Temporal Variability of Global Longitudinal Strain in Stable Patients Undergoing Chemotherapy With Trastuzumab

Global longitudinal strain (GLS) is used to evaluate left ventricular (LV) performance after chemotherapy. Differentiating between reduction in GLS due to clinical change and normal temporal variability in measurement remains a challenge. We quantified interobserver, test-retest variability of GLS by expert observers in relation to variability of GLS quantified for clinical assessment by sonographers in our laboratory. We examined the temporal variability of GLS in 30 patients with normal LV ejection fraction (LVEF >53%) undergoing chemotherapy in the absence of change in medications and clinical symptoms in up to 5 sequential echocardiograms. GLS was quantified using EchoPAC (GE Healthcare, Milwaukee, Wisconsin) and 2-dimensional biplane LVEF was measured from 4- and 2- chamber views. Interobserver test-retest variability of GLS measured in 10 random patients by 2 expert readers was calculated using a one-way analysis of variance. Square root of mean squared error provided the SEM for temporal variability. Baseline LVEF was 59.3 ± 5.1% and remained relatively unchanged over 12 months, p = 0.87. Temporal variability of GLS measured by sonographers was 1.28% and similar to interobserver test-retest variability of GLS measured by expert observers, 1.12% (p = 0.17). Maximum detectable difference in GLS measured by expert observers was similar to that derived from sequential measurements of GLS (3.2% vs 3.6%, respectively). Temporal variability of GLS among clinically stable patients is 1.28% and similar to interobserver test-retest variability of 1.12% measured by expert observers. In conclusion, a reduction in strain >3.2% during sequential echocardiograms under these conditions may be significant.