A. L. Elaimy1, M. Schipper1,2, H. Yin1,2, M. M. Matuszak1, M. M. Dominello3, D. P. Bergsma1, P. A. Paximadis4, M. Zaki5, L. L. Kestin6, J. A. Hayman1, A. F. Dragovic1, and S. Jolly1; 1Department of Radiation Oncology, University of Michigan, Ann Arbor, MI, 2Department of Biostatistics, University of Michigan, Ann Arbor, MI, 3Department of Radiation Oncology, Karmanos Cancer Center, Detroit, MI, 4Department of Radiation Oncology, Corewell Health South, St. Joseph, MI, 5Covenant HealthCare, Saginaw, MI, 6Michigan Healthcare Professionals/GenesisCare USA, Farmington Hills, MI
Purpose/Objective(s): For patients receiving radiation therapy for non-small cell lung cancer (NSCLC), unplanned interruptions prolong the treatment course and may be associated with adverse outcomes. However, little is known regarding patient and treatment planning characteristics that contribute to interruptions. The purpose of this study was to characterize the frequency, type duration and predictors of interruptions in radiation therapy in patients with NSCLC treated with conventional fractionation throughout a statewide quality collaborative. Materials/
Methods: Clinical and dosimetric data as well as frequency and duration of treatment interruptions (> or = 5 days) were prospectively collected by 29 institutions within the Michigan Radiation Oncology Quality Consortium between 2012 and 2024 for patients with NSCLC treated with conventional fractionation using a physician-assessed survey. In version 1 of the survey (2012 to August 2017) data regarding toxicity breaks only were recorded. In version 2 of the survey (September 2017 to 2024) data regarding both any treatment interruption and toxicity breaks were recorded. We modeled the influence of patient, disease and treatment characteristics including mean and other radiation dose metrics for lung, heart and esophagus on the odds of any treatment interruption using multivariate logistic regression. Results: Toxicity breaks were reported in 9% (131/1476) of patients and toxicity breaks greater than 5 days were reported in 3.5% (51/1476) of patients. Any treatment interruption was reported in 18% (154/867) of patients and toxicity breaks were reported in 6% (56/867) of patients enrolled during version 2 of the survey. Stepwise modeling identified a MV model for any toxicity break including heart V10Gy (OR per 10% increase = 1.16, p=0.001), concurrent chemotherapy (Yes vs No = 2.67, p=0.024) and ECOG (OR per 1 point increase = 1.91, p<0.001). In this model, all of the other tested heart dose metrics (mean and V10-V60) were significant predictors for any toxicity break and mean esophagus dose was not a significant predictor for any toxicity break. Stepwise modeling identified a MV model for toxicity break > 5 days including heart V10Gy (OR per 10% increase = 1.12, p<0.001) and ECOG (OR per 1 point increase = 1.87, p <0.001). Additionally, stepwise modeling selected smoking status (p=0.05), ECOG (p<0.01) and lung V10Gy (p<0.01) as jointly significant predictors of any toxicity related break. PTV volume was not associated with treatment interruptions in all 3 models. Conclusion: Both clinical and dosimetric factors are associated with treatment interruptions in patients with NSCLC undergoing conventional fractionation. Efforts should be made to identify patients at increased risk of interruptions to optimize treatment planning and minimize toxicity.
.jpeg "Ameer Elaimy, MD, PhD, BS photo")