Dosimetric parameters related to occurrence of distant metastases and regional nodal relapse after SBRT for early-stage non-small cell lung cancer

Purpose Previous studies have suggested that the dose immediately outside the PTV may impact the incidence of distant metastases after stereotactic body radiation therapy (SBRT) for patients with early-stage non-small cell lung cancer (NSCLC). In particular, Diamant et al. [1,2] reported a correlation between the mean EQD2 of a 30 mm shell around the PTV and both local control and the rate of distant metastases. In this study, we assess this parameter and others in a series of patients with radiographically presumed or biopsy-proven early-stage NSCLC treated at our institution with stereotactic body radiotherapy (SBRT) between 2017 and 2019. Materials/Methods We reviewed the dosimetry, local control, regional nodal relapse, and rate of distant metastases for 304 patients with 325 lesions treated with SBRT at our institution. Dosimetric parameters investigated include the prescribed dose, minimum and mean doses to the PTV, conformity index, and the mean EQD2 to a 30 mm shell around the PTV. Time to each event was defined from date of last fraction of SBRT to date of event, with event-free patients censored at last radiographic follow-up. Univariate (UVA) Cox regression analysis was performed on the collected parameters to assess for correlation with regional nodal relapse and rate of distant metastases. Results There was no significant correlation between the mean EQD2 dose to a 30 mm shell around the PTV and the rate of distant metastases. On UVA Cox proportional hazards analysis, positive predictors of reduced incidence of distant metastases were PTV <22 cc (vs. ≥22 cc, p = 0.01) and GTV <10 cc (vs. ≥10 cc, p < 0.01), with GTV <10 cc also being a positive predictor of reduced incidence of regional nodal relapse (p < 0.01). In the subset of patients treated with 4–5 fractions, mean EQD2 dose to the 30 mm shell around the PTV ≥21 Gy was associated with increased incidence of distant metastases (HR 2.42, 95% CI 1.06–5.53, p = 0.04), differing from prior data from Diamant et al. Conclusions We did not observe a correlation between the rate of distant metastases and dose outside the PTV, as reported by other groups; rather, we noted an opposite trend in patients treated with 4–5 fractions. Our data show additional correlations between distant metastases and tumor size.