Chemoradiation vs. Radiation Alone in the Management of Inoperable cT1-T2N1 Non-Small Cell Lung Cancer, a Propensity-Weighted Hospital Database Analysis.

Purpose/Objective(s) Lobectomy followed by adjuvant chemotherapy is the standard of care for patients with clinically T1-T2N1 non-small cell lung cancer (NSCLC). However, the optimal treatment of patients who cannot undergo or refuse to undergo surgical resection is controversial. Currently, NCCN guidelines recommend concurrent chemoradiation; however, limited evidence exists to substantiate this recommendation for this particular patient population. Many groups advocate for radiation alone with standard or hypofractionated approaches. We hypothesized that concurrent chemoradiation would be associated with longer overall survival (OS) compared to hypofractionated or conventionally fractionated radiation. Materials/Methods We queried the National Cancer Database (NCDB) for patients with cT1-T2N1M0 NSCLC treated non-surgically with definitive radiation or concurrent chemoradiation between 2004 and 2016. We further restricted our cohort to patients receiving RT doses with an EQD2 ≥ 58.4 Gy and > 5 fractions. Patients were divided into four treatment groups: standard fractionated radiation (SFRT), hypofractionated radiation (HFRT), standard fractionated chemoradiation (SFRT+CT), and hypofractionated chemoradiation (HFRT+CT). Chemoradiation was defined as chemotherapy and radiation with start dates within 30 days. To address confounding by indication, inverse probability of treatment weights (IPTW) were constructed regressing sociodemographic, clinical, and facility characteristics on treatment group in a multinomial regression model. Weighted Cox proportional hazard regression models were then used to evaluate the relative hazard of death with SFRT set as the referent treatment group. Results A total of 2,534 patients met inclusion criteria and of these 2,200 patients (57,462.83 person-months) had complete data for all confounders and were included in the analytic sample. Of the 2,200 patients 26.6% received SFRT, 7% received HFRT, 62.1% received SFRT+CT, and 4.2% received HFRT+CT. The median EQD2 (IQR) were: SFRT 66 Gy (60-68), HFRT 66.7 Gy (62.5-71.9), SFRT+CT 65.5 Gy (60-66), and HFRT+CT 67.6 Gy (65.1-72.3). 27% of patients were still alive 36 months after the initiation of RT (17.3% of SFRT, 20.3% of HFRT, 33.2% of SFRT+CT, and 21.8% of HFRT+CT). After IPTW propensity weighting, the probability of death at 3 years was compared between the treatment groups and only SFRT+CT was associated with longer OS. Compared with SFRT, the hazard ratio for all-cause mortality for those treated with SFRT+CT was 0.67 (95% CI: 0.57-0.75), for HFRT was 0.86 (95% CI: 0.61-1.06), and for HFRT+CT was 0.81 (95% CI 0.61-1.06). Conclusion To our knowledge, this is the first large database propensity weighted analysis of patients with cT1-T2N1M0 NSCLC managed non-surgically with definitive radiation or chemoradiation. In this group, treatment with SFRT+CT is associated with longer OS relative to treatment with radiation alone.