Introduction: Neoadjuvant rectal (NAR) score may serve as a surrogate short-term endpoint for overall survival (OS) in clinical trials. This study aims to test the NAR score using a large, national cancer registry.
Methods: National Cancer Database patients with clinical stage II/III rectal adenocarcinoma (RAC) treated with neoadjuvant chemoradiation (CRT) followed by surgery were selected and divided into low-, intermediate-, and high-NAR subgroups. OS outcomes were analyzed using Kaplan-Meier and logistic regression models.
Results: A total of 12 452 patients were selected, of which 5071 (40.7%) were in clinical stage II and 7381 (59.3%) were in clinical stage III; 15.2% had pathologic complete response. The mean NAR score was 10.01 ± 10.61. Six thousand nine hundred and forty-one (55.7%) did not receive adjuvant chemotherapy (AC) and were propensity-matched across NAR subgroups (966 in each group). A significant difference in 5-year OS between low-, intermediate-, and high-NAR groups was observed (85% vs. 76% vs. 68%; p < 0.001). Five thousand five hundred and eleven (44.3%) received AC and 1045 triplets were propensity-matched per NAR groups. A significant difference was again observed for 5-year OS (93% vs. 88% vs. 75%; p < 0.001). Logistic regression confirmed NAR strata as a significant predictor of 5-year OS.
Conclusion: NAR score, as a neoadjuvant response measure, is a strong predictor of 5-year OS, regardless of AC receipt in a heterogenous population of locally advanced RAC patients.
Keywords: NAR score; NCDB; overall survival; rectal adenocarcinoma; surrogate endpoint.
© 2022 Wiley Periodicals LLC.