Skip to content


Open Access

Long-term prognostic significance of late gadolinium enhancement in non-ischemic dilated cardiomyopathy: further evidence from 184 patients

  • Stephanie Lehrke1,
  • Michael Schöb1,
  • Henning Steen1,
  • Helmut Kemmling1,
  • Dirk Lossnitzer1,
  • Philipp Ehlermann1,
  • Constanze Merten1,
  • Grigorius Korosoglou1,
  • Evangelos Giannitsis1 and
  • Hugo A Katus1
Journal of Cardiovascular Magnetic Resonance201012(Suppl 1):O8

Published: 21 January 2010


Cardiovascular Magnetic ResonanceCardiac DeathDilate CardiomyopathyLate Gadolinium EnhancementComposite Endpoint


Non-ischemic dilated cardiomyopathy (DCM) is a major cause for cardiovascular morbidity and premature mortality. Due to the variable clinical course, risk stratification is of paramount importance in these patients to identify those most likely to benefit from aggressive treatment strategies. There is emerging evidence for the prognostic significance of late gadolinium enhancement (LGE) in patients with DCM.


The goal of this study was to further investigate the long-term prognostic significance of LGE in a large cohort of patients with DCM presenting to a tertiary referral center.


Contrast-enhanced cardiovascular magnetic resonance (CE-CMR) was performed in 184 consecutive patients with DCM on a 1.5 T clinical scanner. Significant coronary artery disease had been ruled out in all patients. LV volumes and mass were derived from SSFP cine images. Presence of LGE was determined by two independent observers. Patients were followed for the primary endpoint of cardiac death and a composite endpoint of cardiac death, hospitalization for decompensated heart failure or appropriate ICD firing for a mean of 685 ± 30 days.


LGE was detected in 72/184 patients (39%). Patients with LGE showed a higher NYHA classification (2.2 ± 0.09 vs. 1.9 ± 0.07, p = 0.02) and were more often on oral diuretics (42/72 pts. vs. 45/112 pts., p = 0.02). Presence of LGE was associated with a lower EF (31.6 ± 1.7% vs. 40.4 ± 1.2%, p < 0.001), higher EDV (288.7 ± 12.2 ml vs. 233.6 ml, p < 0.001) and a more pronounced increase of LV mass/BSA (81.9 ± 2.9 g/m2 vs. 70 ± 2.2 g/m2, p < 0.001). There was a trend towards a higher incidence of the primary endpoint in patients with LGE (4/72 vs. 1/112, p = 0.06). The rate of the composite endpoint was significantly associated with the presence of LGE (15/72 vs. 6/112, p = 0.002, Figure 1). When entered into multivariate Cox regression analysis, LGE did not retain an independent predictive value (Table 1).
Figure 1
Figure 1

Event-free survival according to presence of LGE.

Table 1

Predictors of the composite endpoint by Cox regression


HR (95% CI)




EF (%)

0.94 (0.91-0.98)

< 0.001

LGE (y/n)

3.5 (1.4-9)


Age (yrs)

1.06 (1.02-1.09)


Male gender

1.29 (0.43-3.83)


NYHA class

1.84 (0.98-3.46)


Multivariate model


Age (yrs)

1.04 (1.02-1.09)


EF (%)

0.94 (0.91-0.97)


Authors’ Affiliations

Medizinische Universitätsklinik Heidelberg, Heidelberg, Germany


© Lehrke et al; licensee BioMed Central Ltd. 2010

This article is published under license to BioMed Central Ltd.