ABSTRACT
AIM: This study investigated the incidence of pancreatitis and its association with antiretroviral therapy (ART), focussing on stavudine and didanosine. METHODS: EuroSIDA has collected information on pancreatitis since Summer 2001. All identified cases have been verified by the coordinating centre. Individuals were followed from June 2001 or the date of entry into EuroSIDA (whichever occurred later) until a diagnosis of pancreatitis or the last study visit. Factors associated with pancreatitis were investigated using Poisson regression. Cumulative lengths of exposure to didanosine without stavudine, stavudine without didanosine, stavudine with didanosine, and other ART were time-updated variables. Treatment variables were fitted with a 6-month time lag. RESULTS: There were 43 (nine presumptive) pancreatic events in 9678 individuals during 33 742 person-years (incidence 1.27/1000 person-years). The incidence among those with no, 2 or less and over 2 years' exposure to ART including stavudine and didanosine was 1.24, 1.73 and 0.78/1000 person-years, respectively. In multivariable analysis, higher baseline CD4 cell counts were associated with a decreased risk of pancreatitis. There was no evidence of an association of pancreatitis with cumulative exposure to didanosine and stavudine, didanosine without stavudine, stavudine without didanosine, or other ART. CONCLUSION: We observed a low overall rate of pancreatitis in the years 2001-2006, and did not find an association of an increased incidence of pancreatitis with cumulative exposure to antiretroviral agents generally, and to didanosine and stavudine in particular.