Patterns of Practice and Improvements in Survival Among Patients With Stage 2/3 Rectal Cancer Treated With Trimodality Therapy

IMPORTANCE: This study quantifies the trends in trimodality therapy use and its association with pathologic stage and overall survival of patients with rectal cancer at the population level. OBJECTIVE: To describe changes between 2006 and 2016 in the sequence and use of chemotherapy/radiation therapy (C/RT), multiagent (MA) chemotherapy, and total neoadjuvant therapy (TNT) for patients with stage 2/3 rectal cancer and identify associations with pathologic stage and survival over time. DESIGN, SETTING, AND PARTICIPANTS: This retrospective cohort analysis included patient records from the National Cancer Database between 2006 and 2016. Of 110 372 patient records, 77 905 were excluded owing to not receiving trimodality therapy and other predefined exclusion criteria. The final analytic cohort comprised 32 467 patients records treated with trimodality therapy, with 24 297 considered in the survival analysis. Data analysis was performed between June 2020 and December 2021. EXPOSURES: Trimodality therapy was defined as including all of the following: definitive surgery; radiation therapy (RT), alone or in combination with chemotherapy; and neoadjuvant/adjuvant single-agent (SA) or multiagent (MA) chemotherapy independent of RT. MAIN OUTCOMES AND MEASURES: Using Cox multivariable survival analyses across demographics, surgery type, stage, year of diagnosis, and facility type, treatment groups were allocated as the following: group A: TNT (n = 8883 [27%]); group B: preoperative C/RT plus postoperative SA chemotherapy (n = 5967 [18%]); group C: preoperative C/RT plus postoperative MA chemotherapy (n = 12 926 [40%]); and group D: postoperative C/RT plus MA chemotherapy (n = 4689 [14%]). RESULTS: The final analytic cohort comprised 32 467 patients (mean [SD] age at diagnosis, 57.6 [11.6] years; 12 549 [38.7%] women and 19 918 [61.3%] men). Comparing 2016 with 2006, treatment shifted to fewer patients receiving postoperative C/RT (group D) (28% vs 8%; P < .001), and more preoperative C/RT and postoperative MA chemotherapy (group C) (24% vs 45%; P < .001) being used. While clinical stage 2 and 3 distribution remained unchanged, pathologic downstaging was observed to stages 0, 1, 2, and 3: 0.60%, 10%, 31%, and 57% vs 2.8%, 22%, 29%, and 45%, from 2006 to 2015, respectively (P < .001). More recent year of diagnosis was associated with an adjusted hazard ratio of 0.77 (95% CI, 0.67-0.87) for mortality within 36 months after diagnosis (2015 vs 2006). CONCLUSIONS AND RELEVANCE: In this cohort study, the shift toward preoperative C/RT and lower pathologic stage was associated with improved overall survival in stage 2/3 rectal cancers.