The Correlation Between Changes in Cardiac Biomarkers and Cardiac Events in Patients With Non-Small Cell Lung Cancer (NSCLC) Treated With Stereotactic Body Radiation Therapy (SBRT) - Exploratory Analysis of a Phase II Study.

PURPOSE/OBJECTIVE(S): Heart dose in patients receiving thoracic radiotherapy (RT) correlates with cardiovascular events (CVE) and outcomes, but there are currently no established cardiac biomarkers predictive of RT-induced cardiotoxicity. A cardiac risk score (CRS) (10-point score: 0- < 4 = low risk (LR), 4- < 7 = intermediate risk (IR), 7-10 = high risk (HR)) determined by a validated cardiac biomarker panel of 4 proteins (NT-proBNP, osteopontin, KIM-1 and TIMP-1) has been previously shown to predict for CVE in a non-oncology population. We assessed RT-induced changes in the CRS before, during and after RT and hypothesize that biomarker elevations following SBRT correlate with CVE (MI, stroke, CV death, heart failure, arrythmia, valvular surgery), and with heart dose.

MATERIALS/METHODS: Patients with NSCLC were treated with SBRT on an IRB-approved prospective phase II study on dose escalation in patients with large (T1b-T4N0-N1M0) primary NSCLC. Consecutive patients treated from 2013 to 2018 with a history of a prior CVE, a CVE after SBRT, or a mean heart dose (MHD) > 5 Gy were selected for CRS analysis. Patient samples were obtained pre-treatment, on the last day of SBRT and in follow-up (1, 3 and 6 months after SBRT). Two-sample t-tests were performed to compare CRS over time. Subset analyses were performed in patients with or without MHD > 5 Gy and in patients with or without a CVE during follow-up.

RESULTS: A total of 62 samples from 17 patients were analyzed for CRS. The median age was 79 years and 53% were male. Patients were AJCC 8th edition stage IB (29%), IIA (12%), IIB (41%) and IIIA (18%). RT doses were 40 to 45 Gy in 5 fractions (18%), 50 Gy in 5 fractions (70%) or 60 Gy in 8 fractions (12%). Eight (47%) patients had a MHD > 5 Gy. At a median follow-up of 21 months, 11 (65%) patients experienced a CVE after RT. Baseline CRS were LR (41%), IR (41%), HR (12%) or not evaluable (6%). CRS increased from baseline to the last day of SBRT (mean difference (MD) in score was +0.52, P = 0.06). On subgroup analysis, patients who experienced a CVE in follow-up had a greater increase in CRS (MD: +1.101, P = 0.008). There was also a trend of increase in CRS in patients with MHD > 5 Gy (MD: +0.592, P = 0.1). There were no significant changes in the CRS during the follow-up period.

CONCLUSION: Our results from this exploratory prospective analysis demonstrate a significant correlation between an increase in CRS during SBRT and the development of CVE in follow-up after SBRT. Additionally, a trend in association between an increase in CRS during SBRT and MHD was identified, which could indicate an underlying relationship between CRS change during RT and MHD. These findings should be further explored in a larger patient population.