Faculty Bibliography

Bisantrene is a substituted anthracene derivative which preclinically demonstrated a spectrum of activity similar to that of doxorubicin but without associated cardiotoxicity. A Phase I evaluation of the drug has been performed using daily i.v. administrations for 5 days. Sixty courses of treatment were administered to 23 patients at doses from 2.5 to 90 mg/sq m/day. Courses were repeated at 4-week intervals. Dose-limiting toxicities were leukopenia and local cutaneous reactions. The leukopenia was dose related, noncumulative, and of brief duration. Local reactions occurred in 14 of 37 courses administered at doses greater than 60 mg/sq m and in 13 patients resulted in clinical cellulitis of the infused extremity. Gastrointestinal side effects were mild. No alopecia or cardiotoxicity was observed. Two mixed responses were obtained in patients with hypernephromas. Using a daily schedule for 5 days, approximately 40% more drug can be delivered per course than by single-day i.v. administration. However, with this schedule, local cutaneous reactions may prove additionally dose limiting. Phase II studies of Bisantrene in a daily i.v. schedule for 5 days are planned at a dose of 80 mg/sq m/day to be repeated every 4 weeks

Intrahepatic tumor is a major problem in clinical oncology. While direct intravascular infusions provide high local drug concentrations and variable rates of tumor response, they are limited by technical considerations and complications. In this study, we have tested whether high portal venous and hepatic arterial concentrations of 5-fluorouracil (5-FUra) can be achieved by administering drug via peritoneal dialysis. Four patients with metastatic colon carcinoma had a Tenckhoff catheter surgically implanted. During dialysis therapy with 4 mM 5-FUra, simultaneous samples of peritoneal fluid and of portal venous, hepatic venous, and peripheral venous, and arterial blood were obtained, and 5-FUra concentrations were determined. Mean peak portal vein drug concentrations were 60 microM and exceeded the measured concentrations in the other vessels. Total drug exposures as measured by concentration x time (mM x min) during Exchange 1 were: portal, 3.8 +/- 0.65; hepatic vein, 0.97 +/- 0.44; peripheral vein, 0.90 +/- 0.32; and arterial, 1.1 +/- 0.26. During Exchange 7, total drug exposures were: portal, 6.3 +/- 1.4; hepatic vein, 2.5 +/- 1.3; peripheral vein, 2.3 +/- 1.1; and arterial, 2.7 +/- .85. The fraction of i.p. drug that exited the peritoneal cavity through the portal venous system ranged from 0.29 to 1.0. This variation resulted in part from uncertainty in estimating portal blood flow and gastrointestinal drug elimination. Calculated hepatic extraction was 67% (range, 0.23 to 0.89). Extrahepatic metabolism was demonstrated. Measured 5-FUra concentrations compared favorably to values predicted by a pharmacokinetic model for 5-FUra. Dialysis therapy (i.p.) with 5-FUra provides a means of achieving high drug concentrations for treating both i.p. and intrahepatic tumor. Further clinical testing of this route of administration is warranted

A two-compartment physiologic pharmacokinetic model has been developed for 5-fluorouracil (5FU). This model, which incorporates saturable whole body clearance, satisfactorily predicts disappearance kinetics after an intravenous bolus and steady-state levels during constant intravenous infusions. A half-saturating concentration (KM) of 15 microM was determined by comparison of model simulations with literature data. Both hepatic and extrahepatic elimination can be inferred for 5FU, but the exact anatomic or compartmental location of the clearance cannot be determined from the available clinical data. The effect of venous and arterial plasma sampling is discussed. This model has been extended to include intraperitoneal and oral administration of 5FU by the addition of peritoneal fluid and liver compartments

A Phase I study was conducted of 5-fluorouracil administered i.p. in a 2-liter volume of 1.5% Inpersol. The drug was administered via Tenckhoff peritoneal dialysis catheters to ten patients with tumors confined to the i.p. space. Dialysis concentrations ranged from 5 micro M to mM. Complications of the dialysis procedure alone included mild abdominal discomfort and 2 cases of gram-negative bacterial peritonitis, both easily controlled with antibiotics. 5-Fluorouracil caused the same pattern of toxicity as when administered by other routes. There was no local or central nervous system toxicity. Dose-limiting toxicity included pancytopenia and mucositis at a dialysis concentration of 4.5 to 5 mM administered for eight consecutive 4-hr exchanges. There were two documented responses in eight evaluable patients. 5-Fluorouracil concentrations were measured by high-pressure liquid chromatography. Peritoneal fluid concentrations decline in a first-order fashion with a half-life of 1.6 hr. The mean permeability area product was 14 ml/min. A mean of 82% of drug was absorbed in 4 hr. Plasma levels rise over the first 30 to 45 min and decline in a nonlinear fashion. Plasma levels are substantially lower than are peritoneal fluid levels. Mean 4-hr peritoneal fluid concentration was 298 times the simultaneously measured plasma levels. Total body clearance ranged from 0.9 to 15 liters/min and declined with increasing dialysate concentration. We conclude the i.p. route is a relatively safe way to deliver high concentrations and large amounts of drug to the i.p. cavity with a significant pharmacological advantage over conventional routes of administration

We have utilized peritoneal dialysis via semipermanent Tenckhoff catheters to administer methotrexate and 5-fluorouracil to 16 patients with cancer whose tumor was confined to the intraabdominal space. The method of dialysis is described. The therapy was well tolerated by the patients. The major risks were bacterial peritonitis, local drug-induced peritonitis, and the systemic effects of the drugs. Theoretic modeling predicted that far greater concentrations of drug can be safely delivered by this route with less toxicity than when lower concentrations are administered by conventional routes. Furthermore, much higher concentrations should be achieved in the dialysate to which the tumor is exposed than in the systemic plasma. These predictions were confirmed

For patients undergoing intraperitoneal ("belly bath") chemotherapy, it is useful to verify the distribution of infusion fluid. For this reason, 10 patients were examined by computed tomography (CT) after intraperitoneal instillation of dialysate mixed with water soluble contrast material. The extent of intraperitoneal distribution was assessed, and in two patients filling defects, presumably representing tumor, were identified.