Can right ventricular endomyocardial biopsy predict left ventricular fibrosis beforehand in dilated cardiomyopathy?

Abstract

Aims

Myocardial fibrosis of the left ventricle (LV) is a prognostic factor in dilated cardiomyopathy (DCM). This study aims to evaluate whether fibrosis of right ventricular (RV) endomyocardial biopsy (EMB) can predict the degree of LV fibrosis beforehand in DCM.

Methods and results

Fibrosis extent in 70 RV‐EMB specimens of DCM diagnosis was compared with that in the whole cross‐sectional LV of excised hearts in the same patients (52 explanted hearts for transplant and 18 autopsied hearts). The median interval between biopsy and excision was 4.1 (0.13–19.3) years. The fibrosis area ratio of the EMBs and excised hearts were evaluated via image analysis. The distribution of cardiovascular magnetic resonance–late gadolinium enhancement (LGE) in the intraventricular septum was classified into four quartile categories. The fibrosis area ratio in RV‐EMB correlated significantly with that in the short‐axis cut of the LV of excised hearts (r = 0.82, P < 0.0001) and with a diffuse pattern of LGE (r = 0.71, P = 0.003). In a multivariate model, after adjusting for the interval between biopsy performance and heart excision, the fibrosis area ratio in RV‐EMB was associated with that in LV‐excised heart (regression coefficient, 0.82; 95% confidence interval, 0.68–0.95; P < 0.0001).

Conclusions

The fibrosis observed in RV‐EMB positively correlated with the extent of fibrosis in the LV of excised hearts in patients with DCM. The study findings may help predict LV fibrosis, considered a prognostic factor of DCM through relatively accessible biopsy techniques.

Introduction

Idiopathic dilated cardiomyopathy (DCM) is a primary cardiac muscle disease characterised by significant dilatation of both ventricles and wall thinning with poor contractility. Myocardial fibrosis of the left ventricle (LV) is a major key factor in determining DCM prognosis. ¹ , ² , ³ , ⁴ Cardiovascular magnetic resonance (CMR) is the gold standard for the non‐invasive assessment of myocardial fibrosis; however, it cannot sufficiently assess diffuse interstitial fibrosis, which is common in non‐ischaemic DCM. ⁵ , ⁶ Conversely, histopathological assessment enables qualitative macroscopic evaluation of myocardial fibrosis. ⁷ Endomyocardial biopsy (EMB) provides important information such as myocardial inflammation, infiltration, or storage disease that cannot be identified by other means to exclude secondary cardiomyopathy. Then, it plays a role in confirming the presence of non‐ischaemic myocardial fibrosis, providing prognostic information in patients with DCM. ⁸ , ⁹ However, the insight into whether right ventricular (RV) endomyocardial biopsy (EMB) specimens accurately reflect LV histology in DCM remains unclear. ¹⁰ Hence, this retrospective observational study aimed to investigate the validity of using RV‐EMB specimens to predict LV fibrosis beforehand in DCM. According to the availability of a relatively high number of excised hearts and EMB samples from the same patients with DCM, we hypothesised that RV‐EMB histology can predict the extent of LV fibrosis, indicating efficacy for DCM diagnosis.

Methods

Study design

This retrospective, single‐centre, nonrandomised, subserial observational study examined patients with DCM who underwent heart transplantation or autopsy at the National Cerebral and Cardiovascular Center in Japan to determine whether RV‐EMB specimens correspond to LV fibrosis in transplanted or autopsied heart. Patient enrolment in this study is shown in Figure 1 . We included all patients with ventricular dilation and systolic dysfunction without coronary artery disease who underwent RV‐EMB before heart transplantation or autopsy and were ultimately diagnosed with DCM by histologically excluding secondary cardiomyopathy. Initially, we reviewed data from 59 patients with DCM who received heart transplantation for severe heart failure symptoms unresponsive to maximal pharmacological therapy or mechanical support and 72 DCM patients who underwent autopsy between November 1981 and December 2021 to determine the cause of death. Four patients who had undergone cardiac intervention (1 case of coronary artery bypass surgery, 1 case of percutaneous coronary intervention, 2 cases of mitral valve replacement) after EMB were excluded. We also excluded 57 patients who had not undergone EMB at diagnosis and finally included 70 cases (52 explanted hearts for heart transplant and 18 autopsy cases) to correlate RV fibrosis in EMB with LV fibrosis in explanted hearts in the same patients. Nine (50%) of 18 autopsied cases had been registered or considered for heart transplantation before death. The causes of death among all autopsied cases were attributed to worsening heart failure. Table S1 lists the details of autopsy cases.

Figure 1.

Identification of the study population. Flowchart detailing the identification, inclusion, and exclusion of patients. CMR, cardiovascular magnetic resonance; CABG, coronary artery bypass; DCM, dilated cardiomyopathy; LGE, gadolinium myocardial enhancement; PCI, percutaneous coronary intervention.

This study was approved by the ethics committee of the institution (approval No.: R21052) and was conducted in accordance with the principles of the 1964 Declaration of Helsinki, as revised in 2008. Data use was explained on our websites, and the ‘opt‐out’ approach for consent was approved.

Pathological analysis of the endomyocardial biopsy and excised hearts

All transvenous RV biopsies were obtained from the RV aspect of the intraventricular septum (IVS). The fibrosis in EMB was classified as interstitial, perivascular, and replacement fibrosis. Fibrosis with a diffuse distribution within the interstitium indicated interstitial fibrosis, whereas scarring fibrosis corresponds to the replacement of depleted myocytes following cell damage or necrosis defined replacement fibrosis. ⁵ The whole hearts were fixed and sliced by a short‐axis cut at 10 mm intervals from the apex to the base. Using Masson's trichrome staining at the papillary muscle level, we evaluated the fibrosis and its extent microscopically. Subepicardial adipose tissue, including the vessels and pericardium, was excluded from the analysis to avoid overestimating fibrosis. We retrieved the measurement data of the wall thickness of each part (anterior, lateral, posterior wall, and IVS) before fixation. Two pathologists (K. A. and K. H.) who were blinded to the clinical data conducted the histological evaluation.

Histological quantitative analysis

The extent of fibrosis was quantitatively evaluated using a histological analysis method and a dedicated image processing software (WinROOF®, Mitani Corp., Tokyo, Japan) ¹¹ to identify the fibrosis area, which was highlighted in blue using Masson's trichrome staining. ¹² From each RV‐EMB stained Masson's trichrome staining section, all fields of view were imaged using a NanoZoomer‐XR scanner® (Hamamatsu Photonics, Hamamatsu, Japan) at 20‐fold magnification mode scan mode that is virtual whole slide imaging. Excised hearts scanned Masson's trichrome staining section sliced by a short‐axis cut at the papillary muscle level were photographed. Then, all whole slide images were transferred to an analysis platform image processing software for analysis. A diagram of quantitative fibrosis analysis from the Masson's trichrome stained slice. The total myocardial area and fibrosis area marked with blue were obtained semi‐automatically, and the percentage of fibrosis area was automatically determined with the Positive Pixel Count algorithm by dedicated image processing software as the ratio of the total blue‐stained area to the whole myocardial area excluding the endocardium.

The excised hearts were assessed by examining the LV and IVS of a circumferential cross‐sectional heart, respectively (Figure  2 ). IVS includes RV aspect corresponding to usual EMB site. All measurements were blinded to the clinical data.

Figure 2.

Histological quantitative analysis of myocardial fibrosis. (A) RV biopsy specimen; (D) Explanted heart was sliced in transverse and parallel sections at the mid‐wall level. The image processing software classifies the specimen into fibrosis (green) in the RV‐EMB (C) and excised heart (F) corresponding to the IVS (head arrow) or LV of myocardial fibrosis (in blue at the Masson's trichrome staining in the RV‐EMB [B] and in LV in the excised heart [E]). EMB, endomyocardial biopsy; IVS, intraventricular septum; LV, left ventricle; RV, right ventricle.

Cardiovascular magnetic resonance imaging and image analysis

CMR was performed on a 1.5 T scanner (Magnetom Sonata; Siemens, Erlangen, Germany) with a standardised clinical protocol. Late gadolinium enhancement (LGE) imaging was acquired 10–15 min after injecting 0.15 mmol/kg concentration of gadolinium‐based contrast medium (Magnevist; Schering AG, Berlin, Germany), with an inversion‐recovery steady‐state free precession pulse sequence and an inversion time of 300 ms (pixel size of LGE; 1.5 × 1.9 mm). The extent of fibrosis assessed by CMR‐LGE of the mid‐wall IVS pattern was classified into four quartile categories based on LGE distribution: (i) focal, involving less than one‐quarter of the IVS; (ii) linear focal, involving more than one‐quarter and less than one‐half of the IVS; (iii) focal linear, involving more than one‐half and less than three‐quarters of the IVS; and (iv) diffuse, involving more than three‐quarters of the IVS. One certified radiologist blinded to pathologic findings independently evaluated the LGE images for the presence of abnormal enhancements.

Statistical analysis

The statistical data were analysed using JMP Pro version 16.0.1 (SAS Institute Inc., Cary, North Carolina). Considering that most continuous variables were not normally distributed, they are presented as median (interquartile range [IQR]); they were compared using the Mann–Whitney U test or Wilcoxon signed‐rank test, as appropriate. Categorical variables, which are expressed as frequency (%), were compared with χ ² statistics or Fisher's exact test. The clinical and morphological factors associated with LV fibrosis extent were identified using multivariate logistic and linear regression models, which are expressed as a regression coefficient with 95% confidence interval (CI). Variables for the model were selected according to the histological relationship to LV fibrosis extent. A two‐tailed P < 0.05 was considered statistically significant.

Results

Patient and clinical characteristics

Table 1 describes the patients' characteristics. Of these 70 patients, 52 (74.3%) had their hearts explanted for heart transplant and 18 (25.7%) had their hearts autopsied. The median age of patients who underwent heart excision was 48 (from 17 to 84) years, with 45.

Table 1.

Patient and clinical characteristics

	Patients with DCM (n = 70)
	All case	Transplanted case	Autopsied case	P
Number, n (%)	70 (100)	52 (74.3)	18 (25.7)	–
Female, gender, n (%)	12 (17.1)	10 (19.2)	2 (11.1)	0.44
Age at heart transplantation or death, years	48 [36.5–59.3]	45 [32.3–56.0]	58.5 [40.8–71.3]	0.005
Body mass index, kg/m2	21.4 [18.4–24.8]	21.4 [18.5–25.0]	20.2 [17.0–24.6]	0.65
Interval EMB to excised heart, years	4.1 [2.9–6.2]	4.1 [3.0–5.5]	4.3 [1.1–9.7]	0.78
Family history of DCM or SCD, n (%)	14 (20.0)	12 (23.1)	2 (11.1)	0.25
Sinus rhythm, n (%)	54 (77.1)	44 (84.6)	10 (55.6)	0.02
Echocardiogram evaluation at EMB
Left ventricular ejection fraction, %	18.0 [14.8–23.0]	18.0 [14.5–23.0]	18.5 [14.5–24.3]	0.95
Left ventricular end‐diastolic diameter, mm	70.0 [65.0–80.3]	71.5 [66.0–81.8]	67.0 [56.8–79.3]	0.09
Left ventricular end‐systolic diameter, mm	64.0 [57.9–73.3]	66.0 [59.0–77.0]	60.9 [53.8–71.5]	0.11
Left atrial diameter, mm	46.0 [41.0–52.0]	47.0 [41.0–52.0]	45.0 [36.3–55.5]	0.48
Mitral regurgitation: moderate or severe, n (%)	34 (48.6)	26 (50.0)	8 (44.4)	0.68
CMR evaluation, n (%)	15 (21.4)	12 (23.1)	3 (16.7)	0.52
Cardiac resynchronization therapy, n (%)	28 (40.0)	19 (36.5)	9 (50.0)	0.38
Left ventricular assist device, n (%)	59 (84.3)	51 (98.1)	8 (44.4)	<0.001

Values are n (%) or median [first quartile, third quartile]. P value = autopsied case versus transplanted case.

CMR, cardiovascular magnetic resonance; DCM, dilated cardiomyopathy; EMB, endomyocardial biopsy; SCD, sudden cardiac death.

years in those with explanted hearts and 59 years in those with autopsied hearts. The median interval between biopsy and excision was 4.1 (0.13–19.3) years: 4.1 years in those with explanted hearts and 4.3 years in those with autopsied hearts. Furthermore, 14 (20.0%) patients had a family history of DCM or sudden cardiac death, and 59 (84.3%) had previously received an LV assist device implanted in the cardiac apex: 98% of explanted hearts and 44% of autopsied hearts (P < 0.001). The median LV ejection fraction was 18% (6.3–49%), and 34 (48.6%) patients presented with moderate or severe mitral regurgitation. The mean ventricular wall thickness at the mid‐wall level was 8.7 ± 2.1, 8.4 ± 2.4, 7.3 ± 2.2, 7.3 ± 2.5, and 2.8 ± 1.1 mm in the IVS and LV anterior, LV lateral, LV posterior, and RV, respectively. Fibrosis extent of RV‐EMB was negatively associated with the mean ventricular wall thickness at the mid‐wall level in the LV lateral (P = 0.042). The fibrosis extent of RV‐EMB was not statistically associated with the mean ventricular wall thickness in the IVS and LV anterior, LV lateral, LV posterior, and RV.

Histological analysis findings in the endomyocardial biopsy and excised heart

In RV‐EMB specimens, interstitial, perivascular, and replacement fibrosis were found in 70 (100%), 39 (56%), and 29 (41%) patients, respectively. The percentage of fibrosis extent in EMBs and excised hearts in all patients was plotted in ascending order (Figure  3 ). The median fibrosis area ratio of RV‐EMB, IVS, and LV of the whole heart was 16.1% (5.7–47.3%), 14.2% (5.2–46.9%), and 14.9% (6.4–42.3%), respectively. The fibrosis extent in RV‐EMB was greater in autopsy cases than in transplanted cases (22.4 [15.6–26.0] vs. 14.0 [10.3–18.2], P = 0.035) (Table  2 ). The median fibrosis area ratio of RV free wall and RV‐IVS of excised hearts was 10% (7–15%) and 15% (11–21%), respectively. The fibrosis extent of RV‐EMB was more prominent in our patients with DCM compared with the reported DCM cases that did not progress to transplant or autopsy; the median fibrosis area ratio: 16.1% (5.7–47.3%) vs. 8.6% (5.3–12.6%). ¹³ In our cases, the interval between biopsy and excision of transplanted or autopsied was not significantly associated with fibrosis extent of RV‐EMB (P = 0.06 and P = 0.10, respectively). Strong correlations were noted between the fibrosis extent in EMB and IVS or LV of the whole heart (r = 0.76, P < 0.0001 and r = 0.82, P < 0.0001, respectively) (Figure  4 ). The fibrosis extent in IVS strongly correlated with the LV of the whole heart (r = 0.88, P < 0.0001). The fibrosis extent of RV‐EMB vs. RV‐IVS of excised hearts was not significantly changed during follow‐up (the median fibrosis area ratio: 16.1% vs. 15.0%, P = 0.56).

Figure 3.

Fibrosis extent in the RV‐EMB, LV, and IVS of 70 patients with dilated cardiomyopathy. EMB, endomyocardial biopsy; IVS, intraventricular septum; LV, left ventricle; RV, right ventricle.

Table 2.

Fibrosis extent in 70 patients with dilated cardiomyopathy

	Patients with DCM (n = 70)
	All case	Transplanted case	Autopsied case	P
Number, n (%)	70 (100)	52 (74.3)	18 (25.7)	–
Fibrosis extent, %
RV‐EMB, %	16.1 [5.6–47.3]	14.0 [10.3–18.2]	22.4 [15.6–26.0]	0.035
Cross‐sectional LV of excised heart
IVS, %	14.2 [5.2–46.9]	13.6 [9.6–20.8]	17.3 [12.5–22.5]	0.17
LV, %	14.9 [6.4–42.3]	13.9 [10.9–22.7]	20.3 [14.1–27.1]	0.24

DCM, dilated cardiomyopathy; EMB, endomyocardial biopsy; IVS, intraventricular septum; LV, left ventricle; RV, right ventricle.

Values are n (%) or median [first quartile, third quartile]. P value = autopsied case versus transplanted case.

Figure 4.

Positive correlations between fibrosis extent in the RV‐EMB and LV and IVS. Fibrosis extent in the right ventricular endomyocardial biopsy correlated with those in the left ventricle (A) and intraventricular septum (B). EMB, endomyocardial biopsy; IVS, intraventricular septum; LV, left ventricle; RV, right ventricle.

Factors associated with left ventricular fibrosis extent

Using the fibrosis area ratio to indicate the fibrosis extent, we found in a multivariate linear regression model that the fibrosis area ratio of RV‐EMB (regression coefficient, 0.82; 95% CI, 0.68–0.95; P < 0.0001) was significantly associated with the extent of LV fibrosis. Conversely, variables such as the interval between biopsy and excision, age, and LV assist device use were not significantly associated with LV fibrosis.

Radiologic‐pathologic correlation between late gadolinium enhancement and histological fibrosis

Fifteen patients underwent CMR‐LGE before transplantation or autopsy (12 explanted hearts and 3 autopsied cases). CMR was performed at the time of RV‐EMB in 10 Patients. CMR was performed before RV‐EMB in four patients. One patient underwent CMR after RV‐EMB. The degree of LGE of the mid‐wall IVS in the excised hearts positively correlated with RV‐EMB fibrosis extent (r = 0.71, P = 0.003) (Figure 5 A ). Table S2 lists the histological findings and CMR‐LGE pattern in 15 patients, and Figure 5 B,C shows the representative images of the radiologic‐pathologic correlation. One patient showed a discrepancy between the LGE of the IVS and the corresponding histological findings; the excised heart exhibited more advanced fibrosis in the lateral wall than in the IVS.

Figure 5.

Validation of cardiovascular magnetic resonance imaging pattern on the RV‐EMB in 15 DCM cases with LGE. Fibrosis extent in the RV‐EMB positively correlated with the CMR‐LGE pattern (A). An autopsied heart of a 50‐year‐old man: (B) Presence of mid‐wall diffuse LGE in IVS (white arrows). (C) Fibrosis extent was 27.6%, indicating a severe fibrosis with replacement fibrosis in the RV‐EMB. (D) At autopsy, the extent of LV fibrosis was 24.7%. CMR, cardiovascular magnetic resonance; EMB, endomyocardial biopsy; IVS, intraventricular septum; LGE, late gadolinium enhancement; LV, left ventricle.

RV, right ventricle.

Discussion

To our knowledge, this study is the first to assess the fibrosis degree of correlation between RV‐EMB specimens and LV‐excised hearts in the same patients diagnosed with DCM. We found that RV‐EMB specimens can fairly predict beforehand the expected degree of LV fibrosis at the papillary muscle level evaluated via image analysis. In the RV‐EMB and LV excised hearts of our patients with DCM, myocardial fibrosis was extensive, averaging 17.5% ± 8.7% and 18.1% ± 8.6%, respectively. The fibrosis extent was higher in our patients with DCM than the reported normal fibrosis % values. Miles et al. had reported predicted quantification of normal collagen % values as averaging 15.2% (95% CI, 13.2–17.5) and 9.5% (95% CI, 8.2–10.9), respectively, for cardiac tissue composition within the RV and LV from a noncardiac death. ¹⁴

CMR studies have shown that LV fibrosis, especially the image of LGE, is a predictor of DCM mortality. ¹ , ³ , ¹⁵ CMR‐LGE patterns in DCM are typically mid‐wall, patchy focal, and subendocardial in the septum or RV‐LV insertion points. ¹⁶ , ¹⁷ LGE areas in cardiomyopathies correspond to the histopathologic findings of interstitial expansion, including fibrosis, abnormal protein deposition, inflammatory cells, and cardiomyocyte necrosis. ⁵ However, current CMR techniques cannot detect diffuse microscopic fibrosis, such as diffuse interstitial fibrosis; therefore, most patients with DCM will have absent enhancement. ⁵ LGE is seen in only 30% of patients with DCM. ³ Recently, this limitation has been found to be overcome by T1 mapping imaging characterising the extracellular volume fraction; thus, diffuse interstitial fibrosis may already be detected. ¹⁸ , ¹⁹ According to the American Heart Association and the European Society of Cardiology, EMB is considered as a class IIa recommendation, ²⁰ , ²¹ but it is helpful in diagnosing non‐ischaemic cardiomyopathies, especially fibrosis. ²² , ²³ Although the insight into whether RV‐EMB histology obtained from the ventricular septum can reflect the LV ones remains controversial, the differences in the quality of diagnostic RV‐EMB and LV‐EMB are not substantial in patients with myocarditis. ¹⁰ Myocardial fibrosis in DCM is characterised by interstitial and replacement fibrosis, ⁶ but hypertrophic cardiomyopathy is more likely to present with replacement fibrosis, which can be easily recognised by CMR imaging of myocardial fibrosis quantification in the LV mid‐wall. ²⁴ Myocyte loss or myocardial injury generally leads to replacement fibrosis. In contrast, interstitial fibrosis is an adaptive process to hemodynamic and humoral factors caused by heart failure. ⁶ , ²⁵ Extracellular matrix proteins or fibrosis‐related molecules are released into the bloodstream, causing fibrosis. ²⁵ While the EMB histopathology of DCM usually has nonspecific findings, such as hypertrophy and vacuolar changes of myocytes, we can directly evaluate fibrosis, including diffuse interstitial fibrosis that is limited for the assessment by CMR, around cardiomyocytes and foci of micro‐scarring. ²⁶ , ²⁷ Importantly, EMB specimens provide detailed findings of fibrosis. Thus, our study aimed to determine the validity of using RV‐EMB specimens to predict LV fibrosis in DCM. In a previous report, a larger LV diastolic diameter has less interstitial fibrosis in DCM, suggesting that mechanically, the LV may be challenging to dilate due to the stiffness caused by an excess collagen fibre volume. ²⁸ Our results showed that the extent of fibrosis in RV‐EMB correlated with that in LV; however, the fibrosis pattern alone did not correlate with fibrosis in the whole LV. As suggested by CMR studies, fibrosis burden as determined by biopsy serves as a surrogate marker for LV fibrosis in DCM. Our study is unique because we focused on the histological evaluation of fibrosis at different locations and intervals in the same patients. In this study of DCM cases, the RV‐EMB obtained from the septum can possibly reflect LV fibrosis because the IVS wall was thicker than that of the lateral or posterior walls. In addition, 84% of our patients were treated with an LV assist device; however, using this device did not affect the correlation between RV‐EMB fibrosis and LV fibrosis evaluated via image analysis using Masson's trichrome staining. LV assist devices reportedly affect the remodelling of a failing heart, resulting in changes in myocardial fibrosis. ²⁹ , ³⁰ In the reported CMR study, fibrosis findings progress (median interval 1.5 years, increase >1.8% of LV mass or newly found fibrosis) in 18% of DCM patients. ³¹ In our cases, 4 cases (5.7%) had progression, the proportion of more than 10% RV‐IVS fibrosis at transplantation or autopsy from RV‐EMB fibrosis at biopsy. In our study, the reason for the few progression cases might be a result of the unloading effect of LV assist devices.

Our finding, that is, using EMB as a surrogate marker for fibrosis assessment, could be clinically helpful in developing strategies for validating myocardial fibrosis biomarkers in DCM management. Explanted and autopsy samples have been shown to facilitate the advancement of imaging pathology. ³² Our results provide crucial insights into myocardial fibrosis assessment using EMB in patients with DCM, enabling the identification of patients at risk of progression and potentially allowing for intervention in the pathogenetic mechanisms through the use of heart failure treatments, such as the implementation of implantable cardioverter‐defibrillators or mechanical supports, even in patients without contrast‐enhanced CMR‐LGE findings.

Study limitations

First, considering that the EMB samples were obtained from the RV septum, the pathological information is limited to the area of some millimetre squares of subendocardial cardiac tissue. EMB samples have sampling error that restricts the biopsy accuracy in the case of localised fibrosis. Second, all of the cases were transplanted or autopsied, thereby only representing patients in the end stage of DCM. Third, we assessed fibrosis distribution in circumferential short‐axis cuts at the mid‐wall of the papillary muscle levels. Fourth, not all patients underwent genetic testing for DCM‐associated sarcomere protein mutations through DNA sequencing. Despite these limitations, our study has significant implications for evaluating prognostic information in DCM.

Conclusions

Fibrosis on EMB from the RV may help predict LV fibrosis in patients with DCM in advance.

Conflict of interest

The authors declare that there were no conflicts of interest.

Funding

None.

Supporting information

Table S1: Autopsied cases (n = 18).

Table S2: Histological findings and CMR‐LGE pattern (n = 15).

Amemiya, K. , Matsuyama, T. , Ishibashi‐Ueda, H. , Morita, Y. , Matsumoto, M. , Ohta‐Ogo, K. , Ikeda, Y. , Tsukamoto, Y. , Fukushima, N. , Fukushima, S. , Fujita, T. , and Hatakeyama, K. (2024) Can right ventricular endomyocardial biopsy predict left ventricular fibrosis beforehand in dilated cardiomyopathy?. ESC Heart Failure, 11: 1001–1008. 10.1002/ehf2.14642.