Determinants of therapy failure among adults on first-line antiretroviral therapy in Asmara, Eritrea: a multicenter retrospective matched case–control study

Abstract Background Information on treatment failure (TF) in People living with HIV in a data-poor setting is necessary to counter the epidemic of TF with first-line combined antiretroviral therapies (cART) in sub-Saharan Africa (SSA). In this study, we examined the risk factors associated with TF in Asmara, Eritrea from 2001 to 2020. Methods A multicenter, retrospective 1:2 matched (by age and gender) case–control study was conducted in four major hospitals in Asmara, Eritrea on adults aged ≥ 18 years who were on treatment for at least 6 months. Cases were patients who fulfills at least one of the WHO therapy failure criterion during the study period. Controls were randomly selected patients on first-line treatment and plasma viral load < 1000 copies/ml in their latest follow-up measurement. Multivariable logistic regression analysis was conducted to identify risk factors for TF. All P-values were 2-sided and the level of significance was set at P < 0.05 for all analyses. Results Of the 1068 participants (356 cases; 712 controls), 585 (54.7%) were females. The median age at treatment initiation was 46 years [interquartile range (IQR): 39–51]. Median time to combined antiretroviral therapy (cART) failure was 37 months (IQR = 24–47). In the multivariate analysis, factors associated with increased likelihood of TF included initial nucleoside reverse transcriptase inhibitors (NRTI) backbone (Zidovudine + Lamivudine (AZT + 3TC): adjusted odds ratio (aOR) = 2.70, 95% Confidence interval (CI): 1.65–4.41, P-value < 0.001), (Abacavir + lamivudine (ABC + 3TC): aOR = 4.73, 95%CI: 1.18–18.92, P-value = 0.028], and (Stavudine + Lamivudine (D4T + 3TC): aOR = 5.00; 95% CI: 3.03–8.20, P-value < 0.001) in comparison to Emtricitabine and Tenofovir diproxil fumarate (FTC + TDF). Additional associations included prior exposure to cART (aOR = 2.28, 95%CI: 1.35–3.86; P- value = 0.002), record of sub-optimal drug adherence (aOR = 3.08, 95%CI: 2.22–4.28; P < 0.001), ambulatory/bedridden at presentation (aOR = 1.61, 95%CI: 1.12–4.28; P-value = 0.010), presence of comorbidities (aOR = 2.37; 95%CI: 1.36–4.10, P-value = 0.002), duration of cART (< 5 years: aOR: 5.90; 95% CI: 3.95–8.73, P-value < 0.001), and use of SMX-TMP prophylaxis (aOR = 2.00, 95%CI, 1.44–2.78, P-value < 0.001). Conclusion Our findings underscore the importance of optimizing cART adherence, diversification of cART regimens, and interventions directed at enhancing early HIV diagnosis, prompt initiations of treatment, and improved patient-focused monitoring of treatment response.