The efficacy of chemotherapy may be affected by circadian rhythms. Therefore, we tested chronomodulated infusion (administered to coincide with relevant circadian rhythms) of oxaliplatin, fluorouracil, and folinic acid compared with a constant-rate infusion method. The combination of three drugs was delivered for 5-day courses with 16-day intervals.

We expected chronotherapy to increase objective response rate by 20% compared with constant-rate infusion. We tested this effect in a randomised multicentre trial involving patients with previously untreated metastases from colorectal cancer who were enrolled at nine institutions in three countries. 93 patients were asssigned chronotherapy and 93 were assigned constant-rate infusion via multichannel programmable ambulatory pumps. The trial was interrupted when a significant difference in main outcome was reached. All data were analysed by intention to treat.

On enrolment, we found significant imbalances in two characteristics—abdominal gland or bone metastases (constant-rate infusion two patients, chronotherapy ten patients) and relapse from surgically treated metastases (constant-rate infusion seven patients, chronotherapy 22 patients). An objective response was obtained in 47 (51%) of the chronotherapy group, and in 27 (29%) of the constant-rate group (difference 21·5% [95% CI 13·7–31·2], p=0·003). Chronotherapy reduced five-fold the rate of severe mucosal toxicity (14% vs 76%, p<0·0001) and halved that of functional impairment from peripheral sensitive neuropathy (16% vs 31%, difference 15·0% [9·5–25·7], p<0·01). Median time to treatment failure was 6·4 months on chronotherapy and 4·9 months on constant-rate infusion (p=0·006), and 24% of the patients from the constant-rate infusion group received chronotherapy after failure. With a minimum follow-up of 3 years, median survival times and 3-year survival were similar in both groups (15·9 vs 16·9 months and 22% vs 21%, respectively).

Chronotherapy was significantly less toxic and more effective than constant-rate infusion. The results support the concept of temporal selectivity of cancer chemotherapy.