Population pharmacokinetics of intraperitoneal irinotecan and SN-38 in patients with peritoneal metastases from colorectal origin.

ABSTRACT

Peritoneal metastases (PM) are common in patients with colorectal cancer. Patients with PM have a poor prognosis, and for those who are not eligible for cytoreductive surgery (CRS) with or without hyperthermic intraperitoneal chemotherapy (HIPEC), palliative chemotherapy is currently the only option. Recently, we conducted a phase I trial (INTERACT) in which irinotecan was administered intraperitoneally (IP) to 18 patients ineligible for CRS-HIPEC. The primary objective was to evaluate covariates influencing the PK profile of irinotecan and SN-38 after IP administration. Secondly, a population PK model was developed to support the further development of IP irinotecan by improving dosing in patients with PM. Patients were treated with IP irinotecan every 2 weeks in combination with systemic FOLFOX-bevacizumab. Irinotecan and SN-38 were measured in plasma (588 samples) and SN-38 was measured in peritoneal fluid (267 samples). Concentration-Time data were log-transformed and analyzed using NONMEM version 7.5 using FOCE+I estimation. An additive error model described the residual error, with inter-individual variability in PK parameters modeled exponentially. The final structural model consisted of five compartments. Weight was identified as a covariate influencing the SN-38 plasma volume of distribution and GGT was found to influence the SN-38 plasma clearance. This population PK model adequately described the irinotecan and SN-38 in plasma after IP administration, with weight and GGT as predictive factors. Irinotecan is converted intraperitoneal to SN-38 by carboxylesterases and the plasma bioavailability of irinotecan is low. This model will be used for the further clinical development of IP irinotecan by providing dosing strategies.