Prognostic Value of GLS for Early Prediction of Cardiotoxic Chemotherapy

Study Questions:

What is the prognostic performance of left ventricular global longitudinal strain (GLS) in patients receiving potentially cardiotoxic chemotherapy?


The authors performed a meta-analysis of 21 studies comprising 1,782 patients with cancer (13/21 studies on breast cancer, 4/21 on hematologic malignances, and 4/21 on a combination of malignancies), who underwent treatment with anthracyclines with or without trastuzumab and who had echocardiographic assessment of GLS prior to or during chemotherapy and had follow-up for cancer therapy-related cardiac dysfunction (CTRCD). The definition of CTRCD varied, with 12 studies using a decrease in left ventricular ejection fraction. Fifteen studies used the GE EchoPAC software to derive GLS.


The incidence of CTRCD ranged from 9.3% to 43.8%, with a weighted pooled estimate of 21.0%. Four studies examined the association of pretherapy absolute GLS with CTRCD; of which two found no association. For active treatment absolute GLS (nine studies), the high-risk cutoff values ranged from -21.0% to -13.8%, with worse GLS associated with a higher CTRCD risk (odds ratio [OR], 12.27; 95% confidence interval [CI], 7.73-19.47; area under the hierarchical summary receiver operating characteristic curve [HSROC], 0.86; 95% CI, 0.83-0.89). For relative changes versus a baseline value (nine studies), cutoff values ranged from 2.3% to 15.9%, with a greater decrease linked to a 16-fold higher risk of CTRCD (OR, 15.82; 95% CI, 5.84-42.85; area under the HSROC, 0.86; 95% CI, 0.83-0.89). The cut-off value reported varied widely across studies. Last, funnel plot asymmetry suggested publication bias, and most studies were at high risk of bias.


Measurement of absolute GLS and relative change in GLS during therapy was predictive of CTRCD. Publication and study biases preclude identifying an optimal cut-off value for GLS.


This study is a necessary and impactful summary of the literature on the association between GLS and CTRCD. It confirms that a low or decrease in GLS is prognostic for subsequent CTRCD. The meta-analysis, however, highlights the small number of studies on GLS and CTRCD, their heterogeneity, and the high risk of bias, emphasizing the need for large prospective studies with standardized definitions of CTRCD to determine whether GLS is more effective than other clinical and echocardiographic assessments in predicting CTRCD. Last, the importance of a test is in its ability to change our management. It is time to determine how to incorporate the prognostic information of GLS in the management of patients receiving cardiotoxic therapy as a means to optimize their care.

Clinical Topics: Cardio-Oncology, Heart Failure and Cardiomyopathies, Noninvasive Imaging, Prevention, Acute Heart Failure, Echocardiography/Ultrasound

Keywords: Anthracyclines, Breast Neoplasms, Cardiotoxicity, Diagnostic Imaging, Echocardiography, Heart Failure, Prognosis, ROC Curve, Secondary Prevention, Stroke Volume, Ventricular Function, Left

< Back to Listings