Heart failure with reduced ejection fraction (HFREF) is responsible for half of the cases of hospitalization for heart failure (HF). Ischemic heart disease is the main cause of HFREF, but several other conditions can result in systolic ventricular dysfunction.^1-6 The initial evaluation (physical, laboratory, and echocardiography) elucidates the cause in most cases, but many patients remain undiagnosed and are often classified as having idiopathic dilated cardiomyopathy (IDC).⁷

IDC is defined as left ventricular (LV) dilation and systolic dysfunction with no coronary artery disease (CAD) or abnormal ventricular overload.¹ The social and economic impact of adverse events caused by this cardiomyopathy is amplified by more often affecting people in the first few decades of life.² In this context, the establishment of a defined etiology can not only drive different therapeutic management, it can also provide prognostic information for the patient, family, and society.

Cardiac magnetic resonance (CMR) imaging can determine the etiology of certain cardiomyopathies through morpho-functional evaluation and delayed gadolinium enhancement (DGE) presence and pattern.^8-11 The detection of a nonischemic DGE pattern can exclude coronary angiography, saving resources and avoiding complications inherent to an invasive procedure with significant potential for contrast nephropathy. This screening may be even more relevant in HFREF patients since more than half have no significant CAD.^3,4

This study aimed to evaluate the potential contribution of CMR in the diagnosis of HF with systolic ventricular dysfunction without a defined etiology in a referral hospital for cardiovascular diseases.

This study involved the retrospective selection of all CMR examinations performed between January 2017 and June 2018 at Ana Nery Hospital, a tertiary reference center of the Unified Health System in Salvador, BA, Brazil. This service has a Medical Residency Program in Cardiology. The study population consisted of patients with HF and an ejection fraction < 50% (previously determined on transthoracic echocardiography [ECHO]), with no etiology defined at the initial evaluation who were referred to the CMR for diagnostic research (Figure 1).

The exclusion criteria were known history of CAD, myocardial infarction, or previous revascularization; primary valve disease; known history of hypertrophic cardiomyopathy or congenital heart disease; and previously defined cardiomyopathy diagnoses.

A total of 159 patients underwent CMR in the referral period. Of them, 87 (54%) had an indication for diagnostic evaluation of cardiomyopathy (Figure 1).

The study population was predominantly young, with a mean age of 45 ± 16 years and equally distributed by sex. The presence of risk factors for CAD showed that 35% had hypertension, 9% had diabetes, 8% had dyslipidemia, 8% were smokers, and none had a history of stroke or acute myocardial infarction.

As for the functional evaluation by CMR, the mean LV ejection fraction was 32%. There was a high occurrence of regional motility dysfunction and moderate to severe mitral regurgitation. Desynchrony (inter- and intraventricular) was present in 10% of cases, with three patients in functional class (FC) III (with LVEF by CMR of 16–32% in resynchronizer preimplantation evaluation, with one of extensive fibrosis, one of septal fibrosis, and one of absent fibrosis). The others were asymptomatic or had an FC II at the time of the CMR (LVEF by CMR of 48–60% with little or no fibrosis).

Other variables such as mean left atrial diameter and indexed LV diastolic volume are described in Table 1.

A total of 53 (61%) patients had a positive DGE. Of them, the DGE pattern was nonischemic in 44 (50.5%) and ischemic (subendocardial or transmural) in nine (10.3%), three of whom normal coronary arteries, five of whom had presumed Chagas heart disease, and one of whom had amyloidosis.

CMR identified a specific or very suggestive pattern of the HF etiology with ventricular dysfunction in 48 patients (55.3%) (Figure 2). Of the established diagnoses, 38.6% corresponded to diseases with a presumed less unfavorable prognosis (myocarditis, normal CMR, and hypertensive heart disease).

LVEF normalization by CMR (reverse remodeling) was seen in 12 (13.8%) patients who had a previous mean LVEF on ECHO of 40% (the mean interval between the ECHO and CMR exams was 5.5 months). CRM diagnoses in this group were represented by three myocarditis cases and nine undiagnosed cases (four progressed from a dysfunctional ECHO to a normal CMR, one had ischemic DGE with normal scintigraphy, and four had no fibrosis). All reverse remodeling cases were treated with optimized HF medical therapy.

In 39 (44.7%) patients, CMR information was insufficient for a significant diagnostic contribution, of which 33 (37.8%) had a negative DGE, two had junctional DGE, and four ischemic DGE, but with significant CAD excluded by complementary tests and negative Chagas serology (Figure 2). In this group, 15 (46.8%) patients underwent previous coronary angiography, which revealed no coronary obstructions.

This study highlights the importance of CMR in elucidating the etiopathogenesis of IDC without a defined etiological diagnosis. Despite the undetermined initial clinical and complementary tests, CMR helped identify the etiology in 55.3% of cases. In Brazil, there are no data that assertively address this issue. Maron et al.⁸ and Parsai et al.⁹ claim that approximately 50% of cardiomyopathy etiologies remain unknown, with 20–30% being possibly attributable to genetic causes. In such circumstances, CMR can be a useful diagnostic tool in the decision-making process as an intermediate or even definitive step.

The presence of DGE was consistent with the data on with HFREF patients.^3,11,13 It should be noted that the presence of DGE alone is associated with a threefold increase in mortality and hospitalization rates regardless of LVEF.¹³ However, the predominance of the nonischemic DGE pattern did not corroborate the findings of other studies.^3,14,15,16 This may be partly related to the predominance of young patients without CAD as well as the large percentage of patients with myocarditis.

The considerable occurrence of diagnoses with a presumed less unfavorable prognosis highlights the relevance of CMR as an instrument to guide the progressive characteristics of diseases in HF. Defined or suggested diagnoses can assist doctors, patients, and family members with therapeutic/social planning and improve prognostic information.

The NCC diagnosis in 11.5% of patients may not necessarily constitute a specific etiological diagnosis. According to some authors, it can correspond to the common final pathway of several heart diseases¹⁸ and indicate hypertrabeculation instead of genetic disease (which must be associated with other clinical data). Thus, even if it does not show other secondary causes of HF, distinguishing between adaptive hypertrabeculation and true NCC (familial or sporadic, primary or triggered by overlapping heart disease) is a challenge.

Patients without a defined diagnosis and negative DGE may have idiopathic IDC. This is partly because IDC shows no fibrosis on CMR (it classically presents as septal wall linear mesocardial DGE).¹⁹ However, it is worth recalling other causes of ventricular dysfunction and dilation that usually present no DGE, such as alcoholic, peripartum, and Takotsubo cardiomyopathy.^1,8 The correlation with clinical history is greatly important to excluding such secondary cardiomyopathies. The absence of DGE, despite contributing little to the diagnosis, is an important piece of information in the clinical follow-up. These patients have a lower chance of adverse events and a better response to ventricular resynchronization therapy when necessary.²

As for other cases without defined etiology, the junctional DGE found in two patients can be justified by pressure overload secondary to IDC or, less likely, final-stage HCM with dilation.

In this study, CAD was previously excluded by the attending physicians as a cause of cardiomyopathy, but 11% of the tests presented ischemic DGE. Most of them suggested Chagas cardiomyopathy (six patients, later confirmed by new serology); however, four patients had negative serological tests despite positive epidemiology. These cases may correspond to serological false negatives or the lack of a second serology confirmation since there are no other classical causes of segmental fibrosis besides myocardial infarction and Chagas cardiomyopathy, especially in patients from endemic areas.

The absence (or little extension) of fibrosis is associated with a greater chance of reverse remodeling.^1,2 This premise was verified in the present study, with 13.8% of patients presenting normalized ventricular function and, of them, only two had small-sized positive DGE.

In addition to the more favorable prognosis inherent to this group of patients, the relevant role of optimized drug treatment in HF stands out in all patients.

The proportion of patients with normal CMR despite a previous ECHO revealing a reduced ejection fraction and a past history of HF symptoms corresponds to cases of myocarditis with good progression and ventricular function recovery. According to the risk classification of myocarditis, patients with low-risk syndromes (mild to moderate dysfunction without associated ventricular arrhythmias) usually present no symptoms and recover function in 1–4 weeks.¹⁷ However, the great operator-dependency of ECHO was maximized in the present context since these tests were received from different services.

The patients in the present study were a decade younger and had fewer risk factors for CAD than what has been described in the literature for HF patients.^3-6 These data corroborate the probable definition of undetermined diagnoses as being idiopathic IDC, as it is established that this cardiomyopathy often affects people in the first few decades of life.²

In the context of IDC research, CMR may play a restrictive (gatekeeper) role in indicating coronary angiography in HFREF patients, since more than half have no significant CAD.^3,4 Thus, coronary angiography would not be necessary in patients without a defined etiology and nonischemic DGE. In HF, in addition to being susceptible to vascular complications and bleeding, the risk of nephrotoxicity and relevant morbidities should be considered, especially in patients with more severe disease.

The relevance of CMR in the proposed patient profile was researched in a specialized tertiary center. Its usefulness is likely to be even greater in clinical practice, in which non-specialists manage HF.

The performance of CMR in patients without etiologic diagnosis of HFREF is clinically significant since it contributed to the etiology and prognosis of more than 50% of cases. This positive impact was reported in a tertiary cardiology teaching service; thus, CMR may have an even greater role in circumstances other than those presented here since non-specialists often manage HF.