This study is the first to describe the prevalence of TDR among individuals with ARHI outside of a major metropolitan center in the US, and the first to describe this population in the South. Using data only from rigorously evaluated cohorts of ARHI, we found the overall 17.8% prevalence of TDR in NC comparable to previously published estimates of 10–15% among subjects with varying durations of HIV infection, principally from urban areas.1–6
We noted a similar increasing prevalence of NNRTI resistance over time, and a trend toward increasing prevalence of TDR.
Given the proven efficacy27, 28
, cost savings29
, and the availability of fixed-dose, once-daily cART utilizing NNRTIs, the frequency of mutations affecting this class – especially the 8% overall prevalence of K103N – significantly restricts therapeutic options available to the newly infected. In many locations throughout the rural South, public health departments are responsible for providing HIV services. Budget constraints often restrict their use of resistance testing to those felt to be failing a regimen, rather than as recommended for all HIV-positive patients newly entering care.30
Given the high frequency of TDR encountered in this study, approximately 1 out of every 10 new patients empirically
initiated on an NNRTI-based first-line regimen is at risk for early virologic failure.
Our cohort had comparatively fewer MSM, more women, and more black subjects than other studies.1–6
These results suggest that the dynamics of TDR are not necessarily a feature of urban epidemics, and may be consistent across geographic or demographic boundaries. By collecting data through both clinical research and statewide universal surveillance programs conducted in NC’s public HIV testing sites, we obtained a uniquely representative sample of the leading edge of the state’s epidemic. We are confident that our findings are generalizable to the public health system in NC. This effort may represent a model for surveillance of TDR by other states who participate in Centers for Disease Control and Prevention (CDC)-funded surveillance activities that include testing for recent HIV infection.
Our data have several limitations. Assembly of a cohort of ARHI is limited by inherent difficulties in case identification. Although the statewide program for detection of acute infection has improved our ability to find acutely infected individuals, our overall numbers remain relatively small. Sampling during ARHI improves the accuracy of our TDR prevalence estimate, yet it remains possible that some mutations were missed if their frequency fell below the limit of detection of “bulk” sequencing by the time of resistance testing. This would result in an underestimation of TDR prevalence – the same problem faced by TDR studies relying on chronically infected subjects.1, 5, 6
Because we pooled data from multiple sources, differences in data collection methods precluded us from evaluating associations between immunologic markers and TDR. Finally, we chose to use SDRMs as our standard, which may limit direct comparisons of our findings to other studies that utilize the IAS list of drug resistance mutations (IAS-DRMs). A few potentially relevant IAS-DRMs (e.g., the NRTI mutation, E44D) are not included in the more parsimonious SDRM list. However, Green et al
recently showed that, when compared to SDRMs as the reference standard, the overall sensitivity of IAS-DRMs was inferior (93.5%) for detecting TDR.31
Further, because they are not specifically selected to measure TDR, IAS-DRMs may overestimate the prevalence of resistance by as much as 7%.31
In conclusion, the frequency of TDR in NC mirrors that seen in major metropolitan areas, with approximately 10% of new infections having baseline NNRTI class resistance. Our data underscore the importance of obtaining ARV resistance testing prior to initiating therapy whenever possible – regardless of the clinical setting.