Treatment Switch Algorithms for ART Viral Load Monitoring: Implications of Highly-Prevalent Resistance Among Previously Unmonitored Cohort in Malawi

Background:

Despite expanding viral load (VL) testing to resource-limited settings, resistance testing is unavailable. Understanding drug resistance patterns is essential to devise treatment switch algorithms. We evaluated reverse-transcriptase (RT) inhibitor resistance among a previously unmonitored antiretroviral therapy (ART) cohort in Malawi.

Methods:

We enrolled HIV-infected persons at ART clinics in Malawi (03/2013-01/2014). Blood was collected via fingerstick dried blood spot (DBS) cards. We used an in-house nested RT-PCR for two regions of the RT gene from DBS-extracted nucleic acid, followed by population sequencing for persons on 1^st-line ART experiencing virological failure (≥5000 copies/ml) with confirmatory testing to determine resupression (<1000 copies/ml). Resistance was evaluated using Stanford University HIV Drug Resistance Database.

Results:

Among 1,498 enrolled patients, 1,348 (90.3%) were on ART >12 months; 161 (10.8%) had signs of clinical failure. 88 (5.8%) had VL >5,000 copies/ml, 80 of whom were on ART >12 months and 16 with signs of clinical failure. After adherence counseling, 78/88 (88.6%) had confirmatory specimens collected; 13/78 (16.7%) resupressed (<1000 copies/ml). There was no difference in likelihood of resuppression comparing persons on ART <12 vs >12 months (16.4% vs 20%, p=0.8), and a trend towards greater resuppression among those with clinical symptoms (37.5% vs 14%, p=0.1). Among participants with elevated VL, 73/88 (83%) had resistance testing attempted; 57/61 (95%) of amplified sequences demonstrated NRTI (51) and/or NNRTI (55) resistance. There was a trend towards greater resistance among those on ART >12 vs <12 months (95% vs 80%, p=0.2) as well as those with clinical failure vs no clinical failure (100.0% vs 91.5%, p=0.3).

Conclusion:

We observed high rates of ART resistance among patients with elevated VLs in a Malawian cohort of previously unmonitored patients. Our observations might suggest a role for immediate switching given low rates of resuppression and high levels of resistance, but further characterization of the population benefit from immediate switch will require additional evaluation.