Optimizing interstitial delivery of BCNU from controlled release polymers for the treatment of brain tumors

 Two approaches for improving the interstitial administration of carmustine (BCNU) using 3.8% loaded poly(carboxyphenoxypropane-sebacic acid), an implantable biodegradable anhydride which significantly prolongs survival in patients with recurrent malignant gliomas, were evaluated. First, increasing the ratio of carboxyphenoxypropane (CPP) to sebacic acid (SA) in the polymer increases its hydrolytic stability, thus prolonging its half-life in vivo, and extending the period of drug release. A second approach is to increase the dose of drug loaded into the polymer. This study evaluated the relative merits of these two approaches by comparing release kinetics, safety, and efficacy of escalating BCNU doses in polymers with 20:80 and 50:50 ratios of CPP to SA. At the highest dose tested, the 50:50 polymer released BCNU 2.5 times as long in vitro as the 20:80 polymer. Both formulations were nontoxic in rat brains for all BCNU doses tested except 32%. The 20:80 and 50:50 polymers were equally effective in the rat intracranial 9L-glioma model. A dose-response relationship for BCNU was observed (hazard ratio 0.8354 for each mg/kg increase, P<0.001). The two highest loading doses of BCNU improved survival 40-fold (P<0.001). The 20% BCNU-loaded 20:80 polymer achieved the best balance of toxicity and antitumor efficacy, yielding a 75% long-term survival rate. Further evaluation of this polymer in monkeys suggests that it might be used with acceptable toxicity. This study establishes that a dose-escalation strategy for improving BCNU controlled-release polymers is more effective than adjusting the ratio of CPP to SA to prolong drug release. Received: 21 February 1996/Accepted: 14 August 1996