Replacement and Reactive Myocardial Fibrosis in Idiopathic Dilated Cardiomyopathy: Comparison of Magnetic Resonance Imaging with Right Ventricular Biopsy

The presence of focal myocardial fibrosis, also called scar or replacement fibrosis, detected with late gadolinium enhancement (LGE) cardiovascular magnetic resonance imaging (CMR) is a predictor of adverse outcome in patients with idiopathic dilated cardiomyopathy (DCM).1,2 Increased interstitial (reactive) fibrosis determined with histological staining from endomyocardial biopsy (EMB) specimens has been described in DCM and associated with an adverse prognosis and impaired response to therapeutic interventions in selected groups of patients with DCM.3–6 We sought to describe the patterns of focal fibrosis in a group of patients with DCM and correlate focal fibrosis with interstitial fibrosis, myocardial viral load, and inflammation. A pilot study in 10 autopsies was performed to compare interstitial fibrosis from right ventricular (RV) with left ventricular (LV) EMB. Study subjects were 60 consecutive patients with DCM (globally impaired systolic LV function not due to myocardial infarction, significant coronary artery disease, hypertension, valvular disease, or other known causes) referred to our hospital for EMB and CMR. Patients (n = 1) with infarct-like subendocardial or transmural LGE within the territory of a coronary artery and corresponding regional wall motion abnormalities only were not included to avoid the inclusion of patients with an embolic event. Three patients did not agree to participate resulting in a study population of 56. The study was approved by the Hospital Ethics Committee and all patients gave written informed consent. Endomyocardial biopsy from the RV interventricular septum was obtained with a transcatheter bioptome. The collagen volume fraction (CVF) was quantified by one experienced investigator (R.J.S.) blinded to patient details as percentage Sirius red-stained area per total myocardial tissue area.7 Inflammatory cells were quantified with immunohistochemical staining using CD3, CD4, CD8, CD20, CD45, and CD68 antibodies.8 Deoxyribonucleic acid concentrations of Parvovirus B19, Human Herpes virus 6, and Epstein–Barr virus were determined with quantitative polymerase chain reaction.9,10 Cardiovascular magnetic resonance imaging was performed with a 1.5 Tesla scanner. (i) Steady-state free precession cine images were acquired for functional analysis, (ii) dual-inversion black-blood T2-weighted images with fat suppression for the detection of myocardial oedema, and (iii) breath-hold multislice T1-weighted 3D inversion-recovery gradient-echo images to detect focal myocardial fibrosis 10 min after intravenous administration of 0.2 mmol/kg gadolinium–DTPA.11 Inversion times were adjusted to null normal myocardium (200–300 ms) using a Look–Locker sequence. Endocardial and epicardial LV and RV contours were manually traced on short-axis slices to determine end-diastolic volume (EDV), end-systolic volume, ejection fraction (EF), and mass. Late gadolinium enhancement was visually assessed by two independent readers blinded to clinical details and quantified by manual planimetry. Data are expressed as median values with inter-quartile ranges. Differences in LV and RV EDV, EF, LV mass, CVF, viral load, and amount of inflammatory cells between patients with or without focal fibrosis as determined by LGE CMR were calculated by means of the Mann–Whitney test and differences in age by Student's t-test. Receiver-operating characteristic analysis, with corresponding measures of statistical uncertainty 95% confidence intervals (CI), was applied to the LGE and CVF values used to detect interstitial fibrosis. Correlations were calculated by means of the Spearman's rank correlation method. A pilot study to compare interstitial fibrosis from RV EMB (interventricular septum in accordance with the location of EMB from transcatheter sampling) with LV EMB (anterior wall) was performed in 10 autopsies. Patients with fibrosis due to myocardial infarction or hypertrophy were not included. A good correlation between LV (3.2%, 1.7–5.9) and RV interstitial fibrosis (2.2%, 1.7–5.7) was found (r = 0.8, P = 0.5). Patient characteristics are shown in Table 1. Eighteen patients (32%) had an increased LV mass.12 No patient had signs of oedema or pericardial effusion. Focal LV fibrosis on LGE CMR was observed in 22 patients (39%; median extent 5%, 3–14%). Eight patients showed mid-myocardial, three mid- to epicardial, five mid- to transmural, and six more than one LGE pattern (Figures 1 and 2). No patient had focal RV fibrosis. The presence or extent of focal LV fibrosis as measured with LGE CMR did not correlate with age, LV EDV, LV mass, LV EF, RV EDV, or RV EF. The median CVF was 4.9% (2.1–6.4%). Older patients did not show more interstitial fibrosis than younger patients. The CVF was neither correlated with LV EDV, EF, or mass nor with RV EDV or EF. No differences were found between patients with or without focal fibrosis detected with LGE and the amount of CVF (Figure 3): the presence of focal fibrosis was not related to the amount of interstitial fibrosis (area under the curve = 0.51, 95% CI: 0.35–0.67). The amount of focal fibrosis measured with LGE as percentage of LV was also not correlated with CVF. Myocardial viral load was not correlated with focal or interstitial fibrosis. No correlation was found between interstitial fibrosis and inflammatory cells. The numbers of CD3- (P = 0.003) and CD4- (P = 0.02) positive inflammatory cells were significantly differently distributed between patients with and without focal fibrosis. The main findings of the study are that the presence of focal fibrosis in patients with DCM was (i) related to inflammation, but not to interstitial fibrosis, and (ii) both forms of fibrosis were not related to the severity of RV or LV dilation, dysfunction, LV mass, patient age, and viral load. Thus, focal fibrosis might be linked to inflammation rather than interstitial fibrosis or impairment in systolic function. Cardiovascular magnetic resonance imaging has an established role in the detection of focal fibrosis.1,2,13 Although LGE imaging for the detection of focal fibrosis is robust, fast, implemented into standard clinical MR imaging protocols and the images are easily analysed by viewing, scan sequences for MR imaging of diffuse interstitial fibrosis are at an experimental stage with time-consuming image analysis.14,15 A normal LGE CMR study does not exclude increased interstitial fibrosis. The underlying process of developing focal fibrosis in DCM is currently not known. Wu et al.2 proposed that focal fibrosis reflects the transition from compensated to decompensated state and therefore is associated with an adverse prognosis. Since focal fibrosis was not linked to interstitial fibrosis and ventricular function in our study, we postulate that the formation of focal fibrosis might reflect a different pathophysiological process than the formation of interstitial fibrosis as suggested by the correlation between focal fibrosis and inflammation. Conflict of interest: none declared.