Low Concordance between NYHA Classification and Cardiopulmonary Exercise Test Variables in Patients with Heart Failure and Reduced Ejection Fraction

Abstract

Background

The New York Heart Association (NYHA) functional classification is the most commonly used classification system for heart failure (HF), whereas cardiopulmonary exercise testing (CPET) is the gold standard for functional status evaluation in HF.

Objective

This study aimed to analyze correlation and concordance between NYHA classes and CPET variables.

Methods

HF patients with clinical indication for CPET and ejection fraction (EF) < 50% were selected. Correlation (Spearman coefficient) and concordance (kappa) between NYHA classification and CPET-based classifications were analyzed. A p < 0.05 was accepted as significant.

Results

In total, 244 patients were included. Mean age was 56 ± 14 years, and mean EF was 35.5% ± 10%. Distribution of patients according to NYHA classification was 31.2%% class I, 48.3% class II, 19.2% class III, and 1.3% class IV. Correlation (r) between NYHA and Weber classes was 0.489 (p < 0.001), and concordance was 0.231 (p < 0.001). Correlation (r) between NYHA and ventilatory classes (minute ventilation/carbon dioxide production [VE/VCO₂] slope) was 0.218 (p < 0.001), and concordance was 0.002 (p = 0.959). Spearman correlation between NYHA and CPET score classes was 0.223 (p = 0.004), and kappa concordance was 0.027 (p = 0.606).

Conclusion

There was a moderate association between NYHA and Weber classes, although concordance was low. Ventilatory (VE/VCO₂slope) and CPET score classes had a weak association and a low concordance with NYHA classes.

Heart Failure; Prognosis; Exercise Test

Resumo

Fundamento

A classificação funcional da New York Heart Association (NYHA) é o sistema de classificação mais utilizado para a insuficiência cardíaca (IC), enquanto o teste de exercício cardiopulmonar (TECP) é o padrão ouro para a avaliação do estado funcional na IC.

Objetivo

Analisar a correlação e a concordância entre as classes da NYHA e as variáveis do TECP.

Métodos

Foram selecionados pacientes com IC com indicação clínica para TECP e fração de ejeção (FE) < 50%. A correlação (coeficiente de Spearman) e a concordância (kappa) entre a classificação da NYHA e as classificações baseadas no TECP foram analisadas. Um valor de p < 0,05 foi considerado significativo.

Resultados

No total, foram incluídos 244 pacientes no estudo. A idade média foi de 56±14 anos, e a FE média foi de 35,5%±10%. A distribuição de pacientes de acordo com a classificação da NYHA foi a seguinte: classe I (31,2%), classe II (48,3%), classe III (19,2%) e classe IV (1,3%). A correlação (r) entre as classes da NYHA e de Weber foi de 0,489 (p < 0,001), e a concordância foi de 0,231 (p < 0,001). A correlação (r) entre as classes da NYHA e ventilatórias (inclinação da ventilação minuto/produção de dióxido de carbono [VE/VCO₂]) foi de 0,218 (p < 0,001), e a concordância foi de 0,002 (p = 0,959). A correlação de Spearman entre as classes da NYHA e do escore TECP foi de 0,223 (p = 0,004), e a concordância kappa foi de 0,027 (p = 0,606).

Conclusão

Foi identificada uma associação moderada entre as classes da NYHA e de Webber, embora a concordância tenha sido baixa. As classes ventilatórias (inclinação VE/VCO₂) e do escore TECP apresentaram uma associação fraca e uma baixa concordância com as classes da NYHA.

Insuficiência Cardíaca; Prognóstico; Teste de Esforço

Introduction

Despite being a progressive disease, heart failure (HF) does not have a linear course. Hospitalizations due to HF decompensations are independent factors for prognosis. Risk prediction models and prognostic scores will determine the need to escalate specific therapeutic strategies, such as medication change, cardiac resynchronization therapy, implantable cardioverter-defibrillator, ventricular assist device, and cardiac transplantation.¹

The New York Heart Association (NYHA) classification is a well-known, low-cost, simple functional stratification tool for HF with prognostic value.² , ³ It divides patients into 4 different groups according to self-reported dyspnea severity and limitations to physical activities.² , ³ However, the NYHA functional class depends on self-reported symptoms and, therefore, is influenced by the subjectivity of each patient.⁴ , ⁵

Conversely, functional status is assessed objectively by cardiopulmonary exercise testing (CPET), which is a prognostic tool considered to be the gold standard for HF assessment.⁶ , ⁷ Important guidelines define CPET as a class I recommendation for cardiac transplantation and a class IIa recommendation for exercise prescription in this context.⁶ , ⁷

Classically, CPET prognostic evaluation is based on peak oxygen uptake (VO_2peak) measures.⁸ , ⁹ However, other variables such as minute ventilation/carbon dioxide production (VE/VCO₂) slope, heart rate recovery in 1 minute (HRR₁), oxygen uptake efficiency slope (OUES), end-tidal carbon dioxide partial pressure (PetCO₂), and periodic ventilation have demonstrated an independent and incremental prognostic value to VO_2peakin HF.¹⁰ Based on those variables, specific prognostic classifications have been validated, namely Weber classes (VO_2peak), ventilatory classes (VE/VCO₂slope), and CPET score (combining VO_2peak, VE/VCO₂slope, HRR₁, OUES, and PetCO₂).¹¹ - ¹³

Even though the NYHA classification system is widely used, there are few studies correlating NYHA classes with HF prognosis or CPET variables.¹⁴ , ¹⁵ Recently, a systematic review compared NYHA classification and CPET variables, and the variable that was common to all analyzed studies was VO_2peak, yet with much heterogeneity.¹⁴ This study aimed to evaluate correlation and concordance between NYHA classification for HF and CPET-based functional classifications, namely Weber classes, ventilatory classes, and CPET score.¹¹

Methods

This cross-sectional study consecutively recruited patients who underwent CPET for HF evaluation. Inclusion criteria were the following: 1) age ≥ 18 years; 2) confirmed HF diagnosis with ejection fraction (EF) < 50%; and 3) clinical indication for CPET between 2009 and 2019. Exclusion criteria were moderate-to-severe chronic obstructive pulmonary disease, pulmonary hypertension, and/or fibrosis or symptomatic anemia.

CPET variables and demographic data were collected together with clinical information and relevant complementary tests (12-lead resting electrocardiogram and Doppler echocardiogram from the past 3 months). CPET was symptom-limited and was performed at maximal effort with a ramp protocol in a treadmill (Micromed Centurion 300, São Paulo, Brazil) using a Cortex 3b breath-by-breath analyzer (Cortex Inc., Leipzig, Germany). Two-point gas calibration was done before the tests. All techniques followed current guidelines, and a nationally certified physician was responsible for each test.¹⁰

All CPET tests were conducted by the same physician, a cardiologist who specializes in CPET. Before CPET, the same physician in charge of the test determined each patient’s NYHA class according to self-reported limitation to physical activity: (I) no limitation to physical activity; (II) slight limitation to physical activity; (III) marked limitation to physical activity; or (IV) unable to perform any physical activity without discomfort.¹⁶ Then, based on CPET variables, patients were classified into Weber classes, ventilatory classes, and CPET score classes according to their CPET results.¹¹

Weber classification categorizes patients according to their VO_2peakas follows: (A) VO₂> 20 mL.kg^-1.min^-1; (B) VO₂16-20 mL.kg^-1.min^-1; (C) VO₂10-15 mL.kg^-1.min^-1; or (D) VO₂< 10 mL.kg^-1.min^-1.¹² Ventilatory classes use VE/VCO₂slope: (I) VE/VCO₂≤ 29.9; (II) VE/VCO₂30-35.9; (III) VE/VCO₂36-44.9; or (IV) VE/VCO₂≥ 45.¹³ CPET score was calculated for each patient based on the summation of abnormal responses as follows: VE/VCO₂≥ 34 (7 points); HRR₁≤ 6 bpm (5 points); OUES ≤ 1.4 (3 points); PetCO₂< 33 mm Hg (3 points); and VO_2peak≤ 14 mL.kg^-1.min^-1(2 points).¹¹ , ¹⁵ The score is then divided into quartiles: (I) 0-5; (II) 6-10; (III) 10-15; and (IV) > 15.¹¹

Statistical analysis

SPSS version 17.0 (SPSS Inc., Chicago, IL, USA) was used for all statistical analyses. Continuous variables were reported as mean and standard deviation for parametric distribution or as median and interquartile range for nonparametric distribution. Kolmogorov-Smirnov normality test and histogram analysis were used for determination of distribution. Categorical variables were reported as absolute numbers and proportions. Correlation between variables was assessed using Spearman ( s ) or Pearson ( p ) correlation coefficient, and concordance was assessed using kappa (k) coefficient. For all analyses, a p < 0.05 was accepted as statistically significant.

An institutional research ethics committee approved the study protocol. Also, the study respects all national and international regulations for human research.

Results

Patients’ characteristics are described in Table 1 . The sample included 244 patients, mainly men (77.9%), and mean age was 56 ± 14 years. Ischemia was the most frequent etiology (44.4%). Mean EF was 35.5% ± 10%. Patients were on optimized medical therapy as follows: 86.4% angiotensin-converting enzyme inhibitors or angiotensin II receptor blockers, 91.4% beta-blockers, 57.0% aldosterone antagonists, and 53.5% diuretics. Mean VO_2peakwas 19.2 ± 6.7 mL.kg^-1.min^-1, whereas mean VE/VCO₂slope was 39 ± 10. Mean respiratory exchange ratio (RER) was 1.041 ± 0.12 (25% had a RER > 1.10). All tests were interrupted by the effort criteria, and none was interrupted prematurely or due to hemodynamic, arrhythmic, or ischemic criteria. Patients were distributed according to NYHA classification as follows: 31.3% class I, 48.3% class II, 19.2% class III, and 1.3% class IV ( Table 2 ).

Table 1
– General patient demographic, clinical, and cardiopulmonary exercise test characteristics (n = 244)

Variables
Age (mean ± SD)	56±14 years
Gender
Male, n (%)	190 (77.9)
Etiology
Ischemic, n (%)	107 (44.4)
Idiopathic, n (%)	56 (23.2)
Viral, n (%)	30 (12.4)
Chagasic, n (%)	18 (7.5)
Other, n (%)	30 (12.5)
Comorbidities
Hypertension, n (%)	70 (34.7)
Diabetes mellitus, n (%)	43 (21.2)
Coronary artery disease, n (%)	94 (46.3)
Smoking, n (%)	4 (2.0)
Medications used
ACEI or ARB, n (%)	209 (86.4)
Beta-blocker, n (%)	222 (91.4)
MCRA, n (%)	138 (57.0)
Diuretics, n (%)	129 (53.5)
Implantable devices
Pacemaker, n (%)	17 (7.0)
CRT and/or ICD, n (%)	28 (11.5)
VO2peak(mL.kg-1.min-1), mean ± SD	19.2±6.7
Percent of predicted VO2peak(%), mean ± SD	63±20
EF (%), mean ± SD	35.5±10
RER, mean ± SD	1.041±0.12
VE/VCO2slope, mean ± SD	39.0±10.8
PetCO2(mm Hg), mean ± SD	29.8±4.66
HRR1 , median (IQR)	18.0 (15)
SBP at rest, median (IQR)	120 (10)
HR at rest, median (IQR)	74 (22)

• _2peak221

Table 2
– Sample distribution according to subjective and objective classifications, n (%)

	I	II	III	IV
NYHA class	75 (31.2)	116 (48.3)	46 (19.2)	3 (1.3)
VE/VCO2slope	42 (17.2)	70 (28.7)	74 (30.3)	58 (23.8)
CPET score	57 (34.7)	61 (37.2)	36 (22.0)	10 (6.1)
	A	B	C	D
Weber class	95 (39)	55 (22.5)	81 (33.2)	13 (5.3)

• ₂

Figure 1 shows NYHA class distribution according to Weber classes ( Figure 1A ), ventilatory classes (Figure 1B), and CPET score classes (Figure 1C). Correlation (r) between NYHA and Weber classes was 0.489 (p < 0.001), and concordance was 0.231 (p < 0.001). Correlation (r) between NYHA and ventilatory classes was 0.218 (p < 0.001), and concordance was 0.002 (p = 0.959). Finally, correlation (r) between NYHA and CPET score classes was 0.223 (p = 0.004), and concordance was 0.027 (p = 0.606).

Figure 1
– NYHA class distribution, correlation, and concordance according to (A) Weber classes, (B) ventilatory classes (VE/VCO2 slope), and (C) CPET score classes. NYHA: New York Heart Association; VE/VCO2: minute ventilation/carbon dioxide production; CPET: cardiopulmonary exercise test; r: correlation coefficient.

Discussion

In patients with HF with reduced ejection fraction who underwent CPET after clinical indication, we found only a moderate association between NYHA and Weber classes, with a low concordance. However, there was an even lower association or concordance rate between NYHA classification and ventilatory or CPET score classes.

All those functional status classifications have their prognostic value validated for HF.³ , ¹¹ Thus, functional status is the best parameter for risk prediction in those patients.³ , ¹¹ However, as we showed, there was a low concordance between NYHA classification and the 3 classifications based on CPET (which is an objective clinical test). Even though we found a moderate correlation between NYHA and Weber classes, it seems reasonable to hypothesize that subjectivity interferes in NYHA classification risk prediction for HF and has a subsequent impact on therapeutic decisions.

A recent systematic review addressed the correlation between NYHA classification for HF and VO_2peakmeasures (determined by CPET).¹⁴ It found a great heterogeneity in NYHA classes among the included studies.¹⁴ Our findings support those of Lim et al. and reflect a further correlation analysis, as we described the correlation between the subjective NYHA classification and some objective classifications that are based on CPET results, either through a validated score or through ventilatory classes. For example, patients subjectively considered to be in NYHA class I by their attending physicians may have ventilatory class IV VE/VCO₂slope values (poorest prognosis) or be in the poorest prognostic quartile of the CPET score ( Figure 1 ).¹¹ , ¹³

NYHA classification may lead to different interpretations of the same patient from different attending physicians,³ especially when symptoms from intermediate classes (II and III) are reported. In a publication from our group, Ritt et al. demonstrated that patients in Weber class B could be divided into two different prognostic groups when the CPET score was calculated.¹⁵ The groups were then divided into one of higher risk and another of lower risk. However, patients in intermediate NYHA classes are generally those whose functional status is of great importance for decision-making. These decisions include increasing or changing medications, providing surgical indications, or implanting devices (such as cardiac resynchronization therapy or ventricular assist device).¹⁶ In such groups, NYHA classification may not be sensitive enough to address minor but important clinical features. Therefore, an objective, easily reproducible, reliable classification is urgently needed. In patients with NYHA class I or II, CPET may reclassify them to higher risk, and patients with NYHA class III may be reclassified to lower risk, especially those who are candidates to medication changes and/or devices. The use of CPET for this purpose is a matter for future studies.

Our study has some limitations, such as lack of clinical follow-up of our patient sample. We excluded symptomatic anemia, as we focused on clinical diagnostic criteria, but one may argue that asymptomatic anemia may also impact functional capacity. Also, the prevalence of depression was not assessed in our patients, although it may contribute to the lack of effort. Our sample had a mean RER of 1.04; one may argue that a RER > 1.10 is the pattern for achieving acidosis, although in HF some use RER > 1.00 as an acceptable criterion.¹⁷ Although this may impact VO_2peak, it does not impact VE/VCO₂slope, OUES, or HRR₁. New studies addressing a wider population and analyzing clinical outcomes are necessary to a better understanding of the actual prognostic value of each HF classification (NYHA, VE/VCO₂slope, Weber classes, and CPET score). We focused on Weber classes, VE/VCO₂slope classes, and CPET score because all these parameters may be presented as 4-level scale classifications as NYHA; also, VO_2peakand VE/VCO₂slope are the most studied variables in CPET, and other variables from CPET are inserted in the CPET score. However, future studies focusing on specific CPET variables are valuable. Importantly, it remains to be determined whether there is, in fact, an objective CPET-based strategy that is more accurate than the others.

Conclusion

There was a moderate association between the subjective NYHA classification and the objectively measured Weber classes, although concordance was low. The objectively measured ventilatory classes and CPET score classes had a weak association and a low concordance with the NYHA classification.

Referências

• ¹
  Chaudhry SP, Stewart GC. Advanced Heart Failure: Prevalence, Natural History, and Prognosis. Heart Fail Clin. 2016;12(3):323-33. doi: 10.1016/j.hfc.2016.03.001.
• ²
  Holland R, Rechel B, Stepien K, Harvey I, Brooksby I. Patients’ self-assessed functional status in heart failure by New York Heart Association class: a prognostic predictor of hospitalizations, quality of life and death. J Card Fail. 2010;16(2):150-6. doi: 10.1016/j.cardfail.2009.08.010.
• ³
  Caraballo C, Desai NR, Mulder H, Alhanti B, Wilson FP, Fiuzat M, et al. Clinical Implications of the New York Heart Association Classification. J Am Heart Assoc. 2019;8(23):e014240. doi: 10.1161/JAHA.119.014240.
• ⁴
  Yap J, Lim FY, Gao F, Teo LL, Lam CS, Yeo KK. Correlation of the New York Heart Association Classification and the 6-Minute Walk Distance: A Systematic Review. Clin Cardiol. 2015;38(10):621-8. doi: 10.1002/clc.22468.
• ⁵
  Raphael C, Briscoe C, Davies J, Whinnett ZI, Manisty C, Sutton R, et al. Limitations of the New York Heart Association functional classification system and self-reported walking distances in chronic heart failure. Heart. 2007;93(4):476-82. doi: 10.1136/hrt.2006.089656.
• ⁶
  Mehra MR, Kobashigawa J, Starling R, Russell S, Uber PA, Parameshwar J, et al. Listing criteria for heart transplantation: International Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidates--2006. J Heart Lung Transplant. 2006;25(9):1024-42. doi: 10.1016/j.healun.2006.06.008.
• ⁷
  Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure: The Task Force for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC)Developed with the special contribution of the Heart Failure Association (HFA) of the ESC. Eur Heart J. 2016;37(27):2129-200. doi: 10.1093/eurheartj/ehw128.
• ⁸
  Mancini DM, Eisen H, Kussmaul W, Mull R, Edmunds LH Jr, Wilson JR. Value of peak exercise oxygen consumption for optimal timing of cardiac transplantation in ambulatory patients with heart failure. Circulation. 1991;83(3):778-86. doi: 10.1161/01.cir.83.3.778.
• ⁹
  Guimarães GV, Silva MS, d’Avila VM, Ferreira SM, Silva CP, Bocchi EA. Peak VO2 and VE/VCO2 slope in betablockers era in patients with heart failure: a Brazilian experience. Arq Bras Cardiol. 2008;91(1):39-48. doi: 10.1590/s0066-782x2008001300007.
• ¹⁰
  Balady GJ, Arena R, Sietsema K, Myers J, Coke L, Fletcher GF, et al. Clinician’s Guide to cardiopulmonary exercise testing in adults: a scientific statement from the American Heart Association. Circulation. 2010;122(2):191-225. doi: 10.1161/CIR.0b013e3181e52e69.
• ¹¹
  Myers J, Oliveira R, Dewey F, Arena R, Guazzi M, Chase P, et al. Validation of a cardiopulmonary exercise test score in heart failure. Circ Heart Fail. 2013;6(2):211-8. doi: 10.1161/CIRCHEARTFAILURE.112.000073.
• ¹²
  Weber KT, Kinasewitz GT, Janicki JS, Fishman AP. Oxygen utilization and ventilation during exercise in patients with chronic cardiac failure. Circulation. 1982;65(6):1213-23. doi: 10.1161/01.cir.65.6.1213.
• ¹³
  Arena R, Myers J, Abella J, Peberdy MA, Bensimhon D, Chase P, et al. Development of a ventilatory classification system in patients with heart failure. Circulation. 2007;115(18):2410-7. doi: 10.1161/CIRCULATIONAHA.107.686576.
• ¹⁴
  Lim FY, Yap J, Gao F, Teo LL, Lam CSP, Yeo KK. Correlation of the New York Heart Association classification and the cardiopulmonary exercise test: A systematic review. Int J Cardiol. 2018;263:88-93. doi: 10.1016/j.ijcard.2018.04.021.
• ¹⁵
  Ritt LE, Myers J, Stein R, Arena R, Guazzi M, Chase P, et al. Additive prognostic value of a cardiopulmonary exercise test score in patients with heart failure and intermediate risk. Int J Cardiol. 2015;178:262-4. doi: 10.1016/j.ijcard.2014.10.025.
• ¹⁶
  Ponikowski P, Voors AA, Anker SD, Bueno H, Cleland JGF, Coats AJS, et al. 2016 ESC guidelines for the diagnosis and treatment of acute and chronic heart failure, Rev. Esp. Cardiol. 2016;69(12):1167. https://doi.org/10.1016/j.recesp.2016.10.014
  » https://doi.org/10.1016/j.recesp.2016.10.014
• ¹⁷
  Malhotra R, Bakken K, D’Elia E, Lewis GD. Cardiopulmonary Exercise Testing in Heart Failure. JACC Heart Fail. 2016;4(8):607-16. doi: 10.1016/j.jchf.2016.03.022.

• Study Association
  This study is not associated with any thesis or dissertation work.

• Sources of Funding: There were no external funding sources for this study.

Publication Dates

• Publication in this collection
  04 Apr 2022
• Date of issue
  2022

History

• Received
  12 Mar 2021
• Reviewed
  08 Aug 2021
• Accepted
  01 Sept 2021