Maintenance immunosuppression use and the associated risk of avascular necrosis after kidney transplantation in the United States

Background. Avascular necrosis (AVN) after renal transplantation has been largely attributed to the use of corticosteroids. However, other risk factors such as microvascular thrombosis and hyperlipidemia have been well described and may be of increased importance in the era of early steroid cessation and avoidance. We hypothesized that maintenance immunosuppressive medications known to be associated with these risk factors for AVN would also be associated with a higher risk of AVN. Methods. By using the U.S. Renal Data System database, we studied 27,772 primary patients on Medicare who received a solitary kidney transplant between January 1, 1996, and July 31, 2000. Cox proportional hazards regression models were used to calculate adjusted hazard ratios (AHRs) for patient- and transplant-related factors (including allograft rejection) with Medicare claims for AVN. The intensity and duration of corticosteroid use could not be assessed. Results. Among patients who were prescribed sirolimus at discharge, 3.5% of patients who received the combination of sirolimus-cyclosporine A (CsA) demonstrated AVN, compared with 1.4% of patients who received the combination of siroli-mus-tacrolimus (P=0.06 by chi(2)). In Cox regression, CsA use (vs. tacrolimus) (AHR 1.36, 95% confidence interval, 1.09-1.71) was independently associated with an increased risk of AVN. Sirolimus use showed a trend toward significance (AHR 1.59, 95% confidence interval, 0.99-2.56), with no significant interaction with CsA. Conclusions. Compared with other maintenance immunosuppression, AVN was significantly more common after use of CsA prescribed at the time of discharge for renal transplantation. Whether this increased risk of AVN was directly attributable to hyperlipidemia, microvascular thrombosis, or differences in corticosteroid dosing could not be determined.