Predictive Value of Dynamic Tumor Volume Changes in Stage III Non-Small Cell Lung Cancer Treated with Chemoradiation and Consolidative Immunotherapy

Abstract

Concurrent chemoradiation (CRT) followed by Immunotherapy is standard of care in unresectable, locally advanced, non-small cell lung cancer (NSCLC). We hypothesize that on-treatment dynamic changes in tumor volume may predict oncologic outcome in Stage III NSCLC. Stage III NSCLC patients who were treated with definitive CRT of 60 Gy in 30 fractions over 6 weeks with concurrent platinum-based concurrent chemotherapy followed by consolidative immunotherapy were retrospectively reviewed. We manually delineated the gross tumor volumes (GTV) of patients on cone beam computed tomography (CBCT) acquired on day 1, 15, 29 and 43. GTV reduction was quantified as the percent difference in volume on each CBCT as compared to CBCT Day 1. Mid-treatment response (MTR) and End of treatment response (ETR) was defined at day 29 (at 40 Gy) and day 43 (at 60 Gy), respectively. Loco-regional control (LRC) and overall survival (OS) was evaluated by Kaplan-Meier analysis, with log-rank test for groups stratified as per treatment response. Multivariate Cox regression analysis was performed to identify additional prognostic factors. We evaluated 24 consecutively treated patients from 01/2016 to 08/2019, with a median follow-up of 30.5 months. Median age was 69 years (range 51-84). Adenocarcinoma histology was present in 58 % and squamous cell carcinoma in 42% patients. The tumor stage IIIA in 38% and Stage IIIB in 62% patients. All patients received definitive CRT with at least 2 cycles of immunotherapy (median 8 cycles). The median ETR at CBCT43 was 49.5% (range, 1%-84.7%). 1-year cumulative incidence of loco-regional failure (LRF) was 27% versus 10% in patients with ETR <49.5% as compared to patients with greater ETR (p = 0.31). The median MTR at CBCT29 was 32.7% (range, 1%-81%). 1-year cumulative incidence of LRF was 28% versus 9% in patients MTR <32.7% as compared to patients with higher response (p = 0.03). The 3-year actuarial LRC of all patients was 51% which was significantly better in higher MTR (70% vs 30%; p = 0.02, log-rank) as compared to lower MTR. The 3-year actuarial OS for all patients was 45% which was 54% in patients with higher MTR verses 37% in low responders (p = 0.09). On multivariate analysis, age and T-stage were significant factors associated with LRC. Histologic subtype showed no impact on our findings. Stage III NSCLC patients with approximately one-third (33%) reduction of GTV on mid-treatment CBCT during CRT have significantly improved LRC and a trend towards improved OS. This approach may justify early adaptive intensification of RT or application of additional therapies in patients with non-optimal response. Additional large prospective study of CBCT is needed to determine whether treatment can be tailored based on tumor response, to improve outcomes.