An Analysis of Primary Site Radiation, Surgery, and Survival in Metastatic Rectal Cancer: A National Cancer Database Population Study

Abstract

Optimal management of the primary site in metastatic rectal cancer remains undefined. This analysis evaluates patterns of care, primary site treatment, outcomes, and factors associated with survival in metastatic rectal cancer. A National Cancer Database (NCDB) inquiry was performed to identify patients diagnosed with metastatic rectal cancer between 2004 and 2013 who received treatment coded as non-palliative to the primary cancer site. This included either radiation therapy (RT) alone or a combination of RT with primary site surgery. Patients with squamous histology, unknown or palliative surgery, unknown RT dose, or no pelvic RT were excluded. Kaplan Meier method was used to estimate overall survival (OS). Univariate and multivariate analyses (MVA) were conducted to identify associated risk factors. In total, 4,461 patients were analyzed. The majority of patients (79%) were under the age of 70 with a Charlson-Deyo score of 0 (84%). Clinical stage was T1-T2 (7%), T3 (50%), T4 (16%), Tx (27%), N0 (27%), N1 (37%), N2 (13%), Nx (23%). Grade was 1-2 (63%), 3-4 (16.3%), and unknown (20.5%). 93% received chemotherapy. A total of 2,163 of patients received non-palliative RT to the pelvis alone, and 2,298 patients received RT and surgery coded as non-palliative. Patients who had both RT and surgery had superior OS compared to patients who received RT alone (HR 2.4, p<0.001). For patients who received RT and surgery, 92% of patients received 45-60 Gy, 5% received >25 Gy but <45 Gy, and 3% received 25 Gy (3%). There was no difference in OS comparing doses of 25 Gy vs. ≥45 Gy, with five year OS of 42% and 34.2%, respectively (HR 1.1, p=0.65). Patients receiving doses between 25 and 45 Gy with surgery had worse OS (HR 1.9 <0.001). On MVA, the following additional variables were associated with worse OS: male sex (HR 1.2, p=0.005), N2 disease (HR 1.2, p=0.025), poorly differentiated histology (HR 1.7, p<0.001), and treatment at a non-academic center (1.3, p<0.001). Receipt of chemotherapy was associated with improved OS (HR 0.6, p=0.009). Patients with multi-organ metastases had worse OS relative to patients with single organ metastases, or other distributions such as multiple distant lymph node regions, or metastases to one distant organ and one distant lymph node region (HR 1.7, p=0.02). As of 2013, long course radiotherapy (≥45 Gy) is the most frequently delivered dose schedule. Survival amongst patients receiving long and short course (25 Gy) was not statistically different. Long-term survival is attainable in metastatic rectal cancer patients who undergo aggressive multimodality management of the primary site. Future prospective studies are warranted to evaluate the role of aggressive primary site management and optimal radiation doses in patients with metastatic rectal cancer.