What's new?

We report on clinical features of Polish patients with hereditary transthyretin amyloidosis diagnosed with very rare transthyretin variants: p.Ala45Thr, p.Val91Ala, p.Phe53Cys, p.Ala101Val, p.Glu109Lys, and p.Phe53Leu. We described diagnostic pitfalls in transthyretin cardiac amyloidosis that may lead to a serious delay in establishing the diagnosis, such as low sensitivity of ^99mtechnetium^-3,3‑diphosphono‑1,2‑propanodicarboxylic acid scintigraphy in some specific cases and normal level of cardiac biomarkers in patients with intense cardiac uptake on scintigraphy. Our experience underscores the importance of both caution and awareness of the limitations of diagnostic tools, especially in cases that require going beyond the established diagnostic scheme.

Introduction

In recent years, the prognosis of patients with transthyretin cardiac amyloidosis (ATTR‑CA) has improved significantly. With technetium‑based scintigraphy and a noninvasive diagnostic algorithm proposed by Gillmore et al,¹ the diagnostic process has become much less complicated. The presence of intense cardiac uptake on bone scintigraphy using ^99mtechnetium‑3,3‑diphosphono‑1,2‑propanodicarboxylic acid ([^99mTc]Tc‑DPD) in the absence of monoclonal gammopathy greatly facilitates the diagnosis of ATTR‑CA. Bone scintigraphy with [^99mTc]Tc‑DPD has proven to be a useful and reliable technique, which makes histologic diagnosis redundant in the majority of patients. In addition, the advent of new therapies that effectively slow the progression of ATTR‑CA further improves patient prognosis and quality of life.

Considering the fact that ATTR‑CA treatment is more effective in the early stages of the disease, and that some variants in the transthyretin (TTR) gene are responsible for rapidly progressive heart failure (HF), timing of the diagnosis is crucial. Despite advances in the diagnostics of ATTR‑CA and the emergence of new management guidelines, the diagnosis of ATTR‑CA remains challenging in some cases.^2-4 Diagnostic difficulties result mainly from exceptions to established diagnostic rules, heterogeneity of ATTR, and still insufficient knowledge of the pathophysiology of the disease. The presence of essential red flags increases alertness to ATTR‑CA, but has limited sensitivity and specificity. We have become accustomed to the high diagnostic accuracy of [^99mTc]Tc‑DPD bone scintigraphy, almost forgetting its limitations and the fact that an absence of myocardial uptake of the radiopharmaceutical does not always rule out ATTR‑CA.⁵ For some specific TTR variants, for example, p.Val50Met, p.Phe84Leu, or p.Ser97Tyr, [^99mTc]Tc‑DPD bone scintigraphy has very low sensitivity, and the precise mechanism of bone tracer uptake in amyloidosis remains unexplained.^5,6 On the other hand, there are patients with intense cardiac uptake on bone scintigraphy without clinical signs or clinical suspicion of cardiomyopathy, which are usually necessary to trigger the diagnostic procedure.⁵

The aim of this study was to present the clinical features of Polish patients with pathogenic TTR variants diagnosed at our center, as well as to describe difficulties and pitfalls in their diagnostic process. Our experience underscores the importance of both caution and awareness of the limitations of diagnostic tools, especially in the cases that require going beyond the established diagnostic scheme.

Patients and methods

Between August 2018 and May 2024, 252 consecutive patients with suspected CA were routinely evaluated at the National Institute of Cardiology, Warsaw, Poland, including their medical history, physical examination, blood tests, standard 12‑lead electrocardiography (ECG), transthoracic echocardiography, and [^99mTc]Tc‑DPD bone scintigraphy. The suspicion of CA was raised in patients with left ventricular (LV) wall thickness of at least 12 mm and at least 1 red flag or clinical scenario, as described in detail previously.^2-4 Some of the patients had been previously diagnosed with ATTR and referred to our center for further evaluation or confirmation of the diagnosis. Blood tests included N‑terminal pro–B‑type natriuretic peptide (NT‑proBNP) and high‑sensitivity cardiac troponin T (hs‑cTnT) measurements, as well as free light‑chain testing and immunofixation of serum and urine to exclude light‑chain amyloidosis. The TTR gene sequencing was offered to patients with high myocardial uptake of [^99mTc]Tc‑DPD, patients who had first‑degree relatives with hereditary ATTR (hATTR), and individuals with inconclusive results. Genetic analysis was carried out using various platforms available over time. Most patients underwent genetic testing in a local laboratory of the National Institute of Cardiology. Genomic DNA was isolated from whole peripheral blood by standard techniques. The entire coding regions of the TTR gene together with splice sites were sequenced with the Sanger method using a 3130 × L genetic analyzer (Life Technologies, Foster City, California, United States) and the Big Dye Terminator v1.1 Cycle Sequencing Kit (Life Technologies), according to the manufacturer’s instructions. For chromatogram analysis, Variant Reporter 1.1 (Life Technologies) was used.

Results

Of the 252 patients, 56 (22.2%) were diagnosed with ATTR‑CA. A TTR analysis was performed in 48 of 56 patients (85.7%). Fourteen individuals (including 1 woman) were diagnosed with hATTR‑CA, 34 patients were diagnosed with wild‑type ATTR (wtATTR), and the remaining 8 patients refused genetic testing, 6 of whom were over 80 years old at the time of diagnosis (Figure 1). Most of the patients with hATTR‑CA had highly uncommon TTR gene variants: p.Ala45Thr, p.Val91Ala, p.Phe53Cys, p.Ala101Val, p.Glu109Lys, and p.Phe53Leu (Tables 1 and 2). The remaining 4 patients had the p.Val50Met TTR variant. All patients except those with the p.Glu109Lys TTR variant were unrelated. The patient with the p.Val91Ala TTR variant and 1 of the patients with the p.Val50Met TTR variant had a family history of hATTR. Eleven patients (78.6%) with hATTR‑CA presented with mixed phenotype and suffered also from peripheral polyneuropathy.

Typically, the patients diagnosed with hATTR‑CA were younger than those with wtATTR (P <⁠0.001), and had a lower burden of comorbidities. None of the patients in the hATTR‑CA group had arterial hypertension, while in the wtATTR group it was present in 67.6% of the participants (P <⁠0.001). Atrial fibrillation was also more frequently observed in the wtATTR group (P = 0.003). Carpal tunnel syndrome was observed in both groups with the same frequency (P = 0.59).

There was no significant difference between the groups in terms of ECG findings characteristic of CA, such as pseudoinfarct pattern or low QRS voltage (Table 3). Only 4 patients (28.6%) with hATTR‑CA had low QRS voltage in limb leads. Pseudoinfarct pattern in precordial leads was present in 10 individuals (71.4%) with hATTR‑CA. Features of LV hypertrophy on ECG were present extremely rarely: in 1 patient with hATTR‑CA and 5 with wtATTR.

Echocardiography showed increased thickness of LV wall in all patients. The maximal wall thickness varied markedly among the patients, from mildly to massively increased, and was higher in the patients with wtATTR (P = 0.045). Similarly, posterior wall thickness was greater in the wtATTR group (P = 0.01). The dimension of the left atrium was also greater in the patients with wtATTR (P = 0.004), which is consistent with more frequently observed atrial fibrillation in this group. Most of the patients had preserved LV ejection fraction (LVEF). There was no significant difference between the groups in terms of LVEF and diastolic dysfunction (E/e’ ratio). The patients with wtATTR had lower tricuspid annular plane systolic excursion value than those in the hATTR‑CA group (P = 0.04).

Serum NT‑proBNP levels were higher in the wtATTR group than in the patients with hATTR‑CA (P = 0.008). In the hATTR‑CA group, baseline laboratory workup showed increased levels of hs‑cTnT in all but 1 patient. With the exception of 2 patients, all participants with hATTR‑CA had elevated NT‑proBNP levels.

Discussion

The study presents clinical characteristics of patients with TTR variants: p.Ala45Thr, p.Val91Ala, p.Phe53Cys, p.Ala101Val, p.Glu109Lys, and p.Phe53Leu, which are very rare in the hATTR population. Previous reports on the p.Ala45Thr TTR variant described patients of Chinese and Japanese origin with leptomeningeal amyloidosis. Interestingly, no association of this variant with ATTR‑CA or urinary bladder amyloidosis has been previously reported.^7-10 In the case of the p.Phe53Cys TTR variant, only 1 patient with the same variant and surprisingly also of Polish origin has been described so far. The previously reported patient was a woman with vitreous opacities, cardiomyopathy, mild sensory neuropathy, and kidney disease, who underwent liver transplant.¹¹ The p.Val91Ala TTR variant was previously described in families from several different countries including Poland. Visual symptoms were reported in all families, sometimes preceding severe ATTR‑FAP; however, no marked cardiomyopathy was observed.¹² The p.Ala101Val TTR variant is unusually rare. To date, only a few patients with this variant have been presented, including 2 from our center.¹³ A more detailed description of Polish patients with the p.Glu109Lys and p.Phe53Leu TTR variants has been published previously, including information on previous literature reports on these variants.^13,14

In our study, cardiac biomarkers, hs‑cTnT and NT‑proBNP, did not reflect the intensity of cardiac uptake on [^99mTc]Tc‑DPD bone scintigraphy. Two patients with normal or borderline serum levels of the cardiac biomarkers already had high cardiac uptake of the tracer (patient 2 with the p.Val91Ala TTR variant and Perugini grade 3 and patient 14 with the p.Val50Met TTR variant and Perugini grade 2). The lack of association between cardiac biomarker levels and the Perugini grading was mentioned in previous reports, which also indicated that NT‑proBNP and hs‑cTnT are significant prognostic factors in ATTR‑CA, whereas stratification by the Perugini grade has no prognostic significance.^15,16 However, from a diagnostic point of view, the absence of such red flags as elevated levels of cardiac biomarkers may lead to a serious delay in the initiation of clinical investigation to confirm or rule out ATTR‑CA. It is therefore worth noting that transthyretin cardiomyopathy may occur in patients with normal or borderline levels of hs‑cTnT and NT‑proBNP.

On the other hand, the normotensive patient with elevated serum levels of cardiac biomarkers, especially hs‑cTnT, had only mild cardiac uptake (patient 3 with the p.Phe53Cys TTR variant and Perugini grade 1) despite increased myocardial thickness, strongly suggestive of cardiac infiltration with amyloid deposits. In the case of patient 3 with the p.Phe53Cys TTR variant and patient 4 with the p.Ala101Val TTR variant, it was hypothesized that mild or no cardiac uptake may be due to low sensitivity of [^99mTc]Tc‑DPD bone scintigraphy to these rare variants, similarly to what has been reported for several other TTR variants.^5,6 Another explanation for the low sensitivity of bone scintigraphy in these cases is an early stage of the disease, which is inconsistent with other results. The previously reported woman with the p.Phe53Cys TTR variant was diagnosed based on tissue biopsy, so it is currently impossible to find out whether p.Phe53Cys belongs to the group of variants for which [^99mTc]Tc‑DPD bone scintigraphy has very low sensitivity.¹¹ As for the p.Ala101Val TTR variant, 1 patient of Polish descent with a homozygous p.Ala101Val TTR variant and ATTR‑CA has been recently reported to have no myocardial uptake on technetium pyrophosphate ([^99mTc]Tc‑PYP) bone scintigraphy (Perugini grade 0). In the same study, another patient with a heterozygous p.Ala101Val TTR variant was described with positive [^99mTc]Tc‑PYP bone scintigraphy (Perugini grade 3).¹⁷ These findings and our observations suggest that sensitivity of [^99mTc]Tc‑DPD and [^99mTc]Tc‑PYP bone scintigraphy in detecting the homozygous p.Ala101Val TTR variant is extremely low. According to the literature, homozygous patients may develop more severe manifestations of the disease and present symptoms earlier than heterozygous individuals.¹⁷ However, the mechanism by which scintigraphy remains positive in heterozygous patients with the p.Ala101Val TTR variant and negative in homozygous patients remains unclear. Current guidelines³ recommend genetic testing in patients with confirmed ATTR to differentiate between hATTR and wtATTR. However, in the patients with a strong suspicion of ATTR, especially in a population where the TTR variants with low sensitivity on bone scintigraphy occur, TTR gene analysis should be considered, even if the [^99mTc]Tc‑DPD scintigraphy result is negative or inconclusive.