Evaluation of predictors of HIV/AIDS treatment failure in patients with suspected virological and immunological failure from Comprehensive Care Clinic in Nairobi, 2013

Abstract

Global access to antiretroviral therapy (ART) is continuously increasing hence it is estimated that in four years of therapy 5-10% of patients will have their first line of therapy failing, due to development of resistance. Predictors of treatment failure include: plasma viral load above 1000 copies/mL, fall of cluster of differentiation 4 (CD4) count below 350 cells/mm3 in spite of therapy. This study was carried out to evaluate the reliability of these predictors in the detection of treatment failure. Eighty one (81) subjects were recruited from the comprehensive care clinic (CCC), Nairobi, based on the inclusion criteria. The reverse transcriptase region (RT) of the HIV-1 genome was amplified using nested polymerase chain reaction (PCR) and a total of 80 samples were directly sequenced. The sequences were individually blasted into the Stanford HIV Drug Resistance data base insitu tools, where genotypic drug resistance was defined by one or more resistant-related mutations. Isolates showed that the majority of patients who showed extreme treatment failure (CD4 count <100cells/μl and high viral load >100,000 copies/ml) were directly related to mutations such as T215Y, MI84V, M41L, K65KN, K103N, V90I, G190A, Y181C, and H221Y. The positive predictive value (PPV) of viral load was 100% reliable in the prediction of treatment failure however CD4+ count was 89% reliable. The Pearson correlation coefficient of CD4 count and viral load, was observed to be significant at the 0.01 level, r = -0.762 hence R2 = 0.58. The correlation coefficient is indicative of a weak correlation between viral load and CD4+ count. The HIV-1 subtypes in this study were as follows: subtype A1 was the highest 47(58%) subjects, subtype D 22(27%), subtype C 5(6.25%) subjects and subtype G present in one subject respectively. There were two circulating recombinant factor (CRF01_AE and CRF02_AG). The data obtained suggests that viral load is reliable in the detection of treatment failure, however HIV-1 genotypic drug resistance test should be highly considered for patients suspected to be failing therapy before they are switched.