Radiotherapy and anatomically defined pelvic second primary malignancies in 10-year rectal cancer survivors

Abstract Second primary malignancy (SPM) associations after rectal cancer may depend on anatomical endpoint definitions. Using Surveillance, Epidemiology, and End Results data, we examined 24,551 patients who survived at least 10 years after surgery for rectal adenocarcinoma diagnosed during 2000–2012. First SPMs diagnosed at least 11 calendar years after the index year were analyzed using competing-risk methods with approximate year-based event times. Radiotherapy was not independently associated with all-site first SPMs. After correcting broad site categories, 277 events met a conservative anatomical pelvic definition excluding prostate cancer. Radiotherapy was associated with this endpoint (adjusted subdistribution hazard ratio 2.36, 95% confidence interval 1.48–3.75). At 10 years after the landmark, unadjusted cumulative incidence was 2.80% with recorded radiotherapy and 1.07% with no/unknown radiotherapy, a difference of 1.73 percentage points (95% confidence interval 1.22–2.25). Including prostate cancer attenuated the association (1.43, 0.99–2.07). These exploratory findings are definition-dependent associations among selected long-term survivors, not evidence of radiation-induced carcinogenesis or a basis for a new screening policy.