Timing of adjuvant chemotherapy initiation and colon cancer survival.

Abstract

119 Background: The National Quality Forum endorsed initiation of adjuvant chemotherapy within 120 days of diagnosis for stage III colon cancer patients aged < 80 years. Nevertheless, no trials were conducted to establish this 120-day threshold and observational studies used to justify this threshold may be sensitive to recently identified sources of bias. Therefore, we aimed to assess the effect of initiating adjuvant chemotherapy within 120 days on survival among patients with stage III colon cancer. Methods: We used institutional registry data from the JPS Center for Cancer Care (JPS), an accredited Comprehensive Community Cancer Program. Eligible patients were adults aged < 80 years, diagnosed with first primary stage III colon cancer between 2011 and 2015, and received at least part of their first course treatment at JPS. Overall survival was defined as time from cancer diagnosis to death, loss to follow-up, or end of study. We emulated a pragmatic trial and estimated the intention to treat 36-month restricted mean survival difference and 95% confidence limits (CL) for initiating adjuvant chemotherapy within 120 days using a marginal structural model with stabilized inverse probability of treatment weights to reduce confounding bias. Results: Our study population comprised 62 patients, of whom 61% were aged 55 – 64 years, 58% were females, 61% were racial/ethnic minorities, 69% were uninsured, and 61% initiated adjuvant chemotherapy within 120 days after cancer diagnosis. The mean survival after 36 months of follow-up was 31 months (95% CL: 27, 34) for patients who initiated and 31 months (95% CL: 28, 34) for patients who did not initiate adjuvant chemotherapy within 120 days (mean survival difference = -0.10 months, 95% CL: -3.6, 3.4). Conclusions: Our results suggest no meaningful difference in overall survival between initiating or not initiating adjuvant treatment within 120 days of diagnosis for patients with stage III colon cancer, but our estimates are compatible with either a 3-month survival benefit or harm. A larger sample size may provide greater certainty whether the 120-day threshold is a questionable quality measure, as observed in our study.