Reference Ranges of Left Ventricular Strain Measures by Two-Dimensional Speckle-Tracking Echocardiography in Children: A Systematic Review and Meta-Analysis

Levy, P. T.; Machefsky, A.; Sanchez, A. A.; Patel, M. D.; Rogal, S.; Fowler, S.; Yaeger, L.; Hardi, A.; Holland, M. R.; Hamvas, A.; Singh, G. K.

J Am Soc Echocardiogr. 2016 Jan 10; 29(3):209-225 e6

Abstract

BACKGROUND: Establishment of the range of reference values and associated variations of two-dimensional speckle-tracking echocardiography (2DSTE)-derived left ventricular (LV) strain is a prerequisite for its routine clinical adoption in pediatrics. The aims of this study were to perform a meta-analysis of normal ranges of LV global longitudinal strain (GLS), global circumferential strain (GCS), and global radial strain (GRS) measurements derived by 2DSTE in children and to identify confounding factors that may contribute to variance in reported measures. METHODS: A systematic review was launched in MEDLINE, Embase, Scopus, the Cumulative Index to Nursing and Allied Health Literature, and the Cochrane Library. Search hedges were created to cover the concepts of pediatrics, STE, and left-heart ventricle. Two investigators independently identified and included studies if they reported 2DSTE-derived LV GLS, GCS, or GRS. The weighted mean was estimated by using random effects models with 95% CIs, heterogeneity was assessed using the Cochran Q statistic and the inconsistency index (I(2)), and publication bias was evaluated using the Egger test. Effects of demographic (age), clinical, and vendor variables were assessed in a metaregression. RESULTS: The search identified 2,325 children from 43 data sets. The reported normal mean values of GLS among the studies varied from -16.7% to -23.6% (mean, -20.2%; 95% CI, -19.5% to -20.8%), GCS varied from -12.9% to -31.4% (mean, -22.3%; 95% CI, -19.9% to -24.6%), and GRS varied from 33.9% to 54.5% (mean, 45.2%; 95% CI, 38.3% to 51.7%). Twenty-six studies reported longitudinal strain only from the apical four-chamber view, with a mean of -20.4% (95% CI, -19.8% to -21.7%). Twenty-three studies reported circumferential strain (mean, -20.3%; 95% CI, -19.4% to -21.2%) and radial strain (mean, 46.7%; 95% CI, 42.3% to 51.1%) from the short-axis view at the midventricular level. A significant apex-to-base segmental longitudinal strain gradient (P < .01) was observed in the LV free wall. There was significant between-study heterogeneity and inconsistency (I(2) > 94% and P < .001 for each strain measure), which was not explained by age, gender, body surface area, blood pressure, heart rate, frame rate, frame rate/heart rate ratio, tissue-tracking methodology, location of reported strain value along the strain curve, ultrasound equipment, or software. The metaregression showed that these effects were not significant determinants of variations among normal ranges of strain values. There was no evidence of publication bias (P = .40). CONCLUSIONS: This study defines reference values of 2DSTE-derived LV strain in children on the basis of a meta-analysis. In healthy children, mean LV GLS was -20.2% (95% CI, -19.5% to -20.8%), mean GCS was -22.3% (95% CI, -19.9% to -24.6%), and mean GRS was 45.2% (95% CI, 38.3% to 51.7%). LV segmental longitudinal strain has a stable apex-to-base gradient that is preserved throughout maturation. Although variations among different reference ranges in this meta-analysis were not dependent on differences in demographic, clinical, or vendor parameters, age- and vendor-specific referenced ranges were established as well.

Read More on PubMed