The statistical analyses were performed with a StatView computer program (SAS institute, Abacus concepts incorporation, Berkeley, CA, USA). The effects of bolus 5-FU vs continuous 5-FU regimens, chemoradiation vs chemotherapy only, L. rhamnosus GG vs no dietary supplements, and guar gum vs no guar gum were compared using univariate and multivariate inhibitor Imatinib Mesylate logistic regression models. The results are given as odds ratios with 95% confidence intervals. Mann�CWhitney U-test was used to compare the treatment groups with respect to quantitative response variables. Frequency tables were analysed using the ��2 test. RESULTS Patient characteristics and compliance The treatment arms were balanced with gender, the WHO performance status, primary tumour site, Dukes’ stage, and radiation therapy given (Table 1).
The median age at randomisation was 60 (range: 31�C75); 51% of the participants were men. Sixteen (11%) subjects did not complete the scheduled 6 months of adjuvant chemotherapy due to either adverse events (n=7, six of whom received bolus 5-FU), cancer recurrence (n=5), or a concomitant disease (n=4). Two patients (both in the continuous 5-FU group) who did not receive any of the study treatments due to postoperative complications were not included in safety or efficacy analyses leaving 148 patients for these analyses. None of the patients were lost to follow-up. Table 1 Patient characteristics Chemotherapy dose intensity and tolerability The scheduled 5-FU dose intensity, calculated as the percentage of the scheduled dose as milligrams of 5-FU given per square metre per week of the scheduled cumulative dose, was maintained better among the patients who received the simplified de Gramont regimen than among those treated with the Mayo regimen (median 93 vs 78%, respectively; P<0.
0001). The simplified de Gramont regimen was tolerated better than the Mayo regimen (Table 2). Any grade 3 or 4 adverse effect was present in 87% (65 out of 75) of the patients treated with the Mayo regimen as compared with only 45% (33 out of 73) of those treated with the simplified de Gramont regimen, when the vascular access device (VAD)-related toxicity was included in this analysis (P<0.0001, Table 2). Vascular access device-related complications occurred in seven (10%) patients who received continuous Entinostat 5-FU infusions; three were classified as serious. The Mayo regimen was more frequently associated with stomatitis, diarrhoea, and neutropenia than the simplified de Gramont regimen, whereas mild-to-moderate hand�Cfoot syndrome was more common among the patients who received the simplified de Gramont regimen (P<0.0001). There were no treatment-related deaths.