Abstract

Introduction: The proteasome inhibitor bortezomib (VELCADE®) has significant activity against multiple myeloma (MM), and triggers generation of reactive oxygen species (ROS), which are critical in initiating bortezomib-mediated apoptosis. Since ciprofloxacin (CPX) suppresses ROS generation, and is often used as a prophylactic or therapeutic antibiotic, we considered the possibility that CPX might antagonize bortezomib’s anti-tumor efficacy.

Methods: The impact of CPX on bortezomib-mediated effects was evaluated in multiple myeloma cell line models. Additionally, a retrospective analysis was performed of the impact of CPX on clinical outcomes in the phase III APEX trial, which randomized patients with relapsed or refractory MM to receive either bortezomib or dexamethasone.

Results: When CPX was added to bortezomib, it suppressed the proteasome inhibitor’s anti-proliferative effect on RPMI 8226 cells in a dose-dependent fashion, including at CPX concentrations reflective of peak doses achieved in humans. Isobologram analysis yielded combination indices that were consistent with an antagonistic interaction. Flow cytometric studies revealed that CPX significantly inhibited bortezomib-induced specific apoptosis from 61.9±2.3% to 46.7±3.2%. Western blotting and caspase activity assays showed that CPX inhibited the ability of bortezomib to activate caspases-8, -9, and the common effector caspase-3, and reduced cleavage of poly-(ADP)-ribose polymerase in a dose-dependent manner. Additionally, with increasing doses of CPX, bortezomib-triggered activation of heat shock protein-27 was decreased. Microarray analysis of bortezomib+CPX co-treated cells revealed an at least two-fold up-regulation of Heme Oxygenase-1 (HO-1) compared to cells treated with bortezomib alone. HO-1 catalyzes heme degradation, leading to generation of bilirubin, iron, and carbon monoxide (CO), and since bilirubin exerts strong antioxidant effects and CO has anti-apoptotic and cytoprotective actions, this provides a plausible mechanism of action for CPX. To examine the potential clinical relevance of these findings, we compared the response rates of myeloma patients treated with bortezomib in the absence of a quinolone antibiotic, or with one of these agents as a concomitant medication. Among 315 patients treated with bortezomib on the APEX study, those treated with bortezomib without a quinolone (n=222) had an overall response rate (ORR) of 42% (partial response (PR) 35%, complete response (CR) 8%). In the presence of a quinolone (n=93), however, the ORR was only 29% (PR 26%, CR 3%). Notably, quinolones did not reduce the response rate of patients on this study treated with dexamethasone, an agent whose activity is not primarily dependent on ROS.

Conclusions: Studies are being pursued in additional myeloma cell lines and patient samples, as is in vivo modeling, to further validate this hypothesis and proposed mechanism. These findings do, however, support the presence of a mechanism-based interaction between ciprofloxacin and bortezomib that reduces the latter’s anti-tumor activity, and suggest that concomitant use of these agents be avoided clinically. Moreover, they may have significance for the efficacy of other anti-neoplastics that function, in whole or in part, through ROS generation, against myeloma and other tumors.

Author notes

Disclosure:Employment: Drs. Esseltine and Neuwirth are employees of Millennium Pharmaceuticals, Inc. Ownership Interests:; Drs. Esseltine and Neuwirth, as employees of Millennium Pharmaceuticals, Inc., do receive stock options.

Membership Information: Drs. Richardson and Orlowski serve on an advisory board for Millennium Pharmaceuticals, Inc.