Access to Cardiac PET/CT by Sarcoidosis Patients and Cost-Effectiveness Analysis of Cardiac PET/MR compared to the Standard of Care

Abstract

Importance:

Cardiac sarcoidosis is associated with a high mortality rate. Given multiple barriers to obtaining cardiac PET imaging, we suspect individuals with access to this imaging modality are not representative of the Sarcoid patient population, which in the United States are predominantly Black females.

Objective:

To evaluate the demographics of patients with cardiac PET access and the cost-effectiveness of cardiac PET/MR imaging relative to standard of care.

Design:

This is a retrospective, observational study. The demographic information of patients with suspected cardiac sarcoidosis and cardiac PET/CT imaging within a national registry of sarcoidosis were reviewed (n=4,561). An individual-level, continuous, time-state transition model was used for the evaluation of long-term cost-effectiveness for the combined cardiac PET/MR compared to standard of care cardiac MR followed by cardiac PET/CT.

Results:

Patients who underwent cardiac PET in the national registry had 88.35% higher odds of being male (p<0.001) and 43.82% higher odds of being White (p=0.003) than their counterparts who did not have cardiac PET imaging. Combined cardiac PET/MR had overall lower total lifetime costs ($8,761 vs $10,777) and overall improved expected quality of life-years compared to the standard of care (0.77 vs 0.69).

Conclusion and Relevance:

The findings suggest that patients with access to cardiac PET/CT are not representative of the patient population most likely to have cardiac sarcoidosis in this limited study evaluation. Universal insurance coverage should be considered for Cardiac PET imaging as same day cardiac PET and MR imaging has potential long-term cost and quality of life benefit.

Introduction:

Cardiac sarcoidosis, an inflammatory heart condition associated with cardiac issues from arrhythmia to heart failure, is associated with a high mortality rate[1]. In the United States, 25% of deaths in patients with sarcoidosis have been attributed to cardiac involvement[2]. Cardiac sarcoidosis is clinically silent in 25% of sarcoid patients while symptomatic in only 5–10%. A study reviewing 320 autopsy cases of patients with sarcoidosis in Japan found that 46.9% had cardiac involvement; however, only 26.7% had an ante-mortem clinical diagnosis[3], suggesting a significant number are under diagnosed. A more recent epidemiological survey performed in Japan revealed that 23% of 1027 patients with sarcoidosis had cardiac involvement[4]. Although the gold standard for diagnosis is biopsy, false negative rates are high because of the patchy distribution of the granulomatous process.

A demographic review of over 700 newly diagnosed cases of sarcoidosis enrolled in the multicenter ACCESS Trial found that females were more likely diagnosed on average at age 40 years or above, while males were more likely younger at the time of diagnosis[5,6]. Among this study cohort, 64% were female and 44% were Black. Among epidemiological studies conducted in the United States, Black females had the highest prevalence rate for sarcoidosis[7]. One particular study looking at the racial differences in patients with sarcoidosis in Detroit found that Black females represented 39.1 cases out of every 100,000 individuals, followed by 29.8 cases for Black males, 12.1 cases for White females, and 9.6 cases for White males[8]. We expect that the patient population undergoing cardiac PET for sarcoidosis would be reflective of the prevalence rates for sarcoidosis.

The 2006 Japanese Ministry of Health, Labor, and Welfare guidelines recommended that cardiac abnormalities in the setting of a positive sarcoid biopsy from an extra-cardiac organ can be diagnostic for cardiac sarcoidosis. However, these guidelines lack specificity as other causes of cardiac abnormalities are dismissed [9]. The American College of Cardiology (ACC) developed an algorithm for the diagnosis of cardiac sarcoidosis [10], in which an abnormal electrocardiogram (ECG) indicates the need for subsequent evaluation with a cardiac magnetic resonance (cardiac MR) study. If indeterminate, then positron emission tomography (cardiac PET) is recommended for further workup. Cardiac MRI is limited for the diagnosis of cardiac sarcoidosis as the late gadolinium enhancement (LGE) phase of the cardiac MR may not detect early cardiac inflammation and cannot differentiate between scar and inflammation. To accommodate for this limitation, the Lake Louise criteria was proposed where tissue markers suggestive of myocardial inflammation in conjunction with myocardial edema on the cardiac MRI would support a diagnosis of acute myocarditis [11]. Despite this inclusion, cardiac MRI has a sensitivity and specificity of 95% and 85% respectively for the detection of cardiac sarcoidosis [12]. In contrast, Cardiac PET can differentiate between scar and inflammation using the dual imaging tracers of ¹³N ammonia to evaluate for perfusion abnormalities and ¹⁸F FDG to evaluate for myocardial inflammation. Cardiac PET independently has a sensitivity of 84% and specificity of 82% [12].

The novel dual imaging modality, cardiac PET/MR imaging has the potential to further improve diagnostic accuracy at an earlier state in the cardiac sarcoidosis, by combining the sensitivity of cardiac PET with the specificity of cardiac MR. Combined cardiac PET/MR, may potentially be both economically and medically beneficial for sarcoidosis patients compared to the standard of care, which is a separate cardiac MR followed by cardiac PET/CT. Among a cohort patients with suspected cardiac sarcoidosis, cardiac PET/MR was able identify cardiac sarcoidosis in 65% of the 51 patients tested[13]. The combined modality demonstrated a sensitivity of 94% [13], although a meta-analysis evaluating cardiac PET/MR sensitivity and specificity has not been conducted as yet. The main objective of this study was to evaluate access to cardiac PET imaging for the indication of sarcoidosis and determine the cost-effectiveness and quality of adjusted life years (QALYs) of same day cardiac PET/MR imaging relative to the standard of care.

Methods:

Patient demographics

This is a retrospective, observational study with institutional review board (IRB) exemption status (Protocol Number 20–11022976). Cardiac PET/CT imaging with the indication of cardiac sarcoidosis that were completed between January 1, 2018 and October 9, 2020 were identified from the New York-Presbyterian/Weill Cornell Medicine campus along with patient age, gender, race, zip code, and insurance type. Zip code was used as an alternative measure for income by correlating it with publicly available median household income from the 2020 US Census data (American Community Survey series S1903, in 2020 inflation-adjusted dollars). The primary endpoint was the diagnosis of cardiac sarcoidosis as clinically determined and described within the patients’ medical records. All adult patients (>18 years old) who underwent cardiac PET/CT with the indication of cardiac sarcoidosis from January 1, 2018 to October 9, 2020 were included in this study.

The National Sarcoid Registry is a registry of sarcoidosis patients who were willing to be included at the recommendation of their physicians. This registry was graciously shared by the Foundation for Sarcoidosis Research (FSR). Patient age, gender, race, and income of all sarcoidosis patients and sarcoidosis patient with PET were obtained from the registry. Insurance information regarding whether each individual in the registry had private or public insurance was not available from this registry and removed from further analysis.

Descriptive statistics were utilized for this analysis; continuous variables are represented as median (interquartile range) and range (minimum, maximum). Categorical variables are depicted as N (%). Multivariable logistic regression was used to evaluate the independent effect of demographic variables on access to cardiac PET/CT imaging. Adjusted odds ratios and 95% confidence intervals were estimated from the models. All p-values are two-sided with statistical significance evaluated at the 0.05 alpha level. All analyses were performed in R Version 4.1.3 (R Foundation for Statistical Computing, Vienna, Austria).

Health Economics Model

A health economic simulation model was used for the evaluation of long-term cost benefit. An individual-level continuous time state transition model (iCTSTM), a form of Markov modeling where the transition from one health state to another is dependent on co-variates, was used for the simulation. The hesim package[15] was used for implementation of the iCTSTM model on R (version 1.3.1093), which has probabilistic sensitivity analysis integrated. The target population was developed based on the findings described by Arkema and Cozier (2018)[16]. The health states used for this analysis are listed as the labelled boxes in the flowcharts (Figures 1–2). Transition periods were extracted from published literature and the FSR’s national-level registry. Simulations were independently developed and run for the “standard of care” and “cardiac PET/MR” imaging routes described in Figure 2. Standard of care imaging was defined as the separate acquisition of cardiac MR and cardiac PET/CT (¹⁸F FDG and ¹³N-ammonia perfusion imaging). Expected Quality Adjusted Life Years (QALYs) were compared between the standard of care and cardiac PET/MR using two-sample t-tests comparing the mean value of both groups. A p-value less than 0.5 was considered statistically significant.

Figure 1.

Diagram displaying the health states and transition phases used for the simulation. The time intervals and incidence and mortality rates were extracted from published literature[1,2,21–24].

Figure 2.

The standard of care flowchart based on transition rates from the FSR National Registry. For the purposes of this simulation, 50% of the demographic population utilized in the simulation was assigned to the standard of care route while 50% was assigned to the cardiac PET/MR route.

The cost for cardiac PET/MR imaging, defined as the average amount collected from a commercial insurer, was $16,806 at our institution. The cost for the standard of care (SOC), which consisted of a separate cardiac MR ($5,431) and cardiac PET/CT ($11,400) was $16,831. It was assumed that all patients who presented for cardiac MR would go through the cardiac PET imaging health states in the standard of care simulation, however only 34% obtained access based on data from the national sarcoid registry. This assumption was made as approximately 80% of patients incurred changes in management after PET was performed regardless of the MR result[17]. Health-related quality of life measures were extracted from Cox et al (2004)[18]. The annual medical cost for patients with sarcoidosis at the 50^th percentile was $7,659[19], while the annual medical cost increased to $73,346 for sarcoidosis patients with cardiac involvement. The final simulation outcomes were compared as incremental QALYs. To adjust for inflation over time, rates of 6% and 1.5% were utilized for costs and QALYs respectively[20].

The time from initial diagnosis to successful primary management (corticosteroids) was approximated to be 6 months[21]. The proportion of patients who died despite initial intervention was 1%[22]. The proportion of individuals who developed an abnormal EKG/echocardiogram (ECHO) despite successful primary management was 10%[1]. The proportion of patients who survived in 0, 5, 10, and 15 years post-steroid treatment was extracted from Nardi et al (2011)[23] such that there was 100% survival noted at the initial time point 0; 91.5% at 5 years; 84.1% at 10 years; and 78.1% at 15 years. The time from when a patient was found to have an abnormal EKG/ECHO to a secondary intervention (such as another round of steroids or implantable cardioverter defibrillator (ICD) placement) was approximately 1 month[24]. The time from the secondary intervention to when worsening cardiac abnormalities were noted including ventricular tachyarrhythmias was approximately 48 months. The mortality rate after the diagnosis of cardiac sarcoidosis was approximated at 25%[2].

82% of sarcoid patients who had a cardiac PET/MR were likely to have concordant findings on both imaging studies[25]. Among the patients who were referred for cardiac biopsy, 19.2% had a positive biopsy finding[26,27]. 62% of patients who had FDG-PET after a cardiac MR had a positive PET read while 34% had a negative PET read[1].

Results:

Patient demographics

Age, gender, race and income were collected from 98 patients with cardiac PET/CT from our institution (Table 1). 4,561 patients from the FSR’s National Sarcoid Registry were also reviewed for the same data, among whom 491 obtained a cardiac PET (Table 1).

Table 1.

Descriptive statistics of patient demographic information from the single-institutional dataset and the national sarcoid registry.

	Single-Institution	National Sarcoid Registry
Characteristic	N = 981	N = 4911
Age
Median (IQR)	62 (52, 71)	58 (53, 65)
Range	26, 88	6, 83
Unknown		371
Age (Binary)
<65	56 (57%)	87 (72%)
>=65	42 (43%)	33 (28%)
Unknown		371
Age (Categorized)
20–29	2 (2.0%)	0 (0%)
30–39	4 (4.1%)	0 (0%)
40–49	11 (11%)	18 (15%)
50–59	22 (22%)	44 (37%)
60–69	30 (31%)	45 (38%)
70–79	22 (22%)	11 (9.2%)
>=80	7 (7.1%)	1 (0.8%)
Unknown		372
Gender
Female	37 (38%)	287 (58%)
Male	61 (62%)	204 (42%)
Race/Ethnicity
American Indian/Alaska Native	1 (1.0%)	6 (1.2%)
Asian	4 (4.1%)	3 (0.6%)
Black	24 (24%)	47 (9.6%)
Hispanic/Latino	6 (6.1%)	18 (3.7%)
Mixed		10 (2.0%)
Other/Unknown	25 (26%)	8 (1.6%)
White	38 (39%)	399 (81%)
Language
Albanian	1 (1.0%)
Bengali	1 (1.0%)
English	87 (89%)
Korean	1 (1.0%)
Other	4 (4.1%)
Russian	1 (1.0%)
Spanish	3 (3.1%)
Median Household Income
Median (IQR)	85,342 (61,766, 130,968)	73,318 (57,418, 99,054)
Range	28,408, 250,000	24,834, 250,000
Unknown	1	99

¹n (%)

Demographic information was similar between our institution and the national registry for age, race and income. The majority of patients with cardiac PET were above the age of 50 at both our institution (82.1%) and the national registry (85%) with the most common range between 60–69 years of age. The greatest proportion of patients with cardiac PET were of White race at both our institution (39%) and from the national registry (81%).

Demographic information for gender and income was different between our institution and the national registry. The majority of patients with cardiac PET were male at our institution (62%), but female from the national registry (58%). The median household income for patients who had access to cardiac PET was $85,342 from our institution and $73,318 from the national registry, Demographic data from the national registry was compared between those with cardiac PET and those without cardiac PET (Table 2). Both groups have a median age of 58, with the majority being female and White. However, the multivariable logistic regression model demonstrated that males had 88.35% higher odds of undergoing cardiac PET/CT than females, adjusting for race (95% CI: 1.55, 2.28, p<0.001). White patients had 43.82% higher odds of undergoing cardiac PET/CT compared to non-White individuals, when adjusting for gender (95% CI: 1.14, 1.83, p=0.003). The median household income was significantly greater in patients with cardiac PET (p<0.001) at $73,318 compared to $65,883 in patients without cardiac PET (Table 2).

Table 2.

Descriptive statistics of patient demographic information from the national sarcoid registry where individual who received a cardiac PET/CT are compared with those who did not.

	Cardiac PET
Characteristic	Yes, N = 4911	No, N = 4,0711	p-value2
Age			0.56
Median (IQR)	58 (53, 65)	58 (51, 65)
Range	6, 83	5, 90
Unknown	371	2,743
Age (Binary)			0.83
<65	87 (72%)	945 (71%)
>=65	33 (28%)	383 (29%)
Unknown	371	2,743
Age (Categorized)			0.11
20–29	0 (0%)	7 (0.5%)
30–39	0 (0%)	47 (3.6%)
40–49	18 (15%)	242 (18%)
50–59	44 (37%)	403 (31%)
60–69	45 (38%)	427 (32%)
70–79	11 (9.2%)	180 (14%)
>=80	1 (0.8%)	11 (0.8%)
Unknown	372	2,754
Gender			<0.001
Female	287 (58%)	2,971 (73%)
Male	204 (42%)	1,100 (27%)
Race/Ethnicity			0.008
American Indian/Alaska Native	6 (1.2%)	21 (0.5%)
Asian	3 (0.6%)	40 (1.0%)
Black	47 (9.6%)	550 (14%)
Hispanic/Latino	18 (3.7%)	194 (4.8%)
Mixed	10 (2.0%)	155 (3.8%)
Other/Unknown	8 (1.6%)	87 (2.1%)
White	399 (81%)	3,024 (74%)
Median Income			<0.001
Median (IQR)	73,318 (57,418, 99,054)	65,833 (52,080, 85,354)
Range	24,834, 250,000	16,650, 250,000
Unknown	99	903

¹n (%)

²Wilcoxon rank sum test; Fisher’s Exact Test for Count Data; Fisher’s Exact Test for Count Data with simulated p-value (based on 2000 replicates)

Cost-effectiveness analysis

Using the iCTSTM model, the QALYs and total medical costs and expenditures favored the utilization of cardiac PET/MR as demonstrated in Table 3. Cardiac PET/MR not only improved the overall QALY of patients with sarcoidosis but was also associated with lower overall medical costs on average based on a simulation of 10,000 individuals at 500 iterations. The initial costs of cardiac PET/MR imaging were higher, but benefits in earlier detection, diagnosis, and intervention decreased the overall, lifetime medical costs. The average total medical and imaging costs per person was less for the cardiac PET/MR than the current standard of care such that the cost-effective ratio of the standard of care model was higher and less cost-effective at $15,618.84 per QALY while the ratio was $11,377.92 per QALY for the cardiac PET/MR model.

Table 3.

Simulation of costs and QALYs.

Outcome	Cardiac MR followed by Cardiac PET/CT (Standard of Care)	Cardiac PET/MR
QALYs	0.69 (0.56, 0.82)	0.77 (0.65, 0.92)
Costs: Imaging	1,290 (783, 1,807)	2,124 (1,395, 3018)
Costs: Medical	9,487 (7,890, 11,410)	6,637 (5,470, 7901)
Costs: total	10,777 (8,815, 13,025)	8,761 (7,304, 10,374)

Discussion:

Published literature has demonstrated that Black females were the most affected demographic group with sarcoidosis [8,28–30]. A recent review of the published literature analyzing the epidemiology of sarcoidosis demonstrated that there was not only a higher prevalence among Black individuals, but also a higher mortality rate at 16 deaths per one million individuals compared to 1.3 deaths per one million individuals for Caucasian patients [28]. However, according to data at this single institution and within the national registry, cardiac PET/CT imaging was most frequently obtained for individuals who were above the age of 50, White, with a median household income of $85,343 at the single-institution and $73,318 in the national sarcoid registry, both of which are greater than the overall median household income nationwide of $67,521[31]. There are likely multiple factors for why patients with access to cardiac PET/CT were not representative of patients most likely to have cardiac sarcoidosis, although evaluation is limited in this study for reasons discussed below. Awareness of this may be the first step in improving access to cardiac PET for sarcoidosis.

It is difficult to assume that cardiac sarcoidosis only affects patients with higher income, so one barrier for access to cardiac PET/CT is likely cost. The Centers for Medicare and Medicaid Services do not reimburse for cardiac PET/CT if the underlying indication is sarcoidosis [32] and select private insurance will cover the cost if a cardiac MRI is contraindicated or inconclusive. Lack of insurance coverage likely decreases accessibility to cardiac PET/CT for many patients. Additionally cardiac PET/CT may not be available in many institutions and there may be a lack of familiarity with this relatively novel modality among clinicians, both likely related to poor insurance coverage.

Our simulation model demonstrated that the novel dual imaging modality cardiac PET/MR, may potentially be both economically and medically beneficial for sarcoidosis patients compared to the standard of care, which is a separate cardiac MR followed by cardiac PET/CT. The health economic simulation found that cardiac PET/MR was initially more expensive than the standard of care. However, if cardiac PET/MR is integrated in the evaluation of early cardiac sarcoidosis, overall lifetime costs associated with this imaging study would be lower for the healthcare system compared to the standard of care. Diagnosis of cardiac sarcoidosis made at an earlier and reversible stage with cardiac PET/MR may decreased disease progression and complications, costly medical visits and hospitalizations, and invasive intervention. Our model also demonstrated that the quality of adjusted life years was improved for patients who directly underwent cardiac PET/MR as the first step of cardiac sarcoid evaluation compared to the standard of care. Cardiac PET/MR provides reasonable value compared to the standard of care meeting the threshold value of $50,000 per QALY gained[33].

The simulation aims to demonstrate that imaging of cardiac MR and PET performed within a short interval of each other improves outcomes. The simulation assumed all patients who obtained cardiac MR imaging in the standard of care model proceeded to be referred for cardiac PET imaging, however access was available only to one-third of the patients which likely contributed to the delay in diagnosis and poorer QALYs. Thus, the model suggests that cardiac MR and PET performed the same day for diagnostic purposes together can improve overall QALYs and outcomes. In order for this to occur, insurance coverage of cardiac PET must be permitted, allowing individuals being evaluated for cardiac sarcoidosis to obtain a cardiac PET study along with their cardiac MR.

There are several limitations in this study. The population in our institutional dataset and the national registry may differ from the population used to analyze national prevalence rates for sarcoidosis. The demographics of the patients presenting to the single institution are very representative of the institution’s catchment area, a predominantly White and upper middle class neighborhood. Many individuals at this medical institution who would have otherwise not been able to obtain a cardiac PET study through insurance had paid out-of-pocket for obtaining the study, which influences the findings of the single-institution demographic analysis. For this reason, the authors of this study opted to use a national sarcoid registry for further evaluation of patient access to cardiac PET imaging.

Another limitation is the use of the median household income based on zip code which may not correctly characterize the socio-economic level for each individual. Aside from insurance and cost, physicians and patients may also lack availability of cardiac PET imaging in their vicinity and therefore can only partially follow the recommended ACC algorithm. For this reason, there is reporting bias associated with the national registry data. Lastly, the simulation model assumed set parameters for each health state transition based on published literature which may vary in accuracy when compared to specific populations.

In conclusion, we found that patients with cardiac PET/CT at our institution and within the national registry were mostly above the age of 50, White and associated with a median household income greater than the national median. As epidemiologic data demonstrates Black females have the highest prevalence rate of sarcoidosis, our analysis highlights a discrepancy between individuals with sarcoidosis relative to those with access to cardiac PET imaging. Same day imaging with cardiac PET and cardiac MR may potentially be economically and medically beneficial for sarcoidosis patients as overall lifetime costs and the quality of adjusted life years are improved compared to the current standard of care algorithm. It therefore warrants insurance coverage of cardiac PET for the evaluation of sarcoidosis independent of cardiac MR findings.

Highlights:

• Sarcoidosis patients with access to cardiac PET/CT are not representative of the patient population most likely to have cardiac sarcoidosis in this limited study evaluation.

• Same day cardiac PET/MR has long-term cost-effectiveness compared to the standard of care imaging of cardiac MR followed by cardiac PET/CT sequentially.

• Same day cardiac PET/MR had overall improved expected quality of life-years for patients with cardiac sarcoidosis compared to the standard of care.