In the Mitra plus study of prevention of mother-to-child transmission of HIV-1, which included 501 women in Dar es Salaam, Tanzania, triple antiretroviral therapy (ART) was given from late pregnancy throughout breastfeeding up to 6 months postnatally. Here we report findings in a sub-cohort of women with ≤200 CD4cells/μL at enrolment who were continued on ART for life and followed up during 24 months after delivery to determine virologic and immunologic responses, drug resistance and mortality. Blood samples for viral load and CD4 counts testing were collected at enrolment and at 3, 6, 12 and 24 months postpartum. HIV drug resistance testing was performed at 12 months. Data analysis included descriptive statistics and multivariate analysis using Generalized Estimated Equations of 73 women with at least two postpartum assessments. The mortality analysis included 84 women who had delivered. The proportion of women with a viral load ≥400 copies/mL was 97% (71/73) at enrolment, 16% (11/67), 22% (15/69), 61% (36/59) and 86% (48/56) at 3, 6, 12 and 24 months postpartum, respectively. The proportion of women with immunologic failure was 12% (8/69), 25% (15/60) and 41% (24/58) at 6, 12 and 24 months, respectively. At 12 months, drug resistance was demonstrated in 34% (20/59), including 12 with dual-class resistance. Self-report on drug adherence was 95% (64/68), 85% (56/66), 74% (39/53) and 65% (30/46) at 3, 6, 12 and 24 months, respectively. The mortality rate was 5.9% (95% CI 2.5-13.7%) at 24 months. The probability of virologic and immunologic failure was significantly higher among women who reported non-perfect adherence to ART at month 24 postpartum. Following an initial decline of viral load, virologic failure was common at 12 and 24 months postpartum among women initiated on ART for life during pregnancy because of low CD4 cell counts. A high proportion of viremic mothers also had resistance mutations. However, at 24 months follow-up, the mortality rate was still fairly low. Continuous adherence counseling and affordable means of monitoring of the virologic response are crucial for successful implementation of the WHO Option B+ guidelines to start all HIV-infected pregnant women on ART for life.

In many African countries, prevention of mother-to-child transmission of HIV (PMTCT) services are predominantly delivered by nurses. Although task-shifting is not yet well established, community health workers (CHWs) are often informally used as part of PMTCT delivery. According to the 2008 World Health Organization (WHO) Task-shifting Guidelines, many PMTCT tasks can be shifted from nurses to CHWs. The aim of this time and motion study in Dar es Salaam, Tanzania, was to estimate the potential of task-shifting in PMTCT service delivery to reduce nurses’ workload and health system costs. The time used by nurses to accomplish PMTCT activities during antenatal care (ANC) and postnatal care (PNC) visits was measured. These data were then used to estimate the costs that could be saved by shifting tasks from nurses to CHWs in the Tanzanian public-sector health system. A total of 1121 PMTCT-related tasks carried out by nurses involving 179 patients at ANC and PNC visits were observed at 26 health facilities. The average time of the first ANC visit was the longest, 54 (95% confidence interval (CI) 42–65) min, followed by the first PNC visit which took 29 (95% CI 26–32) minutes on average. ANC and PNC follow-up visits were substantially shorter, 15 (95% CI 14–17) and 13 (95% CI 11–16) minutes, respectively. During both the first and the follow-up ANC visits, 94% of nurses’ time could be shifted to CHWs, while 84% spent on the first PNC visit and 100% of the time spent on the follow-up PNC visit could be task-shifted. Depending on CHW salary estimates, the cost savings due to task-shifting in PMTCT ranged from US$ 1.3 to 2.0 (first ANC visit), US$ 0.4 to 0.6 (ANC follow-up visit), US$ 0.7 to 1.0 (first PNC visit), and US$ 0.4 to 0.5 (PNC follow-up visit). Nurses working in PMTCT spend large proportions of their time on tasks that could be shifted to CHWs. Such task-shifting could allow nurses to spend more time on specialized PMTCT tasks and can substantially reduce the average cost per PMTCT patient.

Antibody responses that correlated with reduced risk of HIV acquisition in the RV144 efficacy trial were assessed in healthy African volunteers who had been primed three times with HIV-DNA (subtype A, B, C) and then randomized into two groups; group 1 was boosted twice with HIV-MVA (CRF01_AE) and group 2 with the same HIV-MVA coadministered with subtype C envelope (Env) protein (CN54rgp140/GLA-AF). The fine specificity of plasma Env-specific antibody responses was mapped after the final vaccination using linear peptide microarray technology. Binding IgG antibodies to the V1V2 loop in CRF01_AE and subtype C Env and Env-specific IgA antibodies were determined using enzyme-linked immunosorbent assay. Functional antibody-dependent cellular cytotoxicity (ADCC)-mediating antibody responses were measured using luciferase assay. Mapping of linear epitopes within HIV-1 Env demonstrated strong targeting of the V1V2, V3, and the immunodominant region in gp41 in both groups, with additional recognition of two epitopes located in the C2 and C4 regions in group 2. A high frequency of V1V2-specific binding IgG antibody responses was detected to CRF01_AE (77%) and subtype C antigens (65%). In conclusion, coadministration of CN54rgp140/GLA-AF with HIV-MVA did not increase the frequency, breadth, or magnitude of anti-V1V2 responses or ADCC-mediating antibodies induced by boosting with HIV-MVA alone.

Sera from 622 blood donors collected in 1986 and 1987 in Tanzania were screened for antibodies to HIV-1 by seven different commercial enzyme-linked immunosorbent assay (ELISA) kits. All ELISA-positive sera were tested by Western blot analysis and many of them also by radioimmunoprecipitation assay (RIPA). Sixty-seven sera were confirmed positive. Eight sera, which were repeatedly positive on only one or two of the ELISA kits and showed weak, doubtful reactions on Western blot and RIPA, were considered indeterminate and were not included in the calculations of sensitivity and specificity of the various ELISA kits. The sensitivity of the ELISAs was as follows: Organon Vironostika: low cut-off 74.6%; Organon Vironostika: high cut-off 62.7%; Du Pont: 85.1 %; Pasteur: 78.7%; Abbott: 80.8%; Abbott recombinant: 94.0%; Wellcozyme: 82.1%; Wellcozyme monoclonal: 98.5%. The specificity was as follows: Organon Vironostika: low and high cut-off 100%; Du Pont: 94.7%; Pasteur: 99.3%; Abbott: 100%; Abbott recombinant 98.7%; Wellcozyme: 99.8%; Wellcozyme monoclonal: 98.5%. In conclusion, the two new generation kits tested, Wellcozyme monoclonal and Abbott recombinant, had the highest sensitivity whereas the sensitivity of the first-generation tests was unexpectedly low.