Eric Francois

3545 Background: Sorafenib and irinotecan combination (NEXIRI) showed promising efficacy with a 65% disease control rate (DCR) in pretreated mutated (mt) KRAS mCRC. In this single arm study, CCND1 rs9344 A/A polymorphism had a predictive value (Samalin et al. 2014).This multicentre randomized phase II trial aimed to determine the 2-month progression-free survival rate (2-PFS) of NEXIRI vsIri or Soraf monotherapies in these patients (pts) after failure of all approved active drugs at the time of the study. Methods: Pts PS≤1 with progressive non-resectable mtKRAS (then RAS) mCRC pretreated with irinotecan, oxaliplatin, fluoropyrimidines and bevacizumab (none with regorafenib), were randomized in 3 arms: NEXIRI (biweekly Iri IV 120, 150, 180mg/m² at C3 combined with a fixed dose of Soraf 400mg twice daily) vs Iri (180mg/m²) alone vs Soraf alone, until progression or toxicity, with cross-over to NEXIRI at progression. Primary endpoint was the 2-PFS (RECIST v1.1). Pharmacokinetic, pharmacogenetics and pathologic translational studies were undertaken. Results: We included 173 pts (age 62 [31-82]; PS 0/1: 38/61%) between 2012 and 2014 in 17 French centres. Main results were (median follow-up 17.5 months): See table. Conclusions: We confirmed the NEXIRI regimen efficacy in a randomized study for refractory mtRAS mCRC pts. CCND1 rs9344 may identify patients who benefit from this combination. These results justify comparing NEXIRI to Regorafenib or TAS 102 monotherapies in CCND1 rs9344 A/A pts subgroup. Other results from ancillary studies will be presented at the meeting. Clinical trial information: NCT01715441.

3619 Background: The ACCORD 12 -PRODIGE 2 trial randomly assigned patients to receive radiotherapy (RTH) 45 Gy (CAP-45) with capecitabine or RTH 50 Gy with capecitabine and weekly oxaliplatin(CAPOX-50). The primary objective (complete sterilization of the operative specimen) was not met. This is an update of results at 5 years. Methods: Between November 2005 and July 2008, 598 patients were randomized. Inclusion criteria were: adenocarcinoma of the rectum accessible to digital examination, stage T3 or resectable T4 (or T2 distal anterior rectum) NX M0, < 80 years, PS 0-1. Updated carcinologic results were analyzed using the Kaplan-Meier method and Cox proportional-hazards models. Quality of life was assessed with the anal sphincter conservative treatment questionnaire (ASCT) and by reporting of erectile dysfunction (ED) [satisfaction degree varies from 1(ED) to 7 (normal)]. The evolution of the ED score and selected items from ASCT questionnaire was analyzed with a linear mixed model (LMM). Results: In the ITT population (median follow-up of 5 years), results were in CAP-45 arm and in CAPOX-50 arm, respectively : local recurrence rate 8.8% vs 7.8% (HR for 5y-local RFS 0.92[0.51 – 1.66], p = 0.78), disease free survival (DFS) 60.4% vs 64.7% (p = 0.25, HR = 0.86 [0.66-1,11]), overall survival (OS) 76.4% vs 81.9% (p = 0.056, HR = 0.71[0.50-1 .01]) and specific OS (p = 0.19, HR = 0.75[0.50 – 1.13]). No significant difference was observed in the evolution of the ED score between the two arms (p = 0.29). From ASCT questionnaire, social life disturbance rated from 1 to 4 (greater to no disturbance), with a 5y-median score of 3 (1-4) was not different between the two arms during follow-up (p = 0.15). Moreover, for the global satisfaction score rated from 1 to 7 (unsatisfied to very satisfied), and for the total score, no statistically significant difference was noted (p = 0.43 and p = 0.38 respectively). Conclusions: The quality of life and ED were not different between the two arms after five years of follow-up like the oncologic results as specific OS, DFS and local recurrence. The addition of oxaliplatin to chemoradiation with capecitabin has a minor impact on natural history of the locally advanced rectal cancer. Clinical trial information: NCT00227747.

635 Background: Sorafenib and irinotecan (NEXIRI regimen) showed promising activity with a disease control rate (DCR) of 65% in heavily pretreated mutated (mt) KRAS metastatic colorectal cancer (mCRC) patients in a phase I/II trial (Samalin et al. 2014).This multicentre randomized phase II trial aimed to determine the 2-month progression-free survival rate (2-PFS) of NEXIRI versus irinotecan or sorafenib monotherapy in mtRAS mCRC patients after failure of all approved active drugs at the time of the study. Methods: Patients PS ≤ 1 with progressive measurable and non-resectable mtKRAS (then RAS) mCRC pre-treated with irinotecan, oxaliplatin, fluoropyrimidines and bevacizumab (none regorafenib), were randomized in 3 arms: NEXIRI (irinotecan IV 120 (C1), 150 (C2) and 180mg/m² (C3) if diarrhea grade < 1 in a biweekly regimen combined with a fixed dose of sorafenib, 400mg twice daily) versus irinotecan alone (180mg/m²) versus sorafenib alone until progression or toxicity, with cross over to NEXIRI at progression for the monotherapy arms. The primary endpoint was the 2-PFS (RECIST v1.1). Pharmacokinetic, pharmacogenetics and pathologic translational studies were undertaken. Results: We included 173 patients (median age 62 [31-82]; PS 0/1: 38/61%) between 2012/09 and 2014/07 in 17 French centres. Main results are shown below (median follow-up 17.5 months). Conclusions: We confirmed the NEXIRI regimen efficacy in a randomized study for refractory mtRAS mCRC patients. These results justify comparing this combination to regorafenib or TAS 102 monotherapies in this population. Ancillary studies are ongoing to identify biomarkers. Clinical trial information: NCT01715441. [Table: see text]

490 Background: The ACCORD 12 randomized trial was aiming at improving ypCR by increasing the radiation dose and adding oxaliplatin (OX) to a standard RT 45 Gy with concurrent capecitabine. On short term benefit of OX was not demonstrated (Gérard JP et al. J Clin Oncol. 2010;28:1638-44) at 3 years no significant difference in clinical outcome was achieved (Gérard JP et al. J Clin Oncol. 2012;30:4558-65. François E et al. Radiother Oncol. 2014;110:144-9). This is an update of results at 5 years. Methods: Between 11/2005 and 07/2008, 598 patients were randomized. Inclusion criteria: adenocarcinoma of rectum accessible to digital examination, T3 Nx M0 (or T2 distal anterior rectum). Randomization was between two neoadjuvant treatments chemo radiotherapy: Cap45 (45 Gy + capecitabine) and Capox50 (50 Gy + cape. and oxaliplatin). Main end point was sterilization of the operative specimen. Bowel function for patients treated with anterior resection was analyzed with a global score (1 very poor to 7 excellent). A total of 253 patients received adjuvant chemotherapy usually with FOLFOX: 133 in Cap45, 120 in Capox50. Results: In the ITT population, with a median follow-up of 60 months, updated carcinologic results were calculated using the Kaplan-Meier method. Out of 299 pts randomized in CAP45 and 299 pts in CAPOX50 results were respectively the following 1) 5y-cumulative incidence of Loc. Recurrence : 8.8% vs 7.8% (p = 0.78 HR = 0.92 [0.51-1.66]) 2) 5y-cumulative incidence of dist metastasis : 29.3% vs 27.1% (p = 0.48 HR = 0.89 [0.62-1.54]) 3) 5y-disease free survival rate : 60.4% vs 64.7% (p = 0.25 HR = 0.86 [0.66-1.11]) 4) 5y-overall survival rate : 76.4% vs 81.9% (p = 0.06 HR = 0.71 [0.50-1.01]) 5) bowel function median score : 5.2 [1-7] vs 4.9 [1-7]. Conclusions: At 5 years there was no significant difference in terms of local recurrence or distant metastases rates between both regimens. The new finding is a trend for improved overall survival in the Capox 50 groups may be related to subsequent salvage secondary treatments. At time of meeting the prognostic factors will be available in details and these results put in perspective with other phase III trial testing the role of oxaliplatin in Europe and the US. Clinical trial information: NCT00227747.

Depending on its location or stage, rectal cancer may differ significantly. Before any treatment decision a careful work up is mandatory relying mainly on endoscopy and imaging (MRI). Surgery according to the TME principle is the cornerstone of treatment. Most of the time surgery is associated with external beam radiotherapy often combined with concurrent chemotherapy (capecitabine) according to the neoadjuvant regimen CAP 50 (5 weeks long). It is sometimes possible to escalate safely the dose of irradiation using contact X-ray brachytherapy 50 Kv or Iridium 192 interstitial brachytherapy. Adjuvant chemotherapy may be given in case of pejorative pathological findings but its benefit is not yet proven in contrast with colon cancer. Local recurrences are becoming unusual as is permanent APE surgery with permanent stoma. To reduce the risk of distant metastasis clinical trials are testing first line chemotherapy in T3-4 lesions. For early stage (T2-"small" T3) clinical trials try to achieve organ preservation. Intensification of CAP 50 either with more chemotherapy or radiation dose escalation using contact X-ray aim at achieving a clinical complete response followed by local excision or close surveillance.

676 Background: Organ preservation (avoiding TME surgery) for T2 T3 a-b rectal cancer is a field of active clinical research. Contact X Ray CXB combined with external beam radiotherapy (EBRT) ± concurrent chemotherapy (CRT) is an attractive method to achieve clinical complete response (cCR) and consequently rectal preservation. We report an overview of 120 patients treated with CXB+EBRT over a 25 year period in Lyon since 1986 and then in Nice until 2012. Methods: Between 1986 and 2012, 120 patients presenting rectal adenocarcinoma T2 T3a-b (distal rectum: 87; middle rectum: 33) were treated with CXB +EBRT with conservative intent. In Lyon (1986-2001), 80 patients median age: 73y; T2:52; T3:28) risk were treated using CXB (80-110 Gy/3-4 fr/4-6 weeks) followed by EBRT (39 Gy/13 fr/18 days) and 192 Iridium implant boost (20 Gy). When cCR was achieved, close surveillance was proposed. In Nice (2002-2012), 40 patients (median age 81y; T2:22; T3:18) received CXB same regimen as in Lyon (using new Papillon 50 machine since 2009) + EBRT (45-50 Gy/5weeks) with concurrent chemotherapy (5-FU or Capecitabine). When cCR was achieved close surveillance was proposed or local excision (13 pts). Results: Median follow-up time 58 months in both groups. Local relapse occurred mainly in the 2 first years. Isolated lymph node recurrence

667 Background: Treatment of anal cancer is based on combined radiotherapy and chemotherapy (CT). Some patients present early (2 months) toxicities of grade 3 or greater which can strongly impair quality of life. We aimed to identify predictive factors of toxicities in 106 patients treated in a single center between January 2000 and February 2010. Methods: Retrospective analysis of our databases reported severe local toxicities (grade 3 or higher) during treatment of proven localized epidermoid carcinoma of the anal cancer. The following factors were screened as potential predictive factors: gender, age, irradiation dose, HIV status, type of boost (external vs brachytherapy), circumferential extension, type of chemotherapy, invasion of anal margin, clinical T and N stage, and clinical stage. Results: With a median follow-up of 54.1 months (46.8-61.4) early severe local toxicities occurred in 42 patients (pts) (39.6%) whereas late severe toxicities happened in 21 pts (19.8%). Both early and late toxicities were dominated by proctitis (diarrhea and/or rectal bleeding) (7 pts, 6.6% and 6 pts, 5.6% respectively) and recto-anal epithelial toxicities (27 pts, 25.4% and 17 pts, 16% respectively). Two patients got colostomia because of treatment toxicities (1,8%). Predictive factors of increased early toxicities were as follows: clinical stage III/IV (p=0.04), no brachytherapy boost (p=0.008) and type of CT (no CT, 5.9%, CT, 48.8% p=0.001). Brachytherapy boost and presence of CT retained their independency in multivariate analysis (respectively, p=0.001 and p=0.05). Only HIV positivity (p=0.02) was identified as a predictive factor of late toxicities. HIV positivity (p=0.02), invasion of anal margin (p=0.01) and circumferential extension > 33% (p=0.007) correlated with epithelial ulcer. Invasion of anal margin (p Conclusions: In this cohort absence of brachytherapy boost and CT correlated with more severe early local toxicities whereas HIV positivity was the only predictor of local late toxicities.

550 Background: Preoperative radiochemotherapy (RCT) is the standard of care for patients (pts) with locally advanced rectal adenocarcinoma. However elderly pts may have an increased risk of adverse events after combined modality treatment. The randomized trial ACCORD 12/0405 PRODIGE 2 compared 5 weeks of treatment with radiotherapy 45 Gy/25 fractions (f) with concurrent capecitabine 800 mg/m² twice daily 5 days per week (Cap 45) or radiotherapy 50 Gy/25 f with capecitabine 800 mg/m² twice daily, 5 days per week and oxaliplatin 50 mg/m² once weekly (Capox 50), results of efficacy (complete pathologic response) were not different between the two arms. We analyzed the results of RCT according to pts age. Methods: All eligible pts (n=584) were included in this exploratory analysis. Pts were divided in 2 groups: <70 y and ≥70 y. Toxicity and tumor regression scores were compared between the 2 groups. Results: 442 pts were <70 y and 142 were ≥70 y. Pts characteristics were well balanced between groups (gender, ECOG performance status, primary tumor, histology). Tolerance was worse in pts ≥70 y as shown in the table. Surgical procedures were not different between the 2 groups. Results on histological response were similar between the 2 groups: complete pathologic response was 16.9% (95% CI 13.1 to 20.2%) for pts <70 y and 14.7% (95% CI 9.2 to 21.8%) for pts ≥70 y, (p=0.55) and rates of R0 surgery for pts < 70 y and pts ≥ 70 y were respectively: 90.6% and 88.2%, (p=0.54). Conclusions: As tolerance of elderly pts treated with preoperative RTCT is worse than in younger pts, appropriate therapeutic schedule are warranted for these pts. [Table: see text]