Hiv-1 subtype c-infected individuals maintaining high viral load as potential targets for the “test-and-treat” approach to reduce hiv transmission

The first aim of the study is to assess the distribution of HIV-1 RNA levels in subtype C infection. Among 4,348 drug-naïve HIV-positive individuals participating in clinical studies in Botswana, the median baseline plasma HIV-1 RNA levels differed between the general population cohorts (4.1-4.2 log10) and cART-initiating cohorts (5.1-5.3 log10) by about one log10. The proportion of individuals with high (≥50,000 (4.7 log10) copies/ml) HIV-1 RNA levels ranged from 24%-28% in the general HIV-positive population cohorts to 65%-83% in cART-initiating cohorts. The second aim is to estimate the proportion of individuals who maintain high HIV-1 RNA levels for an extended time and the duration of this period. For this analysis, we estimate the proportion of individuals who could be identified by repeated 6- vs. 12-month-interval HIV testing, as well as the potential reduction of HIV transmission time that can be achieved by testing and ARV treating. Longitudinal analysis of 42 seroconverters revealed that 33% (95% CI: 20%-50%) of individuals maintain high HIV-1 RNA levels for at least 180 days post seroconversion (p/s) and the median duration of high viral load period was 350 (269; 428) days p/s. We found that it would be possible to identify all HIV-infected individuals with viral load ≥50,000 (4.7 log10) copies/ml using repeated six-month-interval HIV testing. Assuming individuals with high viral load initiate cART after being identified, the period of high transmissibility due to high viral load can potentially be reduced by 77% (95% CI: 71%-82%). Therefore, if HIV-infected individuals maintaining high levels of plasma HIV-1 RNA for extended period of time contribute disproportionally to HIV transmission, a modified “test-and-treat” strategy targeting such individuals by repeated HIV testing (followed by initiation of cART) might be a useful public health strategy for mitigating the HIV epidemic in some communities. © 2010 Novitsky et al. This study was conducted according to the principles expressed in the Declaration of Helsinki. The study was approved by the Institutional Review Boards of Botswana and the Harvard School of Public Health. All patients provided written informed consent for the collection of samples and subsequent analysis. Description of the Botswana–Harvard Partnership (BHP) studies has been presented elsewhere [30]. For the purposes of this study, baseline data were used from the following seven BHP cohorts that were monitored including extensive clinical and laboratory follow up for prolonged periods. The time of enrollment to each cohort is shown in Supplementary Table S1. Three types of cohorts were distinguished: general population, MTCT, and cART-initiating cohorts. MTCT cohort BHP004, Mashi study: Prevention of milk-borne transmission of HIV-1C in Botswana (completed). The main goals of this project were two-fold. First, to assess whether the addition of a single dose of maternal nevirapine (NVP) at labor along with zidovudine (AZT or ZDV) from week 34 of gestation provides additional benefit in reducing HIV transmission from mother to child. The study was amended to determine whether maternal NVP (per HIVNET 012 protocol) is necessary in the setting of maternal ZDV from 34 weeks gestation through delivery and single-dose prophylactic infant NVP (at birth) plus ZDV (from birth to 4 weeks) for the reduction of HIV transmission from mother to child. The second goal was to determine the effectiveness and safety of prophylactic AZT to breast-feeding infants to prevent milk-borne HIV transmission. The baseline HIV RNA load in plasma was available for 1,189 Mashi participants. Results of the Mashi study were presented elsewhere [11], [31], [32], [33], [34]. cART-initiating cohort BHP007, Tshepo study: The adult antiretroviral treatment and drug resistance study (completed). The study was an open-label, randomized combination ARV study with a multi-factorial, 3x2x2 design. The factors included a comparison of three NRTI combinations (ZDV/lamivudine (3TC), ZDV/didanosine (ddI), and 3TC/stavudine (d4T)), a comparison of two NNRTIs (NVP and efavirenz (EFV)), and a comparison between two adherence strategies (standard of care (SOC) versus an intensified adherence strategy, SOC plus community-based supervision). The baseline HIV RNA load in plasma was available for 631 Tshepo participants. Results of the Tshepo study were presented elsewhere [35], [36], [37]. General population cohort BHP010, Botsogo study: A natural history of HIV-1 subtype C disease progression study (completed). This observational study gathered data on HIV-1 subtype C disease progression from ARV-naïve HIV-infected individuals with CD4+ cell count ≥400/mm3. The objectives of the study were (i) to determine the kinetics of HIV-1 subtype C disease progression (ii) to estimate the rate of CD4+ cell decline, and (iii) to analyze the time to first HIV-associated or AIDS-defining condition or death in persons with initial CD4+ cell count ≥400/mm3. The baseline HIV RNA load in plasma was available for 444 Botsogo participants. General population cohort BHP011, Dikotlana study: Micronutrient therapy and HIV in Botswana (completed). The study was a randomized, multifactorial, double-blind placebo-controlled trial to determine the efficacy of micronutrient supplementation in improving immune function and preventing early mortality in HIV-1-infected adults whose CD4+ were >350 cells/mm3. The design compared the efficacy of multivitamins, or selenium, or the combination of multivitamins and selenium to a placebo supplementation. The baseline HIV RNA load in plasma was available for 842 Dikotlana participants. MTCT cohort BHP016, Mma Bana study: A randomized trial of ZDV + 3TC + lopinavir/ritonavir vs. ZDV + 3TC + abacavir for virologic efficacy and the prevention of MTCT among breastfeeding women having CD4+>200 cells/mm3 in Botswana (ongoing). This study involved cART initiation by week 28 of gestation in breastfeeding women having CD4+>200 cells/mm3. The third group included pregnant women who received ZDV + 3TC (given as co-formulated Combivir™ or Lamzid™) + NVP as the National Program regimen because they had CD4+<200 cells/mm3. This group also breast-fed their infants. The baseline HIV RNA load in plasma was available for 726 Mma Bana participants. cART-initiating cohort BHP019, Mashi Plus study: The study was designed to determine the response to NVP-containing cART among women who have previously taken single-dose NVP for the prevention of MTCT (completed). The baseline HIV RNA load in plasma was available for 302 Mashi Plus participants. Results of the study were reported elsewhere [32], [38]. cART-initiating cohort BHP026, Bomolemo study: A prospective cohort study evaluating the efficacy and tolerability of tenofovir and emtricitabine (given as co-formulated Truvada™) as the NRTI backbone for first-line cART in treatment-naïve adults (ongoing). The baseline HIV RNA load in plasma was available for 214 Bomolemo participants. Although HIV-1 subtyping was not performed systematically for all individuals included in the seven BHP cohorts analyzed, our previous studies provide strong evidence for the overwhelming dominance of HIV-1 subtype C as the etiologic agent of the HIV/AIDS epidemic in Botswana [27], [39], [40], [41]. According to the HIV Sequence Database at LANL [42], 99.4% of the deposited 1,425 sequences from Botswana belong to HIV-1 subtype C. Therefore, we assume that the vast majority of subjects in this study are infected with HIV-1 subtype C. Both baseline and longitudinal data were used from the eights cohort, BHP012 Tshedimoso study, n = 42, Markers of Viral Set Point in Primary HIV-1C Infection (ongoing). The study was designed to evaluate potential trends between viral load and viral genetic diversity in acute and early HIV-1 subtype C infection, to determine the relationship between virologic parameters and viral set-point, and to identify immunological parameters that correlate with viral set-point in primary HIV-1 subtype C infection. All subjects included in the longitudinal analysis were genotyped and were found to be infected with HIV-1 subtype C. Results of the study were reported elsewhere [27], [39], [41], [43], [44], [45]. The primary infection cohort was comprised of individuals with estimated time of seroconversion. For acutely infected subjects (n = 8) the time of seroconversion was estimated as the midpoint between the last seronegative test and the first seropositive test (within a week in most cases). For recently infected subjects (n = 34) the time of seroconversion was estimated by Fiebig stage assignment as described elsewhere [43], [45]. For a time zero we used the estimated time of seroconversion rather than the estimated time of HIV infection because frequent sampling in this study allowed reliable measurement of the time of seroconversion based on a series of laboratory tests, which can be more accurate than estimation of the time of HIV infection. Time points of sampling and HIV-1 RNA testing in the primary infection study (n = 42) are presented in the Supplementary Figure S1. Individuals whose CD4+ cell count dropped below 200 cells per cubic millimeter or developed opportunistic infection had access to antiretroviral therapy (Combivir (ZDV/3TC) 300/150 mg twice a day plus nevirapine 200 mg twice a day if female, or efavirenz 600 mg every day if male) free of charge, in accordance with Botswana National Treatment Program guidelines. Plasma HIV-1 RNA was quantified by the COBAS Ampli-Prep/COBAS AMPLICOR HIV-1 Monitor Test, version 1.5, according to the manufacturer's instructions as described previously [27]. The method of viral load quantification used in the study has been certified by the Virology Quality Assurance at Rush University, Chicago, IL, as a part of the laboratory proficiency testing. The level of detection was from 50 (1.7 log10) copies/ml for the ultrasensitive method and 400 (2.6 log10) copies/ml for the standard method to 750,000 (5.88 log10) copies/ml. Analysis of individuals with estimated time of seroconversion from the Tshedimoso study included both pre- and post-cART data, which is clearly indicated in the presenting materials. Descriptive statistics (mean and accompanying 95% confidence intervals, median and corresponding inter-quartile range) were quantified using Sigma Stat v. 3.5. Comparisons of continuous outcomes between two groups were based on the Mann-Whitney Rank Sum test. A Spearman rank correlation was used for analysis of potential associations between continuous variables. The Kolmogorov-Smirnov test was used to test whether the distribution of a continuous outcome follows a normal distribution. For the purpose of analysis in this study we defined a “high-viral-load individual” as a subject with plasma HIV-1 RNA levels ≥50,000 (4.7 log10) copies/ml at a given test time-point. We defined the “period of high transmissibility,” or “duration of high viral load,” as the time period during which an HIV-infected individual has plasma HIV-1 RNA ≥50,000 (4.7 log10) copies/ml. For the 14 seroconverters with high early HIV-1 RNA levels in the Tshedimoso study, we estimated the duration of high viral load in the absence of cART using cubic smoothing splines for those with more than 5 data points, and ordinary least squares regression for those with fewer data points. For individuals with increasing HIV-1 RNA levels, the duration of high viral load was imputed as the time from seroconversion to the last observation prior to cART initiation. In the sensitivity analysis, the duration of high viral load for subjects with increasing HIV RNA (n = 3) was estimated using the Kaplan-Meier method. To describe the procedure for estimating the potential reduction in the period of high viral load, we introduce some notation: Let X denote the time when new infections occur and Y denote the duration of high viral load. We assume that X follows a uniform distribution within the testing interval and is independent of Y. The distribution of Y is estimated from the empirical distribution based on the 14 seroconverters. Using τ to denote the length of testing interval, the proportion of individuals with high viral load who can be identified by using repeated HIV testing at τ-month interval is Pr(X+Y> τ), the probability of X+Y being greater than τ. The potential reduction in the period of high HIV transmissibility in individuals with high viral load that can be achieved by repeated HIV testing and ARV treatment was approximated by E(X+Y-τ|X+Y-τ ≥0), the expected value of X+Y-τ when it is positive. Confidence intervals for these two quantities were derived using the bootstrap method [46]. All reported p-values are 2-sided and not adjusted for multiple comparisons.