Pretreatment HIV-drug resistance in Mexico and its impact on the effectiveness of first-line antiretroviral therapy: a nationally representative 2015 WHO survey.

Background: WHO has developed a global HIV-drug resistance surveillance strategy, including assessment of pretreatment HIV-drug resistance. We aimed to do a nationally representative survey of pretreatment HIV-drug resistance in Mexico using WHO-recommended methods.

Methods: Among 161 Ministry of Health antiretroviral therapy (ART) clinics in Mexico, the largest, including 90% of ART initiators within the Ministry of Health (66 in total), were eligible for the survey. We used a probability-proportional-to-size design method to sample 25 clinics throughout the country. Consecutive ART-naive patients with HIV about to initiate treatment were invited to participate in the survey; individuals with previous exposure to ART were excluded. We assessed pretreatment HIV-drug resistance by Sanger sequencing and next-generation sequencing of viruses from plasma specimens from eligible participants with Stanford University HIV Drug Resistance Database methods. We obtained follow-up data for a median of 9·4 months (range 6-12) after enrolment. We investigated possible relations between demographic variables and pretreatment drug resistance with univariate and multivariate logistic regression.

Findings: Between Feb 3 and July 30, 2015, we screened 288 patients in 25 clinics, from whom 264 provided successfully sequenced viruses with no evidence of current exposure to antiretroviral drugs. With the Sanger method, of these 264 participants, 41 (15·5%, 95% CI 11·4-20·5) had pretreatment resistance to any antiretroviral drug and 28 (10·6%, 7·2-15·0) had pretreatment resistance to non-nucleoside reverse transcriptase inhibitors (NNRTIs). At least low-level pretreatment resistance (Stanford penalty score ≥15) was noted in 13 (4 · 9%) of participants to efavirenz and in 23 (8·7%) to the combination tenofovir plus emtricitabine plus efavirenz. With next-generation sequencing, of 264 participants, 38 (14·4%, 95% CI 10·4-19·2) had pretreatment resistance to any antiretroviral drug and 26 (9·8%, 6·5-14·1) had pretreatment resistance to NNRTIs. After median follow-up of 8 months (IQR 6·5-9·4, range 5-11) after ART initiation, 97 (72%) of 135 NNRTI initiators achieved viral suppression (<50 copies per mL) compared with ten (40%) of 25 individuals who started with protease inhibitor-based regimens (p=0·0045). After multivariate regression considering pretreatment resistance and initial ART regimen as composite variables, people starting NNRTIs with pretreatment drug resistance achieved significantly lower viral suppression (odds ratio 0·24, 95% CI 0·07-0·74; p=0·014) than patients without NNRTI resistance.

Interpretation: High levels of pretreatment drug resistance were noted in Mexico, and NNRTI pretreatment drug resistance significantly reduced the effectiveness of first-line ART regimens based on these drugs. Baseline HIV-drug resistance testing for initial ART follow-up and decision making should be considered.

Funding: The Mexican Government and Consejo Nacional de Ciencia y Tecnología.