We found that despite the introduction of novel ART agents in San Francisco starting in late 2007, including raltegravir, maraviroc, and etravirine, the prevalence of transmitted drug resistance in 2008–2009 remained substantial and was not significantly different than in prior years.
To our knowledge, this is the first report of TDR among persons infected with HIV in 2008–2009, after the newest ART medications arrived. Among patients infected in 2005–2007, Hurt et al. reported a 21% prevalence of TDR in North Carolina, similar to our estimate 
. Numerous reports exist on TDR trends among persons with chronic HIV, but the impact of new ART is difficult to assess using these data as they can reflect HIV acquisition years before HIV diagnosis.
Our finding that TDR remained substantial in 2008–2009 could have multiple explanations. It is unlikely that new ART medications have inadequate potency given their demonstrated efficacy in treating drug-resistant HIV 
. Therefore one explanation for our findings is that suboptimal engagement in medical care and/or poor ART adherence could be limiting the penetration of new ART medications among TDR source patients, leading to a persistently elevated TDR rate. However, recent data from San Francisco investigators demonstrates that virologic suppression is improving, and circulating levels of viremia are decreasing, arguing against poor engagement in care or poor adherence as chief explanations for our results 
A second explanation of our findings is that a sizeable fraction of drug-resistant HIV is being transmitted by treatment-naïve persons who themselves acquired TDR. Given that several years typically elapse between HIV infection and diagnosis, ART-naïve individuals have ample time to transmit drug resistance mutations to new recipients. Additionally, persons with early stage HIV are the least likely to be on ART, yet may be responsible for a large fraction of forward transmission events 
. Furthermore, persons with acute/early HIV may have higher seminal HIV RNA levels, and thus higher infectivity 
. In this scenario—where TDR is largely transmitted by ART-naïve persons—the effects of novel ART regimens on TDR rates may be delayed.
The TDR prevalence we observed in recent years differs from that predicted by the mathematical model of Smith et al. 
, who had predicted rising NNRTI resistance from 2008–2009. In contrast, our results suggest a stable or even decreasing prevalence of transmitted NNRTI resistance in San Francisco. While the sophisticated aforementioned model provides important insights into TDR patterns, it did not appear to account for changes in HIV treatment options afforded by the newest ART medications, possibly explaining the differences in results.
Our analysis includes several important limitations. First, our acute/early HIV cohort is not a representative population-based sample. As such, our results may not fully reflect the HIV epidemic in San Francisco, though the demographic characteristics of our cohort subjects closely mirror those of the San Francisco epidemic overall. Since our study focused on one geographic area, there are likely to be differences in TDR rates in other settings, particularly in resource-limited regions where treatment options differ substantially. Second, we lack data on the uptake of new ART medications in San Francisco, information which might best be obtained through clinical databases of treatment records but which was not available for this analysis. If the uptake of new ART agents has been slow, more time may be needed to assess the impact on TDR. Third, our results should be interpreted with caution given the wide confidence intervals surrounding the annual estimates for TDR. Studies with longer duration sampling across broader geographic areas may allow better comparisons of TDR before and after the introduction of new ART agents.
Despite these limitations our data provide potential insights into drug-resistant HIV transmission in a setting with historically high levels of drug resistance. Despite the arrival of several novel ART medications beginning in late 2007, the prevalence of TDR remains substantial at the current time. This emphasizes that early diagnosis and aggressive treatment strategies for patients with drug-resistant HIV remain crucial. Continued surveillance will be essential in fully understanding the impact new ART agents will have on TDR epidemiology.