Background: Recycling nucleos(t)ides (NUCs) is useful in regions where new antiretrovirals are not available. This study compares the effectiveness of NUC-containing regimens as rescue therapy in routine care. Methods: Retrospective, multicentre cohort study (January 2001 to June 2006) of patients with ≥1 virological failure who started therapy with 2 NUCs and 1 non-nucleoside reverse transcriptase inhibitor (NNRTI) or a protease inhibitor (PI). The primary endpoint was the rate of treatment response at 6 months (intention-to-treat [ITT] analysis). Results: We included 719 patients (average of 4 prior regimens over a median 6.1 years). The most frequent NUC pairs were tenofovir plus lamivudine (TDF+3TC; 25%), tenofovir plus stavudine (TDF+d4T; 23%), and stavudine plus didanosine (d4T+ddI; 15%). A boosted PI was used in 68% of total cases. Resistance to both NUCs was more frequent in zidovudine plus lamivudine (AZT+3TC; 22.0%), abacavir plus lamivudine (ABC+3TC; 35.5%), and stavudine plus lamivudine (d4T+3TC; 31.2%). No significant differences were observed in treatment response (overall 65%, P = .67); ddI+3TC (71%) and d4T+3TC (53%) had the highest and lowest response rates, respectively. Median time to failure was shorter with d4T+3TC, d4T+ddI, and ABC+3TC (48, 51, and 58 weeks, respectively; P = .0012). Lower response rates associated with an increasing number of thymidine analog mutations (TAMs) were observed for ABC+3TC (P = .027). Conclusion: The clinical utility of NUCs for rescue therapy is limited and selection should be individualized. Specific combinations (d4T+3TC and d4T+ddI) might be less efficacious.
- virological failure