18.2 Strain Interpretation, Global Longitudinal Strain, and Disease Patterns

Keep the sign and timing straight

CCI task D13 is to evaluate strain imaging. Strain is dimensionless fractional deformation: strain (%) = (L − L₀) / L₀ × 100. If a myocardial fiber shortens from 10 cm to 8 cm, longitudinal strain is (8 − 10)/10 × 100 = −20%. More negative LV longitudinal strain means greater systolic shortening; −12% is worse, not better, than −20%. Current ASE guidance advises retaining the negative sign for GLS and segmental values. If a system displays positive magnitudes, call the metric global longitudinal shortening and state the convention rather than mixing signed and absolute numbers.

End diastole and zero strain should correspond to mitral-valve closure, and default GLS is end-systolic strain at aortic-valve closure. Verify automated event timing on 2-D valve motion or spectral Doppler, especially with conduction delay, pacing, ectopy, or marked preejection time. Peak strain after aortic closure is postsystolic shortening, not ordinary end-systolic function. Curve shape and timing can be more informative than the lowest number in regional disease.

Acquire and validate before reading the bull's-eye

Acquire nonforeshortened apical four-, two-, and long-axis views with the true apex and full myocardium visible, a stable ECG, similar heart rate, and frame rate appropriate to heart rate. Narrow sector width and depth while retaining the complete LV. Avoid immediate postectopic beats; in AF, average representative cycles or use the third of three cycles with similar lengths. Record BP and clinical loading conditions because preload and afterload change strain even when contractility does not.

Trace the endocardium and adjust the region of interest to include full wall thickness without pericardium. An ROI that is too broad lowers magnitude; a subendocardial-only ROI can exaggerate it. Play every tracking overlay. Reject segments that leave myocardium, track papillary muscle, jump across scar, or follow reverberation. If more than three LV segments are uninterpretable, do not report GLS. Foreshortening can falsely increase apical strain and manufacture an apical-sparing map. Report image quality, analysis software/version, layer convention, views, excluded segments, and GLS with its sign.

Interpret global values and serial change

For adult LV myocardial GLS, current consensus categorizes more negative than −18% as normal, −16% to −18% as borderline, and less negative than −16% as abnormal. These are contextual ranges, not universal biological walls: age, sex, blood pressure, loading, body size, acquisition, myocardial versus endocardial analysis, and software matter. GLS complements EF because longitudinal dysfunction may appear while EF is preserved, but a reduced value is not an etiology. Three-dimensional strain and routine radial strain remain insufficiently standardized for routine clinical decisions.

Same-patient follow-up is especially useful. A relative GLS magnitude reduction of about 10%–15% is likely significant when technique and load are comparable. Calculate relative worsening = (|baseline GLS| − |follow-up GLS|) / |baseline GLS| × 100. A change from −20% to −17% is a 3-percentage-point change but a 15% relative reduction in magnitude. First review tracking, vendor/version, BP, rhythm, and volume status; then integrate EF, symptoms, biomarkers, treatment exposure, and other imaging. This signal can support possible treatment-related dysfunction but does not alone prove cardiotoxicity.

Do not pool chamber-specific strain metrics. Acquire RV free-wall strain from an RV-focused apical four-chamber view at about 60–90 frames/s and label whether septum is excluded; current lower limits are approximately −20% in men and −21% in women. LA reservoir strain uses dedicated software, nonforeshortened apical views, and R-R gating. Current ASE consensus describes approximately 30%–60% as normal, 23% as the lower limit of normal, and 23%–30% as borderline abnormal rather than simply normal. These values remain load, rhythm, view, and software dependent. A dedicated chamber ROI is preferable to casually adapting LV software, and serial comparisons must preserve gating and chamber definition.

Use disease patterns as clues

Ischemia or infarction may reduce strain in a coronary distribution, with early systolic lengthening and postsystolic shortening; however, postsystolic shortening is sensitive but nonspecific and also occurs with scar, hypertrophy, conduction delay, and loading change. Segmental strain is less reproducible than GLS, so confirm the pattern with wall motion, ECG, perfusion or coronary data, and serial findings.

In increased wall thickness, impaired strain in hypertrophied septal regions may accompany classic HCM, while apical HCM reduces apical strain. Relative apical sparing with worse basal and mid strain raises the possibility of cardiac amyloidosis, particularly when apical mean strain is more than twice the mean of the rest of the LV, but similar patterns can occur with aortic stenosis, renal disease, hypertensive remodeling, or technical foreshortening. Seek morphology and the appropriate CMR, laboratory, or nuclear pathway rather than declaring amyloid from the map.

Stress cardiomyopathy commonly shows mid-apical, midventricular, basal, or focal impairment extending beyond one coronary territory and should recover on serial imaging; acute coronary syndrome and myocarditis remain alternatives. Hypertension often produces a basal-to-apical gradient, and HFrEF often has diffuse reduction. RV free-wall longitudinal strain requires an RV-focused view and dedicated ROI; do not compare it with RV four-chamber strain that includes septum. Across all diseases, report a pattern as supportive, integrate chamber function and load, and preserve diagnostic uncertainty.