Abstract
The least toxic time (LTT) of irinotecan varied by up to 8 hours according to sex and genetic background in mice. The translational relevance was investigated within a randomized trial dataset, where no LTT stood out significantly in the whole population. 130 male and 63 female eligible patients with metastatic colorectal cancer were randomized to receive chronomodulated Irinotecan with peak delivery rate at 1 of 6 clock hours staggered by 4 hours on day 1, then fixed-time chronomodulated Fluorouracil-Leucovorin-Oxaliplatin for 4 days, q3 weeks. The sex-specific circadian characteristics of grade (G) 3-4 toxicities were mapped with cosinor and time*sex interactions confirmed with Fisher's exact test. Baseline characteristics of male or female patients were similar in the six treatment groups. Main grade 3-4 toxicities over six courses were diarrhea (males vs females, 39.2%; vs 46.0%), neutropenia (15.6% vs 15.0%), fatigue (11.5% vs 15.9%), and anorexia (10.0% vs 7.8%). They were reduced following irinotecan peak delivery in the morning for males, but in the afternoon for females, with statistically significant rhythms (P < .05 from cosinor) and sex*timing interactions (Fisher's exact test, diarrhea, P = .023; neutropenia, P = .015; fatigue, P = .062; anorexia, P = .032). Irinotecan timing was most critical for females, with grades 3-4 ranging from 55.2% of the patients (morning) to 29.4% (afternoon) for diarrhea, and from 25.9% (morning) to 0% (afternoon) for neutropenia. The study results support irinotecan administration in the morning for males and in the afternoon for females, in order to minimize adverse events without impairing efficacy.