Medical Therapy and Clinical Outcomes in Cardiac Sarcoidosis Patients With Systolic Heart Failure

Abstract

Background: Cardiac sarcoidosis (CS) may result in systolic heart failure (heart failure with reduced ejection fraction [HFrEF]), but its response to guideline-directed medical therapy (GDMT) remains uncertain.

Methods and Results: We investigated 881 patients evaluated for CS to identify those with diagnosed CS, left ventricular ejection fraction (LVEF) ≤40% at diagnosis, and follow-up echocardiogram within 11–24 months. Demographics, LVEF, GDMT as quantified by Kansas City Medical Optimization (KCMO) score, and immunosuppressive treatment were recorded. The primary outcome was a composite of event-free survival (unplanned heart failure hospitalization, left ventricular assist device [LVAD]/heart transplant, or death). Seventy-nine (9%) CS patients met the inclusion criteria (35% female, median age 57 years, mean LVEF 30.9%, median New York Heart Association class II [46%], mean number of GDMT agents 1.7, and mean KCMO score 31.8). Most (87%) were treated with immunosuppressive treatment. At follow-up (median 16 months), the mean number of GDMT agents increased to 2.2 (P=0.02), and the mean KCMO score to 70.1 (P<0.001). Mean LVEF improved to 39.9% (excluding LVAD/transplant; P<0.001) and the change in LVEF was correlated with follow-up KCMO score (P<0.001). The primary outcome occurred in 13 (16%) patients and differed by KCMO score (log-rank P<0.001), but not by immunosuppressive treatment (log-rank P=0.36).

Conclusions: GDMT optimization is associated with better cardiac remodeling and clinical outcomes in CS patients with HFrEF.

Cardiac sarcoidosis (CS) is an infiltrative cardiomyopathy characterized by non-caseating granulomatous myocardial inflammation.^1,2 Heart failure (HF) is a well-recognized presentation of CS, occurring in up to 50% of CS patients, although it is less common than arrhythmia or sudden cardiac death, which are the most common presenting findings of the disease.³ Left ventricular dysfunction is the most important predictor of survival in CS, with 10-year transplantation-free cardiac survival of only 53% in CS patients presenting with HF.^3,4 The etiology of HF in CS is not fully understood, and may be related to granulomatous infiltration, myocardial scarring, or arrhythmic dyssynchrony resulting in HF with reduced ejection fraction (HFrEF) or diastolic HF.⁵

Data on HF management in CS are sparse, with current principles of management focusing on providing guideline-directed medical therapy (GDMT) for HF as per national/international societal guidelines for HF management.⁶ However, these management principles are mostly extrapolated from existing studies of HFrEF unrelated to CS. Thus, there is a great need for data describing clinical responses in CS patients with HFrEF treated with GDMT. We examined a large tertiary academic referral center cohort of CS patients with HFrEF to evaluate reverse remodeling in response to GDMT and associated clinical outcomes, and hypothesized that the extent of GDMT support would be associated with improved remodeling and outcomes.

Methods

Study Protocol

This study was performed in accordance with the Declaration of Helsinki and the protocol was approved by the Mayo Clinic Institutional Review Board (IRB ID: 17-000976). A retrospective cohort study design was used to evaluate the electronic medical records of patients diagnosed with CS at 3 tertiary academic referral centers located in 3 different US states.

Cohort Definition

We analyzed our cohort of patients evaluated for CS (January 2000–December 2022) to isolate those diagnosed with CS, HFrEF (left ventricular ejection fraction [LVEF] ≤40%) based on transthoracic echocardiogram (TTE) within 30 days of CS index evaluation, and follow-up TTE 11–24 months after CS index evaluation.⁷ CS diagnosis was categorized by modified Heart Rhythm Society (HRS) criteria as either definite CS (histological evidence in the myocardium), probable CS (histological extracardiac evidence and supporting clinical/imaging evidence), or presumed/clinical CS (cardiac imaging evidence and compatible arrhythmia with or without extracardiac clinical/imaging findings; Supplementary Table).^8,9

Data Collection

We collected baseline demographic data, functional status (New York Heart Association [NYHA] class), pertinent laboratory studies (potassium, estimated glomerular filtration rate by creatinine [eGFR], high-sensitivity cardiac troponin, and N-terminal pro B-type natriuretic peptide [NT-proBNP]), cardiac implantable electronic device (CIED) status, LVEF, left ventricular end-diastolic diameter (LVEDD), positron emission tomography (PET), cardiac magnetic resonance imaging (CMR), immunosuppressive treatment, and GDMT at index evaluation using a combination of manual chart review and regular expression-based textual search of patient chart data.¹⁰ For standardization, doses of angiotensin-converting enzyme inhibitors (ACEi), angiotensin receptor blockers (ARB), angiotensin receptor–neprilysin inhibitors (ARNI), β-blockers, and mineralocorticoid receptor antagonists (MRA) were reported as a percentage of the target dose per the 2022 American Heart Association (AHA)/American College of Cardiology (ACC)/Heart Failure Society of America guidelines for heart failure management.⁷ Where target dose ranges are provided, the lowest target dose threshold was chosen. Loop diuretic doses were converted to furosemide equivalents. In addition, we used the Kansas City Medical Optimization (KCMO) score to describe GDMT optimization in individual patients. The KCMO is a validated score designed to quantify GDMT (range 0–100) as guidelines and GDMT contraindications evolve over time.¹¹ Immunosuppressive treatment (prednisone, mycophenolate, azathioprine, leflunomide/thalidomide, methotrexate, rituximab, cyclophosphamide, and adalimumab/infliximab), with approximate duration of therapy and mean dose, was recorded. Our institutional treatment protocol for prednisone includes an initial dose of prednisone 30 mg daily for 4 weeks followed by incremental reductions every 4 weeks (30, 25, 20, 15, 10, 7.5, and 5 mg daily), although this plan can be individualized, depending on clinical circumstances, by the treating clinician. Prophylaxis for Pneumocystis pneumonia is recommended while the patient is taking ≥15 mg prednisone daily, with either trimethoprim/sulfamethoxazole (TMP-SMX) or an alternative agent as per clinician preference.

At follow-up TTE (obtained 11–24 months after CS index evaluation), LVEF, LVEDD, laboratory studies, and GDMT were recorded.

Remodeling Outcomes

We evaluated the extent of reverse remodeling as defined by: (1) the change in LVEF between index evaluation and follow-up; and (2) achievement of HF with improved ejection fraction (HFimpEF; LVEF >40%) at follow-up.⁷

Clinical Outcomes

The primary clinical outcome was a composite of unplanned HF hospitalization, left ventricular assist device (LVAD) implantation, heart transplant, or death. Secondary outcomes included documented hyperkalemia (clinical or subclinical), all-cause mortality, and arrhythmia. The arrhythmia outcome was defined as event-free survival to the first sustained ventricular tachycardia (VT) or fibrillation (VF) event, or, in patients with CIED, treated arrhythmia events were also collected through device follow-up data available in an integrated device management system.

Fluoro-D-Glucose (FDG)-PET Analysis

When available, PET was analyzed for ¹³N-ammonia uptake and 2-deoxy-2-[fluorine-18] FDG avidity of the myocardium. PET scans were only analyzed if collected after 2016 to eliminate confounding due to historically less rigorous PET protocols. PET scans were only included if collected within 30 days of CS index evaluation (baseline) and 11–24 months after CS index evaluation (follow-up), to align with the TTE inclusion criteria. Where multiple PET scans were available, the closest PET scan to the TTE date was analyzed.

PET scans were performed in accordance with established institutional and published protocols, which are well described in previously published works.^12–14 PET data were abstracted through a combination of regular expression-based textual search of FDG-PET reports and manual abstraction by 2 investigators in the Department of Cardiovascular Medicine (K.Y. and S.R.).¹⁰

Statistical Analysis

The Mayo Clinic database is collected and managed using REDCap electronic data capture tools.^15,16 Statistical analysis was performed using jamovi Ver 2.4.2 (The Jamovi Project, Sydney, NSW, Australia). Baseline data were analyzed using descriptive statistics, including mean±SD and the median with interquartile range (IQR). Normality was tested per variable with the Shapiro-Wilk test. The Wilcoxon rank test was used to compare continuous, non-parametric variables. Chi-squared and Fisher’s exact tests were used to compare unpaired categorical variables. The McNemar and McNemar-Bowker tests were used to compare paired categorical variables. The Kruskal-Wallis test was used to evaluate the relationship between LVEF change at follow-up and the number of GDMT agents at follow-up. Relationships between remodeling outcomes and multiple predictor variables were analyzed by univariable linear regression (for LVEF change) and univariable binomial logistic regression (for HFimpEF). Kaplan-Meier survival and log-rank univariable analyses were used to calculate a survival curve with hazard ratios (HRs) and 95% confidence intervals (CI) for the primary outcome categorized by immunosuppressive treatment and the number of GDMT agents at follow-up. P<0.05 (two-tailed) was selected as the threshold of significance for all analyses.

Results

Cohort Demographics

From the original cohort of 881 patients evaluated for CS, 281 were excluded because CS was later ruled out, 383 were excluded due to a lack of HFrEF at index evaluation, and 138 were excluded due to a lack of TTE available within the outlined time frame (Figure 1). The final cohort analyzed included 79 CS patients with HFrEF (median age 57 years, 35% female). Baseline cohort demographics are presented in Table 1. Most patients had probable CS (47%) and fewer had definite CS (25%) or presumed CS (28%; 46% of whom had negative genetic cardiomyopathy testing). Probable CS patients were diagnosed by extracardiac biopsy (68% lung/mediastinal lymph node, 16% multiple organs, 8% other lymph node, 5% skin, 3% central nervous system), and FDG uptake on PET imaging consistent with CS (32%); late gadolinium enhancement on CMR imaging consistent with CS (11%); unexplained reduced LVEF <40% (24%); unexplained high-grade atrioventricular block (19%); or unexplained sustained VT (14%). Among probable CS and presumed CS patients, 11% and 18%, respectively, had a prior endomyocardial biopsy specimen without findings of sarcoidosis. Overall, the cohort was mostly White, male, with mild obesity, and with comorbidity burden of chronic kidney disease (15%), coronary artery disease (8%), atrial fibrillation (33%), and diabetes (18%). Nearly one-third of patients were current or former smokers. Most patients had good functional capacity (64% NYHA Class I–II). Baseline CMR was available in 46% (n=37) of patients and most (93%) had evidence of late gadolinium enhancement compatible with CS.

Figure 1.

Consort diagram describing the exclusion of patients from the final analyzed cohort. CS, cardiac sarcoidosis; HFrEF, heart failure with reduced ejection fraction; LVEF, left ventricular ejection fraction; TTE, transthoracic echocardiography.

Table 1.

Characteristics of CS Patients With HFrEF at Index Evaluation (n=79)

Median age at CS diagnosis (years)	57
Female sex (%)	35.4
Race (%)
White	91.1
Black	5.1
Asian	3.8
CS diagnostic category (%)
Definite	25.3
Probable	46.8
Presumed	27.8
Probable CS extracardiac biopsy location (%)
Central nervous system	2.7
Lung/mediastinal lymph node	67.6
Multiple organs	16.2
Other lymph node	8.1
Skin	5.4
Negative endomyocardial biopsy specimen (%)
Presumed CS	18.2
Probable CS	10.8
Probable CS diagnostic criteria (%)
Steroid±immunosuppressant-responsive cardiomyopathy or heart block	0.0
Unexplained reduced LVEFA (≤40%)	24.3
Unexplained sustained (spontaneous or induced) VT	13.5
Mobitz Type II second- or third-degree heart block	18.9
Patchy uptake on dedicated cardiac PETB	32.4
LGE on CMRB	10.8
Positive gallium uptakeB	0.0
Presumed CS diagnostic criteriaC (%)
Patchy uptake on dedicated cardiac PET+VT	50.0
Patchy uptake on dedicated cardiac PET+AVB	22.7
LGE on CMRB+VT	13.6
LGE on CMRB+AVB	9.1
Other	4.5
Presumed CS patients with negative genetic cardiomyopathy testing (%)	43.5
Cardiac implantable electronic device (%)	57.0
Implantable cardiac resynchronization therapy defibrillator	44.4
Implantable cardioverter defibrillator	44.4
Permanent pacemaker	11.1
Arrhythmia history (%)
Right bundle branch block	21.5
Left bundle branch block	21.5
First-degree AVB	22.8
Second-degree AVB Type I	2.5
Second-degree AVB Type II	5.1
Third-degree AVB	31.6
Premature ventricular contraction burden >10%	53.2
Non-sustained VT	44.3
Sustained VT/VF	41.8
LGE compatible with CS on CMR within 30 days of index evaluation (n=37)	93.3
Current or former smoker (%)	31.6
Mean BMI (kg/m2)	30.6
CKD Stage III or greater (%)	15.2
CAD, moderate (>40% stenosis) or greater (%)	7.7
Atrial fibrillation (%)	32.9
Diabetes(%)	17.7
NYHA class at index evaluation (%)
I	17.9
II	46.2
III	29.5
IV	6.4

^ADenotes patients who did not meet any other probable CS diagnostic criteria, because all patients in this cohort had reduced LVEF at CS index evaluation. ^BConsistent with CS. ^CSee Supplementary Table. AVB, atrioventricular block; BMI, body mass index; CAD, coronary artery disease; CKD, chronic kidney disease; CMR, cardiac magnetic resonance; CS, cardiac sarcoidosis; HFrEF, heart failure with reduced ejection fraction; LGE, late gadolinium enhancement; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; PET, positron emission tomography; VF, ventricular fibrillation; VT, unexplained sustained (spontaneous or induced) ventricular tachycardia.

GDMT

At index evaluation, patients were on a mean of 1.7 GDMT agents and the mean KCMO score was 31.8 (Table 2; Figure 2A). Most patients were prescribed ACEi/ARB/ARNI (56%) or β-blockers (80%), whereas a minority were prescribed MRA (28%) and sodium-glucose cotransporter 2 inhibitors (SGLT2i; 6%). Nearly half the patients were on a loop diuretic (44%).

Table 2.

HFrEF/CS Metrics and Medications at Index Evaluation and Follow-up

	Index evaluation	Follow-up	P value
Mean LVEF (%; n=68 excluding LVAD/transplant)	30.9	39.9	<0.001A,*
Mean LVEDD (mm; n=68 excluding LVAD/transplant)	60.6	58,1	0.10A
Mean no. abnormal perfusion segments (n=32)	6.4	6.3	0.70B
Mean no. FDG avid segments (n=32)	5.0	2.5	0.007B,*
Mean SUVmax of myocardium (n=32)	4.5	2.4	<0.001B,*
Mean SUVmax of myocardium/SUVmax of blood pool ratio (n=32)	2.0	1.1	<0.001B,*
NYHA class (%)
I	17.9	28.2	<0.001C,*
II	46.2	46.2
III	29.5	23.1
IV	6.4	2.6
eGFR by creatinine (mL/min/1.73 m2; n=51)	61.4	60.6	0.32A
Potassium (mEq/L; n=59)	5.0	4.5	0.03A,*
NT-proBNP (pg/mL; n=36)	1,196.4	1,035.9	0.09A
hs-cTn (ng/L; n=17)	17.4	46.7	0.98A
CIED (%)	57.0	74.7	0.006D,*
Kansas City Medical Optimization score	31.8	70.1	<0.001
ACEi/ARB/ARNI
% Patients on medication	56	80	<0.001D,*
Mean % of target dose	48	89	<0.001A,*
ACEi
% Patients on medication	32	29	0.16D
Mean % of target dose	45	91	<0.001A,*
ARB
% Patients on medication	10	10	1.0D
Mean % of target dose	100	80	0.37A
ARNI
% Patients on medication	15	39	<0.001D,*
Mean % of target dose	40	84	0.009B,*
β-blockers
% Patients on medication	80	87	0.18D
Mean % of target dose	43	77	<0.001A,*
MRA
% Patients on medication	28	39	0.04D,*
Mean % of target dose	88	91	1.0A
SGLT2i
% Patients on medication	6	18	0.007D,*
Loop diuretic
% Patients on medication	44	43	0.83D
Mean daily dose (furosemide equivalents, mg)	55.7	44.3	0.66A
Digoxin
% Patients on medication	5	3	0.16D
Mean daily dose (excluding patients not taking the medication μg)	125.0	125.0	N/A

*P values meeting the threshold of statistical significance. ^AWilcoxon rank test; ^Bpaired Student’s t-test; ^CFisher’s exact test; ^DMcNemar test. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor-neprilysin inhibitor; CIED, cardiac implantable electronic device; eGFR, estimated glomerular filtration rate; FDG, fluorodeoxyglucose; hs-cTn, high-sensitivity cardiac troponin; LVAD, left ventricular assist device; LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction; MRA, mineralocorticoid antagonist; NT-proBNP, N-terminal pro B-type natriuretic peptide; NYHA, New York Heart Association; SGLT2i, sodium-glucose cotransporter 2 inhibitor; SUV_max, maximum standardized uptake. Other abbreviations as in Table 1.

Figure 2.

(A) Guideline-directed medical therapy (GDMT) and (B) left ventricular ejection fraction (LVEF) at index evaluation and follow-up. ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blocker; ARNI, angiotensin receptor-neprilysin inhibitor; KCMO, Kansas City Medical Optimization; MRA, mineralocorticoid antagonist; SGLT2i, sodium-glucose cotransporter 2 inhibitor.

At follow-up, the mean number of GDMT agents increased to 2.2 (P<0.001) and the mean KCMO increased score to 70.1 (P<0.001). Significant increases in the prescription of 3 classes of GDMT were noted: ACEi/ARB/ARNI (80%; P<0.001), MRA (39%; P=0.04), and SGLT2i (18%; P=0.007). In patients prescribed each medication, significant increases in the mean percentage of target doses were seen with ACEi/ARB/ARNI overall (from 48% to 89%; P<0.001), ACEi (from 45% to 91%; P<0.001), ARNI (from 40% to 84%; P=0.009), and β-blockers (from 43% to 77%; P<0.001). Loop diuretic usage or dose did not change. Seven patients were not prescribed any GDMT at follow-up, which was due to heart transplant in 6 patients and LVAD implantation in 1 patient, at a time when ACC/AHA heart failure guidelines did not comment on GDMT recommendations after LVAD implantation.¹⁷

FDG-PET Analysis

Baseline and follow-up FDG-PET data were available for 32 patients (Table 2). On semiquantitative evaluation, abnormal perfusion segments did not change significantly (from 6.4 to 6.3 mean segments; P=0.43), but FDG avid segments decreased (from 5.0 to 2.5 mean segments; P=0.007). On quantitative evaluation, myocardial FDG uptake decreased significantly by both metrics, namely maximum standardized uptake (SUV_max) of the myocardium (from 4.5 to 2.4; P<0.001) and the ratio of SUV_max of the myocardium to the SUV_max of the blood pool (from 2.0 to 1.1; P<0.001).

Remodeling Outcomes

At follow-up (median 16 months), mean LVEF improved from 30.9% to 39.9% (mean [±SD] change in LVEF +8.9±10.6%), excluding patients who received LVAD/transplant (P<0.001; Figure 2B). There was a non-significant trend towards a decrease in LVEDD (from 60.6 to 58.1 mm; P=0.07). HFimpEF was achieved in 44% (n=30) of patients. Associations between primary remodeling outcomes and CS/HF metrics and treatment are presented in Table 3.

Table 3.

Cohort CS/HF Metrics and Treatment Stratified by Primary Remodeling Outcomes

	Remodeling outcome 1: Change in LVEF			Remodeling outcome 2: HFimpEF at follow-up
	Linear regression			Binomial logistic regression
	R2	β coefficient	P value	R2 A	β coefficient	P value
Median age at CS diagnosis (years)	0.02	−0.15	0.29	<0.01	0.01	0.63
Female sex	<0.01	−0.63	0.82	0.01	−0.58	0.27
Race
White	<0.01	Ref.	Ref.	0.01	Ref.	Ref.
Black		−3.08	0.63		−0.43	0.73
Asian		−2.08	0.74		0.95	0.45
CS diagnostic category
Definite	0.04	Ref.	Ref.	0.02	Ref.	Ref.
Probable		−0.88	0.81		0.92	0.22
Presumed		−4.96	0.21		0.80	0.32
Probable CS extracardiac biopsy location (%)
Central nervous system	0.39	26.43	0.009*	0.04	16.65	0.99
Lung/mediastinal lymph node		Ref.	Ref.		Ref.	Ref.
Multiple organs		4.24	0.32		0.09	0.92
Other lymph node		−2.57	0.66		−0.61	0.64
Skin		3.43	0.62		0.09	0.95
Negative endomyocardial biopsy specimen (%)	<0.01	−2.71	0.52	0.02	−1.14	0.19
Probable CS diagnostic criteria (%)
Unexplained reduced LVEFB (≤40%)	0.33	4.08	0.31	0.12	0.18	0.85
Unexplained sustained (spontaneous or induced) VT		−5.30	0.27		<0.01	1.0
Mobitz Type II second- or third-degree heart block		12.59	0.006*		2.2	0.08
Patchy uptake on dedicated cardiac PET		Ref.	Ref.		Ref.	Ref.
LGE on CMR		4.55	0.38		−0.69	0.60
Presumed CS diagnostic criteriaD (%)
Patchy uptake on dedicated cardiac PET+VT	0.17	Ref.	Ref.	0.06	Ref.	Ref.
Patchy uptake on dedicated cardiac PET+AVB		−0.02	0.99		−0.22	0.84
LGE on CMRC+VT		−11.81	0.16		−0.51	0.71
LGE on CMRC+AVB		−7.31	0.46		0.18	0.91
Other		9.18	0.49		16.75	0.99
Presumed CS patients with negative genetic cardiomyopathy testing (%)	0.04	5.14	0.34		0.15	0.86
Cardiac implantable electronic device (%)	<0.01	0.36	0.89	<0.01	−0.41	0.42
Implantable cardiac resynchronization therapy defibrillator	<0.01	1.33	0.71	0.02	0.56	0.42
Implantable cardioverter defibrillator		Ref.	Ref.		Ref.	Ref.
Permanent pacemaker		0.67	0.77		0.69	0.53
Arrhythmia history (%)
Right bundle branch block	0.05	−5.53	0.07	0.02	−0.71	0.24
Left bundle branch block	0.02	3.49	0.24	0.02	0.80	0.16
First-degree AVB	0.06	−6.20	0.05	<0.01	−0.44	0.48
Second-degree AVB Type I	<0.01	−5.52	0.47	0.03	−16.38	0.99
Second-degree AVB Type II	<0.01	1.48	0.79	<0.01	−0.91	0.44
Third-degree AVB	0.02	3.44	0.21	0.02	0.62	0.23
Premature ventricular contraction burden >10%	<0.01	−1.58	0.55	<0.01	0.30	0.54
Non-sustained VT	<0.01	−1.09	0.67	0.02	−0.65	0.19
Sustained VT/VF	0.01	−2.42	0.35	0.02	−0.61	0.21
Current or former smoker	0.05	5.04	0.07	0.04	1.05	0.05
Body mass index	0.01	−0.12	0.47	0.01	−0.03	0.38
CKD Stage III or greater	0.04	−6.11	0.11	<0.01	−0.52	0.49
CAD, moderate or greater	<0.01	0.61	0.90	<0.01	−0.21	0.821
Atrial fibrillation	0.01	2.40	0.40	0.03	0.92	0.09
Diabetes	0.03	5.16	0.14	0.05	1.45	0.05*
LVEF at index evaluation (n=68)	0.01	−0.17	0.33	0.11	0.13	0.005
LVEDD at index evaluation (n=68)	0.02	−0.18	0.24	0.05	−0.06	0.07
NYHA class at index evaluation
I	0.02	Ref.	Ref.	0.05	Ref.	Ref.
II		−0.59	0.86		0.23	0.72
III		2.39	0.52		1.01	0.16
IV		7.60	0.49		16.26	0.99
eGFR by creatinine (mL/min/1.73 m2; n=51)	<0.01	0.05	0.5	<0.01	<0.01	0.74
Potassium (mEq/L; n=59)	<0.01	0.14	0.59	<0.01	−0.05	0.55
NT-proBNP (pg/mL; n=36)	<0.01	<0.01	0.92	<0.01	<−0.01	0.74
hs-cTn (ng/L; n=17)	0.03	−0.17	0.38	0.04	0.05	0.20
Kansas City Medical Optimization score
Index evaluation	0.03	−0.08	0.14	0.02	−0.01	0.25
Follow-up	0.03	−0.05	0.17	<0.01	<−0.01	0.61
ACEi/ARB/ARNI
Index evaluation
Active use	<0.01	−0.05	0.99	<0.01	−0.39	0.44
% of target dose	0.11	−5.51	0.11	0.04	−1.04	0.38
Follow-up
Active use	<0.01	−0.17	0.97	0.04	−1.50	0.08
% of target dose	<0.01	−1.01	0.54	<0.01	−0.15	0.66
ACEi
Index evaluation
Active use	<0.01	−0.87	0.75	<0.01	−0.15	0.76
% of target dose	0.11	−5.51	0.11	0.04	−1.04	0.38
Follow-up
Active use	<0.01	−0.66	0.81	<0.01	−0.19	0.71
Mean % of target dose	0.04	−0.90	0.34	0.02	−0.32	0.52
ARB
Index evaluation
Active use	<0.01	0.17	0.97	<0.01	0.27	0.72
% of target dose	<0.01	−1.44	0.87	<0.01	<0.01	1.0
Follow-up
Active use	0.03	−5.57	0.19	<0.01	−0.75	0.39
% of target dose	0.09	4.10	0.52	0.03	0.75	0.60
ARNI
Index evaluation
Active use	<0.01	2.45	0.49	<0.01	−0.38	0.57
% of target dose	0.07	9.52	0.44	<0.01	0.52	0.81
Follow-up
Active use	0.01	2.35	0.36	<0.01	−0.30	0.54
% of target dose	0.02	3.96	0.51	<0.01	0.40	0.73
β-blockers
Index evaluation
Active use	<0.01	−0.63	0.85	<0.01	0.12	0.85
% of target dose	<0.01	−1.19	0.75	0.03	−1.132	0.16
Follow-up
Active use	0.02	9.12	0.23	<0.01	0.24	0.87
% of target dose	0.05	−4.50	0.08	0.02	−0.63	0.25
MRA
Index evaluation
Active use	<0.01	−0.78	0.78	<0.01	0.05	0.93
% of target dose	<0.01	−2.62	0.85	<0.01	−0.94	0.68
Follow-up
Active use	<0.01	1.04	0.69	<0.01	−0.06	0.91
% of target dose	0.03	−6.59	0.40	<0.01	−0.48	0.76
SGLT2i
Index evaluation
Active use	<0.01	−0.27	0.96	<0.01	−0.18	0.85
Follow-up
Active use	0.01	2.88	0.67	0.01	0.66	0.28
Loop diuretic
Index evaluation
Active use	0.03	−3.99	0.13	<0.01	−0.48	0.34
Daily dose (furosemide equivalents)	0.26	−0.20	0.007*	0.08	−0.04	0.16
Follow-up
Active use	<0.01	−2.10	0.43	<0.01	−0.48	0.34
Daily dose (furosemide equivalents)	0.22	−0.14	0.01*	0.06	<0.01	0.25
Digoxin
Index evaluation
Active use	<0.01	4.69	0.46	<0.01	0.97	0.44
Daily dose (excluding patients not taking the medication)	<0.01	0.04	0.47	<0.01	<0.01	0.44
Follow-up
Active use	0.04	12.52	0.10	0.04	16.88	0.99
Daily dose (excluding patients not taking the medication)	0.04	0.10	0.10	0.04	0.14	0.99
Immunosuppressive treatment
Any immunosuppressive treatment during follow-up	0.03	5.47	0.15	0.02	1.15	0.17
FDG-PET measurements
Index evaluation, n=32
Mean no. abnormal perfusion segments	0.02	−0.42	0.41	0.05	−0.15	0.15
Mean no. FDG avid segments	<0.01	0.13	0.78	0.02	0.08	0.38
Mean SUVmax of myocardium	0.03	4.34	0.30	0.06	1.46	0.12
Mean SUVmax of myocardium/SUVmax of blood pool ratio	<0.01	−0.20	0.86	<0.01	−0.13	0.56
Follow-up
Mean no. abnormal perfusion segments	0.09	−0.87	0.01	0.21	−0.41	0.01*
Mean no. FDG avid segments	0.02	0.39	0.45	0.01	0.08	0.43
Mean SUVmax of myocardium	<0.01	−0.53	0.79	<0.01	0.10	0.80
Mean SUVmax of myocardium/SUVmax of blood pool ratio	0.05	3.70	0.22	0.09	1.36	0.10
Mean change in no. abnormal perfusion segments	<0.01	−0.32	0.60	0.10	−0.29	0.048*
Mean change in no. FDG avid segments	0.01	0.22	0.56	<0.01	<0.01	0.98
Mean change in SUVmax of myocardium/SUVmax of blood pool ratio	<0.01	0.49	0.67	<0.01	0.14	0.57

*P values meeting the threshold of statistical significance. ^AMcFadden’s R². ^BDenotes patients who did not meet any other probable cardiac sarcoidosis (CS) diagnostic criteria, because all patients in this cohort had reduced LVEF at CS index evaluation. ^CConsistent with CS. ^DSee Supplementary Table. AVB, atrioventricular block; HFimpEF, heart failure with improved ejection fraction; PET, positron emission tomography; VF, ventricular fibrillation. Other abbreviations as in Tables 1,2.

Analysis of the first primary remodeling outcome, the change in LVEF, revealed a positive correlation with KCMO score at follow-up (Wilcoxon W=2,332, P<0.001, r_rb=0.988), but no association with the number of GDMT agents at follow-up (P=0.34 by Kruskal-Wallis), immunosuppressive treatment, or FDG-PET measurements.

The second primary remodeling outcome, HFimpEF at follow-up, revealed no significant association with number of GDMT agents at follow-up (P=0.19) or KCMO (by binomial logistic regression; Table 3). However, significant associations were observed with LVEF at index evaluation (R²=0.11, β= 0.13, P=0.005), abnormal perfusion segments at index evaluation (R²=0.21, β=−0.41, P=0.01), and change in abnormal perfusion segments (R²=0.10, β=−0.29, P=0.048).

In a subgroup analysis of only patients receiving immunosuppressive treatment (n=67), mean LVEF improved from 30.1% to 41.2% (P<0.001). The mean change in LVEF was +9.6%, and 27 patients (47%) achieved HFimpEF.

Immunosuppressive Treatment During the Study Period

Most patients received immunosuppressive treatment (85%; n=67) for median 12 months (mean 80% of follow-up time; 25 patients <100% of follow-up time), usually with prednisone (79%; mean dose 16 mg daily) and fewer with mycophenolate (35%; mean dose 1,604 mg daily), methotrexate (27%; mean dose 17 mg weekly), or a tumor necrosis factor-α inhibitor (7%). Of the 67 patients receiving immunosuppressive treatment, 56 (84%) were treated with a combination of prednisone and a steroid-sparing agent. Of the 79 patients in this study, 67% (n=53; 78% of patients treated with prednisone) were on TMP-SMX prophylaxis for Pneumocystis pneumonia. The most common reason for a lack of TMP-SMX prophylaxis was sulfa medication allergy (documented in 5 patients) and alternative medical prophylaxis was favored by some clinicians to avoid hyperkalemia.

Clinical Outcomes

At follow-up, the primary composite outcome occurred in 13 (16%) patients (Figure 3A) and differed significantly by number of GDMT agents at follow-up (univariable log-rank P<0.001; Figure 3B), with the highest risk of events seen in patients on 0 GDMT agents (100%; n=7) and 1 GDMT agent (28%; n=2). All patients on 0 GDMT agents at follow-up met the primary composite outcome by LVAD or heart transplant occurrence. A higher KCMO score was also associated with a decreased risk of the primary composite outcome (HR 0.96; 95% CI 0.94–0.98; log-rank P<0.001). Univariable log-rank analysis of other cohort variables revealed that other significant predictors of the primary composite outcome were CS diagnostic category (definite CS vs. probable, log-rank P=0.003; definite CS vs. presumed CS, P<0.001), LVEF at index evaluation (log-rank P<0.001), NYHA class at index evaluation (log-rank P<0.001), and NT-proBNP (log-rank P=0.02). Notably, there was no difference in the primary composite outcome stratified by immunosuppressive treatment during follow-up (log-rank P=0.36).

Figure 3.

Kaplan-Meier event-free survival curves illustrating (A) primary composite outcome event-free survival (unplanned heart failure hospitalization, left ventricular assist device implantation, heart transplant, or death), with (B) univariable analysis by the number of guideline-directed medical therapy (GDMT) agents at follow-up.

In a subgroup analysis of only patients on immunosuppressive treatment (n=67), the primary composite outcome occurred in 10 (15%) patients and differed significantly by the number of GDMT agents at follow-up, with the highest risk seen in patients on 0 GDMT agents at follow-up (log-rank P<0.001).

Regarding secondary outcomes, documented hyperkalemia was infrequent both in the whole cohort (18%; n=14) and in subgroup analysis of patients treated with TMP-SMX (14%; n=5). No patients died during the follow-up period. At 12 months, Kaplan-Meier estimated event-free survival to first sustained VT or VF was 87% (Supplementary Figure A). No significant relationship between event-free survival to first sustained VT or VF and the number of GDMT agents at follow-up was found (log-rank P=0.26; Supplementary Figure B). The mean number of events was 5.6±22.0, and the median was 0 events. In patients with arrhythmia events (n=18), the median number of events was 11.5 (IQR 23.3 events).

Laboratory Parameters

At index evaluation, the cohort exhibited borderline normal renal function, with mean values of eGFR by creatinine of 61.4 mL/min/1.73 m² (n=51), potassium 5.0 mEq/L (n=59), and abnormal NT-proBNP 1,196 pg/mL (n=36) and high-sensitivity troponin 17.4 (n=17). At follow-up, only potassium changed significantly to 4.5 mEq/L (P=0.03). Subgroup analysis of patients receiving TMP-SMX revealed no change from baseline to follow-up in potassium (from 5.3 to 4.5 mEq/L; P=0.31; n=40) or eGFR (from 60 to 62 mL/min/1.73 m²; P=0.93; n=35).

Discussion

Cardiomyopathy is a common presentation of CS and is typically managed medically with GDMT in the same manner as HFrEF broadly, with the addition of other CS-directed therapies, such as immunosuppressive treatment. However, data on the response to GDMT in CS are sparse and much needed to validate this approach. Here we present the largest retrospective cohort study, to the best of our knowledge, describing the response to GDMT in CS patients with HFrEF on presentation.

New HFrEF at CS diagnosis is a recognized herald of poorer disease prognosis and survival. Several prior studies have described the natural history and response to treatment in CS patients with HFrEF. In a 2017 retrospective study describing the clinical characteristics and outcomes of 73 patients diagnosed with CS by World Association of Sarcoidosis and Other Granulomatous Diseases criteria, 30 patients had HFrEF and most experienced improvement in LVEF after immunosuppressive treatment, but GDMT specifics were not elaborated.¹⁸ A 2018 case series detailed 91 CS patients diagnosed by 2014 HRS expert consensus criteria, of whom 47 (52%) presented with cardiomyopathy and most experienced improvement in LVEF ≥10% during follow-up.³ Most of these patients (91%) were treated with at least 1 GDMT agent during the study, but the study focused on evaluating response to immunosuppressive treatment, so GDMT agents and dosing were outside the study scope.³ Our study offers novel data in this field due to its large cohort size of CS patients with HFrEF, granular GDMT data, and uniformly available follow-up echocardiography, offering insights into the natural history of HFrEF in the first 2 years of treatment. Furthermore, our study incorporates additional collateral information through analysis of FDG-PET to correlate changes in perfusion and inflammation with cardiac remodeling and overall clinical response.

Although GDMT uptitration improved during the follow-up period, implementation of maximal GDMT with all 4 agent classes was rare in our cohort, predominantly due to the recency of SGLT2i approval. ACE/ARB/ARNI or β-blockers were the most common GDMT in use at baseline, and the mean number of GDMT agents increased significantly over the study period, although only reached a mean 2.2 agents (KCMO score 70) at follow-up. In some cases, medication cost may be prohibitive, particularly in the case of ARNI and SGLT2i therapy. Other possibilities for limited uptitration may include per-patient contraindications (although the mean eGFR noted in this cohort suggests that renal dysfunction would not broadly preclude GDMT initiation), limitations in frequent visits for uptitration due to our tertiary referral center practice, or Stage D HF, in which GDMT initiation is more controversial.¹⁹

Significant improvement in HFrEF was seen at follow-up, with mean LVEF rising to 39.9%, and GDMT at follow-up (quantified by KCMO score) was strongly correlated with the quantitative change in LVEF. Although we could not demonstrate this same statistical correlation with the number of GDMT agents at follow-up, the KCMO score is a validated score that accounts for the eligibility of individual patients for GDMT classes and contemporary guidelines, and thus better illustrates this relationship. This is a key clinical finding because CS-specific data demonstrating the efficacy of GDMT are sparse.

Our remodeling outcome analyses revealed that baseline poor prognostic factors for improvement to HFimpEF are more severe systolic dysfunction and abnormal perfusion myocardial segments. At follow-up, interval decrease in abnormal perfusion segments is associated with achievement of HFimpEF, highlighting the remodeling benefits that can be achieved by appropriate CS-directed therapy and improved myocardial perfusion.

We demonstrated that CS patients presenting with HFrEF represent a population at high risk of poor clinical outcomes. Nearly one-quarter of our cohort experienced unplanned cardiovascular hospitalization during the follow-up period, which, given the deliberately chosen 24-month maximum follow-up period, is notably high. Over 10% of patients underwent LVAD implantation or heart transplant within 24 months of index evaluation, further illustrating the high-risk nature of this cohort. Interestingly, no patients died during follow-up, although three-quarters of the cohort had CIED in place by time of follow-up.

Lastly, given the unique component of immunosuppressive treatment and necessity for Pneumocystis pneumonia prophylaxis in this patient population, there is a theoretical concern for hyperkalemia, given that ACEi/ARB/ARNI, MRA, and TMP-SMX all can increase serum potassium by independent and potentially additive mechanisms.²⁰ Only 67% of patients in our cohort were treated with TMP-SMX, but subgroup analysis of these patients revealed a low rate of hyperkalemia, similar to the cohort broadly, and no significant change in serum potassium or eGFR. Thus, our data suggest that with appropriate laboratory monitoring, optimization of GDMT should be attempted even in patients on TMP-SMX because the risk of hyperkalemia or worsening estimated renal dysfunction is low.

Our study is limited by its retrospective nature, small sample size, and relatively short follow-up duration, although the 24-month follow-up cut-off was deliberately chosen to evaluate short-term responses to GDMT after CS index evaluation. The predominantly White racial composition of our cohort and our tertiary care referral-based setting may restrict the external generalizability of our findings. As discussed above, a deficit of patients on all 4 GDMT agents may limit our ability to fully elucidate clinical and echocardiographic responses to GDMT. Furthermore, patients with no GDMT agents often had Stage D HF, limiting the validity of the conclusions regarding the composite outcome in this population. Our study also had too few events of the primary composite outcome relative to predictor characteristics to perform a comprehensive multivariable survival analysis. Most of the patients in our cohort were treated with immunosuppressive treatment, and although we attempted to control for immunosuppressive treatment in multivariable models, our cohort was not sufficiently powered to detect differences in outcomes by immunosuppressive treatment status, because this was outside the scope of our study.

We have demonstrated that cardiac remodeling and LVEF improvement are overall favorable in CS patients presenting with new HFrEF in short-term interval follow-up. More optimal GDMT at follow-up is positively associated with LVEF improvement in CS patients. Patients treated with more GDMT agents experienced improved clinical outcomes consistent with the general HF population, further emphasizing the importance of optimizing GDMT in CS patients with HFrEF.

Sources of Funding

Our institutional database is collected and managed in REDCap software, which is supported by a grant from the Center for Clinical and Translational Science (UL1TR002377).

Disclosures

Our institutional database is collected and managed in REDCap software, which is supported by a grant from the Center for Clinical and Translational Science (UL1TR002377). The authors declare no pertinent conflicts of interest.

IRB Information

The protocol for this study was approved by the Mayo Clinic Institutional Review Board (IRB ID: 17-000976).

Supplementary Files

Please find supplementary file(s);

https://doi.org/10.1253/circj.CJ-24-0205

References

• 1.   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, doi:10.1093/oxfordjournals.aje.a009096.
• 2.   Arkema EV, Grunewald J, Kullberg S, Eklund A, Askling J. Sarcoidosis incidence and prevalence: A nationwide register-based assessment in Sweden. Eur Respir J 2016; 48: 1690–1699, doi:10.1183/13993003.00477-2016.
• 3.   Fussner LA, Karlstedt E, Hodge DO, Fine NM, Kalra S, Carmona EM, et al. Management and outcomes of cardiac sarcoidosis: A 20-year experience in two tertiary care centres. Eur J Heart Fail 2018; 20: 1713–1720, doi:10.1002/ejhf.1319.
• 4.   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.
• 5.   Yafasova A, Fosbøl EL, Schou M, Gustafsson F, Rossing K, Bundgaard H, et al. Long-term adverse cardiac outcomes in patients with sarcoidosis. J Am Coll Cardiol 2020; 76: 767–777, doi:10.1016/j.jacc.2020.06.038.
• 6.   Terasaki F, Azuma A, Anzai T, Ishizaka N, Ishida Y, Isobe M, et al. JCS 2016 guideline on diagnosis and treatment of cardiac sarcoidosis: Digest version. Circ J 2019; 83: 2329–2388, doi:10.1253/circj.CJ-19-0508.
• 7.   Heidenreich PA, Bozkurt B, Aguilar D, Allen LA, Byun JJ, Colvin MM, et al. 2022 AHA/ACC/HFSA guideline for the management of heart failure: A report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2022; 145: e895–e1032, doi:10.1161/CIR.0000000000001063.
• 8.   Birnie DH, Sauer WH, Bogun F, Cooper JM, Culver DA, Duvernoy CS, et al. HRS expert consensus statement on the diagnosis and management of arrhythmias associated with cardiac sarcoidosis. Hear Rhythm 2014; 11: 1304–1323, doi:10.1016/j.hrthm.2014.03.043.
• 9.   Rosenbaum AN, Kolluri N, Elwazir MY, Kapa S, Abou Ezzeddine OF, Bois JP, et al. Identification of a novel presumed cardiac sarcoidosis category for patients at high risk of disease. Int J Cardiol 2021; 335: 66–72, doi:10.1016/j.ijcard.2021.04.022.
• 10.   Wen A, Fu S, Moon S, El Wazir M, Rosenbaum A, Kaggal VC, et al. Desiderata for delivering NLP to accelerate healthcare AI advancement and a Mayo Clinic NLP-as-a-service implementation. NPJ Digit Med 2019; 2: 130, doi:10.1038/s41746-019-0208-8.
• 11.   Khan MS, Chan PS, Sherrod CF, Ikemura N, Sauer AJ, Jones PG, et al. Generalizable approach to quantifying guideline-directed medical therapy. Circ Heart Fail 2024; 17: e011164, doi:10.1161/CIRCHEARTFAILURE.123.011164.
• 12.   Chareonthaitawee P, Beanlands RS, Chen W, Dorbala S, Miller EJ, Murthy VL, et al. Joint SNMMI–ASNC expert consensus document on the role of ¹⁸F-FDG PET/CT in cardiac sarcoid detection and therapy monitoring. J Nucl Med 2017; 58: 1341–1353, doi:10.2967/jnumed.117.196287.
• 13.   Christopoulos G, Jouni H, Acharya GA, Blauwet LA, Kapa S, Bois J, et al. Suppressing physiologic 18-fluorodeoxyglucose uptake in patients undergoing positron emission tomography for cardiac sarcoidosis: The effect of a structured patient preparation protocol. J Nucl Cardiol 2021; 28: 661–671, doi:10.1007/s12350-019-01746-4.
• 14.   Sykora D, Young KA, Elwazir MY, Bois JP, Arment CA, Chareonthaitawee P, et al. The mechanism and natural history of mitral regurgitation in cardiac sarcoidosis. Am J Cardiol 2023; 191: 84–91, doi:10.1016/j.amjcard.2023.01.002.
• 15.   Harris PA, Taylor R, Thielke R, Payne J, Gonzalez N, Conde JG. Research electronic data capture (REDCap)--a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009; 42: 377–381, doi:10.1016/j.jbi.2008.08.010.
• 16.   Harris PA, Taylor R, Minor BL, Elliott V, Fernandez M, O’Neal L, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inform 2019; 95: 103208, doi:10.1016/j.jbi.2019.103208.
• 17.   Hunt SA, Abraham WT, Chin MH, Feldman AM, Francis GS, Ganiats TG, et al. 2009 Focused update incorporated into the ACC/AHA 2005 Guidelines for the diagnosis and management of heart failure in adults: A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines developed in collaboration with the International Society for Heart and Lung Transplantation. J Am Coll Cardiol 2009; 53: e1–e90, doi:10.1016/j.jacc.2008.11.013.
• 18.   Zhou Y, Lower EE, Li HP, Costea A, Attari M, Baughman RP. Cardiac sarcoidosis: The impact of age and implanted devices on survival. Chest 2017; 151: 139–148, doi:10.1016/j.chest.2016.08.1457.
• 19.   Samarendra P. GDMT for heart failure and the clinician’s conundrum. Clin Cardiol 2019; 42: 1155–1161, doi:10.1002/clc.23268.
• 20.   Velázquez H, Perazella MA, Wright FS, Ellison DH. Renal mchanism of timethoprim-induced hperkalemia. Ann Intern Med 1993; 119: 296–301, doi:10.7326/0003-4819-119-4-199308150-00008.