Radiation Pneumonitis Following Lung SBRT: The Role of Circumferential Irradiation to the Anatomic Levels of the Proximal Bronchial Tree

Abstract

Stereotactic body radiation therapy (SBRT) to the lung with dose escalated, hypofractionated treatments can lead to clinically significant lung toxicity in patients including radiation pneumonitis (RP), fistula, and even fatal hemoptysis. Despite adherence to established dosimetric parameters for the lung and proximal bronchial tree (PBT) patients can still develop RP. Prior analysis of our institution’s lung SBRT cohort indicated that high-dose circumferential radiation (HDCR) of the PBT was associated with development of RP. In this study, we sought to determine if the sensitivity of the PBT to HDCR was consistent throughout the structure. A retrospective analysis was conducted for 42 patients who were treated with lung SBRT at one institution between 2010 and 2018. Treatment plans were evaluated for doses delivered to the lung and PBT as well as for presence of circumferential overlap of fractionation relevant isodose lines with the PBT. The isodose lines that defined HDCR were 15, 16, and 18 Gy, which correspond with the PBT D4cc dose limit for 3x18 Gy, 4x12 Gy, and 5x10 Gy fractionation schemes, respectively. The PBT was divided into trachea, primary bronchi, and secondary bronchi to assess if the effect of HDCR varied based on anatomic location. Posttreatment chest imaging and patient charts were evaluated and RP was graded according to the RTOG Toxicity Criteria. This was classified as mild if RTOG grades 0-1 and moderate if RTOG grades 2-3. There were no grades 4-5 toxicity in this group of patients. Statistical analysis was performed using logistic regression. There was a significant association between HDCR of the secondary bronchi and development of RP (OR, 5.0; 95% CI, 1.21-20.6) but no significant association with HDCR of the primary bronchi (OR, 1.38; 95% CI, 0.18-10.8.) There were no patients with HDCR of the trachea. There was no significant association between D4cc (OR, 0.98; 95% CI, 0.90-1.07) or D.03cc (OR, 1.03; 95% CI, 0.99-1.07) to the PBT, mean lung dose (OR, 2.36; 95% CI, 0.67-8.39), lung V5 (OR, 2.2; 95% CI, 0.63-7.66) or lung V20 (OR, 3.07; 95% CI, 0.83-11.4) and grade of RP. This data suggests that circumferential overlap of the secondary bronchi of the PBT with isodose lines corresponding with the D4cc limit plays a role in the development of symptomatic RP. We postulate that this may be due to the smaller diameter of secondary bronchi, which make them more susceptible to luminal stenosis. This should be considered in SBRT treatment planning as conformal dose delivery can result in high doses to a small portion of the airway. Further investigation and validation of these findings with a larger data set is warranted.