Real-World Clinical Impact of Tafamidis on Transthyretin Amyloid Cardiomyopathy　― A Decade of Experience From a Community-Based Hospital (2015–2024) ―

Abstract

Background: Transthyretin amyloid cardiomyopathy (ATTR-CM) is increasingly recognized as a major cause of heart failure in elderly patients with left ventricular hypertrophy. Although tafamidis was approved in 2019 following the ATTR-ACT study, its real-world survival impact in community settings remains unclear.

Methods and Results: This retrospective study analyzed 117 patients diagnosed with ATTR-CM at a single center from 2015 to 2024, with 75 receiving tafamidis and 42 untreated. Among the 83 patients who underwent genetic testing, all had the wild-type genotype. ATTR-CM diagnoses increased significantly after the advent of ^{99 m}Tc-pyrophosphate scintigraphy and tafamidis. Kaplan-Meier analysis showed significantly longer survival in tafamidis-treated patients. Multivariate analysis identified New York Heart Association (NYHA) functional class, left ventricular wall thickness, N-terminal prohormone of brain natriuretic peptide (NT-proBNP) levels, and tafamidis treatment as independent survival predictors. Tafamidis treatment was associated with significantly improved survival in patients who were younger, had a less advanced NYHA functional class, and lower levels of NT-proBNP and troponin T. In contrast, its survival benefits were marginal in patients with older age, higher NYHA functional class, elevated NT-proBNP levels, and increased troponin T levels.

Conclusions: In this real-world cohort, tafamidis treatment was significantly associated with better survival in ATTR-CM patients, particularly when initiated in the early stage. Therefore, early detection and timely initiation of treatment are critical for optimizing clinical outcomes in this increasingly recognized condition.

Transthyretin amyloid cardiomyopathy (ATTR-CM) has been increasingly recognized as a major cause of heart failure in elderly patients with left ventricular hypertrophy.¹ We previously reported that ^{99 m}Tc-pyrophosphate scintigraphy is a reliable diagnostic tool for ATTR-CM, with 18% of patients who were using this method screened testing positive, suggesting a high prevalence of ATTR-CM in real-world clinical practice.²

Based on the results of the ATTR-ACT study,³ tafamidis was approved for the treatment of ATTR-CM, regardless of transthyretin gene mutations, in 2019.⁴ Since then, the number of patients with ATTR-CM treated at Saiseikai Fukuoka General Hospital has increased significantly. However, the effects of tafamidis on the survival of patients with ATTR-CM in real-world clinical practice remain unclear, particularly in community-based hospitals like ours.

The purpose of the present study was to review the clinical management of ATTR-CM at Saiseikai Fukuoka General Hospital over the past decade, beginning with the first diagnosed case in 2015, and to evaluate the survival benefit of tafamidis in real-world clinical practice conducted in a community-based hospital setting.

Methods

Study Design and Patients

This study is a single-center, retrospective observational analysis. The study protocol was approved by the Ethics Committee of Saiseikai Fukuoka General Hospital (2025-2-1). The investigation was conducted in accordance with the principles outlined in the Declaration of Helsinki.

All patients diagnosed with ATTR-CM and treated at Saiseikai Fukuoka General Hospital between 2015 and 2024 were retrospectively analyzed based on their clinical records. The details of the diagnosis and treatment of ATTR-CM were conducted in accordance with the guidelines and statements of the Japanese Circulation Society.^1,4 ATTR-CM was classified as definite if transthyretin amyloid deposition in the myocardium was confirmed by myocardial biopsy, and probable if based solely on positive results from ^{99 m}Tc-pyrophosphate scintigraphy without biopsy. The date of diagnosis was defined as the date of myocardial biopsy for definite cases and the date of ^{99 m}Tc-pyrophosphate scintigraphy for probable cases.

Patients were categorized as ‘tafamidis treated’ if they received tafamidis for more than 2 weeks by the end of 2024; otherwise, they were classified as ‘tafamidis untreated’. One patient who discontinued tafamidis within 2 weeks of initiation was categorized as tafamidis-untreated. For tafamidis-treated patients, the date of the first prescription of tafamidis was recorded as the treatment initiation date.

The duration in days or years from diagnosis was calculated from the date of diagnosis to the date of death, the last outpatient visit, or the end of 2024 for patients confirmed alive in 2025. Similarly, the duration from tafamidis initiation was defined as the period from the date of tafamidis initiation (for tafamidis-treated patients) or the date of diagnosis (for tafamidis-untreated patients) to the date of death, the last outpatient visit, or the end of 2024 for patients confirmed alive in 2025.

Statistical Analyses

All statistical analyses were performed using EZR 1.54 (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria).⁵ Continuous variables are expressed as mean±standard deviation or median [range], unless otherwise specified. Comparison of continuous variables between 2 independent groups was performed using the unpaired Student’s t-test. Categorical data were reported as frequencies and percentages and were compared using a chi-squared or Fisher›s exact test. Survival curves were generated using Kaplan-Meier analysis and compared using the log-rank test. The receiver-operating characteristic analysis was used to obtain cut-off values to convert continuous parameters to categorical variables. A multivariate analysis was performed using the Cox proportional hazards model for the survival time analysis.

All tests were 2 tailed, and a P value of <0.05 was considered to be statistically significant.

Results

Clinical Characteristics of Patients With ATTR-CM

Over the past decade, 117 patients were diagnosed with ATTR-CM, including 92 (79%) with definite diagnoses and 25 (21%) with probable diagnoses. The age at diagnosis ranged from 63.3 to 100.1 years (median 80.7 years). Patients with a probable diagnosis were significantly older than those with a definite diagnosis (89.9±7.1 vs. 78.0±6.2 years; P<0.0001). Although genetic analysis of the transthyretin gene was not performed in all patients, all of those tested (n=83) were confirmed to have the wild-type genotype. Notably, every patient diagnosed before the age of 72.4 years (n=21) underwent genetic analysis and was likewise confirmed to have the wild-type genotype. Therefore, this cohort was likely predominantly composed of patients with the wild-type genotype.

The clinical characteristics of the patients are summarized in Table 1. The mean age at diagnosis was 80.6±8.0 years, with 90 (77%) males and 27 (23%) females. Males were significantly younger at diagnosis (78.4±6.9 years) compared with females (87.9±7.3 years; P<0.001; Figure 1A). Regarding the New York Heart Association (NYHA) functional classification at diagnosis, 87 (74%) patients were classified as Class II, 29 (25%) patients as Class III, and 1 (1%) patient as Class IV. The patient classified as Class IV died of decompensated heart failure during hospitalization shortly after the diagnosis of ATTR-CM. The most common initial manifestation of ATTR-CM was acute decompensated heart failure, observed in 68 (58%) patients. This was followed by the occurrence or recurrence of atrial fibrillation (n=15; 13%), symptoms such as chest discomfort, shortness of breath, or syncope (n=12; 10%), and incidental findings of left ventricular hypertrophy (n=11; 9%). Less frequent initial manifestations included atrioventricular block (n=3; 2.6%), ventricular arrhythmia (n=2; 1.7%), aortic stenosis (n=2; 1.7%), and referrals for evaluation of a heart murmur (n=1; 0.9%), electrocardiogram abnormalities (n=1; 0.9%), left ventricular wall motion abnormalities (n=1; 0.9%), and pleural effusion (n=1; 0.9%).

Table 1.

Baseline Clinical Characteristics of Patients Grouped by Tafamidis Treatment and Survival Status

	Overall (n=117)	Tafamidis			Status
		Treated (n=75)	Untreated (n=42)	P value	Alive (n=93)	Dead (n=24)	P value
Age at diagnosis (years)	80.6±8.0	77.8±6.0	85.6±8.8	<0.001	80.0±7.3	82.7±10.2	0.148
Sex, male	90 (76.9)	66 (88.0)	24 (57.1)	<0.001	71 (76.3)	19 (79.2)	1
NYHA ≥III	30 (25.6)	17 (22.7)	13 (31.0)	0.380	18 (19.4)	12 (50.0)	0.004
LV wall thickness (mm)	14.2±2.4	14.2±2.4	13.9±2.6	0.359	14.1±2.2	14.5±3.2	0.507
LV EDD (mm)	43.3±6.0	43.6±5.0	42.7±7.5	0.453	42.8±5.5	45.3±7.7	0.069
LV ESD (mm)	32.4±6.0	33.0±5.2	31.3±7.2	0.144	31.8±5.7	34.8±6.8	<0.05
LV EF (%)	49.6±12.1	48.3±12.2	52.0±11.6	0.109	50.5±12.1	46.1±11.7	0.110
PCWP (mmHg)	18.2±8.0	18.5±8.4	16.9±6.7	0.469	17.8±8.2	19.6±7.4	0.469
Cardiac index (L/min/m2)	2.01±0.44	1.99±0.42	2.07±0.49	0.522	2.05±0.40	1.84±0.55	0.102
NT-proBNP (pg/mL)	4,523±5,718	2,869±2,838	7,477±8,003	<0.001	3,383±3,857	8,940±8,934	<0.001
eGFR (mL/min/1.73 m2)	50.8±17.3	52.5±14.8	47.8±20.8	0.161	52.1±15.2	45.8±23.4	0.116
Troponin T (ng/mL)	0.067±0.053	0.059±0.047	0.083±0.061	<0.05	0.059±0.046	0.098±0.066	<0.005
Year of diagnosis ≥2019	93 (79.5)	68 (90.7)	25 (59.5)	<0.001	80 (86.0)	13 (54.2)	<0.001

Data presented as mean±SD, or n (%). EDD, end-diastolic dimension; EF, ejection fraction; eGFR, estimated glomerular filtration rate; ESD, end-systolic dimension; LV, left ventricular; NT-proBNP, N-terminal prohormone of brain natriuretic peptide; NYHA, New York Heart Association; PCWP, pulmonary capillary wedge pressure.

Figure 1.

(A) Number of male and female cases by age at diagnosis. (B) Total number of tafamidis-treated and untreated cases by year.

Echocardiography evaluation revealed that the left ventricular wall thickness, measured as the greater of the interventricular septum or posterior wall, was 14.2±2.4 mm. Most (73%) patients had a wall thickness between 12 and 15 mm, 21% had thickness ≥16 mm, and 6% had thickness ≤11 mm (Figure 2A). The end-diastolic and end-systolic left ventricular dimensions were 43.3±6.0 mm and 32.4±6.0 mm, respectively. The left ventricular ejection fraction (LVEF) averaged 49.6±12.1%. LVEF was >50% in 49% of patients, between 40% and 50% in 31%, and <40% in 20% (Figure 2B).

Figure 2.

Distribution of (A) left ventricular wall thickness and (B) left ventricular ejection fraction.

Data from right heart catheterization were available for 72 (62%) patients. The cardiac index was 2.01±0.44 L/min/m², and the pulmonary capillary wedge pressure (PCWP) was 18.2±8.0 mmHg. The distribution of patients based on Forrester’s subsets was as follows: subset I, 17%; subset II, 11%; subset III, 34%; and subset IV, 38%. Most (72%) patients were in a low cardiac output state with a cardiac index <2.2 L/min/m², and approximately half (49%) of the patients exhibited pulmonary congestion with a PCWP >18 mmHg.

At the time of diagnosis, biomarker levels were as follows: NT-proBNP, 4,523±5,718 pg/mL; estimated glomerular filtration rate (eGFR), 50.8±17.3 mL/min/1.73 m²; and troponin T, 0.067±0.053 ng/mL. Only 4 (3%) patients had NT-proBNP levels <300 pg/mL, and 6 (7%) out of 91 patients had troponin T levels <0.014 ng/mL. Chronic kidney disease, defined as an eGFR <60 mL/min/1.73 m², was relatively common, affecting 70% of patients.

Atrial fibrillation was documented in 70 (60%) patients either before or after the diagnosis of ATTR-CM, with 31 (44%) of these patients undergoing catheter ablation. Cardiovascular implantable electronic devices, such as pacemakers or defibrillators, were implanted in 25 (21%) patients. Stroke occurred in 14 (20%) patients, including some cases in which patients were receiving anticoagulant therapy. Surgical aortic valve replacement was performed in 1 patient, while transcatheter aortic valve implantation was performed in 5 patients. Carpal tunnel syndrome was observed in 24 (21%) patients.

Clinical Characteristics of Patients Treated With Tafamidis

Among the 117 patients diagnosed with ATTR-CM, 75 (64%) received tafamidis treatment after 2019. Genetic analysis confirmed that all patients treated with tafamidis exhibited the wild-type genotype.

During the first 4 years of the study, 22 (19%; averaging 5.5 per year) patients were diagnosed with ATTR-CM, and none received tafamidis during that period. In the subsequent 3 years, 24 (20%; averaging 8 per year) patients were diagnosed, and in the final 3 years, 71 (61%; averaging 23.7 per year) patients were diagnosed. A key factor contributing to this marked increase in diagnoses over the past 3 years was an upsurge in referrals from other hospitals, primarily for the initiation of tafamidis treatment at our institution. Consequently, by the end of 2024, 75 of the 117 patients had been treated with tafamidis (Figure 1B).

The interval from ATTR-CM diagnosis to tafamidis initiation ranged from 76 to 1,569 days, with a median of 155 days. Importantly, patients diagnosed before 2019 (n=7) experienced a significantly longer delay in starting tafamidis compared with those diagnosed after 2019 (n=68; 956±338 days vs. 166±68 days; P<0.0001). Furthermore, among the 25 patients diagnosed in 2024, 18 initiated treatment within the same year, while an additional 2 are scheduled to begin in 2025. Consequently, the tafamidis treatment rate was significantly higher among patients diagnosed after 2019 than among those diagnosed before 2019 ([68+2]/93 [75%] vs. 7/24 [29%]; P<0.0001).

As summarized in Table 1, patients treated with tafamidis were significantly younger, had a higher proportion of males, and exhibited lower NT-proBNP and troponin T levels compared with untreated patients. There were no significant differences in NYHA functional class, left ventricular wall thickness, LVEF, PCWP, or cardiac index between tafamidis-treated and untreated patients. Notably, >40% of untreated patients were diagnosed before 2019, whereas >90% of treated patients were diagnosed after 2019.

Univariate Analysis of Survival After the Diagnosis of ATTR-CM

Following their diagnosis of ATTR-CM, a cohort of 117 patients was monitored over a median follow-up period of 532 days (range 4–3,409 days). By the end of 2024, 24 (21%) patients had died; apart from 1 case of suicide, all fatalities were attributable to cardiovascular causes. Compared with survivors, patients who died during follow up exhibited significantly higher levels of NYHA functional class, NT-proBNP, and troponin T at diagnosis (Table 1). In contrast, no statistically significant differences were observed between the groups with respect to age, gender, left ventricular wall thickness, LVEF, PCWP, or cardiac index.

Survival curves were generated using Kaplan-Meier analysis and compared using the log-rank test (Figure 3). Cut-off values for converting continuous parameters into categorical variables were determined using receiver operating characteristic analysis: age 79 years for age at diagnosis, 16 mm for left ventricular wall thickness, 44% for LVEF, 22 mmHg for PCWP, 1.83 L/min/m² for cardiac index, 7,461 pg/mL for NT-proBNP, 42 mL/min/1.73 m² for eGFR, and 0.082 ng/mL for troponin T.

Figure 3.

Survival curves free from all-cause death from the day of diagnosis (A), and grouped by (B) age at diagnosis, (C) gender, (D) New York Heart Association (NYHA) functional class, (E) left ventricular wall thickness, (F) left ventricular ejection fraction (LVEF), (G) pulmonary capillary wedge pressure (PCWP), (H) cardiac index, (I) N-terminal prohormone of brain natriuretic peptide (NT-proBNP), (J) estimated glomerular filtration rate (eGFR), (K) troponin T, and (L) tafamidis treatment. CI, confidence interval.

The 5-year survival rate for the overall population was 53.3% (95% CI 34.6–68.8%). Patients with older age (≥79 years), higher NYHA functional class (≥III), increased left ventricular wall thickness (≥16 mm), lower cardiac index (<1.83 L/min/m²), elevated NT-proBNP levels (≥7,461 pg/mL), reduced eGFR (<42 mL/min/1.73 m²), or elevated troponin T levels (≥0.082 ng/mL) demonstrated significantly worse survival compared with those with more favorable values for these parameters. In contrast, gender, LVEF, and PCWP had no significant effect on survival.

Treatment with tafamidis was associated with significantly better survival outcomes. The median survival for untreated patients was 3.536 years (95% CI 2.066–3.838 years), while for tafamidis-treated patients, the median survival exceeded 6.745 years (P<0.001).

The reason the survival curve for patients with left ventricular wall thickness ≥16 mm crossed that of patients with wall thickness <16 mm is attributed to a specific case: a 73.9-year-old male patient with a left ventricular wall thickness of 19 mm. This patient underwent surgical aortic valve replacement followed by tafamidis treatment and survived for 2,462 days (approximately 6.74 years) after being diagnosed with ATTR-CM.

Multivariate Analysis of Survival After the Diagnosis of ATTR-CM

A multivariate analysis was conducted using the Cox proportional hazards model to evaluate survival time. Explanatory variables included age at diagnosis, NYHA functional class, left ventricular wall thickness, NT-proBNP, eGFR, and tafamidis treatment, as these were statistically significant in the Kaplan-Meier analysis. To maximize the sample size, cardiac index and troponin T, which had missing values, were excluded from the analysis. Continuous variables were used for age at diagnosis, left ventricular wall thickness, NT-proBNP, and eGFR, unlike the Kaplan-Meier analysis. The time-to-event variable was defined as the number of years from diagnosis, and the status variable was all-cause death. The final model was selected using a stepwise method based on P values.

As summarized in Table 2, the determined final model included NYHA functional class, left ventricular wall thickness (mm), NT-proBNP levels (pg/mL), and tafamidis treatment as explanatory variables. NYHA functional class ≥III, increased left ventricular wall thickness, and elevated NT-proBNP levels were associated with significantly higher hazard ratios, while tafamidis treatment was associated with a significantly lower hazard ratio.

Table 2.

Multivariate Analysis of Survival After the Diagnosis

Factor	HR	95% CI	P value
NYHA ≥III	6.462	2.613–15.98	0.0001
Wall thickness (mm)	1.235	1.045–1.461	0.0135
NT-proBNP (pg/mL)/1,000	1.062	1.008–1.118	0.0241
Tafamidis treated	0.251	0.090–0.700	0.0083

CI, confidence interval; HR, hazard ratio; NT-proBNP, N-terminal prohormone of brain natriuretic peptide; NYHA, New York Heart Association.

Given that high troponin T levels were a significant indicator of poor prognosis, a multivariate analysis was performed that included troponin T as an explanatory variable alongside age at diagnosis, NYHA functional class, left ventricular wall thickness, NT-proBNP, eGFR, and tafamidis treatment. Because troponin T was not measured in all patients, this analysis was limited to 91 patients with 20 events. The results indicated that NYHA functional class III or higher and increased left ventricular wall thickness were significantly associated with higher hazard ratios, whereas tafamidis treatment was associated with a significantly lower hazard ratio. Notably, neither NT-proBNP nor troponin T levels were retained in the final model.

Similarly, when the cardiac index was included as an explanatory variable, the analysis was limited to 72 patients with 14 events. In that model, NYHA functional class ≥III, increased left ventricular wall thickness, and elevated NT-proBNP levels were significantly associated with higher hazard ratios, while tafamidis treatment was associated with a significantly lower hazard ratio.

Taken together, these findings indicate that NYHA functional class, left ventricular wall thickness, NT-proBNP, and tafamidis treatment are the 4 major factors determining the survival status of this cohort.

Subgroup Analysis of Survival in Tafamidis-Treated and Untreated Patients

Figure 4 summarizes the survival effects of tafamidis across various patient subgroups. Tafamidis treatment was associated with significantly improved survival among patients with younger age (<79 years), lower NYHA functional class (II), lower NT-proBNP levels (<7,461 pg/mL), and lower troponin T levels (<0.082 ng/mL). In contrast, no significant survival benefit was observed in patients with older age, higher NYHA functional class, elevated NT-proBNP levels, and increased troponin T levels. Notably, tafamidis was associated with improved survival regardless of wall thickness and eGFR. These findings indicate that tafamidis treatment is significantly associated with better survival, particularly when initiated in the early stage of the disease.

Figure 4.

Survival curves free from all-cause death from the day of diagnosis, grouped by tafamidis-treated and untreated patients, stratified by: (A,B) age, (C,D) New York Heart Association (NYHA) functional class, (E,F) left ventricular wall thickness, (G,H) estimated glomerular filtration rate (eGFR), (I,J) N-terminal prohormone of brain natriuretic peptide (NT-proBNP), and (K,L) troponin T.

Correction of Survivor Selection Bias for Tafamidis-Treated Patients

Because tafamidis treatment was not initiated immediately after diagnosis, patients receiving tafamidis might have demonstrated longer survival simply because the treatment was primarily administered to those who had already survived longer. To address this potential survivor selection bias, a multivariate analysis for all-cause mortality was conducted, using the time (in years) from tafamidis initiation as the time-to-event variable.

The analysis revealed a strong interaction between tafamidis treatment and NYHA functional class (Table 3). Among patients with NYHA Class II, higher NT-proBNP levels, increased left ventricular wall thickness, and non-use of tafamidis were significant independent predictors of all-cause mortality. In contrast, none of the explanatory variables analyzed were significant in patients with NYHA Class ≥III.

Table 3.

Multivariate Analysis of Survival After the Tafamidis Initiation

Factor	HR	95% CI	P value
NYHA II
Wall thickness (mm)	1.753	1.262–2.433	0.0008
NT-proBNP (pg/mL)/1,000	1.097	1.028–1.170	0.0053
Tafamidis treated	0.082	0.013–0.504	0.0070
NYHA ≥III
Wall thickness (mm)	1.120	0.867–1.447	0.3972
NT-proBNP (pg/mL)/1,000	1.167	0.967–1.409	0.1067
Tafamidis treated	1.371	0.284–6.615	0.6947

Abbreviations as in Table 2.

As illustrated in Figure 5, tafamidis treatment was associated with improved prognosis in patients with NYHA Class II but not in those with NYHA Class ≥III. This survival benefit was evident regardless of left ventricular wall thickness. However, in patients with high NT-proBNP levels, tafamidis treatment was associated with poorer outcomes, and even among patients with low NT-proBNP levels, the beneficial effect was only marginal. This counterintuitive result may be due to overcorrection for survivor selection bias. Nonetheless, after adjusting for this bias, tafamidis treatment consistently predicted improved survival in patients aged <79 years, with NYHA Class II, preserved eGFR, and lower troponin T levels.

Figure 5.

Survival curves free from all-cause death from the day of tafamidis initiation, grouped by tafamidis-treated and untreated patients, stratified by: (A,B) age, (C,D) New York Heart Association (NYHA) functional class, (E,F) left ventricular wall thickness, (G,H) estimated glomerular filtration rate (eGFR), (I,J) N-terminal prohormone of brain natriuretic peptide (NT-proBNP), and (K,L) troponin T.

Discussion

This study provides valuable insights into the real-world effects of tafamidis on survival in patients with ATTR-CM treated at a community-based general hospital over the past decade. Notably, the number of ATTR-CM diagnoses increased dramatically following the approval of tafamidis in 2019. Multivariate analysis identified NYHA functional class, left ventricular wall thickness, NT-proBNP levels, and tafamidis treatment as independent predictors of survival. Tafamidis was particularly effective in patients with NYHA Class II, but not in those with NYHA Class ≥III. Moreover, the beneficial effect of tafamidis was marginal in older patients or in those with elevated NTproBNP or troponin T levels. These findings underscore the critical importance of early diagnosis and timely initiation of tafamidis to optimize clinical outcomes in ATTR-CM patients.

The observed survival benefit of tafamidis is consistent with findings from the ATTR-ACT study.³ However, our study extends this evidence to a real-world population. In our cohort, all patients treated with tafamidis were confirmed to have the wild-type genotype, in contrast to the ATTR-ACT study where 24% of patients carried a variant genotype. Compared with the ATTR-ACT cohort, patients in our study were older (80.6±8.0 vs. 74.5±7.2 years), included a higher proportion of females (23.1% vs. 9.8%), and exhibited less pronounced ventricular wall thickness (14.2±2.4 mm vs. 16.7±3.8 mm). The proportions of patients with NYHA Class ≥III (25.6% vs. 32.0%) and median NT-proBNP levels (2,642 pg/mL vs. 2,995.9 pg/mL) were comparable between the 2 cohorts. Consistent with the ATTR-ACT study, our results demonstrated that tafamidis provided a survival benefit, particularly in patients with NYHA Class II. In contrast, the benefit was only marginal in patients of older age and those with advanced disease, as reflected by higher NYHA functional class and elevated NT-proBNP or troponin T levels.

Patients who did not receive tafamidis were significantly older, had a higher proportion of females, and exhibited elevated NT-proBNP levels compared with those who received the treatment. This suggests that tafamidis was more frequently prescribed to relatively younger and healthier patients, whereas older and frailer individuals were less likely to receive tafamidis. Consequently, the better prognosis observed in tafamidis-treated patients compared with untreated patients is not entirely unexpected. To account for this potential selection bias, we conducted a multivariate analysis. The results identified NYHA functional class, left ventricular wall thickness, NT-proBNP levels, and tafamidis treatment as independent predictors of survival, whereas age at diagnosis and eGFR were not. Given that tafamidis treatment remained an independent predictor of survival in the multivariate analysis, its beneficial effects cannot be fully attributed to baseline differences in patient characteristics, such as age at diagnosis, NYHA functional class, left ventricular wall thickness, NT-proBNP levels, or eGFR. Furthermore, subgroup analysis indicated that the beneficial effects of tafamidis were predominantly observed in younger, healthier patients, but not in older, frailer individuals, as illustrated in Figure 4 and Figure 5. These findings underscore the therapeutic efficacy of tafamidis, indicating that its survival benefits are not solely attributable to baseline differences in patient characteristics between the treated and untreated groups.

Patients receiving tafamidis might have exhibited longer survival simply because the treatment was primarily administered to those who had already survived longer. Notably, the median time from ATTR-CM diagnosis to tafamidis initiation was 155 days. To address this potential survivor selection bias among tafamidis-treated patients, we conducted a survival analysis starting from the initiation of tafamidis. For tafamidis-untreated patients, survival was assessed from the time of diagnosis rather than randomization, which might have led to an underestimation of the survival benefits of tafamidis compared with randomized studies. Despite this limitation, the results consistently demonstrated that tafamidis treatment significantly improves survival, particularly in patients with NYHA Class II. These findings reinforce that tafamidis is associated with improved prognosis in patients with ATTR-CM in real-world clinical practice, in line with the outcomes reported in the ATTR-ACT study.

Several studies have proposed staging systems for ATTR-CM using baseline cut-off values of NT-proBNP (≥3,000 pg/mL), eGFR (<45 mL/min/1.73 m²), and troponin T (≥0.05 ng/mL).^6–8 In the present study, cut-off values for Kaplan-Meier analysis were determined using receiver-operating characteristic analysis, yielding 7,461 pg/mL for NT-proBNP, 42 mL/min/1.73 m² for eGFR, and 0.082 ng/mL for troponin T. However, even when applying the aforementioned established thresholds to our cohort instead, elevated NT-proBNP, reduced eGFR, and high troponin T remained significantly associated with worse survival outcomes, reinforcing the prognostic value of these staging systems. Furthermore, our observation that tafamidis exhibited only marginal beneficial effects in patients with elevated NT-proBNP, reduced eGFR, or high troponin T further emphasizes the importance of these biomarkers in assessing the disease severity in ATTR-CM.

It is noteworthy that left ventricular wall thickness emerged as an independent prognostic factor in the multivariate analysis, which contrasts with findings from previous studies. This discrepancy may be attributed to the substantially thinner median wall thickness in our cohort (14 mm) compared with other cohorts, such as Mayo Clinic (17 mm),⁶ National Amyloidosis Centre (17 mm),⁷ and Kumamoto University (15.5 mm).⁸ The earlier diagnosis of ATTR-CM in this study, reflected by the less pronounced left ventricular hypertrophy, may have heightened the prognostic significance of wall thickness. As illustrated in Figure 4E,F and Figure 5E,F, tafamidis treatment was associated with a better prognosis regardless of left ventricular wall thickness. However, patients with pronounced left ventricular hypertrophy had a worse prognosis even with tafamidis treatment. Therefore, early diagnosis before the development of marked left ventricular hypertrophy is crucial. This finding further supports the importance of early detection of ATTR-CM to maximize the survival benefits of tafamidis.

This study is not the only one to evaluate the real-world impact of tafamidis on survival in patients with ATTR-CM in Japan. Similar findings have been reported by Kochi University⁹ and Kumamoto University,¹⁰ where patients treated with tafamidis (n=38 and n=125, respectively) showed significantly better prognoses compared with untreated patients (n=44 and n=55, respectively). The proportion of patients receiving tafamidis treatment was 46% in the Kochi cohort, 69% in the Kumamoto cohort, and 64% in our cohort. Among patients treated with tafamidis, the mean age, percentage with NYHA functional class ≥III, and left ventricular wall thickness were 78.1 years, 8%, and 14.6 mm in the Kochi cohort; 75.6 years, 33%, and 16.1 mm in the Kumamoto cohort; and 77.8 years, 23%, and 14.4 mm in our cohort. Compared with the Kochi cohort and our cohort, the Kumamoto cohort appeared to include slightly younger patients with more severe disease, as reflected by a higher percentage of NYHA functional class ≥III and greater left ventricular wall thickness. Despite these differences, the survival curves of tafamidis-treated patients were consistently better than those of untreated patients across the 3 cohorts.

A key strength of the present study was the presentation of Kaplan-Meier survival curves across different patient subgroups. As shown in Figure 4 and Figure 5, these survival curves revealed that the beneficial effects of tafamidis were evident in patients aged <79 years and those with NYHA Class II, eGFR >42 mL/min/1.73 m², NT-proBNP <7,461 pg/mL, and troponin T <0.082 ng/mL. However, the effects were marginal in older patients and those with more advanced NYHA functional class, lower eGFR, elevated NT-proBNP, and increased troponin T levels. Although the cut-off values of these parameters were not absolute, and tafamidis might still have been effective in some patients aged >79 years, the results of the subgroup analysis provided a realistic representation of the patients who derived the maximum benefit from tafamidis treatment.

The present study predominantly included patients with the wild-type genotype. The likelihood of variant ATTR cases being present in our cohort appears to be very low, as all individuals treated with tafamidis (n=75) and all patients aged ≤72.4 years (n=21) were confirmed to have the wild-type genotype through genetic testing (n=83). This contrasts with previous studies that included a substantial proportion of variant ATTR cases. Some may find this observation unexpected, given that a decade ago, the variant form was thought to be more prevalent than the wild-type.¹¹ This discrepancy may, at least in part, reflect the aging population and increased diagnostic awareness following the approval of tafamidis. The availability of tafamidis has facilitated the early detection of ATTR-CM in older patients with heart failure, contributing to a marked rise in diagnoses, as observed in the present study. In contrast, because variant ATTR is caused by specific genetic mutations, its prevalence is unlikely to increase in a similar manner, even in an aging society. Additionally, patients with variant ATTR, who typically present with polyneuropathy at a younger age, are more likely to be referred to university hospitals with specialized neurology departments rather than to community cardiology centers such as ours. Given that most previous reports have originated from university hospitals with established amyloidosis centers, our findings from a community acute-care cardiology setting provide valuable insights, particularly reflecting the real-world clinical landscape of heart failure in an aging population. Considering that no ATTR-CM patients were diagnosed at Saiseikai Fukuoka General Hospital before 2015, many cases may remain undiagnosed unless general physicians become more aware of the widespread presence of the disease.

Study Limitations

This study has several limitations. First, being a single-center, retrospective analysis, it may not fully represent the variability of ATTR-CM management across different clinical settings. Additionally, the study did not evaluate quality of life or functional status improvements, which are critical endpoints in ATTR-CM treatment. Last, prospective multicenter studies are needed to confirm these findings and assess the long-term benefits of tafamidis, especially in the advanced stages of the disease.

Conclusions

Tafamidis provides significant survival benefits in real-world ATTR-CM patients, especially when started early in the disease course. Early detection and timely treatment initiation are crucial for improving outcomes in this increasingly recognized condition in everyday clinical practice.

Acknowledgments

During the preparation of this manuscript, the authors used ChatGPT, an AI language model developed by OpenAI, in order to assist with formatting and language refinement.

Sources of Funding

This study did not receive any specific funding.

Disclosures

T.K. has received honoraria for lectures from Pfizer Japan Inc. The other authors declare no conflicts of interest.

IRB Information

Ethics Committee of Saiseikai Fukuoka General Hospital (2025-2-1).

Data Availability

The deidentified participant data will not be shared.

References

• 1.   Kitaoka H, Izumi C, Izumiya Y, Inomata T, Ueda M, Kubo T, et al. JCS 2020 guideline on diagnosis and treatment of cardiac amyloidosis. Circ J 2020; 84: 1610–1671.
• 2.   Tsutsui Y, Kubota T, Kato S, Nozoe M, Suematsu N, Okabe M, et al. Utility of ^{99 m}Tc-pyrophosphate scintigraphy in diagnosing transthyretin cardiac amyloidosis in real-world practice. Circ Rep 2019; 1: 277–285.
• 3.   Maurer MS, Schwartz JH, Gundapaneni B, Elliott PM, Merlini G, Waddington-Cruz M, et al. Tafamidis treatment for patients with transthyretin amyloid cardiomyopathy. N Engl J Med 2018; 379: 1007–1016.
• 4.   Endo J, Sano M, Izumiya Y, Tsujita K, Nakamura K, Tahara N, et al. A statement on the appropriate administration of tafamidis in patients with transthyretin cardiac amyloidosis. Circ J 2019; 84: 15–17.
• 5.   Kanda Y. Investigation of the freely available easy-to-use software ‘EZR’ for medical statistics. Bone Marrow Transplant 2013; 48: 452–458.
• 6.   Grogan M, Scott CG, Kyle RA, Zeldenrust SR, Gertz MA, Lin G, et al. Natural history of wild-type transthyretin cardiac amyloidosis and risk stratification using a novel staging system. J Am Coll Cardiol 2016; 68: 1014–1020.
• 7.   Gillmore JD, Damy T, Fontana M, Hutchinson M, Lachmann HJ, Martinez-Naharro A, et al. A new staging system for cardiac transthyretin amyloidosis. Eur Heart J 2018; 39: 2799–2806.
• 8.   Nakashima N, Takashio S, Morioka M, Nishi M, Yamada T, Hirakawa K, et al. A simple staging system using biomarkers for wild-type transthyretin amyloid cardiomyopathy in Japan. ESC Heart Fail 2022; 9: 1731–1739.
• 9.   Ochi Y, Kubo T, Baba Y, Sugiura K, Miyagawa K, Noguchi T, et al. Early experience of tafamidis treatment in Japanese patients with wild-type transthyretin cardiac amyloidosis from the Kochi amyloidosis cohort. Circ J 2022; 86: 1121–1128.
• 10.   Takashio S, Morioka M, Ishii M, Morikawa K, Hirakawa K, Hanatani S, et al. Clinical characteristics, outcome, and therapeutic effect of tafamidis in wild-type transthyretin amyloid cardiomyopathy. ESC Heart Fail 2023; 10: 2319–2329.
• 11.   Ministry of Health, Labour and Welfare, Japan. National survey on amyloidosis in Japan (2016 Health and Labour Sciences Research Grant). https://mhlw-grants.niph.go.jp/system/files/2016/162051/201610043B_upload/201610043B0010.pdf (in Japanese).