Recurrence Risk and Survival Following SBRT for Lung Cancer: Impact of Prognostic Factors Used for Patient Selection in a Randomized Trial

Abstract

<h3>Purpose/Objective(s)</h3> Stereotactic body radiation therapy (SBRT) is the standard treatment for early stage, medically inoperable non-small cell lung cancer (NSCLC), attaining excellent local control. However, rates of regional and distant recurrence are relatively high. Current randomized phase 3 trials are investigating the addition of immunotherapy to SBRT in this setting to improve disease control and survival, with varying patient selection criteria. We investigate the influence of the high-risk factors used for patient selection on the ongoing randomized phase 3 SWOG/NRG trial (tumor size ≥2 cm, SUV ≥6.2, and grade 2-3) on disease control and survival in a single institution cohort and hypothesize that patients with high-risk features will have increased rates of recurrence and death. <h3>Materials/Methods</h3> Following IRB approval, we retrospectively analyzed patients treated with SBRT for early stage (T1-3N0M0) NSCLC between 2007 and 2019 at a single institution. Patient and disease specific characteristics and oncologic outcomes were documented. Risk factors including size ≥2 cm, SUV ≥6.2, and grade 2-3 histology were documented for each patient, and total number of high-risk features for each patient was calculated. In-field control (IFC), distant control (DC) and overall survival (OS) were calculated using the Kaplan Meier method, and the influence of high-risk factors was assessed with the log-rank method. <h3>Results</h3> We identified 217 courses of SBRT performed for early-stage NSCLC. All cases had tumor diameter available, 196 had SUV_max prior to SBRT available, and 68 had histologic grade available. Median patient age was 76.2 years (range 48.9-93.1) and median follow up was 35.1 months (range: 1.6-145-3). Tumors were treated to a median of 50 Gy (range: 40 - 60 Gy) in a median of 4 (range: 3-8) fractions. OS decreased with increasing number of risk factors, with 3-year OS of 73.7%, 72.2%, 50.7%, and 44.4% for patients with 0, 1, 2, and 3 high risk factors, respectively (p<0.001). Patients with 0-1 risk factors had 3-year OS of 73.2% as compared to 48.3% for those with 2-3 risk factors (p<0.001). This appeared to be driven by distant, rather than local recurrence: patients with 0-1 high risk factors had 3-year DC of 85.3% as compared to 73.3% P=0.006), but the assessed risk factors were not predictive of IFC. Individually, size ≥2 cm was the most robust risk factor, with 3-year DC of 72.7% versus 89.2% and OS of 61% versus 70.2% for tumors ≥2 cm and <2 cm, respectively. <h3>Conclusion</h3> High-risk features used for patient selection for an ongoing phase 3 trial testing the addition of immunotherapy to SBRT in early-stage NSCLC were highly predictive in aggregate for DC and OS but not LC in a single institution cohort of patients. Several ongoing randomized trials will determine whether immunotherapy reduces recurrence and improves OS in medically inoperable NSCLC treated with SBRT.