Transthyretin Cardiac Amyloidosis in Patients With Severe Aortic Stenosis

Rodney H. Falk; Sharmila Dorbala


Eur Heart J. 2020;41(29):2768-2770. 

Transthyretin (TTR) amyloidosis, once considered a rare cardiac disease, is now recognized as playing a significant role in congestive heart failure in older patients. Among those with an increased wall thickness, congestive heart failure, and preserved ejection fraction, several centres have reported a prevalence of amyloidosis >10%.[1] The exponential growth of patients diagnosed with TTR amyloidosis over the last decade or so arises from an increased awareness, the development of therapeutic drugs, and the use of imaging agents with a high specificity for diagnosing the disease. Specifically, technetium 99mTechnetium-3,3-diphosphono-1,2-propanodicarboxylic acid (99mTc-DPD) imaging (previously used primarily as a bone-imaging agent) has been shown, in Europe, to be avidly taken up by the heart in patients with TTR cardiac amyloidosis. In the USA, a similar agent, 99mTc-pyrophosphate, PYP, is used. The impetus to use 99mTc bone avid tracer cardiac imaging for the diagnosis of TTR amyloidosis came from a multicentre, international study of 1217 patients, in which the finding of a positive scan had a sensitivity and specificity for the diagnosis of cardiac amyloidosis of 70% and 100%, respectively, provided that a plasma cell dyscrasia (a precursor for light chain amyloidosis) had been excluded.[2] That publication, based on data from several major international amyloidosis centres, subsequently was incorporated into international guidelines for the diagnosis of this condition.[3] However, it has been pointed out that by virtue of their expertise in amyloidosis, the prevalence of cardiac amyloidosis in the population studied was far higher than would be anticipated in a general cardiac population, raising the possibility that false-positive tests may be more common than originally believed. Another, often overlooked, conclusion of that paper was that this high specificity was in the group with myocardial uptake equal to or greater than rib uptake (labelled as 'grade 2'), whereas including patients with faint myocardial uptake (grade 1) decreased the specificity to 68%.[2]

Aortic valve stenosis, like TTR amyloidosis, is also a disorder of the elderly. With an ageing population, it is becoming a more prevalent cause of congestive heart failure. The introduction, and widespread acceptance of, transcatheter aortic valve implantation (TAVI) has revolutionized the treatment of aortic valve stenosis in the elderly, permitting valve replacement in this population with less morbidity and mortality than surgical replacement and allowing for replacement in a subgroup in whom surgical replacement was previously considered to be too high a risk[4] or even intermediate risk.[5] Not surprisingly, since TTR amyloidosis and aortic valve stenosis of both diseases of the elderly, these conditions may co-exist (Figure 1). In amyloidosis, the predominant mechanism of heart failure is due to infiltration of the myocardium by amyloid fibrils, resulting, in advanced cases, in a thickened ventricle, decreased left ventricular cavity size, severely reduced left ventricular longitudinal contraction, severe left atrial contractile dysfunction, and profound diastolic dysfunction. In aortic stenosis, left ventricular hypertrophy secondary to the markedly increased left ventricular afterload leads to secondary myocardial changes characterized by left ventricular hypertrophy, diastolic dysfunction, and interstitial myocardial fibrosis with impaired longitudinal contractile function. In very advanced cases of aortic stenosis, stroke volume may be significantly decreased, leading to the condition of 'low-flow, low-output' aortic stenosis. These patients appear to have advanced myocardial disease and a poorer prognosis after aortic valve replacement than those without this pathological condition.[6] Among patients with advanced TTR cardiac amyloidosis, cardiac output is reduced as a result of the mechanisms noted above. The superimposition of aortic valve stenosis, even of so-called moderate intensity, may aggravate heart failure, although the relative contributions of aortic stenosis and the primary myocardial disease may be very difficult to determine. Recent, relatively small studies,[7–9] sometimes in selected groups of aortic stenosis patients, have reported that aortic stenosis and cardiac amyloidosis may co-exist in up to 16% of patients[10] and that outcomes of aortic valve replacement in these patients are poorer than in patients without cardiac amyloidosis.[11,12] This would seem to be an intuitive finding, as symptoms in such patients may have predominantly arisen from the cardiac amyloidosis (and thus would not be expected to improve), and even when there is a component of symptomatic aortic valve disease, severe amyloid-induced myocardial dysfunction still persists once the valve is replaced. Consequently, there has been a reticence among cardiac interventionalists towards performing TAVI in a patient with known TTR cardiac amyloidosis because of perceived futility.

Figure 1.

This schematic demonstrates that aortic stenosis without (top) or with ATTR cardiac amyloidosis (bottom) are age related disorders with increased left ventricular wall thickness from distinct etiologies (myocyte hypertrophy vs. ATTR amyloid infiltration) but common clinical and pathophysiological manifestations. While grade 2/3 myocardial uptake of bone avid radiotracers is highly specific, grade 1 uptake is less specific, for ATTR cardiac amyloidosis (SPECT/CT fusion images shown in the right column).

In the study by Scully et al. in this issue of the European Heart Journal, the question of outcomes after TAVI among patients with aortic stenosis and probable concomitant TTR amyloidosis has been carefully addressed.[13] A total of 200 patients referred for TAVI, age 75 years and older, were prospectively and consecutively evaluated for TTR amyloidosis utilizing DPD scanning as the gold standard for diagnosis. The mean age of the population was 85 years, half were men, and 13% were deemed to have a combination of aortic stenosis and TTR amyloidosis based on a positive DPD scan. Scan positivity was equal to or greater than that of the ribs in 18 patients (9% of the total), and eight (4%) had faint but definite myocardial uptake (grade 1). Eventually, 149 patients (75% of the cohort) underwent TAVI (including 16/18 with positive DPD scans). The authors noted that outcomes did not differ whether or not the patient had had a positive DPD scan, with a 23% mortality at a median follow-up of 19 months post-DPD in the scan-positive group and 21% in the scan-negative group. They conclude that TAVI 'should not be denied to patients with aortic stenosis and amyloidosis' and that the data has 'disproved the preconception that TAVI may be futile in aortic stenosis with amyloidosis.'

While one may debate that there is a universal conception of the futility of aortic valve replacement in patients with co-existent amyloidosis, there does appear to be evidence in other studies that outcomes are poorer when amyloidosis exists.[11] What, then, accounts for the apparently more encouraging findings in the current study? One may argue that the current study is underpowered to detect a difference in outcomes, as the minority of patients had co-existent amyloidosis. While this is likely to be true, it is a larger study than previous publications and consisted of consecutive patients prospectively studied, suggesting a somewhat more accurate conclusion. However, a careful evaluation of the authors' data shows some unusual features which point towards the likelihood that the amyloid deposition in the current population was much less severe than in other studies. The mean septal and posterior wall thickness in the scan-positive group was 14 and 13 mm, respectively, compared with 13 and 11 mm in the aortic stenosis group. While this is a statistically significant difference between the groups, it should be recognized that TTR amyloidosis is a slowly progressive disorder with probably a long subclinical course. In our experience, most patients with symptomatic TTR amyloidosis have a ventricular wall thickness in excess of (and often well in excess of) 15 mm, suggesting that in the current series the degree of amyloid infiltration was relatively minor. This is underscored by the finding that no patient had a DPD uptake equal to or greater than ribs (grade 3), whereas the uptake in those patients labelled as having amyloidosis was less than that of the ribs (grade1) in a third of the patients labelled as having co-existent amyloidosis. Based on the seminal work of Gillmore et al.,[2] a grade 1 DPD uptake, while sensitive for TTR amyloidosis, only has a specificity of 68%. One might therefore question whether these patients, even if they did have TTR amyloidosis, were at an early enough stage for its presence to have little or no effect on the outcome of TAVI. In contrast, most (but not all) of the patients diagnosed with TTR amyloidosis in the setting of transcatheter aortic stenosis valve replacement in other series with worse prognosis had thicker ventricles and more advanced disease.[11] Since the indices of diastolic function in the amyloidosis patients in the current series are within the range that might be expected in severe aortic stenosis, the likelihood is that, in the clinical situation, even the most astute clinician might not consider co-existent amyloidosis. Under such circumstances, there would be no hesitation in performing aortic valve replacement. Indeed, published guidelines on investigation of patients suspected of having cardiac amyloidosis generally recommend investigation when left ventricular wall thickness is increased in the absence of significant valvular disease or poorly controlled hypertension. These patients do not fall into this category. In the case of patients with aortic stenosis, there is no fixed definition as to when to consider co-existent cardiac amyloidosis, but it generally should be suspected with a wall thickness well in excess of 15 mm, concomitant severe diastolic dysfunction, and (often) evidence of increased right ventricular wall thickening. This seems to be a very separate population from that studied here.

Does this mean that the conclusion of the authors that the belief in the futility of TAVI for concomitant aortic valve stenosis and cardiac amyloidosis is, itself, a futile conclusion based on the population they studied? That would be a harsh criticism which would not be merited. However, their population, and their inclusion of patients with grade 1 DPD uptake in their definition of cardiac amyloidosis, should be carefully noted. Their nuanced and careful discussion raises many unanswered questions, including this very one—whether the amyloid patients they described behave the same as those described among TTR amyloid patients without aortic stenosis or whether they represented an earlier stage of cardiac amyloidosis (or even a different manifestation of the disease). They are to be congratulated on a well-designed and interesting study which both advances our knowledge of this fascinating condition and opens up new vistas for research as to why some people are unfortunate enough to develop a disease caused by misfolding of an apparently normal protein, whereas others avoid extensive cardiac amyloid deposition. It also teaches us that we should not rely on imaging alone to make assumptions about prognosis and that while we should not forget cardiac amyloidosis as a contributory factor to heart failure in patients with concomitant aortic valve stenosis, we should also not forget that every patient needs a careful and individual evaluation prior to undertaking aortic valve replacement. Perhaps for those diagnosed with concomitant amyloidosis, stabilizers or silencers of TTR will become the next stage of treatment. But this, as the authors wisely implied, may require separate clinical trials to determine.