Screening for cardiac sarcoidosis: diagnostic approach and long-term follow-up in a tertiary centre

Sarcoidosis is a systemic inflammatory disease of unknown origin that is characterised by the presence of non-caseating granulomas in various organs including the heart [1, 2]. Symptomatic cardiac involvement occurs in 5% of patients with sarcoidosis, although autopsy studies have shown a prevalence of up to 25% [3‐8]. The clinical presentation of cardiac sarcoidosis (CS) varies from an asymptomatic course to overt heart failure (HF) and sudden cardiac death (SCD), and is associated with poor prognosis [9‐11]. Therefore, early diagnosis and treatment are important [12‐16].

The diagnosis of CS is established by histological diagnosis through an invasive endomyocardial biopsy (EMB) with inherent risks and limited diagnostic yield due to sampling error because of the patchy involvement of the disease [17, 18]. Current guidelines aim to diagnose CS without the necessity of a positive EMB and are based on clinical and diagnostic criteria recommended by experts and limited scientific data [10, 11]. As a result, advanced non-invasive imaging modalities such as cardiovascular magnetic resonance imaging (CMR) and ¹⁸F‑fluorodeoxyglucose (¹⁸F‑FDG) positron emission tomography (PET) have been incorporated into these diagnostic guidelines [10, 19, 20]. Although previous research with advanced non-invasive imaging modalities reported similar prevalences of CS as autopsy studies and the sensitivity and specificity of each of these modalities for accurately diagnosing CS have been determined, the most optimal diagnostic approach is still not fully defined [17, 18, 21]. Furthermore, long-term follow-up data of CS patients are limited.

Therefore, the first aim of this study was to evaluate the frequency of CS, and the strengths and limitations of a daily used diagnostic approach for CS in a tertiary centre [10]. Secondly, the long-term outcomes of CS patients based on this diagnostic approach were assessed.

A total of 188 consecutive patients with sarcoidosis and clinical suspicion of CS were eligible for this study. Baseline cardiac analysis including ECG and TTE was performed in 180 patients (Fig. 1). Based on the HRS criteria, CS was diagnosed in 42 patients (23%). Baseline patient characteristics stratified by the presence of CS are summarised in Tab. 1. Mean age was 51 ± 11 years, 48% of the patients were male, and most were diagnosed with sarcoidosis stage I or II.

The aims of this study were to evaluate the diagnostic approach for CS in a tertiary centre and to assess the long-term outcomes of these patients. The main findings were as follows: (1) cardiac symptoms were found to be insufficient to distinguish between cardiac involvement or not, (2) an ECG and TTE without abnormalities made the diagnosis of CS unlikely but could not completely rule out cardiac involvement, (3) advanced non-invasive imaging contributed to the diagnostic yield of the diagnostic approach, and (4) in CS patients without ECG/TTE abnormalities at baseline had a good prognosis during follow-up.

Given the low sensitivity of an invasive EMB and the lack of a gold standard test to diagnose CS, expert opinion-based guidelines have been developed in which advanced non-invasive imaging modalities play an important role [17, 18, 26]. Although several studies have attempted to determine the optimal diagnostic approach for CS, this is still not fully defined. According to the study by Kouranos et al., the initial step of an approach can be based on the presence or absence of cardiac symptoms and/or ECG abnormalities [18]. One of the main reasons for referring sarcoidosis patients for cardiac analysis in the current cohort was the presence of cardiac symptoms. However, our study showed that the presence of cardiac symptoms was not sufficient to distinguish between patients with and those without cardiac involvement. Despite previous studies showing lower sensitivity and specificity of ECG [18, 21], our approach started with the assessment of ECG/TTE abnormalities. In the majority of our patients (88%) without ECG/TTE abnormalities, cardiac involvement could be excluded using the HRS criteria. However, a normal ECG/TTE could not completely rule out cardiac involvement; advanced non-invasive imaging still established cardiac involvement in 13% of these patients. However, follow-up showed these patients had an overall good prognosis as only 1 of them developed sepsis with multi-organ failure and death as a result.

To assess cardiac involvement of sarcoidosis with greater certainty, biomarkers and advanced non-invasive imaging modalities with higher sensitivity and specificity can be incorporated into the diagnostic approach [10, 18, 21, 26‐28]. For example, different studies showed that elevated N-terminal pro-brain natriuretic peptide (NT-proBNP) levels were associated with a higher risk of cardiac involvement and troponin levels were correlated with disease activity, with a normalisation of these levels after treatment with corticosteroids. It has also been described that troponin and NT-proBNP levels can predict long-term outcomes. For example, troponin is associated with fatal arrhythmias, and NT-proBNP is a predictor of heart failure [21, 29]. Therefore, these biomarkers may contribute to increased diagnostic accuracy of the further diagnostic approach. Unfortunately, in our retrospective study, biomarker data were not available in a sufficient number of patients. Furthermore, it is known from the literature that CMR is a useful modality in the assessment of CS due to its ability to characterise the myocardium (i.e. oedema, LGE) in a non-invasive manner. As a result, multiple studies have been able to demonstrate similar prevalences of CS by CMR as by autopsy [4, 6, 10, 17, 18, 21]. Our study showed a prevalence rate of 23%, which was achieved in a selected patient population without the standard performance of CMR/¹⁸F‑FDG PET in each patient. However, clinically silent CS can be missed by applying this approach and may have introduced bias. Therefore, the sensitivity and specificity of both advanced imaging modalities were not calculated. It is likely that if CMR and ¹⁸F‑FDG PET had been performed in every patient, the incidence of (possible) CS could have been slightly higher.

CMR can also contribute to establishing alternative diagnoses if CS can be excluded. In recent years, research has also shown that CMR can contribute to determining the prognosis of CS and deciding whether and when to start therapy (i.e. immunosuppressive therapy or preventive therapy with an ICD in patients with LGE with a higher risk of sustained VT) [17]. In addition, Vita et al. demonstrated that the diagnostic value of determining CS increased when CMR and ¹⁸F‑FDG PET results were combined and that this was also of importance for treatment [26]. Although CMR/¹⁸F‑FDG PET has many possibilities to optimise the diagnostic approach of CS, the availability, costs and expertise needed should be taken into account. Performing advanced cardiac imaging in every patient with extra-cardiac sarcoidosis will be demanding. We still recommend the pulmonary physician screens sarcoidosis patients for cardiac symptoms and refers them to the cardiologist if needed. Based on the clinical presentation and ECG/TTE/biomarkers results, the cardiologist will decide with greater certainty whether further additional testing using advanced cardiac imaging are necessary.

Based on our study, we recommend performing CMR in sarcoidosis patients (with confirmed extra-cardiac granulomatous inflammation on biopsy) who have a high clinical suspicion of CS [29]. CS patients are then monitored by the cardiologist over time, with or without repeated advanced imaging. For patients with normal ECG, TTE, biomarkers and CMR, it is unknown whether and when non-invasive imaging should be repeated. Our recommendation should be applied in case of symptoms, ECG changes or flare-ups of extra-cardiac sarcoidosis. However, we have not investigated this in our current study. No standard routine screening intervals are currently recommended [30]. To validate the findings of this observational single-centre study in a broader, multicentre context, a prospective study should be performed. It is of added value if future studies with a larger population and longer follow-up duration can confirm the findings of a good prognosis in sarcoidosis patients with a normal ECG and TTE. The role of cardiac biomarkers such as NT-proBNP and troponin should be further studied to determine whether they can be used as gatekeeper for additional testing. This may make it easier to decide whether additional imaging is necessary and whether patients can be safely discharged from the cardiology outpatient clinic.

Of the patients with sarcoidosis referred for cardiac evaluation to our tertiary centre, 23% were diagnosed with CS. Cardiac symptoms and ECG and/or TTE abnormalities were of limited value for CS screening. However, by adding CMR/¹⁸F‑FDG PET as advanced non-invasive imaging to our diagnostic approach, a higher diagnostic yield was obtained and other cardiac diseases could be diagnosed in patients without CS. Follow-up showed a good prognosis in CS patients without ECG/TTE abnormalities at baseline.