Skip main navigation

Cardiac Sarcoidosis

Epidemiology, Characteristics, and Outcome Over 25 Years in a Nationwide Study
Originally published 2015;131:624–632



This study was designed to assess the epidemiology, characteristics, and outcome of cardiac sarcoidosis (CS) in Finland.

Methods and Results—

We identified in retrospect all adult (>18 years of age) patients diagnosed with histologically confirmed CS in Finland between 1988 and 2012. A total of 110 patients (71 women) 51±9 years of age (mean±SD) were found and followed up for outcome events to the end of 2013. The annual detection rate of CS increased >20-fold during the 25-year period, reaching 0.31 in 1×105 adults between 2008 and 2012. The 2012 prevalence of CS was 2.2 in 1×105. Nearly two thirds of patients had clinically isolated CS. Altogether, 102 of the 110 patients received immunosuppressive therapy, and 56 received an intracardiac defibrillator. Left ventricular function was impaired (ejection fraction <50%) in 65 patients (59%) at diagnosis and showed no overall change over 12 months of steroid therapy. During follow-up (median, 6.6 years), 10 patients died of a cardiac cause, 11 patients underwent transplantation, and another 11 patients suffered an aborted sudden cardiac death. The Kaplan–Meier estimates for 1-, 5-, and 10-year transplantation-free cardiac survival were 97%, 90%, and 83%, respectively. Heart failure at presentation predicted poor outcome (log-rank P=0.0001) with a 10-year transplantation-free cardiac survival of only 53%.


The detection rate of CS has increased markedly in Finland over the last 25 years. With current therapy, the prognosis of CS appears better than generally considered, but patients presenting with heart failure still have poor long-term outcome.


Described first >80 years ago,1 cardiac sarcoidosis (CS) continues to puzzle clinicians for many reasons, not least for representing a disease for which the cause(s) and innermost nature of which remain unknown.2 Furthermore, its manifestations are protean with a range from silent myocardial granulomas, which may lead to sudden death,36 to symptomatic conduction disturbances, ventricular arrhythmias, and progressive heart failure.79 CS is reported to involve only 2% to 5% of patients with systemic sarcoidosis,710 but both autopsy studies35 and modern cardiac imaging1114 suggest that clinically manifest cases may represent just the tip of the iceberg in CS. There is also much evidence indicating that sarcoidosis can be clinically confined to the heart,3,6,1519 which may delay its detection.17

Clinical Perspective on p 632

Assessing prognosis and planning treatment in CS are no less a challenge. CS is considered a serious condition, with 60% to 75% 5-year survival15,20 most often quoted in the literature. The available outcome data are much more heterogeneous, although the reported prognosis varies from 73% 1-year mortality4 to 100% 10-year survival in the absence of left ventricular (LV) dysfunction.21 The problem here is that, because of the rarity of CS, the prognostic data are derived from small and selected CS populations with varying numbers of cases detected only at autopsy or after transplantation.4,15,20,21 Lack of high-quality research also complicates the planning of treatment. Although steroid therapy is considered mandatory, its efficacy has never been proven beyond limited and uncontrolled observations.2024 For these reasons, it is not surprising that experts in this field disagree on several aspects of the diagnosis and treatment of CS.25

With the uncertainties about CS in mind, we decided to collect data on all cases of histologically confirmed CS seen in Finland over the last 25 years. Our work focused on the epidemiology, characteristics, and long-term outcome of CS. Here, we report data that demonstrate a continuing increase in the detection rate of clinically manifest CS and show that its overall prognosis may be somewhat better than generally considered. Presentation with heart failure seemed to predict poor long-term outcome, however. LV function was commonly impaired and showed little overall change with steroid therapy.


Study Outline

The Myocardial Inflammatory Diseases in Finland (MIDFIN) Study Group is a cardiology research network of the 5 Finnish university hospitals that focuses on CS and giant-cell myocarditis. For the present work, cases of CS from the year 1988 on were identified from the discharge, pathology, and device registries of the member institutions of our network and from 17 central hospitals covering all of Finland. Possible patients were first screened by use of the International Classification of Disease, 10th Revision codes I41.8*D86.8 (sarcoidosis of heart) and D86 (sarcoidosis), combined with one of the following codes: I42 (cardiomyopathy), I44.1 or I44.2 (second- or third-degree atrioventricular block), I45.3 (trifascicular block), I46 (cardiac arrest), I47.2 (ventricular tachycardia), I49.0 (ventricular fibrillation), I49.3 (premature depolarization), I50 (heart failure), or R00.1 (bradycardia). Thereafter, 1 investigator (R.K.) visited each site having potential adult (>18 years of age) CS patients and scrutinized the pertinent hospital charts for data on patient demographics, symptoms, initial and later clinical manifestations, results of diagnostic imaging and laboratory studies, invasive procedures, and details of treatment with drugs and devices. The investigator also studied the records of follow-up visits for results of serial echocardiographic studies, with particular focus on LV ejection fraction (LVEF), as well as for changes in treatment and the occurrence of adverse cardiac events and side effects of treatment. The work started in 2008; thus, the data were fully retrospective for the years 1988 to 2008 but partly prospective thereafter. The last patients were included in February 2012, and adverse events were recorded up to the end of December 2013. The mortality data were double-checked from the official Finnish Population Register in January 2014. Causes of death were determined by hospital chart review. The study had the approval of the national ethical review board (STM/1219/2009). A subpopulation of this nationwide study, the patients seen at Helsinki University Central Hospital between 2000 and 2010, has been reported earlier.17

Inclusion Criteria

As in our previous works,17,26 the criteria for the diagnosis of CS and for inclusion in the present study were sarcoidosis histology in an endomyocardial biopsy or extracardiac histological verification of sarcoidosis associated with both clinical manifestations indicative of a myocardial disease and abnormalities compatible with CS in gadolinium-enhanced cardiac magnetic resonance imaging (Gd-MRI), 18-F-fluorodeoxyglucose positron emission tomography (18F-FDG PET), or echocardiography.7,27,28 As earlier,17,26 the histological diagnosis of sarcoidosis required the presence of nonnecrotizing epithelioid cell granulomas with isolated giant cells and the absence of both myocardial necrosis and more than solitary eosinophils.

Statistical Analysis

Differences between patient groups were assessed with the Mann-Whitney U test, Student t test, and χ2 test or Fisher exact test, as appropriate. Changes in LVEF over the first year of steroid therapy were analyzed with the paired Student t test. Survival curves were plotted by the Kaplan–Meier method, and factors influencing survival were analyzed by the log-rank test and Cox regression analysis. Survival was calculated from the time of first CS manifestation that was considered to be the date of first contact with healthcare professionals as a result of symptoms compatible with CS. The following events were taken as outcome end points in these analyses: cardiac death; a composite of cardiac death and transplantation, whichever came first; and a composite of cardiac death, transplantation, and aborted sudden cardiac death. The analyses were made both in the total population and after the exclusion cases diagnosed only at transplantation or autopsy. In all tests, a 2-tailed value of P<0.05 was considered statistically significant. The analyses were performed with SPSS version 21 for Windows (SPSS Inc, Chicago, IL).


Detection Rate and Prevalence of CS

A total of 110 patients 27 to 69 years of age fulfilled the criteria of CS and were included in the present work. Additionally, 34 patients with a clinical diagnosis of CS were identified at screening (see Methods) but were ultimately excluded because of either a lack of histological proof of sarcoidosis (n=21) or missing confirmative cardiac imaging (n=13). Ninety-seven of the 110 patients came from the 5 university hospitals. There was a remarkable increase over time in the number of new CS diagnoses (Figure 1). The annual detection rate of CS was 0.31 per 105 adults (>18 years of age) between 2008 and 2012. The prevalence of CS in 2012 was 2.2 per 105.

Figure 1.

Figure 1. The number of new cases of cardiac sarcoidosis diagnosed in the 5-year periods between 1988 and 2012.

The Paths to CS Diagnosis

There were variations both across hospitals and over time in the methods used to diagnose CS. All 110 patients had undergone echocardiography, and 66 and 59 patients had undergone 18F-FDG PET and cardiac Gd-MRI, respectively. Coronary angiography or coronary computed tomography had been done in 71 patients, 3 of whom were found to have angiographically significant coronary artery disease in association with CS. These individuals were retained in our analyses. Most patients (92 of 110) had undergone 1 or several (up to 3) endomyocardial biopsy sessions of the right ventricle, LV, or both ventricles. Ultimately, endomyocardial biopsy confirmed the CS diagnosis in 55 patients. In the remaining 55 patients, sarcoid histology was obtained from PET-positive mediastinal lymph nodes in 18 cases and from biopsies of lung (n=11), peripheral lymph nodes (n=8), skin (n=5), liver (n=3), and central nervous system (n=2) in the remaining 29 cases. Finally, 6 cases were diagnosed from native hearts removed at transplantation, and 2 cases were diagnosed only at autopsy. Of the 47 patients with extracardiac confirmation of sarcoid histology, 38 had typical abnormalities on cardiac 18F-FDG PET or Gd-MRI.17,27,28 The remaining 9 patients had global LV dysfunction (LVEF <50%), local thinning or thickening of the basal interventricular septum, or an LV aneurysm at echocardiography.

Patient Characteristics, Modes of Presentation, and Findings at Initial Imaging and Laboratory Studies

At presentation, 71 of the 110 patients (65%) had clinically isolated CS, defined here as cardiac involvement with neither past history nor any current signs or symptoms of extracardiac sarcoidosis by initial clinical examination, routine blood tests, and plain chest x-rays. In the remaining 39 patients (35%), CS was associated with known or clinically manifest extracardiac sarcoidosis at presentation. Seven of the 8 cases identified from cardiac explants or at autopsy were in the subgroup of clinically isolated CS. Table 1 summarizes the patients’ demographics, presenting manifestations, and key findings at the initial imaging and laboratory studies for the whole population and separately for the above subgroups. The median time from the first cardiac manifestation to diagnosis was 9.0 months (range, 0.3–168 months) in the 102 patients diagnosed before transplantation or autopsy. In patients diagnosed after transplantation or at autopsy, the median delay from first manifestation was 41.5 months (range, 5–85 months).

Table 1. Key Characteristics of the CS Population at the Time of Diagnosis (or at First Disease Manifestation in Cases Diagnosed at Transplantation or Autopsy)

All CS Patients(n=110)Clinically Isolated CS(n=71)CS With Known Extracardiac Disease(n=39)P*
Age (mean±SD), y51±951±950±90.463
Female sex, n (%)71 (65)53 (75)18 (46)0.003
First clinical manifestation, n (%)0.016
 Symptomatic atrioventricular conduction block48 (44)34 (48)14 (36)
 Ventricular tachycardia or fibrillation36 (33)27 (38)9 (23)
 Heart failure20 (18)8 (11)12 (31)
 Other6 (6)2 (3)4 (10)
New York Heart Association class, n (%)0.978
 162 (56)40 (56)22 (56)
 233 (30)22 (31)11 (28)
 3–415 (14)9 (12)6 (15)
Findings at 12-lead ECG, n (%)
 No abnormalities15/107 (14)8/69 (12)7/38 (18)0.250
 Second- or third-degree atrioventricular block48/107 (45)34/69 (49)14/38 (37)0.216
 RBBB40/107 (37)30/69 (43)10/38 (26)0.079
 LBBB22/107 (21)16/69 (23)6/38 (16)0.365
Findings at echocardiography, n (%)
 LV dilatation§43/102 (42)27/64 (42)16/38 (42)1.000
 LV systolic dysfunction (LVEF <50%)65/110 (59)49/71 (69)16/39 (41)0.004
 Interventricular septal thinning or thickening63/109 (58)48/70 (69)15/39 (38)0.002
Findings at Gd-MRI, n (%)
 Myocardial late enhancement49/59 (83)36/38 (95)13/21 (62)0.002
Findings at 18F-FDG PET, n (%)
 Focally increased cardiac FDG uptake48/66 (73)34/46 (74)14/20 (70)0.743
 FDG uptake in mediastinal lymph nodes30/42 (71)22/31 (71)8/11 (73)1.000
 FDG uptake outside heart and mediastinum16/39 (41)12/30 (40)4/9 (44)1.000
Laboratory findings, n (%)
 Elevated serum ACE22/90 (24)9/59 (15)13/31 (42)0.005
 Elevated serum lysozyme40/74 (54)21/48 (44)19/26 (73)0.016
 Elevated urinary calcium22/45 (49)9/24 (38)13/21 (62)0.029

Except for age, data are number (%) of patients with positive findings or number of positive findings per number of studied patients. ACE indicates angiotensin-converting enzyme; CS, cardiac sarcoidosis; 18F-FDG PET, 18-F-fluorodeoxyglucose position emission tomography; Gd-MRI, gadolinium-enhanced cardiac magnetic resonance imaging; LBBB, left bundle-branch block; LVEF, left ventricular; LVEF, left ventricular ejection fraction; and RBBB, right bundle-branch block.

*P values for comparison between the 2 subgroups.

Thirty-one patients with ventricular tachycardia and 5 patients with ventricular fibrillation.

Four patients with multiple ventricular premature beats, 1 patient with mitral regurgitation, and 1 patient with pericardial effusion.

§LV diastolic diameter >60 mm in men or >55 mm in women.

Treatment With Drugs and Devices

A total of 102 of the 110 patients (ie, all except those detected at autopsy or transplantation) underwent disease-modifying immunosuppressive therapy. All received steroids with the initial prednisone-equivalent doses varying from 30 to 80 mg/d, being <60 and ≥60 mg in 42 and 60 patients, respectively. Although the dosing schedule was not uniform, the prednisone-equivalent dose was usually tapered to 10 mg/d within 6 months of treatment. In 48 of the 102 patients, steroid use was uninterrupted until the end of follow-up, death, or transplantation. The remaining 54 patients were given steroids intermittently as a result of fluctuation of disease activity or steroid side effects. In 4 patients, corticosteroids were discontinued permanently because of intolerable adverse effects (severe myopathy, psychosis, aseptic joint necrosis, severe insomnia with hypokalemia, and diarrhea). Of other immunosuppressants, azathioprine was used in 50 patients, methotrexate in 6 patients, mycophenolatemofetil in 3 patients, cyclosporine in 2 patients, and infliximab in 1 patient. Symptomatic drugs included β-blockers in 104 patients, angiotensin-converting enzyme inhibitors or angiotensin receptor blockers in 89 patients, diuretics in 48 patients, and amiodarone in 25 patients. An intracardiac cardioverter-defibrillator was implanted in 59 patients (54%), and another 28 patients (25%) received a permanent pacemaker for atrioventricular conduction block.


Follow-up time from the first clinical manifestation of CS varied from 12 to 303 months (median, 79 months). A total of 10 patients died of a cardiac cause, and 6 patients suffered a noncardiac death (1 breast cancer, 1 colon cancer, 1 mesothelioma, 1 stroke, 1 Lewy body dementia, and 1 gastrointestinal perforation). Eleven patients underwent cardiac transplantation, and 2 additional patients were awaiting transplantation at the close of our study in December 2013. Two of the 11 transplanted patients died within 30 days of surgery, and another 2 patients died of allograft failure, without signs of recurrent CS in the graft, 3 and 5 years after transplantation. Thus, 20 of the 110 patients died and 7 patients were alive with a transplanted heart after a median follow-up of 6.6 years. Of the 102 biopsy-diagnosed patients, 19 died or underwent transplantation during follow-up.

Of the 10 cardiac deaths as first outcome event, 9 were sudden (2 in cases diagnosed at autopsy) and 1 was due to terminal heart failure. Another 11 patients experienced an aborted sudden death as the first event; that is, either they were successfully resuscitated from ventricular fibrillation (n=8), or ventricular fibrillation was terminated by an intracardiac cardioverter-defibrillator shock (n=3). Furthermore, an appropriate intracardiac cardioverter-defibrillator treatment for sustained VT was recorded in 15 additional patients during follow-up, but these events were not included in the outcome analyses of the present work. The Kaplan–Meier curves for cardiac survival, cardiac survival free of transplantation, and cardiac survival free of transplantation and aborted sudden death are shown for the total CS population in Figure 2A and for patients who were identified clinically and underwent immunosuppressive treatment in Figure 2B. The Kaplan–Meier estimates of 1-, 5-, and 10-year survival are given in Table 2.

Table 2. Survival Probabilities in All 110 CS Patients and in the 102 Patients Diagnosed Before Transplantation or Autopsy

Cardiac Survival, nCardiac Survival Free of TransplantationCardiac Survival Free of Transplantation and Aborted Sudden DeathCardiac Survival
1- y survival, %11099.1 (94.3–99.9)97.3 (91.6–99.3)89.1 (81.3–93.9)
102100 (95.5–100)99.0 (93.9–99.9)89.2 (81.1–94.2)
5-y survival, %11093.5 (86.7–97.1)90.0 (82.4–94.6)77.7 (68.5–84.8)
10297.0 (90.9–99.2)95.1 (88.4–98.2)82.0 (72.9–88.7)
10-y survival, %11089.3 (81.6–94.2)83.1 (74.5–89.3)70.4 (60.8–78.5)
10292.5 (85.1–96.5)90.6 (82.7–95.2)77.2 (67.6–84.7)

The numbers are Kaplan–Meier survival estimates (95% confidence intervals). CS indicates cardiac sarcoidosis.

Figure 2.

Figure 2. Kaplan–Meier curves for cardiac survival (line 1), cardiac survival free of transplantation (line 2), and cardiac survival free of transplantation and aborted sudden death (line 3) in all 110 patients with cardiac sarcoidosis (A) and in the 102 patients who were identified clinically and received immunosuppressive treatment (B).

Predictors of Outcome

Cardiac Survival Free of Transplantation

In the total population of 110 patients having 21 events, transplantation-free cardiac survival was independent of age (P=0.469, Cox regression), sex (log-rank P=0.163), and the type of CS (log-rank P=0.132) but was strongly related to heart failure as the first clinical manifestation of CS (log-rank P=0.0001; Figure 3A). In patients presenting with heart failure, the 1-, 5-, and 10-year Kaplan–Meier survival probabilities were 90% (95% confidence interval, 66.9–98.2), 75% (95% confidence interval, 50.6–90.4), and 52.5% (95% confidence interval, 29.9–74.2), respectively. Transplantation-free cardiac survival was also related to initial LVEF (P=0.006 by Cox regression and P=0.011 by log-rank test; Figure 3B) and to New York Heart Association class (log-rank P=0.023; Figure 3C).

Figure 3.

Figure 3. Kaplan–Meier curves for transplantation-free cardiac survival in all 110 patients with cardiac sarcoidosis by heart failure (HF) as the first clinical manifestation (A), by left ventricular ejection fraction (EF) at diagnosis (B), and by the initial New York Heart Association (NYHA) class (C).

Analyzed in the clinically diagnosed 102 patients with 13 outcome events, transplantation-free cardiac survival was related only to heart failure as the mode of CS presentation (log-rank P=0.025; Figure 4).

Figure 4.

Figure 4. Kaplan–Meier curves for transplantation-free cardiac survival by heart failure (HF) as the presenting manifestation in the 102 patients with cardiac sarcoidosis who were diagnosed clinically and underwent immunosuppressive therapy.

Cardiac Survival Free of Transplantation and Aborted Sudden Death

There were a total 32 events in 110 patients. Event-free survival was independent of age, sex, and heart failure as the first CS manifestation (log-rank P=0.093) but was related to the type of CS (log-rank P=0.005; Figure 5A) and to LVEF (P=0.017 by Cox regression and P=0.046 by log-rank test; Figure 5B).

Figure 5.

Figure 5. Kaplan–Meier curves for cardiac survival free of transplantation and aborted sudden death in the total population of 110 patients with cardiac sarcoidosis (CS) by the type of CS (A) and by left ventricular ejection fraction (EF) at diagnosis (B). CS+ECS indicates cardiac and extracardiac sarcoidosis; and ICS, isolated cardiac sarcoidosis.

In the 102 clinically diagnosed patients having 24 events, CS type was the only predictor of event-free cardiac survival (log-rank P=0.015; Figure 6).

Figure 6.

Figure 6. Kaplan–Meier curves for cardiac survival free of transplantation and aborted sudden death by the type of cardiac sarcoidosis (CS) in the 102 patients who were diagnosed clinically and underwent immunosuppressive therapy. CS+ECS indicates cardiac and extracardiac sarcoidosis; and ICS, isolated cardiac sarcoidosis.

Changes in LV Function and Atrioventricular Conduction During Steroid Therapy

In the 102 patients receiving immunosuppression, LVEF averaged 44.9±12% at diagnosis and 45.4±11% after on average of 12 months of steroid therapy (P=0.532). Table 3 shows the data on LVEF before and after treatment by subgroups of LV function. The data suggest that patients with severe LV impairment at diagnosis (LVEF <35%) had an improvement in LV function with steroid therapy. Altogether, 74 of the 110 patients (67%) either had LV dysfunction (LVEF <50%) at diagnosis (n=65) or developed it later during the disease course (n=9).

Table 3. LV Function Before and After an Average of 12 Months of Immunosuppressive Treatment in Patients With CS

Initial LV FunctionLVEF Before Treatment, %LVEF After 12 mo of Steroid Therapy, %P*
Normal (LVEF ≥50%), n=4456.8±5.654.9±7.60.145
Moderately impaired (LVEF, 35%–49%), n=3640.9±4.140.8±6.40.979
Severely impaired (LVEF <35%), n=2227.9±4.134.1±8.30.005

Data are mean±SD. CS indicates cardiac sarcoidosis; LV, left ventricular; and LVEF, left ventricular ejection fraction.

*Paired t test.

Initiation of steroids resulted in the recovery of cardiac conduction (defined as <10% ventricular pacing during follow-up) in 7 of 35 patients (20%) with a pacemaker implanted as a result of complete atrioventricular block.

Because the patients not receiving immunosuppression were the ones diagnosed from explantations or at autopsy, analyses of the effect of steroids (yes/no) on outcome were omitted as meaningless. Among patients receiving steroids, there was no difference in transplantation-free cardiac survival by the initial dose of prednisone (<60 versus ≥60 mg/d; log-rank P=0.561) or by the delay from symptom onset to starting steroids (<6 versus ≥6 months; log-rank P=0.867).


With a 25-year nationwide coverage, the present work provides a representative view of the characteristics and outcome of clinically manifest and histologically confirmed CS in an adult white Northern European population. Most conspicuous among our findings were the >20-fold increase in the annual detection rate of CS over the study period, the predominance of disease clinically isolated to the heart at presentation, and the high cumulative frequency of LV dysfunction during the disease course. However, the 10-year estimate of transplantation-free cardiac survival was as high as 91% in patients who were diagnosed clinically and received contemporary immunosuppressive and device therapy.

Epidemiology of CS

With an annual detection rate of 0.31 per 105 adults and with a prevalence of 2.2 per 105 adults, clinically manifest CS remains a very rare condition in Finland today. No figures from elsewhere specific to CS exist for comparison. However, because the reported general incidence of sarcoidosis in Finland is close to the incidence in other white populations,29,30 the present epidemiological data should be even more widely representative. Ethnic differences in the frequency of cardiac involvement are possible, however. An earlier comparative autopsy study revealed myocardial granulomas in 68% of Japanese patients with sarcoidosis versus in 21% of blacks and in 14% of whites.31 We admit that our data most likely underestimate the true frequency of CS because we excluded a number clinically diagnosed cases owing to a lack of confirmatory histology or cardiac imaging and because systematic cardiac screening of patients with extracardiac sarcoidosis was not practiced in Finland during the period of our study. Furthermore, the International Classification of Disease, 10th Revision classification was introduced in Finland no earlier than in 1996, so our screening may have missed solitary patients diagnosed after 1988 but lost to follow-up before 1996. Findings by other groups also suggest that the true prevalence of CS (manifest and silent but detectable cases combined) may be several times higher than the number of clinical diagnoses suggests.35,11–14

The observed increase in the annual detection rate of CS was surprisingly high. We think that improved diagnostic methods, combined with better awareness of CS and heightened resolution in the pursuit of diagnosis,17 may explain this change without a true increase in disease incidence. Of note, the largest step-up in the number of new diagnoses (Figure 1) followed the introduction of Gd-MRI and 18F-FDG PET into the routine clinical assessment of myocardial diseases in our country.

CS Without Clinically Manifest Sarcoidosis in Other Organs

Cardiac involvement without known or clinically apparent extracardiac sarcoidosis (isolated CS) has been reported both in studies of CS populations3,4,6,15,17,32 and in case histories.16,18,19,3336 In our work, 71 of 110 patients (65%) presented with clinically isolated CS, which is close to the number (57%) reported by Okura et al15 from an international registry. Although cardiac involvement is the only clinically apparent manifestation of sarcoidosis at presentation, hidden involvement of other organs can be found by 18F-FDG PET or at autopsy.3,4,6,17,27 In our isolated CS subgroup, silent extracardiac FDG uptake was common (Table 1), but still, 9 of the 30 patients undergoing whole-body PET (30%) had no signs of active sarcoidosis outside the heart. At the utmost clinical case, exclusive involvement of the heart delays the diagnosis until transplantation or autopsy. We had 7 such cases (7% of our study population), whereas in the study by Okura et al,15 a total of 12 of 42 patients (29%) had isolated CS detected at transplantation or autopsy. Similar experiences have been reported by others.35,36 Although the existence of a truly isolated CS can be questioned, we feel that recognition of this entity is important to the cardiology services where these patients present with acute cardiac problems without any clues about the origin from history, chest x-rays, or routine blood tests.

There were conspicuous differences between isolated CS and CS with manifest extracardiac disease (Table 1). Thus, judging from the higher frequency of LV dysfunction and septal abnormalities at echocardiography and from the higher prevalence of late enhancement at Gd-MRI, isolated CS was associated with a female preponderance and a more severe LV involvement. Whereas the difference in sex distribution is difficult to explain, the difference in LV involvement suggests that CS was detected in a more advanced stage of the disease when it was clinically isolated. There were also differences in serum angiotensin-converting enzyme, serum lysozyme, and daily urinary calcium excretion, all pointing toward a heavier burden of granulomatous changes in CS with manifest extracardiac disease. Isolated CS predicted worse event-free survival when aborted sudden death was included in the composite outcome end point, together with cardiac death and transplantation (Figures 5A and 6). One explanation is that there was a marked cluster of early aborted sudden arrhythmic deaths in isolated CS related to the frequency of ventricular arrhythmias at presentation (see Table 1).

Long-Term Outcome

The Kaplan–Meier estimates of cardiac survival shown here (Table 2) suggest an outlook for CS that is somewhat more favorable than what is generally presented and much better than the worst prognostic data present previously. In 1978, Roberts et al4 reported from a clinicopathological study of 113 patients that only 27% of patients were alive 1 year after symptom onset. A British study from the 1980s reported a 40% 5-year survival in CS patients.37 In a study of 95 patients collected from >40 Japanese hospitals, 5-year survival was 60% overall and 75% in the subgroup of patients diagnosed before autopsy.20 In the international multicenter study by Okura et al,15 transplantation-free 5-year survival was 60% overall and 70% in clinically diagnosed patients. In 2 small Japanese studies, 5-year survival exceeded 90%,21,22 and finally, a French study23 reported only 1 CS-related death among 41 patients followed up for nearly 5 years on average. All these works were retrospective in design. Recently, a few small studies11,13,14 have reported outcome data from cardiac screening and prospective follow-up of patients with known extracardiac sarcoidosis. Patel et al12 found a 19% cardiac mortality (4 of 21 patients) in <2 years in patients with LV late enhancement at Gd-MRI. In a similar work, Greulich et al13 followed up a group of 39 patients with cardiac symptoms and LV late enhancement and found an 8% cardiac mortality (3 of 39 patients) over a mean of 2.6 years. In contrast, Mehta et al14 found no mortality or adverse events during 2 years of follow-up in 24 sarcoidosis patients having cardiac symptoms and either PET or Gd-MRI indicative of CS. These reports indicate that CS is a potentially fatal disease, but how fatal it really is remains uncertain owing to the marked disparities in the methods and results of these works.

Although the 5-year survival probability of clinically diagnosed CS appears clearly better in our study (95%) than in the 2 most widely cited previous works15,20 (70%–75%), the data are not directly comparable because we reported transplantation-free cardiac survival, whereas Okura et al15 reported transplantation-free overall survival and Yazaki et al20 reported overall survival. Furthermore, because most of our patients were from a later era, it is possible that, as a result of improved diagnostic methods, they had less advanced cardiac involvement than what was studied in the earlier works.15,20 Differences in the ethnic composition of the study groups also confound any comparisons because Japanese CS patients have been reported to have a markedly higher cardiac mortality than black or white CS patients.31 Finally, there were also differences in treatment, for instance, in the frequency of intracardiac cardioverter-defibrillator implantations, which was highest in our study.

Our survival analyses indicated that heart failure as the first clinical CS manifestation predicted poor transplantation-free cardiac survival, along with impaired LV function at diagnosis. This is understandable because of the prognostic notoriety of heart failure in general, and it agrees well with earlier data showing that dilated LV size20 and impaired LVEF21 predict increased mortality and life-threatening arrhythmias in CS.38 In our study, event-free cardiac survival was impaired in patients with severe LV dysfunction (LVEF <35%), but there was little difference in outcome between patients with moderately depressed and those with normal LV function (LVEF, 35%–50% versus >50%; Figures 3B and 5B). The significance of isolated CS as a predictor of poor outcome was discussed above.

Reflections on Steroid Therapy

Corticosteroid therapy is generally considered mandatory in CS, even though all evidence is based on purely retrospective observations.24 Our work, although larger in size than many earlier studies, could not add much to what is known about steroids and prognosis in CS. We found, like Yazaki et al,20 that long-term outcome was independent of the initial dose of prednisone. Outcome was also independent of the delay from the first disease manifestation to the onset of treatment, but a potential confounder here is that patients with the most severe manifestations are likely to undergo an expedited diagnostic process, leading to shorter treatment delays.

Limited observational data from Japan suggest that administering steroids in CS prevents the development of systolic LV dysfunction if LVEF is initially normal21,22,39 and furthermore that it improves moderately depressed LV function21,40 but has no beneficial effects on severely impaired LV function (LVEF <30%–35%).21,39 Our data, in contrast, indicated an improvement in LV function with immunosuppression in patients with severely impaired LVEF (<35%) but no change if LVEF was normal or only moderately depressed at the start of treatment (Table 3). The weakness of all these works is that it remains unknown what the course of LV function would have been without steroids.


Some of the limitations of our work are self-explanatory such as the retrospective design with its inherent problems, the lack of standardized diagnostic and treatment practice in all participating hospitals, and the change over time in the diagnostic methods. Furthermore, although we tried to collect a comprehensive nationwide clinical series, it is possible that some patients with manifest CS were missed because of deficiencies in both hospital registries and our screening. In the epidemiological sense, it is a major limitation that our series covers only clinically manifest CS. Systematic screening of patients with sarcoidosis for silent cardiac involvement was not done, and CS presenting first as an unexpected sudden death in the community also remained outside the scope of our work. Although with 110 patients our study is one of the largest works on long-term outcome of CS, in a strict statistical sense, the number of patients and events remained small. This limits the interpretation of our survival analyses and may partly explain why there were differences in the statistically significant predictors of different composite end points. The inclusion in survival analyses of cases diagnosed postmortem or after transplantation assumes that there is no really benign form of CS and that the diagnosed cases are representative of all patients who contract the disease. Because this assumption can be questioned, we calculated the survival data also after excluding these cases from the analyses. A similar dual-analysis strategy has also been used in the prior key survival studies in CS.15,20


The number of patients with manifest CS seen annually in Finland increased >20-fold from 1988 to 2012, most likely as a result of improved diagnostic methods and heightened diagnostic activity. The majority of patients had clinically isolated CS that was characterized by female preponderance, more severe LV involvement, and less frequent elevations in the common laboratory markers of sarcoidosis. The 10-year probability of transplantation-free cardiac survival was 83% overall and 91% in patients receiving immunosuppressive treatment. Heart failure at presentation, marked LV dysfunction at diagnosis (LVEF <35%), and isolated CS type predicted impaired event-free outcome. Although the overall prognosis of CS was better than generally considered, LV dysfunction was common and poorly responsive to steroid therapy. Improvement in disease-modifying treatment is clearly needed but presupposes prospective, controlled trials, which, because of the rarity of CS, should be organized internationally.


From February 2012 through August 2014, that is, after the inclusion of the last patient in the present follow-up work, we diagnosed a total of 57 new histologically confirmed cases of CS in Finland. This gives a current annual detection rate of 0.53 per 105 adults, which is clearly more than reported above for the period from 2008 to 2012.


We thank our colleagues in all participating hospitals for help with this study.


Continuing medical education (CME) credit is available for this article. Go to to take the quiz.

Correspondence to Markku Kupari, MD, Division of Cardiology, Heart and Lung Center, Helsinki University Central Hospital, 00029 Helsinki, Finland. E-mail .


  • 1. Bernstein M, Konzelman FW, Sidlick DM. Boeck’s sarcoid: report of a case with visceral involvement.Arch Intern Med. 1929; 44:721–734.CrossrefGoogle Scholar
  • 2. Valeyre D, Prasse A, Nunes H, Uzunhan Y, Brillet PY, Müller-Quernheim J. Sarcoidosis.Lancet. 2014; 383:1155–1167. doi: 10.1016/S0140-6736(13)60680-7.CrossrefMedlineGoogle Scholar
  • 3. Matsui Y, Iwai K, Tachibana T, Fruie T, Shigematsu N, Izumi T, Homma AH, Mikami R, Hongo O, Hiraga Y, Yamamoto M. Clinicopathological study of fatal myocardial sarcoidosis.Ann N Y Acad Sci. 1976; 278:455–469.CrossrefMedlineGoogle Scholar
  • 4. Roberts WC, McAllister HA, Ferrans VJ. Sarcoidosis of the heart: a clinicopathologic study of 35 necropsy patients (group 1) and review of 78 previously described necropsy patients (group 11).Am J Med. 1977; 63:86–108.CrossrefMedlineGoogle Scholar
  • 5. Silverman KJ, Hutchins GM, Bulkley BH. Cardiac sarcoid: a clinicopathologic study of 84 unselected patients with systemic sarcoidosis.Circulation. 1978; 58:1204–1211.LinkGoogle Scholar
  • 6. Tavora F, Cresswell N, Li L, Ripple M, Solomon C, Burke A. Comparison of necropsy findings in patients with sarcoidosis dying suddenly from cardiac sarcoidosis versus dying suddenly from other causes.Am J Cardiol. 2009; 104:571–577. doi: 10.1016/j.amjcard.2009.03.068.CrossrefMedlineGoogle Scholar
  • 7. Kim JS, Judson MA, Donnino R, Gold M, Cooper LT, Prystowsky EN, Prystowsky S. Cardiac sarcoidosis.Am Heart J. 2009; 157:9–21. doi: 10.1016/j.ahj.2008.09.009.CrossrefMedlineGoogle Scholar
  • 8. Dubrey SW, Falk RH. Diagnosis and management of cardiac sarcoidosis.Prog Cardiovasc Dis. 2010; 52:336–346. doi: 10.1016/j.pcad.2009.11.010.CrossrefMedlineGoogle Scholar
  • 9. Blauwet LA, Cooper LT. Idiopathic giant cell myocarditis and cardiac sarcoidosis.Heart Fail Rev. 2013; 18:733–746. doi: 10.1007/s10741-012-9358-3.CrossrefMedlineGoogle Scholar
  • 10. Baughman RP, Teirstein AS, Judson MA, Rossman MD, Yeager H, Bresnitz EA, DePalo L, Hunninghake G, Iannuzzi MC, Johns CJ, McLennan G, Moller DR, Newman LS, Rabin DL, Rose C, Rybicki B, Weinberger SE, Terrin ML, Knatterud GL, Cherniak R; Case Control Etiologic Study of Sarcoidosis (ACCESS) Research Group. Clinical characteristics of patients in a case control study of sarcoidosis.Am J Respir Crit Care Med. 2001; 164(pt 1):1885–1889. doi: 10.1164/ajrccm.164.10.2104046.CrossrefMedlineGoogle Scholar
  • 11. Patel MR, Cawley PJ, Heitner JF, Klem I, Parker MA, Jaroudi WA, Meine TJ, White JB, Elliott MD, Kim HW, Judd RM, Kim RJ. Detection of myocardial damage in patients with sarcoidosis.Circulation. 2009; 120:1969–1977. doi: 10.1161/CIRCULATIONAHA.109.851352.LinkGoogle Scholar
  • 12. Patel AR, Klein MR, Chandra S, Spencer KT, Decara JM, Lang RM, Burke MC, Garrity ER, Hogarth DK, Archer SL, Sweiss NJ, Beshai JF. Myocardial damage in patients with sarcoidosis and preserved left ventricular systolic function: an observational study.Eur J Heart Fail. 2011; 13:1231–1237. doi: 10.1093/eurjhf/hfr099.CrossrefMedlineGoogle Scholar
  • 13. Greulich S, Deluigi CC, Gloekler S, Wahl A, Zürn C, Kramer U, Nothnagel D, Bültel H, Schumm J, Grün S, Ong P, Wagner A, Schneider S, Nassenstein K, Gawaz M, Sechtem U, Bruder O, Mahrholdt H. CMR imaging predicts death and other adverse events in suspected cardiac sarcoidosis.JACC Cardiovasc Imaging. 2013; 6:501–511. doi: 10.1016/j.jcmg.2012.10.021.CrossrefMedlineGoogle Scholar
  • 14. Mehta D, Lubitz SA, Frankel Z, Wisnivesky JP, Einstein AJ, Goldman M, Machac J, Teirstein A. Cardiac involvement in patients with sarcoidosis: diagnostic and prognostic value of outpatient testing.Chest. 2008; 133:1426–1435. doi: 10.1378/chest.07-2784.CrossrefMedlineGoogle Scholar
  • 15. Okura Y, Dec GW, Hare JM, Kodama M, Berry GJ, Tazelaar HD, Bailey KR, Cooper LT. A clinical and histopathologic comparison of cardiac sarcoidosis and idiopathic giant cell myocarditis.J Am Coll Cardiol. 2003; 41:322–329.CrossrefMedlineGoogle Scholar
  • 16. Brown ML, Reeder G, Unni KK, Mullany C. Intraoperative diagnosis of isolated cardiac sarcoid.Heart Lung Circ. 2007; 16:315–317. doi: 10.1016/j.hlc.2006.07.009.CrossrefMedlineGoogle Scholar
  • 17. Kandolin R, Lehtonen J, Graner M, Schildt J, Salmenkivi K, Kivistö SM, Kupari M. Diagnosing isolated cardiac sarcoidosis.J Intern Med. 2011; 270:461–468. doi: 10.1111/j.1365-2796.2011.02396.x.CrossrefMedlineGoogle Scholar
  • 18. Sugizaki Y, Tanaka H, Imanishi J, Konishi A, Yamashita T, Shinke T, Ishida T, Kawai H, Hirata K. Isolated primary cardiac sarcoidosis presenting as acute heart failure.Intern Med. 2013; 52:71–74.CrossrefMedlineGoogle Scholar
  • 19. Nery PB, Keren A, Healey J, Leug E, Beanlands RS, Birnie DH. Isolated cardiac sarcoidosis: establishing the diagnosis with electroanatomic mapping-guided endomyocardial biopsy.Can J Cardiol. 2013; 29:1015.e1–1015.e3. doi: 10.1016/j.cjca.2012.09.009.CrossrefMedlineGoogle Scholar
  • 20. Yazaki Y, Isobe M, Hiroe M, Morimoto S, Hiramitsu S, Nakano T, Izumi T, Sekiguchi M; Central Japan Heart Study Group. Prognostic determinants of long-term survival in Japanese patients with cardiac sarcoidosis treated with prednisone.Am J Cardiol. 2001; 88:1006–1010.CrossrefMedlineGoogle Scholar
  • 21. Chiu CZ, Nakatani S, Zhang G, Tachibana T, Ohmori F, Yamagishi M, Kitakaze M, Tomoike H, Miyatake K. Prevention of left ventricular remodeling by long-term corticosteroid therapy in patients with cardiac sarcoidosis.Am J Cardiol. 2005; 95:143–146. doi: 10.1016/j.amjcard.2004.08.083.CrossrefMedlineGoogle Scholar
  • 22. Kato Y, Morimoto S, Uemura A, Hiramitsu S, Ito T, Hishida H. Efficacy of corticosteroids in sarcoidosis presenting with atrioventricular block.Sarcoidosis Vasc Diffuse Lung Dis. 2003; 20:133–137.MedlineGoogle Scholar
  • 23. Chapelon-Abric C, de Zuttere D, Duhaut P, Veyssier P, Wechsler B, Huong DL, de Gennes C, Papo T, Blétry O, Godeau P, Piette JC. Cardiac sarcoidosis: a retrospective study of 41 cases.Medicine (Baltimore). 2004; 83:315–334.CrossrefMedlineGoogle Scholar
  • 24. Sadek MM, Yung D, Birnie DH, Beanlands RS, Nery PB. Corticosteroid therapy for cardiac sarcoidosis: a systematic review.Can J Cardiol. 2013; 29:1034–1041. doi: 10.1016/j.cjca.2013.02.004.CrossrefMedlineGoogle Scholar
  • 25. Hamzeh NY, Wamboldt FS, Weinberger HD. Management of cardiac sarcoidosis in the United States: a Delphi study.Chest. 2012; 141:154–162. doi: 10.1378/chest.11-0263.CrossrefMedlineGoogle Scholar
  • 26. Kandolin R, Lehtonen J, Kupari M. Cardiac sarcoidosis and giant cell myocarditis as causes of atrioventricular block in young and middle-aged adults.Circ Arrhythm Electrophysiol. 2011; 4:303–309. doi: 10.1161/CIRCEP.110.959254.LinkGoogle Scholar
  • 27. Ohira H, Tsujino I, Ishimaru S, Oyama N, Takei T, Tsukamoto E, Miura M, Sakaue S, Tamaki N, Nishimura M. Myocardial imaging with 18F-fluoro-2-deoxyglucose positron emission tomography and magnetic resonance imaging in sarcoidosis.Eur J Nucl Med Mol Imaging. 2008; 35:933–941. doi: 10.1007/s00259-007-0650-8.CrossrefMedlineGoogle Scholar
  • 28. Youssef G, Beanlands RS, Birnie DH, Nery PB. Cardiac sarcoidosis: applications of imaging in diagnosis and directing treatment.Heart. 2011; 97:2078–2087. doi: 10.1136/hrt.2011.226076.CrossrefMedlineGoogle Scholar
  • 29. Pietinalho A, Hiraga Y, Hosoda Y, Löfroos AB, Yamaguchi M, Selroos O. The frequency of sarcoidosis in Finland and Hokkaido, Japan: a comparative epidemiological study.Sarcoidosis. 1995; 12:61–67.MedlineGoogle Scholar
  • 30. Rybicki BA, Major M, Popovich J, Maliarik MJ, Iannuzzi MC. Racial differences in sarcoidosis incidence: a 5-year study in a health maintenance organization.Am J Epidemiol. 1997; 145:234–241.CrossrefMedlineGoogle Scholar
  • 31. Iwai K, Sekiguti M, Hosoda Y, DeRemee RA, Tazelaar HD, Sharma OP, Maheshwari A, Noguchi TI. Racial difference in cardiac sarcoidosis incidence observed at autopsy.Sarcoidosis. 1994; 11:26–31.MedlineGoogle Scholar
  • 32. Fleming HA. Sarcoid heart disease.Br Heart J. 1974; 36:54–68.CrossrefMedlineGoogle Scholar
  • 33. Meyer T, Lauschke J, Ruppert V, Richter A, Pankuweit S, Maisch B. Isolated cardiac sarcoidosis associated with the expression of a splice variant coding for a truncated BTNL2 protein.Cardiology. 2008; 109:117–121. doi: 10.1159/000105552.CrossrefMedlineGoogle Scholar
  • 34. Galati G, Leone O, Rapezzi C. The difficult diagnosis of isolated cardiac sarcoidosis: usefulness of an integrated MRI and PET approach.Heart. 2014; 100:89–90. doi: 10.1136/heartjnl-2013-304237.CrossrefMedlineGoogle Scholar
  • 35. Donsky AS, Escobar J, Capehart J, Roberts WC. Heart transplantation for undiagnosed cardiac sarcoidosis.Am J Cardiol. 2002; 89:1447–1450.CrossrefMedlineGoogle Scholar
  • 36. Tsai JH, Chou NK, Wang SS, Shun CT. Isolated cardiac sarcoidosis: case experience in heart transplantation.J Formos Med Assoc. 2013; 112:499–500. doi: 10.1016/j.jfma.2011.12.010.CrossrefMedlineGoogle Scholar
  • 37. Fleming HA, Bailey SM. The prognosis of sarcoid heart disease in the United Kingdom.Ann N Y Acad Sci. 1986; 465:543–550.CrossrefMedlineGoogle Scholar
  • 38. Schuller JL, Zipse M, Crawford T, Bogun F, Beshai J, Patel AR, Sweiss NJ, Nguyen DT, Aleong RG, Varosy PD, Weinberger HD, Sauer WH. Implantable cardioverter defibrillator therapy in patients with cardiac sarcoidosis.J Cardiovasc Electrophysiol. 2012; 23:925–929. doi: 10.1111/j.1540-8167.2012.02350.x.CrossrefMedlineGoogle Scholar
  • 39. Yodogawa K, Seino Y, Ohara T, Takayama H, Katoh T, Mizuno K. Effect of corticosteroid therapy on ventricular arrhythmias in patients with cardiac sarcoidosis.Ann Noninvasive Electrocardiol. 2011; 16:140–147. doi: 10.1111/j.1542-474X.2011.00418.x.CrossrefMedlineGoogle Scholar
  • 40. Kudoh H, Fujiwara S, Shiotani H, Kawai H, Hirata K. Myocardial washout of 99mTc-tetrofosmin and response to steroid therapy in patients with cardiac sarcoidosis.Ann Nucl Med. 2010; 24:379–385. doi: 10.1007/s12149-010-0376-8.CrossrefMedlineGoogle Scholar


Cardiac sarcoidosis (CS) is a rare and difficult-to-diagnose disease that challenges clinicians with its heterogeneous manifestations and potentially fatal outcome. In the present nationwide study, we evaluated the epidemiology, characteristics, and long-term outcome of clinically manifest and histologically confirmed CS over the 25-year period from 1988 to 2012 in Finland. A total of 110 patients (64% female; mean age, 51 years) fulfilled the diagnostic criteria and were included in our retrospective work. The annual number of new cases grew >20-fold during the study period. At presentation to the cardiology services, two thirds of patients had clinically isolated CS, indicating that they had neither past history of sarcoidosis nor symptoms or signs of extracardiac involvement at the clinical examination or in routine laboratory tests or chest x-rays. Nearly 80% of the patients presented acutely as a result of either a high-grade atrioventricular conduction block or serious ventricular tachyarrhythmias, most of the rest presenting with heart failure. During the median follow-up of 6.6 years, 10 patients died of a cardiac cause, 11 underwent transplantation, and another 11 experienced an aborted sudden arrhythmic cardiac death. Transplantation-free cardiac survival was 97%, 90%, and 83% after 1, 5, and 10 years of follow-up, respectively. Heart failure as the first manifestation predicted worse outcome. The 2012 prevalence of clinically manifest CS in Finland was 22 patients per 1 million adult population. According to our latest data, 5.3 new cases of CS per 1 million adults are diagnosed and need treatment each year.


eLetters should relate to an article recently published in the journal and are not a forum for providing unpublished data. Comments are reviewed for appropriate use of tone and language. Comments are not peer-reviewed. Acceptable comments are posted to the journal website only. Comments are not published in an issue and are not indexed in PubMed. Comments should be no longer than 500 words and will only be posted online. References are limited to 10. Authors of the article cited in the comment will be invited to reply, as appropriate.

Comments and feedback on AHA/ASA Scientific Statements and Guidelines should be directed to the AHA/ASA Manuscript Oversight Committee via its Correspondence page.