Advanced HIV disease in East Africa and Nigeria, in The African Cohort Study

BACKGROUND

Earlier antiretroviral therapy (ART) may decrease progression to advanced HIV disease (AHD) with CD4 count of <200 cells per cubic millimeter or clinical sequelae. We assessed factors associated with AHD among people living with HIV before and during the "test and treat" era.

SETTING

The African Cohort Study prospectively enrolls adults with and without HIV from 12 clinics in Uganda, Kenya, Tanzania, and Nigeria.

METHODS

Enrollment evaluations included clinical history, physical examination, and laboratory testing. Generalized estimating equations were used to estimate adjusted odds ratios and 95% confidence intervals for factors associated with CD4 count of <200 cells per cubic millimeter at study visits.

RESULTS

From 2013 to 2021, 3059 people living with HIV with available CD4 at enrollment were included; median age was 38 years [interquartile range: 30-46 years], and 41.3% were men. From 2013 to 2021, the prevalence of CD4 count of <200 cells per cubic millimeter decreased from 10.5% to 3.1%, whereas the percentage on ART increased from 76.6% to 100% ( P <0.001). Factors associated with higher odds of CD4 count of <200 cells per cubic millimeter were male sex (adjusted odds ratio 1.56 [confidence interval: 1.29 to 1.89]), being 30-39 years (1.42 [1.11-1.82]) or older (compared with <30), have World Health Organization stage 2 disease (1.91 [1.48-2.49]) or higher (compared with stage 1), and HIV diagnosis eras 2013-2015 (2.19 [1.42-3.37]) or later (compared with <2006). Compared with ART-naive, unsuppressed participants, being viral load suppressed on ART, regardless of ART duration, was associated with lower odds of CD4 count of <200 cells per cubic millimeter (<6 months on ART: 0.45 [0.34-0.58]).

CONCLUSION

With ART scale-up, AHD has declined. Efforts targeting timely initiation of suppressive ART may further reduce AHD risk.

Cumulative exposure to depressive symptoms and all-cause mortality among adults living with HIV in Kenya, Nigeria, Tanzania, and Uganda

OBJECTIVE

:We estimated the effects of cumulative exposure to depressive symptoms on risk of all-cause mortality among people living with HIV in four African countries.

DESIGN

:Analysis of prospective cohort data.

METHODS

:The African Cohort Study (AFRICOS) is a prospective cohort of people receiving care at twelve clinics in Kenya, Nigeria, Tanzania, and Uganda. Every six months from January 2013 to May 2020, participants underwent laboratory monitoring, structured surveys, and assessment of depressive symptom severity using the Center for Epidemiologic Studies Depression Scale (CES-D). All-cause mortality was the outcome of interest. The predictor of interest was a time-updated measure of the percentage of days lived with depression (PDD). Marginal structural Cox proportional hazards regression models were used, adjusting for potential confounders including time-varying alcohol use, drug use, and viral load.

RESULTS

:Among 2520 enrolled participants, 1479 (59%) were women and the median age was 38 (interquartile range [IQR]: 32-46). At enrollment, 1438 (57%) were virally suppressed (<200 copies/mL) and 457 (18%) had CES-D ≥ 16, indicating possible depression. Across 9093 observed person-years, the median PDD was 0.7% (IQR: 0-5.9%) with 0.8 deaths per 100 person-years. Leading causes of death included cancer (18% of deaths) and accidents (14%). Models suggested that each 25% absolute increase in PDD was associated with a 69% increase in the risk of all-cause mortality (HR: 1.69; 95% CI: 1.18-2.43).

CONCLUSIONS

:Cumulative exposure to depressive symptoms was substantially associated with the risk of mortality in this cohort of PLWH in Africa.

Implications of asymptomatic malaria infections on hematologic parameters in adults living with HIV in malaria-endemic regions with varying transmission intensities

OBJECTIVES

HIV and malaria coinfection impacts disease management and clinical outcomes. This study investigated hematologic abnormalities in malaria-asymptomatic people living with HIV (PLHIV) in regions with differing malaria transmission.

METHODS

Study participants were enrolled in the African Cohort Study: two sites in Kenya, one in Uganda, and one in Nigeria. Data was collected at enrollment and every 6 months. Logistic regression estimated odds ratios for associations between HIV/malaria status and anemia, thrombocytopenia, and leucopenia.

RESULTS

Samples from 1587 participants with one or more visits comprising 1471 (92.7%) from PLHIV and 116 (7.3%) without HIV were analyzed. Parasite point prevalence significantly differed across the study sites (P <0.001). PLHIV had higher odds of anemia, with males at lower odds compared to females; the odds of anemia decreased with age, reaching significance in those ≥50 years old. Participants in Kisumu, Kenya had higher odds of anemia compared to other sites. PLHIV had higher odds of leucopenia, but malaria co-infection was not associated with worsened leucopenia. The odds of thrombocytopenia were decreased in HIV/malaria co-infection compared to the uninfected group.

CONCLUSION

Hematological parameters are important indicators of health and disease. In PLHIV with asymptomatic malaria co-infection enrolled across four geographic sites in three African countries, abnormalities in hematologic parameters differ in different malaria transmission settings and are region-specific.

Mortality Among Children Aged <5 Years Living with HIV Who Are Receiving Antiretroviral Treatment - U.S. President's Emergency Plan for AIDS Relief, 28 Supported Countries and Regions, October 2020-September 2022

Globally, children aged <5 years, including those living with HIV who are not receiving antiretroviral treatment (ART), experience disproportionately high mortality. Global mortality among children living with HIV aged <5 years receiving ART is not well described. This report compares mortality and related clinical measures among infants aged <1 year and children aged 1-4 years living with HIV with those among older persons aged 5-14, 15-49, and ≥50 years living with HIV receiving ART services at all clinical sites supported by the U.S. President's Emergency Plan for AIDS Relief. During October 2020-September 2022, an average of 11,980 infants aged <1 year and 105,510 children aged 1-4 years were receiving ART each quarter; among these infants and children receiving ART, 586 (4.9%) and 2,684 (2.5%), respectively, were reported to have died annually. These proportions of infants and children who died ranged from four to nine times higher in infants aged <1 year, and two to five times higher in children aged 1-4 years, than the proportions of older persons aged ≥5 years receiving ART. Compared with persons aged ≥5 years living with HIV, the proportions of children aged <5 years living with HIV who experienced interruptions in treatment were also higher, and the proportions who had a documented HIV viral load result or a suppressed viral load were lower. Prioritizing and optimizing HIV and general health services for children aged <5 years living with HIV receiving ART, including those recommended in the WHO STOP AIDS Package, might help address these disproportionately poorer outcomes.

Impact of weight gain with dolutegravir on antiretroviral adherence and viral suppression in four African countries

OBJECTIVE

We hypothesized that total body weight (TBW) gain after switching antiretroviral therapy (ART) regimen to tenofovir disoproxil fumarate/lamivudine/dolutegravir (TLD) may negatively impact ART adherence and viral load (VL) and therefore sought to examine these associations.

METHODS

The ongoing African Cohort Study (AFRICOS) enrols people with HIV at 12 facilities in Kenya, Nigeria, Tanzania and Uganda supported by The US President's Emergency Plan for AIDS Relief. Among ART-experienced participants who switched to TLD, we used multivariable multinomial logistic regression to examine associations between pre-/post-TLD changes in percentage TBW (≥5% gain, <5% change, ≥5% loss) and changes in self-reported ART adherence (0, 1-2, ≥3 days missed doses in past 30 days) and VL [(<50 copies/mL (undetectable), 50-999 copies/mL (detectable, but suppressed), ≥1000 copies/mL (unsuppressed)].

RESULTS

Among 1508 participants, median time from starting TLD to follow-up was 9 months (interquartile range: 7-11). Overall, 438 (29.1%) participants experienced a TBW gain ≥5%, which was more common among females than among males (32.2% vs 25.2%, p = 0.005) and participants switching from efavirenz [32.0% vs nevirapine (19.9%) and boosted protease inhibitor (20.0%); p < 0.001]. Compared with a TBW change <5% [950 (63.0%) participants], TBW gain ≥5% was not significantly associated with more days with missed ART doses [adjusted odds ratio (aOR) = 0.77, 95% confidence interval (CI): 0.48-1.23] or VL becoming detectable and/or unsuppressed (aOR = 0.69, 95% CI: 0.41-1.16).

CONCLUSIONS

Although a substantial proportion of participants experienced weight gain after switching to TLD, we did not identify a significant impact on adherence or virological outcomes.

Achieving the third 95 in sub-Saharan Africa: application of machine learning approaches to predict viral failure

OBJECTIVE

Viral failure in people with HIV (PWH) may be influenced by multiple sociobehavioral, clinical, and context-specific factors, and supervised learning approaches may identify novel predictors. We compared the performance of two supervised learning algorithms to predict viral failure in four African countries.

DESIGN

Cohort study.

METHODS

The African Cohort Study is an ongoing, longitudinal cohort enrolling PWH at 12 sites in Uganda, Kenya, Tanzania, and Nigeria. Participants underwent physical examination, medical history-taking, medical record extraction, sociobehavioral interviews, and laboratory testing. In cross-sectional analyses of enrollment data, viral failure was defined as a viral load at least 1000 copies/ml among participants on antiretroviral therapy (ART) for at least 6 months. We compared the performance of lasso-type regularized regression and random forests by calculating area under the curve (AUC) and used each to identify factors associated with viral failure; 94 explanatory variables were considered.

RESULTS

Between January 2013 and December 2020, 2941 PWH were enrolled, 1602 had been on antiretroviral therapy (ART) for at least 6 months, and 1571 participants with complete case data were included. At enrollment, 190 (12.0%) had viral failure. The lasso regression model was slightly superior to the random forest in its ability to identify PWH with viral failure (AUC: 0.82 vs. 0.75). Both models identified CD4 + count, ART regimen, age, self-reported ART adherence and duration on ART as important factors associated with viral failure.

CONCLUSION

These findings corroborate existing literature primarily based on hypothesis-testing statistical approaches and help to generate questions for future investigations that may impact viral failure.

High-risk human papillomavirus genotype distribution among women living with and at risk for HIV in Africa

OBJECTIVE

Cervical cancer is a common preventable cancer among African women living with HIV (WLWH). Molecular diagnostics for high-risk human papillomavirus (HR-HPV) genotypes are standard components of cervical cancer screening in resource-rich countries but not in resource-limited settings. We evaluated HR-HPV genotypes among women with and without HIV in four African countries to inform cervical cancer preventive strategies.

METHODS

The African Cohort Study (AFRICOS) enrolled participants with and without HIV at 12 clinics in Tanzania, Kenya, Uganda, and Nigeria. Cervical cytobrush specimens from women were genotyped for 14 HR-HPV types using the multiplex Seegene Anyplex real-time PCR assay. Robust Poisson regression was used to estimate relative risks (RRs) and 95% confidence intervals (CIs) for factors associated with HR-HPV in WLWH.

RESULTS

From January 2015 to March 2020, 868 WLWH and 134 women living without HIV (WLWoH) were tested for HR-HPV with prevalence of 50.9 and 38.1%, respectively ( P  = 0.007). Among WLWH, 844 (97.4%) were antiretroviral therapy (ART)-experienced and 772 (89.7%) virally suppressed 1000 copies/ml or less. The most frequent HR-HPV types among WLWH were HPV-16 (13.5%), HPV-52 (9.5%), and HPV-35 (9.3%). HR-HPV infection was more common among Tanzanian WLWH (adjusted RR: 1.23, 95% CI 1.05-1.44, P  = 0.012). Also, WLWH with CD4 + T cells of less than 200 cell/μl had 1.51-fold increased risk of having HR-HPV (95% CI 1.23-1.86, P  < 0.001).

CONCLUSION

HR-HPV was common in WLWH in four African countries, particularly among women with low CD4 + cell count. Scale up of HPV vaccines and development of vaccines with broader activity against less common HR-HPV types may improve cervical cancer prevention in Africa.

Perinatal Depressive Symptoms and Viral Non-suppression Among a Prospective Cohort of Pregnant Women Living with HIV in Nigeria, Kenya, Uganda, and Tanzania

Depression is common during pregnancy and is associated with reduced adherence to HIV-related care, though little is known about perinatal trajectories of depression and viral suppression among women living with HIV (WLHV) in sub-Saharan Africa. We sought to assess any association between perinatal depressive symptoms and viral non-suppression among WLWH. Depressive symptomatology and viral load data were collected every 6 months from WLWH enrolled in the African Cohort Study (AFRICOS; January 2013-February 2020). Generalized estimating equations modeled associations between depressive symptoms [Center for Epidemiological Studies Depression (CES-D) ≥ 16] and viral non-suppression. Of 1722 WLWH, 248 (14.4%) had at least one pregnancy (291 total) and for 61 pregnancies (21.0%), women reported depressive symptoms (13.4% pre-conception, 7.6% pregnancy, 5.5% one-year postpartum). Depressive symptomatology was associated with increased odds of viral non-suppression (aOR 2.2; 95% CI 1.2-4.0, p = 0.011). Identification and treatment of depression among women with HIV may improve HIV outcomes for mothers.

Coronavirus Antibody Responses before COVID-19 Pandemic, Africa and Thailand

Prior immune responses to coronaviruses might affect human SARS-CoV-2 response. We screened 2,565 serum and plasma samples collected from 2013 through early 2020, before the COVID-19 pandemic began, from 2,250 persons in 4 countries in Africa (Kenya, Nigeria, Tanzania, and Uganda) and in Thailand, including persons living with HIV-1. We detected IgG responses to SARS-CoV-2 spike (S) subunit 2 protein in 1.8% of participants. Profiling against 23 coronavirus antigens revealed that responses to S, subunit 2, or subunit 1 proteins were significantly more frequent than responses to the receptor-binding domain, S-Trimer, or nucleocapsid proteins (p<0.0001). We observed similar responses in persons with or without HIV-1. Among all coronavirus antigens tested, SARS-CoV-2, SARS-CoV-1, and Middle East respiratory syndrome coronavirus antibody responses were much higher in participants from Africa than in participants from Thailand (p<0.01). We noted less pronounced differences for endemic coronaviruses. Serosurveys could affect vaccine and monoclonal antibody distribution across global populations.

Brief Report: Virologic Impact of the Dolutegravir Transition: Prospective Results From the Multinational African Cohort Study

BACKGROUND

The real-world impact on viral suppression of switching from non-dolutegravir-based therapy to tenofovir/lamivudine/dolutegravir (TLD) is not thoroughly characterized in Africa. We described the virologic consequences of switching regimens in the African Cohort Study (AFRICOS), an observational cohort in Nigeria, Kenya, Uganda, and Tanzania.

METHODS

Among antiretroviral-experienced people living with HIV (PLWH) in AFRICOS, we compared viral load (VL) nonsuppression (VL ≥ 1000 copies/mL) among those who switched with those who never switched to TLD, restricting to participants who had at least 1 visit with a recorded VL after the countrywide rollout of TLD. We calculated Kaplan-Meier curves and conducted Cox proportional hazards modeling to estimate adjusted hazard ratios and 95% confidence intervals for factors potentially associated with nonsuppression.

RESULTS

As of September 1, 2021, there were 3108 PLWH enrolled. Among 1576 participants who switched to TLD, 1486 (94.3%) remained suppressed after transition, 12 (0.8%) remained unsuppressed, and 38 (2.4%) lost suppression, compared with 652 (82.1%), 75 (9.4%), and 46 (5.8%), respectively, of 797 participants who did not switch ( P < 0.001). After adjustment for sex, age, study site, and self-reported antiretroviral therapy adherence, virally suppressed participants who did not switch to TLD had significantly higher rates of losing viral suppression compared with those who switched (adjusted hazard ratio: 4.26; 95% confidence interval: 2.72 to 6.68).

CONCLUSIONS

PLWH transitioning to TLD had higher rates of viral suppression compared with those who remained on other regimens. Even within a highly suppressed population, TLD transition provided significant benefits for achieving or maintaining viral suppression.

Non-communicable diseases by age strata in people living with and without HIV in four African countries

Introduction

Non‐communicable diseases (NCDs) are an important driver of morbidity among ageing people living with HIV (PLWH). We examined the composite role of age and HIV status on NCDs in people living with and without HIV.

Methods

The African Cohort Study (AFRICOS) prospectively enrols participants aged ≥15 years with and without HIV at 12 sites in Kenya, Tanzania, Uganda and Nigeria. From 21 January 2013 to 1 September 2021, we assessed participants for renal insufficiency (estimated glomerular filtration rate <60 ml/minute/1.73 m²), elevated blood pressure (BP) (any systolic BP >139 mmHg or diastolic BP >89 mmHg), obesity (body mass index >30 kg/m²), diabetes mellitus (DM) (fasting glucose ≥126 mg/dl or antidiabetic medication) and dysglycemia (fasting glucose ≥99 mg/dl or non‐fasting ≥199 mg/dl). Multivariable logistic regression with generalized estimating equations was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for factors associated with each NCD. The main exposure of interest was a composite of HIV status and age dichotomized around 50 years. All models were adjusted for study site and sex. The renal insufficiency model was additionally adjusted for elevated BP and dysglycemia.

Results and discussion

Of 3761 participants with age data, 557 (14.8%) were age ≥50, 2188 (58.2%) were females and 3099 (82.4%) were PLWH. At enrolment, the prevalence of elevated BP, dysglycemia, renal insufficiency and obesity were n = 128 (26.9%), n = 75 (15.8%), n = 8 (1.7%) and n = 40 (8.4%), respectively, for PLWH ≥50. Compared to people without HIV age <50, PLWH age ≥50 had increased adjusted odds of having DM (OR: 2.78, 95% CI: 1.49–5.16), dysglycemia (OR: 1.98, 95% CI: 1.51–2.61) and renal insufficiency (OR: 6.20, 95% CI: 2.31–16.66). There were significant differences by study site, specifically, participants from Nigeria had the highest odds of elevated BP, dysglycemia and renal insufficiency as compared to Uganda.

Conclusions

There was a high burden of NCDs in this African cohort with differences by geographic region. In order to promote healthy ageing with HIV, screening and treatment for common NCDs should be incorporated into routine HIV care with attention paid to geographic heterogeneity to better allocate resources.

Dissecting drivers of immune activation in chronic HIV-1 infection

Background

Immune activation is a significant contributor to HIV pathogenesis and disease progression. In virally-suppressed individuals on ART, low-level immune activation has been linked to several non-infectious comorbid diseases. However, studies have not been systematically performed in sub-Saharan Africa and thus the impact of demographics, ART and regional endemic co-infections on immune activation is not known. We therefore comprehensively evaluated in a large multinational African cohort markers for immune activation and its distribution in various settings.

Methods

2747 specimens from 2240 people living with HIV (PLWH) and 477 without HIV from the observational African Cohort Study (AFRICOS) were analyzed for 13 immune parameters. Samples were collected along with medical history, sociodemographic and comorbidity data at 12 HIV clinics across 5 programs in Uganda, Kenya, Tanzania and Nigeria. Data were analyzed with univariate and multivariate methods such as random forests and principal component analysis.

Findings

Immune activation was markedly different between PLWH with detectable viral loads, and individuals without HIV across sites. Among viremic PLWH, we found that all immune parameters were significantly correlated with viral load except for IFN-α. The overall inflammatory profile was distinct between men and women living with HIV, in individuals off ART and with HIV viremia. We observed stronger differences in the immune activation profile with increasing viremia. Using machine learning methods, we found that geographic differences contributed to unique inflammatory profiles. We also found that among PLWH, age and the presence of infectious and/or noninfectious comorbidities showed distinct inflammatory patterns, and biomarkers may be used to predict the presence of some comorbidities.

Interpretation

Our findings show that chronic immune activation in HIV-1 infection is influenced by HIV viral load, sex, age, region and ART use. These predictors, as well as associations among some biomarkers and coinfections, influence biomarkers associated with noncommunicable diseases.

Funding

This work was supported by the President's Emergency Plan for AIDS Relief via a cooperative agreement between the Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., and the U.S. Department of Defense [W81XWH-11-2-0174, W81XWH-18-2-0040]. The investigators have adhered to the policies for protection of human subjects as prescribed in AR 70–25. This article was prepared while Michael A. Eller was employed at Henry M. Jackson Foundation for the Advancement of Military Medicine for the U.S. Military HIV Research Program. The views expressed are those of the authors and should not be construed to represent the positions of the US Army or the Department of Defense. The opinions expressed in this article are the author's own, and do not reflect the view of the National Institutes of Health, the U.S. Department of Health and Human Services, or the U.S. government.

Transitioning women to first-line preferred TLD regimen is lagging in Sub-Saharan Africa

INTRODUCTION

In 2019, the World Health Organization (WHO) recommended tenofovir disoproxil fumarate-lamivudine-dolutegravir (TLD) as the preferred first line regimen for adults and adolescents regardless of childbearing status. Nevertheless, final eligibility is determined by local policies which may vary from WHO recommendations. We examined TLD transition by gender across five PEPFAR-supported HIV care programs in sub-Saharan Africa.

METHODS

The African Cohort Study (AFRICOS) enrolls people living with HIV (PLWH) engaged in care in Uganda, Kenya (South Rift Valley and Kisumu West), Tanzania and Nigeria. PLWH with at least one study visit after the country introduced TLD were included. We generated Kaplan-Meier (KM) curves to compare TLD transition by gender from 1) time countries' introduction of TLD and 2) time of TLD eligibility according to local policies.

RESULTS

Among 2.476 participants enrolled through September 2021 at 4 sites in sub-Saharan Africa and eligible to transition to TLD, fewer women (68%) compared to men (80%, p < 0.001) were taking TLD. Kaplan-Meier analysis showed time to transition varied by site, with women in Tanzania transitioning at the same rate as men. In Nigeria, women initially had a slower transition but caught up to men. After adjusting for local policies, women[1] in Kisumu West transitioned at the same rate as men. In South Rift Valley and Uganda, women were less likely to be transitioned.

CONCLUSIONS

Despite TLD being the WHO's preferred regimen since 2019, transition of women to potentially lifesaving TLD has been slower than men at certain clinical sites even after accounting for local eligibility criteria.

Assessment of tuberculosis disease activity in people infected with Mycobacterium tuberculosis and living with HIV: A longitudinal cohort study

BACKGROUND

Early detection of asymptomatic incipient tuberculosis (TB) could improve clinical outcomes and reduce the spread of Mycobacterium tuberculosis (MTB) infection, particularly in HIV endemic settings. This study assessed TB disease activity over 5 years in people living with HIV co-infected with MTB using a surrogate biomarker.

METHODS

Between Jan 1, 2013 and Aug 31, 2018, 2014 people living with HIV were screened annually for active TB using the Xpert MTB/RIF diagnostic assay in 11 clinics in Kenya, Tanzania, Uganda, and Nigeria. Longitudinal blood mononuclear cell samples from 46 selected patients with active and recurrent tuberculosis, latent infection, or incipient TB were further analysed for MTB-specific T-cell activation (defined by CD38 expression) as a well-defined surrogate marker for TB disease covering a total of 1758 person-months.

FINDINGS

MTB-specific CD4 T-cell activation differentiated active, Xpert MTB/RIF positive TB from latent TB with a sensitivity and specificity of 86% and was reduced upon TB treatment initiation. Activated MTB-specific T cells were present in 63% and 23% of incipient TB cases 6 and 12 months before diagnosis of active disease, respectively. Transient increases of MTB-specific T cell activation were also observed in individuals with latent infection, while persistent activation was a hallmark of recurrent TB after the end of treatment.

INTERPRETATION

In most cases, progression to active TB disease started 6-12 months before diagnosis by clinical symptoms and sputum occurrence of bacilli. Blood biomarkers could facilitate early detection of incipient TB, improve clinical outcomes, and reduce the transmission of MTB.

FUNDING

This work was supported by the President's Emergency Plan for AIDS Relief via a cooperative agreement between the Henry M. Jackson Foundation for the Advancement of Military Medicine, Inc., and the U.S. Department of Defense [W81XWH-11-2-0174, W81XWH-18-2-0040] and by the Bundesministerium für Bildung und Forschung (BmBF) through funding of the Deutsches Zentrum für Infektionsforschung (DZIF, TTU-TB personalized medicine TTU 02_813).

Weight gain during the dolutegravir transition in the African Cohort Study

Introduction

Dolutegravir (DTG) has become a preferred component of first‐line antiretroviral therapy (ART) in many settings but may be associated with excess weight gain. We evaluated changes in weight and body mass index (BMI) after switch to single‐tablet tenofovir/lamivudine/dolutegravir (TLD) by people living with HIV (PLWH) in four African countries.

Methods

The African Cohort Study (AFRICOS) prospectively follows adults with and without HIV in Kenya, Uganda, Tanzania and Nigeria. Demographics, ART regimen, weight, BMI and waist‐to‐hip ratio were collected every 6 months. Multivariable Cox proportional hazards modelling was used to estimate hazard ratios and 95% confidence intervals (CIs) for factors associated with developing a BMI ≥25 kg/m². Linear mixed effects models with random effects were used to examine the average change in BMI, weight and waist‐to‐hip ratio.

Results

From 23 January 2013 to 1 December 2020, 2950 PLWH were enrolled in AFRICOS and 1474 transitioned to TLD. In adjusted models, PLWH on TLD had 1.77 times the hazard of developing a high BMI (95% CI: 1.22–2.55) compared to PLWH on non‐TLD ART. Examining change in weight among all PLWH on ART, participants on TLD gained an average of 0.68 kg (95% CI: 0.32–1.04) more than PLWH on other regimens after adjusting for duration on ART, sex, age, study site and CD4 nadir. Among participants who switched to TLD, the average change in weight prior to TLD switch was 0.35 kg/year (95% CI: 0.25–0.46) and average change in weight was 1.46 kg/year (95% CI: 1.18–1.75) in the year following transition to TLD after adjustment for confounders.

Conclusions

Elevated BMI and weight gain among PLWH on TLD are concerning safety signals. Implications for the development of metabolic comorbidities should be monitored, particularly if annual weight gain persists during continued follow‐up after transitioning to TLD.

Clinical similarities and differences between two large HIV cohorts in the United States and Africa

Background

Washington, DC, and sub-Saharan Africa are both affected by generalized HIV epidemics. However, care for persons living with HIV (PLWH) and clinical outcomes may differ in these geographically and culturally diverse areas. We compared patient and clinical site characteristics among adult persons living with HIV (PLWH) enrolled in two longitudinal HIV cohort studies—the African Cohort Study (AFRICOS) and the DC Cohort.

Methods

The DC Cohort is a clinic-based city-wide longitudinal cohort comprised of PLWH attending 15 HIV clinics in Washington, DC. Patients’ socio-demographic characteristics, clinical evaluations, and laboratory data are retrospectively collected from electronic medical records and limited manual chart abstraction. AFRICOS is a prospective observational cohort of PLWH and uninfected volunteers attending 12 select HIV care and treatment facilities in Nigeria, Kenya, Uganda and Tanzania. AFRICOS study participants are a subset of clinic patients who complete protocol-specific visits every 6 months with history and physical examination, questionnaire administration, and blood/sputum collection for ascertainment of HIV outcomes and comorbidities, and neurocognitive and functional assessments. Among participants aged ≥ 18 years, we generated descriptive statistics for demographic and clinical characteristics at enrollment and follow up and compared them using bivariable analyses.

Results

The study sample included 2,774 AFRICOS and 8,420 DC Cohort participants who enrolled from January 2013 (AFRICOS)/January 2011 (DC Cohort) through March 2018. AFRICOS participants were significantly more likely to be women (58.8% vs 27.1%) and younger (83.3% vs 61.1% aged < 50 years old) and significantly less likely to be MSM (only 0.1% of AFRICOS population reported MSM risk factor) than DC Cohort. Similar rates of current viral suppression (about 75% of both samples), hypertension, hepatitis B coinfection and alcohol use were observed. However, AFRICOS participants had significantly higher rates of CD4<200 and tuberculosis and significantly lower rates of obesity, DM, hepatitis C coinfection and syphilis.

Conclusions

With similar viral suppression outcomes, but many differences between our cohorts noted, the combined sample provides unique opportunities to assess and compare HIV care and treatment outcomes in the U.S. and sub-Saharan Africa. Comparing these two cohorts may inform care and treatment practices and may pave the way for future pathophysiologic analyses.

Clinical factors and outcomes associated with immune non-response among virally suppressed adults with HIV from Africa and the United States

A significant minority of people living with HIV (PLWH) achieve viral suppression (VS) on antiretroviral therapy (ART) but do not regain healthy CD4 counts. Clinical factors affecting this immune non-response (INR) and its effect on incident serious non-AIDS events (SNAEs) have been challenging to understand due to confounders that are difficult to control in many study settings. The U.S. Military HIV Natural History Study (NHS) and African Cohort Study (AFRICOS). PLWH with sustained VS (< 400 copies/ml for at least two years) were evaluated for INR (CD4 < 350 cells/µl at the time of sustained VS). Logistic regression estimated adjusted odds ratios (aORs) and 95% confidence intervals (CIs) for factors associated with INR. Cox proportional hazards regression produced adjusted hazard ratios (aHRs) for factors associated with incident SNAE after sustained VS. INR prevalence was 10.8% and 25.8% in NHS and AFRICOS, respectively. Higher CD4 nadir was associated with decreased odds of INR (aOR = 0.34 [95% CI 0.29, 0.40] and aOR = 0.48 [95% CI 0.40, 0.57] per 100 cells/µl in NHS and AFRICOS, respectively). After adjustment, INR was associated with a 61% increase in relative risk of SNAE [95% CI 1.12, 2.33]. Probability of "SNAE-free" survival at 15 years since sustained VS was approximately 20% lower comparing those with and without INR; nearly equal to the differences observed by 15-year age groups. CD4 monitoring before and after VS is achieved can help identify PLWH at risk for INR. INR may be a useful clinical indicator of future risk for SNAEs.

Routine HIV clinic visit adherence in the African Cohort Study

Background

Retention in clinical care is important for people living with HIV (PLWH). Evidence suggests that missed clinic visits are associated with interruptions in antiretroviral therapy (ART), lower CD4 counts, virologic failure, and overlooked coinfections. We identified factors associated with missed routine clinic visits in the African Cohort Study (AFRICOS).

Methods

In 2013, AFRICOS began enrolling people with and without HIV in Uganda, Kenya, Tanzania, and Nigeria. At enrollment and every 6 months thereafter, sociodemographic questionnaires are administered and clinical outcomes assessed. Missed clinic visits were measured as the self-reported number of clinic visits missed in the past 6 months and dichotomized into none or one or more visits missed. Logistic regression with generalized estimating equations was used to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for associations between risk factors and missed visits.

Results

Between January 2013 and March 2020, 2937 PLWH were enrolled, of whom 2807 (95.6%) had initiated ART and 2771 had complete data available for analyses. Compared to PLWH 50+, missed clinic visits were more common among those 18–29 years (aOR 2.33, 95% CI 1.65–3.29), 30–39 years (aOR 1.59, 95% CI 1.19–2.13), and 40–49 years (aOR 1.42, 95% CI 1.07–1.89). As compared to PLWH on ART for < 2 years, those on ART for 4+ years were less likely to have missed clinic visits (aOR 0.72, 95% CI 0.55–0.95). Missed clinic visits were associated with alcohol use (aOR 1.34, 95% CI 1.05–1.70), a history of incarceration (aOR 1.42, 95% CI 1.07–1.88), depression (aOR 1.47, 95% CI 1.13–1.91), and viral non-suppression (aOR 2.50, 95% CI 2.00–3.12). As compared to PLWH who did not miss any ART in the past month, missed clinic visits were more common among those who missed 1–2 days (aOR 2.09, 95% CI 1.65–2.64) and 3+ days of ART (aOR 7.06, 95% CI 5.43–9.19).

Conclusions

Inconsistent clinic attendance is associated with worsened HIV-related outcomes. Strategies to improve visit adherence are especially needed for young PLWH and those with depression.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12981-021-00425-0.

Predictors of All-Cause Mortality among People with HIV in a Prospective Cohort Study in East Africa and Nigeria

Background

Introduction of antiretroviral therapy (ART) has been associated with a decline in human immunodeficiency virus (HIV)-related mortality, although HIV remains a leading cause of death in sub-Saharan Africa. We describe all-cause mortality and its predictors in people living with HIV (PLWH) in the African Cohort Study (AFRICOS).

Methods

AFRICOS enrolls participants with or without HIV at 12 sites in Kenya, Uganda, Tanzania, and Nigeria. Evaluations every 6 months include sociobehavioral questionnaires, medical history, physical examination, and laboratory tests. Mortality data are collected from medical records and survivor interviews. Multivariable Cox proportional hazards models were used to calculate adjusted hazard ratios (aHRs) and 95% confidence intervals (CIs) for factors associated with mortality.

Results

From 2013 through 2020, 2724 PLWH completed at least 1 follow-up visit or experienced death. Of these 58.4% were females, 25.8% were aged ≥ 50 years, and 98.3% were ART-experienced. We observed 11.42 deaths per 1000 person-years (95% CI: 9.53–13.68) with causes ascertained in 54% of participants. Deaths were caused by malignancy (28.1%), infections (29.7%), and other non-HIV related conditions. Predictors of mortality included CD4 ≤ 350 cells/µL (aHR 2.01 [95% CI: 1.31–3.08]), a log₁₀copies/mL increase of viral load (aHR 1.36 [95% CI: 1.22–1.51]), recent fever (aHR 1.85[95% CI: 1.22–2.81]), body mass index < 18.5 kg/m2 (aHR 2.20 [95% CI: 1.44–3.38]), clinical depression (aHR 2.42 [95% CI: 1.40–4.18]), World Health Organization (WHO) stage III (aHR 2.18 [95% CI: 1.31–3.61]), a g/dL increase in hemoglobin (aHR 0.79 [95% CI: .72–.85]), and every year on ART (aHR 0.67 [95% CI: .56–.81]).

Conclusions

The mortality rate was low in this cohort of mostly virally suppressed PLWH. Patterns of deaths and identified predictors suggest multiple targets for interventions to reduce mortality.

Perceived satisfaction with HIV care and its association with adherence to antiretroviral therapy and viral suppression in the African Cohort Study

Background

Increased availability of HIV care over the past decade has dramatically reduced morbidity and mortality among people living with HIV (PLWH) in sub-Saharan Africa. However, perceived and experienced barriers to care, including dissatisfaction with services, may impact adherence and viral suppression. We examined the associations between satisfaction with HIV care and antiretroviral therapy (ART) adherence and viral load suppression.

Methods

The African Cohort Study (AFRICOS) is a prospective observational study conducted at PEPFAR-supported clinics in four African countries. At enrollment and twice-yearly study visits, participants received a clinical assessment and a socio-behavioral questionnaire was administered. Participants were classified as dissatisfied with care if they reported dissatisfaction with any of the following: waiting time, health care worker skills, health care worker attitudes, quality of clinic building, or overall quality of care received. Robust Poisson regression was used to estimate prevalence ratios and 95% confidence intervals (CIs) for associations between satisfaction with care and ART adherence and between satisfaction with care and viral suppression (viral load < 1000 copies/mL).

Results

As of 1 March 2020, 2928 PLWH were enrolled and 2311 had a year of follow-up visits. At the first annual follow-up visit, 2309 participants responded to questions regarding satisfaction with quality of care, and 2069 (89.6%) reported satisfaction with care. Dissatisfaction with waiting time was reported by 177 (7.6%), building quality by 59 (2.6%), overall quality of care by 18 (0.8%), health care worker attitudes by 16 (0.7%), and health care worker skills by 15 (0.7%). After adjusting for age and site, there was no significant difference in viral suppression between those who were satisfied with care and those who were dissatisfied (aPR: 1.03, 95% CI 0.97–1.09). Satisfaction with HIV care was moderately associated with ART adherence among AFRICOS participants (aPR: 1.09; 95% CI 1.00–1.16).

Conclusions

While patient satisfaction in AFRICOS was high and the association between perceived quality of care and adherence to ART was marginal, we did identify potential target areas for HIV care improvement, including reducing clinic waiting times.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12981-021-00414-3.

Transient Reductions in Human Immunodeficiency Virus (HIV) Clinic Attendance and Food Security During the Coronavirus Disease 2019 (COVID-19) Pandemic for People Living With HIV in 4 African Countries

The coronavirus disease 2019 (COVID-19) pandemic and associated public health responses have disrupted daily living activities with economic and health consequences globally. We observed transient decreases in human immunodeficiency virus (HIV) clinic visit adherence and food security among persons living with HIV early in the pandemic, and an increase in viral suppression later in the pandemic.

Severe acute respiratory syndrome coronavirus-2 antibody prevalence in people with and without HIV in rural Western Kenya, January to March 2020

Among 582 participants in Western Kenya who were retrospectively tested from January through March 2020, 19 (3.3%) had detectable SARS-CoV-2 antibodies. The prevalence of detectable SARS-CoV-2 antibodies was similar between participants with and without HIV (3.1% vs. 4%, P = 0.68). One participant reported a cough in the preceding week but others denied symptoms. These may represent cross-reactivity or asymptomatic infections that predated the first reported COVID-19 cases in Kenya.

Prevalence and risk factors associated with HIV and syphilis co-infection in the African Cohort Study: a cross-sectional study

Background

Each year, 5.6 million new syphilis cases are diagnosed globally. Guidelines for people living with HIV (PLWH) in low-income countries (LIC) recommend STI testing for symptomatic persons and those newly diagnosed with HIV; routine STI testing is less clear. Here we provide updated syphilis prevalence and identify co-infection risk factors in PLWH in the African Cohort Study (AFRICOS) to understand these rates as they relate to syndromic treatment.

Methods

AFRICOS is a study enrolling PLWH and HIV-uninfected individuals in four African countries. Participant study enrollment information was used to determine syphilis prevalence and co-infection risk factors. Inclusion criteria consisted of adults 18 years or older receiving care at a participating clinic as a long-term resident who consented to data and specimen collection. Exclusion criteria consisted of pregnancy and/or imprisonment. Screen-positive syphilis was defined as a reactive rapid plasma regain (RPR) upon study enrollment whereas confirmed syphilis included a reactive RPR followed by reactive treponemal test. Multivariate analyses was performed to determine HIV and syphilis co-infection risk factors.

Results

Between 2013 and March 1, 2020, 2939 PLWH enrolled and 2818 were included for analysis. Screen-positive and confirmed syphilis prevalence were 5.3% (151/2818) and 3.1% (87/2818), respectively. When the analysis was restricted to PLWH with an RPR titer of greater than, or equal to, 1:8, 11/87 (12.6%) participants were included. No PLWH and confirmed syphilis had documented genital ulcers. In the multivariate model, participants with confirmed syphilis co-infection were more likely to have none or some primary education [aOR 3.29 (1.60, 6.74)] and consume alcohol [aOR 1.87 (1.16, 3.03)] compared to those without syphilis. Antiretroviral therapy (ART) with suppressed viral load (VL) was protective in the unadjusted model but not adjusted multivariate model.

Conclusions

Our findings show that syphilis rates in sub-Saharan Africa remain elevated where diagnosis remains challenging, and that both lower education level and alcohol consumption are significantly associated with HIV/syphilis co-infection in AFRICOS. Based on our analysis, current STI guidelines targeting testing for African individuals with either new HIV diagnosis or syndromic symptoms may be inadequate, highlighting the need for increased testing and treatment strategies in resource-limited settings.

Temporal trends in self-reported HIV stigma and association with adherence and viral suppression in the African Cohort Study

HIV stigma is a major barrier to HIV care and treatment among people living with HIV (PLWH). Evidence suggests that expansion in antiretroviral therapy (ART) may reduce stigma. However, there are limited longitudinal studies examining temporal trends in HIV stigma in sub-Saharan Africa in the Undetectable = Untransmittable (U = U) era. We longitudinally assessed temporal trends in self-reported experienced stigma and the association of experienced stigma with ART adherence and viral suppression among PLWH enrolled in the African Cohort Study (AFRICOS). AFRICOS is an ongoing cohort study enrolling PLWH in Uganda, Kenya, Tanzania, and Nigeria. As of 1 March 2020, 2937 PLWH enrolled in AFRICOS and had available data. In 2013, 22% of participants reported stigma at the enrollment visit and by 2018 the prevalence decreased to 1% overall and was below 2% for all countries. However, there was not a statistically significant change in stigma prevalence in our longitudinal models. In adjusted models, experiencing stigma was associated with a 0.67 decreased odds of ART Adherence (95% confidence interval (CI): 0.56-0.80) and a 0.64 decreased odds of viral suppression (95% CI: 0.73-0.99). HIV-associated stigma was associated with poor self-reported ART adherence and unsuppressed viral load.

Persons living with HIV in sero-discordant partnerships experience improved HIV care engagement compared with persons living with HIV in sero-concordant partnerships: a cross-sectional analysis of four African countries

Background

Persons living with HIV (PLWH) who are members of sero-discordant and sero-concordant relationships may experience psychological stressors or motivators that affect HIV care. We assessed the association between sero-discordance status, antiretroviral therapy (ART) uptake, and viral suppression in the African Cohort Study (AFRICOS).

Methods

AFRICOS enrolls PLWH and HIV-uninfected individuals at 12 sites in Uganda, Kenya, Tanzania, and Nigeria. At enrollment, we determined ART use through self-report. Viral suppression was defined as HIV RNA < 1000 copies/mL. We analyzed PLWH who were index participants within two types of sexual dyads: sero-discordant or sero-concordant. Binomial regression models were used to estimate prevalence ratios (PRs) and 95% confidence intervals (95% CIs) for factors associated with ART use and viral suppression at study enrollment.

Results

From January 2013 through March 2018, 223 index participants from sero-discordant dyads and 61 from sero-concordant dyads were enrolled. The majority of the indexes were aged 25–34 years (50.2%), female (53.4%), and married (96.5%). Sero-discordant indexes were more likely to disclose their status to partners compared with sero-concordant indexes (96.4% vs. 82.0%, p < 0.001). After adjustment, sero-discordant index participants were more likely to be on ART (aPR 2.8 [95% CI 1.1–6.8]), but no more likely to be virally suppressed. Results may be driven by unique psycho-social factors and global implementation of treatment as prevention.

Conclusions

PLWH in sero-discordant sexual partnerships demonstrated improved uptake of ART compared with those in sero-concordant partnerships. Interventions are needed to increase care engagement by individuals in sero-concordant relationships to improve HIV outcomes.

Assessing the impact of HIV support groups on antiretroviral therapy adherence and viral suppression in the African cohort study

Background

Support groups for people living with HIV (PLWH) may improve HIV care adherence and outcomes. We assessed the impact of support group attendance on antiretroviral therapy (ART) adherence and viral suppression in four African countries.

Methods

The ongoing African Cohort Study (AFRICOS) enrolls participants at 12 clinics in Kenya, Uganda, Tanzania, and Nigeria. Self-reported attendance of any support group meetings, self-reported ART adherence, and HIV RNA are assessed every 6 months. Logistic regression models with generalized estimating equations were used to estimate adjusted odds ratios (aORs) and 95% confidence intervals (95% CIs) for support group attendance and other factors potentially associated with ART adherence and viral suppression.

Results

From January 2013 to December 1, 2019, 1959 ART-experienced PLWH were enrolled and 320 (16.3%) reported any support group attendance prior to enrollment. Complete ART adherence, with no missed doses in the last 30 days, was reported by 87.8% while 92.4% had viral suppression <1000copies/mL across all available visits. There was no association between support group attendance and ART adherence in unadjusted (OR 1.01, 95% CI 0.99–1.03) or adjusted analyses (aOR 1.00, 95% CI 0.98–1.02). Compared to PLWH who did not report support group attendance, those who did had similar odds of viral suppression in unadjusted (OR 0.99, 95% CI 0.978–1.01) and adjusted analyses (aOR 0.99, 95% CI 0.97–1.01).

Conclusion

Support group attendance was not associated with significantly improved ART adherence or viral suppression, although low support group uptake may have limited our ability to detect a statistically significant impact.

Factors associated with sexually transmitted infections among care-seeking adults in the African Cohort Study

OBJECTIVES

Sexually transmitted infections (STIs) are a major cause of morbidity. Understanding drivers of transmission can inform effective prevention programs. We describe STI prevalence and identify factors associated with STIs in four African countries.

METHODS

The African Cohort Study is an ongoing, prospective cohort in Kenya, Nigeria, Tanzania and Uganda. At enrollment, a physical exam was conducted and STI diagnosis made by a clinician using a syndromic management approach. Multivariable logistic regression was used to estimate adjusted odds ratios (aORs) and 95% confidence intervals (95% CIs) for factors associated with an STI diagnosis.

RESULTS

As of June 2020, 3544 participants were enrolled. STI prevalence was 7.7% and did not differ by HIV status (p = 0.30). Prevalence differed by syndrome (3.5% vaginal discharge, 1.5% genital ulcer, 2.1% lower abdominal pain, 0.2% inguinal bubo). The odds of having an STI were higher at all sites compared to Kisumu West, Kenya, and among those with a primary level education or below compared to those with secondary or higher (aOR: 1.77; 95% CI: 1.32-2.38). The odds of an STI diagnosis was higher among participants 18-29 years (aOR: 2.29; 95% CI: 1.35-3.87), females (aOR: 2.64; 95% CI: 1.94-3.59), and those with depression (aOR: 1.78; 95% CI: 1.32-2.38). Among PLWH, similar factors were independently associated with an STI diagnosis. Viral suppression was protective against STIs (aOR: 2.05; 95% CI: 1.32-3.20).

CONCLUSIONS

Prevalence of STIs varied by site with young people and females most at risk for STIs. Mental health is a potential target area for intervention.

113. Advanced HIV Disease Among Adults in the African Cohort Study (AFRICOS)

Background

In the “test and treat” era, early ART may decrease the prevalence of advanced HIV disease (AHD), defined as having a CD4 cell count &lt; 200 cells/µL or World Health Organization (WHO) clinical stage III or IV disease. We assessed trends in AHD and ART coverage and describe factors associated with AHD among adults living with HIV (LWH) across four countries before and during the “test and treat” era.

Methods

The African Cohort Study (AFRICOS) is a prospective cohort enrolling adults at risk for HIV or LWH from 12 facilities in Uganda, Kenya, Tanzania and Nigeria. Clinical history review and laboratory testing were performed at enrollment and every 6 months. Serum cryptococcal antigen screening (CrAg) was performed in a subset with CD4 &lt; 200 at enrollment. Logistic regression was used to estimate odds ratios for factors associated with CD4 &lt; 200.

Results

From January 2013–December 2019, 2934 adults LWH were enrolled (median age 38 years [interquartile range, 31–46 years], 41.5% men). Of 2903 with CD4 results at enrollment, 567 (19.5%) had CD4 &lt; 200. Despite consistent increases in ART coverage since 2016, across all countries the prevalence of AHD did not decline below levels observed in 2013 until 2019. The prevalence of CD4 &lt; 200 did not significantly decline from 11.9% (range 9.1–25.0%) in 2013 to 10.3% (range 0–16%) in 2019, p=0.7, while ART coverage increased from 74.7% (range 68.3–93.8%) in 2013 to 97.5% (range 86–100%) in 2019, p= &lt; 0.01 (Figure 1). Factors associated with a higher risk of CD4 &lt; 200 at enrollment were being enrolled in Tanzania, male sex, age &gt;29 years, having a primary or some secondary education or above, and WHO stage II disease or higher. Factors associated with a lower risk of CD4 &lt; 200 were &gt;1 year since HIV diagnosis and being on ART for at least 6 months (Table 1). Among those with CD4 &lt; 200 at enrollment, the most commonly reported comorbidities included HIV wasting syndrome (9.3%) and tuberculosis (TB) (2.3%); 19 (3.4%) of 564 adults screened were CrAg positive.

Figure 1: Trends in Percentage of Participants with CD4 &lt;200 and ART coverage at Study Enrollment by Country and Year

Conclusion

Despite the scale-up of ART in the era of “test and treat”, AHD prevalence has only recently trended downward. Continued efforts towards early HIV diagnosis and timely ART initiation are needed to reduce the risk for CD4&lt; 200. Strategies to increase TB screening, prophylaxis, and treatment are essential to reduce morbidity.

Disclosures

All Authors: No reported disclosures

1540. Prevalence and Risk Factors associated with HIV and Syphilis Co-infection in the African Cohort Study

Background

Each year, 6 million new syphilis cases are diagnosed globally. Seroprevalence studies in low-income countries (LIC) are limited but is estimated at 3.5-4.6%. Few studies have researched prevalence of sexually transmitted infections (STIs) in people living with human immunodeficiency virus (HIV; PLWH). Current guidelines for PLWH in LIC recommend STI testing for symptomatic persons and those with a new HIV diagnosis, which may lead to high rates of undiagnosed STIs. Here we provide updated STI prevalence rates and risk factors for syphilis co-infection in PLWH in the African Cohort Study (AFRICOS).

Methods

AFRICOS is an ongoing longitudinal study enrolling PLWH in four African countries where participants undergo routine medical exams, sociobehavioral questionnaires, and laboratory extraction for study purposes every 6 months. Enrollment syphilis data was extracted to determine screen-positive and serologically-confirmed syphilis prevalence rates for this study. Bivariate and multivariate analysis were performed to determine risk factors for HIV and syphilis co-infection and reported as adjusted prevalence ratios (APR) with 95% confidence intervals (CI).

Results

Between January 2013 and March 1, 2020, 2883 PLWH enrolled. Prevalence of screen-positive and confirmed syphilis was 5.2% and 3%, respectively. Among PLWH with confirmed syphilis, 58.6% were women, mean age was 37.8 years old (IQR 31.658, 45.011, p = 0.068), and genital ulcers were documented in 1.61% participants. In the multivariate model, participants with confirmed syphilis co-infection were more likely to have none or some primary education [2.65 (1.34, 5.230)], demonstrate impaired cognition [2.1 (1.25, 3.590], and consume alcohol [1.88 (1.19, 2.970] compared to those without syphilis.

Conclusion

In conclusion, our findings suggest that syphilis rates remain elevated at endemic levels in LIC where diagnosis remains challenging. Based on our analysis, current STI guidelines for PLWH in Africa are likely leading to a large proportion of undiagnosed STIs and potentially contributing to community spread. While this study observed that lower education level, alcoholism, and impaired cognition were associated with syphilis co-infection, further studies are needed to investigate these associations.

Disclosures

All Authors: No reported disclosures

976. A Descriptive Retrospective Data Analysis of Maternal Sociodemographic Factors and Access of Healthcare Resources within the African Cohort Study, an Integrated Multicountry Preventative Mother to Child Transmission Program

Background

Global reduction in new infant HIV infections is largely due to the expansion of prevention of mother-to-child transmission (PMTCT) programs. Identification of gaps in healthcare services is paramount in targeting interventions that identify high-risk populations and healthcare barriers that could lead to increased risk of mother to child transmission (MTCT) of HIV.

Methods

HIV infected women from 5 regions of Africa enrolled in the African Cohort Study (AFRICOS) were followed prospectively with assessments performed every 6 months. Sociodemographic factors, pregnancy outcomes, and access of PMTCT resources were reviewed for retrospectively reported pregnancies and those followed prospectively from study enrollment. Statistical analysis compared the impact of clinical factors on infant mortality and preterm delivery.

Results

The study reported 5591 pregnancies from January 2013 to June 2019 of which 5363 were retrospectively reported prior to study enrollment and 228 occurred after enrollment. Pregnancies followed prospectively had higher rates of linkage to PMTCT services prenatally (92.5% vs 6.8%, P< 0.001), intrapartum (64.5% vs 3.5%, P< 0.001), and post-partum (64.5% vs 2.9%, P< 0.001). This group had higher rates of delivery by a skilled birth attendant (93.4% vs 66.7%, P< 0.001) and antiretroviral therapy (ART) prescribed antepartum (96.1% vs 5.5%, P< 0.001) and post-partum (74.6% vs 3.6%, P< 0.001). Both groups had similar rates of prescriptions for intrapartum ART (98.7% vs 97.9%). The majority of women reported ART adherence (96.5%, P< 0.001) which was associated with a decrease in both preterm delivery and infant mortality (adjusted OR 0.24, 95% CI 0.15-0.39). A significant proportion of women followed prospectively reported their infants received ART with good adherence (51.8% vs 0.3% and 93.4% vs 6.3%, respectively P< 0.001).

Conclusion

Participation in AFRICOS increased linkage to PMTCT programs which resulted in increased likelihood of skilled delivery and appropriate ART use for women and their infants. It highlights that linkage to care continues to be a crucial factor in limiting MTCT of HIV especially in resource-limited settings. Limitations in this study exist due to the low number of prospectively followed pregnancies.

Disclosures

All Authors: No reported disclosures

Impact of age on CD4 recovery and viral suppression over time among adults living with HIV who initiated antiretroviral therapy in the African Cohort Study

INTRODUCTION

With increased use of antiretroviral therapy (ART), HIV mortality rates are declining and people living with HIV (PLWH) are surviving longer. We characterized CD4 recovery and viral suppression among adults aged < 50 and ≥ 50 years living with HIV who initiated ART in the African Cohort Study (AFRICOS).

METHODS

Beginning in January 2013, PLWH at twelve clinics in Kenya, Uganda, Tanzania and Nigeria underwent medical history review, CD4 and viral load testing as part of the ongoing African Cohort Study (AFRICOS). ART-naïve PLWH who initiated ART within 30 days of enrollment and had at least one year of follow-up were included in these analyses. To compare ART response in participants < 50 years and ≥ 50 years old, changes in CD4 count and viral load suppression after ART initiation were examined at different time points using linear and binomial regression with generalized estimating equations. Variables for time since ART initiation and the interaction between age group and time on ART were included in the model to evaluate longitudinal changes in CD4 recovery and viral suppression by age.

RESULTS

Between January 2013 and September 2019, 2918 PLHV were enrolled in the cohort. Of these, 443 were ART naïve and initiated on ART within 30 days of enrollment, with 90% (n = 399) aged < 50 years old at ART initiation. At ART initiation, participants aged 50 and older had a higher median CD4 count compared to participants younger than 50 years of age although it did not reach statistical significance (306 cells/mm3, IQR:130-547 vs. 277cells/mm3, IQR: 132-437). In adjusted models examining CD4 recovery and viral suppression there were no significant differences by age group over time. By the end of follow-up viral suppression was high among both groups of adults (96% of adults ≥ 50 years old and 92% of adults < 50 years old).

CONCLUSION

This study found no difference in long-term CD4 recovery or viral suppression by age at ART initiation. We found that particularly among younger adults participants had lower median CD4 counts at ART initiation, suggesting the importance of identifying and putting this population on treatment earlier in the disease course.

Persistent Low-level Viremia Predicts Subsequent Virologic Failure: Is It Time to Change the Third 90?

BACKGROUND

World Health Organization (WHO) guidelines identify human immunodeficiency virus (HIV) viral load <1000 copies/mL as the goal of antiretroviral therapy (ART). However, the clinical implications of viremia below this threshold are unclear in the African context. We examined factors associated with persistent low-level viremia (pLLV) and quantified the risk of subsequent virologic.

METHODS

The African Cohort Study enrolled HIV-infected adults at clinics in Uganda, Kenya, Tanzania, and Nigeria, with assessments every 6 months. We evaluated participants prescribed ART for at least 6 months without virologic failure for pLLV. We used multinomial logistic regression to evaluate associations between prespecified factors of interest and 3 levels of pLLV (<200, 200-499, and 500-999 copies/mL). We used Anderson-Gill extended Cox proportional hazards to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) for viremia category associations with time to failure.

RESULTS

We included 1511 participants with 4382 person-years of follow-up. PLLV <200 copies/mL was observed at 20% of visits while 2% of visits had pLLV 200-499 and 500-999 copies/mL each, with substantial variation by site. Protease inhibitor-containing ART was associated with increased risk of pLLV. Compared to undetectable viral load, pLLV ≥200 copies/mL doubled the risk of developing virologic failure (pLLV 200-499: HR, 1.81 [95% CI, 1.08-3.02]); pLLV 500-999: HR, 2.36 [95% CI, 1.52-3.67]).

CONCLUSIONS

Participants with pLLV ≥200 copies/mL were at increased risk of subsequent virologic failure. Optimized HIV care in this setting should target viral suppression <200 copies/mL.

Pretreatment and Acquired Antiretroviral Drug Resistance Among Persons Living With HIV in Four African Countries

BACKGROUND

Emerging HIV drug resistance (HIVDR) could jeopardize the success of standardized HIV management protocols in resource-limited settings. We characterized HIVDR among antiretroviral therapy (ART)-naive and experienced participants in the African Cohort Study (AFRICOS).

METHODS

From January 2013 to April 2019, adults with HIV-1 RNA >1000 copies/mL underwent ART history review and HIVDR testing upon enrollment at 12 clinics in Uganda, Kenya, Tanzania, and Nigeria. We calculated resistance scores for specific drugs and tallied major mutations to non-nucleoside reverse transcriptase inhibitors (NNRTIs), nucleoside reverse transcriptase inhibitors (NRTIs), and protease inhibitors (PIs) using Stanford HIVDB 8.8 and SmartGene IDNS software. For ART-naive participants, World Health Organization surveillance drug resistance mutations (SDRMs) were noted.

RESULTS

HIVDR testing was performed on 972 participants with median age 35.7 (interquartile range [IQR] 29.7-42.7) years and median CD4 295 (IQR 148-478) cells/mm3. Among 801 ART-naive participants, the prevalence of SDRMs was 11.0%, NNRTI mutations 8.2%, NRTI mutations 4.7%, and PI mutations 0.4%. Among 171 viremic ART-experienced participants, NNRTI mutation prevalence was 83.6%, NRTI 67.8%, and PI 1.8%. There were 90 ART-experienced participants with resistance to both efavirenz and lamivudine, 33 (36.7%) of whom were still prescribed these drugs. There were 10 with resistance to both tenofovir and lamivudine, 8 (80.0%) of whom were prescribed these drugs.

CONCLUSIONS

Participants on failing ART regimens had a high burden of HIVDR that potentially limited the efficacy of standardized first- and second-line regimens. Management strategies that emphasize adherence counseling while delaying ART switch may promote drug resistance and should be reconsidered.

Noninfectious Comorbidity in the African Cohort Study

BACKGROUND

Noninfectious comorbid diseases (NCDs) contribute to morbidity and mortality in human immunodeficiency virus (HIV)-infected populations in resource-rich countries. With antiretroviral therapy (ART) scale-up in Africa, understanding burden NCD informs public health strategy.

METHODS

At enrollment, participants at 11 HIV clinics in Kenya, Uganda, Tanzania, and Nigeria underwent medical history, physical, laboratory, and neuropsychological assessments to identify elevated blood pressure, hypercholesterolemia, dysglycemia, renal insufficiency, and cognitive impairment. Poisson regression models estimated adjusted relative risks (ARRs) and 95% confidence intervals (CIs) for the number of NCDs associated with factors of interest. Logistic regression was used to evaluate each NCD separately among HIV-infected participants.

RESULTS

Among 2720 participants with complete NCD data, 2159 (79.4%) were HIV-infected. Of those, 1426 (66.0%) were taking ART and 813 (37.7%) had at least 1 NCD. HIV infection was associated with more NCDs, especially with ART (ARR, 1.42; 95% CI, 1.22-1.66). In addition to age, body mass index, and program site, ART usage was associated with more NCDs (ARR, 1.50; 95% CI, 1.27-1.78 for virologically suppressed and ARR, 1.38; 95% CI, 1.13-1.68 for viremic) among HIV-infected participants. In participants taking ART, CD4 nadir below 200 cells/mm3 was associated with more NCDs (ARR, 1.43; 95% CI, 1.06-1.93). ART use was independently associated with hypercholesterolemia and dysglycemia. Program site was significantly associated with all comorbidities except renal insufficiency.

CONCLUSIONS

HIV infection was a risk for NCDs, which were common in HIV-infected participants, geographically variable, and largely consistent with metabolic complications of first-line ART.

Predictors and Barriers to Condom Use in the African Cohort Study

Consistent condom use is an inexpensive and efficacious HIV prevention strategy. Understanding factors associated with condom use and barriers to use can inform strategies to increase condom uptake. The ongoing African Cohort Study prospectively enrolls adults at 12 clinical sites in Uganda, Kenya, Tanzania, and Nigeria. At enrollment, participants are asked about condom use at last sex with a regular partner. Robust Poisson regression models were used to evaluate predictors of self-reported condom use. Participants who reported not using condoms were asked to provide reasons. From January 2013 to September 2019, 2482 participants reported having at least one regular sexual partner in the preceding 6 months. Of those, 1577 (63.5%) reported using a condom at last sex. Condom use was more common among older participants, males, HIV-infected participants, and those with an HIV-infected partner. Married participants, those with a partner of unknown HIV status, and those reporting alcohol use were less likely to report condom use at last sex. Condom use at last sex also varied significantly by clinical site. Partner disapproval or refusal to use a condom was a consistent driver of disparities in condom use among participants who were HIV infected, female, and aged 18-24 years. Effective HIV prevention programs should integrate condom education with the tools necessary to negotiate condom use with regular partners.

Decreasing time to antiretroviral therapy initiation after HIV diagnosis in a clinic-based observational cohort study in four African countries

INTRODUCTION

World Health Organization (WHO) guidelines have shifted over time to recommend earlier initiation of antiretroviral therapy (ART) and now encourage ART initiation on the day of HIV diagnosis, if possible. However, barriers to ART access may delay initiation in resource-limited settings. We characterized temporal trends and other factors influencing the interval between HIV diagnosis and ART initiation among participants enrolled in a clinic-based cohort across four African countries.

METHODS

The African Cohort Study enrols adults engaged in care at 12 sites in Uganda, Kenya, Tanzania and Nigeria. Participants provide a medical history, complete a physical examination and undergo laboratory assessments every six months. Participants with recorded dates of HIV diagnosis were categorized by WHO guideline era (<2006, 2006 to 2009, 2010 to 2012, 2013 to 2015, ≥2016) at the time of diagnosis. Cox proportional hazard modelling was used to estimate hazard ratios (HRs) and 95% confidence intervals (95% CI) for time to ART initiation.

RESULTS AND DISCUSSION

From January 2013 to September 2019, a total of 2888 adults living with HIV enrolled with known diagnosis dates. Median time to ART initiation decreased from 22.0 months (interquartile range (IQR) 4.0 to 77.3) among participants diagnosed prior to 2006 to 0.5 months (IQR 0.2 to 1.8) among those diagnosed in 2016 and later. Comparing those same periods, CD4 nadir increased from a median of 166 cells/mm3 (IQR: 81 to 286) to 298 cells/mm3 (IQR: 151 to 501). In the final adjusted model, participants diagnosed in each subsequent WHO guideline era had increased rates of ART initiation compared to those diagnosed before 2006. CD4 nadir ≥500 cells/mm3 was independently associated with a lower rate of ART initiation as compared to CD4 nadir <200 cells/mm3 (HR: 0.32; 95% CI: 0.28 to 0.37). Age >50 years at diagnosis was independently associated with shorter time to ART initiation as compared to 18 to 29 years (HR: 1.38; 95% CI: 1.19 to 1.61).

CONCLUSIONS

Consistent with changing guidelines, the interval between diagnosis and ART initiation has decreased over time. Still, many adults living with HIV initiated treatment with low CD4, highlighting the need to diagnose HIV earlier while improving access to immediate ART after diagnosis.

Biomarkers of Inflammation Correlate With Clinical Scoring Indices in Human Immunodeficiency Virus-Infected Kenyans

Background

In high-income countries, inflammation has been associated with increased morbidity and mortality in human immunodeficiency virus (HIV)-infected individuals despite treatment with antiretroviral therapy (ART). However, these findings may not be generalizable to low-income settings.

Methods

In this cross-sectional study, multivariable linear regression was used to compare 28 inflammatory biomarker levels in HIV-infected and -uninfected participants. Correlations between biomarkers and Veterans Aging Cohort Study (VACS) index, Fibrosis-4 (FIB-4) score, and Framingham risk score were assessed.

Results

Plasma samples from 304 Kenyans were analyzed. Compared to HIV-uninfected controls, virologically suppressed HIV-infected participants had higher levels of CCL5, CXCL10, fatty acid binding protein (FABP) 2, fas ligand (FASLG), matrix metalloproteinase (MMP) 1, MMP7, soluble CD14 (sCD14), and soluble CD163 (sCD163) and lower MMP9 (P < .01). CD4+/HLA-DR+CD38+ (ρ = 0.32; P < .001), sCD14 (ρ = 0.25; P = .004), and sCD163 (ρ = 0.24; P = .006) were correlated with the VACS index. FABP2 was positively correlated (ρ = 0.29; P = .002), whereas MMP1 (ρ = -.32; P < .001) and MMP2 (ρ = -0.28; P = .002) were inversely correlated with the FIB-4 score.

Conclusions

Differences in biomarker levels exist between well-controlled HIV-infected participants on ART and uninfected controls. Some biomarkers are correlated to scoring indices predictive of morbidity and mortality. These biomarkers could serve as prognostic indicators and inform therapeutic development.

1258. The Pregnancy Factor: Differences in the Prevalence of Depression Among Pregnant and Nonpregnant Sub-Saharan African Women Living with HIV

Background

In Sub-Saharan African (SSA) women living with HIV, pregnancy creates unique stressors that may induce or exacerbate depression. We described the prevalence of depression during the perinatal period among women living with HIV that became pregnant (P-WLWH) as compared with those that did not (NP-WLWH).

Methods

We analyzed data from WLWH, age < 45 years, across 11 HIV clinics in Kenya, Nigeria, Uganda, and Tanzania, within the African Cohort Study (AFRICOS). Depression screening was performed via the Center for Epidemiological Studies-Depression (CESD) scale at enrollment and every 6 months in the study. For these analyses, the CESD score was assessed during three periods in time. For P-WLWH, these periods were: (1) enrollment, (2) during pregnancy and (3) the study visit following delivery. For NP-WLWH, three sequential evaluations were analyzed, with visit 1 being enrollment. We compared the prevalence of depression between groups at each period using χ ² analysis and calculated adjusted odds ratios (OR) using logistic regression for variables associated with positive depression screening (CESD score ≥16).

Results

There were 764 WLWH selected for analysis, including 164 P-WLWH and 600 NP-WLWH. At enrollment, P-WLWH tended to be younger than NP-WLWH (mean age=30 vs. 34, P < 0.01). Neither group differed significantly with respect to marital status, employment, education, viral load, or duration of HIV. The crude prevalence of depression during periods 1 thru 3 for P-WLWH and NP-WLWH was: 20% (in both groups, P = 0.99), 10% and 13% (P = 0.26), and 4% and 10% (P = 0.01), respectively (figure). P-WLWH had 74% lower odds [OR = 0.26 [0.09–0.75], P = 0.01] of positive depression screening during period 3 than NP-WLWH, when controlling for age, marital status, educational level, viral load, and duration of HIV.

Conclusion

Surprisingly, our analyses demonstrate P-WLWH were less likely to have positive screening for depression following their pregnancies. These findings may highlight the strength of comprehensive maternal health initiatives in SSA. Additionally, these women will be followed further longitudinally to evaluate patterns of depression in the months following pregnancy.

Disclosures

All authors: No reported disclosures.

HIV virologic failure and its predictors among HIV-infected adults on antiretroviral therapy in the African Cohort Study

INTRODUCTION

The 2016 WHO consolidated guidelines on the use of antiretroviral drugs defines HIV virologic failure for low and middle income countries (LMIC) as plasma HIV-RNA ≥ 1000 copies/mL. We evaluated virologic failure and predictors in four African countries.

MATERIALS AND METHODS

We included HIV-infected participants on a WHO recommended antiretroviral therapy (ART) regimen and enrolled in the African Cohort Study between January 2013 and October 2017. Studied outcomes were virologic failure (plasma HIV-RNA ≥ 1000 copies/mL at the most recent visit), viraemia (plasma HIV-RNA ≥ 50 copies/mL at the most recent visit); and persistent viraemia (plasma HIV-RNA ≥ 50 copies/mL at two consecutive visits). Generalized linear models were used to estimate relative risks with their 95% confidence intervals.

RESULTS

2054 participants were included in this analysis. Viraemia, persistent viraemia and virologic failure were observed in 396 (19.3%), 160 (7.8%) and 184 (9%) participants respectively. Of the participants with persistent viraemia, only 57.5% (92/160) had confirmed virologic failure. In the multivariate analysis, attending clinical care site other than the Uganda sitebeing on 2nd line ART (aRR 1.8, 95% CI 1·28-2·66); other ART combinations not first line and not second line (aRR 3.8, 95% CI 1.18-11.9), a history of fever in the past week (aRR 3.7, 95% CI 1.69-8.05), low CD4 count (aRR 6.9, 95% CI 4.7-10.2) and missing any day of ART (aRR 1·8, 95% CI 1·27-2.57) increased the risk of virologic failure. Being on 2nd line therapy, the site where one receives care and CD4 count < 500 predicted viraemia, persistent viraemia and virologic failure.

CONCLUSION

In conclusion, these findings demonstrate that HIV-infected patients established on ART for more than six months in the African setting frequently experienced viraemia while continuing to be on ART. The findings also show that being on second line, low CD4 count, missing any day of ART and history of fever in the past week remain important predictors of virologic failure that should trigger intensified adherence counselling especially in the absence of reliable or readily available viral load monitoring. Finally, clinical care sites are different calling for further analyses to elucidate on the unique features of these sites.

Diagnosis disclosure to adolescents living with HIV in rural Kenya improves antiretroviral therapy adherence and immunologic outcomes: A retrospective cohort study

Background & aims

Emphasis on adolescent HIV has increased worldwide as antiretroviral treatment has greatly extended life expectancies of HIV-positive children. Few evidence-based guidelines exist on the optimal time to disclose to an adolescent living with HIV (ALHIV); little is known about the medical effects of disclosure. This study looked to determine whether disclosure is associated with improved medical outcomes in ALHIV. Prior work has tended to be qualitative, cross-sectional, and with an emphasis on psychosocial outcomes. This paper addresses the adolescent cohort retrospectively (longitudinally), building upon what is already known about disclosure.

Methods

Retrospective, longitudinal clinical record reviews of ALHIV seen at Kericho District Hospital between April 2004 and November 2012 were performed. Patient demographics and clinical outcomes were systematically extracted. The student’s t-test was used to calculate changes in mean CD4 count, antiretroviral therapy (ART), and cotrimoxazole adherence pre- vs. post-disclosure. Linear regression modelling assessed for trends in those clinical outcomes associated with age of disclosure.

Results

Ninety-six ALHIV (54 female, 42 male) were included; most (73%) entered care through the outpatient department. Nearly half were cared for by parents, and 20% experienced a change in their primary caregiver. The mean time in the study was 2.47 years; mean number of visits 10.97 per patient over the mean time in the study. Mean disclosure age was 12.34 years. An increase in mean ART adherence percentage was found with disclosure (0.802 vs. 0.917; p = 0.0015). Younger disclosure age was associated with significantly higher mean CD4 counts over the course of the study (p = 0.001), and a nonsignificant trend toward a higher mean ART adherence percentage (p = 0.055).

Conclusion

ART adherence and improved immunologic status are both associated with disclosure of HIV infection to adolescent patients. Disclosure of an HIV diagnosis to an adolescent is an important means to improve HIV care.

Kericho CLinic-based ART Diagnostic Evaluation (CLADE): design, accrual, and baseline characteristics of a randomized controlled trial conducted in predominately rural, district-level, HIV clinics of Kenya

Background

Prospective clinical trial data regarding routine HIV-1 viral load (VL) monitoring of antiretroviral therapy (ART) in non-research clinics of Sub-Saharan Africa are needed for policy makers.

Methods

CLinic-based ART Diagnostic Evaluation (CLADE) is a randomized, controlled trial (RCT) evaluating feasibility, superiority, and cost-effectiveness of routine VL vs. standard of care (clinical and immunological) monitoring in adults initiating dual nucleoside reverse transcriptase inhibitor (NRTI)+non-NRTI ART. Participants were randomized (1:1) at 7 predominately rural, non-research, district-level clinics of western Kenya. Descriptive statistics present accrual patterns and baseline cohort characteristics.

Results

Over 15 months, 820 adults enrolled at 7 sites with 86–152 enrolled per site. Monthly site enrollment ranged from 2–92 participants. Full (100%) informed consent compliance was independently documented. Half (49.9%) had HIV diagnosed through voluntary counseling and testing. Study arms were similar: mostly females (57.6%) aged 37.6 (SD = 9.0) years with low CD4 (166 [SD = 106]) cells/m³). Notable proportions had WHO Stage III or IV disease (28.7%), BMI <18.5 kg/m² (23.1%), and a history of tuberculosis (5.6%) or were receiving tuberculosis treatment (8.2%) at ART initiation. In the routine VL arm, 407/409 (99.5%) received baseline VL (234,577 SD = 151,055 copies/ml). All participants received lamivudine; 49.8% started zidovudine followed by 38.4% stavudine and 11.8% tenofovir; and, 64.4% received nevirapine as nNRTI (35.6% efavirenz).

Conclusions

A RCT can be enrolled successfully in rural, non-research, resource limited, district-level clinics in western Kenya. Many adults presenting for ART have advanced HIV/AIDS, emphasizing the importance of universal HIV testing and linkage-to-care campaigns.

Trial Registration

ClinicalTrials.gov NCT01791556

Occurrence of etravirine/rilpivirine-specific resistance mutations selected by efavirenz and nevirapine in Kenyan patients with non-B HIV-1 subtypes failing antiretroviral therapy

Resistance to efavirenz and nevirapine has not been associated with mutations at position 138 of reverse transcriptase. In an evaluation of virologic suppression rates in PEPFAR (President's Emergency Plan For AIDS Relief) clinics in Kenya among patients on first-line therapy (RV288), 63% (617/975) of randomly selected patients on antiretroviral therapy were suppressed (HIV RNA<400 copies/ml). Among those with non-nucleoside reverse transcriptase inhibitor resistance (n = 101), 14 (13.8%) had substitutions at 138 (A, G, K or Q), mutations selected only by etravirine and rilpivirine in subtype B viruses. All 14 patients received efavirenz or nevirapine, not etravirine or rilpivirine, and were predominantly subtype A1. This may be the first report of efavirenz and nevirapine selecting these mutations in these subtypes.