Multidisciplinary Approaches for Diagnosing Underrecognized Transthyretin Amyloidosis in Real-World Practice

Abstract

Background

Transthyretin amyloidosis (ATTR) is a rare but progressive disease. Its heterogeneous clinical presentation often leads to a delayed diagnosis. This study aimed to review the clinical manifestations of ATTR at diagnosis across cardiology and neurology subspecialties and to identify common red-flag symptoms in real-world practice.

Methods

We analyzed consecutive patients diagnosed with ATTR between March 2018 and December 2024. Variant ATTR (ATTR-v) was confirmed by genetic testing, while wild-type ATTR (ATTR-wt) was diagnosed based on the absence of monoclonal protein, a positive ^99mtechnetium-pyrophosphate scan, and no transthyretin gene mutations. All patients underwent clinical assessments, electrocardiography (ECG), and comprehensive echocardiography at diagnosis.

Results

Among the 63 enrolled ATTR patients (median age 64 years, 66.7% men), 27% were diagnosed in cardiology clinics and 73% in neurology clinics. Six had ATTR-wt (all diagnosed by cardiologists), while 57 had ATTR-v (94.7% with the Ala97Ser mutation), predominantly diagnosed by neurologists. Some patients presenting to cardiology clinics had considerable neurological symptoms or a history of bilateral carpal tunnel syndrome in addition to prominent cardiac involvement. Conversely, patients presenting to neurology clinics had significant cardiac involvement comparable to those diagnosed in cardiology clinics. Peripheral neuropathy was the most prevalent red flag, followed by autonomic dysfunction, discordant QRS voltages on ECG, and a history of bilateral carpal tunnel syndrome. Notably, the absence of low-voltage QRS or the presence of left ventricular hypertrophy on ECG did not necessarily exclude ATTR.

Conclusions

Recognizing red-flag symptoms remains the key to identifying ATTR. Early detection of ATTR requires clinical vigilance and multidisciplinary collaboration across subspecialties.

Abbreviations

A, Late diastolic transmitral flow velocity

A97S, Ala97Ser

AS, Aortic stenosis

ATTR, Transthyretin amyloidosis

ATTR-CM, Transthyretin amyloidosis cardiomyopathy

ATTR-PN, Transthyretin amyloidosis polyneuropathy

ATTR-v, Variant transthyretin amyloidosis

ATTR-wt, Wild-type transthyretin amyloidosis

CTS, Carpal tunnel syndrome

E, Early diastolic transmitral flow velocity

e′, Early diastolic mitral annular velocity

ECG, Electrocardiogram

hs-TnT, High-sensitive troponin-T

IQR, Interquartile range

LV, Left ventricular

LVEF, Left ventricular ejection fraction

LVH, Left ventricular hypertrophy

NT-proBNP, N-terminal pro-B-type natriuretic peptide

PYP, Pyrophosphate

Tc, Technetium

TTR, Transthyretin

INTRODUCTION

Transthyretin amyloidosis (ATTR) is a type of systemic amyloidosis characterized by misfolding of transthyretin (TTR), leading to the formation of cross-β-sheet-rich amyloid fibrils that deposit in multiple organs.¹ ATTR can be inherited as an autosomal dominant trait caused by pathogenic variants in the TTR gene (variant ATTR [ATTR-v]) or present without gene mutations (wild-type ATTR [ATTR-wt]). Among the more than 30 known amyloidogenic proteins, ATTR is the most common subtype associated with cardiac involvement, leading to ATTR cardiomyopathy (ATTR-CM) and subsequent major adverse cardiovascular events.² Given the emergence of disease-modifying agents for ATTR-CM in recent years,^3-5 identifying underrecognized cases of ATTR who may benefit from early treatment remains an unmet need in daily clinical practice.

Of the organs affected by systemic amyloidosis, the heart and nervous system are the most frequently involved in ATTR. Therefore, most clinical cases are identified in cardiology and neurology clinics. Unlike the presentation of ATTR-CM in cardiology clinics, patients may present to neurology clinics with predominantly polyneuropathic symptoms and be diagnosed with familial amyloidotic polyneuropathy.⁶ Given the spectrum of clinical manifestations, the phenotypic expression of ATTR can be heterogeneous, varying from predominant cardiac involvement to exclusively neurologic involvement.⁷ Therefore, close collaboration and mutual referral between cardiologists and neurologists are essential, both at the time of initial diagnosis and during ongoing follow-up. Red-flag symptoms and signs thus serve as important clues to help clinicians identify highly suspicious cases that warrant timely referral for an accurate diagnosis.^2,8 However, prior studies have been limited to patients with ATTR-v diagnosed by neurologists and included only a small number of patients with the Ala97Ser (A97S) genotype, which is prevalent in Taiwan.^6,9,10 Therefore, a comprehensive review of patient characteristics from a multidisciplinary perspective is needed. Consequently, the aim of this study was to review the clinical presentations of ATTR at the time of initial diagnosis and compare the prevalence of red-flag symptoms and signs across different subspecialties.

METHODS

Study design and participants

This retrospective cohort study included patients diagnosed with ATTR at a tertiary referral medical center in Taiwan, recruited from cardiology or neurology clinics between March 2018 and December 2024. ATTR-v was diagnosed through TTR gene sequencing, while ATTR-wt was identified using non-biopsy criteria, including the absence of a monoclonal protein, positive cardiac scintigraphy, and no known TTR gene mutations.¹¹ Mutant carriers without evidence of cardiac or neurological involvement were excluded from the study.

Clinical symptoms and signs

Cardiac involvement was defined as heart failure symptoms, arrhythmias, or conduction abnormalities, accompanied by unexplained left ventricular hypertrophy (LVH; a maximal left ventricular wall thickness ≥ 15 mm on echocardiography). Neurological involvement was defined as: (1) the presence of sensory (numbness, dysesthesia, or allodynia) or motor (limb weakness or muscle wasting) symptoms; or (2) signs of dysautonomia (constipation, diarrhea, erectile dysfunction, dysuria, or postural hypotension), documented through a neurologist’s clinical evaluation and/or confirmed by neurophysiological studies (electroneurography, electromyography, skin sympathetic response, or R-R interval variability). A history of bilateral carpal tunnel syndrome (CTS) was defined as a documented history of bilateral nerve decompression surgery or confirmation by electroneurography. Other (non-cardiac, non-neurological) clinical symptoms included unexplained weight loss (decrease of > 10% within six months), lumbar spinal stenosis (ligamentum flavum thickening, confirmed by computed tomography or magnetic resonance imaging), and atraumatic biceps tendon rupture (confirmed by sonographic evidence of a tear in the biceps tendon).

ATTR-CM was defined as isolated cardiac involvement, while ATTR polyneuropathy (ATTR-PN) was defined as isolated neurological involvement. Patients presenting with both cardiac and neurological involvement were classified as having a mixed phenotype. Laboratory data including N-terminal pro-B-type natriuretic peptide (NT-proBNP) and high-sensitivity troponin-T (hs-TnT) measured within six months of the initial diagnosis were recorded.

This study was approved by the Institutional Review Board (2020-05-002C, 2021-08-009BCF), and written informed consent was obtained from each patient. The study adhered to the principles outlined in the Declaration of Helsinki.

Electrocardiography

Electrocardiography (ECG) performed at the time of the initial diagnosis was reviewed. The PR interval was measured from the beginning of the P wave to the start of the QRS complex in patients with sinus rhythm. The diagnosis of LVH by ECG was determined according to the Sokolow-Lyon criteria.¹² Discordance between QRS voltage and left ventricular (LV) wall thickness on echocardiography was defined as a maximal LV wall thickness > 15 mm on echocardiography without meeting the Sokolow-Lyon criteria on ECG. Low QRS voltage was defined as the amplitude of all QRS complexes ≤ 0.5 mV in all limb leads or ≤ 1 mV in all precordial leads.

Echocardiography

A transthoracic echocardiographic study was conducted according to the recommendations of the American Society of Echocardiography.¹³ In brief, maximal LV wall thickness was defined as the thickest LV wall measurement obtained using two-dimensional echocardiography. Transmitral inflow parameters including the ratio of early (E) to late (A) diastolic transmitral flow velocities were measured using pulsed-wave Doppler echocardiography. Early diastolic mitral annular velocity (e′) was measured using tissue Doppler echocardiography. Left ventricular ejection fraction (LVEF) was assessed using the modified Simpson rule. Pericardial effusion was defined as the presence of an echo-free space between the parietal and visceral pericardium at end-diastole, categorized by size as small (< 10 mm), moderate (10-20 mm), or large (> 20 mm).¹⁴

^99mTechnetium-pyrophosphate scintigraphy

The primary physicians determined whether the patients should undergo ^99mTechnetium (Tc)-pyrophosphate (PYP) scintigraphy. Planar imaging and single-photon emission computed tomography were performed to differentiate myocardial PYP uptake from blood pool retention. The extent of myocardial uptake was determined by visual grading as follows: Grade 0 = no myocardial uptake with normal bone uptake; Grade 1 = myocardial uptake that is less than rib uptake; Grade 2 = myocardial uptake that is equal to rib uptake; Grade 3 = myocardial uptake exceeding rib uptake, often accompanied by mild or absent rib uptake. Generally, grades 2 and 3 are considered positive results, supporting the diagnosis of ATTR-CM.¹⁵

Statistical analysis

Continuous variables were reported as median with interquartile range (IQR) and compared using the Mann-Whitney U test for non-normally distributed data. Categorical variables were presented as absolute number and percentage and compared using Fisher’s exact test. All statistical analyses were performed using R software (version 4.0.2, R Foundation for Statistical Computing). All tests were two-sided, with p < 0.05 considered statistically significant.

RESULTS

Distribution of ATTR subtypes by subspecialty

A total of 63 patients with ATTR were enrolled (median age 64 years [IQR: 60-68.5 years], 66.7% men). Among the patients, 17 (27%) were initially diagnosed by cardiologists, and 46 (73%) were diagnosed by neurologists. Six patients were diagnosed with ATTR-wt, and 57 were diagnosed with ATTR-v. Among the patients with ATTR-v, 54 (94.7%) had the A97S genotype, which is the most prevalent genotype in Taiwan. One patient had A45T, one had G66D, and one had T80I (Figure 1).

Figure 1.

Flow chart of study population. A97S, Ala97Ser; ATTR, transthyretin amyloidosis; ATTR-v, variant transthyretin amyloidosis; ATTR-wt, wild-type transthyretin amyloidosis.

Figure 2 illustrates the distribution of ATTR-wt and ATTR-v cases, stratified by the subspecialty at initial diagnosis. ATTR-v was the predominant subtype in both those diagnosed by cardiologists (64.7%) and neurologists (100%). Notably, all ATTR-wt cases were diagnosed by cardiologists, while all cases diagnosed by neurologists were ATTR-v.

Figure 2.

Distribution of the subtypes of ATTR diagnosed from cardiologists and neurologists. ATTR, transthyretin amyloidosis; ATTR-v, variant transthyretin amyloidosis; ATTR-wt, wild-type transthyretin amyloidosis.

Clinical presentations by subspecialty

Table 1 includes the baseline demographics and reason for the first visit by subspecialty. The patients diagnosed by cardiologists were older and included fewer cases of ATTR-v compared to those diagnosed by neurologists. The primary presentations at the initial cardiology visit included heart failure symptoms or the presence of a permanent pacemaker with unexplained LVH. In contrast, those initially visiting neurology clinics mainly presented with distal paresthesia or gastrointestinal disturbances. Notably, 42.9% of the patients had a known family history of ATTR-v, and 22.2% were referred as newly diagnosed cases of ATTR from other hospitals.

Table 1. Demographics and reason for first visit by subspecialties.

Variables	Cardiologists (n = 17)	Neurologists (n = 46)	p value
Demographics
Age, years	67 (64-88)	63.5 (58.3-66.8)	0.004
Men, n (%)	12 (70.6)	30 (65.2)	0.8
ATTR-v, n (%)	11 (64.7)	46 (100).	< 0.001
Reason for first visit
Heart failure, n (%)	11 (64.7)	3 (6.5)	< 0.001
PPM with unexplained LVH, n (%)	3 (17.6)	0 (0)	NA
Distal paresthesia, n (%)	2 (11.8)	35 (76.1)	< 0.001
Gastrointestinal disturbances, n (%)	0 (0)	5 (10.9)	NA
Unexplained weight loss, n (%)	1 (5.9)	3 (6.5)	1.0
Family history of ATTR-v, n (%)	3 (17.6)	24 (52.2)	0.02
Referral as confirmed cases, n (%)	1 (5.9)	13 (28.3)	0.09

ATTR-v, variant transthyretin amyloidosis; LVH, left ventricular hypertrophy; NA, not applicable; PPM, permanent pacemaker.

Figure 3 summarizes the prevalence of the respective red-flag symptoms and signs. Among the red-flag symptoms, peripheral neuropathy was the most prevalent, with approximately 75% of the patients presenting with associated sensory symptoms at the initial diagnosis. It was also the most prevalent red-flag symptom in both the cardiology and neurology cohorts. Autonomic dysfunction including alternating constipation/diarrhea and postural hypotension was also prevalent, presenting in around half of the patients at the initial diagnosis. Discordance between QRS voltage and LV wall thickness and a history of bilateral carpal tunnel syndrome were recorded in more than 40% of the patients.

Figure 3.

Prevalence of the red-flag symptoms and signs stratified by subspecialties.

Cardiac manifestations

There were no significant differences in ECG findings between the two subspecialties, except that pseudo-Q wave was more prevalent in the neurologist cohort (Table 2). Only about 10% of the patients presented with diffuse low QRS voltage on ECG in both subspecialties. In contrast, discordance between QRS voltage and LV wall thickness on echocardiography was identified in around half of the patients in both subspecialties. Of note, around 20% of the patients still met the Sokolow-Lyon LVH criteria. There were no significant differences between the two subspecialties in echocardiographic findings. However, the patients diagnosed by cardiologists had a numerically lower LVEF, higher E/A ratio, higher average E/e′, and higher prevalence of pericardial effusion. All patients presenting with pericardial effusion had only a small amount of fluid accumulation. Notably, none of the patients in this study presented with moderate or severe aortic stenosis (AS). Nearly all of the patients (94.1%) diagnosed by cardiologists had grade 2 or 3 uptake in ^99mTc-PYP scintigraphy, compared to around three-fourths (76.2%) of the patients diagnosed by neurologists. The only patient in the cardiologist cohort without significant uptake on ^99mTc-PYP scintigraphy was identified through family screening of an index patient diagnosed by a cardiologist due to prominent cardiac involvement. Regarding biomarkers, the patients diagnosed by cardiologists had significantly higher levels of NT-proBNP and hs-TnT compared to those identified by neurologists (Table 2).

Table 2. Cardiac presentations by subspecialties.

Variables	Cardiologists (n = 17)	Neurologists (n = 46)	p value
Electrocardiogram
Rhythm			0.002
Sinus rhythm, n (%)	9 (52.9)	41 (89.1)
Atrial fibrillation, n (%)	6 (35.3)	2 (4.3)
Pacing rhythm, n (%)	2 (11.8)	3 (6.5)
Conduction
PR interval, ms	200 (62)	196 (36)	0.6
First-degree AV block, n (%)*	5 (55.6)	16 (39)	0.5
RBBB, n (%)#	3 (20)	5 (11.6)	0.4
LBBB, n (%)#	3 (20)	3 (7)	0.2
QRS complex
Pseudo-Q wave, n (%)	2 (11.8)	20 (45.5)	0.02
Low-voltage, n (%)	2 (11.8)	6 (13)	1.0
Discordance between ECG and echo, n (%)	8 (47.1)	20 (43.5)	1.0
LVH†, n (%)	3 (17.6)	11 (23.9)	0.7
Echocardiogram
Maximal LV wall thickness, mm	16 (4)	15 (5.8)	0.6
LVEF, %	55 (9)	60.3 (13.8)	0.05
E/A	1.5 (1.8)	1.1 (0.8)	0.2
Average E/e′	21.5 (10.2)	17 (12.7)	0.2
Pericardial effusion, n (%)	10 (58.8)	14 (30.4)	0.05
Moderate/severe aortic stenosis, n (%)	0 (0)	0 (0)	NA
PYP scan
Grade 2 or 3 uptake, n (%)	16 (94.1)	32 (76.2)	0.2
n = 59	n = 17	n = 42
Serum biomarkers
NT-proBNP, pg/ml	2238 (2086)	563 (1918)	0.008
hs-troponin T, ng/L	67.5 (29)	43 (32.3)	0.04

* Among patients with sinus rhythm. ^# Among patients with sinus rhythm or atrial fibrillation. ^† According to the Sokolow-Lyon criteria: S wave depth in V1 + tallest R wave height in V5 or V6 > 35 mm.

ATTR-v, variant transthyretin amyloidosis; AV, atrioventricular; E/A, the ratio of early to late diastolic transmitral flow velocity; E/e′, the ratio of early diastolic transmitral inflow velocity to early diastolic mitral annular velocity; ECG, electrocardiogram; hs-troponin T, high-sensitive troponin-T; LBBB, left bundle branch block; LVEF, left ventricular ejection fraction; LVH, left ventricular hypertrophy; NA, not applicable; NT-proBNP, N-terminal pro-B-type natriuretic peptide; NYHA, New York Heart Association; RBBB, right bundle branch block; PYP, pyrophosphate.

Extracardiac manifestations

The patients in the neurologist cohort presented more frequently with sensory, motor, or autonomic dysfunction (Table 3), whereas unexplained weight loss was more prevalent in the patients diagnosed by neurologists. Notably, nearly half of the patients (n = 28, 44.4%) had a history of bilateral carpal tunnel syndrome. Among them, 16 (57.1%) had undergone surgical decompression. The median time from surgery to the diagnosis of ATTR was 15 years. There was no significant difference between the two subspecialties in the proportion of patients undergoing surgery or the time interval from surgery to ATTR diagnosis. In addition, the prevalence of spinal stenosis and biceps tendon rupture was low. Only two patients (both ATTR-v, one in the neurologist cohort and one in the cardiologist cohort) had lumbar spinal stenosis. Only one patient (ATTR-wt, in the cardiologist cohort) had atraumatic biceps tendon rupture.

Table 3. Extracardiac presentations by subspecialties.

Variables	Cardiologists (n = 17)	Neurologists (n = 46)	p value
Sensory symptoms, n (%)	10 (58.8)	38 (82.6)	0.09
Motor symptoms, n (%)	4 (23.5)	19 (41.3)	0.2
Autonomic dysfunction, n (%)	5 (29.4)	24 (52.2)	0.2
Unexplained body weight loss, n (%)	1 (5.9)	15 (32.6)	0.05
Bilateral CTS, n (%)	9 (52.9)	19 (41.3)	0.6
Surgery for CTS, n (%)*	6 (66.7)	10 (52.6)	0.8
Time from the latest surgery to ATTR diagnosis, years	11 (18.8)	15 (6.8)	0.9
Lumbar spinal stenosis, n (%)	1 (5.9)	1 (2.2)	0.5
Atraumatic biceps tendon rupture, n (%)	1 (5.9)	0 (0)	NA

* Among those with bilateral CTS.

ATTR, transthyretin amyloidosis; CTS, carpel tunnel syndrome.

Clinical presentation by clinical phenotype and genotype

Table 4 presents the distribution of red-flag symptoms and comparisons between the three clinical phenotypes. Patients with isolated ATTR-CM (n = 5, 7.9%) had the highest prevalence of cardiac red flags and minimal to no extracardiac red flags. In contrast, patients with isolated ATTR-PN (n = 11, 17.5%) predominantly presented with extracardiac red flags but nearly no cardiac red flags. Importantly, most patients (n = 47, 74.6%) in our study presented with a mixed phenotype, characterized by a high prevalence of both cardiac and extracardiac red flags at their initial diagnosis.

Table 4. Distribution of red flags stratified by clinical phenotypes.

Red flags	ATTR-CM (n = 5)	ATTR-PN (n = 11)	Mixed (n = 47)	p value
Cardiac red flags
Pseudo-Q wave, n (%)	1 (20)	1 (9.1)	20 (44.4)	0.06
Low-voltage, n (%)	2 (40)	0 (0)	6 (12.8)	NA
Discordance between ECG and echo, n (%)	4 (80)	0 (0)	24 (51.1)	NA
Pericardial effusion, n (%)	4 (80)	0 (0)	20 (42.6)	NA
Moderate/severe aortic stenosis, n (%)	0 (0)	0 (0)	0 (0)	NA
Extra-cardiac red flags
Sensory symptoms, n (%)	0 (0)	11 (100)	37 (78.7)	NA
Motor symptoms, n (%)	0 (0)	2 (18.2)	21 (44.7)	NA
Autonomic dysfunction, n (%)	0 (0)	3 (27.3)	26 (55.3)	NA
Unexplained weight loss, n (%)	0 (0)	1 (9.1)	15 (31.9)	NA
Bilateral CTS, n (%)	2 (40)	4 (36.4)	22 (46.8)	0.9
Lumbar spinal stenosis, n (%)	0 (0)	0 (0)	2 (4.3)	NA
Atraumatic biceps tendon rupture, n (%)	0 (0)	0 (0)	1 (2.1)	NA

ATTR-CM, transthyretin amyloidosis cardiomyopathy; ATTR-PN, transthyretin amyloidosis polyneuropathy; CTS, carpel tunnel syndrome; ECG, electrocardiogram; NA, not applicable.

We further subdivided the patients with ATTR-CM and a mixed phenotype into ATTR-v and ATTR-wt subgroups. Table 5 shows comparisons of cardiac and neurological manifestations between these two subgroups. The patients with ATTR-v were more likely to have neurological presentations, including sensory and motor symptoms, dysautonomia, and unexplained weight loss. In contrast, those with ATTR-wt more frequently exhibited cardiac features, such as low-voltage on ECG or discordance between ECG and echocardiography, and pericardial effusion. However, the small number of ATTR-wt cases may have limited the statistical power to detect significant differences.

Table 5. Distribution of red flags among patients with ATTR-CM and mixed phenotype, stratified by genotypes.

Red flags	ATTR-v (n = 46)	ATTR-wt (n = 6)	p value
Cardiac red flags
Pseudo-Q wave, n (%)	20 (45.5)	1 (16.7)	0.4
Low-voltage, n (%)	6 (13.3)	2 (33.3)	0.5
Discordance between ECG and echo, n (%)	24 (52.2)	4 (66.7)	0.8
Pericardial effusion, n (%)	21 (45.7)	3 (50)	1.0
Moderate/severe aortic stenosis, n (%)	0 (0)	0 (0)	NA
Extra-cardiac red flags
Sensory symptoms, n (%)	35 (76.1)	2 (33.3)	0.09
Motor symptoms, n (%)	20 (43.5)	1 (16.7)	0.4
Autonomic dysfunction, n (%)	25 (54.3)	1 (16.7)	0.2
Unexplained weight loss, n (%)	15 (32.6)	0 (0)	NA
Bilateral CTS, n (%)	22 (47.8)	2 (33.3)	0.8
Lumbar spinal stenosis, n (%)	2 (4.3)	0 (0)	NA
Atraumatic biceps tendon rupture, n (%)	0 (0)	1 (16.7)	NA

ATTR-CM, transthyretin amyloidosis cardiomyopathy; ATTR-v, variant transthyretin amyloidosis; ATTR-wt, wild-type transthyretin amyloidosis; CTS, carpel tunnel syndrome; ECG, electrocardiogram; NA, not applicable.

DISCUSSION

From a multidisciplinary perspective, our study illustrates the characteristics of patients diagnosed in either cardiology or neurology clinics in real-world practice. Most of the ATTR cases were identified in neurology clinics, while about a quarter were diagnosed in cardiology clinics. ATTR-v was the predominant subtype diagnosed in both specialties. Notably, ATTR-wt was exclusively diagnosed by cardiologists. Our results also demonstrated that the patients who initially presented to cardiology clinics could have considerable sensory and motor symptoms or a history of bilateral CTS. Similarly, the patients who presented to the neurology clinics could already have significant cardiac involvement comparable to those diagnosed in cardiology clinics. These findings highlight that red-flag symptoms and signs serve as important clues for both subspecialties in identifying ATTR patients.

Red-flag signs of cardiac manifestations

Given the deposition of amyloid fibrils in the interstitial space of the myocardium, low-voltage QRS is a known red-flag sign for ATTR-CM on ECG.^16,17 Prior studies reported that low-voltage QRS was present in approximately 34-45% of patients with cardiac amyloidosis overall,^18,19 while 25-38% of patients with ATTR-v had this finding.^6,18,19 However, only about 10% of our patients presented with low voltage, highlighting that low-voltage QRS is a far less sensitive red-flag sign on ECG. This finding indicates that the absence of low-voltage QRS does not preclude the possibility of ATTR. Notably, approximately half of our patients had discordant QRS voltage patterns on ECG, while around 20% of the patients even met the ECG criteria for LVH based on the Sokolow-Lyon criteria. For both cardiologists and neurologists, discordant QRS voltage remains the most relevant red-flag sign on ECG for detecting ATTR in real-world practice. Our findings also emphasize that meeting the ECG criteria for LVH should not preclude the possibility of detecting ATTR in either subspecialty.

In addition to the findings from ECG, AS has also been shown to be a red-flag sign for ATTR on echocardiography.^8,17,20,21 Several registries of patients with severe AS referred for transcatheter aortic valve replacement reported that approximately 11% of patients were found to have underlying coexistent ATTR.^22,23 Notably, all patients with concurrent severe AS and ATTR had ATTR-wt.^23-26 This high prevalence of ATTR among patients undergoing transcatheter aortic valve replacement underscores the importance of identifying underrecognized cases, especially considering the growing number of patients undergoing the procedure in real-world practice. In another French registry comprising all consecutive patients with ATTR-wt, approximately 10% were found to have concurrent moderate or severe AS.²⁷ However, distinct from the previous studies, no patient in our study presented with moderate or severe AS. This discrepancy may be attributed to the small number of ATTR-wt cases in our registry, as this subtype is more commonly associated with severe AS than ATTR-v. Increased vigilance for concomitant ATTR among older patients with severe AS remains essential for the timely identification of those who may benefit from emerging disease-modifying therapies for ATTR-CM.

Multidisciplinary care with other specialties

In addition to cardiac red-flag symptoms and signs, extra-cardiac manifestations such as neurological manifestations are also important clues for identifying ATTR. With the rapid development of disease-modifying therapies for both ATTR-CM and ATTR-PN, joint evaluation by cardiologists and neurologists is essential to determine appropriate indications for these novel therapies.^28,29 In our study, 74% of the patients exhibited a mixed phenotype, highlighting the importance of close collaboration to ensure timely recognition and management. Polyneuropathy accompanied by red-flag symptoms and signs may suggest ATTR-v even from a neurologist’s perspective, however it should also prompt consideration of other differential diagnoses.³⁰ Therefore, genetic testing remains pivotal in distinguishing ATTR-v from other phenocopies.³⁰ Beyond conventional statistical methods, machine learning algorithms have also identified hidden patterns predictive of a positive genetic diagnosis.³¹ Consistent with our findings, these studies also identified unexplained weight loss, gastrointestinal disturbances, and cardiomyopathy as the key features associated with ATTR-v in patients with chronic axonal sensorimotor polyneuropathy. To enhance clinical suspicion and diagnostic accuracy, a standardized protocol with improved sensitivity and specificity is needed. Achieving this will require more real-world data and the integration of machine learning to address regional variability and improve diagnostic performance.

Apart from the neurological symptoms, orthopedic manifestations also provide an opportunity for diagnosis across different subspecialties. Bilateral CTS was the most prevalent orthopedic manifestation in the present study, accounting for approximately 40% of the patients. This finding highlights that CTS is not uncommon among red-flag symptoms and is frequently observed across both subspecialties. However, our results showed that the median time from the most recent surgery to the diagnosis of ATTR was up to 15 years, significantly longer than the 5 to 10 years reported in previous studies.^32,33 This prolonged time lag compared to Western cohorts highlights the need for clinicians to be more vigilant regarding a history of bilateral CTS, and consider more proactive longitudinal follow-up for these patients.³⁴

Aside from the history of bilateral CTS, the rates of lumbar spinal stenosis (3.2%) and ruptured biceps tendon (1.6%) were low in our cohort. Although a previous study reported that TTR deposition in ligament tissue could be prevalent in patients with lumbar spinal stenosis, none of these patients exhibited definite signs of cardiac amyloidosis.³⁵ In a study on prospective screening for ATTR-CM among patients undergoing spinal stenosis surgery, only 6.3% were found to have ATTR-CM, and all three confirmed cases were ATTR-wt.³⁶ In another study including a referral center cohort of 111 patients with ATTR-wt, the prevalence of ruptured biceps tendon was 33.3%, considerably higher than that observed in our study.³⁷ The lower prevalence of lumbar spinal stenosis and ruptured biceps tendon in our study may be due to the small number of patients with ATTR-wt. Since ATTR-wt was primarily diagnosed in cardiology clinics, these orthopedic red-flag manifestations should be of greater relevance to cardiologists.

In addition to manifested red-flag symptoms and signs, pathological evidence of TTR protein deposition in other organs also offers a clue for orthopedic surgeons to identify potential ATTR-CM. Previous case series have reported that amyloid deposition may first be identified in the prostate and urinary tract.^38-40 Among these studies, 37-80% of detected amyloid deposits were classified as ATTR, and it was the first time ATTR had been identified in 90% of these cases.^38-40 Notably, approximately 80% of these patients were subsequently found to have cardiac involvement.⁴⁰ Compared with the correlations between orthopedic red-flag signs and ATTR-CM, urologic deposition may offer an even more promising avenue for the early detection of ATTR-CM. However, further studies, particularly in collaboration with urologists, are needed to validate these findings in real-world clinical practice.⁴¹

In summary, identifying potential cases of ATTR requires increased vigilance in daily practice and collaboration between different subspecialties. Figure 4 summarizes the multidisciplinary approach in real-world clinical practice. Cardiologists and neurologists remain the main subspecialties to identify potential cases due to the prevalent presentations of cardiac and neurological involvement. For cardiac involvement, the initial encounter can occur in any cardiology subspecialty, such as general cardiologists, interventional cardiologists, electrophysiologists, heart failure specialists, or electrophysiologists. The prevalent extra-cardiac red flags, such as bilateral CTS and dysautonomia as alternating diarrhea and constipation, highlight the important roles of orthopedic surgeons and gastroenterologists, respectively. However, to achieve an accurate diagnosis, collaboration between hematologists, nuclear medicine physicians, and pathologists is mandatory.

Figure 4.

Multidisciplinary approaches for identification of the underrecognized ATTR among different subspecialties. ATTR, transthyretin amyloidosis; GI, gastrointestinal; PCI, percutaneous coronary intervention.

Study limitations

This study was subject to the following limitations. First, the nature of the retrospective review, and the clinical symptoms were recorded from medical records. Although all red-flag symptoms and signs were systematically assessed at initial enrollment, information on operative history, such as dates and etiologies from other hospitals, may be subject to recall bias. Second, although all patients were evaluated by neurologists, only 51 (81%) patients underwent neurophysiological studies such as ENG and EMG. Similarly, patients reporting low back pain did not uniformly receive lumbar computed tomography or magnetic resonance imaging. Consequently, the prevalence of neurological involvement and lumbar spinal stenosis may have been underestimated. Third, we primarily included confirmed cases from the cardiology and neurology clinics of a single tertiary referral center, and 22.2% of the patients were referred as newly diagnosed ATTR cases from other hospitals. We were unable to report the total number of patients screened or followed in both clinics. Nevertheless, our results offer valuable insights into the actual prevalence of red-flag signs and symptoms among confirmed cases in a real-world clinical setting. Fourth, consistent with the real-world prevalence in Taiwan, most cases were ATTR-v, with the A97S mutation being the most common genotype. As such, the generalizability of our findings to other geographic regions may be limited. Fifth, although we included a relatively large cohort of patients with ATTR, ATTR could still be underrecognized in other subspecialties. Collaboration with other subspecialties remains essential to identify hidden cases earlier, before cardiac or neurologic manifestations become apparent.

CONCLUSIONS

ATTR-v was the predominant subtype in both those diagnosed by cardiologists and neurologists, whereas ATTR-wt was exclusively diagnosed by cardiologists. In real-world practice, both cardiologic and neurologic involvement can be prominent in the clinics of both subspecialties. Peripheral neuropathy, autonomic dysfunction, discordance between QRS and wall thickness, and a history of bilateral CTS were the most prevalent red flags, serving as key diagnostic clues for both subspecialties. Early identification of underrecognized ATTR requires heightened clinical vigilance and close multidisciplinary collaboration across subspecialties.

DECLARATION OF CONFLICT OF INTEREST

All the authors declare no conflict of interest.

FUNDING INFORMATION

This study was supported by the Taipei Veterans General Hospital (TVGH), Tri-Service General Hospital (TSGH), and Academia Sinica (AS) Joint Research Program (VTA113-V1-6-2, VTA112-V1-7-2, VTA111-V1-5-2). YT Liu thanks for the support of TVGH by the TVGH, TSGH, AS Joint Research Program (VTA111-V1-5-1, VTA112-V1-7-1, VTA113-V1-6-1) and the Brain Research Center, National Yang Ming Chiao Tung University from The Featured Areas Research Center Program within the framework of the Higher Education Sprout Project by the Ministry of Education (MOE), Taiwan (MOE 111BRC-B408, 112BRC-B408, 113BRC-B408).