Background
Understanding of the extent and patterns of HIV-1 drug resistance in the era of globally widespread dolutegravir therapy is limited, particularly for youth. Data on low-level viremia (LLV; 200-1,000 copies/mL) are sparse, primarily due to technical difficulties with genotyping at low RNA levels, and the evolving global definitions of treatment failure.
Methods
We developed a protocol for LLV genotyping by next generation sequencing (NGS), applied it to specimens from Kenyan children and adolescents longitudinally-followed and failing dolutegravir-based antiretroviral therapy, and compared the extent of NRTI, NNRTI and INSTI resistance between specimens with LLV and those with high-level viremia (HLV), defined as viral load >1000 copies/mL (current WHO viral suppression threshold). Established drug resistance mutations (DRMs) were defined at NGS frequency ≥20%, while minority drug resistance variants were considered at NGS frequency 2-20%. Differences in proportions were estimated using Fisher’s exact test.
Results
In 397 participants (median age 20 years; 49% female; median 14 years on antiretroviral therapy and 4 years on dolutegravir), 77% achieved viral suppression. For established DRMs, of 42 LLV specimens, 37 Reverse Transcriptase (RT) and 38 Integrase (INT) sequences were generated (88%-90% genotyping success), demonstrating overall resistance in 71% (NRTI 38%; NNRTI 68%; major INSTI 0%; accessory INSTI 13%). Of 168 HLV specimens, 157 RT and 159 INT sequences were generated (93%-95% genotyping success), demonstrating overall resistance in 54% (NRTI 26%; NNRTI 49%; major INSTI 0%; accessory INSTI 6%). Minority resistance variants were detected in 53% of LLV specimens (NRTI 11%, NNRTI 46%, major INSTI 8%, accessory INSTI 0%); and 60% of HLV specimens (NRTI 36%, NNRTI 39%, major INSTI 10%, accessory INSTI 6%. NNRTI established DRMs were higher in LLV than HLV (68% vs. 49%; p<0.05), and minority NRTI drug resistance variants were lower in LLV than HLV (OR 0.214, 95% CIs 0.052 to 0.647; p<0.005). Other eight tests were non-significant.
Conclusions
Developing a genotyping protocol for LLV enabled exploration of drug resistance in this challenging and increasingly relevant viral load range. High treatment failure in this vulnerable Kenyan youth population on DTG-based regimens and similar overall resistance prevalence in LLV versus HLV are concerning, enhanced by major INSTI minority resistance variants. Longitudinal follow up to determine the impact is ongoing.
Table or Figure
