Myocardial T-Lymphocytes as a Prognostic Risk-Stratifying Marker of Dilated Cardiomyopathy　― Results of the Multicenter Registry to Investigate Inflammatory Cell Infiltration in Dilated Cardiomyopathy in Tissues of Endomyocardial Biopsy (INDICATE Study) ―

Abstract

Background: Dilated cardiomyopathy (DCM) associated with inflammation is diagnosed by endomyocardial biopsy; patients with this have a poorer prognosis than patients without inflammation. To date, standard diagnostic criteria have not been established.

Methods and Results: This study analyzed clinical records and endomyocardial biopsy samples of 261 patients with DCM (201 males, median left ventricular ejection fraction; 28%) from 8 institutions in a multicenter retrospective study. Based on the European Society of Cardiology criteria and CD3 (T-lymphocytes) and CD68 (macrophages) immunohistochemistry, 48% of patients were categorized as having inflammatory DCM. For risk-stratification, we divided patients into 3 groups using Akaike Information Criterion/log-rank tests, which can determine multiple cut-off points: CD3⁺-Low, <13/mm² (n=178, 68%); CD3⁺-Moderate, 13–24/mm² (n=58, 22%); and CD3⁺-High, ≥24/mm² (n=25, 10%). The survival curves for cardiac death or left ventricular assist device implantation differed significantly among the 3 groups (10-year survival rates: CD3⁺-Low: 83.4%; CD3⁺-Moderate: 68.4%; CD3⁺-High: 21.1%; Log-rank P<0.001). Multivariate Cox analysis revealed CD3⁺ count as a potent independent predictive factor for survival (fully adjusted hazard ratio: CD3⁺-High: 5.70, P<0.001; CD3⁺-Moderate: 2.64, P<0.01). CD3⁺-High was also associated with poor left ventricular functional and morphological recovery at short-term follow up.

Conclusions: Myocardial CD3⁺ T-lymphocyte infiltration has a significant prognostic impact in DCM and a 3-tiered risk-stratification model could be helpful to refine patient categorization.

Dilated cardiomyopathy (DCM) is clinically defined by dilatation of the left or both ventricles with systolic dysfunction, leading to heart failure, refractory arrhythmias, and cardiac death.^1–3 Despite therapeutic advances, some DCM patients still require mechanical support or cardiac transplantation; therefore, the establishment of etiology-based definitive treatment is needed.

The etiology of DCM is heterogeneous, including genetics, inflammation, and cardiotoxic agents.⁴ Particularly, a progression from viral myocarditis to DCM has long been hypothesized.⁵ Myocarditis is an inflammatory condition in the myocardium and can be triggered by virus infection and the subsequent autoimmune response mediated primarily by T-lymphocytes.^6,7 Lymphocytic myocarditis is generally self-limited; however, myocardial inflammation can persist in several cases, causing continuous tissue damage and fibrosis, transits to chronic myocarditis, and finally progressing to DCM.^5,8 The Japanese Circulation Society (JCS) Task Force Committee on Chronic Myocarditis proposed the definition of the disease and the criteria for diagnosis in 1996.⁹ These criteria are also included in the present “Guidelines for Diagnosis and Treatment of Myocarditis (JCS 2009).”¹⁰

Recently, coupled with recognizing the significant role of chronic inflammation in heart failure, DCM associated with inflammation has attracted attention as a potential target for immunosuppressive/immunomodulation therapy.^11–16

At present, only histological diagnosis by endomyocardial biopsy (EMB) allows definite diagnosis of myocardial inflammation, with possible identification of the underlying etiology.^{10,15,17–19} The Dallas criteria proposed in 1986 has provided a histopathological categorization of myocarditis and has been the gold standard for a long period despite much criticism.^20,21 To improve the low sensitivity and objectivity of the traditional Dallas criteria for myocarditis, the World Heart Federation (WHF) proposed quantitative immunohistological criteria to define inflammation in “inflammatory DCM/chronic myocarditis” in 1996 as follows: an infiltrate ≥14 leukocytes/mm² with preferably T-lymphocytes (≤4 macrophages may be included, myocyte necrosis or degeneration is not mandatory).⁷ In 2013, the European Society of Cardiology (ESC) proposed the modified WHF-based criteria for abnormal inflammatory infiltrates, with the additional criterion of CD3-positive (CD3⁺) T-lymphocytes ≥7 cells/mm²;⁶ however, this immunohistological approach has not gained broad consensus yet, nor is well-integrated into the clinical practice.^12,15,22,23 Furthermore, the disease concept and diagnostic criteria of the inflammatory myocardial diseases could vary in the United States, Europe, and Japan.²⁴

Hence, we conducted a multicenter retrospective study named INDICATE (Multicenter Registry to Investigate Inflammatory Cell Infiltration in Dilated Cardiomyopathy in Tissues of Endomyocardial Biopsy) in Japan. First, we immunohistologically evaluated myocardial inflammation in Japanese DCM patients using the ESC criteria as a reference and compared this measure with clinical outcomes. Next, we focused on T-lymphocyte, a leading player of lymphocytic myocarditis and diagnostic mainstay. We examined whether CD3⁺ T-lymphocyte count in the EMB sample could impact long-term outcomes in DCM patients and sought to determine the optimal cut-offs for risk stratification.

Methods

Study Population

This is a multicenter, retrospective, observational cohort study of 261 patients admitted for heart failure who underwent diagnostic EMB and were diagnosed with DCM in 8 participating centers: one National Center (National Cerebral and Cardiovascular Center) and 7 University Hospitals in Japan (Fukushima Medical University Hospital, Gifu University Hospital, Jichi Medical University Hospital, Mie University Hospital, Nagoya City University Hospital, Okayama University Hospital, Saitama Medical University International Medical Center). For registration, the patients with a diagnostic EMB between January 2004 and December 2014 were considered potentially eligible. To avoid the selection bias in this retrospective study design, each participating center registered the consecutive eligible patients within the selected period (median, 6 years), which depended on the center’s availability to register patients. Out of 265 consecutive eligible patients registered, 4 were excluded owing to the absence of myocytes (n=3) or presence of amyloid deposit (n=1) in the EMB samples. Therefore, 261 patients were ultimately analyzed. Supplementary Figure 1 presents the flowchart of participant selection.

Clinical, echocardiographic, and hemodynamic data were collected from patients’ medical records. All patients had undergone echocardiography, coronary angiography, right heart catheterization, and EMB during diagnostic work-up. Left ventricular (LV) ejection fraction (EF) was calculated using the Teichholz formula. Inclusion criteria were LVEF ≤45% and LV diastolic diameter >112% (2 standard deviations above the mean value in Henry’s data) of the predicted value calculated using the formula: 45.3 × (body surface area)^1/3 − (0.03 × age) − 7.2.²⁵ Body surface area was calculated by using the Du Bois formula: 0.007184 × weight^0.425 × height^0.725. Patients with coronary artery disease with >50% stenosis at the main branch, hypertrophic cardiomyopathy, valvular heart disease, left ventricular hypertrophy due to hypertension, a history of uncontrollable or untreated hypertension for ≥1 year before documentation of LV dysfunction, and other secondary cardiomyopathies such as sarcoidosis or systemic disease with cardiac involvement were excluded. Other excluded patients had a history of malignant disease, cardiac surgery, acute myocarditis, active infectious disease or collagen disease, or were on current or prior immunosuppressive medication.

This study was approved by the Ethics Committee of the National Cerebral and Cardiovascular Center (#M27-063) and by the local ethics committees of the participating university hospitals and conducted in accordance with the principles of the Declaration of Helsinki. The committee waived the requirement to obtain informed consent because of the retrospective and observational study design. The “opt-out” approach for consent was approved and a written explanation for data use was provided on the websites so patients could decline participation.

Survival Analysis

The primary endpoint was cardiac death or left ventricular assist device (LVAD) implantation. The initial timepoint for each survival analysis was the EMB date.

Histological and Immunohistochemical Staining for Inflammation and Fibrosis Assessment

Serial unstained glass slides with 4-μm thick sections from preserved buffered formalin-fixed paraffin-embedded EMB tissue from each patient were stained with hematoxylin-eosin (HE) and Masson’s trichrome for conventional histology and fibrosis assessment, and by immunohistochemistry to detect inflammatory cells using an autostainer Leica Bond-III (Leica Biosystems, Wetzlar, Germany) in the National Cardiovascular Center, as previously described.^26,27 Primary antibodies were anti-CD3 (N1580, Dako, Glostrup, Denmark, dilution 1 : 10) for T-lymphocytes, and anti-CD68 (PG-M1, N0876, Dako, dilution 1 : 1,000) for macrophages. Antibody detection and counter-staining with hematoxylin were performed using a Bond Polymer Refine Detection Kit (DS9800; Leica Biosystems). To exclude the possibility of false-positives from the secondary antibody or non-specific immunoglobulin G (IgG) binding, mouse IgG1 antibody (X0931; Dako) was used as negative control.

CD3⁺- and CD68⁺-Cell Count and Fibrosis Area

The diagnosis of inflammation was made according to the most severe finding. To add objectivity, we conducted quantitative analysis. The number of infiltrating CD3⁺-, and CD68⁺-cells in photomicrographs were counted by 2 pathologists (K.O.-O., H.I.-U.) in a blinded manner and shown as number/mm², as described previously.²⁶ Briefly, whole-slide images were scanned with a digital microscope, Aperio Scan-Scope XT (Leica Biosystems), and digital images were viewed using Aperio ImageScope software (Leica Biosystems). Microscopic images were taken in 5 selected high-power fields in a low-magnification whole-view of the digitized picture. CD3⁺- and CD68⁺-cells in blood vessels were excluded from counting. Myocardial fibrosis, except endocardium, was quantitatively measured in Masson’s trichrome-stained sections, as described previously.²⁸ The percentage of fibrosis area was automatically determined with the Positive Pixel Count algorithm version 9 by using Aperio ImageScope, as the ratio of the total blue-stained area to the whole biopsied myocardial area excluding the endocardium.

Association of ESC-Proposed Myocardial Inflammation in EMB Specimens and Prognosis

We determined the frequency of myocardial inflammation in EMB specimens according to the ESC criteria (≥14 leukocytes/mm² including ≤4 monocytes/mm² with the presence of CD3⁺ T-lymphocytes ≥7 cells/mm²) using immunohistochemistry for CD3⁺ T cells and CD68⁺ macrophages.⁶ Then, we analyzed the association with prognosis.

Prognostic DCM Patient Stratification by Myocardial CD3⁺ T-Lymphocyte Count

Next, for refined assessment, we analyzed the association between the CD3⁺ T-lymphocyte count and poor DCM prognosis by stratifying patients into >2 groups as shown in the statistical analysis section.²⁹ We focused on CD3⁺ cells as tissue markers for the following reasons: (1) its pathogenic involvement as a critical mediator of myocarditis, including viral and autoimmune myocarditis in both human and experimental models;^8,15 (2) its increasing use as immunohistological inflammation criteria as emphasized by the ESC;⁶ (3) CD3⁺-dependent determination of inflammation and modest correlations between CD3⁺ and CD68⁺ as shown above; (4) easier interpretation of CD3⁺ T-lymphocytes than CD68⁺ macrophages as pan-macrophage/monocyte markers of heterogeneous populations;^15,16 and (5) clear, easy-to-count, and reliable CD3 immunostaining, widely available in clinical pathology departments.

Association of LV Functional and Morphological Recovery at Short-Term Follow up With CD3⁺ Risk Categories

Finally, to check if the degree of LV functional and morphological recovery at short-term follow up (6 months to 1 year from EMB) differed among CD3⁺ categories, we analyzed the median echocardiographic parameters for LV function (LVEF) and diameters (LV end-diastolic diameter; LVEDD and LV end-systolic diameter; LVESD) at short-term follow up and their delta values.

Statistical Analysis

Continuous variables are expressed as median (interquartile range: IQR), and categorical variables as n (%). Fisher’s exact test, Mann-Whitney U-test, and the Kruskal-Wallis test were used as appropriate. Spearman’s rank correlation analysis was performed to estimate correlations between CD3⁺- and CD68⁺-cell counts, as well as each cell counts and fibrosis area. For the association of CD3⁺ T-lymphocyte count with survival outcome (i.e., incidence of cardiac death or LVAD implantation), the optimal number of cut-off points was determined by the minimum Akaike Information Criterion (AIC) value, and the optimal locations of cut-off points was determined by log-rank tests based on the minimum P value.²⁹ AIC is a fitting index to indicate models’ predictabilities and considering models’ simplicities. Lower AIC values show better models both on predictability and simplicity; refer to the study by Chang et al for more details.²⁹ Event-free survival curves were created by using the Kaplan-Meier method and compared by the log-rank test, post-hoc test (Holm method), and log-rank test for trend among the CD3⁺ categories based on the optimal number of cut-off points determined by the AIC and log-rank test. To investigate associations of myocardial CD3⁺-cell counts with outcome, Cox proportional hazard analyses to compute hazard ratios (HRs) and 95% confidence intervals (CIs) were performed. Covariates were sequentially added to the unadjusted model (Model 1) to assess the effect of covariate adjustments. We added demographic variables (age and sex) and B-type natriuretic peptide (BNP) as established risk factors of heart failure prognosis (Model 2) and other clinical variables as established risk factors (New York Heart Association [NYHA] functional class, ventricular tachycardia [VT] including non-sustained VT [with a duration of <30s] and sustained VT), and histological variables of interest (CD68⁺-cell count and fibrosis area) (Model 3). All P values were 2-tailed; a P value <0.05 was considered statistically significant. Intra- and interobserver variabilities for CD3⁺-cell counting in all patients were assessed by an intraclass correlation coefficient (ICC). All statistical analyses were performed with EZR (Saitama Medical Center, Jichi Medical University; http://www.jichi.ac.jp/saitama-sct/SaitamaHP.files/statmedEN.html),³⁰ a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria, version 2.13.0). More precisely, it is a modified version of R commander (version 1.6-3) designed to add statistical functions frequently used in biostatistics. Additionally, we used R version 4.0.2 when finding the optimal number of cut-off points by AIC and log-rank test.

Results

Patients

A total of 261 DCM patients (201 males, median age 55 years [IQR 40–65]) were analyzed, whose baseline characteristics are shown in Table 1. Sixty-six patients (25%) presented severe heart failure symptoms of NYHA functional class III to IV at baseline. The median value of plasma BNP, LV end-diastolic diameter, and LVEF were 193 (90–494) pg/mL, 65 (60–71) mm, and 28% (22–35%), respectively. The frequency of VT (non-sustained VT and sustained VT) was 26%. At discharge, most patients had been under conventional medical therapy with β-blockers (90%) and angiotensin-converting enzyme inhibitors (ACE-I)/angiotensin II receptor blockers (ARB) (85%).

Table 1. Basic Characteristics, Immunohistochemistry, Fibrosis Area, and Outcomes

	Total, n=261	CD3+ group			P value
		Low <13/mm2, n=178 (68%)	Moderate 13–24/mm2, n=58 (22%)	High ≥24/mm2, n=25 (10%)
Demographics
Age, years	55 (40–65)	55 (42–65)	56 (41–64)	51 (33–66)	0.507
Male, n (%)	201 (77)	143 (80)	43 (74)	15 (60)	0.065
Body mass index, kg/m2	24 (21–26)	24 (21–26)	23 (21–26)	23 (20–29)	0.99
NYHA class III or class IV, n (%)	66 (25)	36 (20)	23 (40)	7 (28)	0.012
Duration of heart failure symptoms, months	4.8 (2.1–38.3)	4.6 (1.8–18.7)	5.9 (2.8–52.7)	14.6 (2.5–60.5)	0.149
Hypertension, n (%)	102 (39)	75 (42)	18 (31)	9 (36)	0.305
Dyslipidemia, n (%)	119 (46)	88 (49)	23 (40)	8 (32)	0.154
Diabetes mellitus (%)	33 (13)	22 (12)	8 (14)	3 (12)	0.955
Atrial fibrillation, n (%)	57 (22)	41 (23)	9 (16)	7 (28)	0.66
Ventricular tachycardia (non-sustained and sustained), n (%)	69 (26)	47 (26)	16 (28)	6 (24)	0.944
Laboratory measurements
BNP, pg/mL	193 (90–494)	190 (81–477)	212 (91–499)	356 (114–522)	0.306
Estimated GFR, mL/min/1.73 m2	70 (56–83)	69 (57–82)	73 (57–93)	63 (46–81)	0.327
C-reactive protein, mg/dL	0.11 (0.05–0.34)	0.11 (0.04–0.34)	0.18 (0.08–0.30)	0.14 (0.06–0.30)	0.57
Echocardiography
LVEDD, mm	65 (60–71)	64 (60–71)	65 (62–71)	66 (59–74)	0.766
LVESD, mm	56 (51–63)	55 (51–62)	57 (51–64)	59 (50–64)	0.812
LVEF, %	28 (22–35)	29 (22–35)	27 (21–36)	25 (22–35)	0.654
Hemodynamics
PCWP, mmHg	10 (7–16)	10 (7–17)	10 (6–18)	14 (6–21)	0.609
RAP, mmHg	4 (2–7)	4 (2–7)	4 (2–6)	6 (2–8)	0.398
Mean PAP, mmHg	17 (13–27)	17 (13–24)	16 (13–30)	20 (15–30)	0.16
Cardiac index, L/min/m2	2.4 (2.1–2.9)	2.4 (2.1–2.9)	2.6 (2.1–3.1)	2.4 (2.2–2.7)	0.827
Medication at discharge
β-blocker, n (%)	236 (90)	162 (91)	51 (88)	23 (92)	0.756
ACE inhibitor or ARB, n (%)	224 (86)	155 (87)	49 (85)	20 (80)	0.603
Loop diuretic, n (%)	184 (71)	121 (68)	44 (76)	19 (76)	0.425
Amiodarone, n (%)	42 (16)	27 (15)	10 (17)	5 (20)	0.816
Warfarin, n (%)	131 (51)	94 (53)	24 (41)	13 (52)	0.28
Statin, n (%)	52 (20)	37 (21)	9 (16)	6 (24)	0.579
Immunohistochemistry
CD3, cells/mm2	8.2 (4.0–15.8)	6.0 (2.7–8.6)	16.5 (14.2–20.0)	32.1 (26.0–52.0)	<0.001
CD68, cells/mm2	26.0 (14.0–38.0)	21.3 (10.1–35.6)	30.0 (22.0–36.0)	56.0 (30.0–64.2)	<0.001
Fibrosis area, %	8.6 (5.3–12.6)	8.5 (5.1–12.0)	8.6 (5.8–12.5)	8.6 (6.6–17.5)	0.217
Outcomes*
Cardiac death or LVAD implantation, n (%)	46 (18)	21 (12)	16 (28)	9 (36)	0.001

Continuous data are presented as median (interquartile range). *Outcomes are defined by the primary composite endpoint of cardiac death or LVAD implantation. ACE, angiotensin-converting enzyme; ARB, angiotensin II receptor blocker; BNP, B-type natriuretic peptide; GFR, glomerular filtration rate; LVAD, left ventricular assist device; LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction; LVSD, left ventricular end-systolic diameter; NYHA, New York Heart Association classification; PAP, pulmonary artery pressure; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure.

Survival Analysis for Cardiac Death or LVAD Implantation

In total, 46 patients (17%) reached the primary composite endpoint of cardiac death (n=29) or LVAD implantation (n=17) during the median follow-up period of 7.4 (3.9–9.8) years after EMB. The survival rate free from cardiac death or LVAD implantation was 89.9% at 5 years and 75.8% at 10 years, respectively.

CD3⁺- and CD68⁺-Cell Count and Fibrosis Area in EMB Specimens

Cell counts in all patients are plotted in ascending order of CD3⁺ count (Figure 1). The median cell counts of CD3⁺ and CD68⁺ were 8.2 (4.0–15.8) and 26.0 (14.0–38.0) cells/mm², respectively, and the median fibrosis area 8.6% (5.3–12.6%) (Table 1). There were modest correlations between the CD3⁺ and CD68⁺ counts (r_s 0.374, P<0.001) (Supplementary Figure 2A), whereas neither CD3⁺ nor CD68⁺ count correlated with fibrosis area (Supplementary Figure 2B,C). Intra- and interobserver ICCs for CD3⁺ count in all patients were 0.98, and 0.82, respectively.

Figure 1.

Myocardial CD3⁺ T-lymphocyte and CD68⁺ macrophage cell counts in endomyocardial biopsy specimens of 261 patients with dilated cardiomyopathy.

Association of ESC-Proposed Myocardial Inflammation in EMB Specimens and Prognosis

According to the ESC criteria for inflammation (≥14 leukocytes/mm² including ≤4 monocytes/mm² with the presence of CD3⁺ T-lymphocytes ≥7 cells/mm²), 48% (124/261) of the DCM patients in our study were categorized as having DCM with inflammation. Determination of inflammation by ESC criteria nearly always depended on CD3⁺ count because up to 96% (251/261) of DCM cases had a CD68⁺ count ≥4 cells/mm² (4 cells=upper limit included in 14/mm²). Kaplan-Meier survival curves demonstrated that these patients had significantly poorer long-term outcomes than those without inflammation (log-rank P<0.001) (Figure 2).

Figure 2.

Kaplan-Meier estimates of cardiac death or left ventricular assist device (LVAD) implantation-free survival in 261 dilated cardiomyopathy (DCM) patients according to inflammation status defined by the European Society of Cardiology (ESC) criteria (log-rank test). Patients with DCM and inflammation had a lower probability of survival than those without.

Photomicrographs of representative cases are shown in Figure 3. Case 1 was categorized as having no inflammation by ESC criteria. Cases 2 and 3 were judged as inflammation positive. Case 2 represents histologically equivocal cases with inflammation identified by immunohistochemistry. Fibrosis area was comparable in all cases (Supplementary Figure 3). Notably, focal inflammation associated with adjacent myocyte necrosis (Case 3 in Figure 3) was rarely observed (4/261, 1.5%). Those patients fulfilled the histological Dallas criteria of myocarditis, where 3 reached the primary endpoint.

Figure 3.

Variations of histological and immunohistological findings in dilated cardiomyopathy patients. Case 1 (CD3⁺: 4/mm², CD68⁺: 32/mm²): no inflammation as per ESC-defined criteria. Case 2 (CD3⁺: 13/mm², CD68⁺: 30/mm²) and 3 (CD3⁺: 124/mm², CD68⁺: 58/mm²): inflammation positive. Case 2 represents histologically equivocal cases with inflammation identified by immunohistochemistry. Case 3 represents rare cases of histologically defined myocarditis with adjacent myocyte necrosis. From top to bottom: Hematoxylin and eosin (HE) stain, immunohistochemistry for CD3 and CD68.

Three-Tier Prognostic DCM Patient Stratification by Myocardial CD3⁺ T-Lymphocyte Count

Based on statistical calculations with AIC values and P values of log-rank tests,²⁹ 2 optimal cut-off points of CD3⁺ T-lymphocyte count were determined for the association with survival outcome: 13/mm² and 24/mm². Accordingly, we divided subjects into 3 groups to risk-stratify for cardiac death and LVAD implantation: CD3⁺-Low, <13/mm² (178, 68%); CD3⁺-Moderate, 13–24/mm² (58, 22%); and CD3⁺-High, ≥24/mm² (25, 10%) (Table 1). Baseline characteristics were comparable among groups except NYHA functional class III or IV. As for tissue markers, CD68⁺ macrophage count significantly differed among groups, whereas fibrosis area did not (Table 1). Primary endpoint occurrence also differed significantly among groups: 21/178 (12%) in CD3⁺-Low, 16/58 (28%) in CD3⁺-Moderate, and 9/25 (36%) in CD3⁺-High (P=0.001).

Cardiac Death or LVAD Implantation-Free Survival and Risk in the Three CD3⁺ Categories

Kaplan-Meier survival curves free from cardiac death or LVAD implantation differed significantly among the 3 CD3⁺ categories (log-rank P<0.001) (Figure 4). Log-rank trend test showed a significant trend of CD3⁺-Low to High with the worst survival in CD3⁺-High (P for trend; P<0.0001), although post-hoc analysis for the log-rank test did not show a significant difference between CD3⁺-Moderate and CD3⁺-High (P=0.133). The survival rates at 10 years were 83.4% for CD3⁺-Low, 68.4% for CD3⁺-Moderate, and 21.1% for CD3⁺-High.

Figure 4.

Kaplan-Meier estimates of cardiac death or left ventricular assist device (LVAD) implantation-free survival in 261 dilated cardiomyopathy patients stratified by the 3 CD3⁺ count categories (log-rank test). Long-term prognosis was incrementally associated with CD3⁺ count. The CD3⁺-High ≥24/mm² group had the worst outcome (P for trend; P<0.0001).

Hazard ratios for cardiac death and LVAD implantation according to the CD3⁺ count are presented in Table 2. In multivariable Cox analysis adjusted for various variables, CD3⁺ count was strongly associated with worse outcomes, with fully adjusted HRs (95% CIs) for 3 tiers of CD3⁺ count of 1 (reference) in CD3⁺-Low, 2.64 (1.33–5.22) in CD3⁺-Moderate, and 5.70 (2.08–15.66) in CD3⁺-High with the highest risk (P for trend; P<0.0001). CD3⁺-High was associated with an HR (95% CI) of 2.16 (0.80–5.85) in Model 3 (P=0.13) when CD3⁺-Moderate was used as the reference. Cox analysis with CD3⁺ count as a continuous variable showed a significant association of a 5-unit increase in CD3⁺ count with cardiac death and LVAD implantation, with an HR (95% CI) of 1.09 (1.01–1.18) (P<0.05).

Table 2. Hazard Ratios for Cardiac Death and LVAD Implantation According to CD3⁺ T-Lymphocyte Count

	CD3+ group					P value for trend	CD3+ count (per 5 unit)	P value
	Low	Moderate	P value	High	P value
CD3+ count, /mm2	<13	13–24		≥24
Number of participants	178	58		25
Number of events	21	16		9
Model 1a	1 [Reference]	2.59 (1.35–4.97)	<0.01	4.84 (2.19–10.69)	<0.0001	<0.0001	1.10 (1.05–1.16)	<0.001
Model 2b	1 [Reference]	2.94 (1.51–5.72)	<0.01	6.22 (2.58–15.02)	<0.0001	<0.0001	1.10 (1.05–1.16)	<0.001
Model 3c	1 [Reference]	2.64 (1.33–5.22)	<0.01	5.70 (2.08–15.66)	<0.001	<0.001	1.09 (1.01–1.18)	<0.05

^AUnadjusted. ^BAdjusted for age, sex, B-natriuretic peptide. ^cAdjusted for covariates in model 2 plus New York Heart Association functional class, ventricular tachycardia, CD68, fibrosis area. LVAD, left ventricular assist device.

For reference, CD68⁺ count was not significantly associated with outcomes in the fully adjusted model (HR 1.00, 95% CI 0.99–1.02, P=0.683. No multicollinearity between CD3⁺ and CD68⁺ count with variance inflation factor [VIF] <10; VIF 1.58 for CD3⁺ count, VIF 1.34 for CD68⁺ count).

Association of LV Functional and Morphological Recovery at Short-Term Follow up With CD3⁺ Risk Categories

Baseline LVEF, LVEDD, and LVESD values did not differ significantly among groups (Table 1). At short-term follow up, the LVEF in the CD3⁺-High group (33 [25–39]%) was significantly lower than that in the CD3⁺-Moderate (40 [30–49]% [P<0.05]) and CD3⁺-Low groups (44 [30–55]% [P<0.01]) (Figure 5, upper panel, left). Further, LV dimensions at short-term follow up were significantly greater in the CD3⁺-High group than in the CD3⁺-Low group (LVEDD: 63 [57–68] mm vs. 56 [50–64] mm, P<0.05; LVESD: 51 [47–59] mm vs. 42 [36–54] mm, P<0.01) (Figure 5, upper panel, middle, and right). Furthermore, temporal analysis showed that the CD3⁺-High group had the worse recovery of LVEF and related dimensions (Figure 5, lower panel).

Figure 5.

Association of left ventricular (LV) functional and morphological recovery at short-term follow-up with CD3⁺ categories. Upper panel, median echocardiographic parameters for LV function (LV ejection fraction [EF]) and diameters (LV end-diastolic diameter [LVEDD] and LV end-systolic diameter [LVESD]) at short-term follow up. Lower panel, temporal changes with delta values of each parameter. The CD3⁺-High group showed the poorest recovery in both LV function and dimensions. *P<0.05, **P<0.01, ***P<0.001.

Discussion

The main findings of this study are the following:

1. In Japan, 48% of DCM patients fell into the category of DCM with inflammation (inflammatory DCM: iDCM) according to ESC criteria.

2. Myocardial CD3⁺ T-lymphocyte count was an independent, incremental prognostic marker for DCM and patients were best risk-stratified into 3 categories, with the optimal cut-off values of 13/mm² and 24/mm².

3. Patients with high numbers of infiltrating T-lymphocytes (CD3⁺-High, ≥24/mm²) had a particularly worse prognosis and poor LV functional and morphological recovery.

Frequency of Inflammatory DCM and Prognosis

“Inflammatory DCM” is often used for a subgroup of patients with DCM associated with inflammation, who have a poorer prognosis than DCM patients without inflammation.⁶

A recent meta-analysis of 61 publications from 2005 to 2014 to immunohistochemically detect inflammatory cardiomyopathy by EMB in clinically ‘suspected DCM or myocarditis’ (total 10,491 patients)²² revealed an overall detection rate of “inflammatory DCM” of 47%, with 13 different diagnostic criteria combining inflammatory cell markers with/without cell adhesion molecules and various thresholds. According to ESC criteria based on the count of T lymphocytes and macrophages, we found that 48% of DCM patients in Japan showed inflammation on EMB in a multicenter study, which is comparable with the meta-analysis and a previous Japanese study reporting inflammation rates of 47% (30/64) by modified Dallas criteria in the myocardium resected at left ventriculoplasty from DCM patients with severe heart failure.³¹

Furthermore, we found that DCM cases with inflammation classified by ESC criteria had worse long-term outcomes, in line with a previous study³² on 181 patients with “suspected myocarditis” showing an association of poor outcomes and immunohistologically positive inflammation classified by modified ESC criteria (>14/mm² CD3⁺ and/or CD68⁺ with enhanced human leukocyte antigen [HLA] class II expression), suggesting more aggressive therapeutic strategies may be needed for these patients. Among etiology-based strategies, anti-inflammation therapies are anticipated. Although recent studies suggest possible beneficial effects of immunosuppressive therapy on LV function or survival for virus-negative inflammatory DCM,^13,14,33 its effectiveness is still controversial, thus a large randomized trial is required for optimal personalized treatments.³⁴

Association of Myocardial CD3⁺ T-Lymphocyte Count and Prognosis in Patients With DCM

In this study, “inflammatory DCM” was diagnosed based on the count of T lymphocytes and macrophages; however, the conventional histology of the EMB ranged from no noticeable inflammation to unequivocal accumulation diagnosis of classic “myocarditis”. As T-lymphocytes are essential for the diagnosis of myocarditis and could play central roles in the pathogenesis of chronic myocarditis and progression to DCM,^5,35 we focused on myocardial CD3⁺ T-lymphocyte count. The survival curves differed significantly in the 3 CD3⁺ categories divided by 2 optimal cut-off points. To our knowledge, this is the first study to show the association between the degree of myocardial T-lymphocyte infiltration and prognosis with a hard endpoint of cardiac death or LVAD implantation with a long-term follow-up period in patients with DCM. Study subjects likely reflect real-world DCM patients who had undergone diagnostic work-up, including EMB and being treated by conventional heart failure therapy, mainly β-blockers and ACE-I/ARB, without knowledge of myocardial viral genome status or disease-causing gene mutation.

Myocardial CD3⁺ Count as a Three-Tiered Risk-Stratifying Marker for DCM and Its Clinical Implication

Furthermore, we found that myocardial CD3⁺ count in patients with DCM is an independent, strong prognostic tissue marker in multivariable analysis. CD3⁺-High ≥24/mm² showed the highest risk (fully adjusted HR 5.70), whereas CD3⁺-Moderate 13–24/mm² had an intermediate risk (fully adjusted HR 2.64) compared to CD3⁺-Low <13/mm². In other words, DCM patients can be risk-stratified by the 3 CD3⁺ categories. Moreover, the CD3⁺-High group was associated with poorer short-term (6 months to 1 year) LV functional and morphological recovery, thus requiring extra attention to improve long-term outcomes. Notably, all 4 cases of authentic myocarditis with myocyte damage belonged to the CD3⁺-High group.

Therefore, we estimate a CD3⁺ ≥24/mm² as a prognosis-oriented criterion for chronic myocarditis with high-grade inflammation in a presumably active state. This level of T-lymphocytic infiltrates will usually be recognizable by HE staining even if presenting a scattered distribution, as the threshold well surpasses the conventional criteria of 5 mononuclear cells per high-power field on HE staining (corresponds to 15/mm² at magnification 400×, field area 0.338 mm²),^36,37 which defines abnormal mononuclear cell infiltrates in the myocardium, as applied in “the Guideline for Diagnosing Chronic Myocarditis” by the Japanese Circulation Society Task Force Committee,⁹ followed by the latest 2009 Guidelines for Diagnosis and Treatment of Myocarditis by the Japanese Circulation Society.¹⁰

The CD3⁺-Moderate 13–24/mm² group requires attention for intermediate prognostic risk, although inflammation in this group seems more heterogenous, presumably including inflammation associated with various stimuli in heart failure.^8,16 Interestingly, the lower cut-off value of 13/mm² nearly corresponds to WHF⁷ and ESC⁶ criteria of 14/mm², addressed as “preferably T-lymphocytes.” In view of temporal LV recovery, which is rather similar to the CD3⁺-Low group, a more detailed subset characterization may be required for better predicting risk in individuals, which needs to be clarified by further studies.

Our data indicate the concept “the more myocardial T-lymphocytes, the worse long-term prognosis” in a DCM setting, supporting the view of harmful T-cell recruitment to the heart with chronic destructive immune reactions being involved in disease progression.^8,15,35 We propose this CD3⁺ 3-tiered risk-stratification as simple, practical prognosis-oriented criteria, wherein the CD3⁺-High subset might be regarded as high-risk chronic myocarditis with relatively high-grade inflammation in immunohistologically defined inflammatory DCM.

Study Limitations

This study was retrospective with a relatively small sample size, particularly in the CD3⁺-High group for survival analysis with 3 CD3⁺ categories, although it was a multicenter design study. False negatives of EMB interpretation may also occur because inflammatory cell infiltration is not always diffuse. Data on serum troponin T levels and pro-inflammatory cytokines, viral genome status, autoantibodies, and disease-causing gene mutation were not available; thus, further comprehensive prospective multicenter studies are warranted.

Conclusions

Myocardial CD3⁺ T-lymphocyte count from EMB was an independent predictive biomarker for poor outcomes in DCM. Our 3-tiered risk-stratification strategy could be helpful for the refined assessment of DCM patients. Particularly, the highest risk patients with ≥24/mm² CD3⁺ T-lymphocytes should be regarded as individuals with chronic myocarditis with high-grade inflammation and poor LV recovery, and be considered top candidates for aggressive treatment, including immunosuppressive therapy.

Acknowledgment

The authors would like to thank Enago (www.enago.jp) for the English-language review.

Sources of Funding

This work was supported by a Grant-in-Aid for Scientific Research from the Japan Society for the Promotion of Science (2646098 to Y.S., JP19H03442 to K.I.-Y), Japan Heart Foundation Research Grant on Dilated Cardiomyopathy (to K.I.-Y) and AMED (under Grant Number 20ek0109476 h0001 to K.I.-Y).

Disclosures

The authors declare that there are no conflicts of interest. T.A. is the current Editor-in-Chief from the Circulation Journal’s Editorial Team.

IRB Information

The Ethics Committee of the National Cerebral and Cardiovascular Center (#M27-063) and local ethics committees of all participating university hospitals approved this study.

Data Availability

Deidentified participant data will not be shared.

Supplementary Files

Please find supplementary file(s);

http://dx.doi.org/10.1253/circj.CJ-21-0529

References

• 1.   Elliott P, Andersson B, Arbustini E, Bilinska Z, Cecchi F, Charron P, et al. Classification of the cardiomyopathies: A position statement from the European Society Of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J 2008; 29: 270–276.
• 2.   Braunwald E. Cardiomyopathies: An overview. Circ Res 2017; 121: 711–721.
• 3.   Kitaoka H, Tsutsui H, Kubo T, Ide T, Chikamori T, Fukuda K, et al. JCS/JHFS 2018 guideline on the diagnosis and treatment of cardiomyopathies. Circ J 2021; 85: 1590–1689.
• 4.   Richardson P, McKenna W, Bristow M, Maisch B, Mautner B, O’Connell J, et al. Report of the 1995 World Health Organization/International Society and Federation of Cardiology Task Force on the definition and classification of cardiomyopathies. Circulation 1996; 93: 841–842.
• 5.   Kawai C. From myocarditis to cardiomyopathy: Mechanisms of inflammation and cell death: Learning from the past for the future. Circulation 1999; 99: 1091–1100.
• 6.   Caforio AL, Pankuweit S, Arbustini E, Basso C, Gimeno-Blanes J, Felix SB, et al. Current state of knowledge on aetiology, diagnosis, management, and therapy of myocarditis: A position statement of the European Society of Cardiology Working Group on Myocardial and Pericardial Diseases. Eur Heart J 2013; 34: 2636–2648; 2648a–2648d.
• 7.   Maisch B, Portig I, Ristic A, Hufnagel G, Pankuweit S. Definition of inflammatory cardiomyopathy (myocarditis): On the way to consensus. A status report. Herz 2000; 25: 200–209.
• 8.   Imanaka-Yoshida K. Inflammation in myocardial disease: From myocarditis to dilated cardiomyopathy. Pathol Int 2020; 70: 1–11.
• 9.   Japanese Circulation Society (JCS) Task Force Committee on Chronic Myocarditis. Guideline for diagnosing chronic myocarditis. Jpn Circ J 1996; 60: 263–264.
• 10.   JCS Joint Working Group. Guidelines for diagnosis and treatment of myocarditis (JCS 2009): Digest version. Circ J 2011; 75: 734–743.
• 11.   Tschope C, Cooper LT, Torre-Amione G, Van Linthout S. Management of myocarditis-related cardiomyopathy in adults. Circ Res 2019; 124: 1568–1583.
• 12.   Hirono K, Takarada S, Okabe M, Miyao N, Nakaoka H, Ibuki K, et al. Clinical significance of chronic myocarditis: Systematic review and meta-analysis. Heart and Vessels 2022; 37: 300–314.
• 13.   Frustaci A, Chimenti C. Immunosuppressive therapy in myocarditis. Circ J 2015; 79: 4–7.
• 14.   Merken J, Hazebroek M, Van Paassen P, Verdonschot J, Van Empel V, Knackstedt C, et al. Immunosuppressive therapy improves both short- and long-term prognosis in patients with virus-negative nonfulminant inflammatory cardiomyopathy. Circ Heart Fail 2018; 11: e004228.
• 15.   Tschöpe C, Ammirati E, Bozkurt B, Caforio ALP, Cooper LT, Felix SB, et al. Myocarditis and inflammatory cardiomyopathy: Current evidence and future directions. Nat Rev Cardiol 2021; 18: 169–193.
• 16.   Adamo L, Rocha-Resende C, Prabhu SD, Mann DL. Reappraising the role of inflammation in heart failure. Nat Rev Cardiol 2020; 17: 269–285.
• 17.   Buja LM, Ottaviani G, Ilic M, Zhao B, Lelenwa LC, Segura AM, et al. Clinicopathological manifestations of myocarditis in a heart failure population. Cardiovasc Pathol 2020; 45: 107190.
• 18.   Sotiriou E, Heiner S, Jansen T, Brandt M, Schmidt KH, Kreitner KF, et al. Therapeutic implications of a combined diagnostic workup including endomyocardial biopsy in an all-comer population of patients with heart failure: A retrospective analysis. ESC Heart Fail 2018; 5: 630–641.
• 19.   Bennett MK, Gilotra NA, Harrington C, Rao S, Dunn JM, Freitag TB, et al. Evaluation of the role of endomyocardial biopsy in 851 patients with unexplained heart failure from 2000–2009. Circ Heart Fail 2013; 6: 676–684.
• 20.   Aretz HT, Billingham ME, Edwards WD, Factor SM, Fallon JT, Fenoglio JJ Jr, et al. Myocarditis: A histopathologic definition and classification. Am J Cardiovasc Pathol 1987; 1: 3–14.
• 21.   Baughman KL. Diagnosis of myocarditis: Death of Dallas criteria. Circulation 2006; 113: 593–595.
• 22.   Katzmann JL, Schlattmann P, Rigopoulos AG, Noutsias E, Bigalke B, Pauschinger M, et al. Meta-analysis on the immunohistological detection of inflammatory cardiomyopathy in endomyocardial biopsies. Heart Fail Rev 2020; 25: 277–294.
• 23.   Leone O, Veinot JP, Angelini A, Baandrup UT, Basso C, Berry G, et al. 2011 consensus statement on endomyocardial biopsy from the Association for European Cardiovascular Pathology and the Society for Cardiovascular Pathology. Cardiovasc Pathol 2012; 21: 245–274.
• 24.   Imanaka-Yoshida K. Tenascin-C in heart diseases: The role of inflammation. Int J Mol Sci 2021; 22: 5828.
• 25.   Henry WL, Gardin JM, Ware JH. Echocardiographic measurements in normal subjects from infancy to old age. Circulation 1980; 62: 1054–1061.
• 26.   Nakayama T, Sugano Y, Yokokawa T, Nagai T, Matsuyama TA, Ohta-Ogo K, et al. Clinical impact of the presence of macrophages in endomyocardial biopsies of patients with dilated cardiomyopathy. Eur J Heart Fail 2017; 19: 490–498.
• 27.   Yokokawa T, Sugano Y, Nakayama T, Nagai T, Matsuyama TA, Ohta-Ogo K, et al. Significance of myocardial tenascin-C expression in left ventricular remodelling and long-term outcome in patients with dilated cardiomyopathy. Eur J Heart Fail 2016; 18: 375–385.
• 28.   Wada Y, Aiba T, Matsuyama TA, Nakajima I, Ishibashi K, Miyamoto K, et al. Clinical and pathological impact of tissue fibrosis on lethal arrhythmic events in hypertrophic cardiomyopathy patients with impaired systolic function. Circ J 2015; 79: 1733–1741.
• 29.   Chang C, Hsieh MK, Chang WY, Chiang AJ, Chen J. Determining the optimal number and location of cutoff points with application to data of cervical cancer. PLoS One 2017; 12: e0176231.
• 30.   Kanda Y. Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 2013; 48: 452–458.
• 31.   Tsukada B, Terasaki F, Shimomura H, Otsuka K, Otsuka K, Katashima T, et al. High prevalence of chronic myocarditis in dilated cardiomyopathy referred for left ventriculoplasty: Expression of tenascin C as a possible marker for inflammation. Hum Pathol 2009; 40: 1015–1022.
• 32.   Kindermann I, Kindermann M, Kandolf R, Klingel K, Bultmann B, Muller T, et al. Predictors of outcome in patients with suspected myocarditis. Circulation 2008; 118: 639–648.
• 33.   Frustaci A, Russo MA, Chimenti C. Randomized study on the efficacy of immunosuppressive therapy in patients with virus-negative inflammatory cardiomyopathy: The TIMIC study. Eur Heart J 2009; 30: 1995–2002.
• 34.   Ammirati E, Frigerio M, Adler ED, Basso C, Birnie DH, Brambatti M, et al. Management of acute myocarditis and chronic inflammatory cardiomyopathy: An expert consensus document. Circ Heart Fail 2020; 13: e007405.
• 35.   Blanton RM, Carrillo-Salinas FJ, Alcaide P. T-cell recruitment to the heart: Friendly guests or unwelcome visitors? Am J Physiol Heart Circ Physiol 2019; 317: H124–H140.
• 36.   Edwards WD, Holmes DR Jr, Reeder GS. Diagnosis of active lymphocytic myocarditis by endomyocardial biopsy: Quantitative criteria for light microscopy. Mayo Clin Proc 1982; 57: 419–425.
• 37.   Foley DA, Edwards WD. Quantitation of leukocytes in endomyocardial tissue from 100 normal human hearts at autopsy: Implications for diagnosis of myocarditis from biopsy specimens of living patients. Am J Cardiovasc Pathol 1988; 2: 145–149.