10011 Background: CRC is primarily a disease of the elderly. The high burden of pre-existing comorbidities alone or in concert with cancer treatment place the older patients with CRC at increased risk of new-onset morbidities, specifically, CVD and CHF. However, the magnitude of risk of new-onset morbidity, and its association with pre-existing comorbidities or treatment remain unknown. Methods: Using SEER-Medicare data, we evaluated individuals diagnosed with incident stage I-III CRC at age ≥66y between 1/1/2000 and 12/31/2011 who had survived ≥2y after diagnosis (n = 57,256; 77% with colon cancer). We compared these to an age, sex-, and race-frequency matched comparison group of non-cancer Medicare patients (n = 104,731). We evaluated new-onset CHF and CVD using competing risk cumulative incidence functions and multivariable Cox regression models. Results: The median age at diagnosis was 77y (66-106y); 45% males; and 85% non-Hispanic white. Median follow-up was 8y (2-14y) from diagnosis of CRC. Treatment included surgery for 99%, chemotherapy for 31%, and radiation for 12%. New-onset morbidity: The 10y cumulative incidence of new-onset CHF and CVD were 43.6% and 58.9%, respectively. After controlling for pre-cancer comorbidities, CRC survivors were at increased risk of new-onset CHF (HR 1.29) and CVD (HR 1.74) (all p < 0.001) compared to controls. Patients receiving radiation (HR 1.29) or 5-FU+oxaliplatin (HR 1.09) were at increased risk of CVD compared to those without those therapies (p < 0.001). Pre-existing diabetes (HR 1.16) and CHF (HR 1.21) independently increased the risk of CVD (p < 0.001). While 5FU+oxaliplatin did not increase the risk of CHF independently (HR 0.97), diabetic patients treated with 5-FU+oxaliplatin were at 1.71-fold increased risk of developing CHF (p < 0.001) when compared with those without pre-existing diabetes. Conclusions: Older CRC survivors are at increased of developing CHF and CVD. Monitoring survivors with a history of exposure to 5FU+oxaliplatin or radiation, and improving management of pre-existing comorbidities may reduce the burden of long-term morbidity for older CRC survivors.
Inference for cumulative incidence functions with informatively coarsened discrete event-time data
