Viewpoints in the Diagnosis and Treatment of Cardiac Sarcoidosis:
Adoption and Modification of Current Guidelines

Introduction 

Sarcoidosis is a multisystem granulomatous disease characterized by the presence of non-caseating granuloma in the involved organ. Except for a rare acute presentation, sarcoidosis is a chronic progressive disease, most frequently involving the lungs. Despite the higher incidence of myocardial involvement in autopsy series, clinically apparent cardiac sarcoidosis is reported to be seen in only 5-10% of patients with systemic sarcoidosis.

Cardiac involvement can be diagnosed directly with endomyocardial biopsy. However, myocardial involvement in sarcoidosis is patchy in nature rather than diffuse. Therefore, the diagnostic yield of endomyocardial biopsy has been reported to be only approximately 20%^. Thus, in most cases, the diagnosis of cardiac sarcoidosis is made when the patient demonstrates histologically confirmed extracardiac sarcoidosis and has non-histologic clues of cardiac involvement. These patients are called the “clinical diagnosis group” in widely cited articles.. The strategy works only in the presence of extracardiac sarcoidosis; however, several studies have suggested the presence of isolated cardiac sarcoidosis in which the diagnosis of extracardiac sarcoidosis cannot be made. Therefore, diagnostic criteria for isolated cardiac sarcoidosis not requiring histologic confirmation of non-caseating granuloma in the myocardium are needed.

In addition to the clinical dilemma in diagnosing cardiac sarcoidosis, various therapeutic issues should be raised for discussion, including the (i) usefulness of steroid treatment in cardiac sarcoidosis, (ii) initial dose of steroid and treatment strategy, (iii) effect of steroid on left ventricular (LV) function and atrioventricular (AV) conduction recovery, and (iv) indication for implantable cardioverter defibrillator (ICD) insertion.

Diagnosis

Clinical need for new criteria for cardiac sarcoidosis

Nowadays, isolated cardiac sarcoidosis is a generally accepted disease condition. Kandolin et al. reported that 63% of cardiac sarcoidosis cases showed isolated myocardial involvement. They only included histologically confirmed cases, including some explanted cases during transplantation or autopsy. Had they included patients with cardiac sarcoidosis in whom histologic confirmation could not be made, the proportion of isolated cardiac sarcoidosis might have increased further.

We prospectively collected 27 patients whose clinical and laboratory findings suggest chronic multifocal patchy myocarditis without other possible causes except cardiac sarcoidosis. A total of 24 patients underwent biopsy. Notably, a single patient may undergo one or two biopsy procedures. Of these patients, 1 underwent liver biopsy, 11 underwent hilar lymph node (LN) biopsy and 21 underwent endomyocardial biopsy. Among these patients, granulomas were found in the liver biopsy of a patient who was suspected to have liver involvement, whereas one patient showed granuloma and another patient showed scattered histiocytes in 11 hilar LN biopsies. Of the 21 patients who underwent endomyocardial biopsy, granuloma was observed only in one patient (1 of 21, 4%).

Hilar LN is the most likely site for the histological diagnosis of extracardiac sarcoidosis. Twenty-five patients were screened for the presence of hilar LN enlargement; however, only 10 patients (40%) showed hilar LN enlargement (presence of any LN ≥ 1 cm in diameter) and granuloma was seen only in one patients. (Table 1).

Table 1. Diagnostic yield of EBUS biopsy

If the diagnosis is maded solely on the basis of most recent guidelines from the Heart Rhythm Society (HRS), only three patients can be diagnosed as having cardiac sarcoidosis; hence, 24 patients were classified as having unknown inflammatory myocarditis. The guidelines suggested by the Japanese Ministry of Health and Welfare (JMH) and later revised by Japanese Society of Sarcoidosis and Other Granulomatous Disorders (JSSOG), allow clinical diagnosis and histologic confirmation for the presence of extracardiac sarcoidosis. However, the known clinical symptoms and signs that suggest a diagnosis of extracardiac sarcoidosis are rather scarce and insufficient. Two patients complained of Raynaud’s phenomenon and another patient was suspected to have interstitial lung disease based on chest CT findings; however, these conditions were insufficient to make a diagnosis of extracardiac sarcoidosis.

The criteria suggested by the JMH/JSSOG and HRS did not consider the presence of isolated cardiac sarcoidosis. The recently suggested electrogram-guided biopsy may increase the yield of endomyocardial biopsy, which may solve the current problem in the diagnosis of isolated cardiac sarcoidosis. However, at present, it seems to be helpful only in finding an abnormal myocardium rather than specifically identifying a granuloma. Therefore, clinically relevant criteria especially in diagnosing isolated cardiac sarcoidosis are needed. In our opinion, for patients suspected of having chronic inflammatory myocarditis and who showed patchy involvement at multiple locations, without other suspected possible causes, diagnosis of probable cardiac sarcoidosis can be reasonably made. We would like to propose a scoring system for this purpose, and we suggest that a patient who met the prerequisites and scored 4 or higher can be diagnosed as having cardiac sarcoidosis (Figure 1).

Figure 1. Criteria for probable cardiac sarcoidosis

Electrocardiogram

In addition to the several non-specific EKG findings, patients with cardiac sarcoidosis may experience symptomatic conduction disturbances or ventricular arrhythmias. Holter monitoring is sometimes needed to reveal the electrocardiographic abnormality that is not apparent with resting 12-lead EKG. In the diagnostic viewpoint, EKG findings have been widely neglected as they have poor sensitivity. The presence of epsilon waves does not necessarily lead to the diagnosis of arrhythmogenic right ventricular dysplasia (ARVD), which is reportedly seen in patients with cardiac sarcoidosis. However, minute EKG findings can sometimes lead to further diagnostic tests and can result in a diagnosis of cardiac sarcoidosis; otherwise, the disease might be missed (Figure 2).

Echocardiogram

Except for very rare cases of increased LV wall thickness mimicking hypertrophic cardiomyopathy, the classic echocardiographic finding is multiple regional wall motion abnormalities (RWMAs) that do not match the coronary artery territory, which indicates non-ischemic multiple RWMAs. Depending on the extent of sarcoid involvement, multiple patchy sarcoid involvement does not always result in multiple RWMAs. Moreover, the non-ischemic nature of RWMAs based on the unusual coronary artery territory is rather difficult to determine even for an expert echocardiographer, especially when the RWMA is confined to a single coronary artery territory. CT coronary angiography and cardiac magnetic resonance imaging (CMR) are helpful in this situation. CMR provides direct evidence of multiple patchy non-ischemic involvement of the myocardium, in contrast to the indirect evidence of the non-ischemic nature of RWMA based on a normal CT coronary angiogram. However, a CT coronary angiogram can provide additional information about the presence of hilar lymphadenopathy and the possible site for biopsy.

Among the RWMA, thinning of the basal anterior septum seems to be rather specific to cardiac sarcoidosis, although the exact specificity of this finding has not been elucidated. However, considering only at this echocardiographic finding may lead to missed diagnosis in a large number of patients, as the reported incidence of this finding is only 20%. In our 27 patients, the involvement of the basal anterior septum, including scarring and thinning (Figure 3), had a much higher incidence of 48% (13 of 27).

Figure 3. Involvement of the basal anterior septum in cardiac sarcoidosis. A,B. Thinning of the anterior septum at the mid-ventricular and basal levels. C. Wall thinning is not apparent. Only high echogenicity is noted. D,E. In addition to wall motion, subendocardial or transmural fibrosis is suspected. F. Wall thinning limited to the basal anterior septum

In addition to the role of echocardiography in the diagnosis of cardiac sarcoidosis, echocardiography is mandatory for monitoring LV function in patients with cardiac sarcoidosis.

Recently, several studies have reported diagnostic ambiguity between cardiac sarcoidosis and ARVD. Therefore, in cases echocardiographically suggestive of ARVD, the possibility of cardiac sarcoidosis should be considered, and histologic or genetic evidence should be sought for the differential diagnosis. In one patient with cardiac sarcoidosis who underwent cardiac transplantation, serial echocardiography before transplantation showed that the right ventricular (RV) morphology became similar to that of ARVD over time, and histological examination of the explanted heart showed fatty tissue replacement mimicking ARVD (Figure 4).

CMR

In addition to the evaluation of wall motion in cine CMR, multiple non-ischemic pattern delayed enhancements (DEs) are useful in the diagnosis of cardiac sarcoidosis (Figure 5). It is not unusual to see DEs in segments without RWMA. Therefore DE evaluation seems more sensitive than other methods used in the past to detect cardiac involvement in sarcoidosis.

Applications of various techniques in CMR have been studied in cardiac sarcoidosis, such as the presence of myocardial edema in T2-weighted images or T1 mapping to further characterize and quantitate myocardial fibrosis. However, despite the large number of publications suggesting the usefulness of these techniques, whether these techniques significantly upgrade the usefulness of CMR in real-world practice remains to be seen.

In addition to its diagnostic utility, the usefulness of CMR in the assessment of arrhythmic risk and evaluation of response to therapy is a hot topic of recent investigations.

18-Fluoro-2-deoxyglucose positron emission tomography (FDG-PET)

Unusual utilization of glucose as an energy source by the myocardium after prolonged fasting or a fatty diet can be seen during ischemic, inflammatory, or neoplastic conditions. Therefore, the presence of multiple patchy hot-uptake in FDG-PET implies multifocal myocarditis. If this finding does not indicate multiple areas of ischemia or multiple metastatic tumors, then sarcoidosis is the most likely condition that can be considered.

In the past, gallium scanning had been used for this purpose; however, nowadays, it is has almost been completely replaced by FDG-PET. The preference for FDG-PET over gallium scan is based on sensitivity and image quality; however, in the aspect of false-positivity, FDG-PET cannot be considered superior to gallium scan. In gallium scan, as the body handles gallium similar to ferric iron, gallium is bound in the area of inflammation or rapid cell division, not in the area of ischemia. However, in FDG-PET, positivity represents unsuppressed or unusual utilization of glucose as an energy source. Therefore, use of glucose by the normal myocardium as an energy source should be adequately suppressed by prolonged fasting or a fatty diet, sometimes with heparin; however, this condition may not be adequately accomplished. Moreover, usage of glucose as an energy source can be seen in ischemia or even in other unknown conditions at the cellular level not associated with epicardial coronary artery disease, such as in hypertrophic cardiomyopathy.

Taking into consideration these limitations and paying careful attention in obtaining images, FDG-PET can provide both morphologic information about patchy involvement and lesion characteristic information related to inflammation. Therefore, we rely more on FDG-PET results than the results of other imaging modalities.

FDG-PET can also be used to evaluate the response to steroids (Figure 6).

Figure 6. Treatment and follow-up scheme

However, in the evaluation of the responsiveness to steroids, caution should be taken concerning the effect of steroid on false-positive FDG-PET results, which is probably associated with the effect of steroid on glucose metabolism (Figure 7).

Serum markers

Serum angiotensin-converting enzyme (ACE) activity is reported to be elevated in approximately 60% of patients with systemic sarcoidosis, which may not be the same as that noted in patients with cardiac sarcoidosis. In one study, the mean value of ACE activity in 47 patients with definite cardiac sarcoidosis (16.6±9.0 IU/L) fell within the normal range ( ≤ 21.4 IU/L).

Despite the disease nature of myocardial necrosis induced by inflammation, the diagnostic utility of conventional measures of evaluating myocardial damage and inflammation - troponin and high-sensitivity C-reactive protein (hsCRP) levels, respectively - has not gained much attention. In our 27 patients, troponin I and hsCRP levels were measured before treatment (in 14 and 16 patients, respectively). Troponin I and hsCRP levels were mildly elevated in 7 of 14 patients (50%, range 0.18-1.45 ng/mL) and in 6 of 16 patients (38%, range 1.59-4.59 mg/L), respectively. (normal values: hsCRP 0-0.5mg/L, troponin I 0-0.028ng/mL). Additional studies are needed on the usefulness of these serum markers in the evaluation of disease activities, and therefore in the follow-up of the disease course and treatment response.

Treatment

Usefulness of steroid treatment in cardiac sarcoidosis

The usefulness of steroids is based on the expectation of their immunologic effects on either active inflammation or the end-result of fibrosis. Among the diseases treated with steroids, either active inflammation or fibrosis dominates. Therefore, use of steroids is not theoretically conflicting. However, granuloma represents an inflammation confined at the center, with fibroblast and collagen encasing and restricting the inflammation, thereby protecting the surrounding tissue. Apart from the on-going active inflammation and the expected favorable effect of steroids during the fibrotic phase, if granuloma formation alone is considered, it cannot be confidently insisted that the fibrosis-preventing effect of steroids is beneficial.

Yazaki et al. showed better long-term survival in patients using steroids. However, their study was done retrospectively in only 95 patients. Furthermore, patients whose cardiac sarcoidosis was diagnosed at autopsy were selected as the non-steroid user group. In another study, Nagai et al. showed a reduced composite end-point of all-cause death, symptomatic arrhythmias, and heart failure requiring admission with steroid therapy. However, no significant differences were found in terms of cardiac death or symptomatic arrhythmias between steroid user and non-users, which seems to indicate that the main effect of steroid therapy is preventing the deterioration of LV function, with the effect of reducing cardiac death or symptomatic arrhythmia being small.

At our institution, we start steroid therapy once the diagnosis of cardiac sarcoidosis is made. However, we limit steroid therapy to patients who showed hot-uptake in FDG-PET, those who did not show RWMAs in a previous study in the absence of hot-uptake in FDG-PET, or those who showed a progressive nature of the disease process. In addition, we also limit steroid therapy to patients whose LV function is relatively preserved (EF >35%), as we think that steroid therapy will be helpful only in maintaining, with little effect in restoring LV function.

Initial dose of steroid and treatment strategy

Two different trends in the dose of initial steroid therapy exist. High-dose of prednisolone (1 mg·kg^-1·day^-1) has been advocated as an initial dose, for minimal treatment failure with steroid therapy. However, in a previous study, the overall survival of patients treated with a high initial dose (>40 mg daily) did not differ from that in those treated with a low dose (<30 mg daily). This study was not a dose-finding study, and dose comparison was done in only 75 patients. Moreover, the survival curves between the high- and low-dose showed differences in the early phases (2-4 years), with similar survivals after 5 years. However, for practicality, we abandoned the initial high-dose policy because the majority of patients could not tolerable a high-dose of steroids, and the steroid dose should be reduced to a low dose range within the first month. In the majority of patients, remission can be induced with low-dose steroids.

The currently suggested scheme of steroid treatment in cardiac sarcoidosis, except for the evaluation of response, largely adapts the scheme used in the treatment of systemic sarcoidosis. The initial dose of therapy is maintained for 2-3 months, which is followed by the evaluation of therapeutic response with FDG-PET. Once there is a therapeutic response, the dose of initial steroid therapy is tapered down to maintenance dose for 3 months and the final dose is maintained for 9-12 months. In our institution, after experiencing a number of patients with successful induction and maintenance with initial one month of low dose steroid therapy and tapering to the maintenance dose during next month, we have been following a similar scheme with a reduced total duration of treatment of 12 months (Figure 5).

However, in the treatment strategy for patients who were confirmed to be non-responsive to steroids based on the FDG-PET scan after the initial steroid therapy, rather than initiating another high-dose immunosuppressive therapy or adding a second-line drug, we prefer not to use steroids, assuming that these patients are non-responders. Among our 27 patients, 2 patients fall into this category and the steroid dose was rapidly tapered off as patients refused further immunosuppressive therapy. LV function was monitored thereafter, which remained stable for 1 year and 2.5 years, respectively, up to present. Even in patients who do not refuse restarting steroid therapy, we do not think that this policy is unethical. Based on the data of Nagai et al., the main expected difference of using steroids is the reduction in admission because of heart failure. Steroid therapy does not seem to reduce cardiac death or symptomatic arrhythmia. Therefore, we treat these patients with conventional heart failure treatment as long as the LV function is maintained.

If the LV function deteriorates, continuing high-dose steroid therapy or adding second-line drugs may be an option. However, even if the patient is responsive to the re-initiation of steroid therapy, the expectation for LV function recovery is not high. Moreover, even in patients who might show improvement in LV function, the magnitude of recovery might not be sufficient to obviate transplantation in the near future.

Effect of steroid on LV function and AV conduction recovery

With regard to the recovery of LV function with steroid therapy, arriving at a conclusion based on current data is rather difficult. However, LV function would likely to be deteriorated without treatment.

Apart from steroid therapy, depending on the LV function and symptomatic status, every conventional up-to-date treatment for heart failure should be applied in these patients.

The more challenging issue is the treatment of conduction disturbance. In contrast to the effect of steroids on LV function, a significant proportion of patients with conduction disturbance recover their AV conduction with steroid therapy, which may render the inserted pacemaker useless. Despite the potential reversibility of heart block, the HRS recommended device implantation under the same indication as that for non-sarcoid patients because reversibility is unpredictable.

Ventricular tachycardia (VT) and sudden death

It is not infrequent to note non-sustained or sustained VT in patients with cardiac sarcoidosis. Studies have shown divergent results on whether steroid therapy exerts a beneficial effect on VT. These results are rather expected from the theoretical viewpoint that only the right combinations between active inflammation and scar formation can result in the right milieu for the trigger and re-entry of VT, and steroids may not only potentially disrupt but may also induce the formation of these milieu.

Ablation therapy can be an option in patients refractory to medical therapy. In the initial reports, despite a successive ablation, the recurrence rates remained high. Even in a recent report suggesting favorable outcome, recurrence rate was not dramatically reduced. Eighteen among 31 patients showed recurrence after ablation therapy, in addition to 2 patients who underwent a second procedure within 3 days. In this study, arrhythmia-free survival is associated with the CMR and PET findings. Therefore, CMR and PET findings could help in selecting more suitable patients for ablation therapy.

One of the most challenging issues in cardiac sarcoidosis is the risk of sudden death and indication for ICD insertion. Although the relationship between sudden death and cardiac sarcoidosis is well known in case reports, the true incidence of sudden death in cardiac sarcoidosis is unknown. In a clinicopathologic study by Silverman et al., seven sudden deaths were noted in 23 patients with cardiac sarcoidosis (7of 23, 30%). However, this incidence is reported among patients who died of sarcoidosis, thereby, inevitably overestimating the true incidence of sudden death.
Studies on the risk stratification in patients with cardiac sarcoidosis have been carried out using various modalities. The presence of DEs in CMR, perfusion defect and abnormal FDG uptake, inducible sustained VT in programmed electrical stimulation, and RV involvement have been evaluated as possible risk stratification factors. However, it is still too early to recommend one or some of these modalities for the decision making on ICD implantation. At present, the HRS guideline should be followed for ICD insertion, in which patients with a history of cardiac arrest are listed at the top. In addition, according to the general device guideline, ICD insertion should be considered in patients with LV EF ≤ 35%. In patients who are candidates for pacemaker insertion or with unexplained syncope, ICD rather than pacemaker insertion might be a better option.