The effect of dose escalation for large squamous cell carcinomas of the anal canal: A National Cancer Database study.

Abstract

791 Background: Chemoradiation allows for organ preservation in patients with anal cancer, but patients with large tumors (T3/T4) continue to have high rates of locoregional recurrence. With conformal radiation techniques, there is interest in dose escalation to improve local control for large tumors. Methods: The National Cancer Database (NCDB) was used to identify patients with anal cancer from 2004-2013 with tumors > 5 cm in size. Adult patients with T3 or T4 squamous cell carcinoma who received definitive chemoradiation were included. Patients with prior resection were excluded. Higher dose was defined as > than or equal to 5940 Gy. Statistical analyses were performed using logistic regression, Kaplan-Meier, and Cox proportional hazards for overall survival (OS). Results: In total, 1349 patients were analyzed with 412 (30.5%) receiving higher dose radiation therapy (RT). Dose in the higher group ranged from 5940 – 7000 Gy. 5-year OS was 58% and 60% for higher and lower dose RT, respectively. On univariate analysis, higher dose RT (HR 0.998, CI 0.805 - 1.239, p = 0.9887) was not associated with a change in OS but older age (HR 1.484, CI 1.193 - 1.844, p = 0.0004), male sex (HR 1.660, CI 1.355 - 2.033, p < 0.0001), comorbidities (HR 1.496, CI 1.183 - 1.893, p = 0.0008), and long RT (HR 1.248, CI 1.016 - 1.533, p = 0.0347) were significantly associated with decreased OS. The results of multivariate analysis are shown in the Table. Conclusions: There was no observed difference in OS for dose escalation of anal cancers > 5 cm in this population-based analysis, but differences in local control cannot be assessed through the NCDB. Males, elderly, and comorbid patients were particularly high-risk populations with poor chemoradiation survival outcomes. Reducing treatment breaks is important for improving outcomes. Whether dose escalation of large tumors may improve local control and colostomy-free survival remains an important question and is the subject of ongoing trials. [Table: see text]