Development of an Objective Scoring System for Endoscopic Assessment of Radiation-Induced Upper Gastrointestinal Toxicity

SIMPLE SUMMARY: High-dose radiation therapy techniques have gained increasing interest in pancreatic cancer treatment, but toxicity to the upper gastrointestinal (GI) organs remains a major concern. We aimed to develop an objective toxicity scoring system to be used during endoscopic evaluation that allows for direct assessment of the stomach and duodenum before and after radiation treatment. Our toxicity scoring takes into account the pathological categories of erythema, edema, ulceration, and stricture to determine radiation-related GI toxicity. We assessed and validated the upper GI toxicity of 19 locally advanced pancreatic cancer trial patients undergoing stereotactic body radiation therapy (SBRT). With future usage, we hope this scoring system will provide objective and reliable assessments of changes in GI toxicity during the radiation treatment of pancreatic cancer and for GI toxicity assessment during radiation therapy research trials. ABSTRACT: We developed and implemented an objective toxicity scoring system to be used during endoscopic evaluation of the upper gastrointestinal (GI) tract in order to directly assess changes in toxicity during the radiation treatment of pancreatic cancer. We assessed and validated the upper GI toxicity of 19 locally advanced pancreatic cancer trial patients undergoing stereotactic body radiation therapy (SBRT). Wilcoxon-signed rank tests were used to compare pre- and post-SBRT scores. Comparison of the toxicity scores measured before and after SBRT revealed a mild increase in toxicity in the stomach and duodenum (p < 0.005), with no cases of severe toxicity observed. Kappa and AC1 statistics analysis were used to evaluate interobserver agreement. Our toxicity scoring system was reliable in determining GI toxicity with a good overall interobserver agreement for pre-treatment scores (stomach, κ = 0.71, p < 0.005; duodenum, κ = 0.88, p < 0.005) and post-treatment scores (stomach, κ = 0.71, p < 0.005; duodenum, κ = 0.76, p < 0.005). The AC1 statistics yielded similar results. With future usage, we hope this scoring system will be a useful tool for objectively and reliably assessing changes in GI toxicity during the treatment of pancreatic cancer and for GI toxicity assessments and comparisons during radiation therapy research trials.