Survival in 2367 zidovudine-treated patients according to use of other nucleoside analogue drugs. The EuroSIDA Study Group.

To evaluate survival according to use of different nucleoside drugs in a routine clinical setting, we studied a large group of zidovudine-treated patients seen in clinics across Europe. A total of 3128 subjects was recruited to the observational, prospective EuroSIDA study in May 1994. These were consecutive patients (up to a predefined limit) seen at outpatient clinics in 37 centers from 16 European countries and followed at 6-month intervals by use of standardized forms completed by clinicians at the respective centers. This report concerns 2367 subjects who began antiretroviral therapy with a regime that included zidovudine either before study entry or during the course of follow-up. Cox proportional hazards models were fitted, with use of other antiretroviral drugs, CD4 count, and date of development of AIDS fitted as time-dependent covariates. Survival times from start of therapy were left truncated at study entry to avoid survival bias. In addition to zidovudine, antiretroviral drugs used included didanosine (ddI) (n = 1119; median 1.6 years after starting zidovudine), dideoxycytidine (ddC) (n = 592; median 1.9 years after starting zidovudine), stavudine (d4T) (n = 241; median 2.9 years after starting zidovudine) and lamivudine (3TC) (n = 33 ; median 2.7 years after starting zidovudine). Of the 2367 patients, 613 died during follow-up. Overall, risk of death was reduced in those zidovudine-treated patients who began at least one other nucleoside analogue drug with or after taking zidovudine (relative hazard [RH], 0.61; 95% confidence interval [CI], 0.51-0.72, adjusting for CD4 count, development of AIDS, and age). Fitting each drug separately, there was a larger association with reduced mortality for starting 3TC (RH, 0.41; 95% CI, 0.28-0.62) than for starting ddl (RH, 0.79; 95% CI, 0.67-0.93), ddC (RH, 0.74; 95% CI, 0.59-0.92) or d4T (RH, 0.67; 95% CI, 0.49-0.91). These results suggest that the beneficial effect of nucleoside combination therapy identified in controlled trials can be seen in routine clinical practice.