(S035) Validity of Current Stereotactic Body Radiation Therapy (SBRT) Dose Constraints for Aorta and Major Vessels

April 15, 2016

From logistic modeling of 625 major vessels, the Radiation Therapy Oncology Group 0813 trial limit of Dmax = 52.5 Gy in five fractions was found to have a 1.2% risk of grade 3–5 toxicity, and the 2008 Timmerman limit of Dmax = 45 Gy in three fractions had a 2.3% risk. Further investigation is warranted, especially for the pulmonary artery, which might not have a dose tolerance as high as other major vessels.

Jinyu Xue, PhD, Gregory Kubicek, MD, Ashish Patel, MD, Benjamin Goldsmith, MD, Tamara LaCouture, MD, Sucha Asbell, MD; UT MD Anderson Cancer Center at Cooper University Hospital

PURPOSE: Major blood vessel toxicity data from stereotactic body radiation therapy (SBRT) are particularly sparse. Therefore, we investigated aorta and major vessel dose-response models to determine tolerance in patients treated with 1 to 5 fractions.

METHODOLOGY: A total of 625 major vessel structures, including aorta, vena cava, pulmonary artery, and pulmonary vein, were analyzed via dose-volume histograms (DVHs). From July 2008 to February 2015, a total of 387 cases with major vessels exposed to 1 to 5 fractions of SBRT with CyberKnife (Accuray, Inc., Palo Alto, CA) in our institution were studied. These data were combined with DVH data on 238 major vessels from the 2014 Nishimura study (J Thorac Oncol 2014;9:1370–6). The median number of fractions was 5, so the linear quadratic model with alpha/beta = 3 Gy was used to convert all doses to 5-fraction equivalent doses prior to any other modeling. DVH Evaluator software (DiversiLabs, LLC, Huntingdon Valley, PA) was utilized to create a logistic dose-response model of the aggregate dataset using maximum likelihood parameter fitting.

RESULTS: The published dataset included three Common Terminology Criteria for Adverse Events version 4 grade 3–5 pulmonary artery complications, and there were no observable major vessel complications in our dataset. The logistic model TD50 for Dmax was 81 Gy, and the slope parameter was 3.13. Models for V25Gy, D4cc, D1cc, and D0.5cc will also be presented. When the 73 pulmonary artery contours from Nishimura (2014) were analyzed alone, the risk was increased by approximately fivefold, but it is not known whether this was due to intrinsically increased sensitivity or attributable to lack of data in the omission of 552 noncomplication data points.

CONCLUSIONS: From logistic modeling of 625 major vessels, the Radiation Therapy Oncology Group 0813 trial limit of Dmax = 52.5 Gy in five fractions was found to have a 1.2% risk of grade 3–5 toxicity, and the 2008 Timmerman limit of Dmax = 45 Gy in three fractions had a 2.3% risk. Further investigation is warranted, especially for the pulmonary artery, which might not have a dose tolerance as high as other major vessels.

Proceedings of the 98th Annual Meeting of the American Radium Society - americanradiumsociety.org