P-325A Phase II trial using a combination of Oxaliplatin, Capecitabine, and Celecoxib with concurrent radiation for newly diagnosed resectable rectal cancer: COX2 expression in relationship to outcomes

Introduction: Improved survival is noted among patients with rectal cancer that achieve pathologic complete response (PCR) after neoadjuvant therapy. The relationship of COX2 expression and outcomes after neoadjuvant therapy is not well defined. In animal models, celecoxib showed increased cytotoxic effects of radiation. A phase II trial combining capecitabine, oxaliplatin and celecoxib with concurrent radiation was conducted to primarily assess the PCR and secondarily determine the surgical down staging (SD) rate, sphincter sparing surgery (SSS) rate, toxicity as well as progression free and overall survival rates. A correlative analysis determined the relationship of these outcomes with COX2 expression in tumor samples at diagnosis. Methods: A total of 37 out of 55 planned patients were enrolled by NM cancer center consortium from 2005 to 2012. Inclusion criteria were: Resectable adenocarcinoma of the rectum within 12 cm of the anal verge, biopsy proven T3-4N1-2M0 based on endoscopic U/S, performance status of 0-2, adequate bone marrow reserve and liver functions. The neoadjuvant chemoradiation treatment was: capecitabine 850 mg/m2 bid Monday through Friday for 5 weeks, weekly oxaliplatin 50 mg/m2 IV, celecoxib 200 mg bid with concurrent 45 gray radiation therapy in 25 fractions over 5 weeks. Evaluable patients had completed at least 2 weeks of the study drugs. The intention-to-treat analysis was used to present the results. Results: Among the 37 enrolled patients, 32 were available for evaluation. Overall the mean age was 52.7 years (range 20-70y), BMI 27.1kg/m2, 59% were Hispanics. Median follow up was 42 months. Overall the PCR was 31% (95%CI: 16-50). Secondary outcomes included: SD 75% (95%CI: 57-89), SSS 56% (95%CI: 38-74). The most common grade 3 and 4 toxicity were diarrhea (9%), abnormal liver function tests (9%), and abdominal pain (6%). Two patients died from severe infection. Relevant absent or grade 1 toxicities were radiation dermatitis (59% and 34%) and rectal mucositis (63% and 31%). PFS was 84% (95%CI: 65-93) and OS at 3 years was 94% (95%CI: 77-98). The relationship to COX2 expression of moderate/high expression versus minimal/low among PCR was 36% vs 25% (p = 0.12), SD 86% vs 100% (p = 1.00) and SSS 75% vs 21% (p = 0.16). Absence of radiation skin toxicity occurred in 69% of moderate high vs 43% minimal/low COX2 expression. Conclusion: This regimen combination showed high rates for PCR, SD and SSS. The PFS and OS are encouraging among this population. Patients that achieved PCR, and minimal skin toxicity were more likely to express moderate/high COX2, although not statistically significant. The use of Celecoxib may increase cytotoxicity by a postulated decreased tumor angiogenesis or by enhancing radiation activity. Improved skin toxicity may be related to anti-inflammatory properties of this drug.