4 101 
Unemployed 12(12.5) 0(0.0) 12 (12.0) 
Formal 
17 (17.7) I (25.0) 18 (18.0) 
employment 
Skilled worker 23 (24 .. 0) 0(0.0) 23 (23.0) 4.173 <0.383) occupation 
Trader 31 (32.3) 3 (75.0) 34 (34.0) 
Agro-worker 13 (13.5) 0(0.0) 13 (13.) 
Total 96 4 100 
109 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.5: Distribution and Association of participant demographic characteristics with the 
reporting status, for Short duration hospital based specimen collection. 
Response to Short appointment approach 
for s~ecimen collection, n {%} 2 
Variables Categories Did not X P 
Reported Total value 
report 
(n = 53) (N = 90) 
{n = 37} 
15 - 19 0(0.0) 6 (16.2) 6 (6.7) 
20 - 24 10 (18.9) 6 (\6.2) 16(17.8) 
25 -29 12 (22.6) 6 (16.2) 18 (20.0) 
30 - 34 11 (20.8) 3 (8.1) 14 (15.6) 
35 - 39 10 (18.9) 4 (10.8) 14 (15.6) 
17.863 
Age (categorised) 40 - 44 6 (11.3) 6 (16.2) 12 (13.3) (0.037)* 
45 - 49 3 (5.7) 3 (8.1) 6 (6.7) 
50 - 54 1 (1.9) 0(0.0) 1 (1.1) 
55 - 59 0(0.0) 2 (5.4) 2 (2.2) 
60 or older 0(0.0) 1 (2.7) 1 (Ll) 
Total 53 37 90 
Unmarried 19 (35.8) 19(51.4) 38 (42.2) 
2. 146 
Marital status Married 34 (64.2) 18(48.6) 52 (57.8) 
(0.143) 
Total 53 37 90 
Christian 50 (94.3) 31 (86.1) 81 (91.0) 
Muslim 3 (5.7) 3 (8.3) 6 (6.7) 3.331 
Religion Other 0(0.0) 2 (5.6) 2 (2.2) (0.189) 
Total 53 36 89 
No formal 
6 (I \.3) 5 (13.5) I] (12.2) 
education 
Primary 8 (IS. I) 5 (13.5) 13 (14.4) 
2.406 
Educational status Junior secondary 3 I (58.5) 24 (64.9) 55 (61.1) 
(0.662) 
Senior secondary 5 (9.4) 3 (8. I) 8 (8.9) 
Post-Secondary 3 (5.7) 0(0.0) 3 (3.3) 
Total 53 37 90 
Unemployed 3 (5.8) 2 (5.6) 5 (5.7) 
Formal 
5 (9.6) 7 (19.4) 12 (\3.6) 
employment 4.046 
Occupation Skilled worker 20 (38.5) 10(27.8) 30 (34.1) (0.400) 
Trader 21 (40.4) 12 (33.3) 33 (37.5) 
Agro-worker 3 (5.8) 5 (\3.9) 8 (9.1) 
Total 52 36 88 
• indicates significant association 
110 
University of Ghana  http://ugspace.ug.edu.gh
However, for the participants who opted for a long duration within which to report to the hospital 
for specimen collection, a statistically significant association was only obtained between age 
(continuous) and reporting status (Spearman Correlation = 0.202; p = 0.022). Those who 
reported, had a median age of 35.0 years, (IQR of 18.0 years), while those who did not report, 
had a median age of 29.0 years (IQR of 16.0 years). However, when the age was categorised it 
became statistically non-significant (Table 4.6). In other words, the reporting for specimen 
collection by each of the strategies was, to a large extent, random and not influenced by the 
demographic characteristics of the participants. 
4.4.2. Participant Characteristics that may have influenced Condom Use 
A chi-square analysis showed that there was a significant relationship between the use of 
condom and educational status, as well as between the use of condoms and marital status (Table 
4.7; p = 0.001). It was determined that more participants than expected with educational status of 
junior secondary or higher stated the use of condom with their male sexual partners while less 
than expected participants with primary or no formal educational stated the use of condoms with 
their male sexual partners. Similarly, more of the unmarried participants than expected stated the 
use of condoms with their male partners while fewer of the married participants than expected 
stated same (Table 4.8). The association of condom use with the occupation of the participants 
was also significant (p = 0.0001) and revealed that for the unemployed, formally employed and 
skilled workers, more participants than expected, stated that they used condom while for the 
traders and agro-workers, less participants than expected reported the use of condoms during 
sexual intercourse. 
1 II 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.6: Distribution and Association of participants' demographic characteristics with 
their reporting status, for Long duration hospital based specimen collection. 
Response to Long duration for specimen 
collection, n {%) 
Categories Did not -l p value 
Variables Reported Total 
= report (n 49) 
{n = (N = 129) 80) 
15 -19 3 (6.1) 6 (7.6) 9 (7.0) 
20-24 6 (12.2) 16 (20.3) 22 (17.2) 
25-29 9 (18.4) 19 (24.1) 28 (21.9) 
30 - 34 5 (10.2) 12 (15.2) 17 (13.3) 
35 -39 6(12.2) 7 (8.9) 13 (10.2) 
Age 13.657 40-44 8 (16.3) 5 (6.3) 13 (10.2) 
(categorised) (0.135) 
45 -49 3 (6.1) 10 (12.7) 13 (10.2) 
50-54 6 (12.2) 4 (5.1) \0 (7.8) 
55 -59 2 (4.1) 0(0.0) 2 (1.6) 
60 or older 1 (2.0) 0(0.0) 1( 0.8) 
Total 49 79 128 
Unmarried II (22.4) 27 (34.6) 38 (29.9) 
2.124 
Marital status Married 38 (77.6) 51 (65.4) 89 (70.1) (0.145) 
Total 49 78 127 
Christian 44 (89.8) 64 (80.0) 108 (83.7) 
Muslim 5 (10.2) 1407.5) 19 (14.7) 2.672 
Religion 
Other 0(0.0) 2 (2.5) 2 (1.6) (0.263) 
Total 49 80 129 
No fomlal 
7 (14.3) 21 (26.9) 28 (22.0) 
education 
Primary 13 (26.5) 14 (17.9) 27 (21.3) 
Junior 
Educational 23 (46.9) 34(43.6) 57 (44.9) 4.576 secondary 
status (0.334) Senior 
5 (10.2) 5 (6.4) 10(7.9) 
secondary 
Post-Secondary I (2.0) 4 (5.1) 5 (3.9) 
Total 49 78 127 
Unemployed 3 (6.8) 2 (2.5) 5 (4.1) 
Formal 
5(11.4) 12(15.2) 17 (13.8) 
employment 
occupation Skilled worker 7 (15.9) 12(15.2) 19 (15.4) 
5.717 
Trader 26 (59.1) 
(0.221) 
37 (46.8) 63 (51.2) 
Agro-worker 3 (6.8) 16 (20.3) 19(15.4) 
Total 44 79 123 
112 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.7: Association between use of condom and demographic characteristics of the 
participants 
Use of condom 
Demographic 
Category Observed (Expected) Association, (p-value) 
Characteristics Spearman, 
No % Yes 0/0 X2 R 
15 -19 12(14.6) 5.0 13 (10.4) 7.7 
20-24 28 (36.9) 11.7 35 (26.1) 20.7-
25 -29 38 (56.8) 15.9 59 (40.2) 34.9-
30 -34 24 (26.9) 10.0 22 (19.1) 13.0 
35 -39 36 (29.9) 15.1 15 (21.1) 8.9 -0.39 52.05 
Age 
40 -44 40 (29.9) 16.7- 11 (21.1) 6.5 «0.001) (0.001) 
45 -49 22 (17.6) 9.2 8 (12.4) 4.7 
SO-54 19 (12.9) 7.9- 3 (9.1) 1.8 
S5 - S9 10 (7.6) 4.2 3 (5.4) 1.8 
60 or Older 10 (5.9) 4.2- 0(4.1) 0.0 
No Formal 
57 (39.8) 24.1- II (28.2) 6.5 
Education 
Primary 51 (4S.1) 21.S 26 (31.9) IS.5 
Educational Junior 30.27 100 (113.5) 42.2- 94 (80.S) 56.0 
status Secondary (0.001) 
Senior 
21 (25.7) 8.9 23 (18.3) 13.7 
Secondary 
Post -Secondary 8 (12.9) 3.4 14 (9.1) 8.3-
Unmarried 65 (86.4) 27.4 82 (60.6) 49.4- 20.33 
Marital status 
Married 172 (150.6) 72.6- 84(105.4) 50.6 (0.001) 
Unemployed 13 (l4.S) 5.6 12(10.5) 7.2 
Formal 
25 (37.2) 10.8 39 (26.8) 23.4-
Employment 26.89 
occupation Skilled Worker 50 (54.1) 21.6 43 (38.9) 25.7 (0.001) 
Trader 104 (99.4) 44.8 67 (71.6) 40.1 
Agro-Worker 40 (26.7) 17.2- 6 (19.3) 3.6 
_ The proportion in the Yes category was significanlly different /Tom that of the No category 
113 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.8: Sexual risk characteristics of the participants' which are associated with the use 
of condom 
Use of condom 
Sexual risk Category Observed (Expected) 
Characteristics "/..,2, (p-value) 
No 0/0 Yes % 
Age at first sexual <20 136 (141.6) 60.4 112 (106.4) 66.3 
intercourse >/=20 89 (83.4) 39.6 57 (62.6) 33.7 1.33 (0.515) 
< II 66 (98.6) 29.5 107· (74.4) 63.3 
11-19 60 (54.7) 26.8 36 (41.3) 21.3 
Sexual age 51.30 (0.001) 
20 - 29 63 (46.7) 28.1 19* (35.3) 11.2 
> 29 35 (23.9) 15.6 7* (18.1) 4.1 
0 41 (35.7) 17.2 20 (25.3) 11.8 
Current number of 196 (201.2) 82.4 148 (142.8) 87.6 Fisher's exact 
male partners 2 0(1.0) 0.0 1 (0.0) 0.6 test's (0.159) 
3 1 (1.0) 0.4 0(0.0) 0.0 
60(51.9) 26.4 30 (38.1)· 18.0 
Lifetime number of 4.3 2 63 (63.4) 27.8 47 (46.6) 28.1 
sexual partners (0.116) 
>/= 3 104 (111.8) 45 .8 90 (82.2) 53.9 
Had any STI within No 126 (110.6) 53.2· 63 (78.4) 37.5 9.692 (0.002) 
the past 10 years Yes 111 (126.4) 46.8 105 (89.6) 62.5· 
No 149 (138.2) 62.3 87 (97.75) 51.5 
Ever had an abortion 4.791 (0.029) 
Yes 90 (100.75) 37.7 82 (71.245) 48.5-
55 (52.84) 6l.1 46(48 .15) 56.1 
Number of abortion 0.445 (0.505) 
>/= 2 35(37.15) 38.9 36 (33.848) 43.9 
• The proportion in the Yes category was significantly different from that of the No category 
Age, analysed as a numerical variable, was significantly but inversely associated with condom 
use (Spearman correlation R = -0.39; p = 0.001). Age remained significant after it was 
categorized (p = 0.001) and showed that more participants than expected within the ages of 15 
114 
University of Ghana  http://ugspace.ug.edu.gh
years to 34 years stated the use of condom. Also, there were significant associations (Table 4.7) 
between the use of condoms and the sexual-age (as a numerical variable) of the participants as 
well as with the lifetime number of male sexual partners (numerical variable). The use of 
condoms decreased among the participants with higher sexual age, while there was an increase in 
the stated use of condoms among the participants with a higher lifetime number of male sexual 
partners. 
However, the association between condom use and the following (analysed as numerical 
variables) were all not statistically significant: the AFSI (p > 0.SI5.), and current number of male 
sexual partners (p > 0.IS9). These variables were categorized and the association with condom 
use were re-analysed by chi-square (Table 4 .8). The sexual age remained significantly 
associated with the state of condom use (X2 = S1.30, p = 0.001). However, the lifetime number 
of sexual partners became statistically not significantly associated (X2 = 4.3, p = 0.116) with the 
stated use of condom (Table 4.8). The AFSI after being categorized as such «19, 19 - 21 , and 
>21 years) remained not significantly associated (p = 0.515) with the stated use of condom. 
Also, after the current number of male sexual partners were categorized into a dichotomous 
variable (currently have or do not have a male sexual partner) there was not a significant 
association (2-sided Fisher 's exact test's p = 0. IS9) with the stated use of condom. 
115 
University of Ghana  http://ugspace.ug.edu.gh
4.5 ADDITONAL FINDINGS - SPECIMEN COLLECTION 
4.5.1 Pre-performance preference of Specimen Collection Method 
Since the preference for specimen collection were not in any way related to the reporting 
strategies, data of all the 253 participants (including the 38 peer recruited) who reported for 
specimen collection were used in this analysis. 
Of this number, 226 participants provided specimen by self and health personnel collection, 7 
participants opted for only health personnel collection and 18 participants opted for only self-
collection. SC specimen of 2 women and HPC specimens of two other women were lost (Figure 
4.2). The major reasons stated by the participants ' pre-performance preference for only health-
personal specimen collection included blindness (one participant) and the feeling that they might 
hurt themselves with the self-specimen collection brush. On the other hand, the major reasons for 
a pre-performance preference for only self-specimen collection were that the participants did not 
want the insertion of the speculum, some were just afraid of the speculum while others resisted 
the speculum insertion. 
4.5.2 Post-performance Preferences for Specimen Collection Method 
Out of the 253 participants who provided at least one specimen, about a half, 50.3% (n = 127) 
preferred the health personnel specimen collection after performing the sample collection, while 
20.2% (n = 51) preferred self-specimen collection after performing (Table 4.9). The others, 
either did not have any post-performance preference (19.4%, n = 49) or did not state their post-
performance preference (9 .9%, n = 24). The commonest reason for preferring self-specimen 
collection post-performance was the experience of slight pain and/or fear for the health personnel 
specimen collection method (Table 4.10). This was stated by 68.6% (n = 35) of participants who 
116 
University of Ghana  http://ugspace.ug.edu.gh
preferred self-specimen collection post-performance. Six other participants (11.8%), each 
preferred self-specimen collection post-performance because of the privacy it provided and how 
simple it was to perform. 
N = 253 
18 provided 7 provided 
only SC only HPC 
Sample samples 
For 2 women, SC were lost 
Figure 4.2: Distribution of the performance of sample collection. 
Table 4.9: Distribution of preferences for specimen collection methods after providing the 
specimen. 
Preference Frequency Percentage 
Self-specimen collection 51 20.2 
Health personnel specimen collection 127 50.3 
I have no preference 49 19.4 
Not reported 25 9.9 
Total 252 100.0 
117 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.10: Reasons for the preference of self-specimen collection. 
Reasons Frequency Percentage 
The health personnel sample collection was with slight pain and lor fear 35 68.6 
Issues of privacy 6 11.8 
The self - specimen collection was simple and easy to perform 6 11.8 
Can't explain 2.0 
Stated no reason .':"> 5.9 
Total 51 100.0 
Of the 127 participants who preferred health personnel specimen collection, the commonest 
reason, stated by 85 (66.9%), was that the health personnel specimen collection was most likely 
better performed than the self-specimen collection they performed, because the nurse and 
medical officers were professionals, experienced and knew what they were about (Table 4.11). 
Sixteen participants (12.6%) stated that SInce they had no knowledge of cervical cancer 
screening, the health personnel specimen collection was preferable. The third most common 
reason, stated by 11.0% (n = 14) of the participants, was that they were not confident they 
performed the self-specimen collection correctly. Other reasons given for the preference of 
health personnel specimen collection by a few of the participants are shown in Table 4.11. 
118 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.11: Reasons for the preference of health personnel specimen collection. 
Reasons Frequency Percentage 
The health personnel would have done them better because they were 85 66.9 
professionals, experienced and know what they were about 
The participants had no idea about cervical screening 16 12.6 
The participants were not confident they performed the self-specimen 
14 11.0 
collection correctly. 
In order to obtain good and/or the right results 2 1.4 
The health personnel would find another problem or unusual 
2 1.4 
occurrences if any around the sampling area 
The Pap smear was not difficult and there was no pain 0.7 
The participant was afraid she was going to hurt herself with the self-
0.7 
specimen collection device 
The participant stated that the self-sampling device is not user- 0.7 
friendly 
The nurse could see what she was doing, but the participant did not 
0.7 
see what she did during self-specimen. 
Had no problem with the health personnel examinations and sample 
0.7 
collection 
Gave no reason 6 4.7 
Total 127 100.0 
4.6 HUMAN PAPILLOMA VIRUS (HPV) TESTING 
4.6.1 HPV Genotype Specific Prevalence 
The quality of the specimens obtained from the 253 women who provided specimens for this 
study were tested by a real-time peR amplification of the human housekeeping gene, RNase H. 
Although 17 of the 244 self-collected and 6 of the 233 health personnel collected specimen were 
not consistently positive with triplicate analyses of the RNase H gene (2 of 3 or I of 3 were 
positive), they were included in the HPV analysis. For the determination of the agreement 
1 19 
University of Ghana  http://ugspace.ug.edu.gh
between the two specimen collection methods, only data generated by the 226 women who 
provided both specimens were used. 
Out of the 46 HPV genotypes tested for, 37 and 26 genotypes were detected with the self and the 
health personnel collected specimens respectively (Table 4.12). Overall, 38 genotypes were 
detected with both methods of specimen collection. The 8 genotypes that were not detected in 
this study were HPVll, HPV13, HPV32, HPV61, HPV71, HPV89, HPV97 and HPV26. Only 
HPV43 was not detected with SC specimen, but was detected with HPC specimen. Also, HPV33, 
HPV44, HPV53, HPV68, HPV69, HPV70, HPV72, HPV73, HPV85, HPV86, HPV90 and 
HPV91 were detected with SC specimens but not with HPC specimens. The six most detected 
HPV genotypes with the self-collected specimen were HPV16 (5.9%; 95% CI 3.0% - 9.0%), 
HPV35 (4 .7%; 95% CI 2.0% - 8.0%), HPV40 (4.7%; 95% CI 2.0% - 8.0%), HPV45 (4.3%; 
95% CI 2.0% - 7.0%), HPV58 (4.0%; 95% CI 2.0% - 7.0%) and HPV18 (3 .6%; 95% CI 1.0-
6.0) (Table 4.12). 
On the other hand, the six most detected HPV genotypes with the health personnel collected 
specimen were HPV35 (2.8%; 95% CI 1.0% - 5.0%), HPV58 (2 .8%; 95% CI 1.0% - 5.0%), 
HPVI6 (2.4%; 95% C] 1.0% - 5.0%), HPVI8 (2.4%; 95% CI 1.0% - 5.0%), HPV56 (2.4%; 
95% CI 1.0% - 5.0%) and HPV45 (2.0%; 95% C] 0.0% - 4.0%). For each of the detected 
genotypes, a higher prevalence was detected with the self-collected sample than was detected 
with the health personnel collected specimen (Table 4.12), however, the differences were not 
significant (p values were between 0.092 and 0.743). 
120 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.12: HPV genotype prevalence determined with each specimen collection method. 
Self-Collected Specimen, n=244 Health Personnel Collected Specimen, n = 230 
HPV Genotype 
n Prevalence 95%CI n Prevalence 95%CI 
HPV16 15 5.9 3.0 - 9.0 6 2.4 1.0 - 5.0 
HPV35 12 4.7 2.0 - 8.0 7 2.8 1.0 - 5.0 
HPV40 12 4.7 2.0- 8.0 4 1.6 0.0 -4.0 
HPV45 11 4.3 2.0 - 7.0 5 2.0 0.0 -4.0 
HPV58 10 4.0 2.0-7.0 7 2.8 1.0 - 5.0 
HPVI8 9 3.6 1.0 - 6.0 6 2.4 1.0 - 5.0 
HPV66 8 3.2 1.0 - 6.0 6 2.4 1.0 - 5.0 
HPV59 7 2.8 1.0 - 5.0 3 1.2 0.0 - 3.0 
HPV6 7 2.8 1.0 - 5.0 3 1.2 0.0-3.0 
HPV31 7 2.8 1.0 - 5.0 2 0.8 0.0 - 2.0 
HPV74 7 2.8 1.0 - 5.0 1 0.4 0.0 - 1.0 
HPV42 5 2.0 0.0 - 4.0 5 2.0 0.0 - 4.0 
HPV52 5 2.0 0.0 -4.0 4 1.6 0.0 -4.0 
HPV51 5 2.0 0.0 - 4.0 3 1.2 0.0 -3.0 
HPV62 5 2.0 0.0 -4.0 2 0.8 0.0 -2.0 
HPV81 5 2.0 0.0 - 4.0 0.4 0.0 - 1.0 
HPV83 5 2.0 0.0 -4.0 0.4 0.0 -1.0 
HPVn 5 2.0 0.0 -4.0 NO 
HPV90 5 2.0 0.0 -4.0 ND 
HPV67 4 1.6 0.0 -3.0 0.4 0.0 - 1.0 
HPV53 4 1.6 0.0 -3.0 ND 
HPV56 3 1.2 0.0 - 3.0 0.4 0.0 - 1.0 
HPV33 3 1.2 0.0 -3.0 ND 
HPV70 3 1.2 0.0 - 3.0 NO 
HPV87 2 0.8 0.0-2.0 2 0.8 0.0 -2.0 
HPV39 2 0.8 0.0 - 2.0 0.4 0.0 - 1.0 
HPV30 2 0.8 0.0 -2.0 0.4 0.0 -1.0 
HPV54 2 0.8 0.0 - 2.0 0.4 0.0 - 1.0 
HPV44 2 0.8 0.0-2.0 NO 
HPV86 2 0.8 0.0 - 2.0 NO 
HPV82 1 0.4 0.0 -1.0 0.4 0.0 -1.0 
HPV84 0.4 0.0 - 1.0 0.4 0.0 - 1.0 
HPV68 0.4 0.0 -1.0 NO 
HPV69 0.4 0.0 - 1.0 NO 
HPV85 0.4 0.0 -1.0 NO 
HPV73 0.4 0.0 - 1.0 NO 
HPV43 ND 0.4 0.0 -1.0 
-ND = not detected; ordcr of listing was, according to HPY prevalence determined by self-collected samples 
121 
University of Ghana  http://ugspace.ug.edu.gh
4.6.2 Overall and Risk Type HPV Prevalence 
The overall prevalence of HPV infection as determined by the 244 self-collected specimen was 
43.1 % (95% CI 38.0% - 51.0%) while that determined by the 233 health-personnel collected 
specimen was 23.3% (95% CI of 19.0% - 31.0%). The difference between these prevalences was 
significant rt = 28.943; P = 0.0001; Table 4.13). 
The prevalence of each risk type group was determined with the 244 self-collected specimens 
and the 233 health personnel collected specimen separately; the prevalence of high risk (HR) 
HPV types (see section 2.8.1 for all the genotypes that make each risk HPV group) detected with 
self-collected specimen was 27.2% (95% CI of 23.0% - 34.0%) and that detected with the 
health-personnel collected specimen was 16.6% (95% CI of 14.0% - 24.0%). Similarly, the 
prevalence of low risk (LR) HPV was 23.3% (95% CI of 19.0% - 30.0%) with self-collected 
specimen and 9.9% (95% CI of 6.0% - 14.0%) with health-personnel collected specimen (Table 
4.13). On the other hand, the prevalence of the probable high risk (PHR) HPV types were very 
low, 4.7% (95% CI of 2.0% - 8.0%) with self-collected specimen and 0.4% (95% CI of 0.0% -
1.0%) with health personnel collected specimen. 
Single HPV genotype infections were detected among 62 (56.9%) of the 109 women who were 
HPV positive with self-collected specimen while 43.1 % of them were infected with multiple 
genotypes of HPV. On the other hand, while single HPV infections were detected among 43 
(72.1 %) of the 59 women who were HPV positive with their health-personnel collected 
specimen. 27.1%  of health-personnel collected specimens were infected with multiple genotypes 
of HPV. A chi-square test of the difference in the proportions of the following; overall HPV 
122 
University of Ghana  http://ugspace.ug.edu.gh
infections, single/multiple, HR, and LR HPV infections determined by the two methods were 
each significant (Table 4.13). 
4.6.3 Agreement between Self and Health Personnel Collected Specimen 
A Cohen's kappa (k) analysis for each of the HPV infection category (Table 4.13) using data on 
the 226 women who provided both SC and HPC showed a significant level of agreement (p = 
0.0001) in the detections of HPV with self and health-personnel collected specimen. However, 
the level of agreement between the two specimen collection methods was poor for both the 
detection of PHR HPV infections and for the detection single HPV infections (kappa values, k = 
0.147 and 0.128 respectively). The agreements between the two collection methods were fair for 
the detection of LR HPV infection, the overall HPV infection and multiple HPV infection, (k = 
0.328, 0.321 and 0.286 respectively). The agreement between the two specimen collection 
methods for HR HPV infections was moderate (k = 0.402). 
As indicated in table 4.14, of the 226 participants who provided both self and health personnel 
collected specimen, 44 (19.5%) were HPV positive with both their HPC and SC specimens; and 
108 (47.8%) were HPV negative with both their HPC and SC specimens. Therefore, 152 (44 + 
108) of the 226 participants had the same results in respect of overall HPV detection by both SC 
and HPC, resulting in an overall HPV detection concordance between SC and HPC specimen of 
67.2%. Furthermore, 13 (5 .8%) of the 226 participants were HPV positive by their HPC 
specimen, but HPV negative by their self-collected specimen. Conversely, 61 (27.0%) of the 226 
participants were HPV positive by their self-collected specimen but HPV negative by their HPC 
specimen. Therefore, the discordance in the overall HPV detection between HPC and SC 
123 
University of Ghana  http://ugspace.ug.edu.gh
specimen was 32.8%. A McNemar's test indicated that the marginal proportions of the overall 
HPV detection were all significantly different in each case (p = 0.0001). 
Table 4.13: HPV prevalences and the extent of agreements between self and health 
personnel collected specimen. 
Self-collection Health Personnel- Cohen's Kappa, 
Infection collection 
Categories 
Prevalence 95% CI Prevalence 95%CI k P value 
Overall HPY 43.1 38.0 - 51.0 23.3 19.0 - 31.0 0.321 0.0001 
HRHPV 27.2 23.0 -34.0 16.6 14.0 -24.0 0.402 0.0001 
LRHPY 23.3 19.0 - 30.0 9.9 6.0 -14.0 0.328 0.0001 
PHRHPV 4.7 2.0-8.0 0.4 0.0 -1.0 0.147 0.0001 
Proportion" Proportionb 
Single 56.9% 72.9% 0.128 0.048 
Multiple infection 43.1% 27.1% 0.286 0.0001 
HR = hi gh risk; LR = low risk; PHR = probable high risk. Prevalence was determined with 244 SC specimens and 
230 HPC specimens. Chi-square and Kappa analysis were performed with 266 pairs ofSC and HPC specimen. '01' 
the 109 HPY positive with SC; b of the 59 HPY positive with HPC 
124 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.14: Concordance and discordance in the detection of HPV infections with self and 
health personnel collected specimen 
Self- Health personnel McNemar's 
HPV 
collected collected specimen test Total Concordance Discordance 
status (p value) 
specimen Negative Positive 
lOS 13 
Negative 121 
(47.S%) (5.8%) 
Overall 61 44 (lOS + 44) (13+61) 
Positive 105 0.0001 
HPV (27.0%) (19.5%) 67.3% 32.S% 
Total 169 57 226 
147 13 
Negative 160 
(65.2%) (5.7%) 
HR 37 29 (148+29) (13 +37) 
Positive 66 0.0001 
HPV (16.3%) (12.8%) 77.9% 22.1% 
Total 184 42 226 
% are of the overall total of participants who performed both methods, n = 226. 
A similar analysis of the concordance and discordance in the detection of high risk (HR) HPV by 
the SC and HPC specimen was performed. Of the 226 participants who provided both SC and 
HPC specimen, 148 participants were HR HPV positive by their SC and HPC specimen, while 
29 participants were HR HPV negative by both their SC and HPC specimens. Therefore, the 
concordance of the two specimen types in the detection of HR HPV was 77.9%. On the other 
hand , while 13 (5.7%) of participants were HR HPV positive by their HPC specimen they were 
HR HPV negative by their SC specimen. Additionally, 37 (16.3%) of the 226 participants were 
HR HPV negative by their HPC specimen, but they were HR HPV positive by their SC 
specimen. Therefore, the discordance between the two types of specimen in respect of the 
detection of HR HPV was 22.1 %. A McNemar's test indicated that the marginal proportions of 
the high risk HPV detection were all significantly different in each case (p = 0.0001). 
125 
University of Ghana  http://ugspace.ug.edu.gh
4.6.4 Test of Hypothesis on HPV Prevalence 
A comparison of the overall prevalence ofHPY by the HPC specimen (23.3%; 95% CI 19.0%-
31.0%) with the hypothesized 26.3% prevalence (Thomas et aI. , 2004) was by a Binomial Test at 
95% confidence level. The p-value obtained by this test was 0.310 and since the p-value was 
more than 0.05, the study failed to reject the null hypothesis (Ho). Therefore, the HPY prevalence 
of the Akuse sub-district of the Eastern Region of Ghana as determined with the HPC specimen 
was not different from the HPY prevalence of26.3% reported by Thomas et aI. , 2004. 
4.6.5 Age Specific Prevalence (ASP) 
The overall HPY, high risk HPY and low risk HPY age specific prevalence (ASP) were 
determined for both self and health personnel collected specimen (Figures 4.3 - 4.5). Age was 
determined to be significantly associated with overall HPY prevalence (X2 = 36.1; p = 0.001), 
H R HPY prevalence (X2 = 26.09; p = 0.002) and with LR HPY prevalence (X2 = 21.49; p = 
0.011). Additionally, the ASP determined with SC specimen were usually higher than those 
determined with HPC specimen. 
T he ASP of overall HPY infections determined with the self-collected specimen was trimodal 
wi th peaks at 20-24 years, 40-44 years and 55-59 years. In respect of the HPC specimen, the 
ASP of the overall HPV infection was bimodal with peaks at 20-24 years and 55-59 years. For 
both SC and HPC specimen, the overall HPY prevalence was the same for the age group 45-49 
years and the age group of > 59 years. The ASP of high risk HPY prevalence (Figure 4.4) was 
similar to that of the overall HPV prevalence described above. That is, for the SC specimen a 
trimodal pattern was obtained with peaks at 20-24 years, 40-44 years and 55-59 years. For the 
126 
University of Ghana  http://ugspace.ug.edu.gh
HPe specimen, the age-specific HR HPY prevalence pattern was bimodal with peaks at 20-24 
years and 55-59 years. 
90 
- Self-Collected Specimen 
80 -.. Ilellltb Personnel Collected Specimen 
70 
~ 
~ 60 
... "=.   
-.; SO ! .:-..  
CI. 40 
> 
=CI. =..   .... -... ...:-.  
30 .. ' . 
0 20 I '.--_ .... ---.-
10 
0 ! 
15-19 20-24 25-29 30-34 35-39 40-44 45-49 50-54 55-59 >59 
Age (years) 
Figure 4.3: Age specific overall HPV infection prevalence obtained with self and health 
personnel collected specimen. 
T he I-IR I-IPY prevalence for the se and HPe specimen were the same for women 45 years and 
older (Figure 4.4). The age-specific prevalence of low risk (LR) HP\V determined with se was 
bimodal with peaks at 20-24 years and 55-59 years while that for the HPe specimen was also 
bimodal but with peaks at 15-19 years and 55-59 years (Figure 4.5). 
127 
University of Ghana  http://ugspace.ug.edu.gh
60 
--Self-Collectcd Specimcn 
- . Health Personnel Collected Specimen 
50 
- 40 ~
~..   
.".d..   
>.  ....  30  
Iloo ., 
> I 
Iloo 
:c 20 I I 
I I 
ill: 
:c 
10 I • 
o I 
15-19  2D-24 25-29 3D-34 35-39 4. .... 4 45-49 50-54 55-59 >59 
Age (yens) 
Figure 4.4: Age specific high risk (HR) HPV infection prevalencc obtained with self and 
health personnel collected specimcn. 
4.7 TEST OF ASSOCIATIONS BETWEEN SEXUAL RISK CHARACTERISTICS AND 
HPV POSITIVITY 
Although the study reported In this thesis was not powered to determine and quantify the 
association between the sexual risk behaviour characteristics and HPY positivity the crude 
assoc iation between these was useful for comparing the two specimen collection methods as to 
whic h of the two was more informative (informativeness was defined as obtaining the 
appropriate directionality of the association, as has been widely reported/excepted for each of the 
ri sk characterist ics). Therefore, multivariable analyses to obtain adjusted odds ratios were not 
performed, which would have been misleading. 
128 
University of Ghana  http://ugspace.ug.edu.gh
80 
--Selr·Collected Specimen 
70 
- .. Hc.llh Personnel Collected Specimell 
60 
-~ 
~..  50 .=<.J   
-.;.;  40 :. 
0.': 
: 30 c>.  
:a:::  
..,j 20 
· -"'--ea 
~ ~ ~ ..... 
10 
~~. . ~ -- -.- - --.- ,I  -- -.' \ 
\ 
\ 
0 • 
15-19 20-24 25-29 30-34 35-39 40-44 45-49 SO-54 55-59 >59 
Age (years) 
Figure 4.5: Age specific low risk (LR) HPV infection prevalence obtained with self and 
health personnel collected specimen. 
4.7.1 Univariable Association between Sexual Risk Characteristics and HPV Infection 
Pos itivity Determined With Self-Collected Specimen 
The odds ratios obtained with the univariable logistic regression analysis for the association 
between the sexual characteristics and HPY positivity determined with self-collected specimens 
were shown in Table 4.15. The AFSI, when analysed as a numerical variable, recorded an odds 
ratio of less than 1.00 for its association with overall HPY positivity, HR HPY positivity, single 
infec tion and multiple infection and an odds ratio of 1.03 (95% CI 0.94 - 1.12) for its association 
w ith LR HPY positivity; however, all the odds ratio were not statistically significant. The 
analys is was repeated with AFSI categorized into two groups, younger than 20 years and 20 
years or older. With the former as the reference group, it was determined that an AFSI of 20 
years or older recorded an odds ratio of less than 1.00 for its association with overall HPY 
129 
University of Ghana  http://ugspace.ug.edu.gh
positivity, HR HPV positivity, single infection and multiple infection while the odds ratio for its 
association with LR positivity was greater than 1.00. Except for the odds ratio for the 
association with HR HPV positivity which was 0.46 (95% CI 0.24-0.85), all the others were not 
significant (Table 4.15). 
Compared to currently (as at the time of the study) not having a male sexual partner, the odds 
ratio for the association between having sexual partner and each of the HPV infection positivity 
was less than 1.00 and not statistically significant (Table 4.15). Similarly, the odds ratio recorded 
for the association between the lifetime number of male sexual partners (LNSP) and each 
category of HPV positivity were all not significant; however, the odds ratios were greater than 
1.00 except for single HPV infection. Categorising LNSP into three groups (1 , 2 and ~3), with 
having I LNSP as the reference group, revealed an odds ratio of greater] .00 for the association 
between having 2 LNSP and each of the categories of HPV positivity except for LR HPV 
positivity [OR = 0.98 (95% CI 0.79 - ] .22)]. On the other hand, the odds ratio for the association 
between having 3 or more LNSP and Overall, LR and single HPV positivity were less than 1.00. 
All these odds ratios were not significant. 
When analysed as a numerical variable, sexual-age was significantly associated with the overall 
HPV positivity (OR = 0.97, 95% CI 0.94 - 0.99) and with single HPV infection positivity (OR = 
0.96; 95% CI 0.93 - 0.99). However, it was not significantly associated with HR, LR and 
multiple HPV positivity, which also recorded odds ratios of less than 1.00 (Table 4 .] 5). 
130 
University of Ghana  http://ugspace.ug.edu.gh
T able 4.15: Crude odds ratio for the association between sexual characteristics and HPV 
infection positivity determined with self-collected specimen 
Odds ratio (95% C I), with Self-Collected Specimen 
Risk Factor Category 
Single Multiple 
Any HPV HRHPV LRHPV 
Infection Infection 
Age at fi rst sexual < 20 Reference 
intercourse 
( categorised) 0.64 0.46* 1.37 0.67 0.60 ~ 20 
(0.37 - 1.09) (0.24 - 0.85) (0.75 - 2.49) (0.36 - 1.27) (0.29 - 1.22) 
No Reference 
Currently have a 
male sexual partner 0.80 0.62 0.80 0.92 0.67 Yes 
(0.38 - 1.67) (0.29 - 1.35) (0.35 - 1.84) (0.37 - 2.29) (0.27 - 1.68) 
Lifetime number of 1.04 1.07 1.01 0.98 1.1 4 
sexual partners (0 .87 - 1.25) (0 .88 - 1.30) (0.82 - 1.24) (0.79- 1.22) (0.90 - 1.45) 
Reference 
Lifetime number of 1.04 1.27 0.78 1.03 1.06 
sel<ual partners 2 (0.50 - 2.15) (0.57 - 2.85) (0.34 - 1.81) (0.44 - 2.38) (0.40 - 2.82) 
(categorised) 
0.87 1.09 0.96 0.76 1.05 
~ 3 (0.45 - 1.71) (0.51 - 2.31) (0.45 - 2.04) (0.35 - 1.67) (0.43 - 2.56) 
0.97* 0.97 0.99 0.96* 0.97 
Sexual-age (0.94 - 0.99) (0 .95 - 1.00) (0.97 - 1.02) (0.93 - 0.99) (0.94 - J. 00) 
< 11 Reference 
0.35* 0.36* 0.27* 0.46* 0.22* 11-19 
Sexual-age (0.18 - 0.67) (0.17 - 0.76) (0.11 - 0.66) (0.21 - 0.99) (0.08 - 0.58) 
(categorised) 0.31· 0.38· 0.32* 0.16* 
20 - 29 0.44* 
(0.15 - 0.63) (0.17 - 0.84) (0.13 - 0.78) (0.20 - 0.98) (0.05 - 0.50) 
0.54 0.65 1.38 0.75 
> 29 0.34 
(0.23 - 1.27) (0.26 - 1.63) (0.58 - 3.29) (0.10 - 1.13) (0.28 - 2.01) 
No Reference 
Contracted STJ 
1.33 1.34 1.1 0 
Yes 1.24 
1.45 
(0 .79 - 2.22) (0.76 - 2.3) (0.61 - 2.00) (0.68 - 2.29) (0.73 - 2.86) 
No Reference 
Condom use 
1.44 1.09 1.48 l.l1 2.03* 
Yes 
(0.86 - 2.41) (0.62 - 1.92) (0.82 - 2.67) (0.60 - 2.06) (1.04 - 3.98) 
* the odds of the association was signi ficant, since the range of95% CI excludes 1.0. 
131 
University of Ghana  http://ugspace.ug.edu.gh
Repeating the analysis with sexual age categorized into the following groups, < 11 years, 11-19 
years, 20-29 years and >29 years, with < 11 years as the reference group, it was determined that 
having a sexual age of between 11 and 19 years as well as between 20 and 29 years were 
sign ificantly associated with each of the categories of the HPV positivity categories, the odds 
ratios were all less than 1.00 (Table 4.15). 
Compared to having not contracted an STI in the past 10 years, having contracted an STI was not 
signi ficantly associated with each of the categories of the HPV positivity with odds ratios greater 
than 1.00. Similar findings were determined for the use of condom. Compared to not using a 
condom, the odds ratio for the association of condom use with each of the categories of HPV 
positivity was greater than 1.00 and not statistically significant (Table 4.15). 
4.7.2 Univariablc Association bctween Sexual Characteristics and HPV Infection 
Positivity Determined with Health Personnel-Collected Specimen 
The odds ratios obtained by the univariable logistic regression analysis for the association 
between the sexual characteristics and HPV positivity determined with health personnel-
collected specimens were shown in Table 4.16. The AFSJ when analysed as a numerical variable 
recorded odds rati o of less than 1.00 for its association with overall HPV positivity, LR HPV 
positiv ity, and multiple infection positivity, while it recorded OR of 1.00 (95% C] 0.91-1.10) and 
1.02 (95% CI 0.93-1.12) for its association with single and HR HPV infection positivity 
respectively. However, these associations were not statistically significant. 
132 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.16: Crude odds ratio for the association between sexual characteristics and HPV 
infection positivity determined with health personnel-collected specimen 
Health Personnel-Collected Specimen, OR (95% CI) 
Sexual 
Category 
characteristics Single Multiple 
Any HPV HRHPV LRHPV 
Infection Infection 
Age at first sexual 0.99 1.02 0.93 1.00 0.96 
I 
intercourse (0.91-1.08) (0.93-1.12) (0.81-1.07) (0.91-1.10) (0.82-1.14) 
Age at first sexual < 20 Reference 
intercourse 
(categorised) 1.1 5 1.1 6 1.05 2: 20 1.00 1.1 9 (0.62-2. 11 ) (0.59-2.31 ) (0.42-2.39) (0.59-2.36) (0 .36-3 .01 ) 
No Reference 
Currently have a 
male sexual partner 0.64 Yes 0.79 0.62 0.48 2.19 
(0 .29-1.41 ) (0.32- 1.96) (0.22-1.79) (0.21-1.11) (0.28-17.37) 
Lifetime number of 1.0 1 1.03 
LNSP 1.03 0.97 
1.1 2 
sexual partners (0 .83 -1.25) (0.82-1 .29) (0.78-1.37) (0.77-1.23) (0.81-1.55) 
Reference 
Lifetime number of 1.00 0.72 
2 1.24 1.02 
0.94 
sexual partners (0.43-2.35) (0.28-1.86) (0.34-4.50) (0.39-2.64) (0.20-4.46) 
(categorised) 
1.05 0.88 1.44 0.96 1.30 
~3 0.48-2.28) (0.38-2.04) (0.45-4.68) (0.40-2.32) (0.33-5.09) 
1.00 0.99 1.00 1.00 0.98 
Sexual-age (0.97-1.02) (0.96- 1.02) (0.96-1.04) (0.97-1 .03) (0.93 -1.03) 
<11 Reference 
0.49 0.62 
11-19 0.32 0.62 0.25 
(0.22-1.07) (0.27-1.46) (0.09-1 .17) (0.26-1.48) (0.05-1.18) 
Sexual-age 
( categorised) 0.38· 0.33 20 - 29 0.42 0.51 
0.15 
(0.15-0.95) (0.11-1.02) (0.1I -1.54) (0.19-1.37) (0.02-1.24) 
1.13 1.27 
> 29 1.09 1.36 0.71 (0.45-2.81 ) (0.47-3.42) (0.33-3.65) (0.49-3.74) (0.14-3.49) 
No Reference 
STI 1.1 4 1.15 0.97 
Yes 1.32 0.78 
(0.63-2.08) (0.58-2.27) (0.42-2.24) (0.66-2.62) (0.28-2.1 8) 
No Reference 
Use condom 1.35 
Yes 1.15 1.08 0.95 1.87 
(0 .63-2.09) (0.55-2.12) (0.59-3.11 ) (0.48-1.89) (0.67-5.27) 
* the odds of the associati on was signi fica nt, since the range of95% CI excludes 1.0. 
133 
University of Ghana  http://ugspace.ug.edu.gh
These associations were determined again with AFSI categorized into two groups, younger than 
20 years and 20 years or older, with the former as the reference group. It was determined that an 
AFSI of::: 20 years as compared to at an AFSI of < 20 years recorded odds ratio slightly above 
1.00 for each category of HPV infection positivity. 
Compared to currently (as at the time of the study) not having a male sexual partner, the odds 
ratio for the association between having a male sexual partner and each of the HPV infection 
positivity was less than 1.00 except for multiple infection positivity, which was 2.19 (85% CI 
0.28-17.37); however all these OR were not statistically significant. Also, the odds ratios 
determined for the association between the LNSP and each category of the HPV positivity were 
slightly greater than 1.00 except for single HPV infection positivity, which recorded an OR of 
0.97 (95% CI 0.77-1.23); but were all not statistically significant. With having I LNSP as the 
reference group, the odds ratios determined for the association of having 2 LNSP with HR and 
multiple HPV positivity were both less than 1.00 while those with overall , LR and single HPV 
infection positivity were slightly greater than 1.00. On the other hand, the odds ratio for the 
assoc iation of having 3 or more LNSP with overall, LR and multiple HPV positivity were greater 
than 1.00. All these odds ratios for the associations with LNSP were all not significant. 
When analysed as a continuous variable, the odds ratio for the associations of the sexual age of 
the participants with their HPV infection positivity were not significant, with odds ratios 
approximately equal to 1.00. Repeating the analysis with sexual age as a categorized variable, 
with < 11 years as the reference group, it was determined that the odds ratio for the association 
of having a sexual age of between 11 and 19 years as well as having a sexual age between 20 and 
29 years with each category of HPV infection positivity were less than 1.00 and not statistically 
134 
University of Ghana  http://ugspace.ug.edu.gh
s ignificant. However, having a sexual age of more than 29 years recoded odds ratio of greater 
1.00 for its association with each category of HPY infection positivity except for the association 
w ith multiple HPY infection positivity. Compared to not having an STI, contracting an STI in 
the past 10 years was not significantly associated with overall (OR = 1.l4; 95% CI 0.63-2.08), 
H R (OR = 1.15; 95% CI 0.58-2.27 and single infection HPY positivity (OR = 1.32; 95% CI 
0.66-2.62). With respect to the use of the condom, it was determined that compared to not using 
condoms, the odds ratio for the association of condom use with each of the categories of HPY 
positivity was greater than 1.00 except for single HPY infection positivity but all were not 
statistically significant. 
4.8 HPY VARIANT ANALYSES 
4.8.1 HPY16 Variants 
Each of the 21 specimen determined by genotyping to be HPV 16 positive, gave positive results 
for the LCR HPV 16 type specific peR for sequencing. However, three (3) of the sequences were 
too short (did not include nucleotides within most of the important variable regions) for further 
a nalys is and therefore were excluded in the nucleotide alignment and phylogenetic tree 
constructed . The sequences of 19 HPV 16 isolates were aligned with the region between positions 
7469 and 7840 of the complete genome of the HPVI6 prototype NC0015626.2 (Kennedy et aI. , 
199 1; Seedorf el al. , 1985). However, the sequences of 2 of the 19 HPV 16 isolates (isolate-866 
a nd isolate-628) were shorter by 13 nucleotides (due to incomplete sequencing) and aligned to 
position 7826 of the prototype. The al ignment also included sequences of selected HPV 16 
135 
University of Ghana  http://ugspace.ug.edu.gh
variants obtained by an NCB] BLAST of the isolates' sequences. These variants were also 
included in the phylogenetic tree constructed (Figure 4.6). 
Nucleotide sequence variations were observed in 30 positions within the LCR positions 7460 -
7840. Of these 30 positions, II were observed exclusively with the HPV 16 isolates obtained in 
this study and were located with the positions 780 I to 7828 (Table 4.17). Specifically, the 
nucleotide positions 7804 and 7807 showed most of the exclusive variation among the study 
isolates; with three isolates each for each variant position. A close look at the overall nucleotide 
seq uence variations showed that for the following isolates, (of this study) isolate-937, isolate-866 
and isolate-618, the observed variations were similar to those of the HPV 16 Africa type I 
variants; specifically, the variation in the sequence of isolate-937 were similar to that of the 
HPVI6 variant BF325_Af-1 (100% coverage and id of 99%) with an additional transversion of 
an A to T at position 7836. On the other hand, isolate-866 was similar to the HPV 16 variants 
KF466576_AFRla and KF466592_AFRlb (99% coverage and id of 100%) but showed 
transversions of a T to G at position 7824 and a G to T at position 7825. Isolate-618 was similar 
to the HPVI6 variants KF466576_AFRla and KF466592_AFRlb (100% coverage and id of 
99%) but showed a transversion of a C to A at position 780 I, a T to G at position 7804 , and a 
transition of an A to G at positions 7807 and 7809. 
A Iso, the sequence variation of the isolates; isolate-500, isolate-612, and isolate-617 were similar 
to that of BF236_Af-2 (100% coverage and id of 99%) but with the following additional 
variations (Table 4. 17); a transition of G to A at position 7815 for isolate-500; another transition 
of an A to G at position 7807 for isolate-612 and a transversion of T to G at position 7804 as 
well as another transversion of C to A at position 7828 for isolate-617. The nucleotide sequence 
136 
University of Ghana  http://ugspace.ug.edu.gh
variations of the isolate-628 were similar to that of the HPVI6 variant KF466652._AFR2b 
(100% coverage and id of 99%) but with the following additional variation, a transversion of a T 
to G at positions 7804 and 7812, and a transition of an A to G at positions 7807 and 7808. The 
nucleotide sequence of isolate-999 matched exactly with that of the HPV16 variant 
KF466626_Af-2a (100% coverage and id of 100%). Also, the nucleotide sequences variations of 
isolate-759, isolate-514 and isolate-51 I were matched exactly with that of the HPVI6 variant 
Rw862_Af-2. Similarly, the nucleotide sequences of the following isolates 765, 609, 606, 603, 
602, 601,569,517 and 607 matched exactly with that ofBF236_Af-2 (100% coverage and id of 
100%). In all, 15% of the isolate were determined to be HPV16 African type 1 variants, all of 
which had additional nucleotide variations not observed in the prototype and other African type 1 
variants included in the alignment. Five percent of the isolates were HPV 16 African type 2a 
variants, while 80% of the isolates were the African type 2b variants, 25% of which had 
additional nucleotide variations not observed in both the prototype and other African type 2b 
variants included in the alignment. 
137 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.17 Nucleotide sequence variations between positions 7469 and 7840 (within the long control region (LCR)) ofHPV16 
variants. 
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 
4 4 4 5 5 5 6 6 6 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 8 8 8 8 8 8 
HPY 16 Isolates .., 
8 8 8 0 2 6 1 6 8 0 2 2 4 6 8 8 0 0 0 0 0 0 I 1 2 2 2 .) 3 3 
1 4 8 6 0 9 0 8 8 2 3 8 9 2 3 0 5 1 4 6 7 8 9 2 5 4 5 8 3 6 7 
NCOOl526 El A A G A G T A C C G T A A T C T C C T T A A A T G T G C G A A 
SiHa cells 
A 
HPV16 
W0122 E2 A C 
Qv02706_EA A C 
Rw677 AAI A A T A C T T 
Qv03545 AA2 A T A C T T 
KF466576 Afla A A T T 
KF466592 Aflb A A A T T 
Rw918 Af-l A A A A T T 
BF325 Af-l A A G A T T 
937 A A G A T TO· 
866 A A A T 11-1-
618 2 A A A T 1iJ· T 
618 A A A T T 
KF466626 Af-2a C A A T A T A T C G 
999 C A A T A T C A T C G 
Rw862 Af-2 C A A T A T T A T C G 
759 C A A T A T T A T C G 
514 C A A T A T T A T C G 
511 C A A T A T T A T C G 
138 
University of Ghana  http://ugspace.ug.edu.gh
table \7. continued 
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 
4 4 4 
Isolates 5 5 5 6 6 6 7 7 7 7 7 7 7 7 8 8 8 8 8 8 8 8 8 8 8 8 8 
8 
HPY 16 
8 8 8 0 2 6 1 6 8 0 2 2 4 6 8 8 0 0 0 0 0 0 1 I 2 2 2 '-"'  3 3 
I 4 8 6 0 9 0 8 8 2 3 8 9 2 -'"'  0 5 1 4 6 7 8 9 2 5 4 5 8 3 6 7 
NCOOl526 EI A A G A G T A e e G T A A T C T C e T T A A A T G T G e G A A 
KF466652. AFR2b C A A T A T T A T e 
628 C A A T A T T iii . ··-a iii . A -BF236 Af-2 e A A T A T T A C G 
765 e A A T A T T A C G 
609 C A A T A T T A e G 
606 C A A T A T T A e G 
603 C A A T A T T A e G 
602 e A A T A T T A C G 
601 e A A T A T T A C G 
569 C A A T A T T A C G 
517 C A A T A T T A C G 
607 e A A T A T T A C G 
800 e A A T A T T B· A C G 
612 C A A T A T T m. A C G 
617 C A A T A T T iii . A M. e G 
• Variation exclusive to isolates ofthis study. o Variations within sequences of known HPV 16 variants DSame nucleotide as 
reference isolate D deletion of nucleotide at that position 
139 
University of Ghana  http://ugspace.ug.edu.gh
-----lJ K F4':<5e-26_f-F'V 16j' J.2a 
r- 15'56 
KF4cee!·2.H F'''' 1e-,·.FP~ 
! ·14 
759 
RN OO2_HPV 6_A t-2 
&Y.l 
eo7 
517 
~e-9 
eOl 
e·Y.1 
003 
eoo 
r---- e<J6o 
7E·5 
BF235_I--FV 16_Af.2 
5 1 
,---------------------- 6~ 
e'17 
,----------&<3 
I .--____r :-:;;-----61&- 2 
10Nl 
L{ K F4CQSe·_I-PV16_AFR1a  
KF4ec=<9t_HP'I/1 o_AFR1t o 
I 18F32: .JFV 1o _A t-l 
l--------l1 937 
1----- R"~1&_I-PV1o_AJ. 1 
S3_Con t~LHF'V16 
'---------~--- NCO:>1526.2_I-FV1 E:....E 1 
l-------W·,':; 22_I-PV16_E2 
t-----i 
0.002. 
F igure 4.6 An evolutionary relationship of HPV16 isolates based on sequences between the 
genome positions 7469 and 7840 within the long control region (LCR). 
The evolutio nary history was inferred using thc Neighbor-Jo ining method (Sai tou and Nei, 1987). The optimal tree 
with the sum of branch length = 0.04 170288 is shown. The tree is drawn to sca le, with branch lengths in the same 
units as those of the evolutionary d istances used to infer the phy logeneti c tree. The evolutionary di stances were 
computed using the Max imum Composite Likelihood method (Tamura et aI. , 2004) and are in the units of the 
number of base substitutions per site. The ana lys is involved 16 nucleotide sequences. Codon positions included were 
I st+2nd+3rd+Noncoding. All positions containing gaps and missing data were eliminated. There were a total of 876 
positions in the final dataset. Evolutionary analyses were conducted in MEGA5 (Tamura et al., 20 II ). 
140 
University of Ghana  http://ugspace.ug.edu.gh
4.8.2 HPV18 Variants 
Of the 15 specimen detected as HPV18 positive by genotyping, 13 were positive by an LCR 
HPV 18 type specific PCR for sequencing. However, one of these 13 gave a poor sequence and 
therefore was excluded in the alignment and phylogenetic tree constructed. The sequences 
obtained with the other 12 isolates were aligned with the region between positions 7464 and 
7839 of the complete genome of the HPV18 Prototype AY262282_AI (Narechania et al., 2005). 
Also included in the alignment were sequences of selected HPV 18 variants, obtained by an 
NCBI BLAST of the isolates' sequences, which were also included in the phylogenetic tree 
construction (Figure. 4.7) . 
Nucleotide variations were determined in 35 positions, 13 of which were exclusively observed 
with the sequences of the study isolates. The positions that showed most of the exclusive 
variat ion among the isolates in this study were 7813 and 7799. Position 78 13 revealed variation 
in 5 isolates, 3 of which were a change to an A and the other 2 were deletions. Position 7799 on 
the other hand revealed variation in 3 isolates, 2 of which were changes to an A and the other one 
was a deletion (Table 4.18). 
Specifically, isolate-858 matched exactly with the HPV 18 prototype A Y262282 _AI (Narechania 
et al. , 2005) with no variations in all positions aligned . However, isolate-613 showed similar 
sequence variation as those ofHPVI8 A3 variant, Qv26861 but with additional variation, which 
were a transition of a C to T at position 7551, a transition of a G to A at position 7563 and a 
transversion of T to G at position 7833. The sequence variations of isolate-917 were similar to 
those of the HPV 18 B I variant Rw750 (100% coverage and id of 99%) but showed additional 
variat ions; transversion of C to G at position 7768, transition of a G to A at positions 7799 and 
141 
University of Ghana  http://ugspace.ug.edu.gh
781 3 and a transition of a T to C change at position 7814. A comparison of the nucleotide 
variation of the isolate-56 1,  isolate-679, isolate-553 and isolate-809 revealed an exact match with 
the nucleotide variations determined for the HPV 18 B2 variant, BF309. However, for the isolate-
591, although it showed a similar nucleotide variation to that of the HPV 18 B2 variant BF309 
(1 00% coverage and id of 99%), an additional transition of a G to A at position 7813 was 
determined . Also, isolate-572 showed the following additional variations compared to that of the 
HPV 18 B2 variant BF309; deletions at positions 7799 and 7813 . 
The isolates 851 and 555 showed similar nucleotide variations with that ofthe HPV18 B3 variant 
BF380 (100% coverage and id of99%); however, isolate-851 revealed an additional transversion 
of a C to A at position 7795, a transition of a G to A at position 7799 and a deletion at 7813, 
whi le isolate-555 revealed a deletion at position 7529. Isolate-627 revealed a nucleotide variation 
simi lar to that of the HPVI8 C variant Qv39775 (100% coverage and 98% identify) but with the 
follow ing additional variations; a transition of a T to C at position 7765, a transversion ofT to G 
at pos ition 7769, a transition of an A to G at position 7770, a transition of a C to T at positions 
7771 and 7773, a transversion of a T to A at position 7774, a transversion of C to G at position 
7800 and a deletion at position 7813 (Table 4.18). 
142 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.18 Nucleotide sequence variations between positions 7464 and 7839 (within the long control region (LCR)) of HPV18 
variants. 
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 
4 4 4 5 5 5 5 5 5 5 5 5 5 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 7 8 8 8 8 
HPV 18 isolates 
7 8 9 1 1 2 3 5 6 6 6 8 9 4 4 5 5 7 8 0 2 3 3 6 6 6 7 7 7 7 9 9 0 1 I 3 
5 6 6 1 2 9 0 I 3 5 7 6 2 3 5 1 8 0 9 4 6 0 6 5 8 9 0 1 3 4 5 9 0 3 4 ~ J 
AY 262282 AI T C C C G C T C G A A C T T C T A A G T C C T T C T A C C T C G C G T T 
HPVI8 ctrll 
858 
Qv29226_AI C [QJ 
Qv3298 1_A2 G A G C C 
Qv2686 1 A3 T A C C T 
613 T A 1.1 c c T iii 
Z135_A4 A C C T 
CUll AS A A A C C C T 
Rw750 BI G A A C A C C G C C T C 
917 G A A C A C C G C C T A c *i1 . ••• .... 
BF309 B2 G A A C A C C G C C T C T A 
56 1 G A A C A C C G C C T C T A 
679 G A A C A C C G C C T C T A 
553 G A A C A C C G C C T C T A 
809 G A A C A C C G C C T C T A 
591 G A A C A C C G C C T C T A 
572 G A A C A C C G C C T C T A 
BF380_B3 G A A A C C G C C T C T A 
85 1 G A A A A C C G C C T C T A 1.+1 · 
555 G A A . A C C G C C T C T A -. 
Qv39775_C G G A A C T A C C C T C T A 
627 G A A C T A C C C T C T A 61 . .. _.- II 
• Variation exclusive to isolates of this study. D variations within sequences of known HPV18 variants D Same nucleotide as reference isolate 
D deletion of nucleotide at that position 
143 
University of Ghana  http://ugspace.ug.edu.gh
Rw7!:O I-F V 18 B 
L..-____ S17 
679 
r-----II-- 8.:.11 
B 8:l<9_ I-FV t:IU 32 
r------e.t~. 
r---- Q'Iff.s6 ~ _I-FV1 8.-";3 
Zl 35·_I-FV la_A4 
- ....-----Q'.G2S<&1 _I-FV ' 8_A2 
Q •. z::t:ne._ H FV 18_ A' 
AY2f/1182.1_HFVl2,_A 1 
L-_~HP''''lJ 8_OOI1 t rol 
Figure 4.7 An evolutionary relationship of the taxa of HPV18 isolates based on sequences 
between the genome positions 7464 and 7839 within the long control region (LCR.). 
T he evol utionary history was inferred using the Neighbor-Joining method (Sa itou and Nei, 1987). The optimal tree 
w ith the sum of branch length = 0.04170288 is shown. The tree is drawn to scale, wi th branch lengths in the same 
units as those of the evolutionary di stances used to infe r the phylogenetic tree. The evolutionary di stances were 
computed usi ng the Maximum Composite Likelihood method (Tamura el a/. , 2004) and are in the units of the 
number of base substitut ions per site. The analysis invo lved 16 nucleotide sequences. Codon positions included were 
I st+2nd+ 3rd+Noncoding. All positions containing gaps and missing data were eliminated. There were a total of 876 
positions in the fi nal dataset. Evolutionary analyses were conducted in MEGA5 (Tamura el a/. , 2011) . 
144 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.19 Nucleotide sequence variations between positions 7074 and 7858 (within the long control region (LCR)) of HPV45 
variants. 
7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 
HPV45 1 1 1 1 1 2 ... .J 3 3 3 4 4 4 4 5 5 5 5 5 6 6 6 6 7 7 7 7 7 7 7 8 
Isolates 1 4 6 6 7 7 2 2 6 8 0 2 2 5 0 0 2 5 6 0 2 6 9 ... ... .J .J 4 7 7 7 7 5 1 2 7 9 
4 3 2 9 ... I 4 0 9 2 I ... 3 4 6 7 7 3 9 9 0 5 4 9 ... 0 ... ... .J .J .J .J 5 9 4 6 .J 8 7 
X74479 AI C G T A T G C A A G T TAT A G C G C G TAT G G C T T A A T T C C A 
820 
460 
583 
632 
831 
880 
1025 
512 •• • 
Qv20214_Al C 
444 C . iii 
Z79 Al rcl 
BF208 A2 @j . T . A L£J . . . . . . l Oj . l£j . . C . A l..Q 
BF134 A3 A . C . @] . .. . ~ A . ~ .. m . Em ... C l]] A 
Qv06560_BI A@] . G A AC . AC .. [Q] . T T. A . . . GA m .. .. G [Q] . . A 
Qv34163 B2 A. Ac rcl AC . IOl T T. A. FC1 . . A ... . . G G A 
• variations exclusive to isolates of this study. o variations within sequences of known HPV45 variants 
DSame nucleotide as reference isolate D deletion of nucleotide at that position 
146 
University of Ghana  http://ugspace.ug.edu.gh
4.8.3 HPV45 Variants 
Of the 18 specimen that were determined to be HPV45 positive by genotyping, 13 gave 
amplification products with a three-primer overlapping LCR HPV 45 specific PCR for 
sequencing. The sequences of 4 of the 13 isolates were determined to have been too short, 
therefore were excluded in the alignment and construction of the phylogenetic tree. The other 9 
sequences were aligned with the genome of the HPV45 prototype X74479.l (Delius and 
Hofmann, 1994) from position 7074 through 7858/1 to 102. Also included in the alignment were 
sequences of selected HPV45 variants obtained by an NCBI BLAST of the isolates' sequences, 
which were also included in the phylogenetic tree construction (Figure 4.8). 
Nucleotide sequence variations were determined in 35 positions, 2 of which were observed 
exclusively with the sequences of the study isolates. Position 97 showed the most variations 
among the isolates in this study; in six isolates 460, 583, 632, 831, 880, 1025 with the nucleotide 
change of A to G as compared to the prototype HPV45 variant. Isolate-820 matched exactly with 
the prototype HPV 45 variant while isolate 512 revealed a nucleotide variation of a transition of a 
C to T at position 7114. The isolate-444 revealed a similar nucleotide variation with the HPV45 
A 1 variant, Qv20214, but with a transition of an A to G at position 97 (Table 4.19). 
145 
University of Ghana  http://ugspace.ug.edu.gh
r--- 512 
-
L--________________ BFl l4_fTV45j.3 
r----------------- Q·£6 .:e':U"FV45_B 
I 
0 .. 002: 
Figure 4.8 An evolutionary relationship of HPV45 isolates based on sequences between the 
genome positions 7074 and 7858 within the long control region (LCR.). 
The evo lutionary hi story was inferred using the Neighbor-Join ing method (Saitou and Nei, 1987). The optimal tree 
with the sum o f branch length = 0.04170288 is shown. The tree is drawn to sca le, with branch lengths in the same 
units as those of the evolutionary distances used to infer the phy logenetic tree. The evolutionary di stances were 
computed using the Max imum Composi te Likelihood method (Tamura el aI. , 2004) and are in the units of the 
number of base substitutions per site. The ana lysis involved 16 nucleotide sequences. Codon positions included were 
1s t+2nd+ 3rd+Noncoding. All positions contai ning gaps and missing data were eliminated. There were a total of 876 
posit ions in the fina l dataset. Evolutionary analyses were conducted in MEGA5 (Tamura el aI. , 2011 ). 
147 
University of Ghana  http://ugspace.ug.edu.gh
4.9 CERVICAL LESIONS AND ABNORMALITIES 
Of the 251 microscopic slides prepared and evaluated for malignant lesions, 212 (84.4%) had 
satisfactory smears while 39 (15.6%) had unsatisfactory smears for evaluation (Table 4.20). 
There was no statistically significant association between the adequacy of the smear and the site 
or location of specimen collection and smear preparation (Table 4 .20). 
Table 4.20: Distribution and association between the location of specimen collection and 
the adequacy of the smears for evaluation 
Adequacy of smear, n(%) Chi-Square 
Specimen collection point 
Unsatisfactory Satisfactory Total (p value) 
Hospital 28 (18.4) 124 (81.6) 152 (60.4) 
Community 88 (88.9) 99 (39.6) 2.092 II (11.1) 
(0.148) 
Total 39(15.5) 212 (84.5) 251 
All the satisfactory smears evaluated (n = 212) were negative for cervical malignant lesion and 
showed no dyskaryosis and intraepithelial lesion. However, 19.5% (n = 41) of them were 
diagnosed with other conditions, the commonest of which was Candida infections (Table 4 .2 1). 
148 
University of Ghana  http://ugspace.ug.edu.gh
Table 4.21: Distribution of diagnosis based on Pap smear evaluation 
Diagnosis Frequency Percentage 
Dyskaryosis 0 0.0 
Acute vaginitis 4 2.1 
Atrophic cervicitis 9 4.2 
Bacterial vaginosis 15 7.4 
Candida infection 13 6.3 
No Abnormality 170 80.0 
detected 
Total 212 100.0 
149 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER FIVE 
DISCUSSION 
5.1 P ARTICIP ANT CHARACTERISTICS 
5.1.1 Distributions demographic Characteristics 
Although the data from the 38 peer-recruited women were potentially going to bias the findings 
of the study, an analysis of the demographic, reproductive and sexual characteristics of the 38 
women indicated that they were not different from the women recruited by the study (Table 4.1). 
Also, when these data were stratified by the reporting strategies, the exclusion or inclusion of the 
38 women did not make any statistical difference. Therefore, demographic, reproductive and 
sexual characteristics of the 38 women were included in the analysis of the overall distributions 
of the demographic, reproductive and sexual characteristics, although the distribution for each 
strategy was also indicated. However, these 38 women were excluded from the analysis of each 
of the individual reporting strategies since these women did not report by any of them (Figure 
4. 1). That is, they did not opt for a reporting time for which it was to be assessed whether they 
reported on time or not. 
T herefore, considering the distribution of the demographic data, the finding that the distribution 
of the religious status of the participants was not uniform among the different reporting strategies 
(Table 4.1) is explainable by the following. These are that there was only one community, known 
a s the Zongo community, where almost all the Muslims in the sub-district were resident. This 
was the first community the study engaged its participants, and with the initial approach of the 
150 
University of Ghana  http://ugspace.ug.edu.gh
study, which recorded a low response rate. No community was approached with more than one 
strategy and therefore, the Zongo community was not approached with any of the other 
strategies. Therefore, the difference in the distribution of the religious status of the women was a 
reflection ofthe inherent non-uniform distribution of Muslims in the sub-district. 
5.1.2 Distribution of Sexual Risk Characteristics 
As indicated in sections 1.3 and 2.5, for a relative small cross-sectional study, as the one reported 
in this thesis, the distribution of the categories of the known risk factors were used as indicators 
of the potential of detecting an HPY infection, since the association of these characteristics and 
H PY infection among the participants in this study could not be appropriately assessed due to the 
re latively small sample size (this study was not powered to determine these associations) . 
T herefore for this study, the question was whether the Akuse sub-district could be said to be at 
high risk for HPY infection, based on the distribution of the sub-categories of each of the known 
sexua l risk factors for HPY infection. If so, will this high risk status be confirmed by the 
prevalence of HPV infection determined in this study? This initial supposition that the Akuse 
sub-district was a high at risk community for the acquisition of HPY infection (section 3.1 -3.2) 
was confirmed by the findings of the study as has been presented in table 4.2. That is, as 
indicated in section 2.5.1.2, AFSI of ~18 years is high ri sk for HPY infection (Erickson et al., 
2 0 13; Luhn et aI. , 2013 ; Bahmanyar et aI. , 2012; The International Collaboration of 
Epidemiological Studies of Cervical Cancer, 2007; Berrington de Gonzalez et aI. , 2004) and 
si nce slightly more than half of the participants of this study, 52.4%, were within this category, 
the sub-district was likely to have slightly more women who are infected with HPY. That is, it 
was a moderate at risk community for HPY infection. Similarly, and as indicated in section 
151 
University of Ghana  http://ugspace.ug.edu.gh
2.5.1.2, the sexual age (the difference between the current age and AFSJ) characteristic' s 
categories of < 11 years and between 11 years and 19 years are high risks for HPV positivity 
(Bahmanyar et al., 2012; Plummer et al., 2012; Lenselink et aI. , 2008). Therefore, the finding 
that 68.7% of the participants of this study were within these sub-categories (Table 4.2) 
suggested that most of the women in the sub-district were more likely to be HPV positive. In 
other words the sub-district was a moderate to high at risk community for HPV infection. 
Furthermore, the finding that almost half(49.1%) of the women in this study were within the ~ 3 
sub-category of lifetime number of male sexual partners, the high risk sub-category for HPV 
positivity (Erickson et aI., 2013; Luhn el af., 2013; Bahmanyar et al., 2012; Almonte el af., 2011 ; 
International Collaboration of Epidemiological Studies of Cervical Cancer, 2007; Shields et aI. , 
2004), further suggest that the sub-district was a moderate at risk community for HPV positivity. 
This suggested moderate to high risk for HPV infection among the women in this sub-district 
was further supported by the findings that all the women were not consistent users of condom 
although 41.3% stated they had ever used condom. This was so because, it is the consistent use 
of condom that protects against HPV infection (Burchell el al., 2006; Winer el aI. , 2006), as 
indicated in section 2.5.1.4. Additionally, the finding that 53.6% of the women had had an STJ 
infection within the past 5-\0 years and that 97.7% of them were infected with Candida provide 
fLirther supportive evidence for a high potential of HPY infection among the women, since a 
history of STJ and an infection with Candida are high risk for HPV positivity (Burchell et ai., 
2006; Sam off, 2005). Overall, the distribution of the categories of sexual risk characteristics 
suggested a moderate to high potential of HPY positivity among the women in the sub-district. 
This suggested high potential was confirmed by the high prevalence of HPV infections detected 
among these women as shown in table 4.14. 
152 
University of Ghana  http://ugspace.ug.edu.gh
The finding that only one woman reported living with HIV, was far below expectations, since the 
national HIV sentinel site closest to the sub-district (Agormanya) had reported the highest HIV 
prevalence (7.8%) in Ghana between the year 2003 and 2011 (within the period this study was 
designed) . Although some participants may not have reported their correct HIV status, this non-
report was not likely the only reason for the very low reported HIV status in this study. The fact 
that the sentinel site closer to the sub-district reported high HIV prevalence may not necessarily 
mean that every community within its catchment area will have a high HIV prevalence. 
5.1.3 Distribution of Reproductive Risk Characteristics 
The assessment of the persistence of high risk HPV infections and the exposure to the 
reproductive risk characteristics are the most useful indicated of the potential of developing 
cervical lesion (Vaccarella et aI., 2006; Bosch et aI., 2008; Koshiol et aI. , 2008). However, in the 
absence of data on HPV persistence, determining the proportion of the women at risk of 
developing pre-cancer lesions based on the distribution of the categories of the reproductive 
characteristics is considered useful in this study (lnternational Collaboration of Epidemiological 
Studies of Cervical Cancer, 2007; Vaccarella et at., 2006). To this end the sub-district was most 
likely a low-at-risk community for cervical lesion development. However, although the study 
was not powered enough for the estimation of the prevalence of cervical lesions, the non-
detection of cervical lesions among the participants of this study, may be considered a 
confirmation of this most likely low-at-risk status (a further discussion on the non-detection of 
cervical lesion is presented subsequently) (Table 4.20). This deduction is supported by the 
distribution shown in table 4.3 and the discussions in section 2.5.2.2. Specifically, a history of 
oral contraception (OC) usage for 5 or more years is a high risk factor for the development and 
153 
University of Ghana  http://ugspace.ug.edu.gh
progression of cervical lesions (Chelimo et aI. , 2013; Luhn et al. , 2013 ; Almonte et aI. , 2011; 
T he International Collaboration of Epidemiological Studies of Cervical Cancer, 2007; 
Castellsague and Munoz, 2003b; Smith et al., 2003) and in this study, it was determined that 
23 .3% of the women had ever used OC in the past 10 years, and only 7.7% had been using it for 
5 or more years. These imply that only a small proportion of the women were at risk of 
developing cervical lesions and therefore, suggested that the sub-district was a community at low 
risk for cervical lesion. Furthermore, and as indicated in section 2.5.2.1 , multiparty of more than 
3 is a high risk characteristic for cervical lesion positivity (Chelimo el at. , 2013; Luhn et at. , 
20 13; Almonte et al., 2011; The International Collaboration of Epidemiological Studies of 
Cervical Cancer, 2007; Castellsague and Munoz, 2003b), therefore, the finding of this study, 
which showed that 38.3% of the women had been pregnant for 5 or more times suggested that a 
small proportion of the women were likely to develop cervical lesion. Overall, the distribution of 
the sub-categories of reproductive characteristics were all suggesting that it was likely to detect 
cervical lesion among a small proportion of the women. However, the distribution of these 
categories of the reproductive risk characteristics, as indicated in sections 2.5 .2 and as discussed 
for the sexual risk characteristics, was used as an indicator of the potential for cervical lesion 
pos itivity am ong the women, since the study was not powered to detect and quantify an 
assoc iation between the risk factors and cervical lesion positivity. 
T he other reproductive risk factors as shown in table 4.3 ; ever had an induced abortion, number 
of induced abortions, ever had a miscarriage and the number of miscarriage, have been shown 
not to be significant risk factors for high grade cervical lesion or cancer in some studies (Deacon 
et al. , 2000) but significant in others (Bahmanyar ef al., 2012). However, the observed 
154 
University of Ghana  http://ugspace.ug.edu.gh
distribution of these risk factors revealed that more than half of the participants had not 
experienced both induced abortions and miscarriage (51.3% and 69.0% respectively) and among 
those who had experienced them, most did, only once (58.4% for induced abortion and 62.6% 
for miscarriage). Therefore, the community herein studied was suggested to be having most of 
the women in a low to moderate risk category for cervical lesions. 
In summary and in respect of the first objective of this study, a moderate to high proportion of 
the women were potentially at risk for HPY positivity, but a small proportion of them were 
potentially at risk for cervical lesion positivity. This seemingly contradictory deductions are 
expla inable by the following facts; that the sample size for this study was not calculated with an 
expected prevalence of cervical lesion but rather with an expected HPY prevalence (that is, the 
study was powered only for the detection of HPY prevalence). This implies that potentially the 
distribution of factors associated with cervical lesions in this study may have been lower then 
what it may have been, since the sample size was smaller than it would have been if the study 
had been powered for the detection of cervical lesion. Additionally, it must be noted that it is 
persistent HPY infection that leads to the development of cervical lesion, not just HPY infection. 
The persistence of the HPY infections was not estimable with these risk factors . Furthermore, the 
possi bility of lower detection related to cytology testing in Ghana, as indicated by a study in 
Ghana (Handlogten et aI., 2014) may have contribute to the low detection in this study. 
5.1.2 Participant Characteristics that may have influenced Condom Use 
As indicated in section 2.5.1.4, the use of condom is a very important primary prevention method 
for STIs in general (WHO; Guideline Development Group, 2014 ; WHO; Chronic Diseases and 
Health Promotion Group, 2006) and for the prevention of HPY infection in particular, it is the 
155 
University of Ghana  http://ugspace.ug.edu.gh
consistent use of condoms that is important (WHO; Guideline Development Group, 2014; WHO; 
Chronic Diseases and Health Promotion Group, 2006; Shields el aI. , 2004; Shepherd el aI., 
2000). Therefore, it is important to determine the demographic characteristics of the participants 
that may have influenced the non-use of condom for the purposes of tailoring education on 
condom use. It can been seen from table 4.7 that to a large extent the expected distribution and 
association (Domfeh et al., 2008) with the use of condom were observed. That is, more young 
women (less than 29 years) than expected used condom, less women older than 30 years and less 
married women than expected used condom. Additionally, table 4.8 shows that fewer women 
than expected with sexual ages higher than 19 years (corresponding to women who were 38 
years and older) and those with one (1) current male sexual partner used condoms. These are 
understandable, in that most women 30 years and older were likely to be married and/or starting 
a family and most married women are less likely to use condom, expect for the few whose 
husbands (in this male dominated society) may agree to the use of condom as a contraceptive. In 
spite of these expected findings, the finding that the use of condom was not consistent was very 
important to cervical cancer prevention, since it is the consistent use of condom that reduces the 
risk of acquiring an HPY infection . Therefore, there is the need for further studies to assess 
whether the inconsistent use of condom was fuelled by limited knowledge on condom use, 
cultural and/or religious limitations, so that more education and other appropriate cultural and 
religion sensitive intervention may be implemented as part of a cervical cancer screening 
programme in the Akuse sub-district and potentially in Ghana. This will encourage the consistent 
use of condom, particularly among men, since in this society, the use of the male condom 
predominates and women often may not have the fina l decision on its use. 
156 
University of Ghana  http://ugspace.ug.edu.gh
5.2 RESPONSE TO THE REPORTING STRATEGIES 
5.2.1 Response Rates 
As discussed in section 4.2 .1, the 38 peer-recruited women were excluded from the related 
analyses, since they did not fit in any of the three reporting strategies used and would have 
biased the estimated response rate. Therefore, the findings show that the overall response rate 
attained in this study (60.4%), was determined with the 377 study recruited participants, and that 
this rate was relatively high. Considering each of the three reporting strategies of this study, 
w hich were, long appointment time (duration) to report to hospital, short appointment time 
(duration) to report to hospital and report at a location within a community), the response rate for 
the strategy requiring the women to report to a location within the community for specimen 
collection (strategy 2), which was 95.1 % (Figure 4.1), implied a very high potential for being an 
effective strategy for improving response to cervical cancer screening activities. However, it is 
worth noting that by the nature of the implementation of this community reporting strategies, the 
participants had a short duration within which to report for specimen collection after completing 
the study questionnaire. Among other reasons, this probably made it easier for that large 
proportion of the participants to follow through to the end of the screening. The other reasons 
include the fact that the participants were allowed to either report in the morning (on their way to 
their regular daily activity) or later in the day (up to 7:00 pm) when returning from their daily 
activities, without it interfering with their regular activities. 
On the other hand, the response rates obtained for the strategy I, short duration and long 
duration, which were 46.6% and 38.5%, respectively, indicated that requesting the participants to 
157 
University of Ghana  http://ugspace.ug.edu.gh
report to a hospital within a short or long duration of time (after a home based recruitment 
process and without providing incentives for the participants), holds a low potential for recording 
high cervical screening coverage in a hospital setting. This is because these response rates were 
lower than the lower limit ofthe global range of response rate for cervical screening, which have 
been reported to be between 48% - 88% (Rossi et al., 2011; Virtanen et aI., 2011; Gok et aI., 
20 I 0; Sanner et al., 2009). These response rates obtained with the strategy 1 suggest that such a 
strategy might not be helpful as a cervical cancer screening strategy for Ghana. Therefore, for the 
planning of cervical cancer prevention and control by screening in Ghana, a strategy that will 
make it possible for women to access screening within their community (particularly at CHPS 
compounds) and at times that will not interfere with their day's activities, such as that used in 
this study, and that the appropriate training is given to the health workers at these CHPS 
compounds, should be given a strong consideration, since its response rate was higher than the 
upper limit of the global range. 
Compared to the response rates reported by recent studies of cervical screening In different 
geographical regions, 48% - 88% (Virtanen et al., 2011; Rossi et at. , 2011; Ook et al., 2010; 
Sanner et al. , 2009), the overall response rate attained in this study was considered relatively 
moderate. This is considered a good response rate because although a similar response rate 
(60.0%) was observed in a similarly designed cervical cancer screening study in the USA, where 
a home-based recruitment was followed with specimen collection (Castle et aI. , 2011 b; Scarinci 
et at., 2010b), the study in the USA provided financial incentive to achieve this high response 
rate, while the study reported in this thesis did not provide any incentive. Incentives were not 
prov ided in the study reported in this thesis, so as to mimic a real world situation in order to 
158 
University of Ghana  http://ugspace.ug.edu.gh
assess the likelihood of a woman, living within these communities, to respond to a regular 
cervical cancer screening activity. Furthermore, although the study in the USA did not indicate 
the duration within which the participants had to report to the hospital after recruitment, the 
results of the study reported in this thesis showed that this duration was very important in 
achieving the high response rate (coverage). This deduction was based on the fact that a 
re latively poor response rate (38.5%) was recorded for the strategy that gave the women a longer 
duration (over a month) within which to report to the hospital after recruitment. 
Furthermore, the determined response rates after calling to remind participant who had not 
reported, 10.3 and 44.4%, suggested that encouraging non-responders to participate in cervical 
screening activities by re-inviting them with a phone call may only contribute marginally to 
improving response rate or coverage. However, this may perform better than re-inviting non-
responders by the sending of letters through the mail because of the less efficient postal system 
and low literacy rate in such communities in Ghana. On the other hand, a study by Rossi et aI., 
(20 II) suggested that re-inviting non-responders with the option to perform the specimen 
collection at home potentially may result in a greater response. This was also emphasized in a 
review of studi es that employed se lf-specimen collection (Snijders et aI. , 2013) . However, since 
the study reported in thi s thesis was not specifically powered to determine the differences in 
these strategies and their respective reporting rates, the observed difference, although significant, 
are only suggestive of the potential influence these strategies may have on the coverage of 
community based cervical cancer screening activities . 
159 
University of Ghana  http://ugspace.ug.edu.gh
5.2.2 Participant Characteristics that may have Influenced Reporting 
In spite of the fact that the study was not powered to compare the response rates, it was 
necessary to determine if any of the participants ' characteristics may have influenced their 
response to each of these strategies (Agurto et al. , 2005). This will be useful in planning a 
cervical cancer prevention and control programme for identifying the group of women to be 
targeted with appropriately tailored education to encourage their participation in cervical cancer 
screening. The association of the demographic characteristics with each of the reporting 
strategies indicated that reporting to a location within the community and reporting after a long 
d uration to the hospital for specimen collection were not influenced by any of the demographic 
characteristics of the participants (all the chi-square p values were > 0.05; Tables 4.4 and 4.6). 
O n the other hand, for the strategy of a short duration within which to report to the hospital , the 
results suggested a possible influence by age, where women between the ages of 20 to 39 years 
reported the most (chi-square p values< 0.05; Tables 4.5). This age related reporting may have 
resulted from the fact that most of the older women were engaged in activities, including those of 
their occupation, and could not make time to visit the hospital during the working hour of a day. 
T hi s may imply that in developing a cervical cancer prevention and control programme for 
Ghana that will involve collection of specimen at the hospital , the planning should consider 
tai loring the reporting time to suit the social and economic activities the women may be engaged 
in, to encourage them to report for cervical cancer screening within a short time after 
recruitment. 
160 
University of Ghana  http://ugspace.ug.edu.gh
5.3 ADDITONAL FINDINGS - SPECIMEN COLLECTION 
5.3.1 Performance and Preference for Specimen Collection 
Although not a specific objective of this study, it is often useful to describe the performance and 
preferences for specimen collection, if it is possible; these form one of the very useful 
information for planning screening activity. The findings of this study that showed that a high 
proportion of the participants, 89.7% (n = 226), opted for both specimen collection methods 
(self-collection and health personnel-collection), was indicative of a high acceptance of both 
specimen collection methods among women in the Akuse sub-district. However, the 7 women 
who did not opt for self-specimen collection did so because one was blind and the others were 
afraid of hurting themselves while using the self-specimen collection brush. On the other hand, 
the 18 women who did not opt for the health-personnel specimen collection indicated that fear 
and difficulties with the use of the speculum were the major reasons for their choice. The 
proportion of the women who performed self-collection (96.4%) and health-personnel specimen 
collection (90.9%) in this study, were slightly higher than those reported for other studies and 
therefore, confirms the high level of acceptance of both methods by the women. For some these 
reported studies the following were noted; a cervical screening campaign in Cameroon reported 
29.0% and 62.0% performed self-collection and performed health personnel collection 
respectively (Berner el a!. , 2013); a study in the Mississippi Delta region of USA reported 80.5% 
and 40.5% of the women performed self-specimen and health personnel specimen collection 
respectively (Castle et a!. , 2011 b) ; a randomised control trial of Mexican women reported 98.0% 
and 87.0%, performed self-specimen and health personnel specimen collection respectively 
(Lazcano-Ponce et aI. , 2011); a study in Uganda among women who did not have access to 
regular screening services reported 80.6% were willing to perform self-collection (Mitchell el 
161 
University of Ghana  http://ugspace.ug.edu.gh
af., 2011). Some other studies, of women who were non-attendees of regular screening 
programmes, have reported slightly lower self-collection performance rates of 64.9% and 39.1 % 
(Virtanen et af., 2011; Gok et af., 2010; Sanner et af., 2009). 
Despite these high performance rates for both methods in this study, the participants clearly 
indicated a higher future preference (50.6%) for health personnel specimen collection and a 
much lower future preference (20.3%) for self-specimen collection, after they had experienced 
both methods. This finding confirm reports of some studies in literature, which showed a similar 
order of preference, but there are other reports by studies that show the converse (Schmeink el 
al., 2011). For example, in a study of women in an urban community, 68.0% preferred health 
personnel collection while 32.0% preferred self-collection, after they had experienced both 
methods (Anhang et al., 2005). 
It is worth noting that differences in study design (that whether a pre-performance or a post-
performance preference was determined) may influence the distribution of preference of 
partic ipants. While the study reported in this thesis and that by Anhang et af. , (2005) showed a 
higher post-performance preference for health personnel specimen collection (discussed above), 
a study of African-American women, who were made to indicate their preference before 
performing one of the two methods, reported a higher pre-performance preference (64.7%) for 
self-collection (Castle el aI. , 20 11 b), suggesting that what the participant may expect before they 
perform the specimen collection might not be what they experience having performed the 
specimen collection. 
162 
University of Ghana  http://ugspace.ug.edu.gh
5.3.2 Reasons for and characteristics that may have influenced Preferences 
The participants of the study reported in this thesis who preferred health personnel-collection 
stated that the health personnel had perfonned the specimen collection satisfactorily because they 
were professionals and experienced. The participants also stated that they were not confident 
about whether they used the self-collection device correctly. These and most of the reasons 
stated by the participants of this study, for their preferences, have been reported by other 
participants who preferred health personnel specimen collection in other studies (Berner et aI. , 
20 13 ; Scarinci el aI. , 2013 ; Petignat and Vassilakos, 2012; Cuzick el al. , 2012; Schmeink el aI. , 
20 11 ; Mitchell et aI. , 2011). On the other hand, the participants who preferred self-specimen 
collection stated that the use of the speculum during the health-personnel specimen collection 
produced slight pain and/or induced fear, the self-specimen collection process was simple and 
easier to perform and that concerns about privacy formed the bases for their preference. These 
reasons were similar to those reported in some other studies by participants who preferred self-
specimen collection (Scarinci et aI., 2013 ; Schmeink el al., 2011 ; Rossi et aI. , 2011). However, it 
was interesting to note that some of the participants were not confident in their performance of 
the self-specimen collection, although they stated that it was simple and easy for them to use . 
The inconsistency between performance and preference suggests the need for educating 
parti cipants on the peculiarity and usefulness of self-specimen collection in order to improve its 
acceptability as part of cervical cancer screening activities. This education should be directed at 
boosting their confidence regarding their performance of the self-specimen collection process. 
It was also determined that the preferences for the collection methods were not associated with 
and therefore were not influenced by the demographics of the participants (p values > 0.05). It 
163 
University of Ghana  http://ugspace.ug.edu.gh
has been shown by other studies that often the demographic of the participants are not 
sign ificantly associated with preference for specimen collection (Gok et aI., 2010). In one of 
such studies, only older age was associated with lesser likelihood to prefer self-collection 
(Mitchell et al. , 2011) . In another study, only higher educational status and ethnicity were 
associated with a higher likelihood to prefer self-collection (Anhang et aI. , 2005). However, the 
participants who had ever had an abortion and those who had ever experienced menstrual 
difficulties were significantly more likely to prefer health personnel-specimen collection (p < 
0.05). This may be so because these women were more likely to either have had a speculum 
examination or a similar medical examination in relation to their reproductive health and they 
therefore were more comfortable with the speculum examination. 
5.4 HPV BURDEN WITH SC AND HPC 
5.4.1 Differences in HPV Prevalences of SC and HPC 
In order to answer the objectives related to how different the HPV prevalences and genotype 
detection obtained with self-collected specimen (SC) was from those obtained with health-
personnel collected specimen (HPC), this cross-sectional study of women aged between 15 years 
and 70 years assessed the distribution of HPV genotypes using the PGMY -09/ 11 PCR and a 
microsphere-based method for the detection and typing of 46 mucosal-associated human 
papillomavirus type (Zubach et ai. , 2011). The findings of the study showed a clear and often 
significate difference between the HPV data obtained with these methods, such that SC HPV 
prevalences were all higher than those of HPC HPV prevalences (Table 4.13). The overall HPV 
detected with self-collected specimen, 43 .1% , was almost twice that obtained with the cervical 
164 
University of Ghana  http://ugspace.ug.edu.gh
swab specimen collected by the health personnel. Similarly, the LR-HPV prevalence obtained 
with self-collected specimen (23.3%) was more than twice that obtained with the health-
personnel collected specimen (9.9%). However, although it is often the case that the high risk 
HPV prevalence obtained with HPC is not significantly different (very comparable) from those 
obtained with SC (Snijders el aI., 2013; Zhao el ai., 2012; Petignat and Vassilakos, 2012; 
Schmeink et aI., 2011 ; Ogilvie et aI., 2010; Petignat et aI., 2007), the prevalence of HR HPV 
obtained with the self-collected specimen in this study (27.2%; 95% CI 25 .0 - 34.0) was 
relatively much and significantly higher than that obtained with the health-personnel collected 
specimen (16.6%; 95% Cl 14.0 -24.0). This finding apart from confirming the widely drawn 
conclusion that self-collection is an appropriated alternative for health-personnel specimen 
collection for HPV testing, leaves us with the following questions; 1) Is it the case among the 
women in these communities that vaginal HR HPV infection is higher than cervical HR HPV 
infections? 2) Is it the case that the health-personnel specimen collection was not able to achieve 
the appropriate performance as compared to the self-collection? Nonetheless, within the limits of 
the performance of these collection methods in this study, significant and higher prevalence of 
HR HPV with the SC gives a better indication of the potential risk status of the women than HPC 
in respect of the burden of HR HPV. This is because although the SC HR HPV infection may be 
in the vagina, such infections may be easily transferred to the cervix when the vagina shads the 
viruses as part of the viral life cycle (WHO; Guideline Development Group, 2014; Bosch et aI. , 
2008), since the external os is exposed to vaginal secretion (WHO; Chronic Diseases and Health 
Promotion Group, 2006; Sankaranarayanan eL aI., 2003). 
165 
University of Ghana  http://ugspace.ug.edu.gh
T he spectrum of HPV detectable with the assay used for this assay, 46 of the 54 genital types 
(Schmitt et 01., 2013; Zubach et aI., 2011) included all the known high risk (HR), probable high 
r isk (PHR) and low risk (LR) HPV types. This implies that the prevalence of HPV reported by 
this study included as many HPV types as possible and therefore gave a good estimate of the 
HPY burden in the Akuse sub-district. Most studies, particularly studies in Ghana, have reported 
HPV data based on detection methods that were/are limited in the number of HPV types they 
cou ld detect (Attoh e/ 01., 2010; Domfeh e/ 01., 2008) and limited in the general detection of 
HPV. This is because, often the general HPY first round primers used in the nested-multiplex 
PCR do show the desired band on the agarose gel detection system, but amplifies the specific 
HPY genotype if present, during the second round of the nested PCR during which the specific 
primers are used (Awua et 01.,2016). Therefore, it is possible that some HPV genotypes that do 
not have their specific primers used in the second round will not be detected. At most, 37 HPV 
types are targeted with most of these methods (Schmitt e/ 01., 2013; Koidl e/ 01. , 2008; Liu et al., 
2009). The wider spectrum of detectability in this study also ensures that the epidemiological, 
clin ical and molecular information of the low risk types do not become limiting to future 
understanding of the role and related changes of all types ofHPV in cervical cancers in Ghana. 
Compared to the HPV prevalence determined for most of Ghana 's neighbouring West African 
countries, in community based studies that used cervical swabs specimen collected by health 
personnel , the 23.3% HPV prevalence obtained with the cervical swab in this study was found to 
be lower. For instance, a community based cross-sectional screening study in Benin, reported an 
HPV prevalence of33.2% (Piras el 01.,2011), while a larger community study in Burkina-Faso 
reported an HPV prevalence of 54.0% among the HIV negative women in the study (Didelot-
166 
University of Ghana  http://ugspace.ug.edu.gh
Rousseau et al., 2006). In a community-based study in Nigeria, an HPY prevalence of 26.3% 
was reported (Thomas et aI., 2004) and in a case-control study in Abidjan, Cote d ' Ivoire, HPY 
prevalence among the population control was 31.1 % (Adjorlolo-lohnson e/ aI. , 2010). However, 
the HPY prevalence reported for a rural Gambia community, 13.4% (Wall et aI. , 2005) and in a 
hospital based study in Ghana, 10.7% (Domfeh et a/., 2008) were lower than that determined in 
this study. The HPY prevalence reported in this study was similar to that reported (21.5%) in a 
meta-analysis of HPY infection among women with normal cytology for the West African 
region, which used 8 studies conducted in Senegal, Nigeria, Guinea, Gambia and Cote d'Ivoire 
(Bruni el al., 20]0). Compared to HPV prevalence reported for some communities in countries 
outside Africa, such as China, the HPY prevalence determined in this study was higher than 
those (13.3% and ]4.3%) reported by two community-based studies (Wu et aI., 2013; Ye e/ aI., 
20] 0). Therefore, the results of this study adds to the existing data suggesting the possible 
existence of geographical differences in HPY prevalence among countries of the West African 
region, although differences in study design (selection of participants), age range of women (for 
one study), methods ofHPY detection and genotyping, and the HIV status of the participant have 
been shown to also contribute marginally to differences in HPV prevalence (Martin et aI. , 2011 ; 
Conesa-Zamora el aI. , 2009). 
A similar pattern of difference were reported in a study of South African adolescents (Adler et 
af. , 2012). However, this higher prevalence with self-collected specimen was explainable with 
the fact that self-collection samples were more likely to have included the vulva, vagina and the 
cervix during the sampling process and therefore more likely to detect more HPY infections than 
the health-personnel collection specimen, which involved only the cervix (Gravitt et al., 2011). 
167 
University of Ghana  http://ugspace.ug.edu.gh
Also, the large difference in the LR HPV prevalence was explainable by the fact that these HPV 
types were more associated with the vaginal epithelium and as such most studies have detected 
more LR HPV with SC specimen (Gravitt et al., 2011; Moscicki et al., 2010). Although HR 
HPVs have been associated with both cervical and vaginal epithelium and they have been 
detected with closer frequencies between SC and HPC specimen (Schmeink et al., 2011), It has 
been shown that if the HPV prevalences were higher for HPC than SC, it meant that the SC were 
not well or were sub-optimally performed (Gage et af., 2011; Gravitt et af., 2011 ; BhatIa et af., 
2009; Petignat et aI., 2007). An optimal performance of SC should involve a full insertion of the 
cytobrush such that it reaches the cervix and a full rotation of the brush, 3-5 times while touching 
the cervix. Another contributor, although minor to this difference, which is the extent of 
exfoliation of the cervical epithelium into the vagina, has been reported in numerous studies and 
reviews, (Gage et al., 2011; Gravitt et al., 20 11; BhatIa et al., 2009; Petignat et af., 2007). 
5.4.2 Age-Specific HPV Prevalence 
As part of determining the differences in HPV prevalences between SC and HPC, the age-
specific HPV prevalence was very important to use. Therefore, for this study, this assessment 
was part of the study objectives. The findings of this study clearly shows that the SC age-specific 
HPV prevalences (i.e. overall, LR, and HR HPV) were higher than those with HPC and that a 
unique pattern of age-specific HPV prevalence by SC was obtained for the Akuse sub-district. 
Specifically, the distributions of the prevalence of the overall and HR HPV with age were 
bimodal for the determination with HPC specimen, but with an additional marginal peak (at 
middle-age) for the determination with the SC specimen (Figures 4.4 - 4.6). However, the 
distribution of LR HPV by age was bimodal for both SC and HPC specimen (Figures 4.4 - 4.6). 
168 
University of Ghana  http://ugspace.ug.edu.gh
Of a higher importance is the generally higher HR HPV prevalence with SC and the additionally 
observed peak at mid age. These implied that the HPV prevalence as determined with the SC 
specimen compared favourably to that determined with HPC specimen and was a better indicator 
for the age groups with the greater burden ofHPV infection. By extension, it may provide a good 
estimation of the age group at a higher risk of developing lower genital tract lesions in future, if 
considered as a specimen collection methods in an HPV testing based cervical cancer screening 
programme in Ghana. 
As has been widely accepted, the high prevalence among younger women (15-29 years) is 
related to sexual behaviour, particular those associated with the initiation of sexual activities 
(Richter et af. , 2013; Clarke et af., 2011; Gage et af., 2011; Bruni et af. , 20 I 0; Chan et af., 2010; 
Onuki et af., 2009). The smaller middle-age (40-44 years) peak of overall HPV prevalence by SC 
specimen has been shown in some cohort studies, with the suggestion that in addition to the 
persistence of previously acquired HPV infection, the acquisition of new HPV infection by these 
women accounts for the peak in prevalence (Baussano et al. , 2013; Gage el al., 2011). The 
second high (major) HPV prevalence peak among women aged 55-59 years determined with 
both SC and HPC specimen has been shown only in selected geographical regions. For instance, 
a very s imilar age-specific overall and HR HPV prevalence distribution has been reported in a 
community based study in China, where an initial major peak was observed among women aged 
20-24 years, small increase among women aged 34-39 years, a second major peak among women 
aged 50-54 years (Ye et af. , 2010) . A study in Nigeria showed a first peak and a second high 
prevalence among women aged between 20-24 years and 55-64 years and a decline above 64 
years (Clarke et af., 2011; Gage et af., 2011). Another study in Nigeria showed two peaks for 
169 
University of Ghana  http://ugspace.ug.edu.gh
women 24-29 years and 55-64 years (Clarke et aI., 2011; Gage et aI. , 2011). A logistic 
regression analysis of the overall, HR and LR HPY prevalence stratified by age for each of the 
risk factors was not possible with the data of this study because the number of participants in the 
subgrouping were too small for most of the cross-tabulations. As such the risk factors that may 
have been associated with the peak prevalence were not identified. With a logistic regression 
analysis in a larger Nigerian study, no risk factor was identified to have had any significant 
association with the high prevalence among the age group of55-64 years (Clarke et aI. , 2011). 
Studies from other regions of the world have shown either a single peak or a bimodal trend with 
a high HPY prevalence among older women (50 years or older). While a large study in Hong 
Kong showed a bimodal trends with the first peaks at a relatively older age (26-30 years) and the 
second peak at a relatively younger age (46-50 years) (Chan et al., 2010), a study in a South 
African community with a high HIY infection rates showed a peak prevalence among women 
less than 25 years and a gradual decline in HPV prevalence with age (Richter et aI., 2013). 
Studies in Denmark, Spain, Italy, Canada and the Netherlands have reported a single peak HPY 
prevalence (20-24 years) with a subsequent a decline with increase in age (Jiang et al., 2013; 
Ogi lvie e/ al., 2013 ; Kjaer et aI., 2008; Franceschi e/ aI. , 2006). A study of Japanese women 
showed a bimodal trend with a first major peak among women aged 15-20 years and a 
continuously raising prevalence in women older than 50 years (Onuki et aI., 2009). A multi-
centre study in China showed a bimodal HR HPV prevalence for all six centres with the first 
peak observed among women < 25 years and a second peak mostly among women 40-44 years 
(Wu et aI. , 2013). 
170 
University of Ghana  http://ugspace.ug.edu.gh
A meta-analysis of the age-specific HPV prevalence for different global regions have shown 
s imilar trends, with a second high peaks among older women (50 years or older) for both Central 
America and West Africa (Bruni et aI., 2010). Although the high prevalence among the older 
women is still not clearly understood, some studies have suggested that the peak could have been 
due to hormonal changes associated with menopause that reactivated latent HPV infection, 
although these studies did not indicate how HPV could remain latent in the cervical tissue. Other 
studies have suggested that this could have been due to changes in sexual behaviour at different 
ages of both the women and their male sexual partners, while other studies have stated that this 
cou ld be a population specific cohort effect. This high prevalence may also result from an 
increased HPV infection due to the reduced protection (local immunity) and increased dryness of 
the female reproductive organ as a result of menopausal changes. For instance, during or after 
menopause, there is a reduction in the secretions of the female reproductive organs which are 
known to protect against infections and provide lubrication that reduce the occurrence/prevent 
microabrasion and subsequent HPV infection (Richter et aI. , 2013; Clarke et aI., 20 II; Gage et 
al., 2011; Bruni et a!. , 2010; Moscicki et al., 2006). The result of the study reported in this thesis 
confirms the existence of this trend but did not investigate the possible factors that may account 
for the trend . 
With respect to the difference in the age-specific HPV prevalence determined with SC and HPC 
specimen, a consistent variation was observed for LR HPV prevalence but not for HR HPV 
prevalence. The variation in the HR HPV prevalence (Figure 4.5), reduced with age, such that 
beyond the age of 49 years the same prevalence were determined for both SC and HPC. This 
observation was not clearly understood as it was expected that the prevalence determined with 
171 
University of Ghana  http://ugspace.ug.edu.gh
SC and HPC was to vary consistently across all age groups as seen for LR HPV prevalence 
(Figure 4.5). Would this imply that HR HPV infection within the study population was common 
in both vaginal and cervical epithelia for women between the ages of 15 years and 44 years, and 
then it becomes more restricted to the cervical epithelia in older women (45-54 years)? In other 
words, is there an increase in HR HPV cervical tissue specificity among older women? On the 
other hand, does the variation in the difference imply that while there was a consistent rate of 
infection of both the cervical and vaginal epithelia by LR HPV with increase in age, the infection 
of the vaginal epithelia by HR HPV reduces with increase in age? Also, this may just be a cohort 
effect or an occurrence by chance. 
Since the HR HPV prevalence determined with SC specimens were higher than that of HPC, it 
is strongly encouraged that SC be used in determining the burden of HR HPV infections in 
Ghana,. However, in light of the fact that this is the first study on SC for determining the burden 
of HPV in Ghana, additional data will strengthen the evidence for the use of Sc. Additionally, 
since most studies on HPV in Ghana and most West African countries report HPV prevalence 
determined with HPC, the HPV data obtained wi th the HPC will be useful for both inter and 
intra-county comparisons. 
5.4.3 Level of Agreement between SC and HPC 
In respect of the level of agreement between the data generated with the specimen collection 
methods, the extent of agreement between HPV prevalences with SC and HPC specimen were 
assessed with a kappa analysis. The findings indicated that the level of agreement between the 
two methods was low to moderate. Specifically, the highest observed level of agreement, was 
moderate and it was with the detection of HR HPV (k = 0.402), while a lower to moderate level 
172 
University of Ghana  http://ugspace.ug.edu.gh
of agreement was observed for the detection of overall and LR HPV (k = 0.321 and 0.328 
respectively). Also, a high level of concordance (77.9%) between the two methods was observed 
with the detection of HR HPV and a moderate level of concordance (67.3%) for the detection of 
the overall HPV (Table 4.13). The McNemar's test indicated that these observed differences in 
prevalences and extent of agreements (concordance and discordance) were significant. 
Other studies have reported high levels of concordance (> 75.0%) for the detection of overall 
HPV infection with either HPC recording a higher prevalence than SC (BhatIa el aI., 2009; 
Petignat el aI., 2007) or SC recording a higher prevalence than HPC (Zhao el al., 2012). 
However, a review had shown that a wide range of levels of agreement (k = 0.24 - 0.96) have 
been reported between SC and HPC for both overall HPV and HR HPV detection (Schmeink et 
al., 2011; Stewart et al., 2007) and that the use of different devices for the collection of SC and 
HPC and populations had an influence on the levels of agreement. Therefore, for the Akuse Sub-
district of the Lower Many Krobo district, SC specimens will be an appropriate alternative to 
HPC specimen for use in determining the burden of HPV infections. 
5.6 ASSOCIA TION OF SEXUAL CHARACTERISTICS WITH HPV POSITIVITY 
There is the need to study the risk factors that may be determinants of HPV infection positivity 
among the population of women alongside the determination of HPV prevalence. Therefore, 
although this study was specifically powered to determine HPV prevalence, data were collected 
to assess the associations between the sexual risk characteristics and the HPV positivity. This 
study compared only the directionality (increased odds or reduced odds) of the associations of 
173 
University of Ghana  http://ugspace.ug.edu.gh
the sexual risk characteristics with HPY positivity (overall, HR, LR, single and multiple HPY 
infection) obtained with SC specimen and those obtained with HPC specimen. This was to 
determine which of the two collection methods will conform to exceptions with regards to the 
directionality of the association. These comparisons were based on the directionality reported by 
studies that determined the associations between sexual risk characteristics and HPY positivity 
and reviews of such studies (Thomas el al., 2004; Johnson el ai., 2012; Bahmanyar el aI., 2012; 
Oakeshott el al. , 2012; Almonte el aI., 2008; Chelimo el al., 2013). Generally, because more 
HPY infections were detected with SC than with HPC and that for a similar sample size, an 
increase in prevalence leads to an increase in the power of a study, the directionality of the 
associations of the sexual risk characteristics with HPY positivity for SC were mostly as 
expected, while those for HPC were less often as expected. Once again, SC was more 
informative as compared to HPC. For instance, an increase in age at first sexual intercourse 
s ignificantly reduced the tendency of being HR HPY and single HPY infection positive with the 
SC specimen, while for HPC specimen, it showed an equal tendency of being HR and single 
HPY positive. Similar trends were obtained for sexual age, as indicated in tables 4.15 and 4.16. 
5.7 HPV VARIANT ANALYSIS 
In determining whether the variants of the detected HPY 16, HPY 18 and HPY 45 were of the 
African lineage and how similar they may be to those detected in other African countries, this 
study obtained a partial nucleotide sequence (- 375 bp) of the long control region (LCR) of the 
genome HPY genotypes. This region of the HPY genome has been shown to be very informative 
174 
University of Ghana  http://ugspace.ug.edu.gh
and therefore very useful in the study ofHPV variants (AI-Awadhi et al., 2013 ; Mendoza et ai., 
2013; Cornet et al., 2012; Tornesello et ai. , 2011). 
All the isolates ofHPV 16 detected in this study were of the Africa lineages, and the commonest 
lineage (80%) was the HPVI6 African 2b (Afr2b). Only I isolate was identified as an HPVI6 
Africa 2a, while the other three isolates were identified as belonging to the HPV 16 African 1 
lineage. The nucleotide position used in this analysis (7460 -7840) was shown not to have had 
enough information to clearly distinguish HPVI6 Afrla and HPV]6 Afrlb since the previously 
described HPV]6 Afrla variants (KF466576) and HPVI6 Afrlb (KF466592) that were included 
in the alignment were clustered closely together in the phylogenetic tree (Figure 4.6). These two 
lineages are distinguished from each other with the substitution at nucleotide position 7438 (A to 
C), which is characteristic to Afr I b (Cornet et ai., 2012), but this position was not part of the 
sequence range obtained in this study. 
A lthough not clearly understood, it was interesting to note that a repeated duplicate sequencing 
of a single DNA specimen produced two the isolates GH618 and GH618 _ 2, which varied at two 
positions (Table 4.17). However, most of the variability identified exclusive to the isolates of 
this study were located within the genome positions 780 I to 7828, which contain the binding 
si tes for the transcription factors , activator protein I (AP I) and Octamer-binding protein I 
(Mendoza et al. , 2013). These may have implication for transcription activities in pathogenic 
pathways. For the 3 women whose SC and HPC specimen were both positive for HPV 16 and had 
good sequences for analysis (three pairs), two pairs of sequences matched exactly. Specifically, 
the iso lates GH511 (obtained with SC specimen) and GH759 (obtained with HPC specimen) 
clustered together as HPV 16 Afr2b lineage, while the isolates GH517 (obtained with SC 
175 
University of Ghana  http://ugspace.ug.edu.gh
specimen) and GH765 (obtained with HPC specimen) clustered together as HPV 16 Afr2b 
lineage. The isolate GH618 (obtained with SC specimen) and isolate GH866 (obtained with HPC 
specimen) did not have their sequences matching exactly but clustered together as HPV 16 Afrl 
lineage. These imply the SC was performed well by the participants and also that the Rover 
sampler was effective at reaching and sampling the cervix uteri. 
Of the 12 HPV18 genotypes sequenced, the African lineage I (Afrl) or the BI and B2 lineages 
according to the new nomenclature were the commonest. The African lineage 2 (Afr2) or the B3 
and C lineages (according to the new nomenclature) were also common. Interestingly, only a few 
were of the European lineage (Eur) or the Al and A3 lineage according to the new nomenclature 
(Figure 4.7). In all , 41.7% (n = 5) of the isolates identified in this study had already been 
described by other studies (Chen e/ al. , 2013 , 2009). Most of the nucleotide variations identified 
exclusively in this study isolates were located within the genome positions 7769 to 7714 and 
7795 to 7814, regions where the binding sites for the transcription factors NFl , YYl and API 
are located. These changes may have implication for transcription activities in pathogenic 
pathways (Mendoza e/ aI. , 2013). For the women whose SC and HPC specimen were positive for 
HPV 18, only one pair had good sequences for comparison, Isolate GH561 (obtained with a SC 
specimen) and isolate 809 (obtained from the HPC specimen of the same woman), matched 
exactly and clustered as African lineage I. 
A ll of the 9 HPV45 genotypes sequenced were determined to belong to the HPV45 lineage Al 
(Africa lineage). For the women whose SC and HPC specimen were both positive for HPV45 , 
on ly two pairs had good sequences and these pairs of sequences, isolate GH583 (obtained with 
176 
University of Ghana  http://ugspace.ug.edu.gh
SC specimen) and GHS31 (obtained with HPC specimen), and GH632 (obtained with SC 
specimen) and GHSSO (obtained with HPC specimen) matched exactly. 
The fact that the HPV IS and HPV 16 isolate were very simi lar to variants isolated from 
specimens collected in Burkina Faso (a country that shares a border with Ghana) but those of 
HPV45 did not match any previously described isolated from the West African region, may be 
informative to the understanding ofHPV genetic variability, transmission and distribution within 
the West African region. Full genome sequencing will throw more light on these issues. 
5.8 CERVICAL LESIONS AND THE ASSOCIATION WITH RISK FACTORS 
Another objective of this study was to determine the prevalence of cervical abnormalities by 
Papanicolaou (Pap) smear screening and evaluate its association with the demographic 
characteristics and risk factors. However, no cervical lesion was detected among the participants 
of this study and therefore the association of cervical lesions with the studied risk factors and 
demographic characteristics could not have been determined. This may not be surprising in a 
study with this small sample size, because in older studies in Ghana with sample sizes of about 
1000 women very low prevalence of cervical lesions, 0.6% low grade lesion, 0.6% high grade 
lesions and 1.6% high grade lesions, were reported (Grace et of., 2006; Nkyekyer, 2000). Also, a 
study in South Africa (a country with high prevalence of cervical cancer) that studied about 
20,600 women reported a 2.42% low grade lesions, I.S high grade lesion and 0.47% diagnosed 
invasive cancers (Fonn et of. , 2002). In Cameroon, a study of over 2000 women identified 2.1%  
low-grade lesions and 1.8% high grade lesion (Marie et 01., 2013). These suggest that the study 
177 
University of Ghana  http://ugspace.ug.edu.gh
reported in this thesis is not powered enough to give a good estimate of the prevalence of 
cervical lesion in this district but then the non-detection of any cervical lesion, leaves us with the 
question, is the community herein studied really a low at risk community for cervical lesions, 
instead of the expected high at risk status based on the distribution of sexual and reproductive 
characteristics, and the determined HPV and STI prevalences?" It may weB mean that Ghana 
needs an active cytology-based screening campaign to determine the prevalence of cervical pre-
cancer lesions. Additionally, the high HPV prevalence determined with SC and HPC but a non-
detection of cervical lesions should be understood in the light of the fact that (as indicated in 
section 2.7) it takes some number of years of HPV persistence for cervical lesion to develop. 
Also, only a small proportion of HPV infection may persist in order to lead to cervical lesion 
(Goldstein el aI., 2009; Kahn, 2009; Schiffman and Wacholder, 2009; Flores et aI., 2006; 
Schiffinan and Castle, 2005) and whether these high HPV infections have persisted long enough 
to lead to cervical lesions was not determined in this study. Additionally, cervical lesion may 
also regress after some time; therefore the non-detection does not imply it never developed. 
5.9 LIMIT A TIONS OF STUDY 
Being one of the very few initial community based studies involving HPV testing and the first to 
evaluate self-specimen collection in Ghana, the information generated in this study of women in 
the Akuse sub-district has provided valuable background information on what to include and 
expect in the implementation of a cervical cancer screening programme in Ghana. However, data 
generated in this study may not completely be free of biases. 
178 
University of Ghana  http://ugspace.ug.edu.gh
Although measures were taken to provide very private interactions during the completion of the 
study questionnaire, there was still the potential for selective response bias with the response to 
sexual and reproductive related questions. This is because some of the women had to be notified 
that the number of children, number of induced abortion and number of miscarriage they 
reported did not tally with the number of total pregnancies they had reported. Also, this study did 
not assess the sexual behaviour characteristic of the participants' male sexual partners which has 
been shown to possibly contribute to the HPV prevalence among women. 
In general, this study was powered to determine the overall and genotype specific HPV 
prevalence and therefore, the findings reporting stratifications of the association of HPV 
infections with sexual risk characteristics should be interpreted with this in mind, as the numbers 
in some subgroups were very small. Due to the fact that all the women in this study were 
determined not to have cervical lesions, the intended investigation of the association of HPV 
infection status as well as the risk characteristics with the presence of cervical lesions was not 
achieved in this study. 
It must also be noted that the HPV variant analysis was performed with partial sequences of long 
control region (LCR) and since full genome analysis may be more desirable for variant analysis 
and HPV classification, the isolates' lineage and or sub-lineage similarities may only be slightly 
modified after a full genome analysis. 
179 
University of Ghana  http://ugspace.ug.edu.gh
CHAPTER SIX 
CONCLUSION AND RECOMMENDATIONS 
6.1 CONCLUSIONS 
This study of women aged 15-70 years was able to show that more than half of the women were 
categorized among the HPV high risk sub-categories of each of the sexual risk characteristics, 
suggesting the Akuse sub-district was a community at high risk for HPV infection. However, 
only a small proportion of the women were categorized among the high risk sub-categories of 
each ofthe reproductive risk characteristics, 
Overall, the reporting of the participant was random among each of the reporting strategies used 
in this study; in other words, no participant characteristic influenced the reporting of the women 
for specimen collection. Therefore, based on the findings of this study, the Akuse sub-district can 
therefore be said to have a high overall HPV prevalence, 43. I % (95% CI 38.0% - 51.0%) with 
SC specimens and 23.3% (95% CI 19.0% - 31.0%) with HPC specimens. Specifically, the 
prevalence obtained with HPC was not greater than the hypothesized 26.3%. These high HPV 
prevalences were particularly among women aged 20-24 years and 55-59 years. Also, the sub-
district can be said to have a moderate to high proportion of multiple HPV infection, with HPC 
specimens and with SC specimens respectively. Generally, the SC method potentially is an 
appropriate alternative to the HPC method for H PY testing as shown by the extent of the 
180 
University of Ghana  http://ugspace.ug.edu.gh
concordance with HPC, age-specific prevalences and the directionality of the association of 
sexual risk characteristics with high risk (HR) HPV positivity. 
The variants of the HPV16, 18 and 45 were predominantly of the African lineage and similar to 
those reported in some West African countries. No cervical lesion (prevalence of and 
associations of cervical lesions with HPV positivity and known risk factors) was detected and 
therefore, assessments or analyses of the associations of cervical lesion and both HPV positivity 
and participant's characteristics were not performed. 
This study presents the first age-specific HPV prevalence data and the first HPV sequence data 
for Ghana. 
6.2 RECOMMENDATION 
6.2.1 Policy 
It is recommended that the Lower Manya Krobo District Health Directorate and Akuse 
Government Hospital , should intensify education on sexual risk behaviours and include 
information about HPV and cervical cancer prevention in the sub-d istr ict's Health Education on 
Reproductive Health. Addit ionally, attention should be pa id to the education on consistent use of 
condom and how it is important in cervical cancer prevention. 
As has been shown in this study, the use of HIV prevalence (due to how it is determined in 
Ghana) will not be an efficient surrogate for determining communities at risk for HPV infection 
and cervical cancer. Therefore the identification of communities at risk by supporting studies 
such as the one reported in this thesis will potentially be useful to prioritizing cervical cancer 
screening programmes even if there be only limited resources. Furthermore, the inclusion of 
181 
University of Ghana  http://ugspace.ug.edu.gh
HPY testing in such studies will be useful for the assessment of communities at risk of cervical 
cancer and therefore, useful for an effective screening programme. It is recommended that self-
specimen collection (SC) and HPY testing are included in the development of a cervical cancer 
screening programme for Ghana. This is because SC provided a comparable HR HPY prevalence 
as that by HPC and with the potential of ensuring high attendance of women at screening, it will 
be key in the determination of communities at risk. Also, the inclusion of self-specimen 
collection will ensure that the epidemiological, clinical and molecular information of both high 
risk and low risk HPY genotypes do not become limiting to future understanding of the changing 
role of these genotypes in cancers and the efficacy of current HPY vaccines to be introduced in 
Ghana and those that might be introduced later. Furthermore, based on the moderate level of 
agreement and the high concordances of the HR HPY prevalence determined with self-collected 
and health personnel collected specimens in this study and the widely reported higher, need for 
training and retraining of personnel, the need for more health personnel and the need for 
attendance at a hospital , in relation to Pap testing, as well as the lack of time, discomfort, 
inconvenience, and cultural objections, a higher coverage in screening activity will be achieved 
in rural areas with self-collection while urban areas have Pap testing systems that will be referral 
si tes for more detailed investigations . Also, the planning of messages for screening education 
should consider the major reasons and misconception regarding the women's preference 
identified in this study. 
6.2.2 Further studies 
In order to determine a very effective strategy for encouraging the participation of women in 
cervical cancer screening activities m Ghana and m order to determine whether different 
182 
University of Ghana  http://ugspace.ug.edu.gh
communities will respond differently to these strategies, further investigation of the strategies 
identified in this study, particularly, the community based specimen collection strategy will help 
achieve effective screening in other communities. A larger study of the age specific HPV 
prevalence that will precisely determine various health characteristics such as immune markers, 
menopausal changes and sexual behaviour of male sexual partners will help throw more light on 
the observed high overall, HR and LR HPV prevalence among women aged 50 years or greater. 
A full genome analysis of the HPV variants identified with additional nucleotide variation in this 
study will help identify if those isolate are new lineages or new sub-lineages of their respective 
HPV genotypes. Additionally, an appropriately powered study (with a large enough samples 
size) is recommended to provide the critical data on the prevalence of cervical pre-cancer lesions 
to appropriately inform the development of a national cervical control programme. 
183 
University of Ghana  http://ugspace.ug.edu.gh
REFERENCE 
Abotchie, P.N., and Shokar, N.K. (2009). Cervical Cancer Screening Among College Students in 
Ghana. International Journal ofG ynecological Cancer, 19, 412-416. 
Adanu, R.M.K. (2002). Cervical cancer knowledge and screening in Accra, Ghana. Journal of 
Women 's Health and Gender-based Medicine, 11, 487-488. 
Adanu, R.M.K., Seffah, J.D., Duda, R., Darko, R., Hill, A., and Anarfi, J. (2010). Clinic visits 
and cervical cancer screening in accra. Ghana Medical Journal, 44, 59-63. 
Adjorlolo-Johnson, G., Unger, E.R., Boni-Ouattara, E., Toure-Coulibaly, K., Maurice, c., 
Vernon, S.D., Sissoko, M., Greenberg, A.E., Wiktor, S.Z., and Chorba, T.L. (20 I 0) . Assessing 
the relationship between HIV infection and cervical cancer in Cote d ' Ivoire: A case-control 
study. BMC 1nfectious Diseases, 10, 242. 
Adler, D.H., Almudevar, A. , Gray, G.E., Allan, 8., and Williamson, A.-L. (2012). High Level of 
Agreement between Clinician-Collected and Self-Collected Samples for HPV Detection among 
South African Adolescents. Journal ofP ediatric and Adolescent Gynecology, 25, 280- 281. 
Afenyadu, D., and Goparaju, L. (2003) . Adolescent Sexual and Reproductive Health Behavior in 
Dodowa, Ghana - CEDPA . 
Agurto, 1., Arrossi, S., White, S., Coffey, P., Dzuba, 1., Bingham, A., Bradley, J. , and Lewis, R. 
(2005). Involving the community in cervical cancer prevention programs. international Journal 
ofG ynecology & Obstetrics, 89, S38- S45. 
Aires, K.A. , Cianciarullo, A.M., Carneiro, S.M., Villa, L.L., Boccardo, E., Perez-Martinez, G., 
Perez-Arellano, I. , Oliveira, M.L.S. , and Ho, P.L. (2006). Production of Human Papillomavirus 
Type 16 L 1 Virus-Like Particles by Recombinant Lactobacillus casei Cells. Applied and 
Environmental Microbiology, 72, 745- 752. 
A I-Awadhi, R. , Chehadeh, W., AI -Jassar, W., AI-Harmi, J. , AI-Saleh, E., and Kapila, K. (2013) . 
Phylogenetic analysis of partial Ll gene of 10 human papillomavirus types isolated most 
commonly from women with normal and abnormal cervical cytology in Kuwait. Archives of 
Virology, 158, 1687- 1699. 
Almonte, M., Albero, G. , Molano, M., Carcamo, c., Garcia, PJ ., and Perez, G. (2008). Risk 
factors for Human Papillomavirus Exposure and Co-factors for Cervical Cancer in Latin 
America and the Caribbean. Vaccine, 26, L16- L36. 
184 
University of Ghana  http://ugspace.ug.edu.gh
Almonte, M., Ferreccio, C , Gonzales, M., Delgado, J.M., Buckley, CH., Luciani, S., Robles, 
S.C., Winkler, J.L. , Tsu, V.D., Jeronimo, J., Cuzick, J., and Sasieni, P. (2011). Risk Factors for 
High-Risk Human Papillomavirus Infection and Cofactors for High-Grade Cervical Disease in 
Peru: International Journal of Gynecological Cancer, 21, 1654-1663. 
Alp AVCI, G. (2012). Genomic organization and proteins of human papillomavirus. Mikrobiyoloji 
Biilteni, 46, 507-515 . 
Anhang, R., Nelson, l.A., Telerant, R., Chiasson, M.A., and Wright, T.C (200S). Acceptability 
of Self-Collection of Specimens for HPV DNA Testing in an Urban Population. Journal of 
Women's Health, 14, 721-728. 
Arbyn, M., and European Commission. Directorate-General Health & Consumer Protection 
(2008). European guidelines for quality assurance in cervical cancer screening (Luxembourg: 
Office for Official Publications oft he European Communities). 
Arrossi, S., Thouyaret, L., Herrero, R., Campanera, A., Magdaleno, A., Cuberli , M., Barletta, P., 
Laudi, R. , and Orellana, L. (2015). Effect of self-collection of HPV DNA offered by community 
health workers at home visits on uptake of screening for cervical cancer (the EMA study): a 
population-based c!uster-randomised trial. The Lancel Global Health, 3, e85-e94. 
Attoh, S., Asmah, R., Wiredu, E.K., Gyasi, R. , and Tettey, Y. (2010). Human papilloma virus 
genotypes in Ghanaian women with cervical carcinoma. East African Medical Journal, 87, 345-
349. 
Awua, A.K., Wiredu, E.K., Osei, Y.D., Sackey, S.T., Asmah, R.H., Tettey, Y. , Gyasi , R.K., 
Attoh, S., Bonney, E.Y., Ampofo, W., and Adjei, A. (2007). Oncogenic genotypes of Human 
Papillomavirus associated with cervical cancer in Ghanaian women. In BooK of Abstracts, 
(Accra: College ofH ealth Sciences, University of Ghana) , p. 
Awua, A.K., Sackey, S.T., Osei, Y.D., Asmah, R.H ., and Wiredu, E.K. (2016). Prevalence of 
human papillomavirus genotypes among women with cervical cancer in Ghana. Irifectious 
Agents and Cancer, 11. 
Bahmanyar, E.R., Paavonen, J. , Naud, P., Salmeron, J., Chow, S.-N., Apter, D., Kitchener, H., 
Castellsague, X. , Teixeira, J.C, Skinner, S.R., Jaisamrarn, D., Limson, G.A., Garland, S.M., 
Szarewski , A., Romanowski , B., Aoki, F. , Schwarz, T.F., Poppe, W.AJ., De Carvalho, N.S., 
Harper, D.M., et al. (2012). Prevalence and risk factors for cervical HPV infection and 
abnormalities in young adult women at enrolment in the multinational PATRICIA trial. 
Gynecologic Oncology, 127, 440-4S0. 
Bais, A.G., van Kemenade, FJ., Berkhof, J. , Verheijen, R.H.M., Snijders, PJ.F., Voorhorst, F., 
Babovic, M., van Ballegooijen, M., Helmerhorst, TJ .M., and Meijer, CJ .L.M. (2007). Human 
papillomavirus testing on self-sampled cervicovaginal brushes: An effective alternative to protect 
nonresponders in cervical screening programs. International Journal of Cancer, 120, ISOS-IS10. 
185 
University of Ghana  http://ugspace.ug.edu.gh
Banura, c., Mirembe, F.M., Katahoire, A.R., Namujju, P.B., and Mbidde, E.K. (2012). Universal 
routine HPV vaccination for young girls in Uganda: a review of opportunities and potential 
obstacles. Infectious agents and cancer, 7,24. 
Barnabas, R.V., Laukkanen, P., Koskela, P., Kontula, 0., Lehtinen, M., and Garnett, G.P. (2006). 
Epidemiology ofHPV 16 and Cervical Cancer in Finland and the Potential Impact of 
Vaccination: Mathematical Modelling Analyses. PLoS Medicine, 3, eI38. 
Baussano, I., Franceschi, S., Gillio-Tos, A., Carozzi, F., Confortini, M., Palma, P., De Lillo, M., 
Del Mistro, A., De Marco, L., Naldoni, c., Pierotti, P., Schincaglia, P., Segnan, N., Zorzi, M., 
Giorgi-Rossi, P., and Ronco, G. (2013). Difference in overall and age-specific prevalence of 
high-risk human papillomavirus infection in Italy: evidence from NTCC trial. BMC Infectious 
Diseases, 13, 238. 
Belinson, J.L., Qiao, Y.L., Pretorius, RG., Zhang, W.H., Rong, S.D., Huang, M.N., Zhao, F.H., 
Wu, L.Y., Ren, S.D., Huang, R.D. , Washington, M.F., Pan, Q.J., Li, L., and Fife, D. (2003). 
Shanxi Province cervical cancer screening study 11: self-sampling for high-risk human 
papillomavirus compared to direct sampling for human papillomavirus and liquid based cervical 
cytology. International Journal oJGynecological Cancer, 13,819-826. 
Bernard, H.-U., Burk, R.D., Chen, Z., van Doorslaer, K., Hausen, H. zur, and de Villiers, E.-M. 
(2010). Classification of papilloma viruses (PVs) based on 189 PV types and proposal of 
taxonomic amendments. Virology, 401, 70- 79. 
Berner, A. , Hassel, S.B., Tebeu, P.-M., Untiet, S., Kengne-Fosso, G., Navarria, 1., Boulvain, M., 
Vassilakos, P., and Petignat, P. (2013). Human Papillomavirus Self-Sampling in Cameroon: 
Women's Uncertainties Over the Reliability of the Method Are Barriers to Acceptance. Journal 
ojL ower Genital Tract Disease, 1 7, 235- 241 . 
Berrington de Gonzalez, A. , Sweetland, S., and Green, J. (2004). Comparison of risk factors for 
squamous cell and adenocarcinomas of the cervix: a meta-analysis. British Journal oJCancer. 
BhatIa, N. , Dar, L., Patro, A.R. , Kumar, P., Kriplani, A. , Gulati , A., Iyer, V.K., Mathur, S.R., 
Sreenivas, V. , Shah, K.V. , and Gravitt, P.E. (2009). Can human papillomavirus DNA testing of 
self-co llected vaginal samples compare with physician-collected cervical samples and cytology 
for cervical cancer screening in developing countries? Cancer epidemiology, 33, 44~50. 
Bosch, F.X., Lorincz, A., Munoz, N., Meijer, C.J.L.M., and Shah, K.V. (2002). The causal 
relation between human papillomavirus and cervical cancer. Journal ojc linical pathology, 55, 
244- 265. 
Bosch, F.X., Qiao, Y.-L., and Castellsague, X. (2006). CHAPTER 2 The epidemiology of human 
papillomavirus infection and its association with cervical cancer. International Journal oj 
Gynecology & Obstetrics, 94, S8-S21. 
186 
University of Ghana  http://ugspace.ug.edu.gh
Bosch, F.X., Burchell , A.N. , Schiffman, M., Giuliano, A.R., de Sanjose, S., Bruni, L. , Tortolero-
Luna, G., Kjaer, S.K., and Munoz, N. (2008). Epidemiology and Natural History of Human 
Papillomavirus Infections and Type-Specific Implications in Cervical Neoplasia. Vaccine, 26, 
Supplement iO, KI - KI6. 
Bosgraaf, R.P. , Verhoef, V.M.J., Massuger, L.F.A.G., Siebers, A.G., Bulten, J ., de Kuyper-de 
Ridder, G.M., Meijer, C.J .M., Snijders, P.J.F. , Heideman, D.A.M., IntHout, J. , van Kemenade, 
FJ. , Melchers, WJ.G., and Bekkers, R.L.M. (2014). Comparative performance of novel self-
sampling methods in detecting high-risk human papillomavirus in 30,130 women not attending 
cervical screening: Novel HPV self-sampling methods compared. International Journal of 
Cancer, n/a-n/a. 
Bradley, J., Barone, M., Mahe, c., Lewis, R., and Luciani, S. (2005) . Delivering cervical cancer 
prevention services in low-resource settings. International Journal ofG ynecology & Obstetrics, 
89, S21 - S29. 
Bruni , L. , Diaz, M., Castellsague, X., Ferrer, E. , Bosch, F.X., and de Sanjose, S. (2010). Cervical 
Human Papillomavirus Prevalence in 5 Continents: Meta-Analysis of I Million Women with 
Normal Cytological Findings. The Journal ofI nfectious Diseases, 202, 1789- 1799. 
Burchell , A.N ., Winer, R.L. , de Sanjose, S., and Franco, E.L. (2006). Chapter 6: Epidemiology 
and transmission dynamics of genital HPV infection. Vaccine, 24, Supplement 3, S52-S61. 
Castellsague, X., and Munoz, N. (2003a). Chapter 3: Cofactors in Human Papillomavirus 
Carcinogenesis-Role of Parity, Oral Contraceptives, and Tobacco Smoking. JNCI Monographs, 
2003, 20- 28. 
Castellsague, X., and Munoz, N. (2003b). Chapter 3: Cofactors in Human Papillomavirus 
Carcinogenesis-Role of Parity, Oral Contraceptives, and Tobacco Smoking. JNCI Monographs, 
2003, 20- 28 . 
Castle, P.E., Solomon, D., Schiffinan, M., Wheeler, C.M. , and for the ALTS Group (2005). 
Human Papillomavirus Type 16 Infections and 2-Year Absolute Risk of Cervical Precancer in 
Women With Equivocal or Mild Cytologic Abnormalities. JNCI Journal oft he National Cancer 
institute, 97, 1066- 1071. 
Castl e, P. E., Stoler, M.H., Wright, T.C. , Jr, Sharma, A., Wright, T.L., and Behrens, C.M. 
(20 1Ia). Performance of carcinogenic human papillomavirus (HPV) testing and HPVI6 or 
HPVI8 genotyping for cervical cancer screening of women aged 25 years and older: a 
subanalysis of the ATHENA study. The lancet oncology, 12, 880- 890. 
Castle, P.E., Rausa, A., Walls, T. , Gravitt, P.E., Partridge, E.E., Olivo, V. , Niwa, S., Morrissey, 
K.G., Tucker, L., Katki , H., and Scarinci, I. (2011 b). Comparative community outreach to 
increase cervical cancer screening in the Mississippi Delta. Preventive Medicine, 52, 452-455. 
187 
University of Ghana  http://ugspace.ug.edu.gh
Castle, P.E., de Sanjose, S., Qiao, Y.-L., Belinson, J.L., Lazcano-Ponce, E., and Kinney, W. 
(2012) . Introduction of Human Papillomavirus DNA Screening in the World: 15 Years of 
Experience. Vaccine, 30, FI17-FI22. 
Chan, P.K.S., Chang, A.R., Yu, M.Y., Li, W.-H., Chan, M.Y.M., Yeung, A.e.M., Cheung, T.-H., 
Yau, T.-N., Wong, S.-M., Yau, e.-W., and Ng, H.-K. (2010). Age distribution of human 
papillomavirus infection and cervical neoplasia reflects caveats of cervical screening policies. 
International Journal of Cancer, I26, 297-30 I. 
Chatterjee, A. (2014). The next generation ofHPV vaccines: nonavalent vaccine V503 on the 
horizon. Expert Review of Vaccines, 13, 1279-1290. 
Chelimo, C., Wouldes, T.A., Cameron, L.D., and Elwood, J.M. (2013). Risk factors for and 
prevention of human papillomaviruses (HPV), genital warts and cervical cancer. Journal of 
Infection, 66,207-217. 
Chen, Z., DeSalle, R. , Schiffinan, M., Herrero, R., and Burk, R.D. (2009). Evolutionary 
Dynamics of Variant Genomes of Human Papillomavirus Types 18, 45, and 97. Journal of 
Virology, 83, 1443- 1455. 
C hen, Z., Schiffinan, M., Herrero, R., DeSalle, R., Anastos, K., Segondy, M., Sahasrabuddhe, 
V.V. , Gravitt, P.E., Hsing, A.W., and Burk, R.D. (2013). Evolution and Taxonomic 
Class ification of Alphapapillomavirus 7 Complete Genomes: HPVI8, HPV39, HPV45 , HPV59, 
HPV68 and HPV70. PLoS ONE, 8, e72565. 
C larke, M.A., Gage, J.e., Ajenifuja, K.O., Wentzensen, N.A., Adepiti, A.C. , Wacholder, S., 
Burk, R.D. , and Schiffman, M. (2011). A population-based cross-sectional study of age-specific 
risk factors for high risk human papillomavirus prevalence in rural Nigeria. Infectious Agents 
and Cancer, 6, 12. 
Clifford, G.M., Smith, J .S., Aguado, T. , and Franceschi, S. (2003). Comparison ofHPV type 
dist ribution in high-grade cervical lesions and cervical cancer: a meta-analysis. British Journal of 
Cancer, 89, 101- 105. 
Clifford , G.M., Gallus, S., Herrero, R., Munoz, N., Snijders, Pol.F. , Vaccarella, S., Anh, P.T.H ., 
Ferreccio, e., Hieu , N.T. , Matos, E., Molano, M., Rajkumar, R., Ronco, G., de Sanjose, S., Shin, 
H.R., Sukvirach, S., Thomas, J.O ., Tunsakul, S., Meijer, C.J .L.M ., and Franceschi, S. (2005a). 
Worldwide distribution of human papillomavirus types in cytologically normal women in the 
International Agency for Research on Cancer HPV prevalence surveys: a pooled analysis. 
Lancet, 366, 991 - 998. 
Clifford, G.M., Rana, R.K. , Smith, J.S., Gough, G., and Pimenta, J.M. (2005b). Human 
Papillomavirus Genotype Distribution in Low-Grade Cervical Lesions: Comparison by 
Geographic Region and with Cervical Cancer. Cancer Epidemiology Biomarkers & Prevention, 
14, 1157- 1164. 
188 
University of Ghana  http://ugspace.ug.edu.gh
Coleman, M.A., Levison, J. , and Sangi-Haghpeykar, H. (2011). HPV vaccine acceptability in 
Ghana, West Africa. Vaccine, 29, 3945- 3950. 
Conesa-Zamora, P. , Ortiz-Reina, S., Moya-Biosca, J ., Domenech-Peris, A. , Orantes-Casado, F., 
Perez-Guillermo, M., and Egea-Cortines, M. (2009). Genotype distribution of human 
papillomavirus (HPV) and co-infections in cervical cytologic specimens from two outpatient 
gynecological clinics in a region of southeast Spain. BMC Infectious Diseases, 9, 124. 
Cornet, I., Gheit, T., Franceschi, S., Vignat, J., Burk, R.D., Syll a, B.S., Tommasino, M., Clifford, 
G.M., and the JARC HPV Variant Study Group (2012). Human Papillomavirus Type 16 Genetic 
Variants: Phylogeny and Classification Based on E6 and LCR. Journal of Virology, 86, 6855-
6861. 
Cox, J.T., Castle, P.E., Behrens, e.M., Sharma, A. , Wright, T.e., and Cuzick, 1. (2013) . 
Comparison of cervical cancer screening strategies incorporating different combinations of 
cytology, HPV testing, and genotyping for HPV 16/18: results from the ATHENA HPV study. 
American Journal of Obstetrics and Gynecology, 208, 184.el - 184.ell. 
C uz ick, J., Clavel, e., Petry, K.-U., Meijer, C.J .L.M., Hoyer, H. , Ratnam, S., Szarewski, A., 
Birembaut, P. , Kulasingam, S., Sasieni, P. , and Iftner, T. (2006). Overview of the European and 
North American studies on HPV testing in primary cervical cancer screening. International 
journal ofc ancer. Jo urnal international du cancer, 119, 1095-110 I. 
C uz ick, J., Bergeron, C., von Knebel Doeberitz, M., Gravitt, P., Jeronimo, J. , Lorincz, A .T. , 
J.L. M. Meijer, e., Sankaranarayanan, R. , J.F. Snijders, P., and Szarewski, A. (2012). New 
Technologies and Procedures for Cervical Cancer Screening. Vaccine, 30, F1 07- F 116. 
Deacon, J.M ., Evans, e.D., Yu le, R., Desai, M., Binns, W., Taylor, e., and Peto, J. (2000). 
Sexual behaviour and smoking as determinants of cervical HPV infection and of CTN3 among 
those infected: a case-control study nested within the Manchester cohort. Britishjournal of 
cancer, 83, 1565- 1572. 
Delius, H., and Hofmann, B. (1994). Primer-directed sequencing of human papillomavirus types. 
Currenl topics in microbiology and immunology, 186, 13- 31 . 
Denny, L. (2012) . Cytological screening for cervical cancer prevention. Best Practice & 
Research Clinical Obstetrics & Gynaecology, 26, 189- 196. 
Denny, L. , Adewole, 1., Anorlu, R., Dreyer, G. , Moodley, M., Smith, T. , Snyman, L. , Wiredu, E., 
Molijn, A. , Quint, W., Ramakrishnan, G., and Schmidt, J. (2014). Human papillomavirus 
prevalence and type di stribution in invasive cervical cancer in sub-Saharan Africa: Cervica l 
Cancer in sub-Saharan Africa. International Journal of Cancer. 
Didelot-Rousseau, M.-N., Nagot, N., Costes-Martineau, V. , Valles, X., Ouedraogo, A., Konate, 
I., Weiss, H.A., Van de Perre, P., Mayaud, P. , and Segondy, M. (2006). Human papillomavirus 
189 
University of Ghana  http://ugspace.ug.edu.gh
genotype distribution and cervical squamous intraepitheliallesions among high-risk women with 
and without HIV-l infection in Burkina Faso. British Journal o/Cancer, 95, 355-362. 
Dillner, J. , Rebolj, M., Birembaut, P., Petry, K.-U., Szarewski, A. , Munk, c., de Sanjose, S., 
Naucler, P., L1overas, B., Kjaer, S., Cuzick, J. , van Ballegooijen, M., Clavel, c., and Iftner, T. 
(2008). Long term predictive values of cytology and human papillomavirus testing in cervical 
cancer screening: joint European cohort study. BMJ, 337, aI754-aI754. 
Domfeh, A. , Wiredu, E., Adjei, A., Ayeh-Kumi, P., Adiku, T., Tettey, Y., Gyasi , R., and Armah, 
H. (2008). Cervical human papillomavirus infection in Accra, Ghana. Ghana medica/journal, 
42, 71-78. 
Doorbar, J. (2013). The E4 protein; structure, function and patterns of expression. Virology, 445, 
80-98. 
Dreer, M., Fertey, J., van de Poel, S., Straub, E., Madlung, J., Macek, B., Iftner, T ., and 
Stubenrauch, F. (2016). Interaction ofNCORlSMRT Repressor Complexes with Papillomavirus 
E81\E2C Proteins Inhibits Viral Replication . PLOS Pathogens, 12, e 1005556. 
Duda, R.B., Chen, G.L., Hill , A.G., Darko, R., Adanu, R.M.K., Seffah, J.D., and Anarfi, J.K . 
(2005) . Screening for Cervical Cancer Still Not Included as Routine Health Care for Women. 
International Journal o/Tropical Medicine, 1, 1-6. 
Dunne, E.F., Datta, S.D., and Markowitz, E.L. (2008). A review of prophylactic human 
papillomavirus vaccines: Recommendations and monitoring in the US. Cancer, J 13,2995-3003. 
Edwin, A.K. (20 I 0). ]s Routine Human Papillomavirus Vaccination an Option for Ghana? 
Ghana Medical Journal, 44, 70- 75. 
Erickson, B.K., Alvarez, R.D., and Huh, W.K. (2013) . Human papillomavirus: what every 
provider should know. American Journal o/Obstetrics and Gynec%gy, 208, 169- 175. 
Ferlay, J ., Shin, H.-R. , Bray, F., Forman, D., Mathers, c., and Parkin, D.M. (20 I 0). Estimates of 
worldwide burden of cancer in 2008: GLOBOCAN 2008. International Journal o/Cancer, 12 7, 
2893- 2917. 
Fertey, J., Ammermann, I., Winkler, M., Stoger, R., Iftner, T. , and Stubenrauch, F. (2010). 
Interaction of the Papillomavirus E8 E2C Protein with the Cellular CHD6 Protein Contributes to 
Transcriptional Repression. Journal 0/ Virology, 84, 9505-9515 . 
Fi rnhaber, c., Le, H. , Petti for, A. , Schulze, D., Michelow, P. , Sanne, I.M., Lewis, D.A., 
Williamson, A.-L., Allan, B., Williams, S., Rinas, A. , Levin, S. , and Smith, J.S . (2009). 
Association between cervical dysplasia and human papillomavirus in HIV seropositive women 
from Johannesburg South Africa. Cancer Causes & Control, 21, 433-443. 
Flannelly, G. (2010). The management of women with abnormal cervical cytology in pregnancy. 
Best Practice & Research Clinical Obstetrics & Gynaecology, 24,51-60. 
190 
University of Ghana  http://ugspace.ug.edu.gh
Flores, R., Papenfuss, M., Klimecki , W.T. , and Giuliano, A.R. (2006). Cross-sectional analysis 
of oncogenic HPV viral load and cervical intraepithelial neoplasia. International journal of 
cancer. Journal international du cancer, 118, 1187-1193. 
Fonn, S., Bloch, B., Mabina, M., Carpenter, S., Cronje, H., Maise, c., Bennun, M., du Toit, G., 
de Jonge, E., Manana, I., and Lindeque, G. (2002). Prevalence of pre-cancerous lesions and 
cervical cancer in South Africa - a multicentre Study. South African Medical Journal, 92, 148-
156. 
Franceschi, S., Herrero, R., Clifford, G.M., Snijders, PJ.F., Arslan, A., Anh, P.T.H., Bosch, 
F.X., Ferreccio, C., Hieu, N.T., Lazcano-Ponce, E., Matos, E., Molano, M., Qiao, Y.-L., 
Rajkumar, R., Ronco, G., de Sanjose, S., Shin, H.-R., Sukvirach, S., Thomas, J.O., Meijer, 
C.J .L. M., et at. (2006). Variations in the age-specific curves of human papillomavirus prevalence 
in women worldwide. International Journal ofC ancer, 119, 2677-2684. 
Franco, E.L., Rohan, T.E., and Villa, L.L. (1999). Epidemiologic Evidence and Human 
Papillomavirus Infection as a Necessary Cause of Cervical Cancer. JNCI Journal oft he National 
Cancer Institute, 91, 506- 511. 
Gaffikin , L. , McGrath, J.A ., Arbyn, M., and Blumenthal, P.O. (2007). Visual inspection with 
acetic acid as a cervical cancer test: accuracy validated using latent class analysis. BMC Medical 
Research Methodology, 7, 36. 
Gage, J.C. , Partridge, E.E., Rausa, A., Gravitt, P.E., Wacholder, S ., Schiffman, M., Scarinci, I. , 
and Castle, P.E. (2011). Comparative Performance of Human Papillomavirus DNA Testing 
Using Novel Sample Collection Methods. Journal ofC linical Microbiology, 49, 4185-4189. 
GAVI Alliance Support (2013). Human papillomavirus vaccine - New and underused vaccines 
support - Types of support - GA VI Alliance. 
Genther, S.M., Sterling, S., Duensing, S. , Munger,. K., Sattler, C., and Lambert, P .F. (2003). 
Q uantitative Role of the Human Papillomavirus Type 16 E5 Gene during the Productive Stage of 
the Viral Life Cycle. Journal of Virology, 77, 2832-2842. 
Giuliano, A.R. , Pa lefsky, J.M., Goldstone, S., Moreira, E.D., Penny, M.E., Aranda, c., Vardas, 
E., Moi, H., Jessen, H., Hillman, R., Chang, Y.-H., Ferris, D., Rouleau, D., Bryan, J., Marshall , 
J .B., Vuocolo, S., Barr, E., Radley, D., Haupt, R.M., and Guris, D. (2011). Efficacy of 
Quadrivalent HPV Vaccine against HPV Infection and Disease in Males. New England Journal 
ofM edicine, 364, 401-411. 
GLOBOCAN, JARC (2008). GLOBOCAN 2008: IARC Cancer Fast Statistics. 
GLOBOCAN, IARC (2012). GLOBOCAN 2012: Estimated Cancer incidence, mortality and 
Prevalence Worldwide in 2012; Fact Sheets. 
191 
University of Ghana  http://ugspace.ug.edu.gh
Gok, M., Heideman, D.A.M., van Kemenade, FJ., Berkhof, J., Rozendaal , L., Spruyt, J .W.M., 
Voorhorst, F., Belien, J .A.M., Babovic, M., Snijders, PJ.F., and Meijer, CJ.L.M. (2010) . HPV 
testing on self collected cervicovaginal lavage specimens as screening method for women who 
do not attend cervical screening: cohort study. BMJ, 340, cl 04O-c 1040. 
Goldstein, M.A., Goodman, A., del Carmen, M.G., and Wilbur, D.e. (2009). Case records of the 
Massachusetts General Hospital. Case 10-2009. A 23-year-old woman with an abnormal 
Papanicolaou smear. The New Englandjournal ofm edicine, 360, 1337-1344. 
Grace, L.e., Allan, G.H., Joseph, S., Richard, M.K.A., Rudolph, D., John, K.A., and Rosemary, 
B.D. (2006). Epidemiology of Cervical Cancer and Dysplasia in a Cross-sectional Study of 
Women in Accra, Ghana. International Journal of Tropical Medicine , 1,6-10. 
Gravitt, P.E., Lacey, J.V., Jr, Brinton, L.A., Barnes, W.A., Kornegay, J.R., Greenberg, M.D., 
Greene, S.M., Hadjimichael, O.e., McGowan, L., Mortel, R., Schwartz, P.E. , Zaino, R. , and 
Hildesheim, A. (2001) . Evaluation of self-collected cervicovaginal cell samples for human 
papillomavirus testing by polymerase chain reaction. Cancer epidemiology, biomarkers & 
prevention: a publication oft he American Associationfor Cancer Research, cosponsored by the 
American Society ofP reventive Oncology, 10, 95- 100. 
Gravitt, P.E., Belinson, J.L. , Salmeron, J., and Shah, K.V. (2011). Looking ahead : A case for 
human papillomavirus testing of self-sampled vaginal specimens as a cervical cancer screening 
strategy. International Journal of Cancer, 129, 517-527. 
Greco, D., Kivi , N., Qian, K. , Leivonen, S.-K., Auvinen, P., and Auvinen, E. (2011). Human 
Papillomavirus 16 E5 Modulates the Expression of Host MicroRNAs. PLoS ONE, 6, e21646. 
Grulich, A.E., van Leeuwen, M.T., Falster, M.O., and Vajdic, e.M. (2007) . Incidence of cancers 
in people with HIV/AIDS compared with immunosuppressed transplant recipients: a meta-
analysis. The Lancet, 370, 59- 67. 
Gulland , A. (2012). Uganda launches HPV vaccination programme to fight its commonest 
cancer. BMJ, 345, e6055-e6055. 
Gustafsson, L. , Ponten, J., Zack, M., and Adami, H.O. (1997). International incidence rates of 
invas ive cervical cancer after introduction of cytological screening. Cancer causes & control: 
CCC, 8, 755- 763. 
Hanahan, D., and Weinberg, R.A . (2000). The hallmarks of cancer. Cell, 100, 57-70 . 
Hanahan, D., and Weinberg, R.A . (2011). Hallmarks of Cancer: The Next Generation. Cell, 144, 
646-674. 
Handlogten, K.S., Molitor, RJ., Roeker, L.E., NarJa, N.P. , Bachman, MJ ., Quayson, S., Owusu-
Afriyie, 0., Adjei , E. , Ankobea, F., Clayton, A., Roberts, L. , MacLaughlin, K., and Ansong, D. 
192 
University of Ghana  http://ugspace.ug.edu.gh
(2014). Cervical Cancer Screening in Ghana, West Africa: Prevalence of Abnormal Cytology 
and Challenges for Expanding Screening. Int J Gynecol Pathol., 33, 197-202. 
Harper, D.M., Franco, E.L., Wheeler, c., Ferris, D.G., Jenkins, D., Schuind, A. , Zahaf, T. , Innis, 
B., Naud, P., De Carvalho, N.S., Roteli-Martins, C.M., Teixeira, J. , Blatter, M .M., Korn, A.P., 
Quint, W., Dubin, G. , and GlaxoSmithKline HPV Vaccine Study Group (2004). Efficacy ofa 
bivalent LI virus-like particle vaccine in prevention of infection with human papillomavirus 
types 16 and 18 in young women: a randomised controlled trial. Lancet, 364, 1757-1765. 
Hu, X., Zhang, Z., Ma, D., Huettner, P.c., Massad, L.S. , Nguyen, L., Borecki, I. , and Rader, J.S. 
(2010). TP53, MDM2, NQO 1, and Susceptibility to Cervical Cancer. Cancer Epidemiology 
Biomarkers & Prevention, 19, 755-761. 
Hunter, M .I., Tewari, K., and Monk, BJ. (2008). Cervical neoplasia in pregnancy. Part 2: current 
treatment of invasive disease. American Journal of Obstetrics and Gynecology, 199, 10- 18. 
Insinga, R.P., Perez, G. , Wheeler, C.M., Koutsky, L.A., Garland, S.M., Leodolter, S., Joura, 
E.A., Ferris, D.G., Steben, M., Hernandez-Avila, M ., Brown, D.R., Elbasha, E., Munoz, N., 
Paavonen, J ., and Haupt, R.M. (201 I). Incident Cervical HPV Infections in Young Women: 
Transition Probabilities for CIN and Infection Clearance. Cancer Epidemiology Biomarkers & 
Prevention, 20, 287- 296. 
International Collaboration of Epidemiological Studies of Cervical Cancer (2007). Cervical 
cancer and hormonal contraceptives: collaborative reanalysis of individual data for 16573 
women with cervical cancer and 35 509 women without cervical cancer from 24 epidemiological 
studies. The Lancet, 370, 1609- 1621. 
Jemal, A. , Bray, F., Center, M.M., Ferlay, J., Ward, E., and Forman, D. (2011). Global cancer 
statistics. CA: A Cancer Journal for Clinicians, 61 , 69- 90. 
Jemal, A. , Bray, F., Forman, D., O'Brien, M., Ferlay, J ., Center, M., and Parkin, D.M. (2012). 
Cancer burden in Africa and opportunities for prevention. Cancer, 118,4372-4384. 
Jiang, Y., Brassard , P., Severini, A., Mao, Y., Li , Y ., Laroche, J ., Chatwood, S., Corriveau, A., 
Kandola, K., Hanley, B., Sobol, I., Ar-Rushdi , M., Johnson, G., Lo, J., Ratnam, S., Wong, T., 
Demers, A ., Jayaraman, G. , Totten, S., and Morrison, H. (2013). The prevalence of human 
papillomavirus and its impact on cervical dysplasia in Northern Canada. Infectious Agents and 
Cancer, 8, 25. 
Johnson, A.M., Mercer, C.H ., Beddows, S., de Silva, N ., Desai, S., Howell-Jones, R. , Carder, c., 
Sonnenberg, P. , Fenton, K.A. , Lowndes, c., and Soldan, K. (2012). Epidemiology of, and 
behavioural ri sk factors for, sexually transmitted human papillomavirus infection in men and 
women in Britain. Sexually Transmitted Infections, 88, 212-217. 
Kahn, J.A. (2009). HPV Vaccination for the Prevention of Cervical Intraepithelial Neoplasia. 
New England Journal of Medicine, 361 , 271-278. 
193 
University of Ghana  http://ugspace.ug.edu.gh
Kapeu, A.S., Luostarinen, T., Jellum, E., Dillner, J., Hakama, M., Koskela, P., Lenner, P., Love, 
A., Mahlamaki, E., Thoresen, S., Tryggvadottir, L., Wadell, G., Youngman, L., and Lehtinen, M. 
(2008). Is Smoking an Independent Risk Factor for Invasive Cervical Cancer? A Nested Case-
Control Study Within Nordic Biobanks. American Journal ofE pidemiology, 169, 480-488. 
Katki, H.A., Kinney, W.K., Fetterman, B., Lorey, T., Poitras, N.E., Cheung, L., Demuth, F., 
Schiffinan, M., Wacholder, S., and Castle, P.E. (2011). Cervical cancer risk for women 
undergoing concurrent testing for human papillomavirus and cervical cytology: a population-
based study in routine clinical practice. The Lancet Oncology, 12, 663-672. 
Kennedy, I.M., Haddow, J.K., and Clements, J.B. (1991). A negative regulatory element in the 
human papillomavirus type 16 genome acts at the level of late mRNA stability. Journal of 
virology, 65, 2093-2097. 
Khan, MJ., Castle, P.E., Lorincz, A.T., Wacholder, S., Sherman, M., Scott, D.R., Rush, B.B., 
Glass, A.G., and Schiffman, M. (2005). The Elevated 1O -Year Risk of Cervical Precancer and 
Cancer in Women With Human Papillomavirus (HPV) Type 16 or 18 and the Possible Utility of 
Type-Specific HPV Testing in Clinical Practice. JNCI Journal oft he National Cancer Institute, 
97, 1072-1079. 
Kim, J.J., and Goldie, SJ. (2008). Health and Economic Implications ofHPV Vaccination in the 
United States. New England Journal ofM edicine, 359, 821-832. 
Kim, J.J ., Song, S., Jin, c., Lee, J., Lee, N., and Lee, K. (2012). Factors Affecting the Clearance 
of High-Risk Human Papillomavirus Infection and the Progression of Cervical Jntraepithelial 
Neoplasia. Journal ofI nternational Medical Research, 40, 486-496. 
Kjaer, S.K., Breugelmans, G., Munk, C., Junge, J., Watson, M., and Iftner, T. (2008) . 
Population-based prevalence, type- and age-specific distribution ofHPV in women before 
introduction of an HPV-vaccination program in Denmark. International Journal ofC ancer, 123, 
1864-1870. 
Koidl, c., Bozic, M., Hadzisejdic, I., Grahovac, M., Grahovac, B., Kranewitter, W., Marth, E., 
and Kessler, H.H . (2008). Comparison of molecular assays for detection and typing of human 
papillomavirus. American Journal of Obstetrics and Gynecology, 199, 144.e 1-144.e6. 
Koshiol, I., Lindsay, L., Pi menta, I.M., Poole, c., Jenkins, D., and Smith, J.S. (2008). Persistent 
Human Papillomavirus Infection and Cervical Neoplasia: A Systematic Review and Meta-
Analysis. American Journal ofE pidemiology, 168, 123-137. 
Kulasingam, S., Hughes, J., Kiviat, N .B., Mao, C., Weiss, N., and Kuypers, J. (2002). Evaluation 
of human papillomavirus testing in primary screening for cervical abnormalities: Comparison of 
sensitivity, specificity, and frequency of referral. JAMA , 288, 1749-1757. 
194 
University of Ghana  http://ugspace.ug.edu.gh
Uiara, E., Day, N., and Hakama, M. (1987). Trends in mortality from cervical cancer in the 
Nordic countries: association with organised screening programmes. The Lancet, 329, 1247-
1249. 
Lazcano-Ponce, E., Lorincz, A.T. , Cruz-Valdez, A., Salmeron, J ., Uribe, P ., Velasco-
Mondragon, E. , Nevarez, P.H., Acosta, R.D., and Hernandez-Avila, M. (20 11). Self-collection of 
vaginal specimens for human papillomavirus testing in cervical cancer prevention (MARCH): a 
community-based randomised controlled trial. The Lancet, 378,1868-1873. 
Leinonen, M., Nieminen, P., Kotaniemi-Talonen, L., Malila, N., Tarkkanen, J. , Laurila, P. , and 
Anttila, A. (2009). Age-Specific Evaluation of Primary Human Papillomavirus Screening vs 
Conventional Cytology in a Randomized Setting. Journal oft he National Cancer Institute , 101, 
1612-1623. 
Lenselink, C.H., Melchers, W.J.G., Quint, W.G.V., Hoebers, A.MJ., Hendriks, J.C.M., 
Massuger, L.F.A.G., and Bekkers, R.L.M. (2008). Sexual Behaviour and HPV Infections in 18 to 
29 Year Old Women in the Pre-Vaccine Era in the Netherlands. PLoS ONE, 3, e3743. 
LEPPERT, P.c. (1995). Anatomy and Physiology of Cervical Ripening. Clinical Obstetrics and 
Gynecology, 38. 
Liu, S.S., Leung, R.C.Y., Chan, K.K.L., Cheung, A.N.Y., and Ngan, H.Y.S. (2009). Evaluation 
of a Newly Developed GenoArray Human Papillomavirus (HPV) Genotyping Assay and 
Comparison with the Roche Linear Array HPV Genotyping Assay. Journal of Clinical 
Microbiology, 48, 758-764. 
Lowy, D.R., Solomon, D., Hildesheim, A., Schiller, J.T., and Schiffman, M. (2008). Human 
papillomavirus infection and the primary and secondary prevention of cervical cancer. Cancer, 
113,1980-1993. 
Lues ley, D., and Leeson, S. (2010). Colposcopy and programme management (Sheffield: NHS 
Cancer Screening Programme) . 
Luhn, P., Walker, J., Schiffman, M., Zuna, R.E., Dunn, S.T., Gold, M.A., Smith, K. , Mathews, 
C., A llen, R.A., Zhang, R., Wang, S., and Wentzensen, N. (2013). The role of co-factors in the 
progression from human papillomavirus infection to cervical cancer. Gynecologic Oncology, 
128, 265- 270. 
Marie, T.P., Sando, Z., Ndoumba, A., Sandjong, I. , Mawech-Fauceglia, P., and Doh, A.S. 
(20 13). Prevalence and Geographical Distribution of Precancerous Lesions of the Uterine Cervix 
in Cameroon. Journal of Cytology & Histology, 4. 
Martfn, P., Kilany, L., Garda, D., Lopez-Garcia, A.M., Martfn-Azana, MJ., Abraira, V. , and 
Bellas, C. (2011). Human papillomavirus genotype distribution in Madrid and correlation with 
cytological data. BMC Infectious Diseases, 11, 316. 
195 rrn~  (' I 'I l. . ) e.' _ I , '-' - n. OF PUB 
C0iJ::.06- /. I 'E' .. LIe '1: ,4 U Ii U G r~A RY 
LE GON 
University of Ghana  http://ugspace.ug.edu.gh
Mayrand, M.-H., Duarte-Franco, E., Rodrigues, I., Walter, S.D., Hanley, J., Ferenczy, A., 
Ratnam, S., Coutlee, F., and Franco, E.L. (2007a). Human papillomavirus DNA versus 
Papanicolaou screening tests for cervical cancer. The New Englandjournal o/medicine, 357, 
1579-1588. 
Mayrand, M.-H., Duarte-Franco, E., Rodrigues, I., Walter, S.D., Hanley, J., Ferenczy, A., 
Ratnam, S. , Coutlee, F., Franco, E.L., and Canadian Cervical Cancer Screening Trial Study 
Group (2007b). Human papillomavirus DNA versus Papanicolaou screening tests for cervical 
cancer. The New Englandjournal o/medicine, 357,1579-1588. 
Mbulawa, Z.Z.A., Coetzee, D., Marais, D.J., Kamupira, M., Zwane, E., Allan, B., Constant, D., 
Moodley, J.R., Hoffman, M., and Williamson, A. (2009). Genital Human Papillomavirus 
Prevalence and Human Papillomavirus Concordance in Heterosexual Couples Are Positively 
Associated with Human Immunodeficiency Virus Coinfection. The Journal 0/I nfectious 
Diseases, 199, 1514-1524. 
McCredie, M.R. , Sharples, KJ., Paul, c., Baranyai, J., Medley, G., Jones, R.W., and Skegg, D.C. 
(2008) . Natural history of cervical neoplasia and risk of invasive cancer in women with cervical 
intraepithelial neoplasia 3: a retrospective cohort study. The Lancet Oncology, 9, 425-434. 
Mejlhede, N., Pedersen, B.V., Frisch, M., and Fomsgaard, A. (2010) . Multiple human papilloma 
virus types in cervical infections: competition or synergy? APMIS: acta pathologica, 
microbiologica, et immunologica Scandinavica, 118, 346- 352. 
Mendoza, L., Picconi, M.A., Mirazo, S., Mongel6s, P., Gimenez, G., Basiletti, J., and Arbiza, J. 
(2013) . Distribution of HPV -16 variants among isolates from Paraguayan women with different 
grades of cervical lesion. International Journal o/Gynecology & Obstetrics, 122, 44-47. 
Mitchell, S., Ogilvie, G., Steinberg, M., Sekikubo, M., Biryabarema, C., and Money, D. (2011). 
Assessing women's willingness to collect their own cervical samples for HPV testing as part of 
the ASPIRE cervical cancer screening project in Uganda. International Journal o/Gynecology & 
Obstetrics, 114, 111 - 115 . 
Moscicki, A.-B. , Schiffman, M., Kjaer, S., and Villa, L.L. (2006). Chapter 5: Updating the 
natural history of HPV and anogenital cancer. Vaccine, 24, Supplement 3, S42-S51. 
Moscicki, A.-B., Widdice, L., Ma, Y., Farhat, S., Miller-Benningfield, S., Jonte, J. , Jay, J., 
Godwin de Medina, c., Hanson, E., Clayton, L. , and Shiboski, S. (20 10). Comparison of natural 
histories of human papillomavirus detected by clinician- and self-sampling. International 
Journal o/Cancer, 127, 1882- 1892. 
Moss, S .. , Gray, A., Marteau, T ., Legood, R., Henstock, E., and Maissi , E. (2004). Evaluation of 
HPVf L BC; Cervical Screening Pilot Studies. 
196 
University of Ghana  http://ugspace.ug.edu.gh
Munoz, N. , Bosch, F.x. , de Sanjose, S., Herrero, R. , Castellsague, X. , Shah, K.V. , Snijders, 
PJ. F., and Meijer, CJ.L.M. (2003). Epidemiologic Classification of Human Papillomavirus 
Types Associated with Cervical Cancer. New England Journal ofM edicine, 348, 518- 527. 
Munoz, N ., Castellsague, X., de Gonzalez, A.B. , and Gissmann, L. (2006). Chapter I : HPV in 
the etiology of human cancer. Vaccine, 24, Supplement 3, S I- S I O. 
Narechania, A., Chen, Z., DeSalle, R., and Burk, R.D. (2005). Phylogenetic incongruence among 
oncogenic genital alpha human papillomaviruses. Journal ofv irology, 79, 15503-15510. 
Neuman, RJ., Huettner, P.c., Li, L., Mardis, E.R., Duffy, B.F., Wilson, R.K. , and Rader, J.S. 
(2000). Association between DQB I and cervical cancer in patients with human papillomavirus 
and family controls. Obstetrics and Gynecology, 95, 134- 140. 
Nkyekyer, K. (2000). Pattern of gynaecological cancers in Ghana. East African medical j ournal, 
77, 534- 53 8. 
Norton, W.E., Fisher, J.D. , Amico, K.R. , Dovidio, J.F. , and Johnson, B.T. (2012). Relative 
Efficacy of a Pregnancy, Sexually Transmitted Infection, or Human Immunodeficiency Virus 
Prevention- Focused Intervention on Changing Sexual Risk Behavior Among Young Adults. 
Journal of American College Health, 60, 574-582. 
Oakeshott, P., Aghaizu, A., Reid, F. , Howell-Jones, R., Hay, P. E., Sadiq, S.T., Lacey, CJ., 
Beddows, S., and Soldan, K. (2012). Frequency and risk factors for prevalent, incident, and 
persistent genital carcinogenic human papillomavirus infection in sexually active women: 
community based cohort study. BMJ, 344, e4168-e4168. 
Ogilvie, G.S., van Niekerk, OJ., Krajden, M., Martin, R.E., Ehlen, T.G., Ceballos, K. , Peacock, 
SJ ., Smith, L.W., Kan, L. , Cook, D.A. , Mei, W., Stuart, G.c., Franco, E.L., and Coldman, AJ . 
(2010). A randomized controlled trial of Human Papillomavirus (HPV) testing for cervical 
cancer screening: trial design and preliminary results (HPV FOCAL Trial). BMC Cancer, J 0, 
111. 
Ogilv ie, G .S., Cook, D.A., Taylor, D.L., Rank, c., Kan, L. , Yu, A., Me i, W., van Niekerk, OJ ., 
Co ldman, AJ ., and Krajden, M. (2013) . Population-based evaluation of type-specific HPV 
prevalence among women in British Columbia, Canada. Vaccine, 3 J, 1129- 11 33. 
O nuki, M., Matsumoto, K. , Satoh, T., Oki, A. , Okada, S., Minaguchi , T ., Ochi , H. , N akao, S ., 
Someya, K., Yamada, N. , Hamada, H., and Yoshikawa, H. (2009) . Human papillomavirus 
infections among Japanese women: age-related prevalence and type-specific risk for cervica l 
cancer. Cancer Science, J0 0, 1312- 1316. 
O rtiz, M., Torres, M., Munoz, L., Fernandez-Garcia, E., Canals, J. , Cabornero, A.I. , Aguilar, E ., 
Ballesteros, J. , del Amo, J. , and Garcia-Saiz, A . (2006). Oncogenic Human Papillomavirus 
(HPV) Type Distribution and HPV Type 16 E6 Variants in Two Spanish Population Groups w ith 
Different Levels ofHPV Infection Risk. Journal ofC linical Microbiology, 44, 1428- 1434. 
197 
University of Ghana  http://ugspace.ug.edu.gh
Palefsky, J.M. (2009). Human papillomavirus-related disease in people with HIV. Current 
Opinion in HIV and AIDS, 4, 52-56. 
Palefsky, J.M ., and Holly, E.A. (2003). Chapter 6: Immunosuppression and Co-infection with 
HIV. JNCI Monographs, 2003, 41-46. 
Parkin, D.M., Bray, F., Ferlay, J., and Pisani, P. (2005). Global Cancer Statistics, 2002. CA: A 
Cancer Journal Jor Clinicians, 55, 74-108. 
Patanwala, I.Y., Bauer, H.M., Miyamoto, J., Park, LU., Huchko, MJ., and Smith-McCune, K.K. 
(2013). A systematic review of randomized trials assessing human papillomavirus testing in 
cervical cancer screening. American Journal oJObstetrics and Gynecology, 208, 343-353. 
PATH; Practical experience (201 Oa). Progress in preventing cervical cancer: Updated evidence 
on vaccination and screening. 
PATH; Practical experience (2010b) . Human Papillomavirus Vaccines : WHO Postion Paper. 
Peh, W.L., Brandsma, J.L. , Christensen, N.D., Cladel, N.M., Wu, X., and Doorbar, J. (2004). 
The Viral E4 Protein Is Required for the Completion ofthe Cottontail Rabbit Papillomavirus 
Productive Cycle In Vivo. Journal oj Virology, 78, 2142-2151. 
Petignat, P. , and Vassilakos, P. (2012). Is It Time to Introduce HPV Self-Sampling for Primary 
Cervical Cancer Screening? JNCI Journal oJthe National Cancer Institute , 104, 166-167. 
Petignat, P. , Faltin, D.L., Bruchim, 1., Tramer, M.R., Franco, E.L., and Coutlee, F. (2007) . Are 
self-collected samples comparable to physician-collected cervical specimens for human 
papillomavirus DNA testing? A systematic review and meta-analysis. Gynecologic oncology, 
105,530-535 . 
P icot, J. , Shepherd, J., Kavanagh, J., Cooper, K., Harden, A., Barnett-Page, E. , Jones, J. , Clegg, 
A., Hartwell , D., and Frampton, G.K. (20 12). Behavioural interventions for the prevention of 
sexually transmitted infections in young people aged 13- 19 years: a systematic review. Health 
Education Research, 27, 495- 512 . 
Piras, F. , Piga, M., De Montis, A. , Zannou, A.R., Minerba, L., Perra, M.T., Murtas, D., Atzori , 
M ., Pittau, M., Maxia, c., and Sirigu, P. (2011) . Prevalence of human papillomavirus infection in 
women in Benin, West Africa. Virology Journal, 8, 514. 
Plummer, M., Peto, J. , Franceschi, S., and; on behalf of the International Collaboration of 
Epidemiological Studies of Cervical Cancer (2012). Time since first sexual intercourse and the 
risk of cervical cancer. International Journal oJCancer, 130,2638- 2644 . 
Q uentin, W., Terris-Prestholt, F., Legood, R., Mayaud, P., Adu-Sarkodie, Y., and Opoku, B.K . 
(20 10). Costs of cervical cancer screening in Ghana. 
198 
University of Ghana  http://ugspace.ug.edu.gh
Quentin, W., Terris-Prestholt, F., Changalucha, J. , Soteli, S., Edmunds, W.J. , Hutubessy, R., 
Ross, D.A., Kapiga, S., Hayes, R. , and Watson-Jones, D. (2012). Costs of delivering human 
papillomavirus vaccination to schoolgirls in Mwanza Region, Tanzania. BMC Medicine, 10, 137. 
Ramanathan, M., and Varghese, J. (2010). The HPV vaccine demonstration projects: we should 
wait, watch and learn. Indian Journal ofM edical Ethics, 7,43-45. 
Richter, K. , Becker, P., Horton, A., and Dreyer, G. (2013). Age-specific prevalence of cervical 
human papillomavirus infection and cytological abnormalities in women in Gauteng Province. 
South African Medical Journal, 103. 
Ronco, G., Segnan, N., Giorgi-Rossi, P., Zappa, M., Casadei, G.P., Carozzi, F., Palma, P.O., Del 
Mistro, A. , Folicaldi, S., Gillio-Tos, A., Nardo, G., Naldoni, e., Schincaglia, P. , Zorzi, M. , 
Confortini, M., and Cuzick, J. (2006). Human Papillomavirus Testing and Liquid-Based 
Cytology: Results at Recruitment From the New Technologies for Cervical Cancer Randomized 
Controlled Trial. JNCI Journal oft he National Cancer Institute, 98, 765-774. 
Ronco, G., Giorgi-Rossi , P., Carozzi, F., Confortini, M., Palma, P.O., Del Mistro, A ., 
Ghiringhello, B., Girlando, S., Gillio-Tos, A., De Marco, L., Naldoni, e., Pierotti, P ., Rizzolo, 
R., Schincaglia, P., Zorzi, M., Zappa, M., Segnan, N., and Cuzick, J. (2010). Efficacy of human 
papillomavirus testing for the detection of invasive cervical cancers and cervical intraepithelial 
neoplasia: a randomised controlled trial. The Lancet Oncology, 11, 249-257. 
Rositch, A.F. , Poole, e., Hudgens, M.G., Agot, K., Nyagaya, E., Moses, S., Snijders, P.J.F., 
Meijer, C.J.L.M., Bailey, R.e., and Smith, J.S. (2011). Multiple Human Papillomavirus 
Infections and Type Competition in Men. Journal ofI nfectious Diseases, 205, 72-81 . 
Rossi, P.G., Marsili, L.M ., Camilloni, L. , Iossa, A. , Lattanzi, A., Sani, C., Di Pierro, C., 
Grazzini, G., Angeloni, C., Capparucci, P., Pellegrini, A. , Schiboni , M.L., Sperati, A. , 
Confortini , M., Bellanova, C., 0 ' Addetta, A. , Mania, E., Visioli , e.B., Sereno, E., and Carozzi, 
F. (2011). The effect of self-sampled I-IPV testing on participation to cervical cancer screening in 
Italy: a randomised controlled trial (ISRCTN96071600). British Journal of Cancer, 104, 248-
254. 
Saitou, N., and Nei, M. (1987). The neighbor-joining method: a new method for reconstructing 
phylogenetic trees. Molecular biology and evolution, 4, 406-425. 
Samoff, E. (2005). Association of Chlamydia trachomatis with Persistence of High-Risk Types 
of Human Papillomavirus in a Cohort of Female Adolescents. American Journal of 
Epidemiology, 162,668--675. 
Sanghvi, 1-1., Limpaphayom, K.K., Plotkin, M. , Charurat, E., Kleine, A., Lu, E., 
Eamratsameekool, W., and Palanuwong, B. (2008). Cervical cancer screening using visual 
inspection with acetic acid: operational experiences from Ghana and Thailand . Reproductive 
Health Matters, 16, 67- 77. 
199 
University of Ghana  http://ugspace.ug.edu.gh
de Sanjose, S., Diaz, M., Castellsague, X., Clifford, G., Bruni, L. , Munoz, N., and Bosch, F.X. 
(2007). Worldwide prevalence and genotype distribution of cervical human papillomavirus DNA 
in women with normal cytology: a meta-analysis. The Lancet Infectious Diseases, 7, 453-459. 
Sankaranarayanan, R., Wesley, R.S., and International Agency for Research on Cancer (2003). A 
practical manual on visual screening for cervical neoplasia (Lyon: International Agency for 
Research on Cancer, World Health Organization). 
Sankaranarayanan, R., Gaffikin, L. , Jacob, M., Sellors, J., and Robles, S. (2005). A critical 
assessment of screening methods for cervical neoplasia. International Journal of Gynecology & 
Obstetrics, 89, S4-S 12. 
Sankaranarayanan, R., Nene, B.M., Shastri, S.S., Jayant, K., Muwonge, R., Budukh, A.M., 
Hingmire, S., Malvi, S.G., Thorat, R ., Kothari, A., Chinoy, R., Kelkar, R., Kane, S., Desai, S., 
Keskar, V.R. , Rajeshwarkar, R., Panse, N., and Dinshaw, K.A. (2009). HPV screening for 
cervical cancer in rural India. The New Englandjournal ofm edicine , 360, 1385-1394. 
Sanner, K., Wikstrom, I., Strand, A. , Lindell, M., and Wi lander, E. (2009). Self-sampling of the 
vaginal fluid at home combined with high-risk HPV testing. British Journal ofC ancer, 101 , 
871-874. 
Sarian, L.O., Hammes, L.S. , Longatto-Filho, A., Guarisi, R., Derchain, S.F.M., Roteli-Martins, 
e., Naud, P., Erzen, M., Branca, M., Tatti, S., de Matos, J.e. , Gontijo, R., Maeda, M.Y.S ., Lima, 
T., Costa, S., Syrjanen, S., and Syrjanen, K. (2009). Increased Risk of Oncogenic Human 
Papillomavirus Infections and Incident High-Grade Cervicallntraepithelial Neoplasia Among 
Smokers. Sexually Transmitted Diseases, 36, 241-248. 
Saslow, D., Solomon, D., Lawson, H.W., Killackey, M., Kulasingam, S.L., Cain, J. , Garcia, 
F.A.R. , Moriarty, A .T., Waxman, A.G., Wilbur, D.e., Wentzensen, N. , Downs, L.S. , Spitzer, M ., 
Moscicki , A.-B., Franco, E.L., Stoler, M.H., Schiffman, M., Castle, P.E., Myers, E.R., and ACS-
ASCCP-ASCP Cervical Cancer Guideline Committee (2012). American Cancer Society, 
American Society for Colposcopy and Cervical Pathology, and American Society for Clinical 
Pathology screening guidelines for the prevention and early detection of cervical cancer. CA: A 
Cancer Journalfor Clinicians, 62, 147- 172. 
Scarinci , I.e., Garcia, F.A.R., Kobetz, E., Partridge, E.E., Brandt, H .M., Bell, M.e. , Dignan, M., 
Ma, G.X. , Daye, J.L. , and Castle, P.E. (2010a). Cervical cancer prevention. Cancer, NA-NA. 
Scarinci , I.e., Garcia, F.A.R., Kobetz, E. , Partridge, E.E., Brandt, H.M., Bell, M.e. , Dignan, M. , 
Ma, G.x. , Daye, J.L. , and Castle, P.E. (2010b). Cervical cancer prevention: New tools and old 
barriers. Cancer, NA-NA. 
Scarinci , I.e., Litton, A.G ., Garces-Palacio, I.e., Partridge, E.E., and Castle, P.E. (2013) . 
Acceptability and Usability of Self-Collected Sampling for HPV Testing Among African-
American Women Living in the Mississippi Delta. Women 's Health Issues, 23, e 123-e 130. 
200 
University of Ghana  http://ugspace.ug.edu.gh
Schellenbacher, e., Roden, R., and Kirnbauer, R. (2009) . Chimeric Ll-L2 Virus-Like Particles 
as Potential Broad-Spectrum Human Papillomavirus Vaccines. Journal oj Virology, 83, 10085-
10095. 
Schiffman, M., and Castle, P.E. (2005). The Promise of Global Cervical-Cancer Prevention. New 
England Journal ojM edicine, 353, 2101 - 2104. 
Schiffman, M., and Kjaer, S.K. (2003). Chapter 2: Natural History of Anogenital Human 
Papillomavirus Infection and Neoplasia. JNCI Monographs, 2003, 14-19. 
Schiffinan, M., and Wacholder, S. (2009). From India to the World - A Better Way to Prevent 
Cervical Cancer. New England Journal ojM edicine, 360, 1453- 1455. 
Schmeink, C.E., Bekkers, R.L.M., Massuger, L.F.A.G., and Melchers, WJ.G. (2011). The 
potential role of self-sampling for high-risk human papillomavirus detection in cervical cancer 
screening: Role of self-sampling for hr-HPV detection. Reviews in Medical Virology, 21, 139-
153. 
Schmitt, M., Oepuydt, e., Benoy, 1., Bogers, J., Antoine, J., Arbyn, M., Pawlita, M., and on 
behalf of the VALGENT Study Group (2013). Prevalence and viral load of 51 genital human 
papillomavirus types and three subtypes. International Journal ojCancer, 132,2395-2403. 
Seedorf, K. , Krammer, G., DUrst, M., Suhai, S., and R6wekamp, W.O. (1985). Human 
papillomavirus type 16 DNA sequence. Virology, 145, 181 - 185. 
Sellors, J.W., Sankaranarayanan, R., and International Agency for Research on Cancer (2003). 
Colposcopy and treatment of cervical intraepithelial neoplasia: a beginner's manual (Lyon, 
France: International Agency jor Research on Cancer). 
Shepherd, J. , Peersman, G. , Weston, R., and Napuli, I. (2000). Cervical cancer and sexual 
lifestyle: a systematic review of health education interventions targeted at women . Health 
Education Research, 15, 681-694. 
Sh ie lds, T.S., Brinton, L.A ., Burk, R.D., Wang, S.S., Weinstein, SJ., Ziegler, R.G. , Studentsov, 
Y.Y ., McAdams, M., and Schiffman, M. (2004). A Case-Control Study of Risk Factors for 
Invasive Cervical Cancer among U.S. Women Exposed to Oncogenic Types of Human 
Papillomavirus. Cancer Epidemiology Biomarkers & Prevention, 13, 1574- 1582. 
Sh ing leton, H.M., Bell, M.e., Fremgen, A., Chmiel, J.S., Russell, A.H. , Jones, W.B. , 
Winchester, D.P., and Clive, R.E. (1995). Is there really a difference in survival of women with 
sq uamous cell carcinoma, adenocarcinoma, and adenosquamous cell carcinoma of the cervix? 
Cancer, 76, 1948- 1955. 
Smith, J.S. , Herrero, R., Bosetti , e., Munoz, N. , Bosch, F.X., Eluf-Neto, J. , Castellsague, X. , 
Meijer, CJ.L.M., Brule, AJ.e.V. den, Franceschi, S., and Ashley, R. (2002a). Herpes Simplex 
201 
University of Ghana  http://ugspace.ug.edu.gh
Tamura, K. , Nei , M., and Kumar, S. (2004). Prospects for inferring very large phylogenies by 
using the neighbor-joining method. Proceedings oft he National Academy ofS ciences oft he 
United States ofA merica, 101, 11030-11035. 
Tamura, K. , Peterson, D., Peterson, N. , Stecher, G. , Nei , M., and Kumar, S. (2011). MEGA5 : 
molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and 
max imum parsimony methods. Molecular biology and evolution, 28, 2731-2739. 
The International Collaboration of Epidemiological Studies of Cervical Cancer (2007). 
Comparison of risk factors for invasive squamous cell carcinoma and adenocarcinoma of the 
cervix: Collaborative reanalysis of individual data on 8,097 women with squamous cell 
carcinoma and 1,374 women with adenocarcinoma from 12 epidemiological studies. 
International Journal ofC ancer, 120, 885- 891. 
Thomas, J .O., Herrero, R. , Omigbodun, A.A., Ojemakinde, K. , Ajayi, 1.0., Fawole, A. , Oladepo, 
0., Smith, J.S. , Arslan, A. , Munoz, N. , Snijders, PJ .F. , Meijer, C.J.L.M., and Franceschi, S. 
(2004). Prevalence of papillomav irus infection in women in Ibadan, Nigeria: a population-based 
study. British Journal ofC ancer, 90, 638- 645. 
Tommasino, M., Accardi , R., Caldeira, S., Dong, W., Malanchi, I., Smet, A., and Zehbe, 1. 
(2003). The role ofTP53 in Cervical carcinogenesis. Human Mutation, 21,307-312. 
Tornesello, M.L. , Losito, S., Benincasa, G ., Fulciniti , F., Botti , G. , Greggi, S. , Buonaguro, L. , 
and Buonaguro, F.M. (2011). Human papillomavirus (HPV) genotypes and HPVI6 variants and 
risk of adenocarcinoma and squamous cell carcinoma of the cervix. Gynecologic Oncology, 121, 
32-42. 
Trott ier, H. (2006). Human Papillomavirus Infections with Multiple Types and Risk of Cervical 
Neoplasia. Cancer Epidemiology Biomarkers & Prevention, 15, 1274-1280. 
Vaccarella, S., Herrero, R ., Dai, M ., Snijders, PJ.F., Meijer, CJ .L.M., Thomas, J.O., Hoang 
Anh, P.T. , Ferreccio, c., Matos, E., Posso, H ., de Sanjose, S., Shin, H.-R., Sukvirach, S., 
Lazcano-Ponce, E., Ronco, G., Rajkumar, R. , Qiao, Y.-L. , Munoz, N. , and Franceschi, S. (2006). 
Reproductive factors, oral contraceptive use, and human papillomavirus infection: pooled 
analysis of the IARC HPV prevalence surveys. Cancer epidemiology, biomarkers & prevention: 
a publication oft he American Associationfor Cancer Research, cosponsored by the American 
Society ofP reventive Oncology, 15, 2148- 2153. 
Vaccarella, S., Herrero, R., Snijders, PJ.F., Dai, M., Thomas, J.O., Hieu, N.T. , Ferreccio, c., 
Matos, E., Posso, H., de Sanjose, S., Shin, H.R., Sukvirach, S., Lazcano-Ponce, E., Munoz, N. , 
Meijer, CJ.L.M ., Franceschi, S., and the IARC HPV Prevalence Surveys (IHPS) Study Group 
(2008). Smoking and human papillomavirus infection: pooled analysis of the International 
Agency for Research on Cancer HPV Prevalence Surveys. International Journal of 
Epidemiology, 3 7, 536-546. 
203 
University of Ghana  http://ugspace.ug.edu.gh
Virus-2 as a Human Papillomavirus Cofactor in the Etiology of Invasive Cervical Cancer. 
Journal oft he National Cancer Institute, 94, 1604- 1613. 
Smith, l .S., Munoz, N. , Herrero, R. , Eluf-Neto, l ., Ngelangel, c., Franceschi , S., Bosch, F.X., 
Walboomers, J.M.M., and Peeling, R.W. (2002b). Evidence for Chlamydia trachomatis as a 
Human Papillomavirus Cofactor in the Etiology of Invasive Cervical Cancer in Brazil and the 
Philippines. Journal ofI nfectious Diseases, 185, 324- 331. 
Smith, l.S., Green, J., de Gonzalez, A.B., Appleby, P., Peto, l. , Plummer, M., Franceschi, S., and 
Beral, V. (2003). Cervical cancer and use ofhonnonal contraceptives: a systematic review. The 
Lancet, 361, 1159-1167. 
Smith, J.S. , Bosetti, c., Munoz, N., Herrero, R., Bosch, F.X., Eluf-Neto, J., Meijer, C.J .L.M., van 
den Brule, AJ.C., Franceschi, S., and Peeling, R.W. (2004). Chlamydia trachomatis and invasive 
cervical cancer: A pooled analysis of the lARC multicentric case-control study. International 
Journal of Cancer, 111, 431-439. 
Snijders, PJ .F., Verhoef, V.MJ., Arbyn, M., Ogilvie, G., Minozzi, S., Banzi, R., van Kemenade, 
FJ ., Heideman, D.A.M ., and Meijer, C.J.L.M . (2013) . High-risk HPV testing on self-sampled 
versus clinician-collected specimens: A review on the clinical accuracy and impact on population 
attendance in cervical cancer screening. International Journal ofC ancer, 132, 2223- 2236. 
Soderlund-Strand, A., Eklund, C., Kemetli, L., Grillner, L., Tornberg, S., Dillner, J. , and Dillner, 
L. (2011) . Genotyping of human papillomavirus in triaging oflow-grade cervical cytology. 
American Journal of Obstetrics and Gynecology, 205, 145.e 1-145.e6. 
Solomon, D., Schiffman, M., Tarone, R., and ALTS Study group (2001). Comparison of three 
management strategies for patients with atypical squamous cells of undetermined significance: 
baseline results from a randomized trial. Journal oft he National Cancer Institute, 93, 293- 299. 
Stewart, D.E., Gagliardi, A., Johnston, M., Howlett, R. , Barata, P. , Lewis, N., Oliver, T. , Mai , 
V., and HPV Self-collection Guidelines Panel (2007). Self-collected samples for testing of 
oncogenic human papillomavirlls: a systematic review. Journal ofo bstetrics and gynaecology 
Canada: JOGC = Journal d'obstetrique et gynecologie du Canada: JOGC, 29, 817- 828. 
Storey, A. , Thomas, M., Kalita, A., Harwood, C., Gardiol , D., Mantovani, F., Breuer, J. , Leigh, 
I.M., Matlashewski, G. , and Banks, L. (1998). Role of a p53 polymorphism in the development 
of human papillomavirus-associated cancer. Nature, 393, 229- 234. 
Straub, E., Dreer, M., Fertey, l., lfiner, T., and Stubenrauch, F. (2014). The Viral E81\E2C 
Repressor Limits Productive Replication of Human Papillomavirus 16. Journal of Virology, 88, 
937- 947. 
Straub, E., Fertey, J., Dreer, M., lfiner, T. , and Stubenrallch, F. (2015). Characterization ofthe 
Human Papillomavirus 16 E8 Promoter. Journal of Virology, 89, 7304- 7313. 
202 
University of Ghana  http://ugspace.ug.edu.gh
Van Calsteren, K. , Vergote, I., and Amant, F. (2005). Cervical neoplasia during pregnancy: 
Diagnosis, management and prognosis. Best Practice & Research Clinical Obstetrics & 
Gynaecology, 19, 611-630. 
Van de Velde, N., Boily, M.-C., Drolet, M., Franco, E.L., Mayrand, M.-H., Kliewer, E.V., 
Coutlee, F., Laprise, J.-F., Malagon, T., and Brisson, M. (2012). Population-Level Impact of the 
Bivalent, Quadrivalent, and Nonavalent Human Papillomavirus Vaccines: A Model-Based 
Analysis. JNCI Journal oft he National Cancer Institute, 104, 1712- 1723. 
Venuti , A., Paolini, F., Nasir, L., Corteggio, A., Roperto, S., Campo, M.S., and Borzacchiello, G. 
(2011). Papillomavirus E5: the smallest oncoprotein with many functions. Molecular Cancer, 10, 
140. 
Veroux, M., Corona, D., Scalia, G., Garozzo, V., Gagliano, M., Giuffrida, G., Costanzo, C.M., 
Giaquinta, A. , Palermo, I. , Zappala, D., Tallarita, T. , Zerbo, D., Russo, R., Cappellani, A., 
Franchina, c., Scriffignano, V. , and Veroux, P. (2009). Surveillance of Human Papilloma Virus 
Infection and Cervical Cancer in Kidney Transplant Recipients: Preliminary Data. 
Transplantation Proceedings, 41, 1191-1194. 
de Villiers, E.-M., and Gunst, K. (2009). Characterization of seven novel human papillomavirus 
types isolated from cutaneous tissue, but also present in mucosal lesions. Journal ofG enera I 
Virology, 90, 1999- 2004. 
de Villiers, E.-M., Fauquet, C., Broker, T.R., Bernard, H.-U., and zur Hausen, H. (2004). 
Classification of papillomav iruses. Virology, 324, 17-27. 
Virtanen, A., Nieminen, P. , Luostarinen, T. , and Anttila, A. (2011). Self-sample HPV Tests As 
an Intervention for Nonattendees of Cervical Cancer Screening in Finland: a Randomized Trial. 
Cancer Epidemiology Biomarkers & Prevention, 20, 1960- 1969. 
Walboomers, J.M ., Jacobs, M.V., Manos, M.M., Bosch, F.X., Kummer, J.A. , Shah, K.V., 
Snijders, PJ ., Peto, J., Meij er, CJ., and Munoz, N. (1999) . Human papillomavirus is a necessary 
cause of invasive cervical cancer worldwide. The Journal ofp athology, 189, 12- 19. 
Wall , S.R., Scherf, C.F., Morison, L., Hart, K.W., West, B., Ekpo, G. , Fiander, A.N., Man, S., 
Gelder, C.M., Walraven, G., and Borysiewicz, L.K. (2005). Cervical human papillomavirus 
infection and squamous intraepitheliallesions in rural Gambia, West Africa: viral sequence 
analys is and epidemiology. British Journal ofC ancer, 93, 1068- 1076. 
Wang, J.W., and Roden, R.B.S. (2013). Virus-like particles for the prevention of human 
papi Ilomavirus-associated malignancies. Expert Review of Vaccines, 12, 129- 141 . 
Wang, S.S., and Hildesheim, A. (2003). Chapter 5: Viral and Host Factors in Human 
Papillomavirus Persistence and Progression. JNC1 Monographs, 2003, 35-40. 
204 
University of Ghana  http://ugspace.ug.edu.gh
Wang, S.S., Sherman, M.E., Hildesheim, A., Lacey, J.V. , and Devesa, S. (2004). Cervical 
adenocarcinoma and squamous cell carcinoma incidence trends among white women and black 
women in the United States for 1976-2000. Cancer, 100, 1035-1044. 
Wang, S.S., Gonzalez, P., Yu, K., Porras, c., Li, Q., Safaeian, M., Rodriguez, A.C., Sherman, 
M.E., Bratti, c., Schiffman, M., Wacholder, S., Burk, R.D., Herrero, R., Chanock, SJ., and 
Hildesheim, A. (2010). Common Genetic Variants and Risk for HPV Persistence and 
Progression to Cervical Cancer. PLoS ONE, 5, e8667. 
Watson-Jones, D., Baisley, K., Ponsiano, R., Lemme, F., Remes, P., Ross, D., Kapiga, S., 
Mayaud, P., de Sanjose, S., Wight, D., Changalucha, J., and Hayes, R. (2012). Human 
Papillomavirus Vaccination in Tanzanian Schoolgirls: Cluster-Randomized Trial Comparing 2 
Vaccine-Delivery Strategies. Journal 0/I nfectious Diseases, 206, 678-686. 
WHO; Chronic Diseases and Health Promotion Group (2006). Comprehensive cervical cancer 
control: a guide to essential practice (Geneva: World Health Organization). 
WHO; Guideline Development Group (2014). Comprehensive cervical cancer control: a guide to 
essential practice (Geneva: World Health Organization). 
WHO/ICO Information Centre on Human Papillomavirus and Cervical Cancer (2010). Human 
Papillomavirus and Related Cancers in Ghana. 
Winer, R.L., Hughes, J.P., Feng, Q. , O'Reilly, S., Kiviat, N.B., Holmes, K.K. , and Koutsky, L.A. 
(2006). Condom Use and the Risk of Genital Human Papillomavirus Infection in Young Women. 
New England Journal o/Medicine, 354, 2645- 2654. 
Wiredu, E.K., and Armah, H.B. (2006). Cancer mortality patterns in Ghana: a 10-year review of 
autopsies and hospital mortality. BMC public health, 6, 159. 
Woodman, C.BJ., Coll ins, S.L, and Young, L.S. (2007). The natural history of cervical HPV 
infection: unresolved issues. Nature Reviews Cancer, 7, 11-22. 
Wu, E.-Q., Liu, B., Cui, J.-F., Chen, W., Wang, J.-B., Lu, L. , Niyazi, M., Zhao, C., Ren, S.-D., 
Li, C.-Q., Gong, x.-Z., Smith, J.S., Belinson, J.L., Liaw, K.-L., Velicer, C., and Qiao, Y.-L. 
(2013). Prevalence of type-specific human papillomavirus and pap results in Chinese women: a 
multi-center, population-based cross-sectional study. Cancer Causes & Control, 24, 795-803 . 
Xi, L.F., Toure, P., Critchlow, C.W., Hawes, S.E., Dembele, B., Sow, P.S., and Kiviat, N.B. 
(2003). Prevalence of specific types of human papillomavirus and cervical squamous 
intraepithelial lesions in consecutive, previously unscreened, West-African women over 35 years 
of age. International Journal o/Cancer, 103, 803-809. 
Ye, J., Cheng, X., Chen, X., Ye, F., Lu, W., and Xie, X. (2010). Prevalence and risk profile of 
cervical human papillomavirus infection in Zhejiang Province, southeast China: a population-
based study. Virology Journal, 7, 66. 
205 
University of Ghana  http://ugspace.ug.edu.gh
Zhao, F.-H., Lin, MJ., Chen, F., Hu, S.-Y., Zhang, R. , Belinson, J.L., Sellors, J.W., Franceschi, 
S., Qiao, Y.-L., and Castle, P.E. (2010). Performance of high-risk human papillomavirus DNA 
testing as a primary screen for cervical cancer: a pooled analysis of individual patient data from 
17 population-based studies from China. The Lancet Oncology, 11, 1160-1171. 
Zhao, F.-H., Lewkowitz, A.K., Chen, F., Lin, MJ., Hu, S.-Y., Zhang, X., Pan, Q.-J., Ma, J.-F., 
Niyazi, M., Li, C.-Q., Li, S.-M., Smith, J.S., Belinson, J.L., Qiao, Y.-L., and Castle, P.E. (2012). 
Pooled Analysis of a Self-Sampling HPY DNA Test as a Cervical Cancer Primary Screening 
Method. JNCI Journal oft he National Cancer Institute, 104, 178-188. 
Zubach, Y., Smart, G., Ratnam, S., and Severini, A. (2011). Novel Microsphere-Based Method 
for Detection and Typing of 46 Mucosal Human Papillomavirus Types. Journal afClinical 
Microbiology, 50, 460-464. 
206 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX I 
Table A.I : Primer sequence for round one genotyping PCR amplification 
Primer designation Primer sequence 
PGMYII-A GCA CAG GGA CAT AAC AAT GG 
PGMYII-B GCG CAG GGC CAC AAT  AAT  GG 
PGMYII-C GCA CAG GGA CAT AAT AAT GG 
PGMYII-D GCC CAG GGC CAC AAC AA T GG 
PGMYII-E GCT CAG GGT ITA AAC AAT GG 
PGMY09-F CGT CCC AAA GGA AAC TGA TC 
PGMY09-G CGA CCT AAA GGA AAC TGA TC 
PGMY09-H CGT CCA AAA GGA AAC TGA TC 
PGMY09-Ia G CCA AGG GGA AAC TGA TC 
PGMY09-J CGT CCC AAA GGA T AC TGA TC 
PGMY09-K CGT CCA AGG GGA TAC TGA TC 
PGMY09-L CGA CCT AAA GGG AA T TGA TC 
PGMY09-M CGA CCT AGT GGA AA T TGA TC 
PGMY09-N CGA CCA AGG GGA TAT TGA TC 
PGMY09-Pa G CCC AAC GGA AAC TGA TC 
PGMY09-Q CGACCCAAGGGAAACTGGTC 
PGMY09-R CGT CCT AAA GGA AAC TGG TC 
HMBOlb GCG ACC CAA TGC AAA TTG GT 
207 
University of Ghana  http://ugspace.ug.edu.gh
Table A.2: Geographic origin, lineage designation, length of complete genome and NCBl 
accession number of HPY types used for the alignment of the sequence obtained in this study. 
Type Isolate Lineage Sublineage NCBI Accession # Length GC% 
HPYI6Ref E (Prototype I ) NCOOl526 7906 0.365 
WOl22b E (Prototype2 ) AF536179 7904 0.368 
Qv02706b E Asian HQ644261 7905 0.366 
BF325b Af Af-I HQ644240 7908 0.363 
Rw918b Af Af-I HQ644293 7908 0.363 
BF236b Af Af-2 HQ644239 7905 0.364 
HPYI6 Rw862b Af Af-2 HQ644292 7904 0.364 
Af Af-2a KF466626 
Af AFRla KF466576 
Af AFRlb KF466592 
Af AFR2b KF466652. 
Rw677b Asian-American AAI HQ644289 7906 0.365 
Qv03545c Asian-American AA2 HQ644263 7906 0.366 
Ref A AI AY262282 7857 40.44 
Qv29226 A AI KC470208 7857 40.41 
Qv32981 A A2 KC470210 7857 40.38 
Qv26861 A A3 KC470212 7857 40.32 
ZI35 A A4 KC470213 7857 40.35 
HPYI8 
CUll A AS GQ180787 7844 40.29 
Rw750 B BI KC470217 7824 40.04 
BF309 B B2 KC470223 7824 40 .07 
Bf380 B B3 KC470228 7844 40 .01 
Qv39775 C KC470229 7837 40 . 14 
Ref A AI X74479 7858 39.63 
Qv20214 A AI EF202156 7858 39.63 
Z79 A AI KC470250 7858 39.60 
HPV45 BF208 A A2 KC470255 7848 39.73 
BFI34 A A3 KC470256 7841 39.64 
Qv06560 B BI EF202163 7841 39.79 
Qv34163 B B2 KC470260 7848 39.73 
208 
University of Ghana  http://ugspace.ug.edu.gh
Isolatc numbcr prc_x dcnotcs location from which samplc was obtained. as follows . AS - Taiwan [I], BF - Burkina 
Faso [2]. IN/ INJP - Thailand [3]. Qv- Costa Rica [4]. R - USA (IIAPI study) [5]. Rw - Rwanda [6]. W - USA 
(WIllS study) [7] , Z - Zambia [8]. 
209 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX II 
SCHOOL OF PUBLIC HEALTH, COLLEGE OF HEALTH SCIENCES 
UNIVERSITY OF GHANA 
A Cross-Sectional Study 
Title: Human Papillomavirus Infection and Cervical Lesions among Women in a High Risk 
Population; in 2 sub-district of the Lower Manya Krobo District, Ghana. 
QUESTIONNAIRE A 
For participant during household survey 
I. Identification No. Community :_ __________  
2. House No. Sampling house ID: _________  ----------
3. Age: ____________________  Religion_ ____________  
4a. Marital status: 
o Married o Unmarried o Cohabiting o in a stable relationship 
4b. If married, are you in a o monogamous o polygamous 
5. Occupation _________________________  
6. Educational status: 
o No formal education o primary o junior secondary Dsenior secondary 
Otertiary 
7. Age at first sexual intercourse ____________ _ 
8. How many male sexual partners have you had since becoming sexually active?_ _____ _ 
9. How many male sexual partners do you currently have?_ ___________  
10. Number of pregnancies ________________________ _ 
11 a. Have you ever had any abortion? D Yes O No 
I I b. If yes, please specify number_ ____ _ 
12a. Have you ever had any miscarriage? 0 Yes o No, 
210 
University of Ghana  http://ugspace.ug.edu.gh
12b: If Yes, please specify number_ __  
13. Do you have menstrual difficulties? D Yes D No 
14. Have you had any sexually transmitted infection within the past 10 years? 
D Yes D No 
15. If yes, please select which, Dsyphilis D gonorrhoea Dcandidacies 
D HIV D others 
16. How often do you use condoms during sex? 
D Every time D Most often D Intermittently D Less frequently D Not at all 
17. Do you use oral contraceptives? D Yes D No 
17b: If yes, how long have you been using it? _________________  
18. How often do you take alcohol? 
D Every day D only at occasions D Not often D Not at all 
19. Do you smoke? D Yes D No 
19b: if yes, how regularly do you smoke? 
D Everyday D Once a week D Once a while 
20. Have you ever heard of cervical cancer? D Yes D No 
If yes, do know about the cause D Yes, D No 
If Yes, please specify the cause_ ________________  
21. Do you known how it could be prevented D Yes D No 
If yes please specify _____________________ _ 
22. Have you ever been tested by a Pap smear? D Yes D No 
If yes, when where ___________  
Thank you very much for participating in this study 
211 
University of Ghana  http://ugspace.ug.edu.gh
QUESTIONNAIRE B 
For Researcher at Sample Collection 
A. Participant ' s Identification No. ____  Community: ----------
B. House No. ------ ---- Sampling house ID: ---------
Date for participant to see Doctor_ __________  
24. Reported to Hospital D Yes DNo Date ---------
IfNo, why? __________________________ _ 
25. Been seen by Doctor for examination D Yes DNo Date --- ------
IfNo, why? ________________________ ___  
26. Samples have been collected D Yes DNo Date ------ ---
IfNo, why? _______ _________________ ___  
212 
University of Ghana  http://ugspace.ug.edu.gh
QUESTIONNAIRE C 
Self-Sampling for Participant after Sample Collection 
A. Participant's Identification No. Community: _________  
B. House No. Sampling house ID:. _________  
27. What is your opinion about the use of self-sampling devices? 
a. easy to use 
b. difficult to use 
28. Which would you prefer? 
a. take your own samples 
b. health personnel to take your sample 
c. I have no preference 
29. If you choose (a.) in 28 above, why? 
a. issues of privacy 
b. issues of culture belief 
c. issues of religious belief 
d. please specify __________________________  
30. If you choose (b.) in 28 above, why? 
a. have no problem with gynaecologist's examination 
b. self-sampling device is not user-friendly 
c. I forgot how to use it 
d. I was afraid I may hurt myself 
e. I lost the device 
f. other reasons please specify ____________________ _ 
213 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX III 
CONSENT FORM 
Title: Human Papillomavirus Infection and Cervical Lesions amongst Women in a High Risk 
Population in 2 District in the Eastern Region, Ghana 
Principal Investigator: Adolf Kofi Awua 
Address: Department of Epidemiology 
School of Public Health 
College of Health Sciences 
University of Ghana 
P.O. BOX LG 13 
Legon-Accra 
Introduction 
This Consent Form contains information about the research named above which is planned to run 
for three years. In order to be sure that you are informed about being in this research, we are 
asking you to read (or have it read to you) this Consent Form. You will also be asked to sign it 
(or make your mark in front of a witness). We will give you a copy of this fornl. This consent 
form might contain some words that are unfamiliar to you. Please ask us to explain anything you 
may not understand. 
Reason for the Research 
214 
University of Ghana  http://ugspace.ug.edu.gh
You are being asked to take part in research to obtain data on the circulating or prevalence of 
Human Papillomavirus and cervical lesion in this district and evaluate the usefulness of a 
combination HPV testing and Pap smear testing for cervical cancer screening]. 
General Information about Research 
Healthy women between the ages of 15 and 65 years would be asked to report to a health centre 
where cervical cell would be obtained from them using a cytobrush or cotton swab by routine 
method and a suspension in Phosphate Buffered Saline (PBS) prepared. These cell suspensions 
will be further analysed for HPV infection by genetic methods (PCR) and cytologic lesion by 
microscopy. Cervical swab samples will be stored for further research purposes in the areas of 
genotyping, HPV variant analysis and testing newer HPV detection methods when the need 
arises 
Your Part in the Research 
If you agree to be in the research, you will be interviewed and you will fill out a questionnaire 
that will seek for your demographic data, education background, sexual behaviour, reproductive 
health status and current knowledge of cervical cancer. You are to note that some of the 
questions are about your private and/or sex life and might be embarrassing and you may not have 
to answer any question if you do not want to. You will be asked to report to a specific health 
centre were cervical cell will be obtained from you by a trained health personnel and also, you 
would be directed as to how to obtain cervical cells by yourself both using cotton swabs or 
cytobrush by routine method. Your part in the research will last for two days. About (500 
women) will take part in this research in two of districts of the Eastern Region of Ghana 
Possible Risks 
No physical, social and psychological risks are anticipated for participating in this study.) 
215 
University of Ghana  http://ugspace.ug.edu.gh
[When applicable: These risks would be the same whether you got tested in this research or were 
tested without taking part in the research. 
Possible Benefits 
Based on the results of the test, subjects who would need to go for further test and or treatment 
would be advised accordingly. The results of the research will be published so as to make it 
available to everybody to whom the information will be useful including national and district 
level policy makers. 
If You Decide Not to Be in the Research 
You are free to decide if you want to be in this research. Your decision will not affect the health 
care you would normally receive. 
Confidentiality 
We will protect information about you and your taking part in this research to the best of our 
ability. Your answered questionnaires and samples will be coded. These codes will only be 
known by the IP for the sake of finding you if needed be. You will not be named in any reports. 
However, the staff of the School of Public Health, University of Ghana, Noguchi Memorial 
institute for Medical Research and the Ghana Atomic Energy Commission may sometimes look 
at your research records. Someone from the IRS might want to ask you questions about being in 
the research, but you do not have to answer them. A court of law could order medical records 
shown to other people, but that is unlikely. 
[When applicable: If you miss a scheduled visit, we may contact you at home by phone, mail or 
in person to schedule another visit and to see if you still want to take part in the research. When 
this contact is made you will not be identified as being in this research .] 
216 
University of Ghana  http://ugspace.ug.edu.gh
Compensation 
You will not be paid, since you do not have to take part in this research. 
Staying in the Research 
When applicable: If you decide to take part in this research, we ask you to use only the research 
(cotton swab or cytobrush that we provide 
Alternatives to Participation 
You do not have to participate in the research in order to receive care or treatments. 
Leaving the Research 
You may leave the research at any time. If you choose to take part, you can change your mind at 
any time and withdraw 
[When applicable] Ifso, please tell the research doctorlclinic staff why you wish to leave] 
Also, you may be asked to leave the research if (list applicable points): 
• the research doctor/clinic staff feels it is best for you, or 
• you are not able to follow the research procedures, or 
• the research is stopped. 
[When applicable] we will tell you if we learn something new about (the research product or 
drug) that could affect your choice to stay in the research. When you are no longer in the 
research, you will still be able to use this clinic' 
If You Have a Problem or Have Other Questions 
Please call Professor R Adanu, (0244238556) if 
• You get sick, or 
• have concerns about your health (become infected), or 
• have questions about the research 
If you are sick or have a health problem due to your participation in this research, you will not 
have to pay for visits to see the research doctor/clinic staff. 
217 
University of Ghana  http://ugspace.ug.edu.gh
VOLUNTEER AGREEMENT FORM 
The above document describing the benefits, risks and procedures for the research title (Human 
Papillomavirus Infection and Cervical Lesions amongst Healthy Women Living within a High 
Risk Population in Ghana) has been read and explained to me. I have been given an opportunity 
to have any questions about the research answered to my satisfaction. I agree to participate as a 
volunteer. 
Date Signature of volunteer 
Date Thumbprint of volunteer 
If volunteers cannot read the form themselves, a witness must sign here: 
I was present while the benefits, risks and procedures were read to the volunteer. All questions 
were answered and the volunteer has agreed to take part in the research. 
Date Signature of witness 
218 
University of Ghana  http://ugspace.ug.edu.gh
Date Thumbprint of witness 
I certify that the nature and purpose, the potential benefits, and possible risks associated with 
participating in this research have been explained to the above individual. 
Date Signature of Person Who Obtained Consent 
, "-" 
[ 0_ , ..... c . --, .- .-". .  ,' .'-.,  IC c:~,_ : 
-'- '-- , ~: ." ( 
. ,- ,-'rrr.:-.UC 
o ~ y 
219 
University of Ghana  http://ugspace.ug.edu.gh
APPENDIX IV 
SEQUENCE DATA 
70 SO g=- 100 110 120' 130 
T T A G TT GGCC T T AAAA G TTT AAA CC T T A T G CC AAA T A T G C AA T T A GG T T A G T T AAAA C AA G CC AAAAA T 
140 1 !SO l eo, 1 70 "'ISO 1QO 200 
A T G T G CC T AA C A G C GG T A TT T AA GG C G TT GG C G C A T A G T AA T T T A TT T T A T A T G A C A C AA T G T A C A T A 
2 1 0 220 2~O :240 250 260 270 
G T O A T T C A G T A G T T G C A C A T A G T G C A G T G T AAAAAA C AA T GG AA T GG T T GG C AA G C A G T G C A GG T C A G 
2S0 200 300 310 320 330 
G AAAA C A GGG A TT T OO C A C O C A T OG C AA G C A GG AAA C O T A C AA G T T T AAA C T A T A O TT G C T O A C A T A O A 
Figure A.I: Nucleotide seuqence chromatogram (trace) of the partial LCR gene of the HPY 16 Isolate GH609 
220 
University of Ghana  http://ugspace.ug.edu.gh
to:) t 10 120 130 40 5-3 160 70 180 
T T CC T G JCCT GCK: T GC TT C-CCAACCA TCCA T G TTTTT T PCPC G C IoC T A T G T G C AAC TA C T G AAT CAC T A T G T A C A T T G T G C A T A T AAA 
l G3 200 2 10 220 230 240 250 260 270 
A T AAA T-A C T A T G C GCCAACGCC TT AAAT  ACCC-C T G T T A GGC ;'C A T A T TTTT G GC TTG TT TTAAC T AACC T AA T T GC A T AT TT GG C A T AAG 
230 2"3 300 310 320 330 340 350 350 
G TTTAAAC TT TTAAGGCCAAC T AAA G T C A CC TTI>I3 TTCA T A C A T G AAC T G T G T AAAGGTTA.'l T C A T A C A TTA TTC A TTTG T A C A GC TG C 
Figure A.2: Nucleotide seuqence chromatogram (trace) ofthe partial LCR gene of the HPY 16 Isolate G HSI7 
221 
University of Ghana  http://ugspace.ug.edu.gh
70 SO gO 10;) 1 "10 "120 130 
T T A A A C C A - T G G C G C G C C T C T T T G G C G C A T A C A A G G C G C A C C T G G T A T T A G T C A T T T T C C T G T C C A G G T 
140 150 1e;) 170 l S0 1 1000 2lX1 
G C G C T A C AA C AA T GG C T T G C A C AA C T A T C T CC A C T CCC T A T G T AA T AAAA C T G C T TT T A GG C A C A T A T T 
:2 1 a 2:20 :230 :240 :2 50 :2 eo :270 
T T A G T C T G TT T TT A C TT AA G C T AA T T G T A T AA TT GG C TT G T A C AA C T A C TT T C A T G T CC AA C A T T C T G T 
:280 2,,0 300 3 1 0 320 330 3"10 
C T A CCC T T AA C A T G AA C T A T AA T A T G A C T AA G C T G T G C A T A C A T A G T T T A T G C AA CC G AAA T A GG TT G 
Figure A.3: Nucleotide seuqence chromatogram (trace) of the partial LCR gene or the HPV 18 Isolate GJ--\56 I 
222 
University of Ghana  http://ugspace.ug.edu.gh
eO "?.;J: 1.30 !iO 20C 210 220 23C 
G TT ~ T A G C G C ACC T GG AC AG G AAAA T O AC T AA T A CC AG G T O C G CC TT G T A T G C G CC AAA G AG G C G C G CC AA T GG TT T AAA 
,..i.e 250 2e.J 27"C 2.30 2~O 300 ~ 0 
A T A T G T A O AA G C AG T O CCC AAAA G A T T AA G T T T T G C AA T A G T G CC A G T G T A C T G T A T T G T G C A C AAAA CC A C A G A C A T AA G 
f\ 
32C ~ 30) 3 ·~O J-SoO 300 ·370 lao 3.0 
CC AAAGG C AA CC GAAA T C GO T T G C A C AG C AAAA T GG A GG A TT G T AGGAT AAAA T GG A T G C T G T AA GG T G T G C A G T T T T A 
... 00 41C "'2'0 .;4.;!C ,&40 ... 50 ~rSo .c.7C 
T A A C T T Q A .... A T A C AGo G A C A A T A T .... T A Go C C C AA C AA ~ C A A C A Ce C A T A ce A C A AA C A C A T A A ':;' C C AAA Q Q C A A C C ~ A .... 
Figur A.4: Nucleotide seuqence chromatogram (trace) of the partial LCR gene of the H PV 18 Isolate G H627 
223