Patterns of Care and Outcomes of Low-Lying Adenocarcinoma and Squamous Cell Carcinoma of the Rectum.

715 Background: To analyze the patterns of care and survival for stage I-III squamous cell (SCC) and adenocarcinoma (AC) of the rectum from 2004-2015 based on receipt of surgery alone (S), neoadjuvant chemoradiation followed by surgery (CRT+S), surgery followed by adjuvant chemoradiation (S+CRT) and definitive chemoradiation (CRT). Methods: The NCDB was explored to identify patients diagnosed with low-lying stage I-III AC and SCC of the rectum from 2004-2015 who received S, CRT+S, S+CRT or CRT. Frequency distributions between the categorical variables were compared using the Chi-square test. Overall survival data and Kaplan-Meier curves were compared via the log-rank test stratifying the groups (AC vs. SCC) by stage and treatment. Cox proportional hazards regression models were performed to assess for covariates that had an impact on survival. Results: There were 22,227 patients included in this study, from which 21,587 (97.1%) were AC and 640 (2.9%) were SCC. Median follow-up for the entire cohort was 54 months. The median RT dose for AC was 5040 cGy and for SCC 5400 cGy. 5-year OS of AC vs. SCC in the entire cohort was 61.6% vs. 56.1%, respectively (p<0.001). The OS for AC was 51.5% for S alone, 71.1% for CRT+S, 65.6% for S+CRT, and 36.2% for CRT (p<0.001). The OS for SCC was 49.2% for surgery alone, 54.2% for CRT+S, 67.1% for S+CRT, and 65.7% for CRT (p=0.227). On multivariable Cox regression for AC, patients who received CRT+S (HR 0.61, p<0.001) or S+CRT (HR 0.67, p<0.001) had improved survival compared to S alone while those who had definitive CRT (HR 1.55, p<0.001) had worse survival. Conclusions: This is the largest case series of rectal SCC in the literature to date. Analysis suggests that SCCs of the rectum are more often treated with definitive CRT but are associated with slightly worse survival than adenocarcinoma, suggesting that perhaps surgery may need an increased utilization in this disease. [Table: see text]