Clinical correlates of in vitro HIV-1 resistance ot zidovudine. Results of the Multicentre Canadian AZT Trial.

ABSTRACT

OBJECTIVE: To describe the rate of development of in vitro HIV resistance to zidovudine (ZDV) and its prognostic implications within the Multicentre Canadian AZT Trial (MCAT). METHODS: HIV-infected subjects in Centers for Disease Control (CDC) stages IIB, III and IVC-2 with CD4 cell counts > 270 x 10(6)/l were treated with ZDV as part of a dose-range study. Participating volunteers underwent prospective clinical and laboratory evaluations at regular intervals. Viral cultures and sensitivity testing were performed every 12 weeks in a predefined subset of 50 volunteers. An isolate was designated ZDV-resistant if it had a median inhibitory concentration (IC50) for ZDV at least 50-fold higher than that of virus isolated from the same subject before initiation of antiviral chemotherapy. The relationship between resistance and subsequent disease progression was studied using the Mantel and Byar method, for which, at each instance of disease progression, 2 x 2 tables classifying progression versus resistance status were constructed. The observed number of progressions was compared with that expected under the null hypothesis using Mantel-Haenszel methods adjusted for baseline CD4CD8 ratio. RESULTS: The Kaplan-Meier estimate for the cumulative development of in vitro resistance was 64% [95% confidence interval (CI), 41-78] at 180 weeks. Baseline CD4CD8 ratio was negatively associated (P = 0.10) with the subsequent development of resistance (proportional hazard, 0.44; 95% CI, 0.17-1.10). After adjusting for baseline CD4CD8 ratio, the numbers of observed and expected progressions following the development of resistance were 15 and 7.6, respectively (P = 0.008). A similar relative risk of progression between resistant and non-resistant states was found in the two CD4CD8 strata; observed and expected progressions were 4 and 2.3 and 11 and 5.2 in the high and low CD4CD8 strata, respectively. CONCLUSIONS: In vitro resistance to ZDV developed in 64% of subjects after 180 weeks of ZDV therapy. Lower CD4CD8 ratio at baseline was associated with faster development of resistance. In addition, the development of resistance was found to be a marker of subsequent disease progression. This association persisted after adjustment for baseline CD4CD8 ratio. Whether in vitro resistance to ZDV is merely a surrogate marker or a determinant of disease progression remains to be established.