348 - Impact of AI-Based Coronary Artery Calcium (CAC) Scores and Heart Dose on Survival in Locally-Advanced Non-Small-Cell Lung Cancer (LA-NSCLC): A Secondary Analysis of RTOG 0617
Presenter(s)
N. Haldar1, Y. Xiao2, D. Wen3, W. Choi1, Y. Vinogradskiy4, J. D. Bradley5, C. Hu6, and M. Werner-Wasik1; 1Dept of Radiation Oncology, Thomas Jefferson University Sidney Kimmel Cancer Center, Philadelphia, PA, 2Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, 3Bunkerhill Health, Bunker Hill, IL, 4Department of Radiation Oncology, Sidney Kimmel Medical College at Thomas Jefferson University, Philadelphia, PA, 5University of Pennsylvania/Abramson Cancer Center, Philadelphia, PA, 6Division of Biostatistics and Bioinformatics, Sidney Kimmel Comprehensive Cancer Center, Johns Hopkins University School of Medicine, Baltimore, MD
Purpose/Objective(s): The RTOG 0617 trial (NCTN00533949) demonstrated that a radiotherapy (RT) dose of 74 Gy with concurrent chemotherapy for LA-NSCLC resulted in worse overall survival (OS) compared to 60 Gy. The reasons for this remain unclear, and cardiovascular morbidity has been implicated. Elevated coronary artery calcium (CAC) scores have been associated with poorer OS following RT for NSCLC. This study retrospectively analyzed RTOG 0617 CT images using an artificial intelligence (AI) algorithm to assess whether high CAC scores contributed to worse OS in the 74 Gy arm.
Materials/Methods: Eligible patients (pts) from RTOG 0617 with available imaging, heart dose data, and RT records, and without cardiac stents, were included. Clinical and demographic data were obtained from the NCI Data Archive. A proprietary, FDA-cleared AI algorithm quantified CAC using the Agatston scoring method. Pts were categorized as having low (CAC < 100) or high (CAC = 100) CAC scores based on established thresholds. OS and progression-free survival (PFS) were the primary and secondary endpoints, respectively. An exploratory analysis examined the impact of heart V30 (=30% vs. <30%) on OS and PFS. Multivariable Cox regression assessed associations between CAC scores, heart dose (V30), and survival.
Results: Of 544 pts, 419 met inclusion criteria, with 278 (66%) having low CAC scores, evenly distributed between the 60 Gy and 74 Gy groups. Regardless of RT dose, high CAC scores were associated with worse OS (HR = 1.50, 95% CI [1.16-1.93], p = 0.0016) but not PFS (HR = 1.12, 95% CI [0.89-1.41], p = 0.31). Among low-CAC pts, 74 Gy was associated with significantly worse OS than 60 Gy (HR = 1.60, 95% CI [1.17-2.19], p = 0.0033). However, among high-CAC pts, OS (HR = 1.15, 95% CI [0.78–1.71], p = 0.48) and PFS (HR = 1.22, 95% CI [0.84–1.77], p = 0.29) did not significantly differ between RT dose groups.
Heart V30 was not correlated with CAC scores (Spearman’s R = 0.028, p = 0.57) and was similarly distributed between low- and high-CAC groups. However, heart V30 =30% was independently associated with significantly worse OS (HR = 1.68, 95% CI [1.28-2.20], p = 0.002). Among low-CAC pts, both high heart V30 (HR = 1.69, 95% CI [1.19-2.40], p = 0.0034) and 74 Gy (HR = 1.56, 95% CI [1.14-2.13], p = 0.006) were associated with worse OS. In contrast, among high-CAC pts, only heart V30 =30% remained significantly associated with worse OS (HR = 1.65, 95% CI [1.07-2.52], p = 0.0220), while RT dose had no significant effect (HR = 1.15, 95% CI [0.78-1.71], p = 0.48).Conclusion: In RTOG 0617, high CAC scores were associated with a worse prognosis, regardless of RT dose. Among low-CAC patients, a 74 Gy dose was associated with significantly poorer OS than 60 Gy, a pattern not observed in high-CAC patients. Heart dose (V30 =30%) was independently associated with worse OS. These findings suggest that dose escalation to 74 Gy may be particularly detrimental for patients with lower coronary risk. Further validation is warranted.