UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The epidemiology and natural history of Hepatitis C infection in a cohort of homosexual men (1982-1998) Craib, Kevin Joseph Patrick 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-90170X.pdf [ 6.52MB ]
Metadata
JSON: 831-1.0091731.json
JSON-LD: 831-1.0091731-ld.json
RDF/XML (Pretty): 831-1.0091731-rdf.xml
RDF/JSON: 831-1.0091731-rdf.json
Turtle: 831-1.0091731-turtle.txt
N-Triples: 831-1.0091731-rdf-ntriples.txt
Original Record: 831-1.0091731-source.json
Full Text
831-1.0091731-fulltext.txt
Citation
831-1.0091731.ris

Full Text

THE EPIDEMIOLOGY AND NATURAL HISTORY OF HEPATITIS C INFECTION IN A COHORT OF HOMOSEXUAL MEN (1982-1998) by KEVIN JOSEPH PATRICK CRAIB Bachelor of Arts (Mathematics), The University of British Columbia, 1978 Master of Mathematics (Statistics), The University of Waterloo, 1983  A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE STUDIES Department of Health Care and Epidemiology  We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA April 15,2004 © Kevin J.P. Craib, 2004  ABSTRACT  Objectives: The objectives of this study were: (1) to estimate the prevalence and incidence of hepatitis C virus (HCV) infection among all gay men residing in Vancouver during the period 1982-98, (2) to identify risk factors for H C V infection and predictors of 'time to H C V seropositivity' in this population, and (3) to determine whether coinfection with H C V adversely influences the natural history of HIV infection in coinfected men. Methods: The Vancouver Lymphadenopathy-AIDS Study (VLAS) has monitored a cohort of homosexual men since November 1982. Serum samples were obtained from 932 men during the period 1982-98, and tested for H C V antibody using EIA1, EIA2, and RIBA. HIV-antibody test results were also available. Data regarding demographic variables, sexual practices, substance use, and history of infectious diseases were obtained from self-administered questionnaires completed during 1982-85. Data regarding HIV-related disease progression were also available including clinical symptoms and signs, physical findings, CD4 cell count, diagnosis of AIDS, and survival. Risk factors for H C V infection were assessed for statistical significance using both cross-sectional comparisons of seropositive and seronegative men and prospective analyses of time to H C V seropositivity. Differences in time to HCV seropositivity, AIDS progression and survival were evaluated by stratified Kaplan-Meier analysis, and tested using the log-rank test. Both logistic regression and Cox proportional hazards regression were used to model the simultaneous effect of several variables on outcomes of interest. All p-values were two-sided. Results: A total of 54 of 932 men (5.8%) tested positive for H C V antibody [95% CI: 4.3%, 7.3%]. H C V prevalence was significantly higher among HIV seropositive men compared to HIV seronegative men (8.8% vs. 2.6%; p<0.001). After 14 years of followii  up, cumulative H C V incidence in the cohort was 7.9% [95% CI: 5.1%, 10.7%]. Annual infection rates ranged from 0 to 1.4 percent during the follow-up period. Men who reported using injection drugs during their lifetime were twenty times more likely to be become H C V seropositive. In prospective analyses, significant elevations in risk were detected for the following sexual practices: 20 or more male sexual partners in the previous 12 months [RR = 3.1, 95% CI: 1.5, 6.3], insertive oral-anal contact [RR = 3.1, 95% CI: 1.1, 8.7], insertive fisting (RR = 2.6, 95% CI: 1.4, 4.8), and receptive anal intercourse [RR = 2.0, 95% CI: 1.0, 3.8]. In multivariate analysis, risk factors that exerted an independent effect on time to HCV seropositivity were injection drug use (p<0.001), HIV seropositivity (p=0.031), and the sexual practice of insertive fisting in combination with insertive oral-anal contact (p=0.038). In terms of HIV-related disease progression, H C V co-infection was not significantly associated with an increase in symptomatic illness, more rapid CD4 cell decline, faster progression to AIDS, or increased mortality. However, we did find strong indication of increased liver and spleen inflammation among co-infected men. Conclusions: Prevalence and incidence of H C V infection are elevated among gay men in Vancouver, and are significantly higher among HIV-seropositive men compared to HIV-seronegative men. Not surprisingly, a history of injection drug use was the most significant risk factor for H C V seropositivity in this population. However, these data also provide evidence of sexual transmission of H C V independent of injection drug use. Sexual practices that result in rectal trauma may play a role in the spread of H C V in this population. With regard to disease progression, we did not detect an adverse affect of H C V infection on HIV disease progression among co-infected individuals but we did observe an adverse influence on progression of liver disease.  iii  TABLE OF CONTENTS  ABSTRACT  ii  TABLE OF CONTENTS  iv  LIST OF TABLES  ix  LIST OF FIGURES  xi  ACKNOWLEDGMENTS  xiii  Chapter 1 Introduction  1  1.1  HEPATITIS C - EPIDEMIOLOGY  1  1.2  HEPATITIS C - N A T U R A L HISTORY  4  1.3  RESEARCH NEEDS A N D STUDY JUSTIFICATION  7  1.4  OBJECTIVES, HYPOTHESES, A N D THESIS ORGANIZATION  10  1.5  SUMMARY  13  1.6  REFERENCES  15  Chapter 2 Background  24  2.1  SEXUAL TRANSMISSION OF H C V IN HETEROSEXUALS  24  2.2  HEPATITS C INFECTION IN HOMOSEXUAL M E N  26  2.3  N A T U R A L HISTORY STUDIES OF H C V / H I V CO-INFECTION  29  2.4  THE VANCOUVER LYMPHADENOPATHY-AIDS STUDY  29  2.4.2  Study Description  29  2.4.2  Follow-up of cohort  30  2.4.3  HCV-antibody  31  2.4.4  Data management  32  2.4.5  Description of statistical methods  33  2.4.6  Rationale for using both cohorts  34  testing procedure and results  iv  2.5  REFERENCES  37  Chapter 3 Prevalence and Incidence Rates of HCV Infection in the VLAS Cohort  42  3.1  FORWARD  42  3.2  INTRODUCTION  43  3.3  METHODS  44  3.4  3.3.2  Study Description  3.3.2  HCV Antibody Testing  3.3.3  HIV Antibody Testing  45  3.3.4  Statistical Methods  45  44 ,  RESULTS  44  46  3.4.2  Demographic characteristics  46  3.4.2  HCV prevalence  46  3.4.3  HCV cumulative incidence  46  3.4.4  HCV incidence density estimates  47  3.4.5  Comparison of HCV and HIV incidence  47  3.5  DISCUSSION  48  3.6  REFERENCES  56  Chapter 4 Risk Factors for HCV Infection in the VLAS Cohort  53  4.1  FORWARD  59  4.2  INTRODUCTION  59  4.3  METHODS  ;  61  4.3.2  Study Description  61  4.3.2  Sources of data  61  4.3.3  Statistical Methods  62  RESULTS  63  4.4.2  HCV Prevalence  63  4.4.2  Demographic characteristics  63  4.4.3  Injection drug use  63  4.4.4  Use of psychoactive drugs  63  4.4.5  Use of tobacco and alcohol  64  4.4.6  Sexual practices  64  4.4.7  History of blood transfusion  64  4.4.8  Multivariate analysis  65  DISCUSSION  65  REFERENCES  73  Chapter 5 Sexual Practices Associated with HCV Seropositivity among Noninjection Drug Users in the VLAS Cohort  77  5.1  FORWARD  77  5.2  INTRODUCTION  77  5.3  METHODS  78  5.3.2  Study Description  78  5.3.2  Study Variables  79  5.3.3  Statistical Methods  79  5.4  RESULTS  80  5.4.2  Injection drug use  80  5.4.2  Demographic characteristics  80  5.4.3  Sexual practices  81  5.4.4  Multivariate analysis  81  5.5  DISCUSSION  82  5.6  REFERENCES  86  Chapter 6 Predictors of Time to First Positive HCV Antibody Test in the VLAS cohort  88  6.1  FORWARD  88  6.2  INTRODUCTION  88  6.3  METHODS  90  6.4  6.3.1  Study Description  90  6.3.2  HCV and HIV antibody testing  90  6.3.3  Statistical Methods  91  RESULTS  92  6.4.1  Demographic characteristics  92  6.4.2  HCV cumulative incidence  92  6.4.3  HCV cumulative incidence stratified by injection drug use  92  6.4.4  HCV cumulative incidence stratified by HIV antibody status  93  6.4.5  HCV cumulative incidence stratified by fisting and oral-anal contact  93  6.4.6  Relationship between HCV infection and other sexual practices  6.4.7  Relationship between HCV infection and history of infectious diseases  6.4.8  Multivariate analysis  93 94 94  6.5  DISCUSSION  95  6.6  REFERENCES  104  Chapter 7 HIV Disease Progression and Mortality among Men in the VLAS who are Co-infected with HCV  106  7.1  FORWARD  106  7.2  INTRODUCTION  106  vii  7.3  7.4  METHODS  107  7.3.1  Study Description  107  7.3.2  Sources of data  108  7.3.3  HCV and HIV antibody testing  108  7.3.4  Laboratory methods  109  7.3.5  Eligibility criteria  109  7.3.6  Outcomes of interest  110  7.3.7  Statistical Methods  110  RESULTS_  111  7.4.1  Demographic characteristics  HI  7.4.2  Symptoms and signs  111  7.4.3  Physical findings  111  7.4.4  CD4 cell decline  112  7.4.5  Progression to AIDS  112  7.4.6  Mortality  113  7.5  DISCUSSION  113  7.6  REFERENCES  121  Chapter 8 General Discussion, Conclusions, and Recommendations  124  8.1  S U M M A R Y OFSTUDY FINDINGS.  124  8.2  UNIQUE CONTRIBUTIONS, IMPACT, A N D IMPLICATIONS  126  8.3  STRENGTHS A N D LIMITATIONS  130  .  8.3.1  Strengths  130  8.3.2  Limitations  132  8.4  CONCLUSIONS  135  8.5  RECOMMENDATIONS  136  8.6  REFERENCES  139  viii  LIST OF TABLES  TABLE 3.1  Comparison of VLAS cohorts with respect to demographic characteristics.  TABLE 4.1  50  Demographic characteristics and HIV antibody status of 39 HCVpositive and 623 HCV-negative men  TABLE 4.2  69  Comparison of HCV-positive and HCV-negative participants with respect to self-reported use of injection drugs, psychoactive drugs, tobacco, alcohol, and other substances.  TABLE 4.3  70  Comparison of HCV-positive and HCV-negative participants with respect to number of sexual partners and sexual practices  TABLE 4.4  71  Multivariate logistic regression model of risk factors for H C V infection 72  TABLE 5.1  Comparison of 22 HCV-positive and 833 HCV-negative men who did not report injection drug use with respect to specific sexual practices 85  TABLE 6.1  Predictors of time to first positive H C V antibody test in the VLAS cohort 101  ix  TABLE 6.2  Estimates of relative risk of H C V infection for self-reported history of infectious diseases among 662 men in the VLAS cohort  TABLE 6.3  Multivariate predictors of time to first positive H C V antibody test in the VLAS cohort  TABLE 7.1  102  103  Demographic and follow-up characteristics of 563 HIV seropositive men stratified by H C V antibody status (46 H C V seropositive and 517 H C V seronegative men)  TABLE 7.2  Comparison of 46 HCV+/HIV+ and 517 HCV-/HIV+ men with respect to clinical symptoms and signs  TABLE 7.3  116  117  Comparison of 46 HCV+/HIV+ and 517 HCV-/HIV+ men with respect to physical findings  118  LIST OF FIGURES  FIGURE 3.1  Cumulative incidence of hepatitis C infection in the VLAS cohort (19831998)  FIGURE 3.2  51  Revised cumulative incidence of hepatitis C infection in the VLAS cohort (1983-1998)  FIGURE 3.3  52  H C V seroconversion in a cohort of 626 initially seronegative men the VLAS cohort (1983-1997)  FIGURE 3.4  ;  Comparison of H C V and HIV cumulative incidence in the VLAS cohort (1983-1998)  FIGURE 3.5  53  54  Comparison of H C V and HIV incidence in the VLAS cohort (1983-1997) 55  FIGURE 6.1  Time to first positive HCV-antibody test among 662 men stratified by injection drug use during lifetime (1983-1998)  FIGURE 6.2  98  Time to first positive HCV-antibody test among 662 men stratified by HIV antibody status (1983-1998)  xi  99  FIGURE  6.3  Time to first positive HCV-antibody test among 662 men stratified by composite variable: insertive fisting and insertive oral-anal contact (1983-1998)  FIGURE  7.1  100  Kaplan-Meier AIDS progression curves for 563 HIV seropositive men stratified by H C V antibody status  FIGURE  7.2  119  Kaplan-Meier curves of cumulative mortality for 563 HIV seropositive men stratified by H C V antibody status  xii  120  ACKNOWLEDGEMENTS I must be honest. I never really thought I would ever finish this project. At least not until I took some time away, dropped all the other competing priorities in my life, and focused on just this one thing. Fortunately, in the late spring and early summer of 2003,1 did get a chance to do exactly this. I am extremely pleased I did. There are a lot of important people in my life who have helped me and encouraged me to complete this personal project, either directly or indirectly. I would like to thank each of these individuals for their much appreciated support and encouragement over the years. Dr. Martin Schechter has been my academic supervisor throughout this endeavor, and he has been my colleague and friend for more than twenty years. The majority of what I have learned about epidemiology during my career has been the result of working directly with Martin. I have always felt very privileged to be able to work with and learn from a world-renowned epidemiologist. I would like to express my gratitude to Dr. Michael O'Shaughnessy for asking me to take on this project back in 1997, and for providing me with the necessary funding to carry out H C V antibody testing on serum specimens obtained from approximately 900 VLAS participants. I also appreciate the patience and understanding that Dr. "O"' has exhibited as I struggled to complete this project. I am indebted to Dr. Chris Sherlock and his laboratory staff for conducting the H C V antibody testing for this study. More importantly, I would like to thank Chris for providing me with much needed encouragement at times when I wanted to abandon this dissertation. I would also like to acknowledge the constant encouragement of my committee member, Dr. Robert Hogg. Mrs. Bonnie Devlin was instrumental in helping me focus my undivided attention on completion of this project. I must thank Bonnie for never letting me forget I needed to finish this thing. Most of all, I want to thank my loving wife Anne for always believing in me and supporting me during the past twenty-five years of our life together. Our children, Jenna and Jonathan, are to be commended for putting up with my antics while I was preparing this dissertation. This accomplishment is truly our success. To my brother Ian - thank you for always being there for me and providing me with a constant link to my past. To my best friend Bill Holden -1 have always enjoyed our daily conversations about golf, music, and life. I appreciate the support you have shown me over the many years of our friendship. To Don Burton and Peter Vann - thanks for sharing some fine musical moments with me during the latter part of this journey. Finally, I would like to dedicate this completed work to the memory of my father Dr. Joseph Stoddart Craib, who passed away on January 30, 2003.  xiii  CHAPTER 1 INTRODUCTION  1.1  HEPATITIS C VIRUS - EPIDEMIOLOGY Hepatitis C is recognized as a global health problem. Hepatitis C virus (HCV) is  a blood-borne pathogen and was first identified i n 1989. H C V belongs to the 1  Flaviviridae family of viruses, and is related to viruses that have been causally linked to yellow fever, hog cholera, and bovine diarrhea. Since its discovery, H C V has been established as an etiologic agent of acute and chronic liver disease, as well as cirrhosis of the liver. It has also been found to be associated w i t h development of hepatocellular carcinoma i n humans. " 2  6  D u r i n g the 1970s and 1980s, medical practitioners i n the United States became increasingly aware that an unidentified agent i n the blood supply was causing a small percentage of b l o o d / b l o o d product recipients to experience post-transfusion hepatitis. In some of these cases, cirrhosis and liver cancer was diagnosed many years later. To distinguish this disease from other k n o w n forms of hepatitis, it was described as non-A, non-B hepatitis. After the development of specific serological tests for antibody to H C V , it was demonstrated that the majority of cases of transfusion-associated non-A, non-B hepatitis were associated with H C V infection. ' In addition, the detection of 2 3  genomic H C V sequences using P C R assays confirmed that H C V viremia was correlated w i t h progression to chronic hepatitis i n non-A, non-B patients. " 7  14  In 1997, the W o r l d Health Organization published its first estimate of the global prevalence of H C V infection. Prevalence data obtained from reporting countries indicated that approximately 3 percent of the world's population were infected with HCV.  1 5 , 1 6  It was estimated that as many as 170 million people may be infected with  H C V . Prevalence appears to be higher among countries i n Africa (5.3%), the Eastern 1  Mediterranean region (4.6%), the Western Pacific Region (3.9%), and Southeast Asia (2.2%). In comparison to these regions, prevalence rates are lower in countries in North America (1.7%) and Europe (1.0%). Countries with prevalence rates above 10% include Egypt (18.1%), Rwanda (17.0%), Bolivia (16.3%), Cameroon (12.5%), Burundi (11.1%), Guinea (10.7%) and Mongolia (10.7%). The following countries have reported prevalence rates between 5% and 10%: Libya (7.9%), Papua New Guinea (7.0%), Gabon (6.5%), Democratic Republic of the Congo (6.4%), Viet Nam (6.1%), Thailand (5.6%), Suriname (5.5%), Ghana (5.4%), and Palestinian Authority (5.2%). The accuracy of current estimates of regional and global H C V prevalence remains in doubt because many countries have not yet conducted population-based studies to assess the exact magnitude of the problem. Moreover, initial estimation of H C V prevalence in many countries has been based on samples of individuals that are not representative of their respective target populations and as a result, these estimates of prevalence are subject to selection and volunteer bias. In addition, some countries have only recently gained access to serologic tests for detecting H C V antibody and viral genomic RNA in serum. H C V antibody assays first became commercially available in 1990. H C V prevalence is lower in the United States and Canada compared to other regions of the world. H C V infection is, however, the most common chronic bloodborne infection in the United States. Prevalence in the general population has been estimated to be 1.8 percent, which corresponds to approximately 4 million infected Americans.  17  Prevalence rates in the United States tend to vary according to gender,  age and ethnicity. For example, among 30 to 49 year-olds, H C V prevalence ranges from 3 to 4 percent. In contrast, among Afro-Americans of similar age, prevalence is  2  estimated to be 10%.  17  H C V infection is currently the primary reason for liver  transplantation i n the United States. It has been estimated that the prevalence of a n t i - H C V positivity is approximately 0.8% (0.68% to 0.94%) i n Canada, 0.96% i n males and 0.53% i n females. '  18 19  Estimates of  prevalence exhibit heterogeneity between provinces. In Newfoundland and Prince E d w a r d Island, H C V prevalence has been estimated to be 0.1% and 0.2%, respectively, compared to 0.9% and 1.4% i n Ontario and British Columbia. H C V prevalence appears highest i n British Columbia, Ontario, Alberta and Quebec. These four provinces account for about 90 percent of all H C V infections i n Canada. H C V infection is now the second most frequently reported disease entity among 47 nationally reported diseases.  20  The majority of H C V transmission worldwide occurs via percutaneous exposure to infected blood. Transmission of H C V is associated w i t h direct percutaneous exposures including transfusion of unscreened blood/blood products, transplantation of organs or tissues from infected donors, and sharing contaminated needles among injection drug users. " Nosocomial transmission of H C V can occur during medical and 21  31  dental procedures from inadequately sterilized instruments, and poor infection control practices. " 32  35  H C V prevalence is higher among health care workers involved i n invasive  procedures and among those who have frequent contact w i t h infected blood. " 36  37  Some  skin piercing practices such as tattooing, body piercing, and acupuncture have also contributed to the spread of H C V , particularly i n developing nations.  38  Iatrogenic infections can also occur. The largest iatrogenic infection scenario k n o w n to date has taken place i n Egypt where approximately 15 to 20% of the general population is estimated to be infected with H C V . This high prevalence rate is thought to be associated w i t h a decades-old strategy to treat schistosomiasis, a parasitic disease. In Egypt, parenteral antischistosomal therapy (PAT) was extensively practiced since the 1920s and discontinued during the late 1980s. P A T typically required 10 to 12 injections  3  and was usually given with reusable syringes. In a recent study conducted by Frank et al, the degree of exposure to PAT was estimated for over 8,000 Egyptians during the period 1961 to 1986, and higher levels of exposure were found to be significantly associated with seroprevalence of antibodies to H C V across four regions of the country.  39  These data provide strong evidence that PAT played a major role in the  diffusion of H C V infection throughout Egypt. Both vertical and horizontal transmissions of H C V have been reported. The rate of transmission of H C V from an infected mother to their newborn infant is generally estimated to be 5%, but this rate is higher among mothers who are co-infected with HIV. " Household contact with another member of the household who is infected with 40  46  H C V has also been implicated as a route of transmission. Family studies suggest that horizontal transmission may be more important than vertical transmission. " 47  49  Sexual contact has been identified as a means of transmitting H C V but the efficiency of transmission in this manner is thought to be low.  30,50  "  Sexual transmission  53  of H C V appears to be infrequent between heterosexual couples in long-term monogamous relationships who have no identifiable percutaneous exposures. " 54  55  Some  epidemiological evidence of sexual transmission of H C V has been provided by observational studies of homosexual men and female sex workers. " H C V 56  57  seroprevalence appears to be higher in these populations compared to the population at large.  1.2  HEPATITIS C - N A T U R A L HISTORY H C V is a single-stranded RNA virus and more than 100 strains have been  identified. These strains have been grouped into at least 6 major genotypes or clades. Geographic clustering of genotypes has been observed globally. Genotypes 1 to 3 exhibit a worldwide distribution, whereas genotypes 4 and 5 are found primarily in 4  Africa. Genotype 6 clusters in Asia. In Canada and the United States, genotype 1 is more common.  16  In the majority of cases, individuals do not exhibit any acute signs or symptoms associated with H C V infection. Twenty to 30 percent of individuals with acute infection become jaundiced, and 10 to 20 percent develop symptoms. " 58  60  Severe hepatic failure  associated with primary H C V infection is rare, but may occur among individuals with compromised immune systems.  3,61  Although acute infection with H C V is usually  asymptomatic, the majority of infected individuals develop chronic infection. Approximately 60 to 80 percent of HCV-infected individuals are asymptomatic and develop chronic H C V infection. The most common symptom associated with chronic H C V infection is fatigue. Other related symptoms include nausea, vomiting, mild fever, arthralgia, myalgias, loss of appetite, abdominal pain, diarrhea, and weight loss. H C V RNA can be detected in the blood of an infected person within a few days after infection, and H C V antibodies can usually be detected within 45 to 90 days. The 62  mean incubation period is estimated to be 6 to 12 weeks.  3,63  Twenty to 30 percent of  infected individuals will exhibit clinical symptoms. About 15 percent of individuals who test H C V RNA positive by serum during acute or post-acute H C V infection eventually become undetectable. This minority of infected individuals appears able to eradicate H C V from their bodies by their immune response.  3,61,64  However, the findings  of Haydon et al. cast doubts that H C V infection can be eradicated and that H C V disease can be self-limiting. In their study, 10 of 12 patients with chronic hepatitis C had H C V RNA detected in their liver but not in their serum.  65  The majority of HCV-infected individuals develop chronic hepatitis C. Data from various observational studies indicate that 75 to 85 percent of acute infections do not resolve and result in a chronic viral infection.  3,64,66  No clinical features of the acute  disease have been found to be predictive of chronicity. In the majority of those with 5  chronic hepatitis C infection, the disease typically progresses slowly without clinical symptoms or physical signs during the first two decades following infection.  67,68  Roughly one-fifth of individuals with chronic hepatitis C develop non-specific symptoms including intermittent fatigue and malaise. Other symptoms may also accompany chronic infection. A recent report by Cacoub et al. found that almost threequarters of HCV-infected individuals exhibited at least one of the following symptoms: arthralgia (joint pain), paresthesia (nerve sensation abnormalities), myalgias (muscle pain), pruritus (itching) and sicca syndrome (dry eyes, skin, and mouth).  69  In many  individuals with chronic hepatitis C symptoms first appear at the time of advanced liver disease. The late sequelae of chronic H C V infection include cirrhosis and hepatocellular carcinoma. Cirrhosis emerges in at least 20 percent of HCV-infected individuals within twenty years of the onset of infection whereas hepatocellular carcinoma occurs in a smaller proportion of individuals (less than 5%) and takes about 10 or more years after onset of cirrhosis to emerge. Risk factors that hasten the rate of progression of chronic hepatitis C to cirrhosis include alcohol consumption, age at time of infection, and the severity of liver histology at the initial bioposy.  16  In this context, the role of other factors  such as gender, viral genotype, co-infection with other viruses like human immunodeficiency virus (HIV) and hepatitis B virus (HBV) is less well understood. Risk factors that are associated with hepatocellular carcinoma are similar to those for cirrhosis and include age over 40 years, duration of infection, and presence of cirrhosis. On the other hand, alcohol consumption and viral genotype have not been shown to be associated with progression to hepatocellular carcinoma.  16  Some individuals with chronic hepatitis C present with extrahepatic sequelae including essential mixed cryoglobulinemia and glomerulonephritis, aplastic anemia, and pulmonary fibrosis. Other extrahepatic sequelae thought to be associated with 6  chronic hepatitis C disease progression include cutaneous manifestations (lichen planus and porphyria cutanea tarda), ocular lesions (Mooren's ulcers), sialadenitis and B-cell lymphoma. ' " 16  69  71  Data from a small number of natural history studies have revealed that approximately 15 to 20 percent of HCV-infected individuals will eventually develop progressive end-stage liver disease. The remainder of infected individuals will die of causes unrelated to liver disease. Chronic liver disease is the tenth leading cause of 72  death among adults in the United States, and accounts for approximately 25,000 deaths annually, or approximately one percent of all deaths.  73  Population-based studies  indicate that 40 percent of chronic liver disease is HCV-related, resulting in an estimated 8,000 to 10,000 deaths each year.  74  Death rates due to H C V are expected to  increase over the next 20 years. Estimates of death rates may underestimate the actual death rate because of underreporting of liver disease and underreporting of deaths due to liver disease on death certificates and in other databases.  75  1.3  RESEARCH NEEDS A N D STUDY JUSTIFICATION The accuracy of current estimates of regional and global H C V prevalence is  questionable because many countries have not yet conducted population-based studies to assess the exact magnitude of the problem. In addition, many countries have only recently gained access to serologic tests for detecting H C V antibody and viral genomic RNA in serum. Initial estimates of H C V prevalence for the most part have been based on samples of individuals that are not representative of their respective target populations. As a result, these initial estimates of prevalence are subject to bias. In Canada, the epidemiology of H C V remains largely unknown. Current estimates of H C V prevalence must be considered as hypotheses because initial estimates are based on fragmented data. Because of this, the true burden of H C V 7  disease i n Canada is still a matter of speculation. A t present, there are no ongoing studies of annual H C V incidence. There is a pressing need to better evaluate the extent and distribution of H C V infection, and to assess the burden of infection and disease in both the short and long term. Recent research i n the United States suggests H C V may lead to a substantial health and economic burden over the next 10 to 20 years . The 75  situation i n Canada w i l l likely be no different. Well designed, population-based observational studies are needed to arrive at more accurate estimates of local, regional and global H C V prevalence. The relative contribution of various routes of H C V transmission also needs to be elucidated. In particular, the role of sexual transmission in the spread of H C V requires clarification. Longitudinal investigations are needed to estimate the true burden of hepatitis C disease, and to estimate morbidity and mortality rates among those infected with H C V , and among those individuals co-infected with other infectious agents such as H I V . Such natural history studies could also be used to identify co-factors and predictors of disease progression i n both HCV-infected and co-infected individuals. D u r i n g the 1980s, i n response to the H I V / A I D S pandemic, a number of cohort studies of HIV-infected individuals were launched i n a number of countries throughout the world. These longitudinal investigations were conducted i n order to gain important knowledge about the epidemiology and the natural history of H I V infection. M a n y of these prospective studies have contributed greatly to our understanding and response to the H I V / A I D S pandemic during the past two decades. One such longitudinal study was conducted in Vancouver, British Columbia. The Vancouver Lymphadenopathy-AIDS Study (VLAS) followed and monitored a cohort of homosexual men from November 1982 to December 1998. The general objective of this longitudinal study was to contribute to our understanding of the natural history of H I V infection including risk factors for seroconversion through to 8  progression to AIDS and survival. The VLAS was a prospective study of gay/homosexual men who were enrolled during two recruitment periods. From November 1982 until December 1984, a total of 729 men were recruited through six primary care practices in central Vancouver. Two additional practices were added and 271 men were enrolled during the period from October 1986 to December 1987. Followup visits occurred approximately every six months until September 1986 after which subjects completed visits on an annual basis until December 1998. During each visit participants completed a self-administered questionnaire which gathered information regarding demographic variables, numbers of sexual partners and sexual practices, history of sexually transmitted diseases, use of tobacco, alcohol, and illicit drugs. In addition, a physical examination and functional inquiry were performed and blood samples were drawn for immunologic and HIV antibody testing. Serum samples were frozen at minus 20 degrees Celsius. Since 1992, plasma specimens from study participants were frozen at minus 80 degrees Celsius. In response to the epidemic of H C V infection in Canada and the need for population-based, long-term, prospective studies, the VLAS has provided a unique opportunity to respond to this need in a well-defined cohort. Finally, the VLAS has provided a cost-effective means to acquire important knowledge concerning the epidemiology and natural history of H C V infection in a well-characterized cohort of homosexual men during the period 1982-98.  9  V  1.4  OBJECTIVES, HYPOTHESES, A N D THESIS ORGANIZATION The general aim of this study was to contribute to our understanding of the  epidemiology of H C V infection and its modes of transmission, and the natural history of H C V infection and its related clinical outcomes in a population-based study of homosexual men. The first objective of this study was to estimate the prevalence and incidence of H C V infection in homosexual men in Vancouver during the period 1983 to 1998. This objective was achieved by utilizing an archive of frozen blood samples obtained from VLAS participants. Cumulative and annual H C V seroconversion rates were estimated using Kaplan-Meier, actuarial, and incidence-density methods. It was estimated that 5 percent of HIV-negative and 15 percent of HIV-positive VLAS participants would be found to be infected with HCV. Based on the HIV serologic status of participants at the time of their last laboratory visit (540 HIV positive and 410 HIV-negative), it was projected that a total of 100 men would be found to be infected with H C V in this cohort. This projection included 20 cases among HIVnegative men and 80 cases among HIV-positive men. It was also estimated that 50 men would be H C V seroconverters, 50 would be seroprevalent, and that 80 would be coinfected with HIV. These numbers of HCV-infected men should provide sufficient statistical power for conducting meaningful cross-sectional comparisons and some prospective analyses. The second objective was to identify risk factors for H C V transmission and assess the relative contributions of various routes of transmission including injection drug use, substance use, sexual practices, co-infection with other sexually transmitted diseases including HIV, history of blood transfusions, demographic characteristics, and other  10  variables. T h i s objective w a s a c h i e v e d b y c o n d u c t i n g nested c o m p a r i s o n s b e t w e e n H C V - i n f e c t e d a n d u n i n f e c t e d m e n . M a n t e l - H a e n s z e l m e t h o d s a n d m u l t i v a r i a t e logistic regression m o d e l i n g w e r e u s e d to assess a n d i d e n t i f y significant r i s k factors for H C V infection. The t h i r d objective w a s to estimate the relative r i s k of H C V t r a n s m i s s i o n for specific sexual practices a m o n g h o m o s e x u a l m e n i n d e p e n d e n t of injection d r u g use. T h i s objective w a s a c h i e v e d b y e x c l u d i n g m e n w h o r e p o r t e d u s i n g injection d r u g s , a n d b y e x p a n d i n g the e l i g i b i l i t y criteria for the c o m p a r a t i v e analysis to i n c l u d e V L A S participants f r o m both recruitment periods. The f o u r t h objective of this i n v e s t i g a t i o n w a s to e x a m i n e p r e d i c t o r s of t i m e to H C V - s e r o p o s i t i v i t y . Because the n u m b e r of i n c i d e n t infections of H C V i n the cohort w a s s m a l l , b o t h i n c i d e n t a n d p r e v a l e n t infections w e r e i n c l u d e d as events of interest i n the p r o s p e c t i v e analyses. R i s k factors of significance that w e r e i d e n t i f i e d i n p r e v i o u s nested c o m p a r i s o n s w e r e assessed as p o t e n t i a l p r e d i c t o r s of t i m e to H C V seropositivity. V a r i a b l e s of interest i n c l u d e d : injection d r u g use, H I V a n t i b o d y status, sexual practices, a n d h i s t o r y of other s e x u a l l y t r a n s m i t t e d diseases. M e t h o d s of s u r v i v a l analysis i n c l u d i n g K a p l a n - M e i e r m e t h o d s , a n d C o x p r o p o r t i o n a l h a z a r d s regression w e r e u s e d to i d e n t i f y i n d e p e n d e n t p r e d i c t o r s of t i m e to H C V s e r o p o s i t i v i t y a m o n g h o m o s e x u a l men i n Vancouver.  F i n a l l y , the fifth objective of this s t u d y w a s to assess the influence of H C V i n f e c t i o n o n the n a t u r a l h i s t o r y of h u m a n i m m u n o d e f i c i e n c y v i r u s ( H I V ) infection. The effects of H C V i n f e c t i o n o n H I V - r e l a t e d o u t c o m e s w e r e e x a m i n e d a m o n g H I V - i n f e c t e d p a r t i c i p a n t s i n the V L A S cohort. T h i s w a s a c h i e v e d b y i n c l u d i n g H I V seropositive p a r t i c i p a n t s f r o m b o t h r e c r u i t m e n t p e r i o d s i n c o m p a r a t i v e analyses. O u t c o m e s of interest i n this i n v e s t i g a t i o n i n c l u d e d s y m p t o m s a n d signs, p h y s i c a l e x a m f i n d i n g s ,  11  CD4 cell count, progression to AIDS, and survival.. Methods of survival analysis were used to analyze these data. This thesis is comprised of eight chapters. The first chapter contains a brief introduction to the epidemiology and natural history of HCV. Chapter 2 presents an overview of previous studies of sexual transmission of H C V in heterosexuals and gay men.  Chapter 2 also contains a detailed description of the VLAS study including the  number of visits completed by the participants and rates of follow-up. Descriptions of data management procedures, statistical methods and approaches to multivariate modeling are also presented in chapter 2. Chapters 3 through 7 address each of the aforementioned objectives in sequential order. In chapter 3, the serological results of hepatitis C virus antibody testing for the VLAS cohort are presented and discussed. Chapters 4 and 5 provide an assessment of risk factors associated with H C V infection including injection drug use, sexual practices, substance use, history of blood transfusions, and HIV infection. Risk factors for H C V transmission that were identified by means of nested comparisons between seropositive and seronegative men in chapters 4 and 5, are re-examined prospectively in chapter 6. The results of an assessment of the influence of H C V infection on the natural history of HIV infection are presented and discussed in chapter 7. Each of chapters 3 through 7 has been written as an independent article and prepared for submission to a scientific journal. Alternatively, the results from one or more these chapters could be combined into a smaller number of scientific articles. The final chapter of this thesis provides a summary of the research findings and describes the strengths and limitations of this research. In addition, chapter 8 outlines the unique contributions and potential impact of the findings of this longitudinal study. The work presented in this thesis was conceived, conducted, and disseminated entirely by the doctoral candidate unless otherwise indicated. 12  1.5  SUMMARY Hepatitis C is recognized as a global health problem. The W o r l d Health  Organization estimates that as many as 170 million people may be infected worldwide. In the United States, it is estimated that 4 million persons are infected with H C V . H C V infection is currently the primary reason for liver transplantation i n the United States. In Canada, it is estimated that 250,000 to 300,000 individuals are infected, and H C V infection is now the second most frequently reported disease entity among 47 nationally reported diseases. The majority of H C V transmission worldwide occurs via percutaneous exposure to infected blood. Persons at risk for H C V infection in developed countries include recipients of previously unscreened blood, blood products and organs, injection drug users, health care workers, individuals undergoing haemodialysis, children born to HCV-infected mothers, individuals who have undergone body piercing or tattooing; and intranasal cocaine users. It is currently believed that sexual and household transmission of H C V is uncommon and that risk of H C V transmission by sexual activity between monogamous partners is low. In Canada, the epidemiology of H C V remains largely u n k n o w n and the true burden of this disease still remains mostly a matter of speculation. There are no ongoing population-based studies of H C V incidence (how do you know?). There is a pressing need to better evaluate the extent and distribution of H C V infection i n Canada, and to assess the burden of infection and disease i n both the short and long term. The relative contributions of various routes of transmission of H C V need to be elucidated. Longitudinal investigations are required to estimate the true burden of hepatitis C disease, and to estimate morbidity and mortality rates among those infected with H C V , and among those individuals co-infected w i t h other infectious diseases such as 13  HIV.  Such natural history studies could also be used to identify co-factors and  predictors of disease progression in both HCV-infected and co-infected individuals. The Vancouver Lymphadenopathy-AIDS Study (VLAS) followed and monitored a cohort of homosexual men from November 1982. In response to the epidemic of H C V infection in Canada and the need for population-based, long-term, prospective studies, the VLAS has provided a unique opportunity to respond to this need. Using an archive of frozen blood samples obtained from VLAS participants during the period 1982 -1998, H C V antibody testing was carried out on these specimens, and estimates of prevalence and annual incidence were calculated. Data regarding demographic variables, numbers of sexual partners and sexual practices, history of sexually transmitted diseases, use of tobacco, alcohol, and illicit drugs were used to identify significant risk factors associated with H C V infection. Longitudinal behavioural and clinical data from the VLAS were used to identify independent predictors of time to H C V seropositivity, as well as the effect of H C V infection on HIV-related disease progression and survival. In summary, the VLAS has provided a cost-effective means to acquire important knowledge concerning the epidemiology and natural history of H C V infection in a wellcharacterized cohort of homosexual men during the period 1983-98.  14  1.6  1.  REFERENCES  Choo Q-L, Kuo G, Weiner AJ, Overby LR, Bradley DW, Houghton M. Isolation of a C D N A clone derived from a blood-borne non-A, non-B viral hepatitis genome. Science 1989;244:359-362.  2.  Alter HJ, Seeff LB. Transfusion-associated hepatitis. In: Zuckerman AJ, Thomas H C (Eds): Viral Hepatitis: Scientific Basis and Clinical Management. Edinburgh, Churchill Livingstone, 1993.  3.  Alter MJ, Margolis HS, Krawczynski K, Judson FN, Mares A, Alexander WJ, Hu PY, Miller JK, Gerber M A , and Samp liner RE. The natural history of community-acquired hepatitis C in the United States. N Engl J Med 1992; 327: 1899-1905.  4.  Hoofnagle JH, Tralka TS. Management of Hepatitis C. Hepatology 1997; 26 (3 Suppl 1).  5.  Bukh J, Miller RH, Kew M C , Purcell RH. Hepatitis C virus RNA in southern African blacks with hepatocellular carcinoma. Proc Natl Acad Sci USA 1993; 90:1848-1851.  6.  IARC Working Group, World Health Organization: Hepatitis C virus. IARC Monograph 1994; 59:165-221.  7.  Okamato H , Okada S, Suigiyama Y, et al. Detection of hepatitis C virus RNA by a two-stage polymerase chain reaction with two pairs of primers deduced from the 5'-noncoding region. Jpn J Exp Med 1990; 60:215-222.  8.  National Institutes of Health Consensus Development Conference Panel Statement: Management of hepatitis C. Hepatology 1997; 156:1427-1430.  15  9.  Garson JA, Tedder RS, Briggs M , et al. Detection of hepatitis C viral sequences in blood donations by nested polymerase chain reaction and prediction of infectivity. Lancet 1990; 335:1419-1422.  10.  Seeff LB, Dienstag JL. Transfusion-associated non-A, non-B hepatitis. Where do we go from here? Gasteroentology 1998; 95:530-533.  11.  Bukh J, Purcell RH, Miller RH: Importance of primer selection for the detection of hepatitis C virus RNA with polymerase chain reaction assay. Proc Natl Acad Sci USA 1992; 89:187-191.  12.  Weiner Aj, Kuo G, Bradley DW, et al. Detection of hepatitis viral sequences in non-A, non-B hepatitis. Lancet 1990; 335:1-3.  13.  Garson JA, Tuke PW, Makris M , et al. Demonstration of viremia patterns in hemophiliacs treated with hepatitis C virus contaminated factor VIII concentrates. Lancet 1990; 336:1022-1025.  14.  Farci P, Alter HJ, Wong D, et al. A long-term study of hepatitis C virus replication in non-A, non-B hepatitis. N Engl J Med 1991; 325:98-104.  15.  Weekly Epidemiological Report, World Health Organization, No. 46, November 1997; 341-344.  16.  Report of a World Health Organization consultation organized in collaboration with the Viral Hepatitis Prevention Board, Antwerp, Belgium. Global surveillance and control of hepatitis C. Journal of Viral Hepatitis 1999; 6:35-47.  17.  Alter MJ, Kruszon-Moran D, Nainan OV, et al. The prevalence of hepatitis C virus infection in the United States, 1988 through 1994. N Engl J Med 1999; 341:556-562.  16  18.  Zou, S., Tepper, M . & Giulivi, A. Current status of hepatitis C in Canada. Canadian Journal of Public Health, 2000; 91,1, S10-S15.  19  Remis, R., Hogg, R., Krahn, M.D., et al. Estimating the number of blood transfusion recipients infected by hepatitis C virus in Canada, 1960-85 and 1990-92. Report to Health Canada (1998).  20.  Health Canada. Notifiable Diseases Annual Summary: 1995. Canada Communicable Disease Report 1997;23S9.  21.  Alter HJ, Jett BW, Poliyo AJ, et al: Analysis of the role of hepatitis C virus in transfusion-associated hepatitis. In: Hollinger FB, Lennon SM, Margolis HS (Eds): Viral Hepatitis and Liver Disease: Baltimore, Williams and Wilkins, 1991; 396-402.  22.  Alter HJ: Post-transfusion hepatitis in the United States. IN: Nishioka K, Suzuki H , Mishiro S, Oda T (Eds): Viral Hepatitis and Liver Disease, Toyko, Springer-Verlag, 1994; 551-553.  23.  Elia GF, Magnani G, Belli L, et al: Incidence of anti-hepatitis C virus antibodies in non-A, non-B post-transfusion hepatitis in an area of northern Italy. Infection 1991; 19:336-339.  24.  Mattsson L, Grillner L, von Sydow M, et al: Seroconversion to hepatitis C virus antibodies in patients with acute post-transfusion non-A, non-B hepatitis in Sweden. Infection 1991; 19:309-312.  25.  Goncales FL Jr., Pedro R de J, Da Silva LJ, et al: Post-transfusion hepatitis in the city of Campinas SP, Brasil. 1. Incidence, etiological agents, and clinicoepidemiological aspects of hepatitis C. Revista do Intituto de Medicina Tropical de Sao Paulo 1993; 35:53-62.  17  26.  Pereira BJG, Wright TL, Schmid C H , Levey AS, for the New England Organ Bank Hepatitis C Study Group: A controlled study of hepatitis C transmission by organ transplantation. Lancet 1995; 345:484-487.  27.  Wreghitt TG, Gray JJ, Allain JP, Poulain J, Garson JA, Deaville R, Maple C, Parameshwar J, Calne RY, Wallwork J, Alexander GJM: Transmission of hepatitis C virus by organ transplantation in the United Kingdom. Journal of Hepatology 1994; 20:768-772.  28.  Shuhart M C , Myerson D, Childs BH, Fingeroth JD, Perry JJ, Snyder DS, Spurgeon CL, Bevan CA, McDonald GB. Marrow transplantation from hepatitis C virus seropositive donors: Transmission rate and clinical course. Blood 1994; 84:3229-3235.  29.  Van Ameijden EJC, Van den Hoek JAR, Mientjes G H C , Coutinho RA. A longitudinal study on the incidence and transmission patterns of HIV, HBV, and H C V infection among injection drug users in Amsterdam. European Journal of Epidemiology 1993; 9:255-262.  30.  Atler MJ. Epidemiology of hepatitis C in the West. Seminars in Liver Disease 1995, 15:5-14.  31.  Neal KR, Jones DA, Killey D, James V. Risk factors for hepatitis C virus infection. A case-control study of blood donors in the Trent Region (UK). Epidemiology and Infection 1994;112:595-601.  32.  Laskus T, Cianciara J, Slusarcxyk J. A follow-up study of an outbreak of non-A, non-B hepatitis in a plasmapheresis unit. Liver 1990; 10:49-53.  18  33.  Rey JF, Halfon P, Feryn JM, Khiri H , Masseyeff MF, Ouzan D. Risque de transmission du virus de l'hepatite C par l'endoscope digestive. Gastroenterologie clinique et biologique 1995; 19: 346-349.  34.  Piazza M , Borgia G, Picciotto L, Nappa S, Cicciarello S, Orlando R. Detection of hepatitis C virus RNA by polymerase chain reaction in dental surgeries. Journal of Medical Virology 1995; 45:40-42.  35.  Merle A , Sagliocca L, Manzillo G, Converti F, Amoroso P, Stazi M A , Ferrigno L, Rapicetta M , Franco E, Adamo B, Palumbo F, Sbreglia C, Pana A , Pasquini P, and the SEIEVA Collaborating Group. Risk factors for acute non-A, non-B hepatitis and their relationship to antibodies for hepatitis C virus: A casecontrol study. American Journal of Public Health, 1994; 84:1640-1643.  36.  Zuckerman J, Clewley G, Griffiths P, Cockcroft A. Prevalence of hepatitis C antibodies in clinical health-care workers. Lancet 1994; 343:1618-1620.  37.  De Browser C, Lecomte A: H C V anibodies in clinical health-care workers (letter). Lancet 1994; 344:962.  38.  Conry-Cantilena C, Vanraden M , Gibble J, Melpolder J, Shakil O, Viladomiu L, Cheung L, Dibisceglie A, Hoofnagle J, Shih JW, Kaslow R, Ness P, Alter HJ. Routes of infection, viremia, and liver disease in blood donors found to have hepatitis C virus infection. New England Journal of Medicine 1996; 334:16911696.  39.  Frank, C, Mohamed MK, Strickland GT, Lavanchy D, Authur RR, Magder LS, Khoby TE, Abdel-Wahab Y, Ohn ESA, Anwar W, Sallam I. The role of parenteral antischistosomal therapy in the spread of hepatitis C virus in Egypt. Lancet 2000; 355:887-891.  19  40.  Zanetti, AR, Tanzi, E, and Newell, ML. Mother-to-infant transmission of hepatitis C virus. Journal of Hepatology 1999;31 (Suppl 1):96-100.  41.  Zanetti, AR, Tanzi, E, Romano, et al. A prospective study on mother-to-infant transmission of hepatitis C virus. Intervirol 1998;41(4-5):208-12.  42.  Michielsen, PP, and Van Damme, P. Viral hepatitis and pregnancy. Acta Gastroenterol Belg 1999;62(l):21-9.  43.  Agha, S, Sherif, LS, Allam, M A , and Fawzy, M . Transplacental transmission of hepatits C Virus in HIV-negative mothers. Res Virol 1998;149(4):229-34.  44.  Ohto, H , Terazawa, S, Sasaki, N , et al. Transmission of hepatitis C virus from mothers to infants. New Eng J Med 1994;330:744-50.  45.  A C O G educational bulletin. Viral hepatitis in pregnancy. Inf J Gynecol Obstet 1998;63:195-202.  46.  Granovsky, MO, Minkoff, HL, Tess, BH et al. Hepatitis C virus infection in the Mothers and Infants Cohort Study. Pediatrics 1998;102:355-359.  47.  Riestra MS, Rodriguez G M , Sanchez SRF, Menendez TF, Suarez GA, Alvarez NC, Perez AR, Rodrigo AL. Intrafamilial spread of hepatitis C. Infection 1991; 19:431-433.  48.  Chuang TT, Liou TC, Young KC, Lin XZ, Lin CY, Shin JS, Wu HL. Intrafamilial transmission of hepatitis C virus: the important role of inapparent transmission. J Med Virol 1994; 42:91-96.  49.  Kiyosawa K,Sodeyama T, Tanaka E, Shimizu S, Furuata S, Miyazaki Y, Akhane Y, Suzuki H,. Intrafamilial transmission of hepatitis C virus in Japan. J Med Virol 1991; 33:114-116.  20  50.  Scully LJ, Mitchell S, Gill P. Clinical and epidemiologic characteristics of hepatitis C in a gasteroenterology/hepatology practice in Ottawa. Canadian Medical Association Journal 1993; 148:1173-1177.  51.  Scotto G, Savastano A M , Fazio V, Conte PE, Ferrara S, Mangano A, Tantimonaco G. Sexual transmission of hepatitis C virus infection. European Journal of Epidemiology 1996;  52.  Tillman HL, Manns MP. Mode of hepatitis C infection, epidemiology, and chronicity ratein the general population and risk groups. Dig Dis Sci 1996; 41: (Suppl) 27S-40S.  53.  Dienstag JL. Sexual and perinatal transmission of hepatitis C. Hepatology 1997; 26: (Suppl). 1) 66S-70S.  54.  Gordon SC, Patel A H , Kulesza GW, et al. Lack of evidence for the heterosexual transmission. American Journal of Gastroenterology 1992; 87:1849-1849-1851.  55.  Osmond D H , Padian NS, Sheppard HW, et al. Risk factors for hepatitis C virus seropositivity in heterosexual couples. Journal of the American Medical Association 1993; 269:361-365.  56.  Osmond D H , Charlebois, Sheppard HW, Page K, Winkelstein W, Moss A, Reingold A. Comparison of risk factors for hepatitis C and hepatitis B in homosexual men. J Infect Dis 1993;167:66-71.  57.  Kihara M , Imai M , Kondo M , Watanabe S, Kihara M , Soda K. Prevalence of hepatitis C virus and HTV infection in Japanese female prostitutes. Nippon Koshu Eisei Zasshi. Japanese Journal of Public Health 1993; 40:387-391.  58.  Dienstag JL. The natural history of chronic hepatitis C and what should do we do about it? Gastroenterology 1997; 112: 651-655.  21  59.  Aach RD, Stevens CE, Hollinger FB, et al. Hepatitis C virus infection in posttransfusion hepatitis. A n analysis with first- and second-generation assays. New England Journal of Medicine 1991; 325:1325-1329.  60.  Koretz RL, Brezina M , Poliyo AJ, et al. Non-A, non-B post-transfusion hepatitis: Comparing C and non-C hepatitis. Hepatology 1993; 17:361-365.  61.  Seeff LB. Natural history of hepatitis C. Hepatology 1997; 26 (3 Suppl 1):21S28S.  62.  Seeff LB. Hepatitis C. Semin Liver Dis 1995; 15:1.  63.  Tremolada F, Casarin C, Tagger A, et al. Antibody to hepatitis C virus in posttransfusion hepatitis. Annals of Internal Medicine 1991; 114:277-281.  64.  Villano SA, Vlahov D, Nelson KE, et al. Persistence of viremia and the importance of long-term follow-up after acute hepatitis C infection. Hepatology 1999; 29:908-914.  65.  Haydon G H , Jarvis L M , Blair CS, Simmonds P, Harrison DJ, Simpson KJ, Hayes PC. Clinical significance of intrahepatic hepatitis C virus levels in patients with chronic H C V infection. Gut 1998;42(4):570-575.  66.  Shakil A O , Conry-Cantilena C, Alter HJ, et al. Volunteer blood donors with antibody to hepatitis C virus: clinical, biochemical, virologic, and histologic features. Annals of Internal Medicine 1995; 123:330-337.  67.  Hoofnagle JH. Hepatitis C: The clinical spectrum of disease. Hepatology 1997; 36 (Suppl. 1) 15S-20S.  68.  Fattovich G, Giustina G, Degos F, et al. Morbidity and mortality in compensated cirrhosis C: a follow-up study of 384 patients. Gastroenterology 1997; 1123:463-472.  22  69.  Cacoub P, Poynard T, Ghillani P, et al. Extrahepatic manifestations of chronic hepatitis C. Arthritis Rheum 1999; 42:2204-2212.  70.  Cacoub P, Renou C, Rosenthal E, et al. Extrahepatic manifestations associated with hepatitis C infection. A prospective multicenter study of 321 patients. The GERMIVIC. Medicine (Blatimore) 2000; 79:47-56.  71.  Koff RS, Dienstag JL, Extrahepatic manifestations of Hepatitis C and the association with alcoholic liver disease. Semin Liver Disease 1995; 15:101-109.  72.  Seeff LB. Natural history of hepatitis C. In: Liang TJ, moderator. Pathogenesis, natural history, treatment and prevention of hepatitis C. Annals of Internal Medicine 2000; 132:299-300.  73.  Dufour MC. Chronic liver disease and cirrhosis. In Everhart JE, ed. Digestive diseases in the United States: epidemiology and impact. US Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Washington, DC: US Government Printing Office, 1994; NIH publication no. 94-1447,615-645.  74.  Recommendations for Prevention and Control of Hepatitis C Virus (HCV) Infection and HCV-Related Chronic Disease. MMWR1998; 47{RR19}:l-39.  75.  Wong JB, McQuillan G M , McHutchison JG, Poynard T. Estimating future hepatitis C morbidity, mortality, and costs in the United States A m J Public Health 2000; 90:1562-1569  76.  Cook D, Sherlock C H , Mak A, et al. Comparison of 5 commercial EIA kits and 2 immunoblot kits for detection of hepatitis C antibody. Conjoint Meeting on Infectious Diseases, Montreal, 1994.  23  CHAPTER 2 BACKGROUND  2.1  SEXUAL TRANSMISSION OF HEPATITIS C VIRUS IN HETEROSEXUALS  The role of sexual transmission in the spread of H C V has been controversial. Controversy exists because the precise magnitude of the risk associated with sexual transmission remains unknown and the findings from previous studies have been inconsistent. This is primarily due to the fact that most studies have used different methods. Many studies were unable to adequately characterize and quantify the risk associated with sexual transmission because they did not collect adequate information from their participants about specific sexual behaviours and practices, nor did they obtain important information about past and present injection drug use, or a blood transfusion history. CDC Sentinel county studies suggest that 15 to 20 percent of individuals with acute hepatitis C infection have a history of sexual exposure in the absence of other risk factors.  1  In Canada, it is estimated that 27 percent of H C V infections have been  acquired by some means other than injection drug use and/or receipt of blood or blood products prior to 1990. The efficiency of sexual transmission among heterosexuals is 2  generally thought to be low, and infection is rare in long-term, monogamous partners. Early investigations of sexual transmission of H C V have collectively shown the risk of inter-spousal transmission to be very low or non existent.  3-6  In the early 1990s, reports began to appear documenting evidence of sexual transmission of H C V . In contrast to previous studies where inter-spousal transmission 7  was low, a study in Japan identified 154 HCV-positive individuals and found that 27 percent of their spouses were H C V positive.  8  24  Almost 90 percent of the individuals with  positive H C V RNA were infected with genotypes identical to those of their spouse. In this study, it was also noted that risk of H C V infection increased with duration of the marriage. Only those individuals who were married for more than 10 years were infected. It should be noted that inter-spousal studies are confounded by household (i.e. non-sexual) contact. Another study conducted in the United States found that risk of H C V infection among heterosexuals increased with multiple lifetime sexual partners. Females in this 9  study whose sexual partners were HCV-positive were at least three times more likely to be HCV-positive compared to females whose sexual partners were HCV-negative. For males, H C V seropositivity had no significant association with the serologic status of their female sexual partners. These results suggest that 'male to female' transmission of H C V is more efficient than 'female to male'. A small number of studies have been carried out among female sex workers to determine whether this population is at higher risk for H C V infection. A study of over 600 sex workers in Taiwan reported an H C V prevalence of 12 percent.  10  A history of  paid sex for at least six months and previous blood transfusion were both significantly associated with HCV-seropositivity in multivariate analysis. Lissen and colleagues studied a total of 312 Spanish female sex workers and 88 of their male clients. All 11  study participants denied prior or current use of injection drugs, and none reported receiving a blood transfusion. H C V prevalence was 6 percent among the sex workers and 7 percent among their male clients. In contrast to the findings of these two studies, a study of Peruvian sex workers reported a prevalence of less than 1 percent among 966 female sex workers who were tested.  12  25  2.2  HEPATITIS C VIRUS INFECTION IN H O M O S E X U A L M E N  The prevalence of H C V appears to be higher among men who have sex with men compared with the general population. During the period 1990 to 1998, there were numerous observational studies of homosexual men that published reports documenting H C V prevalence, and some studies provided assessments of putative risk factors for transmission in this population. In one of the earliest investigations in homosexual men, Melbye and colleagues reported an H C V prevalence of 4.1 percent.  13  Two hundred and fifty-nine male members of a Danish homosexual organization were prospectively monitored in Copenhagen and Aarhus, Denmark from 1981 to 1989. In 1981, the prevalence of H C V was 1.6 percent in the cohort. Between 1981 and 1984, the cumulative incidence rate was 2.5 percent. From 1984 to 1989, no new infections were observed in this cohort. No significant associations were observed between sexual behaviour variables and H C V infection. In 1991, Tedder et al. reported the results of a study of 1074 consecutive patients who had attended an outpatient genitourinary medicine clinic in London, England during November and December 1987. Among homosexual men who participated in 14  the study, the prevalence of H C V infection was 6.9 percent. This estimate was significantly higher than the value for heterosexual participants (1.0%). Homosexual men who reported a higher number of other sexually transmitted infections in their lifetime were significantly more likely to be seropositive than men who did not. The results of a study by Gasparini and colleagues also provided evidence that H C V prevalence is elevated in homosexual men.  15  They studied 259 apparently healthy  men from homosexual men's clubs in the Vena to region of Italy. Injection drug users were excluded from their study. Sera collected during the period 1987-89 were tested 26  for H C V and prevalence was calculated as 4.1 percent. No association was found between H C V seropositivity and anal intercourse or sexual promiscuity. In another Italian study published in 1998, Osella et al. reported a 12.9% prevalence rate of H C V among 228 homosexual men who were attending two clinical centers.  16  While the study  investigators did find hepatitis B virus infection was strongly associated with some sexual practices reported by the participants, there was no association between sexual behaviour and H C V . Osmond and colleagues measured H C V prevalence among 735 homosexual/bisexual men who were participating in two cohort studies in San Francisco. Sera collected during the period between 1985 and 1986 were screened for 17  H C V antibody and a prevalence of 4.6 percent was found. A history of injection drug use or blood transfusions was associated with H C V seropositivity. After adjustment for these variables, H C V infection was marginally associated with having more than 50 sexual partners in the previous year, and engaging in receptive oral intercourse and receptive anal intercourse with more than 25 partners in the previous year. Buchbinder and her colleagues found H C V prevalence to be 9.2% among 435 homosexual men recruited from a municipal STD clinic in San Francisco. Sera for the 18  study were obtained during 1983 and 1984. Injection drug use, anal receptive intercourse, fisting (i.e. insertion of fingers or hand into partner's rectum), having an injection drug user as a sexual partner, history of genital herpes, HIV seropositivity, use of amphetamine and phencyclidine (PCP) were significant univariate risk factors. However, in multivariate analysis, only injection drug use was significantly associated with H C V seropositivity. In another American study, Ndimbie and colleagues assessed H C V infection in 1058 homosexual men who were participating in a natural history study of H P / infection using sera that were obtained during 1984 and 1985. H C V prevalence was 19  27  2.9% and HCV-positive men were more likely to be HIV-positive. Univariate risk factors associated with H C V seropositivity included injection drug use, insertive anal intercourse with ejaculation, HIV infection, history of syphilis, and rectal gonorrhea. In a multivariate statistical model, injection drug use and insertive anal intercourse were association with H C V seropositivity. An Australian study by Bodsworth and colleagues reported H C V prevalence of 7.6 percent among 1038 homosexual men participating in a prospective study established to identify risk factors for AIDS.  20  In HIV-positive men, H C V prevalence  was significantly higher (11.9%) compared to HIV-negative men (4.0%). Univariate risk factors included injection drug use, HIV infection, and history of syphilis, anogenital herpes, and gonorrhea. In multivariate analysis, only injection drug use and HIV infection were significantly associated with H C V seropositivity. When assessing the results of previous studies of H C V infection among homosexual men, several common findings emerge. In all of these studies, HCV prevalence rates were substantially higher than the general populations from which these participants were selected. Prevalence in these studies ranged from 2.9 to 12.9 percent. There is evidence that prevalence of H C V is higher among men who are HIVpositive compared with HIV-negative men. Prevalence of H C V among gay men also appears to be higher in those with other sexually transmitted infections such as genital herpes and rectal gonorrhea. In three American studies and one Australian study that were reviewed, injection drug use was identified as the primary risk factor for H C V acquisition. " In some of these studies, 17  20  specific sexual practices were identified as significant or marginal risk factors in univariate analyses, including number of sexual partners, frequency of anal intercourse, and fisting. These findings suggest that risk of sexual transmission of H C V infection may be increased with specific sexual acts or that they are confounded with injection  28  drug use. In contrast, the majority of European studies have not found any evidence of an association between H C V seropositivity and sexual behaviour.  2.3  N A T U R A L HISTORY STUDIES OF H C V / H I V CO-INFECTION The majority of studies of individuals who are co-infected with H C V and HIV  have concluded that, while the progression of HIV disease is not strongly influenced by H C V infection, hepatitis C progresses more rapidly in individuals with co-infection.  2126  Most of these studies showed that individuals with H C V / H I V co-infection experience more rapid progression to cirrhosis, and have more evidence of extensive liver damage.  27  Staples and colleagues studied time from HIV diagnosis to AIDS, time from  HIV diagnosis to death, and time from AIDS diagnosis to death among H C V / H I V coinfected individuals. They did not find any of these measures of disease progression to significantly differ between H C V / H I V co-infected and HIV-infected groups. In contrast, a study conducted by Sabin and his colleagues reported that hemophiliacs infected with H C V genotype 1 experienced more rapid progression to AIDS and AIDSrelated death compared to other genotypes.  28  2.4  T H E V A N C O U V E R LYMPHADENOPATHY-AIDS STUDY 2.4.1  Study Description  The Vancouver Lymphadenopathy-AIDS Study (VLAS) is a prospective cohort study involving 1,000 homosexual/bisexual men who were enrolled during two recruitment periods. From November 1982 to December 1984, a total of 729 gay and bisexual men aged 18 to 75 were recruited through six general medical practices in Vancouver. Two of the practices provided care almost exclusively to homosexual men, and the other practices had substantial numbers of homosexual male patients. During the period October 1986 to December 1987, two more practices were added to the study 29  and 271 additional men were recruited. Throughout these recruitment periods, any regular patient known to be predominantly or exclusively homosexual was invited by his practitioner to participate in the prospective study. The refusal rate was low, approximately five percent. Only patients who were already enrolled in these medical practices were eligible thereby minimizing self-referral of other participants who might have been aware of this study. All participants gave informed consent. In the first cohort, follow-up visits occurred approximately every six months until September 1986, after which, participants in both cohorts completed visits on an annual basis. During each physician visit participants completed a self-administered questionnaire. Variables of interest included demographic characteristics such as age, ethnicity, income, and education; sexual behavior variables including the number of male sexual partners; frequency of receptive and insertive anal intercourse; frequency of condom use during anal intercourse; and use of tobacco, alcohol, and other drugs. Data concerning sexual practices and substance use were self-reported and described behaviors which occurred during the previous 12 months. In addition, the participants underwent a complete physical examination at each visit and blood samples were drawn for laboratory testing. 2.4.2  Follow-up of cohort  Between October 1982 and December 1984, a total of 729 participants were enrolled in the first cohort. These men were actively followed for a median of 128 months (range 26-180 months). The earliest enrolment visit in this cohort took place in October 1982 and the latest visit occurred during last cycle of physician visits in December 1998. The median number of completed visits was 9 (range 2 to 16). At the conclusion of the ninth cycle of physician visits which occurred during the period from October 1,1988 to December 31,1989, a total of 392 (54%) of the 729 men enrolled in the first cohort had completed the follow-up visit. Ninety-seven (13%) participants had 30  been diagnosed with AIDS, 95 (13%) did not attend the ninth cycle, 87 (12%) had moved away from Vancouver, 43 (6%) could not be contacted, 12 (2%) had transferred to nonstudy physicians and 3 (0.3%) had died from causes unrelated to AIDS. A total of 271 men were enrolled in the second cohort between September 1986 and December 1987. These men were followed for a median of 75 months (range 6-128 months). In this cohort, the earliest visit occurred in September 1986 and the latest visit occurred in December 1998. The median number of physician visits completed by these men was 5 (range: 2 to 10). As of December 31 1989, a total of 197 (73%) of the 271 men had completed the follow-up visit, 25 (9%) had been diagnosed with AIDS, 18 (7%) had moved away from Vancouver, 12 (4%) could not be contacted, 9 (3%) did not attend the ninth cycle,9 (3%) had transferred to non-study physicians and 1 (0.3%) had died from causes unrelated to AIDS. 2.4.3  HCV antibody testing procedure and results  All VLAS participants who completed two or more physician visits and provided blood specimens underwent H C V antibody testing. Antibody testing was carried out using stored sera obtained during the period 1982 to 1998, and the testing of these specimens was conducted between August 1999 and March 2000. H C V antibody was assayed by enzyme immunoassay (EIA), using a modified algorithm to confirm seropositivity. This testing algorithm was based on a study carried out in the University of British Columbia Diagnostic Virology and Reference Laboratory and the B.C. Centre for Disease Control Virology Laboratory. Variations of this algorithm have 29  been widely used in diagnostic virology laboratories throughout Canada. Briefly, a micro-particle enzyme immunoassay (Abbott AxSYM H C V Version 3.0 MEIA) was used to screen all sera. Negative sera were tested no further. Positive samples, as well as samples with an indeterminate result, were further tested using a synthetic peptidebased EIA (United Biomedical Inc. H C V EIA 4.0). Sera that tested positive on both EIAs 31  were classified as positive and tested no further. Sera that gave discordant results were further tested by immunoblot (RIBA III; Ortho Diagnostic Systems). Sera that were negative by RIBA were classified as negative. Sera were available for 932 of the 1000 VLAS participants. Using the testing algorithm described above, a total of 54 men were classified as H C V antibody positive. Thirty-nine infections were identified among 662 members of the first cohort, whereas 15 infections were detected among 270 participants in the second cohort. Among seropositive men in the first cohort, the numbers of prevalent and incident infections were 26 and 13, respectively. In this group, the earliest date of H C V seropositivity occurred in January 1983 and the latest date in February 1997. All 15 H C V infections identified in the second cohort were deemed prevalent. The dates of H C V seropositivity for members of the second cohort took place between October 1986 and January 1988. 2.4.4  Data management  All questionnaire, laboratory and serology data were stored on a Sun Microsystems computer. Because no identifying information on any participant was allowed to leave his physician's office, all participants were identified by a numeric code. Data for participants were stored in databases that were designed to facilitate cross-referencing. For example, the primary database contained all questionnaire and laboratory data, serologic information (HCV and HIV antibody test results), dates of seroconversion for seroincident men, dates and types of diagnoses for AIDS cases, and dates of death. The secondary database contained information regarding stored blood specimens. The third database contained information regarding completed follow-up visits. All databases were maintained and updated on an ongoing basis by study personnel. All statistical analyses presented in this dissertation were conducted using SAS statistical software (version 6.12).  30  32  2.4.5  Description of statistical methods  Methods of statistical analysis used in this dissertation were dependent on the type of data being analyzed and the study design being used. For cross-sectional comparisons of demographic and behavioural variables, both parametric and nonparametric statistical methods were used to analyze the data. Bivariate categorical data were analyzed using Pearson's chi-squared test. Fisher's exact test was used when twenty-five percent or more of the expected cell frequencies in a contingency table were less than five. Unadjusted relative risk estimates were computed using the odds ratio, and 95 percent confidence intervals were approximated using test-based limits proposed by Miettinen.  31  Bivariate comparisons of categorical and numeric variables  were conducted using Student's t-test and Wilcoxon's rank-sum test. Multivariate logistic regression analysis was used to model the independent association of several risk factors with H C V infection. Survival analytic methods were used to investigate and describe the time to H C V seropositivity among members of the first cohort. The critical event in these analyses was defined as 'time to H C V seropositivity'. Time zero was taken to be January 1983 for prevalent men, and defined as the midpoint between the last negative and first positive H C V antibody test for seroincident men. Cumulative incidence curves were generated, stratified over various putative risk factors and compared via the log-rank test. To model the simultaneous effect of several variables progression to H C V seropositivity, Cox proportional hazards analysis was used. A parsimonious approach was undertaken with respect to multivariate model building in this dissertation. This approach was undertaken because of the relatively small number of H C V infections identified in this study. When there is a small number of outcome events per independent variable in a multivariate model, the results from 33  the fitted regression model may not be accurate or precise.  32,33  In logistic regression  analysis, the number of outcome events is defined by the smaller number of binary outcomes (e.g. H C V seropositive vs. H C V seronegative) of the dependent variable. In proportional hazards analysis the number of outcome events is defined by the count of "failure" events (e.g. H C V infection). When the number of outcome events per variable analyzed is low, the accuracy, precision, and significance of the regression coefficients estimated by these methods will become untrustworthy. The results of proportional hazards analysis and logistic regression should be interpreted with caution when fewer than 10 events per variable are analyzed.  34,35  Since the number of outcome events in  cross-sectional comparisons and prospective analyses was less than 40, multivariate models presented in this dissertation contained no more than three independent variables. All demographic and behavioural variables that were found to be significantly or marginally associated with H C V seropositivity were considered for inclusion in multivariate models. Both forward selection and backward elimination procedures were used to select variables for final inclusion in multivariate models. 2.4.6  Rationale for using both cohorts  The decision to include or exclude the second cohort in a specific sub-analysis was made entirely by the candidate. The inclusion of both cohorts for some analyses described in this dissertation was not done to adjust or alter the overall findings in any way. The use of both cohorts was chosen for some sub-analyses for purpose of comparison, and to demonstrate consistency between estimates of H C V prevalence and incidence. The exclusion of the second cohort from some sub-analyses was primarily due to a lack of comparability between questionnaire items pertaining to sexual practices during the two different periods of participant recruitment. In Chapter 3, results are presented separately for members of the cohort who were enrolled during the first recruitment period (Cohort #1: November 1982 to December 34  1984) and for those who entered the prospective study during the second period of recruitment (Cohort #2: September 1986 to December 1987). The inclusion of cohort#2 in this sub-analysis was primarily for purpose of comparison. Men who were enrolled during the second recruitment period were less likely to be Caucasian, more likely to report an annual income greater than $10,000 and more likely to have attended college or university. Despite these demographic differences and the different time periods of recruitment, the point estimates of H C V prevalence were almost identical. Furthermore, combining the two cohorts resulted in increased precision of our overall estimate of the true population prevalence. In Chapter 4, a cross-sectional comparison between HCV-positive and HIVnegative men is presented. This sub-analysis was restricted to the first cohort. This was done in order to make use of the earliest available behavioural data, and because the format of sexual behaviour questions was markedly different from those used during the second period of recruitment. Throughout the years of follow-up of the VLAS cohort, individual questionnaire items were revised considerably to reflect new knowledge. As our knowledge of HIV/AIDS epidemiology expanded, the questions regarding sexual behaviour were revised and expanded. During the second recruitment period, questions concerning specific forms of male-to-male contact were asked separately for regular (i.e. at least once per month) and casual (i.e. less than once a month) sexual partners. No distinction was made between regular and casual partners during the first recruitment period. In addition, the measures of frequency of specific sexual practices differed between the two recruitment periods. Another reason for excluding the second cohort from the sub-analysis in Chapter 4 relates to behaviour change. After the discovery of HIV, numerous epidemiological studies of gay men were successful in identifying specific sexual practices and other behaviours that were significantly associated with increased risk of infection. Dissemination of preventive 35  public health messages in the gay community was followed shortly by reports of reductions in high-risk sexual behaviours, increased condom use, and lower HIV incidence rates. VLAS participants who were enrolled during the second recruitment period entered the study after the discovery of HIV and had the benefit of receiving these prevention messages designed to reduce the frequency high-risk behaviours. Consequently, their responses to questions about high-risk sexual behaviours could have been influenced by these prevention messages. Responses to questions concerning sexual behaviours, injection drug use, and substance use could be influenced by the participant's knowledge of their HIV or H C V antibody status. However, in the first cohort, responses to these questions were obtained two to three years before the participant's HIV antibody testing took place, and over fifteen years prior to H C V antibody testing. Thus responses to questions concerning sexual behaviours and substance use were not likely influenced by the participant's knowledge of their HIV or H C V antibody status. The sub-analyses described in Chapters 5 and 6 also excluded participants from the second cohort for the same reasons as in Chapter 4. In Chapter 7, both cohorts were included in the sub-analysis because data regarding symptoms and signs, physical exam findings, CD4 cell counts, diagnosis of AIDS, and survival were available for both cohorts, and the questionnaire items for these variables were directly comparable.  36  2.5  REFERENCES  1.  Alter MJ. Epidemiology of hepatitis C. Hepatology 1997; 26:62S-65S.  2.  Remis R. Epidemiology of hepatitis C infection in Canada. Presented at the First Canadian Conference on Hepatitis C, Montreal, Quebec, May 1-4,2001.  3.  Everhart JE, Di BA, Murray L M et al. Risk for non-A, non-B (type c) hepatitis through sexual or household contact with chronic carriers. Ann Intern Med 1990; 112:544-545.  4.  Eyster ME, Alter HJ, Aledort L M et al. Heterosexual co-transmission of hepatitis C virus (HCV) and human immunodeficiency virus (HIV) Ann Intern Med 1991; 115:746-748.  5.  Bresters D, Mause BE, Reesink HW et al. Sexual transmission of hepatitis C. Lancet 1993; 342:210-211.  6.  Hallam NF, Feltcher ML, Read SJ et al. Low risk of sexual transmission of hepatitis C virus. J Med Virol 1993; 40:251-253.  7.  Healey CJ, Smith DB, Walker JL et al. Acute hepatitis C infection after sexual exposure. Gut 1995; 36:148-50.  8.  Akahane Y, Kojima M , Sugai Y et al. Hepatitis C virus infection in spouses of patients with type C chronic liver disease. Ann Intern Med 1994; 120:749-752.  9.  Thomas DL, Zenilman JM, Alter HJ et al. Sexual transmission of hepatitis C among patients attending sexually transmitted diseases clinic in Baltimore - an analysis of 309 sex partners. J Infect Dis 1995; 171:768-75.  10.  Wu LC, Lin HC, Jeng FS, et al. Prevalence, infectivity, and risk factor analysis of hepatitis C virus infection in prostitutes. J Med Virol 1993; 39:312-317.  37  11.  Lissen E, Alter HI, Abad M A et al. Hepatitis C virus infection among sexually promiscuous groups and heterosexual partners of hepatitis C virus infected index cases. .Eur J Clin Microb Infect Dis 1993; 12:82-831.  12.  Hyams KC, Phillips IA, Tejada A, et al. Three year incidence study of retroviral and viral hepatitis transmission in a Peruvian prostitute population. J Acquir Immune Defic Syndr 1993; 6:1353-1357.  13.  Melbye M , Biggar RJ, Wantzin P, Krogsgaard K, Ebbesen P, Becker NG. Sexual transmission of hepatitis C virus: cohort study (1981-9) among European homosexual men. BMJ 1990; 301:210-212.  14.  Tedder RS, Gilson RJC, Briggs M , Loveday C, Cameron C H , Garrson JA, Kelly GE, Weller TVD. Hepatitis C virus: evidence for sexual transmission. BMJ 1991; 302:1299-1302.  15.  Gasparini V, Chiaramonte M , Moschen ME, Fabris P, Altinier G, Majori S, Campello C, Trivello R., Hepatitis C infection in homosexual men: A seroepidemiological study in gay clubs in North-East Italy. Eur J Epidemiol 1991; 7:665-669.  16.  Osella AR, Massa M A , Joekes S, et al. Hepatitis B and C virus sexual transmission among homosexual men. A m J Gastroenterol 1998; 93(l):49-52.  17.  Osmond D H , Charlebois, Sheppard HW, Page K, Winkelstein W, Moss A , Reingold A. Comparison of risk factors for hepatitis C and hepatitis B in homosexual men. J Infect Dis 1993; 167:66-71.  18.  Buchbinder SP, Katz M H , Hessol N A , Liu J, O'Malley PM, Alter MJ. Hepatitis C virus infection in sexually active homosexual men. Journal of Infection 1994; 29:263-269.  38  19.  Ndimbie OK, Kingsley LA, Nedjar S, Rinaldo CR. Hepatitis C virus infection in a male homosexual cohort: risk factor analysis. Genitourin Med 1996; 72:213216.  20.  Bodsworth NJ, Cunningham P, Kaldor J, Donavan B. Hepatitis C virus infection in a large cohort of homosexually active men: independent associations with HIV-1 infection and injecting drug use but not sexual behaviour. Genitourin Med 1996; 72:118-122.  21.  Wright TL, Hollander H , Pu X, et al. Hepatitis C in HIV-infected patients with and without AIDS: prevalence and relationship to patient survival. Hepatology 1994; 20:1152-1155.  22.  Sanchez-Quijano A, Andreu J, Gavilan F, et al. Influence human immunodeficiency virus type 1 infection on the natural course of chronic parenterally acquired hepatitis C. Euro J Clin Microbiol Infect Disl995; 14949953.  23.  Zylberberg H , Pol S. Reciprocal interactions between human immunodeficiency virus and hepatitis C virus infections. Clin Infect Dis 1996; 23:1117-1125.  24.  Soto B, Sanchez-Quijano A, Rodrigo L, et al. Human immunodeficiency virus modifies the natural history of chronically parenterally acquired hepatitis C with an unusually rapid progression to cirrhosis. J Hepatol 1997; 26:1-5.  25.  Piroth L, Duong M,Quantin C, et al. Does hepatitis C co-infection accelerate clinical and immunological evolution of HIV-infected patients? AIDS 12: 381388.  39  26.  Staples CT, Rimland D, Dudas D. Hepatitis C in the HIV (human immunodeficiency virus) Atlanta V.A. (Veterans Affairs Medical Center) cohort study (HAVACS): the effect of co-infection on survival, Clin Infect Dis 29:150-154.  27.  Garcia-Samaniego J. Soriano V, Castilla J, et al. Influence of hepatitis C virus genotype and HIV infection on histological severity of chronic hepatitis C. A m J Gastroenterol 1997; 92:1130-1134.  28.  Sab in A C , Telfer P, Philips A N , et al. The association between hepatitis C virus genotype and human immunodeficiency virus disease progression in a cohort of hemophilic men. J Infect Dis 199; 175:164-166.  29.  Cook D, Sherlock C H , Mak A, Littlewood RA, Karakas L, Middleton PJ. Comparison of 5 commercial EIA kits and 2 immunoblot kits for detection of hepatitis C antibody. Conjoint Meeting on Infectious Diseases, Montreal, 1994.  30.  SAS (r) Proprietary Software Release 6.12. Copyright (c) 1989-1996 by SAS Institute Inc., Cary, N C , USA. Licensed to University of British Columbia Computing Center, Site 0011073001.  31.  Miettinen O. Estimability and estimation in case-referent studies. A m J Epidemiol 1976; 103:226-35.  32.  Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariable models. Annals of Internal Medicine 1993; 118:201-210.  33.  Harrell Fe, Lee KL, Matchar DB, Reichert TA. Regression models for prognostic prediction: Advantages, problems, and suggested solutions. Cancer Treat Rep 1985; 69:101-1077.  40  34.  Concato J, Peduzzi P, Holford TR, Feinstein AR. Importance of events per independent variable in proportional hazards analysis: I. Background, goals, and general strategy. Journal of Clinical Epidemiology 1995; 48:1495-1501.  35.  Peduzzi P, Concato J, Feinstein AR, Holford TR. Importance of events per independent variable in proportional hazards analysis: II. Accuracy and precision of regression estimates. Journal of Clinical Epidemiology 1995; 48:1503-1510.  41  CHAPTER 3 PREVALENCE A N D INCIDENCE OF H C V INFECTION IN A C O H O R T OF H O M O S E X U A L M E N IN V A N C O U V E R (1983-98)  3.1  FORWARD In this chapter, the serological results of hepatitis C virus antibody testing for the  entire VLAS cohort are presented and discussed. For purpose of presentation, the results are shown separately for members of the cohort who were enrolled during the first recruitment period (Cohort #1: November 1982 to December 1984) and for those who entered the prospective study during the second period of recruitment (Cohort #2: September 1986 to December 1987). Both cumulative incidence and incidence density estimates are used to quantify the rate at which new H C V infections spread through the population of homosexual males in Vancouver during the period between 1983 and 1998. Prevalence and incidence data are summarized using both tabular and graphical methods. The candidate identified and selected serum specimens from archived samples for H C V antibody testing. The selected specimens were delivered to Dr. Chris Sherlock's laboratory where they were tested. The candidate managed the serological data, linked the H C V test results to questionnaire data, performed statistical analyses, summarized and interpreted the findings, and wrote the summary report.  42  3.2  INTRODUCTION The World Health Organization has estimated that three percent of the world's  population (approximately 170 million) is infected with Hepatitis C Virus (HCV). In 1  the United States, it is estimated that four million persons are infected. H C V has been 2  established as an etiologic agent of acute and chronic liver disease and cirrhosis of the liver, and has been found to be associated with development of hepatocellular carcinoma. " Furthermore, the onset of cirrhosis or liver failure may be accelerated by 3  7  co-infection with HIV leading to increased morbidity and mortality. " 8  9  In Canada, prevalence of H C V has been estimated to be 0.8% in the general population, and is assumed to be higher among males (0.96%) compared to females (0.53%).  Higher prevalence rates have been reported among injection drug users in  10-11  Vancouver , among street youth in Montreal and Ottawa , and among dialysis 12  13  patients in Alberta.  15  14  However, the true burden of hepatitis C virus (HCV) infection in  Canada remains largely unknown. Population-based prevalence studies are lacking and there are no ongoing studies investigating the annual incidence of HCV. Previous studies conducted outside Canada have reported elevated H C V prevalence rates among men who have sex with men compared to the general population. " To date, no prevalence/incidence studies have been reported in the 16  22  homosexual male population in Canada. The primary objective of this study was to estimate the prevalence and annual incidence of hepatitis C virus infection in the male homosexual population in Vancouver during the period from 1983 to 1998. We also investigated H C V incidence in parallel with HIV incidence during the same time period.  43  3.3  METHODS 3.3.2  Study Description  The Vancouver Lymphadenopathy-AIDS Study (VLAS) followed and monitored a cohort of homosexual men since November 1982. From November 1982 to December 1984, over 700 men were recruited through six general practices in central Vancouver. Two medical practices were added to the study and approximately 300 additional participants were enrolled during the period from October 1986 to December 1987. Follow-up visits occurred approximately every six months until September 1986 after which subjects completed visits on an annual basis. Active follow-up of this cohort continued until December 31,1998. At each visit, participants provided a blood sample for routine laboratory and serological testing. Serum samples were cryopreserved at minus 20° C. Since 1992, plasma specimens from study participants were cryopreserved at minus 80° C. All participants who completed two or more physician visits during the observation period and who provided blood specimens underwent H C V and HIV antibody testing. Testing was carried out using stored serum specimens obtained from study participants during the period 1982 to 1998. H C V antibody testing of these specimens was carried out between August 1999 and March 2000. 3.3.2  HCV Antibody Testing  H C V antibody was assayed by enzyme immunoassay (EIA), using a modified algorithm to confirm positivity. This testing algorithm was based on a study carried out in the University of British Columbia Diagnostic Virology and Reference Laboratory and the B.C. Centre for Disease Control Virology Laboratory.  23  Variations of this  algorithm have been widely used in diagnostic virology laboratories throughout 44  Canada. Briefly, a micro-particle enzyme immunoassay (Abbott AxSYM H C V Version 3.0 MELA) was used to screen all sera. Negative sera were tested no further. Positive samples and those with an indeterminate result were further tested using a synthetic peptide-based EIA (United Biomedical Inc. H C V EIA 4.0). Sera that tested positive on both EIAs were classified as positive and tested no further. Sera that gave discordant results were further tested by immunoblot (RIBA III; Ortho Diagnostic Systems). For this study, the H C V antibody testing was done on the specimens collected between November 1982 and December 1998. 3.3.3  HIV Antibody Testing  The National Reference Laboratory in Ottawa, using the enzyme-linked immunosorbent assay (ELISA) carried out HIV antibody testing. Positive results were verified using Western blot. 3.3.4  Statistical Methods  Estimates of H C V prevalence were obtained using data from both separate and combined cohorts. Ninety-five percent confidence intervals were also calculated. Both cumulative incidence and incidence density were used to quantify the burden and spread of H C V in the VLAS cohort. To study cumulative incidence, we examined time to the first positive HCV-antibody test for all men who were enrolled during the first recruitment period. In this analysis, time zero was defined as January 1,1983. Participants were right-censored at the date of their last follow-up visit. A sub-analysis of 'time to the first positive HCV-antibody test' was also performed. In this sub-analysis, men who were found to be H C V seroprevalent at the time of their first laboratory visit were excluded. This analytic approach was undertaken to reduce prevalence bias in the estimate of the population cumulative HCV incidence rate. Kaplan-Meier product-limit methods were used to estimate cumulative and annual incidence. For the incidence density approach, incidence was expressed in 45  terms of person years of observation. Using this approach, person years were calculated from January 1,1983 for cohort#l and from September 1,1986 for members of cohort #2. These incidence density estimates were not calculated as estimates of the true values but for purposes of comparison between cohorts.  3.4  RESULTS 3.4.2  Demographic characteristics  Demographic characteristics of the VLAS cohort members were stratified by recruitment period and presented in table 3.1. Men who were enrolled during the first recruitment period were more likely to be Caucasian, less likely to report an annual income above ten thousand dollars, and less likely to have attended college or university than men belonging to cohort #2. The mean age at enrolment was similar between the two cohorts. 3.4.2  HCV prevalence  Serum samples were available for a total of 932 men and 54 (5.8%) of these men tested positive for H C V antibody [95% CI: 4.3%, 7.3%]. Prevalence of H C V was similar in the two cohorts. A total of 39 (5.9%) of 662 men who were enrolled during the first recruitment period tested H C V positive [95% CI: 4.1%, 7.7%]. Among 270 participants who were enrolled during the second recruitment period, a total of 15 (5.6%) men tested positive [95% CI: 2.8%, 8.3%]. Overall, approximately 1 in 17 men in the VLAS cohort tested positive for H C V antibody. 3.4.3  HCV cumulative incidence  Figure 3.1 shows the Kaplan-Meier curve for 'time to first positive H C V antibody test' for men who were enrolled during the first recruitment period. These 662 men were followed for an average of 99 months (or 8.25 years). As previously mentioned, a total of 39 events were ascertained in cohort #1 during the observation period. As seen 46  in figure 3.1, after 14 years of follow-up cumulative H C V incidence was 7.9% [95% CI: 5.1%, 10.7%]. It was noted the incidence curve rose dramatically during the first 12 to 24 months of follow-up. This increase was likely related to a significant number of prevalent infections in the cohort during this period. Further H C V testing of all available serum specimens obtained from the 39 H C V positive men was conducted in order to distinguish between prevalent and incident infections. Further testing revealed that 2 out of 3 HCV-positive cases were prevalent infections (26 prevalent and 13 incident). Figure 3.2 shows the cumulative incidence curve when these 26 prevalent men were excluded from the analysis. This analysis included 626 men who were HCV-negative as of January 1,1983. The revised estimate of the cumulative incidence in the population of homosexual men in Vancouver was 4.0% [95% CI: 1.5%, 6.5%]. H C V incidence is shown for five consecutive 3-year intervals in figure 3.3. During the period between 1983 and 1997, annual H C V seroconversion rates ranged from 0 to 1.4 percent. 3.4.4  HCV incidence density estimates  A total of 662 men in cohort #1 were followed for a total of 5,474 person years. Incidence density was estimated as 7.1 per 1,000 person years [95% CI: 4.9, 9.4]. In comparison, 270 men who joined cohort #2 were followed for 1,890 person years, and 15 were H C V positive. This yielded an incidence density estimate of 7.9 per 1,000 person years [95% CI: 3.9,11.9]. The entire cohort of 962 men was followed for a total of 7,364 person years and the incidence density estimate was 7.3 per 1,000 person years based on 54 seropositive men in the cohort [95% CI: 5.4, 9.3]. 3.4.5  Comparison of HCV and HIV incidence  Comparative Kaplan-Meier curves for H C V and HIV are shown in Figure 3.4. A total of 445 men were HIV-negative as of January 1,1983. One hundred and thirty-six HIV seroconversions were documented during follow-up. HIV cumulative incidence 47  was approximately 8 times higher than H C V incidence during the observation period. After 14 years of follow-up, cumulative HIV incidence was 34.8% [95% CI: 29.8%, 39.7%} compared to 4.0% [95% CI: 1.5%, 6.5%] for HCV. Rates of H C V and HIV seroconversion during 1983-97 are compared in Figure 4.5 The rate of HIV infection was highest during the period 1983-85 and declined significantly thereafter. In comparison, annual H C V infection was significantly lower throughout the duration of the study. Despite the 'closed cohort effect', HIV rates remained at 1 to 4 percent during the 3-year time intervals from 1989 onward. However, the pattern of incidence for H C V infection was different. Unlike HTV, there was no evidence of a decline in H C V incidence during the latter years of the study. Incidence ranged from 0.8% to 1.8% during four of the five 3-year intervals shown in Figure 4.5. The highest H C V seroconversion rate occurred during the period 1995-97.  3.5  DISCUSSION Using stored serum specimens, we documented a high prevalence of H C V  among homosexual men residing in Vancouver during 1983-98. The prevalence observed in our study was approximately 6 percent, and this value was consistent with rates of H C V prevalence reported in other studies of homosexual men conducted in other geographic locations.  17,19,22  Prevalence observed in the Vancouver cohort was 7  times higher than the estimated H C V prevalence in the general Canadian population, and 6 times greater than prevalence estimated for Canadian males.  10,11  Our study  confirms findings from previous observational studies that H C V prevalence is substantially higher among men who have sex with men compared to the general population. To our knowledge, this is the first population-based, prospective study to report rates of H C V incidence in homosexual men in Canada. Annual infection rates among 48  men enrolled in the study remained less than one percent for each year of observation with the exception of 1997. The cumulative incidence rate observed in the Vancouver cohort (about 4 percent) was in agreement with the results reported by a Danish study of homosexual males. The pattern of H C V incidence was stable and remained low 16  throughout the duration of the study. However, there was no evidence of a decline in H C V infections despite the closed cohort effect. This observation was even more prominent when comparing trends between H C V and HIV incidence during 1983-98. Rates of HIV infection were significantly higher among homosexual men in Vancouver throughout these years and tended to decrease over time whereas H C V rates remained fairly stable. We have documented elevated prevalence and incidence of H C V infection in a cohort of homosexual men in Vancouver during 1983-98. During this period, we found approximately 1 in 17 men to be H C V positive. The reasons for this observed elevation in prevalence among gay men need to be further explored including the relative contribution of various routes of H C V transmission. In addition to injection drug use, the role of sexual transmission in the spread of H C V in this population requires clarification. Documentation of specific risk factors is necessary in order to elucidate various ways H C V can enter the human body, and to identify specific sexual practices that carry the greatest risk of H C V transmission.  49  Table 3.1: Comparison of V L A S cohorts with respect to demographic characteristics.  Cohort #1 Variable  n(%)  Cohort #2 /  'n(%)  p-value  Ethnicity Caucasian  632  (97.5)  251  (94.4)  16  (2.5)  15  (5.6)  < $10,000  161  (24.3)  50  (18.4)  > $10,000  501  (75.7)  221  (81.6)  291  (44.0)  162  (59.8)  371  (56.0)  109  (40.2)  Non-Caucasian  0.016*  Annual Income 0.052  Education level Attended College or UniversityGrade 12 or less  <0.001  Age at enrollment (in years) Mean Standard deviation * Based on Pearson's chi-squared test t Based on t-test for independent samples  50  32.1  32.2  7.0  6.3  0.713  +  igure 3.1: Cumulative Incidence of Hepatitis C Infection in the VLAS cohort (1983-1998)*  * Time to first positive test (seroprevalent men included) Cohort #1: n = 662 participants (39 events)  [6  '  v  0  1  1  1  1  12  24  36  48  1  1  1  1  1  i  1  1  1  1  1  60 72 84 96 108 120 132 144 156 168 180 Months since January 1983  51  Figure 3.2: Revised Cumulative Incidence of HCV Infection in the VLAS cohort (1983-1998)* 1098-  * Time to first positive test (seroprevalent men excluded) Cohort #1: n = 636 participants (13 events)  76543210<^ 0 12  24  36  48  60  72  84  —I  1  1  Months since January 1983  52  1  1  1  1  96 108 120 132 144 156 168 180  53  Figure 3.4: Comparison of H C V and HIV cumulative incidence in the VLAS cohort (1983-98) 40 H I V (n=445,136 events)  10  0  —-H  H C V (n=639, 13 events)  v  0  u  " r "I 12 24  I1  "  ' i 1 1 1 36 48  1  60  r  -  1  2  72  -  ! — i1 — i 1— i 1 — i — 1 1 i —1 i — 1i — 1i — 11 84 96 108 120 132 144 156 168 180  Months since January 1983  54  Figure 3.5: Comparison of H C V and H I V incidence i n the V L A S cohort (1983-97)  55  3.6  1  REFERENCES  Report of a World Health Organization consultation organized in collaboration with the Viral Hepatitis Prevention Board, Antwerp, Belgium. Global surveillance and control of hepatitis C. Journal of Viral Hepatitis 1999; 6:35-47.  2  Alter MJ, Kruszon-Moran D, Nainan OV, et al. The prevalence of hepatitis C virus infection in the United States, 1988 through 1994. N Engl} Med 1999; 341:556-562.  3  Alter HJ, Seeff LB. Transfusion-associated hepatitis. In: Zuckerman AJ, Thomas H C (Eds): Viral Hepatitis: Scientific Basis and Clinical Management. Edinburgh, Churchill Livingstone, 1993.  4  Alter HJ, Margolis HS, Krawczynski K, et al. The natural history of communityacquired hepatitis C in the United States. N Engl} Med 1992; 327:1899-1905.  5  Hoofnagle JH, Tralka TS. Management of Hepatitis C. Hepatology 1997; 26 (3 Suppl 1).  6  Bukh J, Miller RH, Kew MC, et al. Hepatitis C virus RNA in southern African blacks with hepatocellular carcinoma. Proc Natl Acad Sci USA 1993; 90:1848-1851.  7  IARC Working Group, World Health Organization: Hepatitis C virus. IARC Monograph 1994; 59:165-221.  8  Daar ES, Hynn H , Donfield S, et al. The relationship between hepatitis C virus and HIV infection in hemophiliacs. Program and abstracts of the 7 Conference on th  Retroviruses and Opportunistic Infections; January 30-February 2, 2000; San Francisco, California; Abstract 280. 9  Ragni M. Impact of HIV on progression to end-stage liver disease in HCV co-infected hemophiliacs. Program and abstracts of the 7 Conference on Retroviruses and th  Opportunistic Infections; January 30-February 2, 2000; San Francisco, California; Abstract 281.  56  10.  Remis R, Hogg R, Krahn MD, et al. Estimating the number of blood transfusion recipients infected by hepatitis C virus in Canada, 1960-85 and 1990-92. Report to Health Canada (1998).  11.  Zou S, Tepper M, Giulivi A. Current status of hepatitis C in Canada. Canadian Journal of Public Health 2000; 91(1):S10-S15.  12.  Strathdee S, Patrick D, Currie S, et al. Needle exchange is not enough: lessons from Vancouver injecting drug use study. AIDS 1997; 11:F59-F65.  13.  Roy E et al. Hepatitis B and C among street youth in Montreal - final report. Submitted to LCDC (1997), Division of HIV/AIDS and Division of Bloodborne Pathogens.  14.  Slinger R, El Saadany S, Tepper M , et al. Seroprevalence of and risk factors for hepatitis C and hepatitis B in street youth in Ottawa, Canada. Paediatr Child Health 1999;4(Suppl B):48B.  15.  Sandhu J, Preiksaitis JK, Campbell PM et al. Hepatitis C prevalence and risk factors in the northern Alberta dialysis population. Am J Epidemiol 1999;150(l):58-66.  16.  Melbye M , Biggar RJ, Wantzin P, Krogsgaard K, Ebbesen P, Becker NG. Sexual transmission of hepatitis C virus: cohort study (1981-9) among European homosexual men. BMJ 1990; 301:210-212.  17.  Gasparini V, Chiaramonte M , Moschen ME, Fabris P, Altinier G, Majori S, Campello C, Trivello R., Hepatitis C infection in homosexual men: A seroepidemiological study in gay clubs in North-East Italy. Eur J Epidemiol 1991; 7:665-669.  18.  Osella AR, Massa M A , Joekes S, et al. Hepatitis B and C virus sexual transmission among homosexual men. A m J Gastroenterol 1998; 93(l):49-52.  57  19.  Osmond D H , Charlebois, Sheppard HW, Page K, Winkelstein W, Moss A, Reingold A. Comparison of risk factors for hepatitis C and hepatitis B in homosexual men. J Infect Dis 1993; 167:66-71.  20.  Buchbinder SP, Katz M H , Hessol N A , Liu J, O'Malley PM, Alter MJ. Hepatitis C virus infection in sexually active homosexual men. Journal of Infection 1994; 29:263-269.  21.  Ndimbie OK, Kingsley LA, Nedjar S, Rinaldo CR. Hepatitis C virus infection in a male homosexual cohort: risk factor analysis. Genitourin Med 1996; 72:213-216.  22.  Bodsworth NJ, Cunningham P, Kaldor J, Dona van B. Hepatitis C virus infection in a large cohort of homosexually active men: independent associations with HIV-1 infection and injecting drug use but not sexual behaviour. Genitourin Med 1996; 72:118-122.  23.  Cook D, Sherlock C H , Mak A, Littlewood RA, Karakas L, Middleton PJ. Comparison of 5 commercial EIA kits and 2 immunoblot kits for detection of hepatitis C antibody. Conjoint Meeting on Infectious Diseases, Montreal, 1994.  58  CHAPTER 4 RISK FACTORS FOR HEPATITIS C VIRUS INFECTION IN T H E V L A S C O H O R T  4.1  FORWARD This chapter has been prepared as a manuscript for submission to a scientific  journal. The title of the manuscript is "Risk factors for hepatitis C virus infection in a cohort of homosexual men." Co-authors of this study include Drs. Mark Tyndall, Christopher Sherlock, Robert Hogg, Michael O'Shaughnessy and Martin Schechter. Drs. Schechter, Hogg, and Sherlock are members of the supervisory committee. Dr. O'Shaughnessy provided funding for the hepatitis C antibody testing of stored blood specimens from the VLAS cohort, and Dr. Sherlock supervised the virology laboratory staff that conducted the testing of the specimens. The candidate is the first author on this manuscript, and formulated the study design, and performed all data collection and management, statistical analysis, as well as writing the initial draft and final manuscript.  4.2  INTRODUCTION The World Health Organization has estimated that three percent of the world's  population (approximately 170 million) is infected with Hepatitis C Virus (HCV). In 1  the United States, it is estimated that four million persons are infected. In Canada, 2  H C V infection ranks second among nationally reported diseases with an estimated 225,000 to 300,000 individuals currently infected.  3  H C V has been established as an etiologic agent of acute and chronic liver disease and cirrhosis of the liver, and has been found to be associated with development of  59  hepatocellular carcinoma.  4-8  Furthermore, the onset of cirrhosis or liver failure may be  accelerated by co-infection with HIV leading to increased morbidity and mortality.  910  In both the United States and Canada, the majority of H C V infections has been associated with injection drug use or with transfusion with blood or blood products prior to 1990.  2,11/12  In contrast to the efficient transmission by parenteral exposure, there  is conflicting evidence regarding the sexual transmission of H C V despite the presence of viral particles and H C V RNA in saliva and genital fluids.  1314  Some evidence of sexual transmission of H C V has been provided by studies of heterosexuals and female sex workers. " In addition, CDC Sentinel county studies 15  19  have suggested that 15 to 20 percent of patients with acute hepatitis C infection have a history of sexual exposure in the absence of other risk factors. However, other studies 20  of heterosexuals have shown that transmission rates are low among monogamous, stable sexual partners of individuals with chronic H C V infection. " 21  23  Evidence of sexual transmission of H C V has also been provided by studies of homosexual men.  24  A study in Italy conducted among homosexual males who did not  use injection drugs, reported that prevalence of H C V was eleven times higher compared to male blood donors from the same area (18.9% vs. 0.6%). An American 25  study concluded that H C V infection in the male homosexual population is associated with sexual practices such as insertive anal intercourse. In contrast, other studies have 26  concluded that sexual practices appear to play a minor role in its transmission, and sexual transmission is rare in this population. " 27  30  The objectives of this study were to determine the prevalence of H C V infection in a cohort of sexually active homosexual men and to identify risk factors associated with  60  H C V infection including injection drug use, sexual practices, substance use, history of blood transfusions, and HIV infection.  4.3  METHODS 4.3.1  Study Description  The Vancouver Lymphadenopathy-AIDS Study (VLAS) recruited 729 homosexual men aged 18 to 75 into a longitudinal study over the period November 1982 to December 1984. Participants were recruited through six primary care practices in central Vancouver. Follow-up visits occurred approximately every six months until September 1986, after which, participants completed visits on an annual basis. During each physician visit, participants completed a self-administered questionnaire that elicited information regarding demographic variables, sexual practices, substance use and other variables. In addition, blood samples were drawn for immunologic and HIV antibody testing. HIV antibody testing was routinely carried out for all participants throughout the follow-up period. Serum samples were stored at minus 20° C. 4.3.2  Sources of data  The latest available serum specimen was selected for each participant for H C V antibody testing. A total of 662 specimens were available for antibody testing, and these specimens were tested between August 1999 and March 2000. A total of 67 men did not have specimens available for H C V antibody testing. These participants either failed to provide blood samples during the period of follow-up, or their specimens had been depleted as a result of previous laboratory testing. H C V antibody was assayed by enzyme immunoassay (EIA) using a modified algorithm to confirm positives. This algorithm was based on a study carried out in the University of British Columbia  61  Diagnostic Virology & Reference Laboratory and the British Columbia Centre for Disease Control Virology Laboratory.  31  To assess differences i n sexual behaviors and other risk factors for H C V infection, we compared HCV-positive and HCV-negative men using data obtained from their enrolment questionnaire. This questionnaire was completed between November 1982 and December 1984. We chose this study visit for comparison because this was the earliest visit from which information on demographics, sexual behaviour, and history of recreational and injection drug use was available. Behavioural data from the enrolment questionnaire was used for these analyses for both prevalent and incident infections. For prevalent infections, putative risk behaviours could have occurred at any time prior to the enrolment visit. Data from the enrolment questionnaire was used for incident infections because comparable questionnaire items were not available from subsequent physician visits.  Data regarding history of blood transfusions were not available from the enrolment questionnaire and were obtained from two follow-up visits, which occurred within twelve months of the enrolment visit. 4.3.3  Statistical Methods  Categorical variables were compared between HCV-positive and HCV-negative men using Pearson's chi-squared test. Contingency tables that contained one or more expected counts of less than five were analyzed by Fisher's exact test. Comparisons of quantitative variables were carried out using Wilcoxon's rank-sum test. Odds ratios (unadjusted and adjusted) and 95 percent confidence intervals were calculated for all variables of interest. Variables that were significantly or marginally associated with H C V infection i n bivariate analyses were considered for inclusion i n multivariate 62  statistical models. Multivariate logistic regression analysis was used to model the independent association of several risk factors with H C V infection. 4.4  RESULTS 4.4.1  HCV Prevalence  A total of 39 of 662 men with available sera were identified as HCV-positive, yielding a prevalence of 5.9 percent [95% CI: 4.1, 7.7]. Prevalence of H C V infection was significantly higher among HrV-positive men compared to HIV-negative men (p<0.001). Thirty-one of 352 (8.8%) HIV-positive participants were HCV-positive [95% CI: 6.6,11.0], whereas only 8 of 310 (2.6%) HIV-negative men were identified as HCVpositive [95% CI: 1.4, 3.8]. 4.4.2  Demographic characteristics  HCV-positive men were similar to HCV-negative men with respect to demographic characteristics including age, ethnicity, and annual income (Table 4.1). Although HCV-positive men were less likely to report having attended university or college, this difference was not statistically significant. However, HCV-positives were significantly more likely to be infected with HTV compared to HCV-negative men. 4.4.3  Injection drug use  Fewer than ten percent of study participants reported using injection drugs. A total of 41 (6.2%) reported injection drug use during their lifetime. Injection drug use was a highly significant risk factor for H C V infection (Table 4.2). Men who reported using injection drugs during their lifetime were thirty times more likely to be H C V positive. 4.4.4  Use of psychoactive drugs  HCV-positive men were more likely than HCV-negative men to report using the following psychoactive drugs during the previous twelve months: methylene-dioxy amphetamine (MDA), lysergic acid diethyl amide (LSD), amphetamine, and cocaine 63  (Table 4.2). M e n w h o reported using one or more of these drugs during the past 12 months were 2.5 times more likely to be H C V positive [95% CI: 1.3, 5.1]. The prevalence of heroin use among study participants was low (<2%) and was similar i n both groups. 4.4.5  Use of tobacco and alcohol  M e n who were HCV-positive were twice as likely to report using tobacco (Table 4.3). Over ninety percent of study participants reported consuming alcohol and over seventy percent reported using marijuana during the previous 12 months. N o significant differences between HCV-positive and HCV-negative men were observed with regard to alcohol or marijuana use. HCV-positive men were twice as likely to report using nitrite inhalants or poppers, but this association was of marginal significance. 4.4.6  Sexual practices  HCV-positive and HCV-negative men were compared w i t h respect to numbers of male sexual partners and specific sexual practices (Table 4.3). Elevated numbers of male sex partners over the lifetime and during the previous year were both positively associated w i t h H C V infection i n univariate analyses. Positive associations were also noted for active oral-anal contact (rimming), and insertive fisting (i.e. insertion of fingers or hand into partner's rectum). M e n who reported engaging i n these sexual practices were three times more likely to be H C V positive. Weaker positive associations were observed w i t h respect to frequency of receptive oral sex, receptive fisting, and passive oral-anal contact. N o significant associations were detected w i t h respect to frequency of insertive oral sex or frequency of insertive anal intercourse. 4.4.7  History of blood transfusion  A total of 610 of the 662 participants (92.1%) completed at least one follow-up questionnaire i n the year following their enrolment into the study. Participants who completed one or more of these follow-up visits were asked if they ever received a 64  blood transfusion. N o significant difference was observed between HCV-positive and HCV-negative men w i t h respect to their history of blood transfusions (p=0.796). Three of 35 HCV-positive men (8.6%) reported receiving a blood transfusion i n their lifetime, compared to 57 of 557 (9.9%) i n the HCV-negative group. 4.4.8  Multivariate Analysis  The results of the multivariate logistic regression model are shown i n Table 4.4. Independent risk factors for H C V infection included injection drug use (adjusted OR=27.3; 95% CI: 12.4, 60.4), HIV-infection (adjusted OR=2.4; 95% CI: 1.0, 5.8), and insertive fisting (adjusted OR=2.2; 95% CI: 1.0, 4.8). When HIV-infection was excluded from this multivariate model, both injection drug use and insertive fisting remained statistically significant but the estimates of risk shown above d i d not change appreciably (data not shown). 4.5  DISCUSSION These data indicate an overall prevalence rate of H C V seropositivity of 5.9  percent among the 662 men tested. Our estimate is similar to prevalence rates reported in studies of gay and homosexual men i n different geographic locations and conducted during comparable time periods.  25-29  Prevalence rates i n these studies ranged from 2.9  to 9.2 percent. Our finding that H C V prevalence was significantly higher among H I V positive men is also consistent w i t h the findings of other studies. - 26  28  29  Our study confirms the role of injection drug use as the primary risk factor for H C V infection among gay and homosexual men. This finding is concordant with the results of previous studies, which found both univariate and multivariate associations between injection drug use and hepatitis C infection in this population. " 26  29  However,  there was a low prevalence of self-reported injection drug use among our study participants. Almost one-half of the HCV-positive men i n our study d i d not report a history of injection drug use. 65  After controlling for both injection drug use and HIV-antibody status, we found an association between H C V infection and insertive fisting. A study of homosexual men recruited from a municipal STD clinic in San Francisco, reported an association with H C V positivity and fisting, but no association was observed with respect to oralanal contact.  29  Sexual practices such as insertive fisting may cause damage to the mucosal barrier of the rectum, and result in tearing and bleeding. Lesions produced by such trauma could serve as a source or passage for pathological micro-organisms like HCV. Moreover, individuals who engage in this sexual practice in conjunction with other sexual practices such as oral-anal contact, may increase their likelihood of contact with HCV-infected blood or fluids. We found an elevated risk for H C V infection among men who reported using psychoactive drugs. Previous studies have reported positive associations between H C V infection and use of amphetamines and cocaine (26).  26,29  Our study of homosexual  men found increased risk among users of LSD and tobacco in univariate analyses, but these risks did not remain statistically significant after adjustment for other risk factors. In contrast to two studies which failed to detect an association between H C V positivity and HIV infection,  25,27  our data may suggest that concurrent HIV infection  may facilitate sexual transmission of HCV, independent of injection drug use. Other H C V prevalence studies of homosexual men conducted in San Francisco, Pittsburgh, and Sydney, Australia have also reported HTV infection as a significant risk factor for HCV-positivity.  26,28,29  Only one of these studies found HIV infection to be  associated with H C V positivity in multivariate analysis.  28  The failure of HIV infection  to retain significance in multivariate analysis in these studies could be due to residual confounding. We detected no association between HCV-infection and a history of blood 66  transfusion among our participants. This finding is in contrast to a study of homosexual men conducted by Osmond et al, which reported history of injection drug use and blood transfusion to be independent risk factors for H C V positivity.  27  However, our results are compatible with the those reported by Ndimbie et al.  26  who  found similar percentages of HCV-positive and HCV-negative men receiving a blood transfusion during the previous five years (5.3% vs. 2.9%). An important limitation in assessing whether a specific sexual practice is associated with H C V transmission is the inadequate control of confounders or differential misclassification within the variables under consideration for inclusion in the multivariate statistical model. Although insertive fisting remained significantly associated with H C V seropositivity after adjustment for injection drug use and HIV seropositivity in our multivariate model, the possibility must be acknowledged that insertive fisting was not directly associated with H C V infection but rather with other residual lifestyle factors not fully captured in the multivariate model. For example, if insertive fisting was associated with other parenteral exposures such as tattooing or body piercing this could explain its retention in the multivariate model. We were unable to assess the risk associated with tattooing, ear or body piercing, and needle borrowing/sharing, because these variables were not included as items in our enrolment and early, follow-up questionnaires. In addition, we did not evaluate occupational and nosocomial exposures. The role of horizontal exposure in the spread of H C V was not addressed in this study. For prevalent infections in this study, risk behaviours of interest could have occurred at any time prior to their enrolment visit. It is also possible that for seroincident men, high-risk behaviours could have occurred after their enrolment visit. As a result, our estimates of risk associated with specific sexual practices might be distorted. 67  The results of this study are based on prevalent H C V infections and self-reported behaviours. Responses to questions concerning sexual behaviours, injection drug use, and substance use may be influenced by the participant's knowledge of their HIV or H C V antibody status. However, in our study, responses to these questions were obtained two to three years before the participant's HIV antibody testing took place, and over fifteen years prior to H C V antibody testing. Neither HIV nor H C V antibody testing was available prior to 1984. Thus responses to questions concerning sexual behaviours and substance use were not likely influenced by the participant's knowledge of their antibody status in this study. In order to assess the effect of HTV infection on H C V transmission it is important to establish the temporal sequence of these infections in longitudinal studies. Distinguishing between prevalent and incident cases of H C V infection will permit estimation of the incidence rate of H C V seroconversion in risk groups of interest. Prospective analyses of risk factors identified in previous cross-sectional studies will help to identify and confirm predictors of H C V seroconversion. We are presently addressing these objectives in the Vancouver Lymphadenopathy-AIDS Study cohort. We conclude that hepatitis C virus can be spread by sexual transmission among gay and homosexual men, independent of injection drug use. We speculate that sexual practices that cause rectal trauma in combination with oral-anal contact, may increase the risk of exposure to HCV-infected blood and fluids.  68  Table 4.1: Demographic characteristics and HIV antibody status of 39 HCV-positive and 623 HCV-negative men  Variable  HCV-positive n (%)  HCVrnegative . ;,n'(%)  - p-value*  39 (100)  593 (97)  0.616  10 (26)  151 (24)  0.843  17 (44)  354 (57)  0.106  31 (79)  321 (52)  <0.001  Years  Years  40  40  26-51  21-69  Ethnicitv Caucasian  f  Annual Income < $10,000  Education College/University  HIV antibody status Positive  Age (at time of H C V testing) Median Range * Based on Pearson's chi-squared test t Based on Fisher's exact test £ Based on Wilcoxon's rank sum test  69  .713*  Table 4.2: Comparison of HCV-positive and HCV-negative participants with respect to self-reported use of injection drugs, psychoactive drugs, tobacco, alcohol, and other substances.  , • HCV-  Variable  1  1  Injection drug use during lifetime  positive , n(%)  ,v',Hoy- negative - n (%)  p-value  Odds Ratio r  95% CI  * tr  20  (51)  21  (3)  <0.001*  30.2  17.3, 52.6  amphetamine (MDA)  25  (64)  228  (37)  <0.001  3.0  1.6, 5.9  Amphetamine  11  (28)  84  (14)  0.013  2.5  1.2, 5.0  Tobacco  26  (67)  296  (48)  0.027  2.1  1.1,4.2  diethylamide (LSD)  19  (49)  158  (26)  0.002  2.0  1.4,5.1  Cocaine  18  (46)  180  (30)  0.029  2.0  1.1,3.9  Heroin  1  (3)  8  (1)  0.429*  2.0  0.3,15.7  Nitrite inhalants  28  (72)  341  (56)  0.052  2.0  1.0, 4.1  Marijuana  32  (82)  470  (77)  0.480  1.4  0.6, 3.1  Alcohol  34  (87)  554  (91)  0.407*  0.7  0.3,1.9  Methylenedioxy-  Lysergic acid  * Based on Fisher's exact test  70  Table 4.3: Comparison of HCV-positive and HCV-negative participants with respect to number of sexual partners and sexual practices  Variable  HCVpositive n (%)  HCVp-value negative n(%); •  Odds Ratio  95% CI .  Number of male sexual partners in previous vear > 20 partners  29  (74)  294  (48)  0.002  3.1  1.5,6.3  32  (82)  391  (64)  0.024  2.5  1.1,5.7  35  (90)  443  (73)  0.019  3.3  1.2, 8.9  37  (95)  522  (85)  0.100  3.2  0.8,12.4  19  (49)  162  (27)  0.003  2.6  1.4, 4.9  9  (23)  81  (13)  0.087  2.0  0.9,4.2  31  (79)  401  (66)  0.076  2.0  0.9, 4.4  29  (74)  444  (73)  0.818  1.1  0.5, 2.3  24  (62)  274  (45)  0.043  2.0  1.0,3.8  23  (59)  339  (55)  0.670  1.2  0.6,2.2  Number of male sexual partners in lifetime > 100 Oral-anal contact (active) Yes Oral-anal contact (passive) Yes Insertive fisting Yes Receptive fisting Yes Oral sex (receptive) > 25% of encounters Oral sex (insertive) > 25% of encounters Receptive anal intercourse > 25% of encounters Insertive anal intercourse > 25% of encounters  71  Table 4.4: Multivariate logistic regression model of risk factors for H C V infection  Variable Injection drug use (during  Unadjusted OR  95% CI  Adjusted OR  95% CI  30.2  17.3, 52.6  27.3  12.4, 60.4  3.6  1.7, 7.7  2.4  1.0, 5.8  2.6  1.4, 4.9  2.2  1.0, 4.8  lifetime - yes vs. no)  HIV infection (yes vs. no)  Insertive fisting (yes vs. no)  72  4.6  1.  REFERENCES  Report of a World Health Organization consultation organized in collaboration with the Viral Hepatitis Prevention Board, Antwerp, Belgium. Global surveillance and control of hepatitis C. Journal of Viral Hepatitis 1999; 6:35-47.  2.  Alter MJ, Kruszon-Moran D, Nainan OV, et al. The prevalence of hepatitis C virus infection in the United States, 1988 through 1994. N Engl J Med 1999; 341:556-562.  3.  Tepper M . The epidemiology of AIDS in Canada. Presented at the Update on Liver Disease and Hepatitis: Issues and Controversies, Scottsdale, Arizona, June 3-8.1998; http://www.hepnet.com/hepc/uldh98/tepper.html  4.  Alter HJ, Seeff LB. Transfusion-associated hepatitis. In: Zuckerman AJ, Thomas HC (Eds): Viral Hepatitis: Scientific Basis and Clinical Management. Edinburgh, Churchill Livingstone, 1993.  5.  Alter HJ, Margolis HS, Krawczynski K, et al. The natural history of communityacquired hepatitis C in the United States. N Engl} Med 1992; 327:1899-1905.  6.  Hoofnagle JH, Tralka TS. Management of Hepatitis C. Hepatology 1997; 26 (3 Suppl 1).  7.  Bukh J, Miller RH, Kew M C , et al. Hepatitis C virus RNA in southern African blacks with hepatocellular carcinoma. Proc Natl Acad Sci USA 1993; 90:1848-1851.  8.  IARC Working Group, World Health Organization: Hepatitis C virus. IARC Monograph 1994; 59:165-221.  9.  Daar ES, Hynn H , Donfield S, et al. The relationship between hepatitis C virus and HIV infection in hemophiliacs. Program and abstracts of the 7 Conference on th  Retroviruses and Opportunistic Infections; January 30-February 2, 2000; San Francisco, California; Abstract 280.  73  10.  Ragni M. Impact of HIV on progression to end-stage liver disease in HCV coinfected hemophiliacs. Program and abstracts of the 7 Conference on Retroviruses and th  Opportunistic Infections; January 30-February 2, 2000; San Francisco, California; Abstract 281. 11.  Chaudhary, RK, Mo T. Antibody to hepatitis C virus in risk groups in Canada. Can J Infect Dis 1992;3:27-29  12.  Prevention and control of hepatitis C: Guidelines and recommendations. Can Commun Dis Rep 1995;21S2  13.  Liou TC, Chang TT, Young KC, Lin XZ, Lin CY, Wu HL. Detection of H C V RNA in saliva, urine, seminal fluid, and ascites. J Med Virol. 1992;37:197-202.  14.  Rey D, Fritsch S, Schmitt C, Meyer P, Lang JM, Stoll-Keller F. Quantitation of hepatitis C virus RNA in saliva and serum of patients coinfected with H C V and human immunodeficiency virus. J Med Virol. 2001;63:117-119.  15.  Corona R, Prignano G, Melee A, Gentili G, Caprilli F, Franco E, Ferrigno L, Giglio A, Titti F, Bruno C, Verani P, Pasquini P. Heterosexual and homosexual transmission of hepatitis C virus: relation with hepatitis B virus and human immunodeficiency virus type I. Epidemiol. Infect. 1991;107:667-672.  16.  Osmond D H , Padian NS, Sheppard HW, Glass S, Shiboski SC, Reingold A. Risk factors for hepatitis C virus seropositivity in heterosexual couples. JAMA.1993;269:361-365  17.  Akhane Y, Kojima M , Sugai Y, Sakamoto M , Miyazaki Y, Tanaka T, Tsuda F, Mishiro S, Okamoto H , Miyakawa Y, Mayumi M . . Hepatitis C virus infection in spouses of patients with type C chronic liver disease. Ann Intern Med 1994;120:748-752.  74  18.  Thomas DL, Zenilman JM, Alter HJ, et al. Sexual transmission of hepatitis C virus among patients attending sexually transmitted diseases clinics in Baltimore - an analysis of 309 sex partnerships. J Infect Dis. 1995;171:768-775  19.  Kihara M , Imai M , Kondo M, Watanabe S, Kihara M , Soda K. Prevalence of hepatitis C virus and HIV infection in Japanese female prostitutes. Nippon Koshu Eisei Zasshi. Japanese Journal of Public Health 1993; 40:387-391.  20.  Alter MJ. Epidemiology of hepatitis C. Hepatology 1997;26:62S-65S.  21.  Gordon SC, Parel A H , Kulesza GW, Barnes RE, Silverman AL. Lack of evidence for the heterosexual transmission of hepatits C. Am J Gastroenterol 1992;87:1849-51.  22.  Tong MJ, Lai PPC, Hwang S-J, et al. Evaluation of sexual transmission in patients with hepatitis C infection. Clinical and Diagnostic Virology 1995;3:39-47.  23.  Brettler DB, Mannucci PM, Gringeri A, Rasko JE, Forsberg AD, Rumi MG, Garsia RJ, Rickard KA, Colombo M. The low risk of hepatitis C virus transmission among sexual partners of hepatitis C-infected hemophilic males: an international, multicenter study. Blood 1992;80:540-543.  24.  Tedder RS, Gilson RJC, Briggs M , Loveday C, Cameron C H , Garrson JA, Kelly GE, Weller IVD. Hepatitis C virus: evidence for sexual transmission. BMJ 1991;302:12991302.  25.  Gasparini V, Chiaramonte M , Moschen ME, Fabris P, Altinier G, Majori S, Campello C, Trivello R., Hepatitis C infection in homosexual men: A seroepidemiological study in gay clubs in North-East Italy. Eur J Epidemiol 1991;7:665-669.  26.  Ndimbie OK, Kingsley LA, Nedjar S, Rinaldo CR. Hepatitis C virus infection in a male homosexual cohort: risk factor analysis. Genitourin Med 1996;72:213-216.  75  27.  Osmond D H , Charlebois, Sheppard HW, Page K, Winkelstein W, Moss A, Reingold A. Comparison of risk factors for hepatitis C and hepatitis B in homosexual men. J Infect Dis 1993;167:66-71.  28.  Bodsworth NJ, Cunningham P, Kaldor J, Donavan B. Hepatitis C virus infection in a large cohort of homosexually active men: independent associations with HIV-1 infection and injecting drug use but not sexual behaviour. Genitourin Med 1996;72:118-122.  29.  Buchbinder SP, Katz M H , Hessol N A , Liu J, O'Malley PM, Alter MJ. Hepatitis C virus infection in sexually active homosexual men. Journal of Infection 1994;29:263-269.  30.  Melbye M , Biggar RJ, Wantzin P, Krogsgaard K, Ebbesen P, Becker NG. Sexual transmission of hepatitis C virus: cohort study (1981-9) among European homosexual men. BMJ 1990; 301:210-212.  31.  Cook D, Sherlock C H , Mak A, Littlewood RA, Karakas L, Middleton PJ. Comparison of 5 commercial EIA kits and 2 immunoblot kits for detection of hepatitis C antibody. Conjoint Meeting on Infectious Diseases, Montreal, 1994.  76  CHAPTER 5 SEXUAL PRACTICES ASSOCIATED WITH H C V POSITIVITY A M O N G N O N INJECTION D R U G USERS IN T H E V L A S C O H O R T  5.1  FORWARD I n this chapter, results of a c o m p a r i s o n b e t w e e n H C V - p o s i t i v e a n d H C V -  negative m e n w h o d i d not use injection d r u g s are p r e s e n t e d a n d d i s c u s s e d . The q u a n t i f i c a t i o n a n d e s t i m a t i o n of risk associated w i t h specific s e x u a l practices that cause rectal t r a u m a a n d increase the l i k e l i h o o d of e x p o s u r e to H C V - i n f e c t e d b l o o d a n d f l u i d s is e x p l o r e d m o r e t h o r o u g h l y . T h i s w a s a c h i e v e d b y e x c l u d i n g m e n w h o r e p o r t e d u s i n g injection d r u g s . T h e m a t e r i a l that is presented i n this chapter is i n t e n d e d to serve as a n a d d e n d u m to the results of the p r e v i o u s chapter. T h e c a n d i d a t e f o r m u l a t e d the s t u d y d e s i g n a n d h y p o t h e s i s , p e r f o r m e d a l l data m a n a g e m e n t , c o n d u c t e d statistical analyses, s u m m a r i z e d a n d i n t e r p r e t e d the f i n d i n g s , a n d w r o t e the m a n u s c r i p t .  5.2  INTRODUCTION T h e role of sexual contact i n the s p r e a d of hepatitis C v i r u s has b e e n  c o n t r o v e r s i a l . C o n t r o v e r s y has existed because of inconsistencies a m o n g p r e v i o u s studies. T h e efficiency of sexual t r a n s m i s s i o n a m o n g heterosexuals is generally t h o u g h t to be l o w , a n d i n f e c t i o n is rare i n l o n g - t e r m , steady partners. S o m e evidence of sexual t r a n s m i s s i o n of H C V has b e e n p r o v i d e d b y o b s e r v a t i o n a l studies of m e n w h o h a v e sex w i t h m e n . H o w e v e r , m o r e data are n e e d e d to i d e n t i f y h i g h - r i s k practices that are associated w i t h H C V t r a n s m i s s i o n b e t w e e n sexual partners.  77  In the previous chapter, it was concluded that H C V could be spread by sexual contact among gay and homosexual men, independent of injection drug use. Sexual practices such as insertive fisting may cause damage to the mucosal barrier of the rectum, and result i n tearing and bleeding. Lesions produced by such trauma could serve as a source or passage for pathological micro-organisms like H C V . Moreover, individuals w h o engage i n this sexual practice i n conjunction w i t h other sexual practices such as oral-anal contact (or rimming) may increase their likelihood of contact w i t h HCV-infected blood or fluids. The objective of this investigation was to estimate the relative risk of H C V transmission for specific sexual practices that cause rectal trauma and involve oral-anal contact among homosexual men who were not injection drug users.  5.3  METHODS 5.3.1  Study Description  The Vancouver Lymphadenopathy-AIDS Study (VLAS) was an observational study of 729 gay men who were recruited through their general practitioners i n central Vancouver between November 1982 and December 1984. Follow-up visits occurred approximately every six months until September 1986, after which participants completed visits on an annual basis. D u r i n g each visit, participants completed a selfadministered questionnaire, which gathered information regarding demographic variables, injection drug use, and sexual practices. In addition, blood samples were drawn for H C V and HTV antibody testing. For this study the H C V antibody testing was done on the specimen collected from the latest visit.  78  5.3.2  Study Variables  Demographic variables of interest in this investigation included age, ethnic group, education and annual income. Participants in the study were asked whether they had ever used injection drugs during their lifetime. Sexual behaviour variables of interest included numbers of male sexual partners during the participant's lifetime and the past year. Specific sexual practices that were studied included fisting (i.e. insertion of fingers or hand into the partner's rectum), oral-anal contact, and anal intercourse. We also studied the sexual practice of fisting in combination with oral-anal contact between sexual partners. This composite variable was of particular interest because it involved rectal trauma of one sexual partner, and potential oral exposure of the other partner to their sexual partner's HCV-infected blood or fluids. Study participants were asked to indicate whether they engaged in these sexual practices with their male partner(s) during the preceding 12 months. All demographic and behavioural data were obtained from the questionnaire administered at the time of the first physician visit.  5.3.2  Statistical Methods  Bivariate comparisons between HCV-positive and HCV-negative men were carried out using Pearson's chi-square test for categorical variables. Fisher's exact test was used when more than 25 per cent of the expected cell frequencies were less than five. Unadjusted relative risk estimates were computed using odds ratios, and 95% confidence intervals were approximated using test-based limits. Multivariate comparisons were conducted using logistic regression analysis. For all comparisons, participants with missing or unknown values for the variable(s) of interest were excluded from the analysis. All reported p-values are two-sided.  79  5.4  RESULTS  5.4.2  Injection drug use  In total, 662 VLAS participants had serum specimens available and were tested for H C V antibody. Of these, 39 (5.9%) men tested positive. Previous injection drug use was reported by 41 (6.1%) of 662 men, and by 20 of those who were HCV-positive. Men who reported using injection drugs during their lifetime were 30.2 times more likely to be H C V positive (95% CI: 17.3, 52.6). A total of 19 (48.7%) of the 39 men who tested positive did not report using any lifetime injection drugs use. To assess the risk of specific sexual practices for H C V transmission, we excluded men who reported using injection drugs in the study. Thus, comparisons between HCV-positive and HCVnegative men with respect to sexual practices were based on 621 participants. Among men who did not report using injection drugs in their lifetime, 19 (3.1%) were HCVpositive.  5.4.2  Demographic characteristics  HCV-positive men were similar to negative men with respect to demographic characteristics including age at enrolment (33 vs. 32 years; p=0.629), Caucasian ethnicity (100% vs. 97%; p=0.999), annual income less than $10,000 (16% vs. 24%; p=0.586), and post-secondary education (37% vs. 42%; p=0.632). A higher percentage of HCV-infected men were HIV-positive compared to uninfected men and this difference was marginally significant (74% vs. 51%; p=0.053).  80  5.4.3  Sexual practices  A higher percentage of HCV-positive men reported having 20 or more sexual partners during the previous 12 months (68% versus 48%; p = 0.078) yielding an odds ratio of 2.4 [95% CI: 0.9, 6.1]. We compared H C V positive and negative men with respect to specific sexual practices (Table 5.1). A higher percentage of HCV-positive men reported having practiced receptive anal intercourse in more than 25% of sexual encounters during the previous 12 months (63% versus 44%; p = 0.099) giving rise to an odds ratio of 2.2 (95% CI: 0.8, 5.6). Both active oral-anal contact and insertive fisting were found to be significantly associated with increased risk of H C V infection. Men who reported these sexual practices were 6.9 and 3.1 times more likely to be HCVpositive, respectively. In addition, those who reported engaging in both practices with their sexual partners during the previous 12 months were 10.6 times more likely to be H C V positive compared to men who reported neither practice [95% CI: 2.0, 56.8}. No significant differences between the two serologic groups were observed with respect to frequency of insertive anal intercourse (p=0.832), the practice of receptive fisting (p=0.728), or receptive oral-anal contact (p=0.999).  5.4.4  Multivariate Analysis  All behavioural variables found to be significantly or marginally associated with HIV seropositivity were considered for inclusion in a multivariate logistic regression model. Using a stepwise approach to model building, the only variable that retained statistical significance in multivariate analysis was insertive fisting (p=0.012). Because of the marginal association of HCV-positivity with the number of sexual partners during the previous 12 months, and because this variable was likely to be confounded with the various sexual practices studied, we carried out further analysis with 81  adjustment for number of sexual partners. After such adjustment, insertive fisting (AOR=2.6,95% CI: 1.0,6.8) remained marginally associated with H C V positivity. 5.5  DISCUSSION We examined the prevalence of HCV-antibody in an observational study of  homosexual men in Vancouver, with the specific objective to estimate the relative risk of H C V infection for specific sexual practices independent of injection drug use. Our study confirms that H C V prevalence is elevated in homosexual males. Prevalence of H C V in homosexual men has generally been found to be significantly higher than that observed in the general population. Our study also confirms that injection drug use is 18  the primary risk factor for H C V acquisition in men who have sex with men.  4-8  It is important to recognize that not all H C V seropositive cases observed in this cohort were related to injection drug use. Of the 39 H C V seropositive men identified in this study, 20 (51%) were related to injection drug use but 19 (49%) were not. The effect of injection drug use on the spread of H C V is related to both the relative risk and the proportion of the population who use injection drugs. In this observational study, the prevalence of injection drug use was 6 percent and the relative risk associated with injection drug use was equal to 30. Using these estimates, we calculated the attributable risk for injection drug use to be 0.635. We estimate that approximately 64 percent of H C V infections occurring among homosexual men in Vancouver could be eliminated by reducing the risk associated with injection drug use. However, about 36 percent of H C V seropositive cases identified in this study were attributable to other risk factors other than injection drug use. The role of sexual contact in the global spread of H C V should not be underestimated. For example, if 35 to 45 percent of all H C V infections were attributable to sexual contact, then this would account for 59.5 to 76.5 million infections worldwide  82  (WHO global estimate of number of HCV-infected is 170 million). Sexual behaviour may be an important mode of spread of H C V if the pool of asymptomatic and infectious carriers is large. Because H C V infection becomes chronic in the majority of individuals and sub-clinical hepatitis may be common, there is reason to believe that in some countries and sub-populations, this carrier pool could be large. Even a low level of sexual transmission may result in a substantial 'attributable' risk.  5  Of considerable interest in our analysis was the association of specific sexual acts with increased risk of H C V positivity among men who did not use injection drugs. Significant elevations in risk were detected in univariate analyses for the practice of insertive fisting, insertive oral-anal contact, and receptive anal intercourse. The highest elevation in risk was observed for men who practiced both insertive fisting and insertive oral-anal contact with their sexual partners. After adjustment for number of partners, elevations in risk persisted for these sexual practices. To our knowledge, this is the first observational study of homosexual men to exclude injection drug users and document specific sexual practices as significant risk factors for H C V acquisition. The sexual practices identified in our study involve rectal trauma and potential exposure to a sexual partner's infected blood or fluids. Both anal intercourse and fisting may cause damage to the mucosal lining of the rectum, and result in tearing and bleeding. Lesions produced by receptive anal intercourse could serve as a passage for the semen of an HCV-infected partner. Injury to the rectum caused by insertive fisting could serve as a source of HCV-infected blood for a partner who also engages in oralanal contact. Chmiel and colleagues, who examined numerous types of sexual behaviour between homosexual men, have reported similar findings with respect to HIV. In their observational study, the factor most strongly associated with prevalent 9  HIV infection was rectal trauma.  83  There are a number of caveats regarding these data which should be discussed. Although individual participants provided estimates of their number of sexual partners during the previous 12 months, the number of HCV-infected sexual partners to which an individual was exposed during this period was not taken into account. The extent of exposure to each partner (duration and type of exposure) and the degree of infectivity of each partner are unknown. In addition, there may be recall problems associated with the reporting of past sexual behavior. Although the effect of these recall problems is difficult to estimate, several studies have shown that interview-administered and selfadministered questionnaires can provide reasonably reliable data concerning sexual behavior in homosexual men when the recall period is relatively short.  10-11  In our study,  information regarding sexual practices was obtained from the questionnaire at the enrollment visit. All participants in this study were unaware of their H C V status at the time of the questionnaire completion. This might have reduced the amount of recall bias in our study. Our findings provide evidence that H C V can be spread by sexual contact among gay and homosexual men in the absence of injection drug use. Sexual practices that cause rectal trauma and involve oral-anal contact can increase the risk of exposure to HCV-infected blood and fluids. These findings not only have important implications for sexually active homosexual men, but heterosexuals as well. Sexual practices that result in vaginal or rectal trauma may play a role in the spread of H C V and other sexually transmitted diseases in the heterosexual population. Large population-based observational studies are needed to confirm these findings and determine which sexual practices place individuals at increased risk of contracting HCV.  84  Table 5.1: Comparison of 19 HCV-positive and 602 HCV-negative men who did not report injection drug use with respect to specific sexual practices  i j HCVpositive n (%)  HCVnegative n (%)  p-value  Odds Ratio  95% CI  Insertive  10 (53)  325 (55)  0.832  0.9  0.4, 2.3  Receptive  12 (63)  260 (44)  0.099  2.2  0.8, 5.6  Insertive  18 (95)  466 (72)  0.029  6.9  1.2, 39.8  Receptive  17 (89)  502 (85)  0.999  1.5  0.3, 6.5  Insertive  10 (53)  173 (26)  0.012  +  3.1  1.3, 7.4  Receptive  3(16)  77 (13)  0.728  1.2  0.4, 4.4  Sexual Practice Anal intercourse*  Oral-anal contact  +  Fisting  t  * Percentage of sexual encounters that included this practice exceeded 25 percent. t Based on Fisher's exact test  85  5.6  1.  REFERENCES  Melbye M , Biggar RJ, Wantzin P, Krogsgaard K, Ebbesen P, Becker NG. Sexual transmission of hepatitis C virus: cohort study (1981-9) among European homosexual men. BMJ 1990; 301:210-212.  2.  Tedder RS, Gilson RJC, Briggs M , Loveday C, Cameron C H , Garrson JA, Kelly GE, Weller 1VD. Hepatitis C virus: evidence for sexual transmission. BMJ 1991; 302:1299-1302.  3.  Gasparini V, Chiaramonte M , Moschen ME, Fabris P, Al tinier G, Majori S, Campello C, Trivello R., Hepatitis C infection in homosexual men: A seroepidemiological study in gay clubs in North-East Italy. Eur J Epidemiol 1991; 7:665-669.  4.  Osella AR, Massa M A , Joekes S, et al. Hepatitis B and C virus sexual transmission among homosexual men. A m J Gastroenterol 1998; 93(l):49-52.  5.  Osmond D H , Charlebois, Sheppard HW, Page K, Winkelstein W, Moss A , Reingold A. Comparison of risk factors for hepatitis C and hepatitis B in homosexual men. J Infect Dis 1993; 167:66-71.  6.  Buchbinder SP, Katz M H , Hessol N A , Liu J, O'Malley PM, Alter MJ. Hepatitis C virus infection in sexually active homosexual men. Journal of Infection 1994; 29:263-269.  7.  Ndimbie OK, Kingsley LA, Nedjar S, Rinaldo CR. Hepatitis C virus infection in a male homosexual cohort: risk factor analysis. Genitourin Med 1996; 72:213216.  86  8.  Bodsworth NJ, Cunningham P, Kaldor J, Donavan B. Hepatitis C virus infection in a large cohort of homosexually active men: independent associations with HIV-1 infection and injecting drug use but not sexual behaviour. Genitourin Med 1996; 72:118-122.  9.  Chmiel JS, Detels R, Kaslow RA, et al. Factors associated with prevalent human immunodeficiency virus (HIV) infection in the Multicenter AIDS Cohort Study. A m J Epidemiol 1987; 126:566-577.  10  Spanier GB. Use of recall data in survey research on human sexual behavior. Soc Biol 1976; 23:244-253.  11.  Coates RA, Soskolne CL, Calzavara, L et al. The reliability of sexual histories in AIDS-related research: Evaluation of an interview-administered questionnaire. Can J Pub Health 1986; 77:343-348.  87  CHAPTER 6 PREDICTORS OF TIME T O FIRST POSITIVE H C V ANTIBODY TEST IN T H E VLAS C O H O R T  6.1  FORWARD In this chapter, results of prospective analyses of 'time to event' data are  presented and discussed. Risk factors for H C V transmission that were identified by means of nested comparisons between seropositive and seronegative men in previous chapters, are re-examined prospectively. HIV serologic status and a number of selfreported sexually transmitted diseases are also assessed as potential predictors of time to H C V seropositivity in the VLAS cohort. These objectives were achieved by utilising longitudinal serologic data from members of the cohort who were enrolled during the first recruitment period, and by focusing on the dependent variable 'time to H C V seropositivity from January 1,1983'. Methods of survival analysis were used to analyze the data. The candidate's contributions to this work consisted of formulation of the study design, data management, performing statistical analysis, summarizing and interpreting the data, and writing the manuscript.  6.2  INTRODUCTION The prevalence of H C V infection appears to be elevated among men who have  sex with men. Previous observational studies of gay men in different geographic locations have reported prevalence rates ranging from approximately 3 to 13 percent.  1  Cross-sectional comparisons of HCV-infected and uninfected men have revealed injection drug use as the predominant mode of transmission in this population. " 4  88  7  Evidence for sexual transmission of H C V among gay has been provided by several cohort studies conducted in the United States. " Osmond et al. found that H C V 4  6  infection was marginally associated with more than 50 sexual partners a year.  4  Buchbinder and colleagues reported receptive anal intercourse, fisting, having a sexual partner with a history of injection drug use, a self-reported history of genital herpes, and being HIV positive as significant univariate risk factors for H C V infection.  5  However, injection drug use was the only significant risk factor in their multivariate analysis. In another study, Ndimbie et al. identified insertive anal intercourse with ejaculation, douche or enema before receptive anal intercourse, self-reported history of rectal gonorrhea, as well as syphilis as statistically significant univariate risk factors. In 6  contrast, other cross-sectional studies of gay men did not find a significant association between sexual practices and H C V seropositivity.  3,8  Previous studies of H C V transmission in homosexual men have all been crosssectional. In these studies, H C V antibody has been determined at a single point in time and comparisons of risk factors between infected and uninfected men are subsequently conducted. Such analyses are useful for generating hypotheses but are potentially misleading because they do not focus on the event of seroconversion. For this reason, risk factors for H C V acquisition identified in previous cross-sectional studies need to be re-examined prospectively. The objective of this investigation was to examine predictors of time to HCV-seropositivity in a cohort of homosexual men in Vancouver. Because the number of incident infections of H C V in the cohort was small, both incident and prevalent infections were included as events of interest in prospective analyses. Variables of interest in this study included: injection drug use during lifetime, HIV antibody status, sexual practices, and self-reported history of other sexually transmitted diseases.  89  6.3  METHODS 6.3.1  Study Description  The Vancouver Lymphadenopathy-AIDS Study (VLAS) has previously been described in detail. " In summary, during the period November 1982 to December 9  10  1984, over 700 homosexual men were recruited and enrolled in a prospective study through six primary care practices located in central Vancouver. At each physician visit, participants completed self-administered questionnaire, which contained items pertaining to demographic characteristics, sexual practices, and history of sexually transmitted diseases including gonorrhea, syphilis, and genital herpes. Participants also provided blood samples for serologic testing. Participants who entered the VLAS between November 1982 and December 1984, and for whom H C V and HIV serologic results were available for at least one follow-up visit were eligible for this analysis. Thus, the serologic status of each participant was based on at least two and as many as seventeen successive tests for H C V and HIV antibody. 6.3.2  HCV and HIV antibody testing  H C V antibody was assayed by enzyme immunoassay (EIA), using a modified algorithm to confirm positivity. This testing algorithm was based on a study carried out in the University of British Columbia Diagnostic Virology and Reference Laboratory and the B.C. Centre for Disease Control Virology Laboratory. Briefly, a micro-particle 11  enzyme immunoassay (Abbott AxSYM H C V Version 3.0 MEIA) was used to screen all sera. Negative sera were tested no further. Positive samples, as well as samples with an indeterminate result, were further tested using a synthetic peptide-based EIA (United Biomedical Inc. H C V EIA 4.0). Sera that tested positive on both EIAs were classified as positive and tested no further. Sera that gave discordant results were further tested by immunoblot (RIBA III; Ortho Diagnostic Systems). For this study, the 90  H C V antibody testing was done on specimens collected between January 1983 and December 1998. HIV antibody testing was conducted using the enzyme-linked immunosorbent assay (ELISA) and equivocal results were further tested using the Western blot technique. 6.3.3  Statistical methods  Methods of survival analysis were used to analyze the data. The event of interest was time to the first positive H C V antibody test. The starting point (or time zero) for this prospective analysis was defined as January 1,1983. Participants who consistently remained HCV-antibody negative were considered as right-censored at the time of their most recent test result. The time to the first seropositive test was calculated as the number of months elapsed from January 1983 until HCV-antibody was detected. Cumulative rates of time to H C V seropositivity were plotted using Kaplan-Meier methods. In these plots, prevalent and incident H C V infections were distinguished from one another. Data regarding risk factors were obtained from the enrolment questionnaire and categorized using previously published levels.  10  Tests of association  between putative predictors and subsequent H C V seropositivity were performed by comparing the cumulative rates of seropositivity associated with the strata of the predictor variable using the log-rank test. Risk ratios and 95 percent confidence intervals were obtained for variables of interest. Cox proportional hazards regression was used to obtain both unadjusted estimates of relative risk, and to assess the independent effect of predictors on time to HCV seropositivity in the cohort. Variables that were identified as either significantly or marginally associated with time to HCV seropositivity in unadjusted analyses were considered for inclusion in multivariate Cox regression models. Forward stepwise regression was used to model the effects of selected predictors on time to seropositivity. All p-values were two-sided.  91  6.4  RESULTS 6.4.1  Demographic characteristics  A total of 662 VLAS participants met the entry criteria for statistical analysis. The median duration of follow-up of these men was 91 months (inter-quartile range: 47 to 168). With regard to demographic characteristics, the majority of men were Caucasian (98%) with a median age of 31 years at entry into the study (range: 17 to 60). Forty-four percent of men attended college or university, and 24 percent reported having annual incomes less than $10,000. 6.4.2  HCV cumulative incidence  Thirty-nine events of H C V seropositivity were documented during the observation period, which yielded a crude cumulative rate of 5.9 percent. The product limit estimate of the probability of becoming H C V seropositive during the 15-year period of observation was 7.9% [95% CI: 5.1%, 10.7%]. The 39 events of H C V seropositivity were comprised of 26 prevalent and 13 incident infections. Prevalent infections were identified on the first laboratory visit occurring between February 1983 and November 1985. Among incident infections, the earliest event of H C V seropositivity occurred in May 1983, and the latest event occurred in February 1997. 6.4.3  HCV cumulative incidence stratified by injection drug use  Figure 6.1 shows the Kaplan-Meier curves for time to first positive HCVantibody test stratified by self-reported injection drug use. The cumulative incidence rate of seropositivity was considerably higher among those who reported injection drug use during their lifetime compared to those who did not (p<0.001). Among 41 men who reported injection drug use, the product limit estimate of the cumulative rate of HCV infection was 52.0 percent (standard error = 8.4%) at 15 years, compared to 4.8 percent (standard error = 1.3%) among 621 men who did not inject drugs. Of the 19 events observed in the group that did not use injection drugs, 8 events were documented as 92  'incident' H C V infections compared to 5 of 20 events that occurred in the group that did not inject drugs. As seen in figure 6.1, six of the eight incident infections occurring in men who did not inject drugs, were documented in 1984 or later. Incident infections continued to occur throughout the observation period among men in this group. 6.4.4  HCV cumulative incidence stratified by HIV antibody status  Figure 6.2 shows the Kaplan-Meier curves for time to first positive HCVantibody test stratified by HIV antibody status. The majority of H C V infections occurred in men who were HTV seropositive, 31 (79%) versus 8 (21%) infections in HIV seronegative men. The cumulative rate of H C V infection was significantly higher among men who were HIV seropositive compared to men who were HIV-negative (p<0.001). The product limit estimate of cumulative H C V infection in the HIV-infected group was 14.3% (standard error = 3.4%) at 15 years, compared to 2.6% (standard error = 0.9%) in the HIV-negative group. 6.4.5  HCV cumulative incidence stratified by fisting and oral-anal contact  Significant disparities between cumulative rates of H C V infection were also observed between strata for a number of sexual practices that were assessed. Figure 6.3 displays the cumulative event curves stratified according to whether the participant reported engaging in both insertive fisting and insertive oral-anal contact with their sexual partner(s) during the previous 12 months. Among men who did engage in this combination of sexual practices, the estimate of the cumulative rate of H C V infection at 15 years was 16.4% compared to 5.2% in the group that did not (p<0.001). 6.4.6  Relationship between HCV infection and other sexual practices  Other sexual practices that were significantly associated with increased risk of H C V seropositivity included: 20 or more male sexual partners in the previous year (p<0.001), anal receptive intercourse with more than 25 percent of sexual encounters (p=0.030), insertive fisting (p=0.003), and insertive oral-anal contact (p=0.024). A 93  marginal elevation in risk was detected for receptive fisting (p=0.074). No association with time to H C V seropositivity was observed with regard to anal insertive intercourse (p=0.705), receptive oral-anal contact (p=0.120), insertive oral-genital contact (p=0.100) or receptive oral-genital contact (p=0.784). For the purpose of brevity, Kaplan-Meier curves have not been illustrated for all sexual practices that were assessed. Table 6.1 shows unadjusted estimates of relative risk and 95 percent intervals for injection drug use, HIV-antibody status, and several sexual practices that were significantly associated with time to H C V seropositivity. Men who reported using injection drugs during their lifetime were about 20 times more likely to have tested seropositive during the observation period [95% CI: 10.9, 39.4]. HIV-seropositive men were almost 4 times more likely to have tested H C V antibody positive during followup. For each of the sexual practices shown in table 6.1 there was an approximate two to three-fold increase in risk of H C V infection. 6.4.7  Relationship between HCV infection and history of infectious diseases  The relationship between lifetime history of other infectious diseases and time to first positive H C V antibody test was also investigated. Table 6.2 displays the relative risk estimates for 10 self-reported infectious diseases. Men who reported a history of non-specific urethritis were 2.5 times more likely to test seropositive during the observation period compared to men without such a history (p=0.013). A marginal increase in risk of H C V infection was noted for a history of gonorrhea (p=0.055). No significant elevation in risk was observed for any of the other infectious diseases that were reported. 6.4.8  Multivariate Analysis  Because injection drug use, sexual practices, and history of infectious diseases were likely confounded with each other, multivariate analysis was conducted using Cox regression to model the simultaneous effect of these variables on time to H C V 94  seropositivity. All variables found to be significantly or marginally associated with H C V infection in univariate Cox regression models were considered for inclusion in the final model (p<0.10). Table 6.3 shows the final multivariate model of this analysis. The risk factors that exerted an independent effect on time to H C V seropositivity were injection drug use (p<0.001), HIV seropositivity (p=0.031), and the sexual practice of insertive fisting in combination with insertive oral-anal contact (p=0.038).  6.5  DISCUSSION Previous cross-sectional studies of gay men have consistently implicated  injection drug use as the primary risk factor for H C V seropositivity. " Other risk factors 4  7  that have been less consistently associated with H C V infection include HIV seropositivity, sexual practices, and history of other sexually transmitted infections. " 4  6  In contrast, some observational studies have not found evidence of an association between sexual behaviour and HCV seropositivity among gay men.  3,8  In our prospective study of homosexual men in Vancouver, we have confirmed that injection drug use is the primary risk factor for H C V infection in this population. Other risk factors found to be independently associated with H C V seropositivity in multivariate analysis included HIV seropositivity, and the sexual practice of insertive fisting in combination with insertive oral-anal contact. To our knowledge, this is the first prospective study of gay men to report a specific sexual practice as a significant predictor of time to H C V seropositivity independent of injection drug use. Our finding that HIV antibody positive men were more likely to be infected with HCV, independent of injection drug use, is consistent with the results reported by Bodsworth et al. While a number of other studies have also reported elevated HCV infection rates among HIV seropositive gay men in univariate analysis, HIV antibody status did not retain significance in multivariate analysis. " 4  95  6  Univariate analysis revealed several sexual practices to be associated with risk of H C V seropositivity in our cohort including twenty or more sexual partners in the prior year, frequency of receptive anal intercourse, insertive fisting, and insertive oral-anal contact. Both insertive fisting and receptive anal intercourse can result in rectal trauma and oral-anal contact can result in exposure to a sexual partner's HCV-infected blood or fluids. It is noteworthy thata history of non-specific urethritis and gonorrhea were either statistically or marginally associated with H C V seropositivity in univariate analysis. These sexually transmitted infections could also promote infection with H C V by disruption of rectal or mucosal barriers. It should be acknowledged there is a considerable amount of overlap between the results presented in this chapter and results presented in chapters 4 and 5. In previous chapters, behavioural data were obtained from the enrolment visit and crosssectional comparisons were carried out between H C V seropositive and seronegative men to identify significant risk factors associated with H C V acquisition. In this chapter, we adopted a different analytic approach and focused on comparisons of the distributions of time to first positive HCV-antibody test between strata of the predictor variable of interest. Behavioural data for these prospective analyses were also obtained from the enrolment questionnaire. Thus the sources of behavioural data were the same for both cross-sectional and prospective analyses. There are limitations that should be acknowledged with regard to our analysis. The number of incident infections of HCV observed in the cohort (n=13) did not provide enough statistical power to adequately assess comparisons of interest and draw generalizable conclusions. In order to increase the statistical power for the prospective analysis, H C V seroprevalent men were also included. Men who were H C V seropositive at their enrolment visit were likely at highest risk for H C V infection. The inclusion of the 26-seroprevalent men in this report likely resulted in a group that was probably less 96  homogenous with regard to their risk factors. Secondly, the results of our study are based on both prevalent and incident H C V infections, and self-reported behaviours. It is possible that responses to questions concerning sexual behaviours, and injection drug use may be influenced by the participant's knowledge of their H C V or HIV antibody status. However, in our study, responses to these questions were obtained two to three years before the participant's HIV antibody testing took place, and over fifteen years prior to H C V antibody testing. HCV antibody testing for this study was carried out during 1999 and 2000. Responses to questions concerning sexual behaviours and injection drug use in our study could not have been influenced by the participant's knowledge of their H C V antibody status. We conclude that injection drug use was the primary risk factor for H C V infection in homosexual men in Vancouver during 1983-98. Our findings also provide confirmatory evidence that H C V can be spread by sexual contact among gay and homosexual men independent of injection drug use. Sexual practices that cause rectal trauma in combination with oral-anal contact increase the risk of exposure to HCVinfected blood and fluids.  97  Figure 6.1: Time to first positive HCV-antibody test among 662 men stratified by injection drug use during lifetime (1983-98)  Injection drug users (n=41, 20 events)  N o I D U - Seroprevalent N o I D U - Seroincident A  I D U - Seroprevalent  A  I D U - Seroincident  Did not use injection drugs (n=621, 19 events)  4ES 0  12  24  36  48  60  72  84  96  108  Months since January 1983  98  120  132  144  156  168  180  Figure 6.2: Time to first positive HCV-antibody test among 662 men stratified by H I V antibody status (1983-98) HIV negative (HCV prevalent) HIV negative (HCV incident) A  HIV positive (HCV prevalent)  A  HIV positive (HCV incident)  HIV antibody positive (n=352, 31 events)  Log-rank p-value < 0.001  HIV antibody negative (n=310, 8 events)  T  12  24  36  48  60  72  84  96  108  Months since January 1983  99  120  132  144  156  T  168  180  Figure 6.3: Time to first positive HCV-antibody test among 662 men stratified by composite varable: insertive fisting and insertive oral-anal contact (1983-98) N o (HCV prevalent)  •  No (HCV incident) _  Yes (HCV prevalent)  A  Yes (HCV incident)  Log-rank p-value < 0.001  No (n=484, 20 events)  —  —  12  —  24  —  36  —  48  —  —  60  72  84  96  i 108  Months since January 1983  100  i 120  i 132  i 144  i 156  i 168 180  Table 6.1: Predictors of time to first positive H C V antibody test in the V L A S cohort  Risk Ratio 20.5  195% C1J 10.9, 39.4  HIV antibody positive  3.8  1.7, 8.3  20 or more male sexual partners  3.1  1.5,6.3  Insertive oral-anal contact (yes vs. no)  3.1  1.1,8.7  Insertive fisting and insertive oral-  2.9  1.5,5.4  Insertive fisting (yes vs. no)  2.6  1.4, 4.8  Receptive anal intercourse during  2.0  1.0, 3.8  Variable Injection drug use during lifetime  during the previous 12 months  anal contact (yes vs. no)  more than 25% of sexual encounters  101  Table 6.2: Estimates of relative risk of H C V infection for self-reported history of infectious diseases among 662 men in the V L A S cohort  Risk Ratio 2.3  [95% CI] 1.2,4.5  Gonorrhea  2.2  1.0, 5.0  Scabies  1.7  0.9,3.2  Hepatitis A virus  1.6  0.6,4.0  Hepatitis B virus  1.5  0.7, 2.9  Syphilis  1.4  0.8,2.8  Herpes zoster  1.2  0.6, 2.4  Genital warts  0.7  0.4,1.4  Pubic lice  0.7  0.3,1.5  Epstein-Barr virus  0.5  0.2,1.7  Infectious disease Non-specific urethritis  102  Table 6.3: Multivariate predictors of time to first positive H C V antibody test in the VLAS cohort  Parameter Estimate 2.845  Standard Error . 0.324  Chi-square p-value <0.001  Adjusted Risk Ratio 17.2  HIV antibody positive  0.883  0.410  0.031  2.4  1.1,5.4  Insertive fisting and insertive  0.680  0.328  0.038  2.0  1.0, 3.8  Variable Injection drug use during  95% CI , 9.1, 32.5  lifetime (yes vs. no)  oral-anal contact (yes vs. no)  103  6.6  1.  REFERENCES  Melbye M , Biggar RJ, Wantzin P, Krogsgaard K, Ebbesen P, Becker NG. Sexual transmission of hepatitis C virus: cohort study (1981-9) among European homosexual men. BMJ 1990; 301:210-212.  2.  Tedder RS, Gilson RJC, Briggs M , Loveday C, Cameron C H , Garrson JA, Kelly GE, Weller IVD. Hepatitis C virus: evidence for sexual transmission. BMJ 1991; 302:1299-1302.  3.  Gasparini V, Chiaramonte M , Moschen ME, Fabris P, Altinier G, Majori S, Campello C, Trivello R., Hepatitis C infection in homosexual men: A seroepidemiological study in gay clubs in North-East Italy. Eur J Epidemiol 1991; 7:665-669.  4.  Osmond D H , Charlebois, Sheppard HW, Page K, Winkelstein W, Moss A, Reingold A. Comparison of risk factors for hepatitis C and hepatitis B in homosexual men. J Infect Dis 1993; 167:66-71.  5.  Buchbinder SP, Katz M H , Hessol N A , Liu J, O'Malley PM, Alter MJ. Hepatitis C virus infection in sexually active homosexual men. Journal of Infection 1994; 29:263-269.  6.  Ndimbie OK, Kingsley LA, Nedjar S, Rinaldo CR. Hepatitis C virus infection in a male homosexual cohort: risk factor analysis. Genitourin Med 1996; 72:213216.  7.  Bodsworth NJ, Cunningham P, Kaldor J, Donavan B. Hepatitis C virus infection in a large cohort of homosexually active men: independent associations with HIV-1 infection and injecting drug use but not sexual behaviour. Genitourin Med 1996; 72:118-122.  104  8.  Osella AR, Massa M A , Joekes S, et al. Hepatitis B and C virus sexual transmission among homosexual men. A m J Gastroenterol 1998; 93(l):49-52.  9.  Boyko WJ, Schechter MT, Jeffries E, Douglas B, Maynard M , O'Shaughnessy MV. The Vancouver Lymphadenopathy-AIDS Study: 3. Relation of HTLV-III sero-positivity, immune status and lymphadenopathy. CMAJ 1985; 133(1): 2832.  10.  Schechter MT, Boyko WJ, Douglas B, et al. The Vancouver LymphadenopathyAIDS Study: 6. HIV seroconversion in a cohort of homosexual men. CMAJ 1986; 135(12): 1355-60.  11.  Cook D, Sherlock C H , Mak A, Littlewood RA, Karakas L, Middleton PJ. Comparison of 5 commercial EIA kits and 2 immunoblot kits for detection of hepatitis C antibody. Conjoint Meeting on Infectious Diseases, Montreal, 1994.  105  CHAPTER 7 HIV DISEASE PROGRESSION A N D MORTALITY A M O N G M E N IN T H E VLAS C O H O R T W H O ARE CO-INFECTED WITH H C V  7.1  FORWARD In this chapter, the results of an assessment of the influence of H C V infection on  the natural history of human immunodeficiency virus (HIV) infection are presented and discussed. The effects of H C V infection on HIV-related outcomes are examined among HIV-infected participants in the VLAS cohort. Outcomes of interest in this investigation included symptoms and signs, physical exam findings, CD4 cell count, progression to AIDS, and survival.  The candidate was responsible for formulating the study design, data management, statistical programming and analyses, summarizing and interpreting the data, and writing the manuscript.  7.2  INTRODUCTION Both hepatitis C virus (HCV) and human immunodeficiency virus (HIV) are  global health problems. Because both viruses share similar routes of transmission, coinfection with H C V and HTV is common. In the United States, it is estimated that 240,000 people are infected with both H C V and HIV. It is estimated that 11,000 1  individuals are co-infected in Canada. Compared to the general population, the 2  prevalence of co-infection is higher among injection drug users, and homosexual men. " 1  106  9  The majority of studies of individuals who are co-infected with H C V and HIV have concluded that HIV disease progression is not strongly influenced by H C V infection, but hepatitis C progresses more rapidly in individuals with co-infection. " 9  14  Individuals with H C V / H I V co-infection experience more rapid progression to cirrhosis, and have more evidence of extensive liver damage.  16  Staples et al. studied time from  HIV diagnosis to AIDS, time from HIV diagnosis to death, and time from AIDS diagnosis to death among H C V / H I V co-infected individuals. ^ They did not find any of 14  these measures of disease progression to significantly differ between H C V / H I V coinfected and HIV-infected groups. The objective of this study was to determine whether co-infection of H C V and HIV influences the natural history of HIV infection in a cohort of HIV-infected homosexual men in Vancouver. HIV-related outcomes were compared between men who were co-infected with H C V and HIV and those who were infected only with HIV. Outcomes of interest for these comparisons included clinical symptoms and signs, physical findings, CD4 cell count, progression to AIDS, and survival.  7.3  METHODS 7.3.1  Study Description  The methods and aims of the Vancouver Lymphadenopathy-AIDS Study (VLAS) have been described previously.  16,17  Briefly, the VLAS has monitored a cohort of 1,000  homosexual men since November 1982. From November 1982 to December 1984, a total of 729 men were recruited and enrolled through six general practices in central Vancouver. Two medical practices were added to the study and 271 additional participants were enrolled during the period October 1986 to December 1987. Followup visits occurred approximately every six months until September 1986, after which subjects completed visits on an annual basis. 107  7.3.2  Sources of data  During each study visit, a physical examination was performed and a functional inquiry was administered. The physical exam included an assessment of the size of the liver and spleen (normal, enlarged, small), and measurement of lymph nodes at each of the following sites: occipital, scalene, upper neck, supraclavicular, axilla, tonsillar, groin, and antecubital. Generalized lymphadenopathy was defined as the presence of lymph nodes greater than 1 centimeter in diameter at two or more extrainguinal sites. The physician administered the functional inquiry. Study participants were asked whether they experienced any of the following symptoms and/or signs during the past year: shortness of breath (unrelated to smoking), cough (unrelated to smoking), night sweats (at least two times per week for one month), fever (for at least 1 week), unintentional weight loss, excessive fatigue, arthralgias (multiple joint pains), diarrhea (for at least 1 week), oral thrush, oesophageal Candida, herpes simplex (cold sores), herpes zoster (shingles), and presence of skin infections. Data regarding symptoms, signs, and physical findings were obtained from follow-up visits that occurred subsequent to the date of the first positive H C V and HIV antibody tests. For each participant, we determined the number of visits in which a specific symptom, sign, or physical finding was reported.  7.3.3  HCV and HIV antibody testing  Both HIV and H C V antibody testing were conducted using serum specimens. HIV antibody tests were performed at the Laboratory Centre for Disease Control in Ottawa. The enzyme-linked immunosorbent assay (ELISA) test was used with equivocal results tested by Western blot. H C V antibody tests were performed at the University of British Columbia Diagnostic Virology and Reference Laboratory at St Paul's Hospital in Vancouver. H C V antibody was assayed by enzyme immunoassay 108  (EIA), using a modified algorithm to confirm seropositivity. Variations of this 18  algorithm have been widely used in diagnostic virology laboratories throughout Canada. Briefly, a micro-particle enzyme immunoassay (Abbott AxSYM H C V Version 3.0 MEIA) was used to screen all sera. Negative sera were tested no further. Positive samples as well as samples with an indeterminate result, was further tested using a synthetic peptide-based EIA (United Biomedical Inc. H C V EIA 4.0). Sera that tested positive on both EIAs were classified as positive and tested no further. Sera that gave discordant results were further tested by immunoblot (RIBA III, Ortho Diagnostic Systems). Sera that were negative by RIBA were classified as negative. For this study, the H C V antibody testing was done on specimens collected between January 1983 and December 1998. 7.3.4  Laboratory methods  Laboratory methods of the VLAS have been described previously. Serial 16  measurements of CD4 cell count were obtained for the majority of HIV-infected participants. For HIV-infected men with two or more CD4 cell counts following enrolment, the annual rate of CD4 cell decline was estimated using the method of least squares. 7.3.5  Eligibility criteria  All HIV seropositive men who were enrolled in the VLAS during the period November 1982 to December 1986, and who completed at least two visits during the period January 1983 to December 1998, were eligible for statistical analysis. This group included men who were HIV seropositive at enrolment, denoted 'HIV seroprevalent', and those who seroconverted during the observation period, denoted 'HIV seroincident'. These men were further classified and denoted as 'HCV seroprevalent' and 'HCV seroincident'.  109  7.3.6  Outcomes of interest  The diagnosis of AIDS in the cohort was defined according to criteria of the Centers for Disease Control (CDC), in Atlanta. Cases of AIDS were ascertained 19  through physician reports and verified by record linkages with a national registry. Deaths of cohort members were also determined by physician reports and through record linkages with British Columbia Vital Statistics. 7.3.7  Statistical methods  Bivariate categorical data (e.g. co-infection and symptoms) were analyzed using Pearson's chi-squared test. Fisher's exact test was used when 25 percent or more the expected values in a contingency table were less than 5. Rates of CD4 cell decline were compared between groups using Student's t-test for independent samples. To analyze data regarding progression to AIDS and mortality, we used statistical methods for survival analysis, which take into account varying lengths of observation for study participants. Differences in AIDS progression and survival rates between H C V / H I V co-infected and HIV-infected men were evaluated by stratified KaplanMeier analysis, and tested using the log-rank test. The starting point for these prospective analyses was defined as the date of the first positive HIV and HCV antibody test for seroprevalent men and the mid-point between the last negative and first positive antibody test result for seroincident men. The events of interest in these analyses were 'time to AIDS diagnosis' and 'time to death'. Participants who were event-free were right-censored as of December 31, 1998. For all statistical comparisons, participants with missing or unknown values for the variable(s) of interest were excluded from the analysis. All reported p-values are two-sided.  110  7.4  RESULTS 7.4.1  Demographic characteristics  A total of 563 men were eligible for statistical analysis. This group included 408 (72%) HIV seroprevalent men and 155 (28%) HIV seroincident men. A total of 46 (8%) of the 563 men who met the entry criteria were H C V antibody positive. This group was comprised of 35 (76%) H C V seroprevalent and 11 (24%) H C V seroincident men. Table 7.1 shows the demographic characteristics and follow-up features of the cohort stratified by H C V antibody status. H C V seropositive and H C V seronegative men were similar with respect to average age at enrolment (p=0.988), ethnicity (p=0.785), education (p-0.788), income (0.616), and average duration of follow-up (p=0.405). 7.4.2  Symptoms and signs  Table 7.2 displays a comparison of the co-infected and control groups with respect to clinical symptoms and signs. No statistically significant differences were observed for any of the individual clinical symptoms included as part of a functional inquiry. However, a marginal association was observed between H C V infection and diarrhea. H C V seropositive men were more likely than H C V seronegative men to report the occurrence of this symptom during the observation period were (53% vs. 41%; p=0.097). 7.4.3  Physical findings  Table 7.3 shows the comparative results of physical findings between men who were co-infected with H C V and those who were not. As seen in this tabular summary, significantly more H C V co-infected men were diagnosed with an enlarged liver (31% vs. 13%; p<0.001) and an enlarged spleen (24% vs. 11%; p=0.011). There was no significant difference between the groups with respect to the occurrence of generalized lymphadenopathy (p=0.527).  Ill  7.4.4  CD4 cell decline  We studied HIV disease progression as measured by the number of CD4 cells measured at baseline, and the rate of CD4 cell decline. The H C V co-infected and comparison groups did not significantly differ with respect to baseline CD4 count (p=0.495) or the rate of CD4 cell decline (mean annual decline: 72 versus 82 cells per year; p=0.870). 7.4.5  Progression to AIDS  A total of 315 cases of AIDS were documented in the cohort during the observation period January 1983 to December 1998. Of these, 27 cases occurred among the H C V seropositive men and 288 cases in H C V negative men. The crude AIDS attack rate was similar in both groups (58% vs. 56%; p=0.696). Figure 7.1 shows the KaplanMeier curves of progression to AIDS among co-infected men and those infected with HIV only. After 14 years of follow-up, the estimated cumulative AIDS progression rate was 64 percent [95% CI: 49%, 78.8%] in the co-infected group compared to 58.5 percent [95% CI: 53.9%, 63.1%] in men who were not co-infected. Cumulative AIDS progression was slightly higher among co-infected men but the difference was not statistically significant (p=0.451). We also carried out a sub-analysis in which the co-infected group was restricted to only H C V seroprevalent men (n=33). This restriction was implemented to ensure that the diagnosis of AIDS did not occur prior to the date of H C V infection among H C V seroincident men. In this sub-analysis, the cumulative AIDS progression rate after 12 years of follow-up was 67.7 percent [95% CI: 51.2%, 84.2%] in the co-infected group compared to 55.5 [95% CI: 50.9%, 60.0%] percent in the comparison group. Although the AIDS progression rate was higher among co-infected men, this difference did not achieve statistical significance (p=0.243).  112  7.4.6 Mortality During the observation period, a total of 265 deaths in the cohort were reported. Twenty-four deaths occurred in the H C V seropositive group and 241 deaths were reported among H C V seronegative men. Crude death rates did not significantly differ between the respective groups (52% vs. 47%; p=0.469). Kaplan-Meier mortality curves are displayed in Figure 7.2. The estimated cumulative mortality rate at 14 years was 55.0 percent [95% CI: 39.7%, 70.2%] in men who were co-infected with HCV. In comparison, among men who were not co-infected, the 14-year cumulative mortality rate was 49.0 percent [95% CI: 44.3%, 53.7%]. Although the 14-year cumulative mortality rate was higher among men who were H C V seropositive, overall the difference was not statistically significant (p=0.314). 7.5  DISCUSSION In this longitudinal study of HIV seropositive men in Vancouver, we found that  co-infection with H C V did not significantly influence the natural history of HIV infection. In particular, HCV-HIV co-infection was not significantly associated with an increase in symptomatic illness, more rapid CD4 cell decline, faster progression of HIV disease to AIDS, or increased mortality. To our knowledge, this is the first longitudinal study in Canada to compare HIV disease progression rates at 14 years among individuals who are co-infected with H C V with men who are infected with HIV only. Another important finding of our prospective study is that mortality did not significantly differ between H C V seropositive and H C V seronegative men. To date, comparable data have not been reported in this population. We found no significant difference between H C V co-infected men and those who were infected with HIV only with regard to the occurrence of numerous non-specific, clinical symptoms and signs. These data provide evidence that H C V infection does not result in any significant increase of symptomatic illness among HIV-infected 113  individuals. However, we did find strong indication of increased liver and spleen inflammation as evidenced from our physical findings among H C V co-infected men. This result is consistent with previous studies that have reported evidence of increased liver damage among H C V / H I V co-infected individuals.  15  To assess HTV disease progression we first compared the rate of CD4 cell decline in the two serologic groups of interest. We did not find any significant difference between baseline CD4 cell counts and rates of CD4 cell decline of H C V seropositive and seronegative men our cohort. These findings are in agreement with the results of a previous cross-sectional study and an Italian cohort study of H C V co-infection in HTV seroconverters.  20,21  In contrast, another cross-sectional study reported that CD4  lymphocyte counts were higher among HCV-positive compared to HCV-negative patients.  22  The rate of progression to AIDS was not significantly different among HCVseropositive and seronegative men in our study. This result corroborates the findings of other prospective studies of H C V co-infection. '  20 21  In addition, we found no significant  increase in the rate of mortality among men co-infected with H C V in our cohort. This result is in agreement with the findings of an earlier observational study, which also reported H C V infection did not shorten survival among individuals with HTV infection.  9  There are several limitations that should be addressed concerning the findings of this investigation. First, the absolute number of HCV-seropositive men in the cohort was relatively small. Consequently, our estimates of cumulative AIDS progression and mortality in this group were subject to considerable variability. This was evidenced by much wider interval estimates of these parameters in the HCV-seropositive group compared to the HCV-seronegative group. Second, we are unable to adequately comment on progression of H C V disease in this cohort because HCV-related outcomes 114  were not well studied. For example, no prospective data regarding liver function were available (e.g. liver enzymes). Third, we are unable to comment on the causes of death among H C V co-infected men at this time. It is quite possible that significantly more HCV-related deaths occurred in the co-infected group. The results of a recently conducted death record linkage with British Columbia Vital Statistics are currently under review and will be the subject of a future report. In conclusion, this prospective study of gay men in Vancouver has demonstrated the natural history of HIV disease is not strongly influenced by H C V infection. These findings provide confirmation of results reported in previous observational studies. The prevalence of HCV-HIV co-infection in homosexual men appears to be elevated but H C V does not appear to affect HIV progression or survival.  115  Table 7.1: Demographic and follow-up characteristics of 563 HIV seropositive men stratified by HCV antibody status (46 HCV seropositive and 517 HCV seronegative men)  Characteristic  HCVseropositive  ' *' • (%)  \  HCVseronegative,  p-value - *  Ethnicitv Caucasian  95  96  0.785*  27  23  0.616  56  53  0.788  Years  Years  32  32  6  6  Months  Months  92  103  15-192  20-193  Annual Income < $10,000 Education Attended College/University  Age (at enrolment) Mean Standard deviation  Duration of follow-up* Median Range  .988+  .405  f  * Based on Fisher's exact test + Based on Student's t-test X Duration between date of first positive HIV-antibody test and participant's death date or December 31, 1998 if participant was alive  116  Table 7.2: Comparison of 46 HCV+/HIV+ and 517 HCV-/HIV+ men with respect to clinical symptoms and signs  Clinical symptom and sign* Fatigue  HCVseropositive  HCVseronegative  p-value  43  52  0.247  (%)  (%)  Multiple joint pains  34  32  0.771  Diarrhea  53  41  0.097  Unintentional weight loss  36  34  0.764  Shortness of breath  22  24  0.835  Night sweats  29  29  0.970  Fever  27  21  0.364  Cough (unrelated to smoking)  36  36  0.996  Herpes simplex  67  62  0.494  Oral thrush  47  36  0.170  Herpes zoster  24  27  0.702  Oesophageal Candida  16  23  0.287  Skin infections  69  73  0.575  *Reported at one or more follow-up visits subsequent to the date of the first positive H C V / H I V antibody tests  117  Table 7.3: Comparison of 46 HCV+/HIV+ and 517 HCV-/HIV+ men with respect to physical findings  HCVseropositive  IICVseronegative  Enlarged liver  31  13  <0.001  Enlarged spleen  24  11  0.011  Generalized lymphadenopathy  62  67  0.527  Physical finding*  (%)  .  -  (%)  p-value  * Reported at one or more follow-up visits subsequent to the date of the first positive H C V / H I V antibody tests  118  Figure 7.1: Kaplan-Meier AIDS progression curves for 563 HIV seropositive men stratified by HCV antibody status •  H C V seropositive (n=46, 27 events)  •  H C V seronegative (n=517, 288 events)  Log-rank p-value = 0.451  60  72  84  96  108 120 132  Months since first positive HIV antibody test  119  144  156 168  Figure 7.2: Kaplan-Meier curves of cumulative mortality for 563 HIV seropositive men stratified by HCV antibody status •  HCV seropositive (n=46, 24 deaths)  •  HCV seronegative (n=517, n=241 deaths)  Log-rank p-value = 0.391  0+-  168  12 Months since first positive H I V antibody test  120  7.6 1.  REFERENCES Sulkowski MS, Mast EE, Seeff LB, et al. Hepatitis C virus as an opportunistic infection in persons infected with human immunodeficiency virus. Clin Infect Dis 2000; 30:S77-84.  2.  Remis R. Epidemiology of hepatitis C infection in Canada. Presented at the First Canadian Conference on Hepatitis C, Montreal, Quebec, May 1-4, 2001.  3.  Dieterich DT. Activity of combination therapy with interferon alfa-2b plus ribavirin in chronic hepatitis C patients co-infested with HIV. Semin Liver Dis 1999; 19:87-94.  4.  Mathews C, et al. Prevalence of H C V in an HIV cohort at a university clinic, Digestive Disease Week; Abstract 121, San Diego, 2000.  5.  Sherman KE, et al. Hepatitis C prevalence in HIV-infected patients: a crosssectional analysis of the US adult clinical trials group. 10 International th  Symposium on Viral Hepatitis and Liver Disease, Atlanta, 1999. 6.  Buchbinder SP, Katz M H , Hessol N A , Liu J, O'Malley PM, Alter MJ. Hepatitis C virus infection in sexually active homosexual men. Journal of Infection 1994; 29:263-269.  7.  Ndimbie OK, Kingsley LA, Nedjar S, Rinaldo CR. Hepatitis C virus infection in a male homosexual cohort: risk factor analysis. Genitourin Med 1996; 72:213216.  8.  Bodsworth NJ, Cunningham P, Kaldor J, Donavan B. Hepatitis C virus infection in a large cohort of homosexually active men: independent associations with HIV-1 infection and injecting drug use but not sexual behaviour. Genitourin Med 1996; 72:118-122.  121  9.  Wright TL, Hollander H , Pu X, et al. Hepatitis C in HIV-infected patients with and without AIDS: prevalence and relationship to patient survival. Hepatology 1994; 20:1152-1155.  10.  Sanchez-Quijano A , Andreu J, Gavilan F, et al. Influence human immunodeficiency virus type 1 infection on the natural course of chronic parenterally acquired hepatitis C. Euro J Clin Microbiol Infect Disl995; 14949953.  11.  Zylberberg H , Pol S. Reciprocal interactions between human immunodeficiency virus and hepatitis C virus infections. Clin Infect Dis 1996; 23:1117-1125.  12.  Soto B, Sanchez-Quijano A , Rodrigo L, et al. Human immunodeficiency virus modifies the natural history of chronically parenterally acquired hepatitis C with an unusually rapid progression to cirrhosis. J Hepatol 1997; 26:1-5.  13.  Piroth L, Duong M,Quantin C, et al. Does hepatitis C co-infection accelerate clinical and immunological evolution of HIV-infected patients? AIDS 12: 381388.  14.  Staples CT, Rimland D, Dudas D. Hepatitis C in the HIV (human immunodeficiency virus) Atlanta V.A. (Veterans Affairs Medical Center) cohort study (HAVACS): the effect of co-infection on survival, Clin Infect Dis 29:150-154.  15.  Garcia-Samaniego J. Soriano V, Castilla J, et al. Influence of hepatitis C virus genotype and HIV infection on histological severity of chronic hepatitis C. A m J Gastroenterol 1997; 92:1130-1134.  122  16.  Boyko WJ, Schechter MT, Jeffries E, Douglas B, Maynard M , O'Shaughnessy MV. The Vancouver Lymphadenopathy-AIDS Study: 3. Relation of HTLV-III sero-positivity, immune status and lymphadenopathy. CMAJ 1985; 133(1): 2832.  17.  Schechter MT, Boyko WJ, Douglas B, et al. The Vancouver LymphadenopathyAIDS Study: 6. HIV seroconversion in a cohort of homosexual men. CMAJ 1986; 135(12): 1355-60.  18.  Cook D, Sherlock C H , Mak A, Littlewood RA, Karakas L, Middleton PJ. Comparison of 5 commercial EIA kits and 2 immunoblot kits for detection of hepatitis C antibody. Conjoint Meeting on Infectious Diseases, Montreal, 1994.  19.  Centers for Disease Control Revision of the CDC surveillance case definition for acquired immunodeficiency syndrome. MMWR 1987,36 (suppl):l-5.  20.  Llibre JM, Garcia E, Aloy A, et al. Hepatitis C virus and progression of infection due to human immunodeficiency virus. Clin Infect Dis 1993; 16:182.  21.  Dorrucci M,Pezzotti P, Phillips AS. Coinfection of hepatitis C virus with human immunodeficiency virus and progression to AIDS. J Infect Dis 1995; 172:1503-1508.  22.  Quan M C , Krajden M , Grigoriew GA, et al. Hepatitis C virus infection in patients infected with the human immunodeficiency virus. Clin Infect Dis 1993; 17:117-119.  123  CHAPTER 8 GENERAL DISCUSSION, CONCLUSIONS, A N D R E C O M M E N D A T I O N S  8.1  S U M M A R Y OF SUDY FINDINGS In this study, we have documented elevated prevalence and incidence of H C V  infection in a cohort of homosexual men residing in Vancouver. During the period 1983-98, approximately 1 in 17 men in the cohort or about 6 percent were H C V antibody positive. Prevalence observed in this cohort was seven times higher than the estimated H C V prevalence in the general Canadian population, and six times greater than prevalence estimated for Canadian males.  1,2  This result is consistent with prevalence  rates reported by other observational studies of gay men.  3-6  Annual infection rates among men enrolled in this prospective study remained on average, at slightly less than one percent for each year of observation with the exception of 1997, in which it was slightly higher. The pattern of H C V incidence observed in this cohort was stable and remained so throughout the duration of the observation. There was no evidence of a decline in H C V infections despite the closed cohort effect. This pattern was even more prominent when comparing H C V and HIV incidence trends between 1983 and 1998. Rates of HIV infection were significantly higher among homosexual men in Vancouver throughout this time period and tended to decrease. In comparison, H C V rates remained fairly stable. This study confirms the role of injection drug use as the primary risk factor for H C V infection among gay and homosexual men. However, less than ten percent of the study cohort reported using injection drugs. Almost one-half of the HCV-positive men identified in our study did not report a history of injection drug use. After controlling for both injection drug use and HIV-antibody status, we found an association between H C V infection and insertive fisting. 124  Among non-injection drug users, we investigated specific sexual practices as putative risk factors for H C V infection including fisting, oral-anal contact, and anal intercourse. We also estimated the relative risk associated with the sexual practice of insertive fisting in combination with oral-anal contact. This composite variable was of special interest because it involved rectal trauma of one sexual partner, and potential oral exposure of the other partner to their sexual partner's HCV-infected blood or fluids. Significant elevations in risk were detected for each of the individual sexual practices: insertive fisting, insertive oral-anal contact, and receptive anal intercourse. The highest elevation in risk was observed for men who practiced both insertive fisting and insertive oral-anal contact with their sexual partners. H C V positive men in our study were also more likely to report higher numbers of sexual partners during the previous 12 months. After adjustment for number of partners, significant elevations in risk persisted for the each of the sexual practices mentioned above. We did not observe a significant association between HCV-infection and a history of blood transfusion among our participants. We were unable to assess the risk associated with tattooing, body piercing, needle borrowing/sharing, and condom use, because these variables were not included as items in most of our survey questionnaires. We did not find that co-infection with H C V significantly influenced the natural history of HIV infection. H C V co-infection was not significantly associated with an increase in symptomatic illness, more rapid CD4 cell decline, faster progression of HIV disease to AIDS, or increased mortality among HIV-infected men. In addition, we found no significant difference between H C V co-infected men and the comparison group with regard to the occurrence of numerous non-specific, clinical symptoms and signs. However, we did find strong indication of increased liver and spleen  125  inflammation as evidenced from the physical findings among H C V co-infected men in the cohort.  8.2  UNIQUE CONTRIBUTIONS, IMPACT, A N D IMPLICATIONS The series of articles comprising this dissertation has contributed to our  understanding of the epidemiology of H C V infection, and the natural history of HCVHIV co-infection and its related outcomes. Using frozen serum specimens from a cohort study of homosexual men in Vancouver, we were able to carry out H C V antibody testing for approximately 900 participants and estimate the prevalence and incidence of H C V infection among all gay men residing in Vancouver during the period 1983 to 1998. Using self-administered questionnaire data obtained from these men, we were able to identify specific sexual practices that are associated with increased risk of H C V acquisition in this population. Significant risk factors that were identified in cross-sectional comparisons were further examined and confirmed by conducting prospective analyses of 'time to event' data. Long-term follow-up of this cohort enabled us to assess the influence of H C V infection on outcomes associated with human immunodeficiency virus (HIV) infection including progression to AIDS and survival. The results of this epidemiological study have provided important data for assessing the burden of H C V in the population of gay men in Vancouver. We have documented elevated prevalence and incidence of H C V infection in this population during 1983-98. This finding highlights the need for improved health intervention programs that are designed to control H C V infection in homosexual men. To our knowledge, this is the first population-based, prospective study to report cumulative incidence rates of H C V infection in Canada. In addition, this is the first study to compare H C V and HIV incidence rates in gay men.  126  The results of this study have also addressed an urgent need to identify behavioural risk factors associated with H C V infection. Not surprisingly, our findings confirm that injection drug use is the primary risk factor for H C V infection in this population. Of greater interest in this research, was to determine the strength of the association between specific sexual acts and risk of H C V infection among individuals who did not use injection drugs. This study provides epidemiological evidence that H C V can be spread by sexual contact in the absence of injection drug use. After adjustment for number of sexual partners, significant elevations in risk persisted for insertive fisting, insertive oral-anal contact, and receptive anal intercourse. The highest elevation in risk was observed for men who practiced both insertive fisting and insertive oral-anal contact with their sexual partners. Sexual practices such as insertive fisting can cause damage to the mucosal barrier of the rectum, and result in tearing and bleeding. Lesions produced by such trauma can serve as a source or passage for pathological microorganisms like HCV. Moreover, individuals who engage in this sexual practice in conjunction with other sexual practices including oral-anal contact may increase their likelihood of contact with HCV-infected blood or fluids. To our knowledge, this is the first observational study of homosexual men to document specific sexual practices as significant risk factors for H C V acquisition among men who did not use injection drugs. The findings of this research have contributed to our understanding of the influence of H C V infection among men who are also infected with HTV. We found that co-infection with H C V did not significantly influence the natural history of HIV disease. However, we did find evidence of increased liver and spleen inflammation among coinfected individuals. These physical findings are indicative of chronic viral hepatitis. To our knowledge, this is the first longitudinal study of HIV seropositive men in Canada to estimate HTV-related disease progression rates at 14 years among individuals 127  who are co-infected with HCV, and to identify evidence of liver damage in this population. There are a number of implications that arise from these findings. First, the elevated prevalence of H C V infection, and the low but stable incidence observed throughout the period 1983-98, suggests H C V infections continue to occur in this population. In addition to injection drug use, sexual behaviour may be an important mode of spread of H C V among gay men. Prevalence of HCV infection was significantly higher among injection drug users in the VLAS cohort compared to those who did not inject drugs (20/41=48.8% versus 19/621=2.9%; p<0.001). Therefore, H C V prevalence observed among non-injection drug users in the Vancouver cohort was approximately 3 times greater than the estimate for Canadian males.  2  It is important to recognize that not all H C V seropositive cases observed in this cohort were related to injection drug use. Of the 54 H C V seropositive men identified in this study, 32 (59%) were related to injection drug use but 22 (41%) were not. The effect of injection drug use on the spread of H C V is related to both the relative risk and the proportion of the population who use injection drugs. In this study, the prevalence of injection drug use was 6 percent and the risk ratio associated with injection drug use was equal to 21. Using these estimates, we calculated the attributable risk percent for injection drug use. We found 56 percent of H C V infections occurring in the population of homosexual men in Vancouver could be eliminated by reducing the risks associated with injection drug use. However, up to 44 percent of H C V seropositive cases identified in this cohort were attributable to factors other than injection drug use. Our finding that specific sexual practices were significant risk factors for H C V seropositivity among homosexual men has implications concerning sexual transmission of H C V among heterosexuals. Sexual practices that result in vaginal or rectal trauma 128  may also play a role in the spread of H C V and other sexually transmitted diseases in the heterosexual population. For example, the practice of 'dry sex' is common in many African countries, and this practice may be associated with an increased prevalence of sexually transmitted diseases among heterosexuals in this geographic location. The 7  practice of dry sex refers to the drying and tightening of the vagina for sexual intercourse. This practice has been adopted in some cultures to increase sexual pleasure. " The methods for this practice include insertion of a substance into the 8  11  vagina, drying with cloths, vaginal douching, and drinking preparations believed to cause drying effects on the vagina. " Some of these methods used for dry sex can cause 8  11  an inflammatory response and epithelial damage. " Genital lesions have been shown 8  12  to increase the risk of HIV transmission. In an observational study by Beksinska and 13  colleagues, dry sex practices were reported by 60 percent of men and 40 percent of women. Dry sex practice was associated with an increased prevalence of self-reported 7  sexually transmitted diseases among men in the study. The relationship between dry sex and risk of H C V transmission has not been investigated. The role of sexual contact in the global spread of H C V should not be underestimated. If 35 to 45 percent of all H C V infections (WHO global prevalence estimate = 170 million people) were attributable to sexual contact, then this would account for 59.5 to 76.5 million infections worldwide. In order to develop appropriate primary prevention messages regarding the potential risk of acquiring H C V by sexual transmission it is important to have knowledge of the sexual practices in the targeted population. Larger observational studies are needed in both developed and developing countries to elucidate the role of sexual transmission of H C V and to identify specific sexual practices that place individuals at increased risk of exposure.  129  Our assessment of men who were co-infected w i t h H C V and H I V i n this cohort indicates that progression of H I V disease is not strongly influenced by H C V infection. However, co-infected men showed significantly more evidence of liver inflammation. Similar results have been reported by other observational studies. " 14  20  H C V co-infection  among HIV-infected individuals is increasingly becoming a common problem. It is estimated that one-third to one-half of HIV-infected individuals i n industrialized countries are also infected w i t h H C V .  2 1  In the next ten to twenty years, a marked  increase i n morbidity and mortality related to H C V disease is likely to occur among HIV-infected individuals.  8.3  STRENGTHS A N D LIMITATIONS 8.3.1  Strengths  This is a timely study. The burden of H C V on the health care system i n Canada is expected to increase substantially i n the next decade. Therefore, it is important to acquire an improved understanding of H C V infection through population-based research. Our study has directly addressed some urgent issues facing H C V epidemiological research i n Canada. This longitudinal investigation targeted the population of gay men, estimated the burden of H C V i n this population, and identified high-risk sexual practices that are associated with increased risk of acquiring H C V . This study also assessed the effect of H C V co-infection on both HIV-related and H C V related outcomes. To our knowledge, no other study i n Canada has addressed all of these research objectives at one time. This study has several methodological strengths. The major strength of this research lies i n the fact the findings are based on the results of a prospective study. The prospective component of this study permitted the calculation of 14-year cumulative incidence rates of H C V infection, and also allowed comparisons of incidence rates 130  between groups of men who reported specific behavioural factors of interest and those who did not. Because this was a longitudinal study with active follow-up of its participants, we were able to assess HIV-related disease progression and mortality among H C V co-infected individuals. Outcomes of interest including diagnosis of AIDS and death were verified through record linkages with both national and provincial registries. Another methodological strength of this study was that both cross-sectional and prospective statistical analyses were used to estimate the risk of H C V infection for specific sexual practices. The results that were obtained using these two analytic approaches were consistent with one another. The results of this study are based on prevalent H C V infections and self-reported behaviours. Responses to questions concerning sexual behaviours, injection drug use, substance use, and history of infectious diseases may be influenced by the participant's knowledge of their H C V or HTV antibody status. However, in our study, responses to these questions were obtained two to three years before the participant's HIV antibody testing took place, and over fifteen years prior to H C V antibody testing. Neither HIV nor H C V antibody testing was available prior to 1984. Thus responses to questions concerning sexual behaviours and substance use could not have been influenced by the participant's knowledge of their antibody status in this study. This research study has provided a cost-effective means of acquiring important knowledge about the epidemiology and natural history of HCV. This was accomplished by having access to epidemiological data from a well-characterized cohort of homosexual men. The VLAS is the largest and longest-running populationbased natural history study of HIV infection in Canada. It continues to serve as a valuable resource for answering important research questions about HTV and HCV infection as well as other infectious diseases.  131  8.3.2  Limitations  There are a number of limitations that should be acknowledged concerning the interpretation of the results of this study. First, the serologic results were based on frozen specimens obtained during the period 1983-98, and it is possible that some degradation of early specimens might have occurred. This degradation may have interfered with ELISA testing for H C V antibody. The presence of HIV infection can also diminish the accuracy of H C V antibody assays. There is an increased risk of receiving false-negative results from H C V screening antibody tests in people with HIV infection.  16,22,23  There exists a possibility the presence of both of these factors could have  led to under-estimation of the true rate of H C V prevalence and incidence in this study.  The observed number of H C V seropositive men in our cohort was fifty percent lower than what was hypothesized in the preliminary planning phase of this study. The small number of incident H C V infections in the cohort failed to provide adequate statistical power to assess comparisons of interest in this group alone. In order to increase statistical power for comparisons of interest, H C V seroprevalent men were also included in our prospective analyses. It is likely that men who were H C V seropositive at enrolment were also at highest risk for H C V infection. The inclusion of these seroprevalent men likely resulted in a sample that was less homogenous with respect to risk factors. Sera were not available for H C V antibody testing for a number of VLAS participants. Approximately ten percent of participants either failed to provide blood samples during the period of follow-up, or their specimens had been depleted as a result of previous laboratory tests. To determine whether persons who did not have serum were different from those who did, a comparative analysis of demographic and behavioural variables was undertaken. Participants who did not have sera for HCV 132  antibody testing tended to be younger, were less likely to have attended college or university, and more likely to report an annual income of less than ten thousand dollars. Participants without sera were also more likely to be an injection drug user. The groups were similar with respect to the self-reported sexual behaviours that were studied. The results of this comparative analysis suggest that men in this cohort who did not undergo H C V antibody testing were at higher risk for infection. Consequently, the results of this longitudinal study likely led to under-estimation of the true prevalence and incidence of H C V infection in the population of gay men in Vancouver. Due to the small number of men in the cohort who were co-infected with HCV and HIV, our estimates of cumulative AIDS progression and mortality in this group were subject to considerable variability. This was evidenced by much wider interval estimates of these parameters in the HCV-seropositive group. Because of these small numbers, we were unable to draw firm conclusions as to whether people with HIV more easily acquire HCV, and conversely, whether people with H C V more easily acquire HTV. The results of this study are based on self-reported behaviours and the potential for misclassification and recall bias exists. Responses to questions concerning sexual behaviours, and injection drug use may be influenced by the participant's knowledge of their H C V or HIV antibody status. However, in our study, responses to these questions were obtained two to three years before the participant's HIV antibody testing took place, and over fifteen years prior to H C V antibody testing. H C V antibody testing for this study was carried out during 1999 and 2000. Responses to questions concerning sexual behaviours and injection drug use in our study could not have been influenced by the participant's knowledge of their H C V antibody status. Although individual participants provided estimates of their number of sexual partners during the previous 12 months, the number of HCV-infected sexual partners to 133  which they were exposed during this period was not taken into account. The extent of exposure to each partner (duration and type of exposure) and the degree of infectivity of each partner are unknown. In addition, there may be recall problems associated with the reporting of past sexual behavior. The effect of these recall problems is difficult to estimate. However, several studies have shown that interview-administered and selfadministered questionnaires can provide reasonably reliable data concerning sexual behavior in homosexual men when the recall period is relatively short.  24,25  In our study,  information regarding sexual practices was obtained from the questionnaire at the enrollment visit. As mentioned previously, all participants in this study were unaware of their H C V status at the time of the questionnaire completion. This might have reduced the amount of recall bias in our study. Another important limitation of this study involves the interpretation of associations between H C V transmission and specific sexual practices. The possibility must be acknowledged that sexual practices identified as significant risk factors in this study were not directly associated with H C V infection but rather confounded with other unmeasured risk factors. For example, we were unable to assess the risk associated with several putative parenteral and non-parenteral risk factors in this study. Tattooing, ear or body piercing, and needle borrowing/sharing, were not included as variables in our enrolment and early, follow-up questionnaires. Snorting of cocaine and sharing of paraphernalia used for this practice (e.g. straws) were not studied. In addition, we did not evaluate occupational and nosocomial exposures. The role of horizontal exposure in the spread of H C V was not addressed in this study. We are unable to comment in any great detail on progression of H C V disease in this cohort because HCV-related outcomes were not well studied. No prospective data regarding liver function were available. We are also unable to comment on the causes of death among H C V co-infected men at this time. It is quite possible that significantly 134  more HCV-related deaths have occurred i n the co-infected group. The results of a recently conducted record linkage study w i t h British Columbia Vital Statistics are presently being reviewed and w i l l be the subject of a future report. The men who participated i n this prospective study may not have been representative of the target population of all gay men i n Vancouver. Homosexual men who choose to visit a primary care physician, whose practice is comprised of predominantly homosexual men, may be different from the "general" male population. M a n y male homosexuals do not openly acknowledge their sexual orientation and hence were not eligible to participate i n this cohort study. Finally, it is important to recognize the findings of this study are based on a cohort of gay men that were recruited almost twenty years ago. Data regarding sexual behaviour and substance use obtained during this time period may no longer be representative of current behavioural practices of younger gay men i n Vancouver. This limitation underscores the need for continued research i n this population to determine the risk of H C V transmission associated with current behavioural practices. 8.4  CONCLUSIONS Four principal conclusions have been drawn from the results of this study. First,  we conclude the prevalence and incidence of H C V infection is elevated among gay men in Vancouver. Prevalence and incidence of H C V are both significantly higher among HIV-infected men compared to HIV-negative men i n this population. These findings highlight the need for improved primary health intervention programs that are designed to reduce and control H C V infection i n this community. Second, we conclude that injection drug use is the primary risk factor for H C V infection among homosexual men i n Vancouver. The role of injection drug use i n the spread of H C V has been well established. While the relative risk associated w i t h injection drug use was extremely high in this study, the prevalence of self-reported injection drug use was low among the 135  participants. Approximately one-half of men who were H C V antibody positive in this study did not use injection drugs. Third, we conclude that H C V can be spread by sexual transmission independent of injection drug use. Sexual practices that cause rectal trauma in combination with oral-anal contact increase the risk of exposure to HCV-infected blood and fluids. Fourth, we conclude H C V infection does not adversely affect HIV disease progression among co-infected individuals but does adversely influence progression of liver disease.  8.5  RECOMMENDATIONS The results and findings of this study give rise to several recommendations  regarding H C V research initiatives in Canada. First, surveillance of H C V needs to be dramatically improved. Current estimates of H C V prevalence must be considered as hypotheses because these estimates are based on fragmented data. There is a pressing need to better evaluate the extent and distribution of H C V infection, and to assess the burden of infection and disease in both the short and long term. In order to determine accurately the true burden of H C V in Canada, it is recommended that population-based prevalence and incidence studies be conducted in regions throughout our entire country. These observational studies need to be targeted at both high and low risk populations. It is also crucial these investigations include prospective monitoring of participants as part of their research protocols. In Canada, it is estimated that between 10 and 40 percent of HCV-infected individuals contracted the disease by a mode of transmission that is unknown or cannot be positively identified. Further research is urgently required to elucidate the role of 1  sexual contact in the spread of H C V in both homosexual and heterosexual populations. To help achieve this objective, it is recommended that future observational studies incorporate behavioural research components as part of their research agendas. 136  The results of this study suggest that specific sexual practices may place individuals at increased risk of contracting HCV. It is recommended that primary prevention programs and intervention programs that are designed to reduce and control H C V infection should address the potential risk of exposure to H C V related to high-risk sexual behaviours including number of sexual partners and sexual practices that cause rectal or vaginal trauma. Previous observational studies of gay men have been limited in terms of their statistical power to identify specific sexual practices as significant risk factors for H C V infection. Small numbers of H C V seropositive men in these studies have made it difficult to identify specific sexual practices as independent risk factors in multivariate analysis. For example, in an observational study reported by Buchbinder et al., numerous sex acts were identified as significant risk factors in univariate analyses but only injection drug use was significant in multivariate analysis. Similar difficulties 4  have been encountered in other observational studies of gay men.  3,5,6  In order to advance our understanding of the epidemiology of HCV it is recommended that an effort be made to establish an international collaborative project with researchers in Canada, the United States, Australia, and the Netherlands. During the 1980s, in response to the HIV/AIDS pandemic, cohort studies of HIV-infected individuals were launched in each of these countries. These longitudinal investigations were conducted in order to gain important knowledge about the epidemiology and the natural history of HIV infection. Like the VLAS, many of these studies have carried out H C V testing later using frozen blood specimens obtained from their study participants many years ago. A pooled data set comprised of demographic and behavioral variables for such individuals would be extremely useful in helping to identify specific sexual practices that are associated with H C V transmission. In addition, a pooled data set containing virologic, immunologic, treatment and endpoint variables from these 137  cohorts could be used to identify patterns of H C V and HIV disease progression, to estimate the duration from H C V seroconversion to critical events and diagnoses, and to investigate behavioral cofactors related to disease progression.  138  8.6  1.  REFERENCES  Remis R. Epidemiology of hepatitis C infection in Canada. Presented at the First Canadian Conference on Hepatitis C, Montreal, Quebec, May 1-4,2001.  2.  Zou S, Tepper M , Giulivi A. Current status of hepatitis C in Canada. Canadian Journal of Public Health 2000; 91(1):S10-S15.  3.  Osmond D H , Charlebois, Sheppard HW, Page K, Winkelstein W, Moss A, Reingold A. Comparison of risk factors for hepatitis C and hepatitis B in homosexual men. J Infect Dis 1993; 167:66-71.  4.  Buchbinder SP, Katz M H , Hessol N A , Liu J, O'Malley PM, Alter MJ. Hepatitis C virus infection in sexually active homosexual men. Journal of Infection 1994; 29:263-269.  5.  Ndimbie OK, Kingsley LA, Nedjar S, Rinaldo CR. Hepatitis C virus infection in a male homosexual cohort: risk factor analysis. Genitourin Med 1996; 72:213-216.  6.  Bodsworth NJ, Cunningham P, Kaldor J, Donavan B. Hepatitis C virus infection in a large cohort of homosexually active men: independent associations with HIV-1 infection and injecting drug use but not sexual behaviour. Genitourin Med 1996; 72:118-122.  7.  Beksinska MF, Rees HV, Kleinschmidt, et al. The practice and prevalence of dry sex among men and women in South Africa: a risk factor for sexually transmitted infections? Sex Transm Inf 1999; 75:178-180.  8.  Brown JE, Ayowa OB, Brown RC. Dry and tight: sexual practices and potential AIDS risk in Zaire. Soc Sci Med 1993; 37:989-994.  9.  Civie D, Wilson D. Dry sex in Zimbabwe and implications for condom use. Soc Sci Med 1996; 42:91-98.  139  10.  Sandala L, Lurie P, Sunkutu MR et al. "Dry sex" and HIV infection among women attending a sexually transmitted diseases clinic in Lusaka, Zambia. AIDS 1995; 9(Suppl l):S61-68.  11.  Runganga A, Pitts M , McMaster J. The use of herbal and other agents to enhance sexual experience. Soc Sci Med 1992; 35:1037-1042.  12.  Brown RC, Brown JE, Ayowa OB. The use and physical effects of intravaginal substances in Zairean women. Sex Transm Dis 1993; 20:96-99.  13.  Kreiss JK, Coombs R, Plummer F, et al. Isolation of human immunodeficiency virus from genital ulcers in Nairobi prostitutes. J Infect Dis 1989; 160:380-384.  14.  Wright TL, Hollander H , Pu X, et al. Hepatitis C in HIV-infected patients with and without AIDS: prevalence and relationship to patient survival. Hepatology 1994; 20:1152-1155.  15.  Sanchez-Quijano A, Andreu J, Gavilan F, et al. Influence human immunodeficiency virus type 1 infection on the natural course of chronic parenterally acquired hepatitis C. Euro J Clin Microbiol Infect Disl995; 14949-953.  16.  Zylberberg H , Pol S. Reciprocal interactions between human immunodeficiency virus and hepatitis C virus infections. Clin Infect Dis 1996; 23:1117-1125.  17.  Soto B, Sanchez-Quijano A, Rodrigo L, et al. Human immunodeficiency virus modifies the natural history of chronically parenterally acquired hepatitis C with an unusually rapid progression to cirrhosis. J Hepatol 1997; 26:1-5.  18.  Piroth L, Duong M,Quantin C, et al. Does hepatitis C co-infection accelerate clinical and immunological evolution of HIV-infected patients? AIDS 12: 381-388.  140  19.  Staples CT, Rimland D, Dudas D. Hepatitis C in the HIV (human immunodeficiency virus) Atlanta V.A. (Veterans Affairs Medical Center) cohort study (HAVACS): the effect of co-infection on survival, Clin Infect Dis 29:150-154.  20.  Garcia-Samaniego J. Soriano V, Castilla J, et al. Influence of hepatitis C virus genotype and HIV infection on histological severity of chronic hepatitis C. A m J Gastroenterol 1997; 92:1130-1134.  21.  Sulkowski MS, Mast EE, Seeff LB, et al. Hepatitis C virus as an opportunistic infection in persons infected with human immunodeficiency virus. Clin Infect Dis 2000; 30:S77-84.  22.  George S. et al. Antibody negative H C V infection in HIV positive individuals. Abstract 127. 10 International Symposium on Viral Hepatitis and Liver Disease, th  Altlanta 2000. 23.  Galli R, Mo T, Braitstein P, Sherlock C H , Yip B, Comanor L, Zierman R, O'Shaughnessy MV, Harrigan R, Montaner JSG. Prevalence of HCV RNA in antiH C V negative, antiretroviral experienced patients from a Canadian cohort. 14th International AIDS Conference, Barcelona, Spain, July 2002.  24.  Spanier GB. Use of recall data in survey research on human sexual behavior. Soc Biol 1976; 23:244-253.  25.  Coates RA, Soskolne CL, Calzavara, L et al. The reliability of sexual histories in AIDSrelated research: Evaluation of an interview-administered questionnaire. Can J Pub Health 1986; 77:343-348.  141  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0091731/manifest

Comment

Related Items