Biventricular endomyocardial biopsy in patients with suspected myocarditis: Feasibility, complication rate and additional diagnostic value

Background

Endomyocardial biopsy (EMB) remains the gold standard in the diagnosis of myocarditis. The diagnostic and therapeutic value of this procedure, its safety and the choice of the ventricular side (left versus right) of EMB are still a matter for discussion. Previous retrospective studies have shown that the use of biventricular EMB was associated with a higher diagnostic yield compared to selective univentricular EMB (1, 2). The study of Stiermaier et al. examined prospectively the feasibility, safety, and diagnostic performance of biventricular EMB in patients with suspected myocarditis.  

Methods

A total of 136 patients with suspected myocarditis were investigated. The suspicion of myocarditis was based on symptoms suggesting inflammatory heart disease (e.g. dyspnea, palpitations, chest pain, etc.), signs of recent or ongoing myocardial damage (wall motion abnormalities with left ventricular dysfunction, ECG-abnormalities or elevated troponin), and history of a systemic viral infection after exclusion of significant coronary artery disease by coronary angiography. Biventricular EMB was performed in 132 patients. The specimens underwent histological, immunohistological, and molecular pathological examination according to the World Health Organization/ International Society and Federation of Cardiology Task Force Report (3). Acute myocarditis was diagnosed by evidence of myocyte necrosis and myocardial inflammation based on immunohistochemical proof of ≥ 14 infiltrating immune cells/mm² (CD3+T-lymphocytes and/or CD68+ macrophages). In case of chronic myocarditis (formerly called borderline myocarditis), acute myocyte injury could not be detected, but inflammation and interstitial fibrosis became evident. All EMB were tested for viral genome by nested (reverse transcriptase-) polymerase chain reaction. Evidence of viral genomes in the absence of inflammation was diagnosed as latent virus persistence. Healed myocarditis was characterized by multifocal fibrosis/scarring without inflammation. Dilated cardiomyopathy was primarily diagnosed in association with additional angiographic and CMR data.

Results

Out of the 136 investigated patients with suspected myocarditis, in 4 patients LV-EMB was not performed (in 3 patients due to left ventricle thrombus and in 1 patient due to pericardial tamponade, which occurred during RV-EMB). In 5 patients, EMB of either the left or right ventricle were insufficient for adequate histopathological analyses, resulting in 127 patients (96%) with completely investigated biventricular EMB.

A median of 6 samples were taken from each ventricle. Out of 136 RV-EMB, one major complication occurred (pericardial tamponade requiring surgical revision) corresponding to a major complication rate of 0.7%. During left ventricular biopsy, no major but two minor complications (supraventricular arrhythmia, atrioventricular block) were documented.

Myocarditis was diagnosed in 97 patients (71.3%): chronic myocarditis in 92 subjects (67.6%) and acute myocarditis in 5 patients (3.7%). Out of 127 patients with BV-EMB, myocarditis was diagnosed in 89 of these subjects (70.1%)

While 67 patients (75.3%) fulfilled the diagnostic criteria of myocarditis in both ventricles, myocarditis was diagnosed in the left ventricle only in 16 patients (18%) and in the right ventricle in 6 patients (6.7%), respectively. Consequently, selective RV/LV-EMB would have resulted in missing the diagnosis in 18% and 6.7% of the cases, respectively. In 37 of 38 patients with the final diagnosis other than myocarditis, LV- and RV-EMB demonstrated similar results which led to a congruent diagnosis.

A latent virus persistence without myocardial inflammation was shown in 9.6% (n=13) of all  patients in whom EMB was performed and a healed myocarditis in 2 cases (1.5%). The diagnosis of a dilated cardiomyopathy was confirmed in 14% of the patients. Hypertrophic cardiomyopathy was documented in 2 patients and an amyloidosis in one subject. In 2 patients (1.5%), no pathological findings were documented. Viral genome was detected in 45 patients (36%) (mainly PVB 19 and HHV6) with a single LV presence in 10 patients (22.2%) and a single RV presence in 3 patients (6.7%). As a result, selective RV- or LV-EMB would have missed viral genomes in 22.2% and 6.7 % of the patients, respectively. In 3 patients with PVB 19 genome-detection in both ventricles, RV-EMB led to the additional finding of EBV in 2 patients and of HHV6 in one subject.

The detection of infiltrating immune cells (CD3+ T-lymphocytes and CD68+ macrophages) correlated highly in the overall cohort and in patients with evidence of inflammation in both ventricles but not in patients with single ventricular evidence of myocarditis.

Conclusion

This prospective study shows that biventricular EMB performed by experienced interventionalists is a safe procedure with a low complication rate. BV-EMB improves the diagnostic yield of EMB especially for the detection of myocarditis or viral genome. Selective LV-EMB seems to be superior to selective RV-EMB to confirm or exclude a suspected myocarditis.

Comments

The non-invasive diagnosis of myocarditis and other infiltrative or storage cardiac diseases, such as amyloidosis can be challenging. Endomyocardial biopsy, however, remains  the gold standard in the diagnosis of inflammatory heart diseases.

The present prospective study of Stiermeier et al. demonstrated the diagnostic superiority of biventricular EMB in comparison to a selective right or left ventricular biopsy and examined the safety of a biventricular biopsy procedure.

These results are in line with the retrospective study by Yilmaz et al. (1), in which myocarditis was diagnosed in both ventricles in 73.4% of patients with biventricular EMB. Presence of myocarditis was detected in 18.7% of the LV-EMB and in 7.9% of the RV-EMB only. Considering also other cardiac diseases as amyloidosis, sarcoidosis as well as hypertrophic und restrictive cardiomyopathy, diagnostic EMB results were achieved significantly more often in patients with biventricular EMB (79.3%) compared to those who underwent a selective LV- or RV-EMB (67.3%). The major complication rate (hemopericardium/tamponade with pericardiocentesis or stroke) for LV-EMB was 0.64% and for RV-EMB 0.82%, respectively.

In the study of Chimenti C et al. (2), the periprocedural major complication rate (perforation with/without cardiac tamponade, embolization) was 0.33% for LV-EMB and 0.45% for RV-EMB out of 4221 patients who underwent diagnostic EMB because of suspected myocarditis or other non-ischaemic cardiomyopathies. In 2396 patients who received a biventricular myocardial biopsy, LV-EMB showed key histopathological findings in 96.3% of the patients, whereas in case of RV-EMB diagnostic histopathological  results were documented in 71.4% only.

Stiermeier et al. also confirmed the safety of biventricular EMB with a quite low complication rate, which appears to be under 1% in the hands of experienced interventionalists.

Interestingly, Chementi C et al. (2) were able to show in their study that the diagnostic yield of LV-EMB in case of echocardiographically-proven structural and functional abnormalities (e.g. increased wall thickness, ventricular dilation or hypokinesia) solely in the left ventricle is significantly higher when compared with RV-EMB (98% versus 53%). Whenever both ventricles were affected, a diagnosis could be made in 98.1% of the LV-EMB and 96.5% of the RV-EMB, respectively.

To determine which ventricle is affected and should subsequently be biopsied, a Cardiovascular Magnetic Resonance (CMR) study performed before EMB might be helpful. In the study by Yilmaz et al. (1), no diagnostic benefit of EMB taken from regions showing late gadolinium enhancement (LGE) could be documented. Yet, it remains to be consider that other relevant hyperemia- and edema-detecting sequences that are included in the Lake Louis criteria for the diagnosis of myocarditis (4), were not carried out in all these patients. For this reason, a prospective study investigating the value of a CMR guided biopsy adhering to the Lake Louise criteria is desirable. The question whether increasing the number of specimen from the right or left ventricle leads to a change in the diagnostic value or to a higher complication rate should also be investigated in future studies.