Stereotactic Body Radiation Therapy for Apical Lung Tumors: Dosimetric Analysis of the Brachial Plexus and Preliminary Clinical Outcomes

Dosimetric constraints of the brachial plexus have not yet been well-established for patients undergoing stereotactic body radiation therapy (SBRT). This study evaluated long-term experience with the treatment of early-stage apical lung tumors with SBRT and reports on dosimetric correlates of outcome.Between 2009 and 2018, a total of 78 consecutive patients with 81 apical lung tumors underwent SBRT for T1-3N0 non-small cell lung cancer. Apical tumors were those with tumor epicenter superior to the aortic arch. The brachial plexus (BP) was anatomically contoured according to the Radiation Therapy Oncology Group atlas. Patient medical records were reviewed retrospectively to determine incidence of brachial plexus injury (BPI) and a normal tissue complication probability model was applied to the dosimetric data.Five patients (6.4%) reported neuropathic symptoms consistent with BPI and occurred a median 11.9 months after treatment (range, 5.2-28.1 months). Most common dose and fractionation in those developing BPI were 50 Gy in 5 fractions (4 patients). Symptoms consisted of pain in 2 patients (40.0%), numbness in the hand or axilla in 4 patients (80.0%), and ipsilateral hand weakness in 1 patient (20.0%). In the overall cohort the median BP Dmax (EQD23 Gy) was 5.13 Gy (range, 0.18-217.2 Gy) and in patients with BPI the median BP Dmax (EQD23 Gy) was 32.14 Gy (range, 13.4-99.9 Gy). The normal tissue complication probability model gave good fit with an area under the curve of 0.75 (odds ratio, 7.3; 95% confidence interval, 0.8-68.3) for BP Dmax (EQD23 Gy) threshold of 20 Gy.Significant variation exists in the dose delivered to the brachial plexus for patients treated by SBRT for apical lung tumors. The incidence of neuropathic symptoms in the post-SBRT setting was appreciable and prospective clinical correlation with dosimetric information should be used to develop evidence-based dose constraints.