Background:  Recent studies indicate that transmission of HIV-1 with antiretroviral drug-resistance mutations may be increasing in the United States and that newly diagnosed persons may have increased risk of treatment failure. Additionally, increased travel and migration may influence the prevalence of HIV subtypes not commonly found in the United States. Using data from the U.S. Variant, Atypical, and Resistant HIV Surveillance system, we estimated the prevalence of antiretroviral drug resistance mutations and the distribution of subtypes.

Methods:  In 11 states (409 sites), PR and RT sequence data were identified from specimens drawn from March 2003 to October 2006 from newly diagnosed, confidentially tested, and drug-naïve persons (diagnosis dates:  January 2003 to October 2006). Major antiretroviral drug resistance mutations were defined according to the International AIDS Society (IAS) -USA guidelines, and subtypes were derived using the Stanford online HIV Drug Resistance Database. These data were merged with demographic and clinical data reported to the national HIV case surveillance system.

Results:  Of 3130 specimens suitable for analysis, antiretroviral drug resistance mutations were observed in 327 (10.4%) persons. The range in participating states was 6.3 to 13.0%. Drug-resistance mutations to nucleoside reverse transcriptase inhibitors (NRTI), non-NRTI (NNRTI), and protease inhibitors (PI) were found in 111 (3.6%), 217 (6.9%), and 75 (2.4%) persons, respectively. The predominant mutations were M41L for NRTI (45.1%), K103N for NNRTI (70.1%), and L90M for PI (40.0%). Multi-drug resistance mutations were noted in 60 (1.9%) persons, ranging from 0 to 0.3% across states, with 44 (1.4%) persons having drug-resistance mutations to at least 1 drug in 2 classes and 16 (0.5%) having 3 class antiretroviral drug resistance mutations. Subtype B was observed in 2971 (94.9%) persons and non-B subtypes or recombinant forms were found in 158 (5.1%) persons. Among the latter, subtype C and CRF02_AG were most prominent accounting for 62 (2.0%) and 45 (1.4%) persons, respectively. The range of non-B subtypes and recombinant forms by state varied from 0 to 1.6%. Antiretroviral drug resistance mutations were observed in 13 (8.2%) persons with non-B subtypes or recombinant forms and in 314 (10.6%) persons with subtype B, though this difference was not statistically significant.

Conclusions:  These data represent the largest and most diverse U.S. population-based sample reported to date, and have implications for initial treatment regimens for persons newly diagnosed with HIV. These findings also emphasize the importance of ongoing molecular HIV surveillance in the United States and will be useful for future evaluations of trends associated with the transmission of drug-resistance mutations and subtype distribution.