The introduction of HIV antiretroviral medication (ARVs) in Africa has resulted in substantial reductions in morbidity and mortality. This project is studying the impact of ARVs on community level incidence in the Rakai Community Cohort Study (RCCS) in Uganda, the impact of immunologic monitoring, and potential delays in detecting virologic failure on transmitted and acquired genotypic ARV resistance. We have shown complete elimination of transmission among discordant couples on ARVs and continue to scale up treatment as prevention. Our combination prevention study has proven the impact of effectively implemented programs on HIV incidence in the general Rakai population. Our study was the first to provide evidence of the beneficial impact of combination prevention on rural communities in sub-Saharan Africa. Between 2004 when ART was introduced and 2016, ART coverage increased up to 69% and the proportion of HIV positive individuals with viral suppression increased from 42% in 2009 to 75% in 2016. Male circumcision increased from a baseline of 15% in 1999 to 59% in 2016. Combined with behavioral risk reduction counseling, these interventions resulted in a 42% decline in HIV incidence between 2006 and 2016. We have also monitored the impact of combination prevention among high HIV-risk fishing communities bordering Lake Victoria. Between November 2011 and February 2017, HIV testing coverage increased from 68% to 95%; male circumcision coverage increased from 35% to 59%, ART coverage increased from 16% to 75%, and HIV viral load suppression in all HIV-positive persons increased from 30% to 78%. Baseline HIV incidence measured in 2012 was 3.43 per 100 person-years. By 2016-2017, HIV incidence was 1.58 per 100 person-years, a decline of 48% (adjusted incidence rate ratio (aIRR) 0.52; 95% confidence interval (CI) 0.34 to 0.79). Despite this success, challenges remain at achieving epidemic control and we continue to describe important implementation challenges as we scale up combination prevention. We assessed migration patterns using data collected between August 2011 and January 2015 from the RCCS. 29% (n=6718) of participants who migrated over 2 years were significantly more likely to be young and female. Compared to long-term residents, risk of HIV-infection was significantly elevated in women and men in the first two years following migration (women: adjIRR=1.92, 95%CI: 1.52-2.43; men: adjIRR=1.75, 95%CI: 1.33-2.33). While HIV incidence significantly declined among residents and non-recent in-migrants with scale-up of combination HIV prevention, it did not decline among recent in-migrants. HIV-infected migrants, largely women, are less likely to use ART and differentially move into hotspot fishing communities. We find that migrants moving into hotspots (prevalence40%) had significantly higher HIV prevalence than migrants moving elsewhere, but that out-migrants from hotspots dispersed broadly, contributing minimally to HIV burden in individual destination locations. One concern with increased use of ARVs in sub-Saharan Africa is the extent by which viral resistance will develop over time among the non-clade B HIV-1-infected individuals. We measured the levels of transmitted antiretroviral drug resistance among 51 recently infected RCCS seroconverters between November 2012 and October 2013. We found low rates of transmitted antiretroviral drug resistance with only 2 individuals (4%) having resistance to NNRTIs and no resistance found to NRTIs or PIs. In a study done in collaboration with our colleagues at the University of Cape Town, baseline drug resistance among pregnant women initiating ART was low, observed in 12/167 (7%) of women. However, viremia post-treatment initiation was common in these women, and by measuring antiretroviral drug levels prior to viremia we were able to show that 90% of viremic episodes were due to non-adherence to ART in this population as opposed to pre-treatment drug resistance suggesting that public health efforts to support drug adherence as opposed to screening for pre-treatment drug resistance would be more important in this setting. Viral load monitoring (VLM) is currently being scaled up in Uganda after a decade of relying on immunologic and clinical monitoring. This provides a unique tool to investigate adherence challenges and also improve on the quality of care delivery by focusing resources on those most in need. We have examined the use of early VL monitoring to identify patients most at risk for long term success and failure on first line ART. The 12 month VL measurement was found to be highly predictive of long term outcomes, In adjusted analysis, compared to clients with VL<400 copies/ml at 12 months, the adjusted hazard ratios of VF were 6.54 (95% CI=2.7-15.9) in clients with VLs of 400-1000 copies/ml, 7.82 (95% CI=2.7-22.4) in clients with VLs of 1001-2000 copies/ml and 26.81 (95% CI=15.4-46.6) in clients with VLs greater than 2000 copies/ml (p-value<0.001). VL monitoring could facilitate differentiated care strategies currently promoted by WHO and PEPFAR by identifying individuals who could benefit from intensive monitoring to circumvent adherence challenges. We have documented challenges facing ART programs in Uganda including delayed switch to second line therapy as VL monitoring has been scaled up. Among 3,287 HIV-infected persons who initiated ART between 2004 and 2011, of whom 173 met the criteria for virologic failure 6 or more months after ART initiation. 121 (70%) switched to second line ART. The median timing of switching to second line ART was 7.8 months after virologic failure (IQR=3.3-15.2). Cumulative incidences of switching at 6, 12 and 24 months after virologic failure were 33.2% (95% CI=26.2 41.0), 49.8% (95% CI=42.3 57.9) and 71.9% (95% CI=64.2 79.2) respectively. We observed earlier switching to second line among patients with more advanced disease (lower CD4, higher VL) suggesting that clinicians were still relying on clinical and immunologic parameters. Patients not switched to second line therapy were significantly more likely to die, with an adjusted mortality of 11.1% compared to those switched at 1.6% (p-value<=0.005). Among those switched, the longer the time interval between virologic failure and regimen switch, the more patients experienced CD4 decrease and/or further increase in VL during that interval. Drug sharing has also been identified among some clients on ART in Rakai. In a qualitative study done in our fishing communities, participants reported frequent ART sharing within occupational networks, but no selling. Mobility was the principal driver of ART sharing and was associated with other barriers to treatment access including stigma, fear of negative health provider interactions, and transportation. Early mortality in the first year of ART therapy continues to challenge the health benefits of treatment particularly among patients presenting with late change disease. We hypothesized that as treatment guidelines changed in Uganda to earlier initiation, significant mortality declines would occur. We conducted a retrospective cohort analysis on 6800 HIV-infected adults aged 18+ years who started ART between 2005 and 2016 at the urban-based Mengo Hospital HIV clinic to examine trends in mortality over time. The proportion of patients initiating ART with CD4 count 100 cells/ul decreased from 46% to 27% (p <0.001) and those initiating with WHO stage III or IV disease significantly decreased from 50% to 22% (p<0.001) over this time. Mortality in the first year of ART decreased by 70% over this corresponding time period, from 8.7/100 person-years to 2.5/100 person-years (p<0.001). Changing treatment guidelines and strategies to implement earlier initiation of ART have had a dramatic impact on early mortality in our setting.
Showing the most recent 10 out of 207 publications
