Feasibility of Left Ventricular Global Longitudinal Strain Measurements from Contrast-Enhanced Echocardiographic Images.

BACKGROUND

Although left ventricular global longitudinal strain (GLS) is an index of systolic function recommended by the guidelines, poor image quality may hamper strain measurements. While contrast agents are commonly used to improve endocardial visualization, no commercial speckle-tracking software is able to measure strain in contrast-enhanced images. This study aimed to test the accuracy of speckle-tracking software when applied to contrast-enhanced images in patients with suboptimal image quality.

METHODS

We studied patients with a wide range of GLS values who underwent transthoracic echocardiography. Protocol 1 included 44 patients whose images justified use of contrast but still allowed noncontrast speckle-tracking echocardiography (STE), which was judged as accurate and used as a reference. Protocol 2 included 20 patients with poor image quality that precluded noncontrast STE; cardiac magnetic resonance- (CMR-) derived strain was used as the reference instead. Half the manufacturer recommended dose of a commercial contrast agent (Definity/Optison/Lumason) was used to provide partial contrast enhancement. Higher than normal mechanical indices (0.6-0.7) and lowest frequency range for maximal penetration settings were used for imaging. GLS was measured (Epsilon) with and without contrast-enhanced images and by CMR-derived feature tracking (TomTec). Comparisons included linear regression and Bland-Altman analyses.

RESULTS

The contrast STE analysis failed in 4/64 patients (6%). Manual corrections were needed to optimize tracking with contrast in all patients. GLS measurements were in good agreement between contrast and noncontrast images (r = 0.85; mean GLS in the contrast images, -12.9% ± 4.7%; bias, 0.34% ± 2.4%). Good agreement was also noted between contrast STE- and CMR-derived strain (r = 0.83; mean, GLS -13.5% ± 4.0%; bias, 0.72% ± 2.5%).

CONCLUSIONS

We found that GLS measurements from contrast-enhanced images are feasible and accurate in most patients, even in those with poor image quality that precludes strain measurements without contrast enhancement.