Pharmacokinetic strategies to improve the safety and efficacy of intraperitoneal chemotherapy

Ovarian cancer is one of the leading causes of death in women. Although clinical studies have demonstrated that intraperitoneal (IP) chemotherapy provides superior anti-tumor effects compared to intravenous (IV) chemotherapy, IP chemotherapy has not been adopted as the standard, first-line treatment for advanced ovarian cancer. Two well-identified problems limit the clinical success of IP chemotherapy: poor penetration of drug into peritoneal tumors and dose-limiting systemic toxicity. This dissertation attempts to develop and test pharmacokinetic strategies to overcome these problems with the goal of improving the safety and efficacy of IP chemotherapy.;Based on the mechanistic determinants of drug exposure in tumors, we hypothesized that co-administration of antiangiogenic agents with IP chemotherapy will decrease drug removal from tumor, increase drug concentrations in peritoneal tumors, and increase the efficacy of IP chemotherapy. The strategy was evaluated in silico with the use of a hybrid PBPK model developed for topotecan. Simulations predicted that tumor blood flow reductions, as potentially achieved by antiangiogenic therapy, would lead to substantial increases in tumor topotecan concentrations after IP chemotherapy. Pharmacokinetic studies conducted with tumor-bearing mice supported the hypothesis and demonstrated that animals receiving combined IP topotecan and antiangiogenic therapy had ∼6.5-fold higher (P=0.0015) tumor topotecan concentrations compared with animals receiving IP topotecan alone. In subsequent therapeutic studies, the combined administration of an antiangiogenic agent and IP chemotherapy was found to produce superior anti-cancer effects in mouse models of human ovarian cancer, using 4 chemotherapeutic agents (topotecan, cisplatin, paclitaxel and carboplatin), in 3 separate studies. Thus, this work demonstrates that the proposed combination of antiangiogenic therapy and IP chemotherapy significantly improves the therapeutic efficacy of IP chemotherapy.;This dissertation has also evaluated an inverse targeting strategy to minimize the dose limiting systemic toxicity of IP topotecan chemotherapy. The approach employs systemic co-administration of anti-topotecan antibodies, with IP administration of topotecan, to increase the ratio of peritoneal to systemic cytotoxicity. In order to investigate the proposed inverse targeting strategy in silico, a PBPK model of topotecan disposition was constructed. The model was developed based on data generated from IV topotecan doses of 5, 10 and 30 mg/kg, and the model was validated using (i) plasma concentration data following topotecan doses of 1, 1.25, 15, 20 and 80 mg/kg IV and 20 mg/kg IP and (ii) tissue concentration data following 1 and 20 mg/kg IV topotecan doses. To predict the effect of 8C2, a high affinity anti-topotecan monoclonal antibody, on the plasma and tissue PK of topotecan, two mathematical models were developed. Model 1 combined the PBPK model for topotecan with a two compartment model of 8C2 disposition, and model 2 combined a PBPK model for IgG disposition with the PBPK model for topotecan disposition. The predictive performance of both models was tested by comparing simulated data with observed data. The pharmacokinetics of topotecan were subsequently linked to a toxicodynamic model to develop an integrated model of topotecan disposition and toxicity. This model was then used to conduct a series of toxicodynamic simulations to investigate the effect of SC 8C2 administration on the dose-limiting systemic toxicity of topotecan. An in vivo study was then conducted to compare model predictions to observed results. The mathematical model predicted that subcutaneous administration of 8C2 would reduce the extent of topotecan-induced weight loss, at nadir, by 59%, which was quite similar to the observed 50% reduction in nadir weight loss (20+/-8% weight loss for topotecan alone vs. 10+/-8% weight loss for topotecan + 8C2, p=0.1).;Chapter 7 utilizes a mathematical model of antibody-ligand binding and disposition to conduct a series of population simulations and fitting experiments. Analyses were performed to evaluate relationships between data quality and the accuracy and precision of estimation of pharmacokinetic parameters. The results demonstrated that parameter estimation was sensitive to the species of antibody and ligand measured (e.g., bound vs. free vs. total), the number of measured species, and the accuracy and precision of the concentration data. The simulations will be useful in guiding strategies for assay development and pharmacokinetic analyses for immunotoxicotherapies.