1398

A novel, low temperature-sensitive liposome containing doxorubicin (DOX-LTSL; ThermoDox, Celsion, Inc.) has been designed to release drug in the temperature range of 40-42oC when coupled with tumor bed hyperthermia. We previously demonstrated that use of this formulation resulted in improved drug delivery and efficacy in a human xenograft model (Cancer Research 2000;60:6950-7). The current study describes the first phase I trial of the formulation in 21 dogs with solid tumors presenting to two veterinary clinics. Since the normal dog body temperatures exceed those of humans (and hence could lead to immature DOX release from the liposome) the patients’ bodies were cooled prior to and during infusion. DOX-LTSL dose was escalated using a standard phase I design at dose levels of 0.60, 0.7, 0.93 and 1.0 mg/kg. The drug was administered intravenously by 30 minute intravenous infusions every 3 weeks. Tumor beds were subjected to hyperthermia (target temperature 39.5-42 oC) starting immediately post-infusion for 1 hour. Plasma (total DOX) and tumor microdialysate (free DOX) pharmacokinetics were evaluated in the first cycle of therapy in all patients, and in subsequent cycles in some. Samples were analyzed by HPLC with laser-based fluorescence detection. Pharmacokinetic parameters were generated using a standard, two-stage approach. Dose-limiting toxicity (myelosuppression) was reached at the 1 mg/kg level. Plasma DOX concentrations were typically above the assay sensitivity until the 24 hr post-dose sample. Mean systemic DOX clearance was 15.24 L/hr/kg, varied between patients over a wide range, and was not associated with dose level. One explanation for the variability is likely due to differences in the dogs’ actual body temperatures (35-39oC), as an inverse correlation was evident between the plasma DOX exposure and the rectal body temperature measured at the end of infusion. (Higher body temperatures would result in additional free DOX and thus a lower AUC due to increased elimination and extravascular distribution.) No evidence of induction or inhibition of DOX systemic clearance was evident in the patients assessed for up to 3 cycles of therapy. DOX was typically first detectable (> 50 pg/mL) in tumor microdialysate samples by the end of infusion, maximal within 90 minutes post-infusion, and persisted for the microdialysis sampling period (120 minutes). Maximal tumor microdialysate concentrations were up to 18 fold above the assay sensitivity. Between-patient variability appeared higher than that observed within an individual patient’s tumor (2 sites evaluated per patient). These in vivo data suggest the LTSL formulation when combined with tumor hyperthermia, provides a unique approach to enhancing tumor selectivity for a cytotoxic drug. Supported by NIH grant: CA42745

[Proc Amer Assoc Cancer Res, Volume 46, 2005]