Right Atrial Dose Is Associated with Worse Outcome in Patients Undergoing Definitive Stereotactic Body Radiation Therapy for Central Lung Tumors

SIMPLE SUMMARY: The clinical consequences of irradiating the cardiac substructures during stereotactic body radiation therapy (SBRT) remains unclear. We evaluated 83 lung cancer patients who underwent SBRT for early stage lung cancer. Using specialized software, we generated structures for fourteen cardiac substructures and evaluated radiation dose parameters for each. Among these parameters, the dose to 45% (D45%) of either the right atria or ventricle was associated with worse non-cancer associated survival with an identified cutoff value of 890 cGy and 564 cGy for each, respectively. Via these cutoffs, the D45% to the right atria, not the right ventricle, was associated with worse non-cancer associated and overall survival. Based on these findings, reducing the dose to the right atria during SBRT may improve patient outcomes in at risk patients. ABSTRACT: The consequence of cardiac substructure irradiation in patients receiving stereotactic body radiation therapy (SBRT) is not well characterized. We reviewed the charts of patients with central lung tumors managed by definitive SBRT from June 2010–April 2019. All patients were treated with five fractions, typically either 5000 cGy (44.6%) or 5500 cGy (42.2%). Via a multi-patient atlas, fourteen cardiac substructures were autosegmented, manually reviewed and analyzed using dosimetric parameters. A total of 83 patients were included with a median follow up of 33.4 months. Univariate Cox regression analysis identified a D45% dose to the right atria and ventricle for further study. Sequential log-rank testing evaluating an association between non-cancer associated survival and D45% dose to the right atria or ventricle and association was employed, identifying candidate cutoff values of 890.3 cGy and 564.4 cGy, respectively. Kaplan–Meier analysis using the reported cutoff values found the D45% right atria constraint to be significantly associated with non-cancer associated (p ≤ 0.001) and overall survival (p ≤ 0.001) but not the right ventricle constraint. Within a multivariate model, the proposed right atria D45% cutoff remained significantly correlated with non-cancer associated survival (Hazard’s Ratio (HR) ≤ 8.5, 95% confidence interval (CI) 1.1–64.5, p ≤ 0.04) and OS (HR ≤ 6.1, 95% CI 1.0–36.8, p ≤ 0.04). In conclusion, a dose to D45% of the right atria significantly correlated with outcome and the candidate constraint of 890 cGy stratified non-cancer associated and OS. The inclusion of these findings with previously characterized relationships between proximal airway constraints and survival enhances our understanding of why centrally located tumors are high risk and potentially identifies key constraints in organ at risk prioritization.