Trends in primary surgical resection and chemotherapy for metastatic colorectal cancer, 2000-2016.

Abstract

e16000 Background: When, whether and in whom resection of the primary tumor for patients with metastatic colorectal cancer (CRC) is indicated remains incompletely elucidated, although a recent randomized trial (JCOG1007, presented at GI ASCO 2020) showed no survival benefit in resection of asymptomatic primary tumors in this population. Given the emergence of multiagent chemotherapy, surgery may be used less, but patterns of care for metastatic CRC have not been reported. As such, we sought to evaluate trends in use of primary surgical resection and chemotherapy among patients with metastatic CRC. Methods: Patients diagnosed with metastatic CRC between 2000-2016 were identified from the Surveillance, Epidemiology and End Results (SEER) registry. Multivariable logistic regression defined odds of undergoing primary surgical resection, with year of diagnosis as the primary independent variable. The cohort analysis was also stratified by primary site (colon versus rectum), age ( < 50 vs. >50 years) and whether patients also underwent resection of metastatic sites (yes versus no). The secondary endpoint of interest was receipt of any chemotherapy, also assessed by multivariable logistic regression. Results: Among 99,835 patients with metastatic CRC, 55,527 (55.7%) underwent resection of their primary tumor. The odds of undergoing primary surgery decreased with later year of diagnosis, with patients diagnosed in 2016 61.1% less likely to undergo surgery than those diagnosed in 2000 (AOR 0.39, 95% CI 0.36-0.42, p < 0.0001; absolute percent 62.3% to 43.8%). Black patients and women were also less likely to have surgery (p < 0.001). Similar trends by year for primary surgery were observed among each of the subgroups, although patients with colon primary, young adults (age < 50), and patients also undergoing metastatectomy were more likely to undergo primary surgery (p < 0.001 for all). In contrast, the odds of receiving chemotherapy increased dramatically with later year of diagnosis, with patients diagnosed in 2016 221% more likely to receive chemotherapy than those diagnosed in 2000 (AOR 2.21, 95% CI 2.04 to 2.40, p < 0.0001; 45.5% to 64.0%). Conclusions: From 2000-2016, we observed a sharp decline in the rate of primary surgical resection for patients with metastatic CRC, while use of chemotherapy increased over the same period. Prospective studies are needed to define the optimal local treatment for patients with metastatic CRC, perhaps with stratification by molecular and clinical characteristics, in order to optimize both cancer-specific and symptomatic outcomes.