Expanding test and treat for HIV prevention in MSM in Kisumu, Kenya
The Anza mapema study
PI: Fredrick O. Otieno (NRHS)
Investigators: Robert C. Bailey (UIC), Susan Graham (UW), Supriya Mehta (UIC), Donath Emusu (CDC) and Boaz Otieno Nyunya (CDC)
Duration: September 2013 – September 2017
Study Sites: Anza Mapema Study Clinic, Tom Mboya Estate
Anza Mapema CBD Clinic
Sponsor: US Centres for Disease Control and Prevention
DFiD though EHPSA
Purpose: The goal of this project was to identify and test 700 gay and bisexual men who have sex with men (GBMSM), to assess the success of a comprehensive package of find, test, link and retain in care (FTLR) interventions delivered to HIV-infected GBMSM, and to retain those who test negative in a peer-led behavioral risk reduction program that includes retesting every 3 months.
Significance: This study was to generate evidence to inform feasibility, acceptability, and expected outputs necessary to guide PEPFAR programming for expansion of counseling, testing, and comprehensive care and treatment as effective HIV prevention interventions adapted for GBMSM in sub-Saharan Africa.
Design: This was an observational cohort study. Over 6 months, 700 GBMSM were to be tested and counseled for HIV. Outreach occurred using existing peer networks, snowballing, and Time Location Recruitment (TLR) strategies. Those who met the eligibility criteria were asked to enroll in the study. HIV-positive participants were recruited into care and offered ART, regardless of CD4 count, then followed up for retention in care and treatment adherence. HIV-negative participants were recruited to a risk reduction program with quarterly counseling, linkage to support groups, and repeat HIV counseling and testing every 3 months. Each participant was followed for 12 months, during which structured HIV behavioral assessments, medical examinations, and collection of biological samples for STI/HIV tests were conducted. Participants seroconverting in the study were followed monthly for an additional 12 months post seroconversions. Mid-way and exit qualitative interviews with participants explored barriers to retention in periodic testing and HIV prevention. Participants who did not complete 12 months of follow-up were contacted to explore reasons for no longer participating. Approximately 30 clinicians and peer outreach staff were interviewed to get their views on unmet needs and program improvement.
Study Population: Men who report oral or anal sex with a man in the past 6 months, are ≥18 years of age, and are residents of Kisumu, irrespective of HIV infection status.
Study Size: A total of approximately 700 men were enrolled, plus approximately 30 staff.
Study Aims: Aim 1: To find and test 700 GBMSM for HIV
Aim 2: To link and retain HIV-positive GBMSM in HIV care including ART
Aim 3: To link and retain HIV-negative GBMSM in a non-ART care and risk reduction program including retesting every 3 months post-enrollment
Results: Between August 2015 and September 2016 we screened 1,012 individuals of whom 75% were eligible. Of those eligible, we enrolled 94%, withdrew 44 with 8 refusing participation despite being eligible. Our retention was average 80% [Range 77% – 81%]. Among the 712 men enrolled, 636 were HIV-negative at baseline, 22 were known HIV-positive and out of care, and 55 were newly diagnosed HIV-positive at baseline. The median age was 24 years, and 44% of all men were >25 years of age. Sex work was predominant in this population with 65% of the participants engaging in sex work at different times with different frequencies. Condom use for all sex acts was high with condom use during anal intercourse also being very high.
Cumulative prevalence of urogenital Chlamydia Trachomatis (CT) or Neisseria Gonorrhoea (NG) at baseline was 12.4%, month 6 was 7.1% and month 12 was 9.4%. There was approximately 3.8% rectal NG at BL, but 50% we co infected with urogenital. Co-infection of urogenital NG/CT was 10% at BL, just 3% at 6 months and 7% at 12 months. At baseline, 25% of men who were infected with urogenital NG and/or CT reported symptoms – urethral discharge or dysuria.
Among 75 GBMSM who were HIV-positive at baseline, 13 did not initiate ART during the Anza Mapema Study, while seven received ART from clinics not affiliated with the study. Among the 55 men who initiated ART through the Anza Mapema Study, 44 returned for their month 6 follow-up visit and 51 returned for their month 12 visit. While 24.5% of the men were virologically suppressed at baseline, 73.2% and 62.7% men were virologically suppressed at month 6 and month 12, respectively. Median CD4 count was 485 at baseline, 639 at moth 6 and 601 at month 12.
Of 635 HIV-negative participants enrolled in the study, 13 men seroconverted over 554.1 person-years (py). This gave is an estimated HIV incidence of 2.3 (95% confidence interval [CI] 1.4-4.0) per 100 py. Reviewing baseline sociodemographic characteristics for correlation, no baseline sociodemographic or behavioral predictor evaluated was associated with HIV incidence. But on time dependent variables during follow up, time-updated report of receptive or versatile sex role and of condomless receptive anal intercourse were significant predictors in unadjusted analyses with receptive or versatile sex having a hazards ratio of 3.65, 95% CI 1.14-11.65) and condomless receptive anal intercourse having a hazards ratio of 4.49, 95% CI 1.54-13.09
Conclusion: Anza Mapema recorded very high retention rates exceeding the 70% threshold recommended by CDC for best-evidence risk reduction interventions. This provides an opportunity for best practices on retaining MSM in HIV intervention cohorts targeting them.
As in other GBMSM cohorts, HIV incidence was associated with condomless receptive anal intercourse and receptive or versatile positioning thus these should be included in screening for PrEP to further reduce incidence.
Unemployment and lack of financial stability was associated with high rates of virologic failure thus structural intervention addressing financial empowerment needs to be explored.
This study had a high prevalence and incidence of NG and CT despite offering treatment for the same. Rectal NG/CT are also a considerable contributor to the overall burden of NG/CT in this population.