Phenotypic and genotypic resistance patterns of HIV-1 isolates derived from individuals treated with didanosine and stavudine

OBJECTIVE: To evaluate the phenotypic susceptibilities and genotypic resistance patterns to both didanosine and stavudine of baseline and follow-up HIV-1 isolate pairs, derived from antiretroviral naive subjects treated with this dual nucleoside combination. DESIGN AND METHODS: Phenotypic drug susceptibility testing was performed in peripheral blood mononuclear cells on 34 viral isolate pairs derived from patients participating in the BMS AI-460 trial. Sequencing of the complete reverse transcriptase of 36 study isolate pairs, baseline and follow-up, was performed using standard dideoxy techniques. RESULTS: The mean fold change in susceptibilities to didanosine was 1.6 (P= 0.278) and to stavudine 1.9 (P= 0.002, Wilcoxon's signed rank test). Mutations classically associated with zidovudine resistance were observed to emerge in 7 out of 36 isolates, T215Y/F (four), M41L +T215Y/F (two) and D67N (one). Other mutations observed included the A62V, V751, F77L, F116Y, Q151 M multinucleoside resistance complex (one), the Q151M mutation (two) and the rare V75T mutation (two). No mutations classically associated with didanosine exposure and resistance were observed. No relationship was evident between the emergence of zidovudine associated mutations and the level of phenotypic resistance to either stavudine or didanosine or between the emergence of zidovudine associated mutations and changes in plasma HIV RNA levels. CONCLUSION: These comprehensive data demonstrate modest (< twofold) mean reductions in didanosine and stavudine susceptibilities at follow-up. The emergence of zidovudine associated mutations in this retroviral-naive population treated with combination didanosine and stavudine therapy is notable. Furthermore, the emergence of these mutations and of the Q151 M multinucleoside resistance complex raise concerns for potential nucleoside analog cross-resistance. The potential mechanisms driving the selection of the zidovudine associated mutations in the setting of didanosine and stavudine therapy and the relevance of these findings to current three and four drug regimens merit further evaluation.