Author: Sebag-Montefiore D.; Adams R.; Gollins S.; Samuel L.M.; Glynne-Jones R.; Harte R.; West N.; Quirke P.; Myint A.S.; Bach S.P.; Parsons P.; Falk S.; Dhadda A.S.; Misra V.; Brown N.; Brown G.; Harrison M.; White L.; Duggan M.; Lopes A.
Abstract: Background: Phase II studies reported high pathological complete response (pCR) rates and acceptable toxicity using irinotecan and fluoropyrimidine chemoradiation in LARC (ISRCTN:09351447). Method(s): This phase III, multicentre, open-label trial funded by Cancer Research UK, randomly assigned (1:1) patients with MRI defined LARC threatening or involving resection margins without metastases, to pre-operative radiotherapy (RT) 45Gy/25 fractions combined with either capecitabine 900mg/m2 (CRT) or 650 mg/m2 bd weekdays with Irinotecan iv once-weekly 60mg/m2 weeks 1-4 (IrCRT). The primary endpoint is disease-free survival (DFS). Secondary endpoints include treatment compliance, safety and pCR. Result(s): 75 UK sites randomised 564 eligible patients from Oct/11 to July/18; 284 to CRT and 280 to IrCRT. 370 (66%) male; median age 61 years (range:29-83). Staging in both arms was similar: mrT3 (432/564(77%), mrT4 (89/564(16%); mrCRM involved (275/564(49%); threatened <=1mm (215/564(38%). Compared with CRT, IrCRT patients were less likely to receive 45Gy RT (207/276(75%) vs 251/283(89%), p < 0.001) or receive >=90% capecitabine dose in 188/276(68%) vs 253/283(89.4%)p < 0.001). A total of 204/276(74%) received >=90% irinotecan dose. The grade 3-4 gastrointestinal adverse event rate was 21%(58/276) with IrCRT and 12%(34/283) with CRT (p = 0.004). Patients receiving IrCRT had significantly more diarrhoea 38/276(13.8%) vs 10/283(3.5%)p < 0.001) and neutropenia 27/276(9.8%) vs 3/283 (1.1%) p < 0.001). Two CRT and three IrCRT patients experienced a treatment related death. 237/276(86%) IrCRT and 241/283(85%) CRT patients had surgery. The median time from end of RT to surgery(10.6 weeks), the surgical procedure APE 262/478(55%), AR 189/478(40%), Hartmann's 10/478(2%); and the surgical complications(any event) 38%(181/478) were similar in both arms. The pCR rate is available in > 95% patients and is 20.2% (46/228) for IrCRTvs.17.4%(40/230) for CRT (p = 0.45), A > 84% CRM-ve resection rate is similar in both arms. Conclusion(s): For patients with MRI defined high risk LARC low rates of CRM involvement were observed in both arms reflecting high quality multidisciplinary care. The addition of irinotecan did not significantly improve the pCR rate, was associated with a decrease in the RT and capecitabine compliance and a higher rate of adverse events. Surgical procedure or complications were unaffected. Longer follow-up is required to assess DFS and translational data.
