- Incidence and Risk Factors for Severe Adverse Events in a Prospective Cohort of HIV-Infected Adults Started on a Protease Inhibitor-Containing Therapy (Oral 56)
Authored by X. Duval, V. Journot, T. May, C. Charlois, A. Waldner, C. Merle de Boever, C. Barennes, J.M. Ragnaud, G. Chêne, C. Leport and the APROCO Study Group
Dr. Leport presented the results of a descriptive study examining the incidence of serious adverse events (SAEs) and their associated risk factors during two years of follow-up on the APROCOC cohort (ANRS -- The French National Agency for AIDS Research). The study enrolled protease inhibitor-naive subjects who began protease inhibitors (PIs) between April 1997 and June 1998.
Data was collected on December 31, 1999, with the follow-up period defined as day 0 through first protease inhibitor interruption. SAEs (grade 3 or 4 events, except lipodystrophy) were prospectively reported and validated by two clinical experts; the likelihood of an SAE resulting from antiretroviral (ARV) treatment was classified as related or unrelated to PI and/or other ARV. A total of 550 SAEs were identified in the 1,155 subjects. Seventy-seven percent of the total cohort was composed of men with a mean age of 38 years who were followed for a median of 23 months. The median CD4 was 288, and the HIV viral load was 4.4 log.
At baseline, 44 percent of the subjects initiated an indinavir (IDV, Crixivan)-containing regimen, 25 percent nelfinavir (NFV, Viracept), 16 percent ritonavir (RTV, Norvir), 11 percent saquinavir (SQV, Invirase, Fortovase) and 4 percent other PI regimens.
During a median follow-up of 23 months (2,037 patient-years), 550 adverse events were reported, 235 events were related to ARV treatment and 169 events were related to PI treatment. Overall, 20 percent of the adverse events occurred within the first month and 51 percent within the first four months after the initiation of therapy.
The most frequent SAEs related to PIs involved the liver (31 percent), the kidneys (18 percent), the hematological system (14 percent), metabolic parameters (11 percent) and cardiovascular system (3 percent). In a multivariate proportional hazard model, a creatinine clearance less than 70 ml/min [HR=2.11, 95 percent confidence interval (CI): 1.21-3.67, P=0.007], an increase of AST (for additional 100 UI/l, HR=1.60, CI: 1.31-1.96, P=0.0001), hepatitis C and/or B coinfection (HR=2.56, CI: 1.76-3.72, P=0.0001), an HIV viral load more than 5 log (100,000 copies/mL) (HR=1.54, 95 percent CI: 1.06-2.25, P=0.03), and indinavir initiation at baseline (HR=1.69, CI: 1.16-2.44, P=0.008) were associated with a higher risk of SAEs. Even if one excluded those with abnormal liver function tests, co-infection with hepatitis was still an independent predictor for developing an SAE.
The investigators concluded that SAEs are frequent after the initiation of PIs -- especially during the first month -- and are related both to host (e.g., if a patient is co-infected with hepatitis) and treatment characteristics. The early occurrence of an SAE may be attributed to either pharmacokinetics (increased drug concentration is responsible for toxic side effects such as nephrolithiasis secondary to indinavir) or hypersensitivity reactions (e.g., transaminitis). The incidence of SAEs clearly plateaus after the first four months, stressing the importance of monitoring closely for such events during the initiation of therapy. I do not believe any of the results reported were surprising -- this study supports most experts' advice to frequently obtain complete blood counts and chemistry panels frequently during the first months of therapy.