To identify clinical and dosimetric factors associated with vagus nerve (VN) and recurrent laryngeal nerve (RecLN) injury following stereotactic ablative radiation therapy (SABR) in the chest.

We examined the clinical courses and SABR plans of 67 patients treated for T1 or T2 non-small cell lung cancer of the upper right or left lung, including 2 who developed vocal cord paresis (VCP) following treatment. After developing a contouring atlas for the VN and RecLN in the thorax, dose to those structures was retrospectively determined for each patient, and we identified 12 patients whose treatment imparted significant dose to either nerve and who were assessable for more than 12 months follow-up. Biologically effective doses using linear-quadratic (LQ) and linear quadratic-linear (LQ-L) modeling were correlated with VN and RecLN toxicity.

Of 12 patients, 2 developed VCP. The first underwent repeat SABR and received a cumulative single fraction equivalent dose (alpha/beta = 3; SFED3) of 37.4 or 64.5 Gy to the VN and 13.7 or 15.3 Gy to the RecLN (by LQ or LQ-L modeling, respectively). This was the highest VN dose and fifth highest RecLN dose in the cohort. The second had rheumatoid arthritis and connective tissue disease and received a SFED3 of 16 Gy to the VN and 19.5 Gy to the RecLN (by both LQ and LQ-L modeling). This was in the upper tertile of VN and RecLN doses for the cohort.

Following SABR for non-small cell lung cancer, VCP was associated with high cumulative dose to the VN in 1 patient and a moderately high dose to the VN and RecLN in another patient with rheumatoid arthritis and connective tissue disease. Particularly in the setting of reirradiation or connective tissue disease, potential toxicity to the VN or RecLN should be considered.