Due to both the element of active granulomatous inflammation, and the heterogenous involvement of the left ventricle (LV) and/or right ventricle (RV), CS may behave differently than other cardiomyopathies when it comes to risk of ventricular arrhythmias and SCD. Accordingly, a different set of rules may govern SCD risk and ICD indications than the standard LVEF < 35 % that is often used for non-sarcoidosis cardiomyopathies. While CS patients with EF < 35 % should receive a primary prevention ICD in accordance with accepted device guidelines [6], many CS patients with EF > 35 % may also be candidates for primary prevention ICD implantation.

While a lower LVEF is associated with appropriate ICD therapy in several retrospective studies, many patients with only mildly impaired LV function also had substantial risk of ventricular arrhythmias and resultant appropriate ICD shocks and anti-tachycardia pacing [7–9]. In fact, in the study by Kron et al. most primary and secondary prevention patients who received appropriate ICD therapies had an LVEF > 35 %, suggesting that CS patients with mild or moderately reduced LVEF may still be at substantial risk for ventricular arrhythmias [7]. Furthermore, in the study by Betensky et al. 7 of the 17 patients (41 %) with appropriate ICD therapy had LVEF > 35 % [9]. While these studies noted that a number of patients with mild and moderately reduced LVEF received appropriate ICD therapies for VT/VF, Schuller et al. reported that in their primary prevention cohort, no patient with normal left and right ventricular systolic function (LVEF > 50 % and RVEF > 40 %) required ICD therapies over the follow up period [8]. It is from these data that a Class IIa recommendation for consideration of ICD implantation in patients with LVEF 36–49 % and/or RV ejection fraction <40 %, after optimal medical therapy and initiation of immunosuppression if indicated was based.

As discussed in Chaps.​ 5, 6, and 7, CMRI and 18-FDG myocardial PET appear to be the imaging modalities with the highest sensitivity and specificity in the detection of CS [10–14]. CMRI, in particular, has emerged as the test of choice at many centers in the evaluation of CS, owing in part to a lower false positive rate. CMRI uses T-2 weighted signal and early gadolinium images to detect acute inflammation and late gadolinium enhancement (LGE, also known as delayed contrast enhancement) to assess for fibrosis or scar. Preliminary data also suggest that surveillance with CMRI can assess the efficacy of steroid therapy [15, 16].

Importantly, CMRI has a role in risk stratification and prognosis. In a recent study of 155 consecutive patients with systemic sarcoidosis who underwent CMR for workup of suspected CS, LGE was present in 39 patients (25.5 %), and its presence was associated with a Cox hazard ratio of 31.6 for death, aborted sudden cardiac death, or appropriate ICD discharge [17] over a median 2.6 year follow-up period. Of the twelve patients with sudden cardiac death or appropriate ICD discharge in this study, all had LGE present on CMRI. Sarcoidosis patients without LGE, even those with LV dilatation and severely impaired LVEF, did not experience SCD, suggesting CMRI may provide prognostic data beyond that conferred by ejection fraction alone.

While CMRI may have a higher specificity, ¹⁸ F-fluoro-2-deoxyglucose positron emission tomography (¹⁸F-FDG PET) appears to detect active inflammation in CS with a slightly higher sensitivity [18, 19]. However, as the uptake of ¹⁸F-FDG is seen in other inflammatory myocardial diseases, PET is non-specific for CS and must be interpreted in the appropriate clinical context. Like CMRI, PET carries prognostic value – in a study of 118 patients with suspected CS, 60 % had abnormal cardiac PET findings. Over a median follow-up of 1.5 years, abnormal PET findings were associated with a hazard ratio of 3.9 for death or VT, and an abnormal PET remained a significant predictor of death or VT even after multivariate analysis to adjust for ejection fraction and other clinical variables [13], indicating PET offers prognostic value beyond EF alone.

Electrophysiologic study can add useful diagnostic information, particularly if CMRI or PET are inconclusive or unable to be obtained. Electroanatomical mapping, or “voltage-mapping,” of the right ventricle can help identify areas of scar, which can aid in the diagnosis of CS in the right clinical context. There have also been limited studies investigating the role for programmed electrical stimulation (PES) for the inducibility of sustained monomorphic ventricular tachycardia as a means of risk stratification of sudden cardiac death in patients with CS. One such study of 76 patients with CS undergoing PES showed that 6 of 8 patients (75 %) with inducible VT went on to have ventricular arrhythmias and ICD therapies, compared to 1 patient with sudden cardiac death amongst the 68 patients (1.5 %) who had been non-inducible by PES [20]. In another study of 32 patients with CS, 4 of 6 patients (67 %) with inducible VT went on to have appropriate ICD therapies while 2 of 20 patients (10 %) who were non-inducible went on to have ventricular arrhythmias or sudden cardiac death [21]. While these studies suggest a potential role for electrophysiologic testing in risk stratification, these data ultimately need to be replicated in larger cohorts. Furthermore, these studies are limited by mean follow-up periods of 5 years and 2.6 years, respectively; and, in light of CS as a progressive disease, the long-term prognostic value of a single negative EP study is unclear.