Abstract
Background
Prognostic staging systems in transthyretin amyloid (ATTR) cardiomyopathy were developed before the introduction of disease‐modifying treatment. We therefore evaluated contemporary risk stratification according to different staging systems in patients with ATTR cardiomyopathy.
Methods
All patients with newly diagnosed ATTR cardiomyopathy and baseline biomarker measurements from 2019 to 2023 were included from Mayo Clinic, Rochester, MN. Prognostic outcomes were calculated with Kaplan–Meier analysis for 3e staging systems in the main analysis: the National Amyloidosis Centre (NAC) staging system; a modified Mayo staging system using hsTNT (high‐sensitivity troponin T) cutoff >65 ng/L; the extended NAC model with a fourth stage (NT‐proBNP [N‐terminal pro‐B‐type natriuretic peptide] >10 000 ng/L).
Results
The study population consisted of 441 patients with a median age of 76 years, 93% were men, 85% were initiated on tafamidis, and 94% had wild‐type ATTR. Median follow‐up was 3.4 years, and 109 deaths occurred during follow‐up. The 3‐year survival for the Mayo stages were 92.3% (88.7–95.9) for stage I, 71.6% (63.0–81.3) for stage II, and 43.5% (32.4–58.5) for stage III. For the NAC system, 3‐year survival were 92.0% (89.0–95.7), 67.5% (59.0–77.3), and 47.0% (34.7–63.5) for stages I, II, and III, respectively.
The 3‐year survival by the extended NAC system was 92.0 (69.0–79.2), 69.0% (60.1–79.2), 53.2% (39.9–70.8) and 34.4% (17.4–68.0) for stages I, II, III, and IV, respectively.
Conclusions
In a contemporary cohort of patients with ATTR cardiomyopathy, the NAC and modified Mayo staging systems remained equally effective in delineating clinically meaningful prognostic groups, whereas the extended NAC system identified a stage IV subgroup with high mortality.
Keywords: amyloidosis, prognosis, staging, transthyretin
Subject Categories: Cardio-Oncology, Cardiomyopathy, Cardiorenal Syndrome, Heart Failure
Nonstandard Abbreviations and Acronyms
- ATTR
transthyretin amyloidosis
- ATTR‐CM
transthyretin amyloidosis cardiomyopathy
- ATTRwt‐CM
wild‐type transthyretin amyloidosis cardiomyopathy
- hsTNT
high‐sensitivity troponin T
- NAC
National Amyloidosis Centre
- NYHA
New York Heart Association
Clinical Perspective.
What Is New?
In a contemporary transthyretin amyloid cardiomyopathy cohort with, both the National Amyloidosis Centre and modified Mayo staging system stratified patients into clinically meaningful prognostic groups, and survival was similar for patients classified into the same numerical stage across systems, indicating comparable prognostic performance.
The extended National Amyloidosis Centre model further identified a stage IV subgroup with markedly increased mortality.
What Are the Clinical Implications?
The established staging systems remain highly reliable, discriminatory prognostic tools among contemporary predominantly tafamidis‐treated patients with transthyretin amyloid cardiomyopathy.
Transthyretin amyloidosis cardiomyopathy (ATTR‐CM) affects an estimated 150 000 individuals in the United States as of 2025. 1 Accurate and clinically applicable staging systems are essential for guiding both prognostication and clinical management of patients with ATTR‐CM. 2 , 3 , 4
Two of the most established prognostic models are the National Amyloidosis Centre (NAC) staging system and the Mayo staging system. 5 , 6 The NAC system has seen broader adoption in ATTR‐CM and has been considered to provide modestly superior prognostic discrimination in ATTR‐CM. 7 However, this has been challenged in a study concluding that a modified Mayo staging system may offer superior prognostic accuracy compared with the NAC system in contemporary patients with wild‐type ATTR‐CM (ATTRwt‐CM). 8 Moreover, an extended version of the NAC system has recently been proposed, incorporating a fourth stage defined by NT‐proBNP (N‐terminal pro‐B‐type natriuretic peptide) >10 000 ng/L, which was shown further to identify a subgroup with particularly high 1‐year mortality. 9 Nonetheless, all these staging systems were developed in disease‐modifying treatment‐naïve patients. Collectively, these observations highlight the need to reevaluate the prognostic implications of established staging systems in the context of modern therapeutic strategies that likely have improved survival in ATTR‐CM. 10
In this study, we aimed to characterize and compare prognostic stratification in the era when tafamidis was available for routine clinical use according to (1) the NAC system, (2) a modified Mayo system, (3) the extended NAC system, and (4) an exploratory 3‐biomarker classification that combines the NAC and the modified Mayo system.
METHODS
Study Population and Design
We conducted a retrospective study of all patients diagnosed with ATTR‐CM at Mayo Clinic in Rochester, Minnesota, or referred there within 90 days of diagnosis between January 2019 and December 2023.
Patients were identified from an institutional database of all patients with ATTR‐CM at Mayo Clinic. All patients were manually verified to meet current diagnostic criteria for ATTR‐CM. 11 Patients without biomarker measurements within ±120 days of the diagnosis date and those who had undergone heart transplantation before ATTR‐CM diagnosis were excluded.
During the study period, 705 patients were identified in the institutional database with ATTR. After medical record review, the diagnosis of ATTR‐CM was confirmed in 568 patients where 441 had the complete biomarker data within ±120 days of diagnosis and were included in the final study population.
Ethical approval was obtained from the Mayo Clinic Institutional Review Board. The study was classified as minimal risk and conducted under a waiver of informed consent, applicable to patients who had authorized the use of their medical records for research. The data that support the findings of this study are available from the corresponding author upon reasonable request.
Data Sources
Baseline demographics, comorbidities, clinical status, biomarkers, ECG, and echocardiographic data were electronically extracted from the medical records on the date closest to the date of diagnosis. Comorbidities were electronically collected at baseline using International Classification of Diseases, Ninth Revision (ICD‐9) or Tenth Revision (ICD‐10) codes. Initiation of tafamidis was confirmed by medical record review. Date of tafamidis initiation was defined either as the day of insurance approval or the prescription date in the medical record.
Staging Systems and Biomarkers
The NAC staging system and the Mayo staging system were applied according to the previously defined cut‐points, except that the recent harmonized hsTNT (high‐sensitivity troponin T) cutoff of 65 ng/L was used. 12
The definitions and thresholds for each applied staging system were as follows:
NAC staging system: NT‐proBNP (cut‐point >3000 ng/L), estimated glomerular filtration rate eGFR (cut‐point <45 mL/minute per 1.73 m2).
Modified Mayo staging system: NT‐proBNP (cut‐point >3000 ng/L), hsTNT (cut‐point 65>ng/L).
Extended NAC staging system (extNAC): A fourth stage was added to the original NAC staging system for patients who had a NT‐proBNP of >10 000 ng/L regardless of eGFR level. 9
Combined NAC–Mayo staging system (NAC–Mayo): One point was assigned for each abnormal biomarker, defined as NT‐proBNP >3000 ng/L, eGFR <45 mL/minute per 1.73 m2, or hsTNT >65 ng/L. This resulted in 4 stages: stage 1 (0 points), stage 2 (1 point), stage 3 (2 points), and stage 4 (3 points).
The biomarkers were measured using the following methods: hsTnT with the Elecsys Troponin T Gen 5 assay (Roche Diagnostics) and NT‐proBNP with Roche Diagnostics, version 12.0. eGFR was calculated using the Modification of Diet in Renal Disease formula, 13 based on serum creatinine by the Creatinine Plus version 2 assay (Roche Diagnostics). All biomarker measurements remained consistent throughout the study period.
Outcome
All‐cause mortality was the primary outcome. Mortality data were extracted from the electronic medical record, or from Accurint database if vital status was not confirmed in 2025 and the individual was not known to be deceased previously. 14
Statistical Analysis
Continuous data were assessed for normality using visual inspection of histograms and Q‐Q plots. Continuous variables were expressed as means±SDs for normally distributed data, or medians and interquartile range (IQR) for skewed distributions.
The agreement between the modified Mayo and the NAC staging system was evaluated using weighted Cohen's kappa statistics with a 95% CI.
Survival probabilities were estimated with Kaplan–Meier estimates, and log‐rank were used to test difference in survival probabilities. Survival curves were restricted to a 3‐year follow‐up to ensure a reasonable number of patients at risk at each time point. Kaplan–Meier estimates of survival probability were calculated at 1‐, 2‐, and 3‐years for all staging systems and presented by stage. Unadjusted and age‐adjusted hazard ratios (HRs) of all‐cause mortality between stages were estimated using Cox proportional hazards regression. Proportional hazards assumptions were evaluated using Schoenfeld residuals and found acceptable. The model concordance index was calculated and reported as a measure of each staging system's discriminative ability. 15
We performed exploratory age‐stratified analyses of prognostic outcomes for the Mayo and NAC staging systems, dividing the cohort at 75 years (approximately the median age), as age showed significant interaction with mortality for both staging systems. We performed an ATTRwt‐subgroup analysis of the prognostic outcomes for the Mayo and NAC staging systems.
P values <0.05 were considered statistically significant. The statistical analysis was performed with BlueSky Statistics software v. 10.3 (BlueSky Statistics LLC, Chicago, IL, USA).
RESULTS
Characterization of the Study Population
The baseline characteristics of the study population are displayed in Table 1. The mean age was 76.3 years, and 93% were men and 94% were of White descent. The median level of the 3 biomarkers was hsTNT 48 ng/L (IQR, 33–69), eGFR 57 mL/minute per 1.73 kg2 (IQR, 45–69), and NT‐proBNP 1801 ng/L (IQR, 942–3520). The interval between diagnosis and biomarker measurement was −8.3±26.7 days for eGFR, −11.1±30.6 days for hsTNT, and −9.4±29.8 days for NT‐proBNP, indicating that most measurements were obtained shortly before the diagnostic date. The majority of the cohort had ATTRwt‐CM (94%), and 6% had variant ATTR‐CM. The Val122I was the most frequent mutation, accounting for one third of patients with variant ATTR‐CM. Tafamidis was initiated in 85% of patients, with a median time to initiation of 29 days from diagnosis (IQR, 6–71 days). The proportion of patients who initiated tafamidis during follow‐up according to the Mayo staging system was 89.1% for stage I, 79.5% for stage II, and 79.5% for stage III. For the NAC staging system, the corresponding proportions were 89.1%, 78.6%, and 81.4%, respectively. The mean time to first visit at Mayo Clinic was 9±22 days from the time of diagnosis in the study population.
Table 1.
General Characteristics of the Study Population by ATTR‐CM Type
| Total (n=441) | ATTRv‐CM (n=27) | ATTRwt‐CM (n=414) | P value | |
|---|---|---|---|---|
| Age, y | 76.3 (8.3) | 65.1 (10.3) | 77.0 (7.6) | <0.01 |
| Male sex | 411 (93.2%) | 18 (66.7%) | 393 (94.9%) | <0.01 |
| White | 413 (93.7%) | 13 (48.1%) | 400 (96.6%) | <0.01 |
| Body mass index, kg/m2 | 28.3 (25.3 to 31.1) | 25.8 (23.6 to 29.2) | 28.4 (25.4 to 31.2) | 0.11 |
| Systolic blood pressure, mm Hg | 124.8 (19.3) | 121.1 (14.1) | 125.1 (19.6) | 0.31 |
| Atrial fibrillation | 209 (47.7%) | 5 (18.5%) | 204 (49.6%) | <0.01 |
| Hypertension | 146 (33.3%) | 2 (7.4%) | 144 (35.0%) | 0.03 |
| Severe left‐sided valve disease or prosthetic valve | 28 (6.4%) | 0 (0.0%) | 28 (6.8%) | 0.16 |
| Diabetes | 66 (15.1%) | 1 (3.7%) | 65 (15.8%) | 0.09 |
| Pacemaker | 69 (15.8%) | 4 (14.8%) | 65 (15.8%) | 0.89 |
| Implantable cardiac defibrillator | 30 (6.8%) | 2 (7.4%) | 28 (6.8%) | 0.91 |
| Biomarkers | ||||
| N‐terminal pro‐brain natriuretic peptide, ng/L | 1801 (942 to 3530) | 1603 (410 to 2671) | 1826 (971 to 3612) | 0.04 |
| Estimated glomerular filtration rate, mL/min per 1.732 | 57.1 (45.0 to 69.1) | 62.0 (54.5 to 83.5) | 56.6 (44.6 to 68.8) | 0.01 |
| High‐sensitivity troponin T, ng/L | 48.0 (33.0 to 69.0) | 46.0 (20.5 to 55.5) | 48.0 (33.0 to 69.0) | 0.07 |
| National Amyloidosis Centre stage, I/II/III, % | 58.0/28.6/13.3 | 74.1/25.9/0 | 57.0/28.7/14.3 | 0.08 |
| Mayo stage, I/II/III, % | 58.0/25.4/16.6 | 81.5/0/18.5 | 56.5/27.1/16.4 | 0.01 |
| Echocardiographic measures | ||||
| LV ejection fraction, % | 54.0 (45.0 to 60.0) | 55.0 (36.0 to 61.0) | 54.0 (45.0 to 60.0) | 0.25 |
| Global longitudinal strain, % | −12.0 (−14.0 to −9.0) | −11.5 (−15.8 to −9.0) | −12.0 (−14.0 to −10.0) | 0.52 |
| Interventricular septum diameter, mm | 15.8 (3.3) | 15.6 (3.2) | 15.8 (3.3) | 0.80 |
| LV mass index, g/m2 | 150.1 (42.3) | 148.8 (43.5) | 150.2 (42.2) | 0.88 |
| Left atrial volume index, mL/m2 | 48.0 (42.0 to 57.0) | 44.0 (37.5 to 50.0) | 48.0 (42.0 to 58.0) | <0.01 |
| ECG measures | ||||
| PR interval, ms | 193.0 (170.0 to 220.5) | 167.0 (148.0 to 186.0) | 196.0 (173.5 to 227.0) | <0.01 |
| QRS interval, ms | 106.0 (92.0 to 138.0) | 94.0 (84.0 to 120.0) | 107.0 (92.0 to 138.5) | 0.05 |
| Functional test | ||||
| 6‐minute walk test, m | 393.2 (309.3 to 475.5) | 498.3 (400.0 to 539.4) | 390.1 (304.8 to 469.4) | 0.01 |
| Disease‐modifying treatment initiation follow‐up | ||||
| Tafamidis | 375 (85.0%) | 21 (77.8%) | 354 (85.5%) | 0.28 |
Values are mean±SD, median (interquartile range) or n (%).
ATTR‐CM indicates transthyretin amyloidosis cardiomyopathy; ATTRv‐CM, variant transthyretin amyloidosis cardiomyopathy; and ATTRwt‐CM, wild‐type transthyretin amyloidosis cardiomyopathy.
Patients with variant ATTR‐CM were younger and had lower disease burden than those with ATTRwt cardiomyopathy.
The median follow‐up time in the study population was 3.4 years (IQR, 2.3–4.7). A total of 109 (24.7%) patients died during follow‐up in the total study population. The overall 3‐year survival in the study population was 78.7% (95% CI, 74.7–83.1).
ATTR‐CM Survival by the NAC and Mayo System Stages
For the NAC and the Mayo staging systems, the overall agreement in stage classification was high with a kappa statistic of 0.76 (95% CI, 0.72–0.80) (Table 2).
Table 2.
Agreement Across the Modified Mayo System Stages and NAC System Stage by Cohen's Kappa Statistics
| NAC stage | |||||
|---|---|---|---|---|---|
| 1 | 2 | 3 | Total | ||
| Mayo stage | 1 | 223 | 33 | 0 | 256 |
| 2 | 33 | 61 | 18 | 112 | |
| 3 | 0 | 32 | 41 | 73 | |
| Total | 256 | 126 | 59 | 441 | |
Weighted kappa=0.76 (95% CI, 0.72–0.80). NAC indicates National Amyloidosis Centre.
The 3‐year survival for the Mayo staging system was 92.3% (95% CI, 88.7–95.9) for stage I, 71.6% (95% CI, 63.0–81.3) for stage II, and 43.5% (95% CI, 32.4–58.5) for stage III (Figure 1A; Table S1). For the NAC staging system, the 3‐year survival was 92.0% (95% CI, 89.0–95.7) for stage I, 67.5% (95% CI, 59.0–77.3) for stage II, and 47.0% (95% CI, 34.7–63.5) for stage III (Figure 1B; Table S1). Using disease stage I as the reference, the age‐adjusted HR of all‐cause mortality for the modified Mayo staging system was 2.43 (95% CI, 1.50–3.93; P≤0.01) for stage II and 5.34 (95% CI, 3.24–8.80; P≤0.01) for stage III (Figure 2; Table S1). For the NAC staging system, the age‐adjusted HR was 3.00 (95% CI, 1.87–4.82; P≤0.01) for stage II and 4.12 (95% CI, 2.40–7.05; P≤0.01) for stage III. The model concordance was 0.77 for the age‐adjusted Mayo staging model and 0.76 for the age‐adjusted NAC staging system model (Table S1).
Figure 1. ATTR‐CM survival by staging systems.
A, ATTR‐CM survival by stages of the modified Mayo staging system Kaplan–Meier curves showing overall survival probabilities by the modified Mayo staging system stages for all patients with ATTR‐CM. B, ATTR‐CM survival by stages of the modified NAC staging system. Kaplan–Meier curves showing overall survival probabilities by National Amyloidosis Center stages in all patients with ATTR‐CM. ATTR‐CM indicates transthyretin amyloidosis cardiomyopathy; and NAC, National Amyloidosis Centre.
Figure 2. Age‐adjusted all‐cause mortality by staging systems in ATTR‐CM.
Adjusted hazard ratios are presented with 95% CIs. extNAC represents the extended NAC staging system with a fourth stage of NT‐proBNP >10 000 ng/L. Stage 1=no abnormal biomarkers, stage 2=1 abnormal biomarkers, stage 3=2 abnormal biomarkers, stage 4=NT‐proBNP >10 000 ng/L. NAC–Mayo represents the combined 3‐biomarker model using the traditional cut‐points for all 3 biomarkers. Stage 1=no biomarkers abnormal, stage 2=1 abnormal biomarker, stage 3=2 abnormal biomarkers, stage 3=all 3 biomarkers abnormal. aHR indicates age‐adjusted hazard ratio; ATTR‐CM, transthyretin amyloidosis cardiomyopathy; extNAC, extended National Amyloidosis Centre staging system; NAC, National Amyloidosis Centre; and NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide.
The age‐stratified prognostic outcomes according to the NAC and the modified Mayo staging system are presented in Table 3. Most patients aged ≤75 years were classified as stage I in both staging systems (72%–73%) and exhibited excellent 3‐year survival, ranging from 96% to 97%. Among patients aged ≤75 years, Mayo stage II was not associated with an increased risk of all‐cause mortality compared with Mayo stage I (HR, 1.31 [95% CI, 0.42–4.06]; P=0.64). Among patients aged >75 years, survival estimates were largely consistent between the modified Mayo and NAC staging systems; nonetheless, the HRs across stages and the model concordance were slightly higher with the modified Mayo staging system, but wide CIs limited the estimates.
Table 3.
Stratified Analysis of Survival and All‐Cause Mortality by Staging System
| Strata | Modified Mayo staging system | NAC staging system | ||||||
|---|---|---|---|---|---|---|---|---|
| No. | 3‐y survival, % | HR | P value | n | 3‐y survival, % | HR | P value | |
| Age≤75 y | ||||||||
| Stage 1 | 145 | 96.3 (92.5–99.5) | Reference | Reference | 147 | 97.5 (94.2–100.0) | Reference | Reference |
| Stage 2 | 35 | 94.3 (86.2–100.0) | 1.31 (0.42–4.06) | 0.64 | 43 | 83.4 (71.8–96.7) | 7.55 (3.02–18.87) | <0.01 |
| Stage 3 | 21 | 57.0 (38.0–85.4) | 11.12 (4.37–28.31) | <0.01 | 11 | 49.9 (26.2–94.9) | 17.60 (5.56–55.75) | <0.01 |
| Model concordance: 0.70 (0.59–0.81) | Model concordance: 0.76 (0.65–0.86) | |||||||
| Age>75 y | ||||||||
| Stage 1 | 111 | 87.2 (80.6–94.4) | Reference | Reference | 109 | 84.8 (77.7–92.6) | Reference | Reference |
| Stage 2 | 77 | 61.2 (50.4–74.5) | 3.04 (1.74–5.31) | <0.01 | 83 | 59.4 (48.7–72.5) | 2.42 (1.43–4.09) | <0.01 |
| Stage 3 | 52 | 37.8 (25.2–56.7) | 5.18 (2.88–9.29) | <0.01 | 48 | 46.1 (32.7–65.0) | 3.35 (1.89–5.93) | <0.01 |
| Model concordance: 0.69 (0.63–0.75) | Model concordance: 0.64 (0.59–0.70) | |||||||
Hazard ratios are presented with 95% CI. HR indicates hazard ratio; and NAC, National Amyloidosis Centre.
ATTR‐CM Survival by the Extended NAC System Stages
In the extended NAC staging system (ie, extNAC), 256 patients were stage I, 115 were stage II, 50 were stage III, and 20 were stage IV.
The survival probabilities for each stage are visualized in Figure 3 and summarized in Table S2. The 3‐year survival by the extNAC was 92.0 (95% CI,88.5–95.7) for stage I, 69.0% (95% CI, 60.1–79.2) for stage II, and 53.2% (95% CI, 39.9–70.8) in stage III, and 34.4% (95% CI, 17.4–68.0) for stage IV. The age‐adjusted HR of all‐cause mortality for the extNAC was 2.86 (95% CI, 1.77–4.63; P≤0.01) for stage II versus I, 3.51 (95% CI, 1.98–6.22; P≤0.01) for stage III versus I, and 8.54 (95% CI, 4.27–17.06; P≤0.01) for stage IV versus I. The model concordance for the age‐adjusted model was 0.77.
Figure 3. ATTR‐CM survival by stages of the extended NAC staging system.
Kaplan–Meier curves showing overall survival probability for the extended NAC stages for all patients with ATTR‐CM. extNAC represents the extended NAC model, where a fourth stage is introduced to the traditional system for patients with NT‐proBNP >10 000 ng/L. ATTR‐CM indicates transthyretin amyloidosis cardiomyopathy; extNAC, extended National Amyloidosis Centre staging system;NAC, National Amyloidosis Centre; and NT‐proBNP, N‐terminal pro‐B‐type natriuretic peptide.
ATTR‐CM Survival by the Combined NAC–Mayo Staging System
In the combined NAC–Mayo staging system (ie, NAC–Mayo), 223 patients were in stage I, 109 in stage II, 68 in stage III, and 41 in stage IV.
The survival probabilities for each stage are visualized in Figure 4 and summarized in Table S3. The 3‐year survival with the NAC–Mayo were 93.6% (95% CI,90.1–97.2) for stage I, 75.5% (95% CI,67.1–85.0) for stage II, 59.5% (95% CI,48.1–73.5) for stage III, and 42.3% (95% CI,28.3–63.1) for stage IV. The age‐adjusted HR of all‐cause mortality for each stage of the NAC–Mayo was 3.16 (95% CI, 1.84–5.42, P≤0.01) for stage II versus I, 4.16 (95% CI, 2.36–7.34, P≤0.01) for stage III versus I, and 5.79 (95% CI, 3.09–10.86, P≤0.01) for stage IV versus I. The model concordance for the age‐adjusted model was 0.77.
Figure 4. ATTR‐CM survival by stages of the exploratory NAC–Mayo staging system.
Kaplan–Meier curves showing overall survival probability between the combined NAC–Mayo staging system for all patients with ATTR‐CM. NAC–Mayo represent the 4‐stage 3‐biomarker model using both high‐sensitivity troponin T, N‐terminal natriuretic peptide and estimated glomerular filtration rate with traditional cut‐points for abnormality. Stage 1=no abnormal biomarker, stage 2=1 abnormal biomarker, stage 3=2 abnormal biomarkers, stage 4=all three biomarkers abnormal. ATTR‐CM indicates transthyretin amyloidosis cardiomyopathy; and NAC, National Amyloidosis Centre.
ATTRwt Subgroup Survival by the NAC and Mayo System Stages
The median follow‐up of the ATTRwt‐CM subgroup was 3.43 years (IQR, 2.32–4.73), in which 105 all‐cause deaths were observed. Survival and mortality risk estimates for the ATTRwt subgroup were comparable to those observed for the overall cohort (visualized in Figure 5A and 5B and summarized in Table S4).
Figure 5. ATTRwt survival by staging systems.
A, ATTRwt survival by stages of the Mayo staging system. Kaplan–Meier curves showing overall survival probability between Mayo staging system stages for all patients with ATTRwt. B, ATTRwt survival by stages of the NAC staging system. Kaplan–Meier curves showing overall survival probability between NAC staging system stages for all patients with ATTRwt. ATTRwt‐CM indicates wild‐type transthyretin amyloidosis cardiomyopathy; and NAC, National Amyloidosis Centre.
DISCUSSION
In this study of patients with ATTR‐CM diagnosed in the contemporary era, with the availability of specific disease‐modifying therapy, we found that both the NAC and the modified Mayo staging systems remained effective for prognostic stratification. We also found that they were similar effective at stratifying patients into clinically relevant prognostic groups, indicating that the 2 staging systems offer comparable clinical utility for prognostic stratification. In addition, we observed a tendency suggesting that the NAC system provided somewhat better risk discrimination across stages among patients aged ≤75 years, whereas the modified Mayo staging system appeared to offer marginally better risk stratification among patients >75 years. Finally, we confirmed that the extNAC system identified a subgroup at particularly high risk (stage IV) among contemporary patients, and that the exploratory combination of the NAC and Mayo staging models could suggest that incorporating all three biomarkers (NT‐proBNP, hsTNT, and eGFR) might provide some additional prognostic granularity within stages compared with the established staging systems. Nonetheless, no clear separation in survival between the stages was observed, and the clinical significance of the system's increased complexity remains uncertain.
Previous ATTR‐CM Studies on Prognostic Biomarkers
The original NAC staging system was proposed in 2017 in a cohort with available measurements of TnT/hsTnT, eGFR, and NT‐proBNP in a treatment‐naïve patient population. 5 The study showed comparable prognostic performance between systems combining NT‐proBNP with either hsTnT or eGFR but favored the NT‐proBNP and eGFR combination due to better standardization and broader clinical availability of eGFR. In our contemporary cohort of patients, we found that these 2 staging systems continue to provide similar and effective risk stratification across all stages, despite the introduction of disease‐modifying treatment. This finding contrasts with a recent study that suggested that only the modified Mayo staging system, and not the NAC staging system, distinguished mortality risk between low‐ and intermediate‐risk patients with ATTR‐CM receiving disease‐modifying treatment. 8 However, the sample size and duration of follow‐up in the study were modest, which may have reduced the statistical power to detect differences between groups. Also, a recent multicenter study of patients initiated on tafamidis demonstrated clear risk separation across NAC stages, with the 3‐year survival approximately inferred from the published Kaplan–Meier curves as 80% for stage I, 55% for stage II, and 25% for stage III, supporting that the NAC staging system remains an effective prognostic tool in contemporary patients, and challenging the conclusion of the previous study. 16 In the multicenter study, the Mayo staging system was not applied, which precluded direct comparison between the NAC and Mayo staging systems. 16 However, collectively, these findings seem to suggest that despite recent advances in the therapeutic management of patients with ATTR‐CM, mortality risk remains substantial and that this risk is effectively stratified into clearly delineated, clinically relevant groups by both the original NAC and the modified Mayo staging systems.
Age is an important prognostic factor in ATTR‐CM, and it was also observed to affect the risk discrimination between stages among the NAC and the modified Mayo staging system. 17 The NAC staging system seemed to provide slightly better risk stratification in younger patients, whereas the modified Mayo system appeared to provide marginally better stratification in older individuals. One possible explanation for any age‐related differences among these staging systems could be that serum creatinine‐based eGFR estimates can be prone to inaccuracies in older, frail patients due to the influence of nonrenal factors, such as reduced muscle mass, lower physical activity, and dietary variations, on eGFR. 18 , 19 This mechanism could be another potential factor explaining the prior study's observations that only the modified Mayo staging system reliably distinguished between low‐ and intermediate‐risk patients with ATTRwt. That study cohort included patients with a mean age of 82 years, and the modified Mayo staging system might have performed slightly better in an older patient cohort as a result. However, another study with treatment‐naïve patients with ATTRwt with similar mean age reported superior risk stratification using NT‐proBNP and eGFR compared with NT‐proBNP and troponin I (≥0.11 ng/mL), and therefore whether the prognostic performance of eGFR attenuates with higher age in ATTR‐CM is uncertain.
The extended NAC staging system was IId to identify a subgroup of patients with particularly high 1‐year mortality. 9 In our cohort, only a few patients were classified as stage IV (n=20); nevertheless, we observed a 1‐year survival of 70% for those classified as stage IV, which is consistent with the estimate that was reported for the same stage group in the original cohort. Notably, in that study, tafamidis‐treated patients were not included, whereas ∼50% of stage IV patients in our cohort initiated tafamidis therapy. Though tafamidis is not expected to change 1‐year survival, we also found similar 3‐year survival estimates between the 2 studies, which might suggest that tafamidis treatment has not substantially affected survival in this group. However, caution is warranted when comparing survival across different time periods and patient cohorts. Nonetheless, in the ATTR‐ACT (Safety and Efficacy of Tafamidis in Patients With Transthyretin Cardiomyopathy) study, the mortality benefit of tafamidis was most pronounced and most certain in patients with less advanced disease, and separation of the survival curves emerged only during the second year of follow‐up, which might explain the comparable survival observed in this patient group. 10
The combined NAC and Mayo staging model suggested that incorporating all 3 biomarkers (NT‐proBNP, hsTNT, and eGFR) could provide some additional prognostic granularity within stages. Nevertheless, the differences were subtle, and an important strength of the original NAC staging system is its simplicity, which has supported widespread clinical adoption. Therefore, whether a potential for further prognostic granularity from a more complex 3‐biomarker system outweighs its potential drawbacks remains uncertain. However, in settings where routine measurement of all 3 biomarkers is available, future studies could explore whether risk prediction with all 3 biomarkers could improve risk stratification compared with current models. Ultimately, improving prognostication in ATTR‐CM remains critical, especially given the high cost of specific disease‐modifying agents.
Limitations
The study was a single‐center, retrospective study with potential selection bias.
Approximately 20% of those with ATTR‐CM did not undergo a complete biomarker assessment and were excluded, which may have introduced bias to our estimates if those excluded differed systematically from the study population.
Another limitation was that biomarkers could be measured within 120 days of the diagnosis date. This was permitted as Mayo Clinic is also a referral center, and therefore, delays occur between diagnosis and the first visit. Most patients in the study were diagnosed locally, and the time to biomarker measurement from diagnosis was short in the majority. However, potential differences in biomarker values between the time of diagnosis and the time of measurement cannot be ruled out.
Our cohort consisted predominantly of older male patients with ATTRwt‐CM, reflecting the typical disease distribution in clinical practice but limiting generalizability to younger patients, women, and those with variant ATTR‐CM. Accordingly, the findings should be interpreted with caution when extrapolated to these subgroups.
Because tafamidis treatment was initiated after diagnosis, and we lacked information on treatment adherence and withdrawal, we did not adjust or stratify our results for tafamidis treatment, as this could introduce bias to our estimates.
Our sample size limited the precision of the 3‐year survival estimates for the 4‐stage models, particularly for stage IV, resulting in wide CIs.
Data on supportive heart failure therapy, diuretic dose, and New York Heart Association class were not available, precluding assessment of staging system performance using these clinically relevant parameters and stratification by these important clinical variables. 20
Finally, we did not perform external validation, calibration, or reclassification analyses, which limits formal assessment of contemporary model performance.
CONCLUSIONS
In a contemporary cohort of patients with ATTR‐CM, including patients receiving specific disease‐modifying treatment, the NAC and modified Mayo staging systems remained equally effective in delineating clinically meaningful prognostic groups, whereas the extended NAC system identified a stage IV subgroup with high mortality.
Sources of Funding
The study was supported by a grant from the “Fonden af Laegevidenskabens Fremme,” “Reinholdt Jorcks Fond,” “Aarhus University Travel Grant,” and a scholarship grant by Aarhus University. The funding had no influence on the study.
Disclosures
Dr Ezzeddine has received research grants from Pfizer Inc; consulting fees from Alnylam Pharmaceuticals and Medscape from WebMD. Dr Dispenzieri has received research support from Celgene, Takeda, Janssen, Pfizer, and Alnylam; and has received consulting fees from Intellia Therapeutics and Akcea Therapeutics. Dr Poulsen has received consultancy fees from BridgeBio. Dr Grogan has received research support from Alnylam, Eidos Therapeutics, Pfizer, and Prothena; and has received consulting fees and honoraria (paid to Mayo Clinic) from Akcea Therapeutics, Alnylam, Eidos Therapeutics, Pfizer, and Prothena. The remaining authors have no disclosures to report.
Supporting information
Tables S1–S4
STROBE statement
Acknowledgments
The authors used ChatGPT and Grammarly to improve sentence structure, grammar, and wording during article preparation. The authors have critically reviewed and validated all content and take full responsibility for its accuracy and integrity.
This article was sent to Timothy C. Wong, MD, MS, Associate Editor, for review by expert referees, editorial decision, and final disposition.
Supplemental Material is available at https://www.ahajournals.org/doi/suppl/10.1161/JAHA.125.048585
For Sources of Funding and Disclosures, see page 11.
REFERENCES
- 1. Ruberg FL, Maurer MS. Cardiac amyloidosis due to transthyretin protein: a review. JAMA. 2024;331:778–791. doi: 10.1001/jama.2024.0442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Garcia‐Pavia P, Rapezzi C, Adler Y, Arad M, Basso C, Brucato A, Burazor I, Caforio ALP, Damy T, Eriksson U, et al. Diagnosis and treatment of cardiac amyloidosis: a position statement of the ESC Working Group on myocardial and pericardial diseases. Eur Heart J. 2021;42:1554–1568. doi: 10.1093/eurheartj/ehab072 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Pregenzer‐Wenzler A, Abraham J, Barrell K, Kovacsovics T, Nativi‐Nicolau J. Utility of biomarkers in cardiac amyloidosis. JACC Heart Fail. 2020;8:701–711. doi: 10.1016/j.jchf.2020.03.007 [DOI] [PubMed] [Google Scholar]
- 4. Feng KY, Loungani RS, Rao VN, Patel CB, Khouri MG, Felker GM, DeVore AD. Best practices for prognostic evaluation of a patient with transthyretin amyloid cardiomyopathy. JACC CardioOncol. 2019;1:273–279. doi: 10.1016/j.jaccao.2019.11.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Gillmore JD, Damy T, Fontana M, Hutchinson M, Lachmann HJ, Martinez‐Naharro A, Quarta CC, Rezk T, Whelan CJ, Gonzalez‐Lopez E, et al. A new staging system for cardiac transthyretin amyloidosis. Eur Heart J. 2017;39:2799–2806. doi: 10.1093/eurheartj/ehx589 [DOI] [PubMed] [Google Scholar]
- 6. Grogan M, Scott CG, Kyle RA, Zeldenrust SR, Gertz MA, Lin G, Klarich KW, Miller WL, Maleszewski JJ, Dispenzieri A. Natural history of wild‐type transthyretin cardiac amyloidosis and risk stratification using a novel staging system. J Am Coll Cardiol. 2016;68:1014–1020. doi: 10.1016/j.jacc.2016.06.033 [DOI] [PubMed] [Google Scholar]
- 7. Cappelli F, Martone R, Gabriele M, Taborchi G, Morini S, Vignini E, Allinovi M, Di Gioia M, Bartolini S, Di Mario C, et al. Biomarkers and prediction of prognosis in transthyretin‐related cardiac amyloidosis: direct comparison of two staging systems. Can J Cardiol. 2020;36:424–431. doi: 10.1016/j.cjca.2019.12.020 [DOI] [PubMed] [Google Scholar]
- 8. Müller ML, Spethmann S, Messroghli D, Brand A, Mattig I, Hahn K, Landmesser U, Heidecker B. Accuracy of established prognostic staging systems for cardiac transthyretin amyloidosis in the tafamidis era. JACC Adv. 2025;4:101568. doi: 10.1016/j.jacadv.2024.101568 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Nitsche C, Ioannou A, Patel RK, Razvi Y, Porcari A, Rauf MU, Bandera F, Aimo A, Emdin M, Martinez‐Naharro A, et al. Expansion of the National Amyloidosis Centre staging system to detect early mortality in transthyretin cardiac amyloidosis. Eur J Heart Fail. 2024;26:2008–2012. doi: 10.1002/ejhf.3354 [DOI] [PubMed] [Google Scholar]
- 10. Maurer MS, Schwartz JH, Gundapaneni B, Elliott PM, Merlini G, Waddington‐Cruz M, Kristen AV, Grogan M, Witteles R, Damy T, et al. Tafamidis treatment for patients with transthyretin amyloid cardiomyopathy. N Engl J Med. 2018;379:1007–1016. doi: 10.1056/NEJMoa1805689 [DOI] [PubMed] [Google Scholar]
- 11. Kittleson MM, Maurer MS, Ambardekar AV, Bullock‐Palmer RP, Chang PP, Eisen HJ, Nair AP, Nativi‐Nicolau J, Ruberg FL. Cardiac amyloidosis: evolving diagnosis and management: a scientific sithent from the American Heart Association. Circulation. 2020;142:e7–e22. doi: 10.1161/CIR.0000000000000792 [DOI] [PubMed] [Google Scholar]
- 12. Muchtar E, Kumar SK, Gertz MA, Grogan M, AbouEzzeddine OF, Jaffe AS, Dispenzieri A. Staging systems use for risk stratification of systemic amyloidosis in the era of high‐sensitivity troponin T assay. Blood. 2019;133:763–766. doi: 10.1182/blood-2018-10-875252 [DOI] [PubMed] [Google Scholar]
- 13. Levey AS, Coresh J, Greene T, Stevens LA, Zhang YL, Hendriksen S, Kusek JW, Van Lente F. Using standardized serum creatinine values in the modification of diet in renal disease study equation for estimating glomerular filtration rate. Ann Intern Med. 2006;145:247–254. doi: 10.7326/0003-4819-145-4-200608150-00004 [DOI] [PubMed] [Google Scholar]
- 14. Accessed April, 2025. http://www.accurint.com.
- 15. Gönen M, Heller G. Concordance probability and discriminatory power in proportional hazards regression. Biometrika. 2005;92:965–970. doi: 10.1093/biomet/92.4.965 [DOI] [Google Scholar]
- 16. Masri A, Bhattacharya P, Medoff B, Ejaz AU, Elman MR, Chandrashekar P, Ives L, Santos AM, Teruya SL, Zhao Y, et al. A multicenter study of contemporary long‐term Tafamidis outcomes in transthyretin amyloid cardiomyopathy. JACC CardioOncol. 2025;7:282–293. doi: 10.1016/j.jaccao.2024.12.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Lane T, Fontana M, Martinez‐Naharro A, Quarta CC, Whelan CJ, Petrie A, Rowczenio DM, Gilbertson JA, Hutt DF, Rezk T, et al. Natural history, quality of life, and outcome in cardiac transthyretin amyloidosis. Circulation. 2019;140:16–26. doi: 10.1161/CIRCULATIONAHA.118.038169 [DOI] [PubMed] [Google Scholar]
- 18. Carrero J‐J, Fu EL, Sang Y, Ballew S, Evans M, Elinder C‐G, Barany P, Inker LA, Levey AS, Coresh J, et al. Discordances between creatinine‐ and cystatin C–based estimated GFR and adverse clinical outcomes in routine clinical practice. Am J Kidney Dis. 2023;82:534–542. doi: 10.1053/j.ajkd.2023.04.002 [DOI] [PubMed] [Google Scholar]
- 19. Nankivell BJ, Nankivell LFJ, Elder GJ, Gruenewald SM. How unmeasured muscle mass affects estimated GFR and diagnostic inaccuracy. EClinicalMedicine. 2020;29‐30:100662. doi: 10.1016/j.eclinm.2020.100662 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Cheng RK, Levy WC, Vasbinder A, Teruya S, De Los Santos J, Leedy D, Maurer MS. Diuretic dose and NYHA functional class are independent predictors of mortality in pithnts with transthyretin cardiac amyloidosis. JACC CardioOncol. 2020;2:414–424. doi: 10.1016/j.jaccao.2020.06.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Tables S1–S4
STROBE statement