Impact of Tumor Location & Dosimetric Predictors for Chest Wall Toxicity in Single Fraction SBRT for Stage I Non-Small Cell Lung Cancer

Abstract

Single fraction stereotactic body radiation therapy (SF-SBRT) is an acceptable regimen for the treatment of peripheral Stage I Non-Small Cell Lung Cancer (NSCLC). Rates of chest wall toxicity (CWT) reported by phase II trials did not stratify by distance of tumor to chest wall (CW) and dosimetric parameters to limit CWT are not well defined. We sought to determine the relationship of tumor location and dosimetric parameters with CWT in SF-SBRT for peripheral Stage I NSCLC. An institutional review board-approved prospective SBRT registry of 1,462 patients (pts) was used to identify pts treated with 30 Gy or 34 Gy in one fraction, on or off relevant protocols. Tumors were ≤ 5 cm, node-negative, and ≥ 2 cm from the proximal tracheo-bronchial tree. The ribs and CW were retrospectively contoured (CW as a 3 cm soft-tissue structure between the sternum and vertebral body). Gross tumor volume was measured as abutting, ≤ 1 cm, 1-2 cm or > 2 cm from the CW. CWT was graded according to CTCAE 3.0 criteria. Rates of CWT were compared using unpaired t-test. Logistic regression analysis was used to identify tumor and dosimetric parameters associated with CWT. This study included 138 pts treated with SF-SBRT to 146 lesions. Median follow-up was 23.8 months (34.7 months for living pts). There were 80 pts (55%) treated with 30 Gy and 66 pts (45%) treated with 34 Gy. The rate of CWT was 30.6% for lesions abutting CW, 8.2% for lesions ≤ 1 cm from CW, 3.8% for lesions 1-2 cm from CW, and 5.7% for lesions > 2 cm from CW. Grade ≥ 3 CWT for the whole cohort was modest (1.4%). Tumor abutment (OR 6.5; p=0.0005), BMI (OR 1.1; p= 0.02), rib D1cc (defined as dose to 1 cc) (OR 1.01 per Gy; p= 0.03), CW D1cc (OR= 1.08 per Gy; p=0.03), and CW D5cc (OR 1.10 per Gy; p= 0.01) were significant predictors for CWT on univariate analysis. Tumor abutment was the only significant predictor for CWT (OR 7.5; p= 0.007) on multivariate analysis. CWT remained low until a dose threshold and then increased more rapidly. CWT occurred in only 1 out of 40 pts with CW D5cc < 18 Gy (defined as dose to 5 cc less than 18 Gy), while our model suggested a 15% risk of CWT with D5cc of 27.2 Gy to the CW. CWT occurred in only 1 out of 39 pts with rib D1cc < 19 Gy, while our model suggested a 15% risk of CWT with D1cc of 30.2 Gy to the rib. The rate of CWT is significantly associated with distance from the CW. When considering only lesions adjacent to the CW, the rate of CWT in this series (30.6%) does not appear to exceed rates in the published fractionated SBRT literature (20-33%). This suggests location adjacent to the CW should not be a contraindication to SF-SBRT. Rib D1cc and CW D1cc and D5cc may be used as predictors of CWT rates, as noted in this model. As most CWT is low-grade and self-limited, these dosimetric parameters should be utilized as a guideline, rather than an absolute constraint.