655   Do Protease Inhibitors Increase the Risk for Coronary Heart Disease Among HIV Positive Patients—Follow-Up Through June 2000.

D. Klein*, L. Hurley, M. Sorel, and S. Sidney.
Kaiser Permanente, Oakland, CA.

Background:It may be too soon to know if elevated lipid levels associated with protease inhibitors (PI) lead to higher rates of coronary heart disease (CHD) among HIV⁺persons. PI use is relatively new, and an increase in CHD events due to high lipids may not be seen in the short-term. This ongoing prospective observational study monitors the rate of CHD events as PI exposure accumulates.

Methods:Hospital events for CHD (ICD9 410—414, primary discharge diagnosis) were identified among 4541 HIV⁺members of the Kaiser Permanente Northern California HMO (cases) and among 41K random age- and sex-matched members not known to be HIV⁺(controls). Follow-up (FU) began 1/1/96 or, for cases, at HIV diagnosis, whichever came later. FU ended at the earliest of member termination or 6/30/00. For cases, person- time was assigned to either non-PI FU (days of FU preceding any PI use) or PI FU (days of FU following first PI use). Cases could contribute person-time to both PI and non-PI FU. Cases and controls with prior CHD events were excluded. Age-adjusted (1990 US census) incidence rates and 95% confidence intervals (CI) were calculated. CHD risk factors were assessed among cases with events.

Results:4541 HIV⁺cases contributed a total of 14,703 person-years (PY) of FU (median 4.3, mean 3.2) and had 53 CHD events—3147 additional PY and 11 new events over what was reported from this study in Feb. 2000. 4,534 cases contributed 7292 PY of non-PI FU (median 1.0, mean 1.6) and had 23 CHD events. 2,837 cases contributed 7412 PY of PI FU (median 2.8, mean 2.6) and had 30 CHD events. Among cases, the adjusted overall incidence rate was 5.5 CHD events per 1000 PY (CI: 3.8, 7.3). The rate was 5.2 (CI: 2.7, 7.7) for non-PI FU and was 5.8 (CI: 3.4, 8.2) for PI FU. The rate among controls was 2.8 (CI: 2.5, 3.1)—about half the rate among cases. Multiple risk factors were present among the cases with events.

Conclusions:While PIs have been widely used for only 4 years, our data continue to suggest that PI use does not increase short-term risk for CHD. The higher rate of CHD events among HIV⁺persons compared to controls is yet to be explained. Risk reduction management is warranted in all patients with multiple CHD risk factors.