Search

Gastro-intestinal (GI) side effects are dose limiting for pelvic & abdominal radiation therapy (RT). Advanced-age is hypothesized to be an aggravating factor in severe radiation toxicity, but this has not been well defined. We sought to quantify the effect of age on grade IV / V gastro-intestinal complications of RT by combining information from two US population-based databases. The absolute number of complications was obtained by querying the National Inpatient Database (NIS) of the Healthcare Cost and Utilization Project, which contains discharge data from a 20% stratified sample of all US non-federal hospitals. Inclusion criteria were inpatient episodes between 2015-18, with listed diagnoses of both 'Malignant neoplasm of rectum' (ICD-10-CM Code C20), and 'Gastroenteritis and colitis due to radiation' (ICD-10-CM Code K52.0). A grade IV complication was defined as a hospitalization, alive on discharge; whereas grade V was defined as a hospitalization culminating in death. The relative age-distribution of patients with rectal cancer treated with RT was obtained from the SEER database. Relative toxicity incidence by age group was obtained by normalizing the absolute number of toxic events obtained from NIS with the age-distribution treatment data from SEER. Statistical techniques included best-fit linear regression and multivariate logistic regression. Within the NIS database there were 31,878 hospitalizations with a listed diagnosis of rectal cancer (median age 63 years), of whom 290 had grade IV (median age 66) and 10 grade V (median age 76.5) radiation gastroenteritis / colitis. There were no grade 4 toxicities under the age of 20, and no grade 5 toxicities under the age of 70. After adjusting for the baseline age distribution of rectal cancer patients receiving radiation derived from SEER data, the relative risk (RR) of grade IV toxicity was 0.44 for subjects aged 20-39, and 0.53 for subjects aged 40-59 and 1.0 for 60-79 (the comparator). The RR for grade V toxicity was zero for <60, 1.0 for 60-79 (the comparator), and 1.26 for subjects aged 80-99. Best-fit linear regression line for grade IV toxicity was defined by [0.0084* age(yrs)* + 0.2], and for grade V toxicity by [0.024* age(yrs)* - 0.87]. On multivariate logistic regression, amongst all those with grade 4-5 radiation gastroenteritis / colitis, age was the only clinical predictor of death (dosimetric factors were not available). We demonstrate that the risk of acute severe radiation-induced bowel toxicity in rectal cancer patients rises linearly with age. Modelling demonstrates that the gradient is far steeper for grade V toxicity than grade IV toxicity. Importantly, our results facilitate the incorporated of age into existing NTCP models of bowel toxicity. Age-adjusted radiation dose constraints for the bowel should be adopted. [ABSTRACT FROM AUTHOR]