324   Virological and Clinical Outcome of NNRTI-Containing Regimens for 1932 Patients in EuroSIDA.

A. N. Phillips*, C. Pradier, A. Lazzarin, B. Clotet, F. Goebel, P. Hermans, F. Antunes, B. Ledergerber, O. Kirk, and J. Lundgren.
Royal Free and Univ. Coll. London, UK and Hvidovre Hosp., Copenhagen, Denmark.

Background:Few data are available which relate use of NNRTI-containing regimens in clinical practice with virological and clinical outcome.

Methods:We identified all patients in EuroSIDA (a pan-European clinic-based prospective study) who began a regimen including either nevirapine or efavirenz (not both) after July 1997 (when efavirenz was first used) and for whom pre-therapy and maximum viral load, together with pre-therapy and nadir CD4 count, were known. In all cases this was the first use of an NNRTI. Virological failure was defined as two consecutive values >500 cps/ mL after starting the regimen (>6 months after start of the regimen if baseline viral load was >500 cps/mL).

Results:1202 patients initiated nevirapine and 730 efavirenz. Respectively, median start dates were Aug. 98 & Feb. 99; median CD4 count baseline, 255 & 251/mm³, nadir 120 & 100/mm³; viral load baseline 4.0 & 3.7 cps/ mL, max ever 4.8 & 4.9 cps/mL; mean no. of previous PIs, 1.4 & 1.8, previous nucleosides, 3.3 & 3.5; mean no. of PIs in regimen, 1.3 & 1.3, no. of nucleosides in regimen, 2.4 & 2.3; previous AIDS, 35% & 42%. 525 patients experienced virological failure during follow-up. In a Cox model, less PIs and nucleosides previously used (p = 0.0001 and p = 0.009, respectively), higher CD4 nadir (p = 0.002), lower viral load baseline (p = 0.0005) and no previous AIDS diseases (p = 0.006) were associated with lower risk of virological failure. Those on efavirenz (vs. nevirapine) experienced a relative hazard (RH) of 0.58 (95% CI 0.46—0.73; p < 0.0001). 123 patients experienced new AIDS or death during follow-up. After similar adjustment, the RH for efavirenz (vs. nevirapine) was 0.52 (95% CI 0.33—0.84; p = 0.007).

Conclusions:The difference in outcome between those using nevirapine and efavirenz could reflect differences in efficacy of the drugs, but, despite the similarity between groups at baseline, bias cannot be completely excluded as an explanation. Replication of these findings in randomized trials and other cohort studies is required.