Investigating the treatment phenotypes of cardiac sarcoidosis: A prospective cohort study

Abstract

Introduction

Data indicates there are 4 main pulmonary sarcoidosis duration/treatment phenotypes: asymptomatic, acute (disease duration <1–2 years), chronic and advanced. There are no data about disease duration/treatment phenotypes of cardiac sarcoidosis patients. Our study had 2 main aims (i) to assess the response to corticosteroids and (ii) to assess the incidence of relapse after a one-year course of corticosteroids (thereby classifying patients as acute or chronic treatment phenotype).

Methods

Consecutive, treatment naive patients with CS were prospectively recruited and treated with 0.5 mg/kg prednisone, to a maximum dose of 40 mg/day. Patients had a follow-up PET after 3–6 months of therapy (PET 2). In the responders (PET definition of response) the prednisone was then weaned and stopped after 12 months. Three months after stopping, the PET was repeated to look for disease relapse (PET 3).

Results

Twenty-one consecutive patients were included, and all patients showed a reduction in cardiac FDG uptake after 3–6 months and 19/21 (90.5 %) met the PET definition of response. Of these, 12/19 (63.1 %) relapsed after prednisone was stopped. There were no serious adverse effects during the trial of therapy cessation and there were no later relapses in the 7 non-relapsers during over 4 years of subsequent follow-up.

Conclusion

The initial response rate to prednisone was high with all patients showing a reduction in FDG uptake and 19/21 meeting a PET definition of >25 % response. Secondly, a trial of therapy discontinuation was able to classify 7/19 patients as acute treatment phenotype and 12/19 as chronic.

1. Introduction

Sarcoidosis is a multisystem, inflammatory granulomatous disease of unknown etiology. It is characterized by systemic infiltration with non-caseating granulomas occurring in any organ in the body including the heart, liver, spleen, eyes, and kidneys [1]. Autopsy analysis and cardiovascular magnetic resonance (CMR) studies have demonstrated a 20–47 % rate of clinically silent cardiac involvement of patients with sarcoidosis [2], [3], [4]; while 3–5 % showed clinically manifest involvement [1].

The treatment of cardiac sarcoidosis (CS) remains a challenge as there are few data to guide physicians. A recent systematic review of corticosteroids in CS identified 34 papers reporting outcomes. The best data related to atrioventricular (AV) conduction recovery and corticosteroids appeared to be beneficial [2]. Despite this scarcity of data, most expert opinion and guidelines have been proponents of treating CS [3], [4]. For example the 2021 European Respiratory Society clinical practice guidelines on treatment of sarcoidosis recommended the use of glucocorticoids (with or without other IST) in patients with CS and evidence of functional cardiac abnormalities, including heart block, dysrhythmias, or cardiomyopathy (strong recommendation, very low quality of evidence) [3]. However many unknowns remain including whether all CS patients respond to corticosteroids, and what the optimal dose is. Also there is much interest in trying to understand the different phenotypes of sarcoidosis, and a variety of phenotypic categorizations have been developed [5], [6]. There are considerable data on the phenotypes of pulmonary sarcoidosis suggesting there are 4 main disease duration/treatment phenotypes: asymptomatic, acute (disease duration <1–2 years), chronic and advanced [6]. There are no data about disease duration/treatment phenotypes of patients with CS.

Our study had 2 main aims (i) to systematically and prospectively assess the response to corticosteroids and (ii) to assess the incidence of relapse after a one-year course of corticosteroids (thereby classifying patients as acute or chronic phenotype). We also examined potential predictors of relapse and tracked adverse events during the trial of therapy cessation and whether initial non-relapsers subsequently relapsed during long-term follow-up.

2. Methods

2.1. Study population

This study is a single centre sub-study of the Cardiac Sarcoidosis Multi-Center Prospective Cohort Study (CHASM-CS; NCT01477359). All patients in CHASM-CS at the University of Ottawa Heart Institute were included in this study. The protocol was approved by the local institutional ethics committee and all patients signed informed consent. Consecutive patients who met all the following criteria were included in the sub-study:

• (i)
  Presentation with clinically manifest CS defined as one or more of the following clinical features:

  • -
    advanced conduction system disease (sustained Mobitz II AV block or third-degree AV block)
  • -
    sustained VT of unknown etiology
  • -
    ventricular dysfunction (left ventricular ejection fraction (LVEF) < 50 % and/or right ventricular ejection fraction (RVEF) < 40 %)
• (ii)CS diagnosed based on the Heart Rhythm Society (HRS) criteria [11]
• (iii)Abnormal myocardial ¹⁸F-FDG uptake on PET scan in a pattern consistent with active CS [7]
• (iv)Plan to initiate treatment with systemic corticosteroids
• (v)No previous treatment for sarcoidosis in any organ.

2.2. Therapy and follow-up

All patients were treated with 0.5 mg/kg prednisone up to a maximum dose of 40 mg once daily. All patients had a follow up ¹⁸F-FDG-PET scan after 3–6 months of therapy (PET 2). Patients were then classified as cardiac responders or non-responders:

• Cardiac responder: ≥25 % reduction in LV SUV max [8], [9] between PET 1 and PET 2 AND clinical report of major improvement in overall cardiac ¹⁸F-FDG uptake.

• Cardiac non-responder: <25 % reduction in LV SUV max [8], [9] between PET 1 and PET AND clinical report of less than major improvement in overall cardiac ¹⁸F-FDG uptake.

In the responders the prednisone was then weaned over 6–9 months to complete 12 months of therapy and stopped. Three months after stopping, the ¹⁸F-FDG-PET was repeated (PET 3) to look for disease relapse and patients were classified as relapsers or non-relapsers defined as follows:

• Cardiac relapser: ≥25 % increase in LV SUV max between PET 2 and PET 3 AND clinical report of major increase in overall cardiac FDG uptake

• Cardiac non-relapser: <25 % increase in LV SUV max between PET 2 and PET 3 AND clinical report of no major increase in overall cardiac FDG uptake.

Fig. 1 shows details of the study flow.

Fig. 1.

Progress of study participants.

2.3. ¹⁸F-FDG PET scanning

Full details on the imaging protocols and image interpretation methods have been published [12]. In brief, the imaging protocol included a whole-body acquisition (from the vertex to mid-thighs), approximately 60 min after the intravenous injection of 5 MBq/kg of ¹⁸F-FDG and followed immediately by a dedicated cardiac acquisition with ECG gating for evaluation of LV function. To achieve adequate suppression of physiological myocardial ¹⁸F-FDG uptake, all patients were instructed to follow a low-carbohydrate, fat-rich, protein-permitted diet the day before the examination, followed by a fast of at least 12 h immediately before the examination. Unless contraindicated, patients also received a low dose (15 IU/kg) of intravenous unfractionated heparin before ¹⁸F-FDG injection. All patients also underwent a rest myocardial perfusion imaging study using ECG-gated PET-CT with either ⁸²Rb or ¹³N-ammonia on the same day.

2.4. PET scan analysis

All images were interpreted by an experienced nuclear cardiologist/nuclear medicine specialist with significant experience in reporting CS scans. Readers were blinded to clinical data and to whether patients were on corticosteroids or not. Sites of disease involvement were defined as “active” when abnormal FDG uptake in a pattern consistent with sarcoidosis was present [7]. The number of discrete areas with increased FDG uptake was counted. Maximum standardized uptake value (SUV_max) of the LV was measured on axial images, while the mean SUV of the LV (LV mean SUV) and its coefficient of variation (COV) were calculated using FlowQuant automated software (Ottawa, ON, Canada). Right ventricular ¹⁸F-FDG uptake was recorded as present or absent. Individual organ extra-cardiac ¹⁸F-FDG uptake was recorded as present or absent. Resting perfusion defects were scored on a 4-point scale and the summed rest score (SRS) was calculated using the 17-segment model as described previously [10].

2.5. Additional definitions

• Extra-cardiac responder: no abnormal ¹⁸F-FDG uptake in any organ

• Extra-cardiac relapse: abnormal ¹⁸F-FDG uptake in ≥one organ (in patients who were prior extra-cardiac responders).

• PET 1: Pre prednisone PET (PET 1)

• PET 2: PET done after 3–6 months of prednisone 0.5 mg/kg prednisone up to a maximum dose of 40 mg once daily PET (dose continued to time of PET 2).

• PET3: PET done 3 months after cessation of prednisone

• Serious cardiac events: one or more of cardiac death, heart failure hospitalization, episode of sustained ventricular arrythmia.

2.6. Statistics

Categorical variables are presented using percentages or frequencies, and continuous variables using means (±standard deviation) or medians (25th, 75th percentiles), when appropriate. We compared categorical variables using the chi-square test (or Fisher's exact test when appropriate), and continuous variables using one-way analysis of variance or Kruskal-Wallis test for normally and non-normally distributed variables, respectively. Statistical significance was defined as p < 0.05.

3. Results

3.1. Study population

Twenty-one patients were included (see Table 1). The mean age was 55.9 ± 8.5 years-old, 19/21 (90.5 %), were white, 10/21 (47.6 %) were female and 17/21 (81.0 %) presented with AV block.

Table 1.

Baseline demographics of the cohort included in the study.

Characteristics	Total (n = 21)
Age at presentation of CS (years)	55.9 ± 8.5
Race (Caucasian)	19 (90.5 %)
Sex (female)	10 (47.6 %)
Days between presentation and starting prednisone	98.3 ± 121.0
Diabetes	3 (14.3 %)
Presenting cardiac feature
VT or cardiac arrest	5 (23.8 %)
Sustained Mobitz II and/or third-degree AV block	17 (81.0 %)
Both	2 (9.5 %)
Other	1 (4.8 %)

Values are n/n, mean ± SD or n (%).

CS – cardiac sarcoidosis; VT – ventricular tachycardia; AV - atrioventricular.

3.2. Assessment of cardiac response after 3–6 months of prednisone, comparing PET 1 and PET 2

The mean time between starting prednisone and first follow-up PET (PET 2) was 136.0 ± 107.3 days. All patients had a reduction in cardiac FDG uptake, and 19/21 (90.5 %) patients were classified as responders. In one of the non-responders, LV SUV_max was 6.38 pre-prednisone (PET 1) and 5.47 post (PET 2) (14.26 % reduction) and in the other non-responder, LV SUV_max was 7.41 pre-prednisone (PET 1) and 5.83 post (PET 2) (23.33 % reduction). Mean reduction in LV SUV_max in the responders was 6.14 ± 4.31, with mean percentage reduction of 59.4 ± 15.4 %. Additional details of the PET variables pre- and post-prednisone in the 19 responders are shown in Table 2. Changes in LV SUV_max are shown in Fig. 2.

Table 2.

Serial fluorodeoxyglucose-positron emission tomography results in 19 cardiac responder patients.

	Pre-prednisone PET 1	During prednisone PET 2	Comparing PET 1 with 2 (p value)	Post-prednisone PET 3	Comparing PET 2 with 3 (p value)
Cardiac data (n = 19)
Discrete LV areas with increased FDG uptake	4.89 ± 3.09	1.11 ± 1.85	<0.001	3.21 ± 3.03	0.016
LV SUVmax	9.47 ± 5.03	3.33 ± 1.24	<0.001	6.54 3.88	0.001
Mean LV SUV	2.81 ± 2.04	1.59 ± 0.40	0.015	2.16 0.73	0.017
RV FDG uptake	7/18	0/18	0.016	3/18	0.25
SPRS	2.42 ± 3.63	2.68 ± 3.70	0.80	2.95 ± 3.55	0.53
EF	47.6 ± 14.8	47.2 ± 10.1	0.49	48.4 ± 14.4	0.11
Whole body FDG uptake (n = 18a)
Lungs	9/18	0/18	0.004	6/18	0.031
Spleen	4/18	0/18	0.13	1/18	1.00
Liver	2/18	0/18	0.50	2/18	0.50
Bone	2/18	0/18	0.50	3/18	0.25
Brain	0/18	0/18		0/18
Skin/subcutaneous	1/18	0/18	1.00	0/18
Lymph nodes
Thoracic	16/18	2/18	<0.001	18/18	<0.001
Abdominal	11/18	1/18	0.002	13/18	<0.001
Neck	8/18	0/18	0.008	7/18	0.016
Any	17/18	3/18	<0.001	18/18	<0.001
Mean number of extra-cardiac organ involvementb	1.94 ± 0.87	1.00 ± 0	0.10	1.74 ± 1.00	0.18

Values are n/n, mean ± SD or n (%).

FDG-PET - fluorodeoxyglucose-positron emission tomography; SUV - standardized uptake value; SPRS - summed perfusion rest score; CS - cardiac sarcoidosis.

^aOne patient did not have whole body scan.

^bAmongst patients with at least one organ involvement, considering any lymph node involvement as one organ.

Fig. 2.

LV SUV_max is shown for non-relapsers (n = 7) and relapsers (n = 12), at time of 3 PET scans. Individual patient data, in dashed lines, as well as mean ± SD, in solid black, are shown. The first PET was performed before treatment, the second was performed 3–6 months after prednisone initiation, and the third performed after prednisone discontinuation.

3.3. Assessment of cardiac relapse in initial responders, after stopping prednisone, comparing PET 2 and PET 3

The mean total duration of corticosteroids in the 19 responders was 389.7 ± 106.0 days. The post-steroid cessation PET (PET 3) was performed a mean of 134.3 ± 75.4 days after prednisone was stopped. This PET showed that 12/19 (63.1 %) of patients had relapsed after prednisone was stopped. The mean increase in LV SUV_max in the relapsers was 5.06 ± 3.05, p < 0.001 with mean % increase of 164.4 ± 132.3 % (comparing PET 2 with PET3) (see Fig. 2). Additional details of the PET variables comparing PET 2 and PET 3 are shown in Table 2. Fig. 2 illustrates the individual patient changes in LV SUV_max between all 3 PET studies.

3.4. Assessment of extra-cardiac response after 3–6 months of prednisone, comparing PET 1 and PET 2

All patients (20, as one patient did not have whole body scans) had reduction on extra-cardiac ¹⁸F-FDG uptake and 17/20 (85.0 %) had no extra-cardiac FDG uptake after 3–6 months of continued prednisone (see Table 3).

Table 3.

Patient characteristics stratified by relapse after prednisone withdrawal.

	Non-relapser (n = 7)	Relapser (n = 12)	p value
Age at presentation of CS (years)	56.7 ± 11.2	56.8 ± 7.0	0.99
Race (Caucasian)	7 (100)	10 (83.3)	0.51
Sex (female)	3 (42.9)	5 (41.7)	1.00
Prior history of extra-cardiac sarcoidosis	1 (14.3)	1 (8.3)	1.00
Hypertension	1 (14.3)	7 (58.3)	0.15
Diabetes	1 (14.3)	2 (16.7)	1.00
Presentation cardiac manifestations
Mobitz II and/or third-degree AV block	5 (71.4)	11 (91.7)	0.52
VT/sudden cardiac arrest	1 (14.3)	3 (25.0)	1.00
Both	1 (14.3)	2 (16.7)	1.00
Other	1 (14.3)	0	0.37
Time intervals between PETs and corticosteroids
Days between PET 1 and starting corticosteroids	11.9 ± 147.0	27.9 ± 15.5	0.78
Days between starting corticosteroids and PET 2	162.1 ± 169.5	120.7 ± 49.7	0.43
Durations of corticosteroid therapy	439.3 ± 157.1	360.8 ± 49.5	0.12
Days between stopping corticosteroids and PET 3	112.7 ± 54.6	146.8 ± 85.0	0.36

Values are n/n, mean ± SD or n (%).

PET - positron emission tomography; CS - cardiac sarcoidosis; VT – ventricular tachycardia; AV – atrioventricular.

3.5. Assessment of extra-cardiac sarcoidosis relapse after stopping prednisone, comparing PET 2 and PET 3

All patients (18, as one patient did not have whole body scans) had extra-cardiac relapse (new) or increased extra-cardiac ¹⁸F-FDG uptake on post prednisone cessation PET (PET 3).

3.6. Adverse events during period of prednisone cessation

There were no serious cardiac events during this period in either group. Also, the there was no significant change in LVEF during trial of therapy cessation (47.6 % ± 14.8 on PET 2, 48.4 % ± 14.4 on PET 3, p = 0.11).

3.7. Assessing potential predictors of relapse

Table 3 shows patient characteristics of 19 initial responders stratified by relapse/non relapse status; there were no significant differences. Table 4 shows ¹⁸F-FDG PET scan findings stratified by relapse/non relapse status and again there were no significant differences. Fig. 2 shows serial cardiac LV SUV_max in relapsers and non-relapsers. Fig. 3 shows images from representative patients.

Table 4.

¹⁸F-FDG PET scan data stratified by relapse after prednisone withdrawal.

	Non-relapser (n = 7a)	Relapser (n = 12)	p value
Cardiac PET 1
Focal/focal on diffuse LV pattern	7 (100)	12 (100)
Discrete LV areas with increased FDG uptake	5.29 ± 3.90	4.67 ± 2.67	0.69
LV SUVmax	9.51 ± 6.92	9.44 ± 3.91	0.98
LV mean SUV	3.31 ± 2.92	2.56 ± 1.54	0.48
RV FDG uptake	1 (14.3)	6 (50)	0.17
SRSS	1.14 ± 1.68	3.17 ± 4.28	0.25
LVEF	47.0 ± 18.1	47.9 ± 14.0	0.91
Whole body PET 1
Neck LN	2 (33.3)	6 (50)
Thoracic LN	5 (83.3)	11 (91.7)	1.00
Abdominal LN	3 (50)	8 (66.7)	0.63
Any LN	5 (83.3)	12 (100)	0.33
Lungs	4 (66.7)	5 (41.7)	0.62
Spleen	3 (50)	1 (8.3)	0.08
Liver	0	2 (16.7)	0.53
Bone	0	2 (16.7)	0.53
Brain	0	0
Skin/subcutaneous (PET findings)	0	1 (8.3)	1.00
Mean number of extra-cardiac organ involvement	2.00 ± 0.63	1.92 ± 1.0	0.86
Cardiac PET 2
Focal/focal on diffuse LV pattern	0	4 (33.3)	0.25
Discrete LV areas with increased FDG uptake	1.43 ± 2.30	0.92 ± 1.62	0.58
Improvement in number of discrete LV areas with increased FDG uptake	6 (85.7)	12 (100)	0.37
LV SUVmax	2.82 ± 1.0	3.63 ± 1.31	0.18
LV mean SUV	1.45 ± 0.40	1.68 ± 0.37	0.23
RV FDG uptake	0	0
Improvement in RV FDG uptake	1 (14.3)	6 (50)	0.17
SPRS	3.43 ± 4.58	2.25 ± 3.22	0.52
LVEF	44.8 ± 13.6	48.5 ± 8.1	0.49
Whole body PET 2
Neck LN	0	0
Thoracic LN	2 (33.3)	0	0.10
Abdominal LN	0	1 (8.3)	1.00
Any LN	2 (33.3)	1 (8.3)	0.25
Lungs	0	0
Spleen	0	0
Liver	0	0
Bone	0	0
Brain	0	0
Skin/subcutaneous	0	0
Reduction in number of affected organs	6 (100)	12 (100)

Values are n/n, mean ± SD or n (%).

FDG-PET - fluorodeoxyglucose-positron emission tomography; LV – left ventricle; SUV - standardized uptake value; SPRS - summed rest score; LVEF – left ventricle ejection fraction; LN – lymph node.

^aNote: one patient did not have whole body PET scan.

Fig. 3.

a. 63 years-old male, presented with syncope and A-V block. A. Cardiac and systemic sarcoidosis on PET1. B. Complete response to corticosteroids on PET 2. C. Disease relapse on PET 3.

b. 49 years-old male, presented with complete A-V block. A. Cardiac and systemic sarcoidosis on PET1. B. Complete response to corticosteroids on PET 2. C. PET 3 showing no relapse of cardiac sarcoidosis after therapy cessation.

3.8. Long term outcomes in non–relapsers

Patients have been followed for a mean of 3.7 ± 1.3 years after prednisone cessation. Details of the follow-up are show in Table 5. Five of the seven patents have had more than one further follow-up PET scan. The most recent PET was done 3.7 ± 1.3 years after prednisone cessation. In all 5 patients there was no abnormal cardiac uptake and in 2/4 there was no extra-cardiac uptake (no whole-body scan done in the fifth patient).

Table 5.

Details of clinical follow-up of non-relapsers.

Patient number	Presentation	Age at presentation	Years between stopping corticosteroids and last FU	LVEFa before starting prednisone (%)	Years between stopping and most recent LVEF	Most recent LVEFa (%)	VT	HF hospitalization	Therapy
1	VT	48.5	5.5	52	5.4	53	No	No	No
2	HB	42.9	4.4	54	4.5	54	No	No	No
3	SND	74.6	4.2	55	3.7	54	No	No	Yes, for pulmonary disease
4	HB	67.4	2.9	32	3.4	25	Yes	Yes	No
5	HB	49.5	3.3	56	3.1	48	No	No	No
6	HB	59.7	4.4	53	3.9	53	No	No	No
7	HB	54.4	1.5	55	0.3	55	No	No	No
Mean ± SD			3.7 ± 1.3	51 ± 8	3.5 ± 1.6	49 ± 11

VT – ventricular tachycardia; HB – heart block; SND – sinus node dysfunction; LVEF – left ventricle ejection fraction; HF – heart failure.

^aLVEF from echo.

There has been no change in echocardiogram LVEF during FU (51 ± 9 % pre therapy and 49 ± 11 %, p = 0.58 most recently), and the most recent echo was done a mean of 3.5 ± 1.6 years after stopping prednisone. Six of seven patients have been clinically stable. The other patient (#4) has had recurrent VT and one heart failure hospitalization. He has had two FU PETS scans and neither have shown cardiac relapse.

4. Discussion

The main findings of our study are firstly, the initial response rate to prednisone was high with all patients showing a reduction in FDG uptake and 19/21 meeting a PET definition of >25 % response. Secondly, the cardiac relapse rate after prednisone discontinuation was also high. However, a trial of therapy discontinuation was able to identify a subset of 7/19 patients who did not require chronic therapy and are classified as acute phenotype. Furthermore, there were no later relapses during 3.7 years of subsequent follow-up. Hence these 7/19 patients can be classified as having an acute phenotype of CS and the other 12 as having a chronic disease phenotype. Importantly there were no serious adverse effects during the trial of therapy cessation. Finally, there were no clinical or imaging findings that could predict which patients were going to have relapse (i.e., have chronic disease phenotype).

4.1. Response to corticosteroids in cardiac sarcoidosis

We found that all 21 patients had a reduction in FDG uptake in the LV myocardium; 19/21 (90.4 %) patients met (pre-defined) definition of PET response to prednisone. Only two other groups have looked at a PET definition of steroid responsiveness in patients with CS. Okada et al. used the same PET definition as in our study and the results are similar to ours. They found there was a significant reduction in myocardial ¹⁸F-FDG uptake in 28/32 (88 %) patients, the interval between PET scans was a median of 5 (IQR: 4 to 19) months [8]. In contrast, Shelke treated 15 Indian patients with prednisone 1 mg/kg per day for 3 months then tapered to 10 mg/day over 6 months [11]. A follow-up PET was done after 125.8 ± 54.2 days of therapy and showed complete resolution of LV ¹⁸F-FDG uptake in only 4 patients, a decrease in 5 and an increase in in FDG uptake 6 patients. Possible explanations as to why Shelke's results are different include possible racial differences in steroid responsiveness, medication non-compliance, issues with patient preparation in the second PET scan of some patients, leading to false positive FDG uptake.

4.2. Relapse rates of cardiac sarcoidosis after prednisone withdrawal

Our study's second major finding is that 12 of 19 (63.1 %) patients relapsed on a PET scan performed a mean of 140.4 ± 81.3 days after steroid discontinuation. Hence these 7/19 patients can be classified as having an acute phenotype of CS and the other 12/29 as having a chronic disease phenotype. We believe that this is the first study to look at CS treatment phenotypes systematically and prospectively, after a period of complete therapy cessation, based on serial PET scans. There has been one retrospective study which partially examined this question. Rosenthal described 9 patients who discontinued immunosuppression after initial complete resolution of cardiac inflammation on PET [12]. Immunosuppression was stopped after a mean duration of 32.0 ± 9.7 months. Patients underwent a repeat ¹⁸F-FDG-PET scan 8.4 ± 2.4 months after therapy cessation. CS recurrence by PET occurred in 8 of 9 (88.9 %) patients. However, it should be noted that mean LV SUV_max was only 1.73 (SD not stated) and only one patient had LV SUV_max > 2. Two of nine patients had cardiac events (episodes of ventricular arrhythmia) during therapy cessation. They did not report whole body PET findings and there was no change in LVEF during the period of discontinuation [12].

Chapelon-Abric reported on a clinical definition of relapse in 59 patients with CS. They defined relapse as the “reappearance of abnormalities on ECG, holter, echocardiography and at least one other imaging method (e.g., thallium scanning, MRI) on previously healed lesions.” [13] Over a median follow-up 60 months relapse of CS was diagnosed in 23 (39 %) patients. At the time of relapse, corticosteroids and immunosuppressive therapy (IST) had been stopped in five patients. They did not report how many of the 36 patients who did not relapse continued prednisone and/or IST [13].

4.3. Response to corticosteroids in extracardiac sarcoidosis

All 21 patients had reduction on extra-cardiac ¹⁸F-FDG uptake after steroid therapy and 16/21 (81 %) had complete resolution of extra-cardiac activity. We believe this is the first report of extracardiac steroid responsiveness using a PET scan definition of response. Goldman and Judson defined true corticosteroid failure in sarcoidosis as ongoing evidence of active granulomatous inflammation, causing symptoms, while receiving >30 mg/day of prednisone [14]. Using this definition they considered that this is likely to be very rare and our data would support this conclusion [14].

4.4. Relapse rates in extracardiac sarcoidosis after prednisone withdrawal

We found that all 19/19 patients had extra-cardiac relapse on the third PET scan. There are more data on relapse rates after therapy cessation in patients with extracardiac sarcoidosis, compared to the literature related to CS relapse. However, the literature has many limitations: firstly, the lack of an agreed definition of relapse (many reports use clinical definitions, some have used imaging-based definitions, others both); secondly, many studies included heterogeneous sarcoidosis populations; thirdly, different criteria for discontinuing therapy; and finally, variable follow up duration. Likely reflecting these many issues, the reported rates of extra-cardiac sarcoidosis relapse vary widely from 3.4 % to 75 %; relapse rates are higher when imaging is included in the definition [15], [16], [17], [18], [19], [20].

Gottlieb et al. defined relapse as recurrence of symptoms of sufficient severity to warrant re-initiation of treatment with corticosteroids, following remission without treatment lasting for >1 month [16]. They observed a relapse rate of 76/103 (74 %) [16]. Johns et al. reported a clinical relapse rate of 75% [17] and Rizzato et al. 30/82 (36.6 %) during a mean follow-up of over 3 years [18]. In contrast, Hunninghake found a clinical relapse rate of only 5/37 (13.5 %), although they do not state how long the patients were followed for [19]. A multi-centre study followed 500 patients from 10 clinics for 5 years, and of these 353/500 (70.6 %) required treatment [15]. Amongst this population, 136/353 (38 %) had treatment stopped with no subsequent re-initiation. However, they did not report the number of patients who had to reinitiate therapy after a trial of cessation [15].

In general, reported relapse rates are lower for patients with newly diagnosed disease compared to chronic disease and the following studies illustrate this. Pietinalho randomized patients with stage I and stage II newly detected (<3 months) pulmonary sarcoidosis to prednisone for 3 months followed by inhaled budesonide for 15 months or placebo tablets/inhaler therapy [20]. Patients were followed for 5 years. Of the patients who were initially treated with active therapy, only 2/65 (3.4 %) had symptom relapse requiring oral corticosteroids between 18 months and 5 years. In the placebo arm, 16/75 (21.3 %) of patients required active therapy for symptoms after 81 months [20]. In contrast Vorselaars et al. followed 47 patients after stopping infliximab (increasingly used as third line therapy in many forms of chronic sarcoidosis). Relapse was defined as the need for retreatment due to worsening of symptoms in combination with evidence of disease activity on a PET scan [21]. After a mean of 7.8 months, 29/47(62 %) experienced a relapse. In a similar study, Paneslinas showed a relapse rate of 86 % in 14 patients [22].

4.5. Limitations

The primary limitation is the small sample size. Also, 19/21 of the patients were Caucasian and the findings cannot be extrapolated to other patient populations. Thirdly, none of patients had significant LV dysfunction so unclear whether these patients would respond similarly. Finally the findings cannot be extrapolated to other therapies.

5. Conclusions

The main findings of our study are firstly, the initial response rate to prednisone was high with all patients showing a reduction in FDG uptake and 19/21 meeting a PET definition of >25 % response. Secondly, the cardiac relapse rate after prednisone discontinuation was also high. However, a trial of therapy discontinuation was able to identify a subset of 7/19 patients who did not require chronic therapy and are classified as acute phenotype. Furthermore, there were no later relapses during 3.7 years of subsequent follow-up. Hence these 7/19 patients can be classified as having an acute phenotype of CS and the other 12 as having a chronic disease phenotype. Importantly there were no serious adverse effects during the trial of therapy cessation. Finally, there were no clinical or imaging findings that could predict which patients were going to have relapse (i.e., have chronic disease phenotype). Our findings suggest that therapy guided by serial PET scanning can identify a subset of patients with clinically manifest cardiac sarcoidosis who do not require chronic therapy. Our findings need to be replicated in additional patient populations.

Sources of funding

The Canadian Institutes of Health Research (D. Birnie, NPI) for the Cardiac Sarcoidosis Cohort Study (CHASM-CS) (grant no. 148965, NCT01477359).

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: There are no known conflicts of interest associated with the principal or co-investigators of this research study or any member of their immediate family. RSB is or has been a consultant for- and receives grant funding from GE Healthcare, Lantheus Medical Imaging, and Jubilant DraxImage. The other members of this research team have nothing to disclose.