Predictors of High-Grade Radiation Pneumonitis and Overall Survival Following Radiochemotherapy for Locally Advanced Non-Small Cell Lung Cancer: Analysis of Lung and Heart Dose Parameters and the Impact of Treatment Technique

Abstract

The effect of lung V20Gy and mean lung dose on the pneumonitis (RP) risk is well established. Recently, cardiac dose has been observed as a predictor of both RP and overall survival. Volumetric modulated arc therapy (VMAT) achieves highly conformal dose distributions at the expense of large low-dose volumes. As the use of VMAT continues to grow, we hypothesize that low radiation doses to heart and lungs are correlated with an increased risk of RP ≥ grade 3. Scoring of low grade RP can be subjective; RP≥3 was therefore chosen as an objective and clinically significant endpoint for this study.105 patients with locally advanced non-small cell lung cancer (median age 61 years, 62% male, 60% white) underwent conventionally fractionated radio(chemo)therapy to a median dose of 63Gy (45-70.2Gy). 3DCRT was used in 36%, IMRT in 51% and VMAT in 13%. A retrospective IRB-approved analysis of 25 clinical (gender, race, DLCO, diabetes, statin use, smoking history), radiographic (emphysema, ILD on pretreatment CT) and radiotherapy dose- and technique-related factors was performed to identify predictors of RP≥3 and overall survival (OS). Following testing of all variables for statistical association with RP and OS using univariate analysis (UVA), a forward selection algorithm was implemented for building a multivariate predictive model with limited sample size for logistic regression of RP and OS on all variables. ROC analysis was performed for variables significant on MVA.RP≥3 was diagnosed in 10/105 (9.5%) patients. 3-year OS rate was 43.4% with a median survival of 28.5 months. On UVA, predictors for RP≥3 were diabetes, PTV volume, upper lobe location, VMAT, lung V5Gy (%), lung Vspared5Gy (ml), lung V20Gy (%), heart V5Gy (% and ml). Predictors for OS were VMAT, lung V20Gy (%), mean heart dose (Gy), heart V5Gy/30Gy/60Gy (each in % and ml). On MVA, VMAT was the only significant predictor for both RP≥3 (P = 0.042) and OS (P = 0.013). Lung V5Gy (%, P = 0.073) and lung V20Gy (%, P = 0.08) were borderline significant for predicting RP≥3; AUC values were 0.86 and 0.83, respectively. Lung V20Gy (%, P = 0.09) and mean heart dose (Gy, P = 0.05) were borderline significant for predicting OS; AUC values were 0.60 and 0.68, respectively. Patients with RP≥3 had a median survival of 9.96 months compared to 29.5 months with RP < 3 or no RP (P = 0.02). Corresponding 2-year OS rates were 26.7% and 56.9%.In this study, VMAT was the only factor that was significantly correlated with both RP≥3 and OS. Lung V5Gy and lung V20Gy were excellent predictors of RP≥3 on ROC analysis. Avoiding RP is important not only as a toxicity per se, but also as a risk factor for poor survival. When using VMAT, caution needs to be taken to particularly reduce lung and heart V5Gy doses in addition to other well-established dose constraints.