No abstract

BackgroundWe describe the accumulation of HIV-1 drug resistance and its effect on the activity of next-line components in patients with virological failure (HIV-1 RNA >1000 copies/mL) after 1 year (t1) of first-line antiretroviral therapy (ART) not switching to second-line drugs for one additional year (t2) in low-middle income countries (LMIC).Methods and resultsWe selected 48 patients from the DREAM cohort (Maputo, Mozambique); their median pre-ART CD4+ cell count was 165 cells/μl. At t1 patients were receiving ART since a median of 12.2 months (mainly zidovudine/lamivudine/nevirapine), their median HIV RNA was 3.8 log10 copies/mL, 43 (89.6%) presented at least one resistance-associated mutation (RAM), most frequently for lamivudine/emtricitabine, nevirapine and efavirenz. Resistance to tenofovir, was 10% at 1 year and higher than 20% at 2 years, while projection at 3 years was >30%. At t2, 42 (89.4%) had a predicted low-level or higher resistance to at least 1 s-line drug. At t1, the frequency of RAM in patients with a lower adherence to pharmacy appointments (<95%) was significantly lower (12/20, 60% for NRTI and 14/20, 70% for NNRTI) than in those with a better adherence (26/28, 92.8% for NRTI and 25/28, 89.3% for NNRTI) (OR 0.12, 95% CI 0.02–0.63, p = 0.012 and OR 0.28, 95% CI 0.06–1.29, p = 0.103, respectively). Overall thymidine analogue mutations (TAMs) accumulation rate was 0.32/year, 0.50/year in the subgroup with HIV RNA >10,000 copies/mL; NNRTI RAM accumulation rate was 0.15/year, 0.40/year in the subgroup with HIV RNA >10,000 copies/mL.ConclusionsWhile the activity of NNRTIs is compromised early during failure, tenofovir and zidovudine activity are reduced more frequently after 1 year of documented virological failure of thymidine analogue-based first-line ART, with RAMs accumulating faster in patients with higher viral loads. The present observation may help informing decisions on when to switch to a second line ART in patients on virological failure in LMIC.

Objective: Tuberculosis (TB) is the leading cause of death in HIV-positive people. In Kenya, 140 000 new TB cases occurred in 2019, and 13 000 HIV-positive patients died due to TB. The objective of this study was to investigate the role of high-sensitivity C-reactive protein (HS-CRP) in TB diagnosis and the prediction of mortality in HIV-positive patients. Methods: The IDEA-TB Study enrolled HIV-positive adult patients attending three DREAM centres in Kenya who were suspected of having TB. A lateral flow urine lipoarabinomannan assay (LF-LAM), serum HS-CRP, and GeneXpert MTB/RIF assay (Xpert MTB/RIF) were performed. Six-month survival was evaluated. Results: A total of 574 patients were enrolled. The median (interquartile range) age, body mass index, and CD4 count were 45 years (37-54 years), 20.5 kg/m 2 (18.5-23.69 kg/m 2 ), and 477 cells/mL (290-700 cells/ mL), respectively. TB was confirmed in 87 (15.2%) patients. Concordance between the Xpert MTB/RIF and LF-LAM tests was 87.1%. HS-CRP was higher in TB patients (35.39 mg/l vs 9.21 mg/l). Malnutrition and elevated HS-CRP were associated with TB: odds ratio (OR) 2.5 (95% confidence interval (CI) 1.14-5.72) and OR 6.6 (95% CI 3.87-11.52), respectively. Nine (1.6%) patients died during follow-up. No single factor was associated with mortality. Only the combination of malnutrition and elevated HS-CRP was highly predictive of death (odds ratio (OR) 9.8, 95% CI 1.88-50.95); the association was stronger in TB patients (33.3% vs 1.0%; OR 47.6, 95% CI 7.03-322.23). Conclusion: TB diagnosis in HIV-positive patients remains challenging. HS-CRP could play a role in predicting early mortality in symptomatic patients.

Background. The use of dried blood spots (DBS) for HIV1 viral load quantification can greatly improve access to viral monitoring for HIVinfected patients receiving treatment in resourcelimited settings. Objectives. To evaluate and validate HIV viral load measurement from DBS in subSaharan Africa, with a reliable, allautomated, standard commercial assay such as the Abbott m2000. Methods. A total of 277 DBS were collected in different health centres in Malawi and Mozambique and analysed for viral load determination using the Abbott m2000 assay with the corresponding plasma samples as gold standard. Samples were extracted using the m2000SP automatic extractor and then processed as the plasma samples using the specific 1.0 mL HIVRNA DBS protocol. Results. Among samples with detectable HIVRNA the correlation between viral load obtained from the paired 131 plasma and DBS samples was high (r=0.946). Overall, viral load values between DBS and plasma differed by less than 0.5 log unit in 90.1% of cases and by less than 1 log unit in 100% of cases. Using a threshold of 1 000 copies/mL (defining virological failure in resourcelimited settings), sensitivity was 94.2% and specificity 98.6%, and both positive and negative predictive values were high (98.5% and 94.5%, respectively). Conclusion. DBS extracted and processed using the Abbott automated system can be reliably used in resourcelimited setting to diagnose virological failure.