Global Longitudinal Strain: A Potential Noninvasive Tool for Early Detection of Radiation-Induced Cardiac Dysfunction in Patients With Lung Cancer Receiving Thoracic Radiation Therapy.

Abstract

Purpose: Radiation-induced cardiac dysfunction (RICD) is a competing cause of morbidity and mortality in patients receiving thoracic radiation therapy (RT). Currently, there are no clinically-validated approaches for early detection of RICD at a time point that affords the potential for mitigation. The goal of this study was to evaluate the potential of global longitudinal strain (GLS) derived from standard-of-care echocardiogram (ECHO) for the early detection of RICD and to assess the association between adverse GLS changes and survival in patients receiving thoracic RT.

Methods and materials: A retrospective review of changes in GLS was carried out in patients with primary or secondary lung cancer who received standard-of-care thoracic RT with a mean heart dose of ≥5 Gy and had measurable GLS on ECHOs performed before and after RT. Changes in 2-chamber (2C), 3-chamber (3C), and 4-chamber (4C) GLS and peak average GLS after RT (relative to pre-RT baseline) were quantified. Survival probabilities were estimated in patients with normal versus abnormal GLS.

Results: Thirty-eight patients had measurable GLS before and after RT. Abnormal GLS (defined as <18% or >15% relative decline in GLS after RT from a normal baseline value) was present in 31.6% of patients before RT and 57.9% of patients after RT (P = .012). On paired comparisons, the absolute median reduction (IQR) in 2-chamber, 3-chamber, 4-chamber, and average GLS after RT relative to pre-RT baseline was 1.90 (4.43), 3.00 (3.83), 2.50 (3.63), and 2.25 (3.53), respectively, all P < .001. No statistically significant change in left ventricular ejection fraction was noted after RT. Patients with abnormal GLS after RT had significantly worse survival than those with normal GLS on univariable analysis (P = .049). Despite the small sample size of the study, the survival detriment in patients with abnormal GLS after RT strongly trended toward significance on multivariable analysis (P = .063).

Conclusions: Adverse changes in GLS are detectable on standard-of-care ECHOs and precede significant changes in left ventricular ejection fraction in this cohort of high-risk patients with primary and secondary lung cancer receiving thoracic RT. Thus, ECHO-derived GLS has the potential to serve as an early and noninvasive marker of RICD in this patient population and may enable early adoption of GLS-guided cardioprotective therapy, which has been shown to mitigate cardiac dysfunction in patients with cancer receiving cardiotoxic treatments.