Retrospective analysis of renal toxicity in breast cancer patients following zoledronic acid or pamidronate treatment

Abstract

564 Background: In breast cancer (BC) patients (pts), reports suggest that some i.v. bisphosphonates (BPs) are associated with an incremental renal toxicity risk. Few studies have investigated this risk in clinical practice. We retrospectively analyzed rates of renal toxicity in BC pts treated with zoledronic acid (ZO), pamidronate (PA), or no BP treatment (NT). Methods: Pts were identified through medical and pharmacy claims and laboratory data from >35 managed care plans between 01/2000 and 12/2004. Inclusion criteria: female pts with ≥2 claims for BC, no gap in health plan coverage, and serum creatinine measurements available from before and after first ZO or PA infusion. BC pts receiving NT provided a separate reference group. Observations were analyzed from first infusion to health plan disenrollment (BP monotherapy) or before BP switch. Renal impairment was defined as a ≥25% decrease in glomerular filtration rate (GFR,abbreviated MDRD formula) from baseline. Andersen-Gill model, adjusted for demographics and baseline comorbidities, was used to analyzer multiple events in a given pt, as physician practice patterns indicate that administration of both BPs continues even after the development of a first nephrotoxic event. Results: 312 ZO pts, 166 PA pts, and 2,862 NT pts were included. ZA and PA groups had comparable baseline characteristics, renal-related comorbidities, and baseline GFR. Both BP groups were more frequently exposed to chemotherapies and nephrotoxic drugs than the NT group (p<0.0001). ZO and PA groups had similar renal impairment rates that were significantly higher than the NT group and increased with treatment duration (ZO 15.8%, PA 15.6%, NT 7.3% in pts treated for <6 months, increasing to ZO 42.0%, PA 40.0%, NT 16.9% in pts treated for ≥18 months). Mean renal impairment events per person-year were significantly higher for ZO than PA (0.93 vs 0.81, p=0.019). Multiple-event analysis showed a significantly higher renal impairment risk for ZO than PA (hazards ratio=1.4, p<0.0001). Conclusions: In this study on BC pts, renal impairment was more commonly seen in ZO and PA pts than in NT pts. ZO and PA pts have similar renal impairment risk, but ZO pts are more likely to experience multiple impairment events after repeated administrations. [Table: see text]