Elucidating factors of geographic variation in early-onset colorectal cancer survival among men

Authors: Rogers CR, Moore JX, Henson A, Qeadan F, Huntington MS, Holowatyj AN

Category: Cancer Health Disparities
Conference Year: 2020

Abstract Body:
Background: Despite reductions in overall colorectal cancer (CRC) burden, incidence rates are rising among individuals age<50 years. Previous studies have identified regional differences in early-onset CRC (EO-CRC) incidence by geographic region; however, no studies have examined geographic differences in CRC mortality specifically among patients age younger than 50 years at diagnosis. We aimed to determine the geospatial distribution of EO-CRC, and total variance explained of personal and county level factors with EO-CRC hazard among men. Methods: EO-CRC “hot spots” (counties with high EO-CRC mortality rates) were derived from CDC data, years 1999-2017. EO-CRC hot spots data were linked to NIH/NCI SEER program data among men aged 15-49 years at CRC diagnosis from 1999-2016. Cox proportional hazards models were used to compare CRC-specific survival probability (and hazard) among men in hot spots versus non-significant counties. To estimate the total variance in EO-CRC survival explained by personal and county level factors, we used a generalized R-square. Results: We identified 232 counties as geographic hot spots for EO-CRC—214 (92%) of which were in the South. Among SEER patients, 32,447 men (1,009 in hot spots; 31,438 in non-significant counties) were diagnosed with EO-CRC. Men residing in hot spots had higher hazard of CRC-specific death compared to men in non-significant counties (HR, 1.22; 95% CI, 1.11-1.35) after adjusting for age, race, stage, grade, surgery, and marital status. AJCC stage independently explained 29.5% of EO-CRC survival, and all factors together explained about 35% of the variation in EO-CRC survival. Conclusion: Survival after EO-CRC diagnosis is significantly worse among men residing in hot spot counties, explained mainly by person-level late-stage diagnosis. Further studies of CRC-related health disparities among non-Hispanic Black men diagnosed with EO-CRC are needed and could have significant implications for early detection and care. Examination of individual-level health behaviors and clinical characteristics among men diagnosed with EO-CRC is warranted to explore gene-environment interactions associated with geographic variation in survival and to tailor clinical algorithms for CRC early detection.

Keywords: men’s health; race; SEER; smoking; young-onset