Automated Global Longitudinal Strain Assessment in Long-Term Survivors of Childhood Acute Lymphoblastic Leukemia.

ABSTRACT

There is limited evidence that supports the use of the global longitudinal strain (GLS) in long-term cardiac monitoring of childhood acute lymphoblastic leukemia survivors (CLSs). Our aim was to assess the utility of automated GLS to detect left ventricular systolic dysfunction (LVSD) in long-term CLSs. Asymptomatic and subclinical LVSD were defined as LVEF < 50% and GLS < 18.5%, respectively. Echocardiographic measurements and biomarkers were compared with a control group. Inverse probability weighting was used to reduce confounding. Regression models were used to identify factors associated with LVEF and GLS in the survivors. Ninety survivors with a median follow-up of 18 (11−26) years were included. The prevalence of LVSD was higher using GLS than with LVEF (26.6% vs. 12.2%). The measurements were both reduced as compared with the controls (p < 0.001). There were no differences in diastolic parameters and NT-ProBNP. Survivors were more likely to have Hs-cTnI levels above the detection limit (40% vs. 17.2%, p = 0.006). The dose of anthracycline was associated with LVEF but not with GLS in the survivors. Biomarkers were not associated with GLS or LVEF. In conclusion, LVSD detection using automated GLS was higher than with LVEF in long-term CLSs. Its incorporation into clinical routine practice may improve the surveillance of these patients.