Chemotherapy delivery in early-onset colorectal cancer is impacted by urban versus rural settings in Colorado.

Abstract

19 Background: While colorectal cancer (CRC) incidence has been decreasing overall, incidence in adults under age 50 has been rising both nationally and across Colorado. National guidelines have adapted to recognize this trend, but knowledge gaps regarding early-onset CRC remain. This group has previously established rising early-onset CRC incidence in Colorado, as well as later stage at diagnosis and worse prognosis. This study examines demographics, treatment, survival, and concomitant disease among patients < 50 and diagnosed with CRC across a multicenter healthcare system in Colorado in order to better understand early-onset CRC. Methods: We analyzed 1,192 CRC cases in patients < 50 from the Colorado Health Data Compass database and cross-referenced these cases with the Colorado Central Cancer Registry to examine for association of gender, race/ethnicity, BMI, zip code, insurance type, stage, and concomitant medical conditions with overall survival and oncologic treatment modality. Logistic regressions were used to evaluate the relationships between chemotherapy and the important variables while adjusting for covariates, with 95% confidence intervals reported. Cox proportional hazard regressions were used to evaluate the relationships between overall survival and the important variables while adjusting for covariates, with 95% confidence intervals reported. Results: Overall, early-onset CRC in our population was 2.8% stage I, 6.3% stage II, 12.7% stage III, 18.5% stage IV, and 59.8% were unstageable or no stage recorded. Surgical treatment was associated with a 52% improvement in overall survival. Chemotherapy treatment was not associated with any survival changes in stage II disease. Radiation treatment, insurance type, number of concomitant conditions, geographic location, and race/ethnicity were not associated with overall survival. Upon exploring chemotherapy relationships, patients from urban areas were significantly more likely to receive chemotherapy than patients from rural areas, while adjusting for ethnicity, inpatient encounters, and stage (OR 5.55, p = 0.001). Conclusions: While less is known about CRC in patients < 50, trends are emerging that are not as well described in a traditional CRC population. The difference between urban and rural patients and their rates of chemotherapy points to a potential disparity in access to care, particularly when considering advanced stage CRC. This persisted through insurance type, indicating that distance from an infusion center, rather than ability to pay, is more likely to be driving this relationship. Further research is needed to determine if these findings are generalizable to the larger early-onset CRC population, and how access may be affecting care for this population.