Impact of carrier solutions on pharmacokinetics of intraperitoneal chemotherapy.

Abstract

In the treatment of gastrointestinal malignancies with dissemination to peritoneal surfaces the principal advantage of intraperitoneal chemotherapy over intravenous chemotherapy is the high drug concentration achieved locally with low systemic toxicity. This advantage can be optimized by maintaining a large area of contact between the chemotherapy solution and the surfaces within the abdomen and pelvis over a prolonged time period. Using a rat model we compared the pharmacokinetics of two drugs infused intraperitoneally, 5-fluorouracil and gemcitabine, in five different carrier solutions. A total of 120 Sprague Dawley rats were randomized into groups according to the carrier solution and the drug administered. Rats were given a single dose of intraperitoneal 5-fluorouracil (20 mg/kg) or gemcitabine (12.5 mg/kg) in 0.1 ml/g body weight of each carrier solution. The carrier solutions used varied in their tonicity (0.3%, 0.9% or 3% sodium chloride), or were isotonic and varied in molecular weight (0.9% sodium chloride, 4% icodextrin and 6% hetastarch). With the hypotonic, isotonic and hypertonic sodium chloride solutions, only 5-fluorouracil was used. Each group was further randomized according to the intraperitoneal dwell period (1, 3 or 6 h). At the end of the procedure the rats were killed, the peritoneal fluid was withdrawn completely and the blood was sampled using a standardized protocol. The volume of the peritoneal fluid was recorded, and the drug concentrations in the peritoneal fluid and plasma were determined by high-performance liquid chromatography. Measurements of peritoneal fluid volume showed a more rapid clearance of hypotonic and isotonic sodium chloride solutions from the peritoneal cavity as compared to hypertonic sodium chloride and high molecular weight solutions. When comparing the remaining intraperitoneal volumes at 6 h, the differences were statistically significant for both 5-fluorouracil and gemcitabine when hetastarch (P < 0.0001 and P = 0.0004) and icodextrin (P = 0.002 and 0.008) were compared with isotonic sodium chloride solution. Similarly, there was a significant difference in the volumes recorded at 6 h when hypotonic (P < 0.0001) and isotonic sodium chloride solutions (P = 0.0002) were compared with hypertonic sodium chloride solution. The concentrations of chemotherapy in the different carrier solutions varied little. The total amount of drug in the peritoneal cavity decreased with all solutions and more quickly with 5-fluorouracil than with gemcitabine. There was a significant difference in the total intraperitoneal 5-fluorouracil between hypotonic and isotonic sodium chloride solutions at 1 h (P = 0.0003) and 3 h (P = 0.0043), as well as between the isotonic and hypertonic sodium chloride solutions at 1 h (P = 0.03) and 3 h (P < 0.0001). Similarly, there was a significant difference in the total peritoneal gemcitabine at 6 h between icodextrin and isotonic sodium chloride solution (P = 0.01) and between hetastarch and isotonic sodium chloride solution (P = 0.05). There were no significant differences in plasma 5-fluorouracil and plasma gemcitabine concentrations obtained with the five solutions. These findings show that the clearance of 5-fluorouracil and gemcitabine from the peritoneal cavity can be significantly modified by varying the tonicity or the molecular weight of the carrier solution. Peritoneal fluid clearance was slower with hypertonic sodium chloride and high molecular weight solutions and this resulted in a reduced clearance of chemotherapy. By using a high molecular weight carrier solution the exposure of intraperitoneal cancer cells to gemcitabine was prolonged and drug availability at the peritoneal surface was increased. Similarly, by using a hypertonic carrier solution the exposure to 5-fluorouracil was prolonged and drug availability at the peritoneal surface was also increased.