Cardiac Scintigraphy and Screening for Transthyretin Cardiac Amyloidosis: Caveat Emptor!!

A growing body of evidence now convincingly demonstrates that transthyretin cardiac amyloidosis (ATTR-CA) is an underdiagnosed cause of heart failure with preserved ejection fraction (HFpEF).¹ With the advent and widespread availability of FDA approved therapies for ATTR-CA and evidence suggesting enhanced efficacy if administered early in the course of the disease, there has been an intense focus upon “screening” for ATTR-CA that leverages non-invasive cardiac scintigraphy with bone-seeking tracers to identify early disease. The attraction of scintigraphy using technetium-based bone-avid radiotracers includes low cost, relative ease of administration and interpretation, availability, minimal radiation dose, and avoidance of a cardiac biopsy. While such an approach seems logical and appropriate based upon data highlighting the excellent performance characteristics of cardiac scintigraphy with evidence-based multi-societal guidelines endorsing its appropriate use², we have significant concerns regarding the widespread and inappropriate application of scintigraphy in populations that do not have features suggestive of ATTR-CA (see Figure 1). We, and others, have observed an alarmingly increasing rate of false positive results seen at amyloidosis specialty referral centers.³

Figure 1:

Suggested steps for optimal use of cardiac scintigraphy to diagnose ATTR-CA: (1) Clinicians should identify clinical features suggestive of the phenotype of cardiac amyloidosis, in order to, (2) increase pre-test probability, which will enhance positive predictive value and avoid false positive results, and (3) assess for monoclonal proteins to exclude light chain (AL) amyloidosis, employ SPECT imaging to confirm myocardial retention of isotope and use delayed imaging (e.g. after 3 hours) especially in lower prevalence population undergoing testing to minimize the confounding impact of blood pool tracer retention on scan interpretation.

The seminal paper that delineated a non-biopsy approach for the diagnosis of ATTR-CA evaluated two retrospective cohorts of subjects gleaned from 9 international referral centers for cardiac amyloidosis.⁴ One cohort included 374 subjects who underwent scintigraphy and an endomyocardial biopsy (the gold standard diagnostic test) in whom the prevalence of ATTR-CA was 70%. An additional 843 subjects underwent cardiac scintigraphy without an endomyocardial biopsy. Collectively, in the 1,217 subjects studied, with an overall prevalence of ATTR-CA of 44%, the diagnostic performance of positive scan result ( semi-quantitative grade 2 or 3) was excellent conferring a sensitivity of 90.4%, specificity of 97.1%, negative predictive value (NPV) of 92.9%, and positive predictive value (PPV) of 96%. When a positive scan result was combined with the absence of monoclonal proteins by serum and urine testing (thereby excluding light-chain (AL) amyloidosis, which was the most common cause of false positive scan), the specificity and PPV were 100%. It is critical to note that when using scintigraphy for non-biopsy diagnosis, uptake noted on planar imaging must be verified as myocardial (and not in the ventricular cavity or “blood pool”) by single photon emitted computed tomography (SPECT). Based upon these data, clinical providers have rightfully embraced this non-biopsy approach for the diagnosis of ATTR-CA in populations deemed to be high-risk such as older adults with cardiovascular conditions (e.g. HFpEF, atrial fibrillation, conduction disease), especially when accompanied by other non-cardiac ATTR manifestations, such as carpal tunnel syndrome or spinal stenosis. The multi-societal guidelines that outline appropriate utilization of scintigraphy clearly stipulate the proper clinical scenarios for use as well as the scenarios in which the test should not be applied. Unfortunately, our experience is that providers are not adhering to this well-vetted guidance resulting in increasing misdiagnoses, particularly in the context of planar imaging alone.

It is critical to recognize that diagnostic tests are often validated in populations with a high prevalence of the disease of interest, often with advanced phenotypes. Cardiac scintigraphy for ATTR-CA is no exception. The problems arising from spectrum and bias in the initial evaluation of diagnostic tests were predicted by Ransohoff and Feinstein⁵ to result in lower values for PPV and NPV over time when a test is subsequently applied to populations with lower prevalence and less severe disease. Indeed, they note that initial observations of diagnostic performance “often produced misleading results such that tests initially regarded as valuable were later rejected as worthless.” Such events are all too familiar in the practice of cardiovascular medicine. In the assessments of non-invasive stress testing for significant coronary artery disease identification, initial patient populations with known multi-vessel coronary artery disease (e.g. severe spectrum) and those with a high prevalence of disease were studied. When stress testing was performed in lower disease prevalence populations, the same issues were observed as PPV declined as patients with less severe coronary artery disease were studied.

Thus, while strong proponents of the appropriate utilization of cardiac scintigraphy as a non-invasive test to diagnose ATTR-CA, we are compelled to issue this warning. If one applies the test characteristics determined from a population with a prevalence of disease of ~50% to a population in which the prevalence of disease is 5% or less (screening), the PPV will decline from 96% to 62.1% (Figure 1). Accordingly, a large percentage of positive imaging results will in truth be false positive. Thus, for appropriate utilization, the population evaluated with scintigraphy must have features suggestive of the known clinical phenotype of ATTR-CA. We suggest that this includes older patients (over age 60 years), increased ventricular wall thickness (at least 12 mm or greater) and a non-dilated left ventricle along with other red flags (apical sparing, orthopedic manifestations [e.g. carpal tunnel syndrome, lumbar spinal stenosis or a biceps tendon rupture], persistently positive troponin, atrio-ventricular conduction block or a peripheral or autonomic neuropathy). Cardiac scintigraphy testing for ATTR-CA should not be performed in patients with a dilated cardiomyopathy, severe ischemic heart disease, or those with dilated left ventricles without increased wall thickness, or another known cause of heart failure. Furthermore, cardiac scintigraphy is preferably performed with delayed (e.g. 3 hours) imaging to minimize the impact of tracer retained in the blood pool, must include SPECT imaging and must be accompanied by testing for a plasma cell disorder to exclude light-chain amyloidosis. Failure to follow all these recommendations will not only result in misdiagnoses but incur significant clinical as well as financial consequences owing to high treatment co-pays and inappropriate downstream testing, while undermining the confidence among clinicians of scintigraphy as a method to achieve non-invasive ATTR-CA diagnosis even if applied appropriately. Cardiac scintigraphy affords great potential as a diagnostic test for early-stage ATTR-CA but utilization comes with the risk of great peril. All responsible parties including government regulators, industry partners, insurance payors, and academic societies should coordinate and work together to guide clinicians down the proper path.