Association of Lung Stereotactic Body Radiation Therapy Conformality Index and Posttreatment Radiation Pneumonitis in Early Stage Non–small Cell Lung Cancer

Abstract

Previous studies of lung stereotactic body radiotherapy (SBRT) for non-small cell lung cancer (NSCLC) have identified the total volume of lung 20Gy (V20) and mean lung dose (MLD) as important variables for limiting radiation pneumonitis (RP). Conformality index of the 50% prescription isodose volume (CI50) is an important variable in SBRT planning but has not been previously correlated to RP. We hypothesize that adherence to recommended CI50 guidelines for patients undergoing lung SBRT will result in decreased incidence and severity of RP compared to patients with minimal deviations and unacceptable deviations. We retrospectively identified 168 patients treated between 2006 and 2016, with >3 months follow up from lung SBRT treatment (40-60 Gy in 5 fractions with 50 Gy in 5 fractions as the most common) for early stage NSCLC. CTCAE V4.0 toxicity grades were used to classify RP. Clinically significant RP was defined as grade ≥2 toxicity. Using CI50 guidelines based on volume of PTV, patients were separated into three groups: acceptable, minor deviation, and unacceptable based on Radiation Therapy Oncology Group clinical trial guidelines. The relationship of CI50 groupings to increasing toxicity was evaluated with a univariable ordinal logistic regression model. Median follow up was 21.8 months (range 4.1-115 months). The rate of grade ≥2 RP for the entire cohort was 12.5%. 34 patients had acceptable CI50, 101 had minor deviations, and 33 had unacceptable deviations. Among patients with acceptable CI50, 5 (14.7%) had clinically significant RP. Among patients with minor deviations, 11 (10.9%) had clinically significant RP. Among patients with unacceptable CI50, 5 (15.2%) had clinically significant RP. The only patient with grade 3 pneumonitis was in the group with unacceptable CI50. Those with minor deviations had 0.71 (95% CI: 0.23-2.21) times the odds for clinically significant toxicity compared to those with acceptable CI50. Those with major deviations had 1.04 (95% CI: 0.27-3.97) times the odds for higher toxicity compared to those with acceptable CI50. Adhering to recommended CI50 values alone does not significantly decrease the incidence of clinically significant RP in patients with NSCLC treated with SBRT. Nevertheless, CI50 is an important treatment planning parameter to maintain a compact intermediate dose level which will correspond to other known predictors of RP such as V20 and MLD. Further research is warranted to clinically validate CI50 goals for future clinical trials.