Rhein is a non parametric estimate technique, power calculation is not possible. However, assuming exponential distribution, a null hypothesis of PFS 6.2 months, and alternative hypothesis of PFS 10.6 months, a two sided test will have a 90% power to detect this diVerence if 45 patients were enrolled in the study. To be conservative, 50 patients were enrolled to ensure adequate power. Results Fifty patients from Princess Margaret Hospital in Toronto, Canada, were enrolled over 18 months, from December 2006 to June 2008. The median age was 58, and the majority of patients were ECOG 0. Thirty three had a colon primary, 16 had a rectal primary, and 1 had a colorectal cancer not otherwise speciWed. Seven patients received prior adjuvant chemotherapy. The median number of treatment cycles administered was 12, and the median follow up time was 13 months.Adverse custom peptide synthesis events are listed in Table 2. The most common adverse events included fatigue, handfoot syndrome, and diarrhea. The most common grade 3 or greater adverse events included hand foot syndrome, neutropenia, and diarrhea. The only grade 4 toxicity was grade 4 neutropenia which occurred in 4 patients.
There was one death from a case of small intestinal obstruction in which the patient declined surgical intervention. Dose modiWcations were common, with 56% of patients requiring dose reductions of capecitabine, and 36% of patients requiring dose reductions of irinotecan, Anastrozole within the Wrst 6 months of treatment. The dose of bevacizumab was modiWed in 36% of patients due to weight changes of greater than 10%. Discussion There have been signiWcant improvements in outcomes for mCRC with the incorporation of multiple new active agents. Several studies have demonstrated the beneWt of adding bevacizumab to 5 FU based chemotherapy, but there are limited data on the use of bevacizumab in combination with capecitabine and irinotecan. In our trial, a regimen consisting of dose modiWed Nattokinase capecitabine and irinotecan with bevacizumab showed promising activity and was well tolerated.
Although results from initially phase II studies were promising, two subsequent randomized phase III trials raised signiWcant concern about the toxicity of the capecitabine and irinotecan combination. The BICC C study is a phase III study comparing FOLFIRI, mIFL, and CapeIRI. The CapeIRI arm is associated with signiWcant toxicity, including a rate of grade 3/4 diarrhea of 47.5% and grade 3/4 neutropenia of 31.9%. The EORTC 40015 study compared FOLFIRI with CAPIRI with or without celecoxib and the study was closed early due to higher than expected patient deaths, especially on the CAPIRI arm. In the current study, a dose capecitabine XELIRI regimen was used, with the dose of irinotecan reduced from 250 mg/ m2 to 200 mg/m2 for all patients and the capecitabine dose reduced from 1,000 mg/m2 to 750 mg/m2 twice daily in patients over the age of 65. Two other studies of XELIRI and bevacizumab have been recently published or presented, and they also utilized lower doses of irinotecan. Garcia Alfonso and colleagues performed a phase II study the combination of biweekly capecitabine and irinotecan with bevacizumab. This regimen was relatively well tolerated with the most frequentgrade 3/4 adverse events being diarrhea, asthenia, vomiting, and nausea.