1. Mycoplasma genitalium and risk of HIV acquisition among women in Zimbabwe Sue Napierala 1, Barbara Van Der Pol 2, Cynthia Kwok 3, Helen Weiss 1, Tsungai Chipato 4, Ariane van der Straten 5, Robert Salata 6, Charles Morrison 3 1 London School of Hygiene and Tropical Medicine, London, United Kingdom, 2 Indiana University School of Medicine, Indianapolis, United States, 3 Family Health International, Research Triangle Park, United States, 4 UZ-UCSF Collaborative Research Programme, Harare, Zimbabwe, 5 RTI International, San Francisco, United States, 6 Case Western University, Cleveland, United States

2. Mycoplasma genitalium (MG) First isolated in 1980 First isolated in 1980 Smallest known prokaryote, 580kb, lacks a cell wall Smallest known prokaryote, 580kb, lacks a cell wall Very difficult to culture Very difficult to culture Development of PCR in early 1990’s allowed for study of disease association Development of PCR in early 1990’s allowed for study of disease association Association with urethritis in men, cervicitis in women Association with urethritis in men, cervicitis in women Prevalence in the range of 1- 18% in healthy control groups, and 11-33% in HIV+ populations Prevalence in the range of 1- 18% in healthy control groups, and 11-33% in HIV+ populations Pollack, JD. Trends Microbiol. 1997; 5:413-419

3. Mycoplasma genitalium Sexually transmitted Sexually transmitted Emerging epidemiological evidence of an association with HIV Emerging epidemiological evidence of an association with HIV Odds ratio of HIV infection associated with M. genitalium infection in 19 studies

4. Parent Study Longitudinal study to examine hormonal contraceptive use and HIV acquisition Longitudinal study to examine hormonal contraceptive use and HIV acquisition 2296 HIV- women recruited from family planning clinics in Zimbabwe between 1999-2002 2296 HIV- women recruited from family planning clinics in Zimbabwe between 1999-2002 Cervical swabs collected every 3 months for 15-24 months Cervical swabs collected every 3 months for 15-24 months Baseline demographics Women aged 18-35 years Women aged 18-35 years 93% living with partner 93% living with partner 75% reported 1 lifetime sex partner 75% reported 1 lifetime sex partner 69% had completed more than 9yrs school 69% had completed more than 9yrs school

5. Natural history study Objective To describe the prevalence and natural history of MG infection among HIV- women in Zimbabwe To describe the prevalence and natural history of MG infection among HIV- women in Zimbabwe To describe factors associated with incident MG infection among HIV- women in Zimbabwe To describe factors associated with incident MG infection among HIV- women in ZimbabweMethods Women were randomly selected from the parent study, retrospective MG testing at baseline and quarterly for 1 year Women were randomly selected from the parent study, retrospective MG testing at baseline and quarterly for 1 year 285 women contributed data for at least 2 visits 285 women contributed data for at least 2 visits

6. Natural history study 20/285 (7.0%) had an MG infection over the course of the year Of these, 2 women had MG infection detected at more than one visit

7. Natural history study Cox regression analysis of risk factors associated with incident MG infection Incidence rate: 7.66% (95% CI=4.52-11.98%) Incidence rate: 7.66% (95% CI=4.52-11.98%)

8. Case-control study Objective To establish whether women who acquired HIV infection during follow up were more likely to be infected by MG prior to HIV seroconversion To establish whether women who acquired HIV infection during follow up were more likely to be infected by MG prior to HIV seroconversionMethods 155 women who HIV seroconvertered were each matched with 2 controls 155 women who HIV seroconvertered were each matched with 2 controls Matching criteria: Matching criteria: - Age - Time of follow-up - Composite STI variable: 1=CT, GC or BV positive; otherwise=0 otherwise=0 MG testing at 3-months prior to, and at HIV detection visit (or equivalent visits for HIV-negative controls) MG testing at 3-months prior to, and at HIV detection visit (or equivalent visits for HIV-negative controls)

9. Case-control study MG infection in cases and controls 0.02.04.06.08 HIV-positive cases HIV-negative controls MG at HIV detection visit MG prior to HIV detection visit MG at (equivalent of) HIV detection visit MG prior to (equivalent of) HIV detection visit 3.2% 1.4% 7.2% 7.3%

10. Case-control study Univariate and multivariable analysis of risk factors for HIV infection present at visit before serological detection of HIV, or equivalent visit for nonseroconverters Univariate and multivariable analysis of risk factors for HIV infection present at visit before serological detection of HIV, or equivalent visit for nonseroconverters

11. Case-control study Univariate and multivariable analysis of risk factors for HIV infection present at HIV detection visit, or equivalent visit for nonseroconverters Univariate and multivariable analysis of risk factors for HIV infection present at HIV detection visit, or equivalent visit for nonseroconverters

12. Summary This is the first study to examine the temporal relationship between MG and HIV acquisition This is the first study to examine the temporal relationship between MG and HIV acquisition Prevalence of MG was low in this general population of HIV- women in Zimbabwe Prevalence of MG was low in this general population of HIV- women in Zimbabwe Borderline association between MG and HIV prior to and at the HIV detection visit was observed, highlighting the need for further study of the MG-HIV relationship Borderline association between MG and HIV prior to and at the HIV detection visit was observed, highlighting the need for further study of the MG-HIV relationship Future analysis of this case-control study will include 60 additional HIV cases and 120 controls from Uganda Future analysis of this case-control study will include 60 additional HIV cases and 120 controls from Uganda Future epidemiological studies should address a possible causal association through longitudinal study designs and randomized control trials of MG prevention and control Future epidemiological studies should address a possible causal association through longitudinal study designs and randomized control trials of MG prevention and control

13. Acknowledgments We would like to thank the following people and institutions: UCSF AIDS Research Institute, Innovative AIDS Pilot Awards Program UCSF AIDS Research Institute, Innovative AIDS Pilot Awards Program Family Health International Family Health International Case Western Reserve University Case Western Reserve University UZ-UCSF Collaborative Research Programme UZ-UCSF Collaborative Research Programme Barbara Van Der Pol Barbara Van Der Pol Yaw Agyei Yaw Agyei Marshall Mujoma Marshall Mujoma Participants of the HC-HIV study, Zimbabwe Participants of the HC-HIV study, Zimbabwe