HIV-1 reverse transcriptase and protease resistance mutations selected during 16-72 weeks of therapy in isolates from antiretroviral therapy-experienced patients receiving abacavir/efavirenz/amprenavir in the CNA2007 study.

OBJECTIVE: To determine HIV-1 reverse transcriptase (RT) and protease (PRO) mutations selected in isolates from antiretroviral therapy (ART)-experienced patients receiving an efavirenz/abacavir/amprenavir salvage regimen. METHODS: Open-label, single arm of abacavir, 300 mg twice daily, amprenavir, 1200 mg twice daily and efavirenz, 600 mg once daily, in ART-experienced patients of which 42% were non-nucleoside reverse transcriptase inhibitor-naive. The virology population examined consisted of all patients who took at least 16 weeks of study drugs (n=74). Plasma population sequencing was carried out at baseline and last time point at which patients were still taking the three study drugs + other ART. The median follow-up was 48 weeks (range week 16-72). RESULTS: Baseline (n=73) and on-therapy (n=49) genotypes were obtained. By 48 weeks, 51% of isolates had > or = 3 non-nucleoside reverse transcriptase inhibitor (NNRTI) mutations. NNRTI mutations selected on therapy were K103N (51%), substitutions at position 190 (17/49, 35%): G to A (n=11) / S (n=4) / E (n=1) and T (n=1); L100I (37%) and V1081 (20%) mutations. P225H was not observed in this study. L100I and G190A/S/E/T mutations were rarely detected in the same viral population and baseline Y181C favoured the G190 mutations (OR=8.9, P<0.001), rather than the L100I. The NRTI mutations selected were in accordance with abacavir known resistance profile, no new TAMs were observed, new L74V or I mutations developed in 39 and 16% of isolates, respectively, however, new M184V mutations were only detected in isolates from two patients, one of whom had added lamivudine + didanosine. M184V was common at baseline (55%) and maintained in 22/27 (81%) isolates (five of these 22 added lamivudine or didanosine, or both). The PRO mutations selected were in accordance with the distinct resistance profile of amprenavir compared with other protease inhibitors. Mutations D30N, G48V, N88D/S, L90M and 154V were de-selected, and mutations I50V, I or V to 54M/L, I84V, M46I/L, L33F, I47V as well mutations at position 10 were observed in 20/49 (41%) isolates. CONCLUSION: Prior NNRTI and NRTI therapy influences the pathway of resistance to efavirenz. In this study, the prevalence of mutations selected by efavirenz were different from those described in less ART-experienced patients. Baseline Y181C was associated with the development of mutations at position 190, but not L100I or K103N. In this patient population, abacavir with efavirenz preferentially selected for L74V but not for thymidine analogue mutations. M184V was rarely selected and was maintained in only 77% of patients who did not add lamivudine or didanosine. Finally, amprenavir-specific mutations were selected in the background of other primary protease inhibitor mutations, confirming the distinct resistance profile of amprenavir.

Mutations in HIV-1 reverse transcriptase during therapy with abacavir, lamivudine and zidovudine in HIV-1-infected adults with no prior antiretroviral therapy.

OBJECTIVE: To evaluate HIV-1 reverse transcriptase (RT) drug resistance in patients receiving abacavir, lamivudine and zidovudine therapy. METHODS: In a randomized, double-blind study, 173 antiretroviral treatment-naive HIV-1-infected adults received abacavir/lamivudine/zidovudine or lamivudine/zidovudine for up to 48 weeks. After week 16, patients could switch to open-label abacavir/lamivudine/zidovudine, and those with plasma HIV-1 RNA (vRNA) > 400 copies/ml could add other antiretrovirals. From weeks 11 to 48, samples with vRNA > 400 copies/ml were collected for genotyping and phenotyping. RESULTS: At baseline, 90% of isolates were wild-type (WT). At week 16, vRNA was > 400 copies/ml in seven of 72 (10% patients receiving abacavir/lamivudine/zidovudine and in 41 of 66 (62%) receiving lamivudine/ zidovudine. At week 16, the genotypes in isolates from the abacavir/lamivudine/zidovudine group were M184V alone (n = 3 cases), WT (n = 3) and M184V plus thymidine analogue mutations (TAMs) (n = 1). The genotypes in isolates from the lamivudine/zidovudine group were M184V alone (n = 37), WT ( n= 1) and M184V plus TAMs (n = 3). In the four cases where M184V plus TAMs were detected some mutations were present at baseline. Despite detectable M184V in 74% of patients on lamivudine/zidovudine, addition of abacavir with or without another antiretroviral therapy resulted in a reduction in vRNA, with 42 of 65 (65%) patients having week 48 vRNA < 400 copies/ml (intent-to-treat with missing = failure). At week 48, the most common genotype was M184V alone in the abacavir/ lamivudine/zidovudine group (median vRNA 1-2 log,10 below baseline), and M184V with or without TAMs in patients originally assigned to lamivudine/zidovudine. At week 48, phenotypic results were obtained for 11 isolates for patients from both arms, and all had reduced susceptibility to lamivudine but all remained sensitive to stavudine, all protease inhibitors and all non-nucleoside reverse transcriptase inhibitors. Three, three and two isolates had reduced susceptibility to abacavir, didanosine and zidovudine, respectively. CONCLUSIONS: Abacavir retained efficacy against isolates with the M184V genotype alone. TAMs did not develop during 48 weeks of abacavir/lamivudine/zidovudine therapy and were uncommon when abacavir was added after 16 weeks of lamivudine/zidovudine therapy. Limited mutations upon rebound on this triple nucleoside combination allows for several subsequent treatment options.