Fibrosis location, fibrosis burden, and myocardial strain in patients with non-ischemic dilated cardiomyopathy: association with adverse clinical outcomes

Risk stratification in non-ischemic dilated cardiomyopathy (NIDCM) remains difficult once replacement fibrosis is already present. We examined whether the anatomical distribution of late gadolinium enhancement (LGE), the amount of LGE, and cardiovascular magnetic resonance (CMR) derived global strain were associated with adverse clinical outcomes in patients who were all LGE positive. This retrospective single-center cohort included 90 adults with NIDCM and LGE on CMR. Fibrosis burden was expressed as the fibrosis index, which was defined as the percentage of LGE mass relative to the left ventricular mass. Fibrosis location was classified as septal only, LV free-wall only, or combined septal and free wall. Fibrosis pattern was classified as mid-wall, subepicardial, or mixed mid-wall and subepicardial. The primary endpoint was a composite of all-cause death, heart transplantation, sudden cardiac death or aborted sudden cardiac arrest, and appropriate implantable cardioverter–defibrillator (ICD) therapy. Cox proportional hazards regression was the primary analysis. Restricted mean survival time (RMST) was used where the proportional hazards assumption was not met. The primary endpoint occurred in 37 of 90 patients (41.1%). Over a median follow-up of 31 months (interquartile range 17.5 to 47.0), combined septal and free-wall fibrosis was associated with the primary endpoint compared with septal-only fibrosis (hazard ratio [HR] 2.24, 95% confidence interval [CI] 1.06 to 4.74, P = 0.035), and this persisted after adjustment for age and sex (HR 2.30, 95% CI 1.06 to 4.96, P = 0.034). At 36 months, the RMST difference was 3.4 months shorter for combined involvement (95% CI 0.2 to 6.7 months shorter, P = 0.032). Fibrosis index was not associated with the primary endpoint (HR 1.03 per 1% increase, 95% CI 0.97 to 1.08, P = 0.324), and neither were global longitudinal, circumferential, or radial strain. Fibrosis pattern and fibrosis location were concordant in 84 of 90 patients (93.3%, Cramer V 0.911). When appropriate ICD therapy was removed from the composite, the association with combined involvement was no longer statistically significant (HR 1.84, 95% CI 0.74 to 4.59, P = 0.192). Among patients with NIDCM who already have replacement fibrosis, fibrosis involving both the septum and the LV free wall identified a higher-risk subgroup, whereas the quantity of fibrosis and global strain did not. The signal was carried largely by arrhythmic events, and these unconfirmed findings are hypothesis generating.