- Email: [email protected]
Etiology of troponin I elevation in patients with congestive heart failure and low clinical suspicion of myocardial infarction
Resuscitation 63 (2004) 195–201
Etiology of troponin I elevation in patients with congestive heart failure and low clinical suspicion of myocardial infarction George O. Angheloiu a , Reginald P. Dickerson b , Keyvan Ravakhah b,∗ b
a Department of Cardiovascular Medicine, The Cleveland Clinic Foundation, Cleveland, OH, USA Department of Medicine, Huron Hospital, Cleveland Clinic Health System, Cleveland, OH44112, USA
Received 9 February 2004; received in revised form 26 May 2004; accepted 26 May 2004
Abstract Background: Increased troponin I (TnI) has been detected in congestive heart failure (CHF) patients. We demonstrate that occurrence of LVH, and systolic and diastolic blood pressure values (SBP and DBP, respectively) correlate with increasing levels of serum TnI in a selected CHF population. Methods: Our team reviewed the medical records of 40 consecutive patients admitted with CHF symptoms, low clinical suspicion of myocardial infarction (MI), and serum TnI measured upon admission. Twenty numerical and categorical parameters were screened. Patients were divided in to seven non-exclusive groups by increasing cutoff values of serum TnI (0, 0.1, 0.2, 0.3, 0.4, 0.6, and 1.5 ng/ml). In each group the average value of numerical parameters and the occurrence of categorical ones were calculated. The correlation between these values and the seven TnI cutoffs were analyzed. Results: Occurrence of electrocardiographic left ventricular hypertrophy (ECG-LVH), and average values of SBP and DBP significantly correlated with the TnI cutoffs by means of multiple regression analysis (P = 0.001, 0.02 and 0.007, respectively). Conclusion: ECG-LVH and values of SBP and DBP significantly correlate with increasing cutoff levels of serum TnI in hospitalized patients with CHF and low clinical suspicion of MI. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Troponin I; Congestive heart failure; Left ventricular hypertrophy
Resumo Contexto: Tem sido detectado um aumento de troponina I (Tn I) em doentes com Insuficiˆencia Card´ıaca Congestiva (CHF). Demonstramos que a ocorrˆencia de LVH, e os valores de pressão arterial sistólica e diastólica (SBP e DBP, respectivamente) correlacionam-se com n´ıveis elevados de Tn I numa população seleccionada com CHF. Métodos: A equipa reviu os registos médicos de 40 doentes admitidos consecutivamente com sintomas de CHF, baixa suspeição cl´ınica de enfarte do miocárdio (MI), e TnI sérica medida na admissão. Foram revistos vinte parˆametros numéricos e categóricos. Os doentes foram divididos em sete grupos não exclusivos aumentando os valores de cutoff da TnI sérica (0, 0.1, 0.2, 0.3, 0.4, 0.6 e 1.5 ng/ml). Em cada grupo foi calculado o valor médio dos parˆametros numéricos e a ocorrˆencia de valores categóricos. Foram analizados a correlação entre estes valores e os sete cutoffs de TnI. Resultados: A ocorrˆencia de hipertrofia ventricular esquerda electrocardiográfica (ECG-LVH), e valores médios de SBP e DBP correlacionam-se de forma significativa com os cutoffs por meios de análise de regressão múltipla (P = 0.001, 0.02 e 0.007, respectivamente). Conclusão: ECG-LVH e os valores de SBP e DBP correlacionam-se significativamente com n´ıveis de cutoff elevados de TnI em doentes hospitalizados com CHF e baixa suspeição cl´ınica de MI. © 2004 Elsevier Ireland Ltd. All rights reserved. Palavras chave: Troponina I; Insuficiˆencia Card´ıaca Congestiva; Hipertrofia Ventricular Esquerda
Resumen Antecedentes: Se ha detectado troponina I (TnI) aumentada en pacientes con insuficiencia card´ıaca congestiva. Demostramos que la ocurrencia de hipertrofia ventricular izquierda (LVH), valores de presión sistólica y diastólica (SBP yDBP, respectivamente) se correlacionan con niveles crecientes de TnI en una población seleccionada con CHF. Métodos: nuestro equipo revisó los registros médicos de 40 pacientes
∗ Corresponding author. E-mail address: [email protected] (K. Ravakhah).
0300-9572/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2004.05.018
196
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
consecutivos admitidos con s´ıntomas de CHF, y escasa sospecha cl´ınica de infarto de miocardio (MI), y se midió TnI sérica al ingreso. Se midieron 20 parámetros numéricos y categóricos. Los pacientes se dividieron en 7 grupos no exclusivos por valores de corte crecientes de TnI sérico (0, 0.1, 0.2, 0.3, 0.4, 0.6 y 1.5 ng/ml). En cada grupo se calculó el valor promedio de los parámetros numéricos y la ocurrencia de los categóricos. Se analizó la correlación entre estos valores y los 7 cortes en TnI. Resultados: La ocurrencia de LVH electrocardiográfica, y valores promedio de SBP y DBP se correlacionaron significativamente con los cortes de TnI por medio de análisis de regresión múltiple (P = 0.001, 0.02 y 0.007, respectivamente). Conclusión: LVH en el ECG y los valores de SBP y DBP se correlacionan significativamente con cortes crecientes en TnI sérico en pacientes hospitalizados con CHF y escasa sospecha de MI © 2004 Elsevier Ireland Ltd. All rights reserved. Palavras chave: Troponina I; Falla cardiaca congestiva; Hipertrofia ventricular izquierda
Troponin I (TnI) is an important diagnostic and prognostic marker in patients with acute coronary syndromes [1,2]. Investigational evidence indicates that levels of TnI greater than the normal range are present in patients with congestive heart failure (CHF) [3], could be associated with etiologies other than acute coronary syndromes and predict negative outcomes in patients with this diagnosis [4,5]. In the same time left ventricular remodeling was found to be a cause for elevated TnI in patients with severe non-ischemic CHF [6], and TnI above upper normal limit of 0.5 ng/ml was associated with left ventricular hypertrophy in outpatient hypertensive patients [7]. In the present study we investigated what clinical or laboratory variables could correlate with increasing serum TnI cutoff values in patients admitted with a CHF diagnosis [8], and low clinical suspicion of myocardial infarction (MI).
ria [10]. LV ejection fraction was considered abnormal for values lower than 40%. To analyze the relative contribution to the TnI variance of all 20 parameters recorded, of both numerical and categorical type, we divided the patients investigated into seven non-exclusive groups in a step-wise manner, using increasing cutoff values of serum TnI. Each group included all patients with TnI levels greater than or equal to a particular cutoff: 0 (n = 40, all patients), 0.1 (n = 24), 0.2 (n = 19), 0.3 (n = 15), 0.4 (n = 9), 0.6 (n = 6) and 1.5 ng/ml (n = 4). For every categorical parameter studied we calculated its occurrence in each group of patients defined as above, while for numerical parameters we calculated their average value. Linear regression analysis was performed between the occurrence and, respectively, average values of the 20 parameters in these seven groups and the TnI cutoffs that defined the groups. To find which parameters contributed independently to the serum TnI variance, we performed multiple regression analysis using the variables that
2. Methods
Table 1 Parameters recorded in the study
1. Introduction
We define low clinical suspicion of myocardial infarction as the absence of chest pain, ischemic electrocardiographic findings (ST segment elevation or depression of 0.1 mV or greater in two adjacent leads), and previous coronary artery interventions [9]. We reviewed the charts of 40 consecutive patients with this clinical picture and CHF symptoms admitted between August and December 1999 in a network of three community teaching hospitals, after obtaining the necessary IRB approval. We studied 20 clinical and laboratory parameters recorded in the first 48 h after admission, of both numerical and categorical type, as listed in Table 1. Two-dimensional echocardiogram was performed in all patients and cardiac catheterization in three. Serum TnI was measured using a Dade-Behring RXL Dimension instrument (Dade Behring, Deerfield, IL) at least three times at 4-h interval, starting upon admission. The lower reference limit of this instrument is 0.1 ng/ml. Peak serum TnI value was considered to be the highest value measured in each patient, and all statements made in this study refer to this value. To diagnose electrocardiographic left ventricular hypertrophy (ECG-LVH) we used the Sokolow-Lyon QRS voltage crite-
Parameter group
Parameter
Demographic
Gender Race Age H/o diabetes H/o hypercholesterolemia H/o hypertension H/o smoking H/o stroke H/o PVD H/o aortic stenosis Blood pressure valuea LV wall motion abnormalities LV inter-ventricular septum LV posterior wall LV ejection fraction LV diameter ECG-LVH Pulmonary edema Creatinine Hemoglobin
Clinical
2D-echo
ECG Chest X-ray Blood tests
ECG, electrocardiogram; ECG-LVH, electrocardiographic left ventricular hypertrophy; H/o, history of; LV, left ventricle. a Recorded as the initial value in the emergency department, or on the floor in the case of direct admissions.
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
positively correlated with the TnI cutoff values by linear regression. Subsequently, we tested whether there was a direct correlation between the individual numerical parameters that had an independent contribution to the TnI variance and the correspondent individual serum TnI levels. An arbitrary value of 0 ng/ml was used for TnI in all cases where TnI was lower than 0.1 ng/ml. For categorical parameters that had an independent contribution to the TnI variance, we felt necessary to compare the occurrence of these parameters between patients with TnI levels greater than each of the TnI cutoff analyzed and patients with TnI lower than the same cutoff values. Comparisons were performed for all TnI cutoffs analyzed with the exception of 0 ng/ml. 3. Statistics Linear and multiple regression analysis for correlation characterization were used as detailed above, and between-groups comparison was performed by means of Fisher exact test. P-values lower than 0.05 were considered significant. 4. Results Electrocardiogram and two-dimensional echocardiography were performed in all 40 patients, and cardiac catheterization in 3. Serum TnI levels of patients included in the study are depicted in Fig. 1. There were significant correlations between increasing serum TnI cutoff values (0, 0.1, 0.2, 0.3, 0.4, 0.6, and 1.5 ng/ml) and several parameters as illustrated in Table 2. The most significant correlations were documented with occurrence of ECG-LVH and African-American race,
197
Table 2 Results of linear regression between serum TnI cutoffs and 13 parameters investigated Parameter
Pearson’s r-value
P-value
African-American race History of diabetes Hypercholesterolemia Smoking Creatinine SBP DBP ECG-LVH Pulmonary edema IVS LV ejection fraction LV end-diastolic diameter Patients with WMA
0.94 −0.81 0.57 0.83 0.81 0.99 0.99 0.95 0.87 0.92 0.06 0.01 –0.80
<0.001 –a 0.17 0.012 0.018 <0.001 <0.001 <0.001 0.005 0.002 0.58 0.83 –a
DBP, diastolic blood pressure; ECG-LVH, electrocardiographic left ventricular hypertrophy; IVS, interventricular septum; SBP, systolic blood pressure; WMA, wall motion abnormalities. a No P-value was calculated for negative Pearson’s factor, since this would imply a counter-physiological correlation between increased TnI values and the parameters analyzed.
and average systolic (SBP) and diastolic blood pressure (DBP; < 0.001). No positive correlation was detected between serum TnI levels and the presence of increased LV end-diastolic diameter, low ejection fraction or wall motion abnormalities. When introduced in a multiple regression algorithm, ECG-LVH occurrence and average SBP and DBP had independent contributions to serum TnI variance (P = 0.001, 0.02, and 0.007, respectively). Fig. 2 illustrates the linear regression relation between increasing TnI cutoff values and these three parameters. Individual TnI values correlated with correspondent individual SBP and DBP values as illustrated in Fig. 3 (P = 0.0004 and 0.006, respectively). Occurrence of ECG-LVH was significantly greater in five groups of patients with high TnI values in comparison to the correspondent groups of patients with TnI lower than the cutoffs analyzed (Table 3). Three patients of the four with serum TnI equal to 1.5 ng/ml or greater had a cardiac catheterization performed. Two of them did not present atherosclerotic lesions, and a third one had a 50% stenosis in a secondary obtuse marginal branch. A total of nine patients (23%) had wall motion abnormalities on echocardiogram, with five among the patients with serum TnI values of 0.1 ng/ml or greater (21%). We performed a second analysis to test the correlation between serum TnI values and the parameters investigated after excluding these nine patients, and highly significant correlations persisted with ECG-LVH occurrence and average SBP and DBP (P = 0.008, 0.004, and 0.002, respectively). 5. Discussion
Fig. 1. Scatter plot of peak serum troponin I values in 40 hospitalized CHF patients with low suspicion of myocardial infarction.
In the present study we investigated the possible existence of clinical or laboratory parameters that would corre-
198
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
Fig. 2. Linear regression correlation of serum TnI cutoff with occurrence of electrocardiographic left ventricular hypertrophy (ECG-LVH) (a), and average values of systolic (SBP) (b) and diastolic blood pressure (DBP) (c) in groups of patients defined by increasing TnI cutoff values.
late with elevated TnI values in hospitalized CHF patients with low clinical suspicion of MI. In order to perform a complete search we studied 20 variables of both numerical and categorical nature (Table 1). To be able to compare both types of parameters we took a step-wise analysis approach in the first part of our investigation, and identi-
fied which parameters contribute in an independent manner to the TnI variance. We subsequently confirmed the resulting parameters, either by correlating the numerical variables with the individual TnI values, or by comparing the occurrence of categorical parameters in groups of patients with TnI greater than the cutoffs analyzed versus the
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
199
Fig. 3. Linear regression correlation between individual TnI levels and correspondent individual values of systolic (SBP) (a) and diastolic blood pressure (DBP) (b). Table 3 Comparison of ECG-LVH occurrence between groups of patients with TnI levels greater than or equal to six TnI cutoffs and correspondent groups with TnI lower than the same values TnI cutoff (ng/ml)
1.5 0.6 0.4 0.3 0.2 0.1
Patients with TnI ≥ cutoff
Patients with TnI < cutoff
n
ECG-LVH occurrence
n
ECG-LVH occurrence
4 6 9 14 19 23
4 4 6 8 8 9
36 34 31 26 21 17
7 7 5 3 3 2
(100%) (67%) (67%) (57%) (42%) (39%)
(19%) (21%) (16%) (12%) (15%) (12%)
P-value
0.003 0.04 0.007 0.004 0.05 0.06
ECG-LVH, electrocardiographic left ventricular hypertrophy; n, number of patients; TnI, troponin I.
correspondent groups with TnI levels lower than the same values. Our results indicate that occurrence of ECG-LVH in CHF patients with low clinical suspicion of myocardial infarction independently correlated with increasing TnI cutoff values (Fig. 2). We also demonstrated that ECG-LVH occurrence was significantly higher in groups of patients with TnI
greater than or equal to 1.5, 0.6, 0.4, 0.3, and 0.2 ng/ml in comparison to the correspondent groups of patients with TnI lower than these values (Table 3). This comparison had a strong tendency for being significant at the TnI cutoff value of 0.1 ng/ml (Table 3). This findings are in agreement with two previous reports, which argued that left ventricular remodeling was a cause of
200
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
elevated TnI in patients with severe non-ischemic CHF [6], and that elevated TnI was associated with left ventricular hypertrophy in outpatient hypertensive patients [7]. Several clinical and experimental studies could suggest a causal relationship between LVH and elevated serum TnI levels. Four clinical studies performed in patients without coronary artery disease showed that left ventricular mass correlated with the extent of coronary flow reserve impairment, and of ischemia as measured by exercise thallium perfusion imaging [11–14]. An animal study by Van Eyk et al. [15] showed that global ischemia can result in TnI breakdown and release. In light of these studies, the association between ECG-LVH and elevated TnI can be explained by a link between LVH, global myocardial ischemia, cardiac TnI breakdown, and resulting release of serum TnI. We also determined that average SBP and DBP in groups of patients defined by increasing TnI cutoffs, independently, correlated with the TnI cutoff values used (Fig. 2). At the same time individual SBP and DBP values recorded in each patient significantly correlated with the correspondent individual TnI levels (Fig. 3). An explanation for these results is suggested by an experimental study by Feng et al. [16], where elevated LV preload resulted in TnI degradation. It is reasonable to think that an increase in afterload induced by high blood pressure might have as end-result a similar process of TnI degradation. This inference makes intuitively logical the association between elevated values of SBP and DBP and increased TnI. There was no positive correlation between elevated serum TnI and the occurrence of low ejection fraction, increased LV end-diastolic diameter, or wall motion abnormalities (Table 2), although the small number of patients included in the study may have prevented these associations to occur. We made considerable efforts to insure that acute coronary events were not a cause of TnI elevation in our study. Patients were included in the study only in the absence of chest pain, ECG ischemic changes, and previous coronary artery interventions [9]. Three of the four patients with TnI greater than 1.5 ng/ml were subjected to cardiac catheterization, and no or trivial coronary artery disease was detected. All patients had a two-dimensional echocardiogram performed. There were a total of nine patients with wall motion abnormalities, with five among patients with serum TnI of 0.1 ng/ml or greater. Although numerous studies documented the presence of asynergic myocardial regions in patients with non-ischemic dilated cardiomyopathy [17], wall motion abnormalities are generally considered a sign of coronary artery disease. To insure that our results were not biased by the eventual presence of coronary artery disease, we excluded in a second analysis the nine patients with wall motion abnormalities. Highly significant correlations between TnI levels and the presence of ECG-LVH and values of SBP and DBP persisted even after these patients were excluded. Nonetheless, temporary intra-coronary thrombi caused by non-obstructive coronary plaques could be reason for short-lived release of serum TnI, and neither echocardiography nor angiography,
the methods employed in our study, could detect this type of atherosclerotic disease. Our study suggests that ECG-LVH occurrence, and SBP and DBP values independently correlate with increasing TnI cutoff levels in hospitalized patients with CHF symptoms and low clinical suspicion of myocardial infarction. Increased serum TnI is a highly specific and sensitive marker for myocardial damage [9], and early identification of its etiology could be paramount in adopting the proper inpatient medical management and therapy. Caution should nevertheless be exercised when interpreting these results, since a limited number of patients were included in the study, and the work directed towards ruling out coronary atherosclerosis was rather not exhaustive. Larger studies, eventually accompanied by a more invasive strategy of excluding coronary artery disease, should follow up.
Acknowledgements We gratefully acknowledge Drs. Burton C West, Hsien Wang, Michael Kalus and Robert E. Botti for valuable suggestions and technical support, and Mrs. Joyce Callahan for editorial assistance.
References [1] Antman EM, Tanasijevic MJ, Thompson B, et al. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. N Engl J Med 1996;335:1342–9. [2] D’Costa M, Fleming E, Patterson MC. Cardiac troponin I for the diagnosis of acute myocardial infarction in the emergency department. Am J Clin Pathol 1997;108:550–5. [3] Missov E, Calzolari C, Pau B. Circulating cardiac troponin I in severe congestive heart failure. Circulation 1997;96:2953–8. [4] Khan IA, Tun A, Wattanasauwan N, et al. Elevation of serum cardiac troponin I in noncardiac and cardiac diseases other than acute coronary syndromes. Am J Emerg Med 1999;17:225–9. [5] Horwich TB, Patel J, MacLellan WR, et al. Cardiac troponin I is associated with impaired hemodynamics, progressive left ventricular dysfunction, and increased mortality rates in advanced heart failure. Circulation 2003;108:833–8. [6] Logeart D, Beyne P, Cusson C, et al. Evidence of cardiac myolysis in severe nonischemic heart failure and the potential role of increased wall strain. Am Heart J 2001;141:247–53. [7] Siciliano M, Mettimano M, Dondolini-Poli A, et al. Troponin I serum concentration: a new marker of left ventricular hypertrophy in patients with essential hypertension. Ital Heart J 2000;1:532–5. [8] McKee PA, Castelli WP, McNamara PM, et al. The natural history of congestive heart failure: the Framingham study. N Engl J Med 1971;285:1441–6. [9] Alpert JS, Thygesen K, Antman E, et al. Myocardial infarction redefined--a consensus document of The Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol 2000;36:959– 69. [10] Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and limit leads. Am Heart J 1949;37:752–8.
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201 [11] Heupler S, Lauer M, Williams MJ, et al. Increased left ventricular cavity size, not wall thickness, potentiates myocardial ischemia. Am Heart J 1997;133:691–7. [12] Houghton L, Frank MJ, Carr AA, et al. Relations among impaired coronary flow reserve, left ventricular hypertrophy and thallium perfusion defects in hypertensive patients without obstructive coronary artery disease. J Am Coll Cardiol 1990;15:43– 51. [13] DePuey EG, Guertler-Krawczynska E, Perkins JV, et al. Alterations in myocardial thallium-201 distribution in patients with chronic systemic hypertension undergoing single-photon emission computed tomography. Am J Cardiol 1988;62:234–8. [14] Salcedo EE, Marwick TM, Korzick DH, et al. Left ventricular hypertrophy sensitizes the myocardium to the development of ischemia. Eur Heart J 1990;11:72–8.
201
[15] Van Eyk JE, Powers F, Law W, et al. Breakdown and release of myofilament proteins during ischemia and ischemia/reperfusion in rat hearts: identification of degradation products and effects on the pCa-force relation. Circ Res 1998;82:261–71. [16] Feng J, Schaus BJ, Fallavollita JA, et al. Preload induces troponin I degradation independently of myocardial ischemia. Circulation 2001;103:2035–7. [17] Waller TA, Hiser WL, Capehart JE, et al. Comparison of clinical and morphologic cardiac findings in patients having cardiac transplantation for ischemic cardiomyopathy, idiopathic dilated cardiomyopathy, and dilated hypertrophic cardiomyopathy. Am J Cardiol 1998;81:884–94.
Etiology of troponin I elevation in patients with congestive heart failure and low clinical suspicion of myocardial infarction George O. Angheloiu a , Reginald P. Dickerson b , Keyvan Ravakhah b,∗ b
a Department of Cardiovascular Medicine, The Cleveland Clinic Foundation, Cleveland, OH, USA Department of Medicine, Huron Hospital, Cleveland Clinic Health System, Cleveland, OH44112, USA
Received 9 February 2004; received in revised form 26 May 2004; accepted 26 May 2004
Abstract Background: Increased troponin I (TnI) has been detected in congestive heart failure (CHF) patients. We demonstrate that occurrence of LVH, and systolic and diastolic blood pressure values (SBP and DBP, respectively) correlate with increasing levels of serum TnI in a selected CHF population. Methods: Our team reviewed the medical records of 40 consecutive patients admitted with CHF symptoms, low clinical suspicion of myocardial infarction (MI), and serum TnI measured upon admission. Twenty numerical and categorical parameters were screened. Patients were divided in to seven non-exclusive groups by increasing cutoff values of serum TnI (0, 0.1, 0.2, 0.3, 0.4, 0.6, and 1.5 ng/ml). In each group the average value of numerical parameters and the occurrence of categorical ones were calculated. The correlation between these values and the seven TnI cutoffs were analyzed. Results: Occurrence of electrocardiographic left ventricular hypertrophy (ECG-LVH), and average values of SBP and DBP significantly correlated with the TnI cutoffs by means of multiple regression analysis (P = 0.001, 0.02 and 0.007, respectively). Conclusion: ECG-LVH and values of SBP and DBP significantly correlate with increasing cutoff levels of serum TnI in hospitalized patients with CHF and low clinical suspicion of MI. © 2004 Elsevier Ireland Ltd. All rights reserved. Keywords: Troponin I; Congestive heart failure; Left ventricular hypertrophy
Resumo Contexto: Tem sido detectado um aumento de troponina I (Tn I) em doentes com Insuficiˆencia Card´ıaca Congestiva (CHF). Demonstramos que a ocorrˆencia de LVH, e os valores de pressão arterial sistólica e diastólica (SBP e DBP, respectivamente) correlacionam-se com n´ıveis elevados de Tn I numa população seleccionada com CHF. Métodos: A equipa reviu os registos médicos de 40 doentes admitidos consecutivamente com sintomas de CHF, baixa suspeição cl´ınica de enfarte do miocárdio (MI), e TnI sérica medida na admissão. Foram revistos vinte parˆametros numéricos e categóricos. Os doentes foram divididos em sete grupos não exclusivos aumentando os valores de cutoff da TnI sérica (0, 0.1, 0.2, 0.3, 0.4, 0.6 e 1.5 ng/ml). Em cada grupo foi calculado o valor médio dos parˆametros numéricos e a ocorrˆencia de valores categóricos. Foram analizados a correlação entre estes valores e os sete cutoffs de TnI. Resultados: A ocorrˆencia de hipertrofia ventricular esquerda electrocardiográfica (ECG-LVH), e valores médios de SBP e DBP correlacionam-se de forma significativa com os cutoffs por meios de análise de regressão múltipla (P = 0.001, 0.02 e 0.007, respectivamente). Conclusão: ECG-LVH e os valores de SBP e DBP correlacionam-se significativamente com n´ıveis de cutoff elevados de TnI em doentes hospitalizados com CHF e baixa suspeição cl´ınica de MI. © 2004 Elsevier Ireland Ltd. All rights reserved. Palavras chave: Troponina I; Insuficiˆencia Card´ıaca Congestiva; Hipertrofia Ventricular Esquerda
Resumen Antecedentes: Se ha detectado troponina I (TnI) aumentada en pacientes con insuficiencia card´ıaca congestiva. Demostramos que la ocurrencia de hipertrofia ventricular izquierda (LVH), valores de presión sistólica y diastólica (SBP yDBP, respectivamente) se correlacionan con niveles crecientes de TnI en una población seleccionada con CHF. Métodos: nuestro equipo revisó los registros médicos de 40 pacientes
∗ Corresponding author. E-mail address: [email protected] (K. Ravakhah).
0300-9572/$ – see front matter © 2004 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.resuscitation.2004.05.018
196
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
consecutivos admitidos con s´ıntomas de CHF, y escasa sospecha cl´ınica de infarto de miocardio (MI), y se midió TnI sérica al ingreso. Se midieron 20 parámetros numéricos y categóricos. Los pacientes se dividieron en 7 grupos no exclusivos por valores de corte crecientes de TnI sérico (0, 0.1, 0.2, 0.3, 0.4, 0.6 y 1.5 ng/ml). En cada grupo se calculó el valor promedio de los parámetros numéricos y la ocurrencia de los categóricos. Se analizó la correlación entre estos valores y los 7 cortes en TnI. Resultados: La ocurrencia de LVH electrocardiográfica, y valores promedio de SBP y DBP se correlacionaron significativamente con los cortes de TnI por medio de análisis de regresión múltiple (P = 0.001, 0.02 y 0.007, respectivamente). Conclusión: LVH en el ECG y los valores de SBP y DBP se correlacionan significativamente con cortes crecientes en TnI sérico en pacientes hospitalizados con CHF y escasa sospecha de MI © 2004 Elsevier Ireland Ltd. All rights reserved. Palavras chave: Troponina I; Falla cardiaca congestiva; Hipertrofia ventricular izquierda
Troponin I (TnI) is an important diagnostic and prognostic marker in patients with acute coronary syndromes [1,2]. Investigational evidence indicates that levels of TnI greater than the normal range are present in patients with congestive heart failure (CHF) [3], could be associated with etiologies other than acute coronary syndromes and predict negative outcomes in patients with this diagnosis [4,5]. In the same time left ventricular remodeling was found to be a cause for elevated TnI in patients with severe non-ischemic CHF [6], and TnI above upper normal limit of 0.5 ng/ml was associated with left ventricular hypertrophy in outpatient hypertensive patients [7]. In the present study we investigated what clinical or laboratory variables could correlate with increasing serum TnI cutoff values in patients admitted with a CHF diagnosis [8], and low clinical suspicion of myocardial infarction (MI).
ria [10]. LV ejection fraction was considered abnormal for values lower than 40%. To analyze the relative contribution to the TnI variance of all 20 parameters recorded, of both numerical and categorical type, we divided the patients investigated into seven non-exclusive groups in a step-wise manner, using increasing cutoff values of serum TnI. Each group included all patients with TnI levels greater than or equal to a particular cutoff: 0 (n = 40, all patients), 0.1 (n = 24), 0.2 (n = 19), 0.3 (n = 15), 0.4 (n = 9), 0.6 (n = 6) and 1.5 ng/ml (n = 4). For every categorical parameter studied we calculated its occurrence in each group of patients defined as above, while for numerical parameters we calculated their average value. Linear regression analysis was performed between the occurrence and, respectively, average values of the 20 parameters in these seven groups and the TnI cutoffs that defined the groups. To find which parameters contributed independently to the serum TnI variance, we performed multiple regression analysis using the variables that
2. Methods
Table 1 Parameters recorded in the study
1. Introduction
We define low clinical suspicion of myocardial infarction as the absence of chest pain, ischemic electrocardiographic findings (ST segment elevation or depression of 0.1 mV or greater in two adjacent leads), and previous coronary artery interventions [9]. We reviewed the charts of 40 consecutive patients with this clinical picture and CHF symptoms admitted between August and December 1999 in a network of three community teaching hospitals, after obtaining the necessary IRB approval. We studied 20 clinical and laboratory parameters recorded in the first 48 h after admission, of both numerical and categorical type, as listed in Table 1. Two-dimensional echocardiogram was performed in all patients and cardiac catheterization in three. Serum TnI was measured using a Dade-Behring RXL Dimension instrument (Dade Behring, Deerfield, IL) at least three times at 4-h interval, starting upon admission. The lower reference limit of this instrument is 0.1 ng/ml. Peak serum TnI value was considered to be the highest value measured in each patient, and all statements made in this study refer to this value. To diagnose electrocardiographic left ventricular hypertrophy (ECG-LVH) we used the Sokolow-Lyon QRS voltage crite-
Parameter group
Parameter
Demographic
Gender Race Age H/o diabetes H/o hypercholesterolemia H/o hypertension H/o smoking H/o stroke H/o PVD H/o aortic stenosis Blood pressure valuea LV wall motion abnormalities LV inter-ventricular septum LV posterior wall LV ejection fraction LV diameter ECG-LVH Pulmonary edema Creatinine Hemoglobin
Clinical
2D-echo
ECG Chest X-ray Blood tests
ECG, electrocardiogram; ECG-LVH, electrocardiographic left ventricular hypertrophy; H/o, history of; LV, left ventricle. a Recorded as the initial value in the emergency department, or on the floor in the case of direct admissions.
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
positively correlated with the TnI cutoff values by linear regression. Subsequently, we tested whether there was a direct correlation between the individual numerical parameters that had an independent contribution to the TnI variance and the correspondent individual serum TnI levels. An arbitrary value of 0 ng/ml was used for TnI in all cases where TnI was lower than 0.1 ng/ml. For categorical parameters that had an independent contribution to the TnI variance, we felt necessary to compare the occurrence of these parameters between patients with TnI levels greater than each of the TnI cutoff analyzed and patients with TnI lower than the same cutoff values. Comparisons were performed for all TnI cutoffs analyzed with the exception of 0 ng/ml. 3. Statistics Linear and multiple regression analysis for correlation characterization were used as detailed above, and between-groups comparison was performed by means of Fisher exact test. P-values lower than 0.05 were considered significant. 4. Results Electrocardiogram and two-dimensional echocardiography were performed in all 40 patients, and cardiac catheterization in 3. Serum TnI levels of patients included in the study are depicted in Fig. 1. There were significant correlations between increasing serum TnI cutoff values (0, 0.1, 0.2, 0.3, 0.4, 0.6, and 1.5 ng/ml) and several parameters as illustrated in Table 2. The most significant correlations were documented with occurrence of ECG-LVH and African-American race,
197
Table 2 Results of linear regression between serum TnI cutoffs and 13 parameters investigated Parameter
Pearson’s r-value
P-value
African-American race History of diabetes Hypercholesterolemia Smoking Creatinine SBP DBP ECG-LVH Pulmonary edema IVS LV ejection fraction LV end-diastolic diameter Patients with WMA
0.94 −0.81 0.57 0.83 0.81 0.99 0.99 0.95 0.87 0.92 0.06 0.01 –0.80
<0.001 –a 0.17 0.012 0.018 <0.001 <0.001 <0.001 0.005 0.002 0.58 0.83 –a
DBP, diastolic blood pressure; ECG-LVH, electrocardiographic left ventricular hypertrophy; IVS, interventricular septum; SBP, systolic blood pressure; WMA, wall motion abnormalities. a No P-value was calculated for negative Pearson’s factor, since this would imply a counter-physiological correlation between increased TnI values and the parameters analyzed.
and average systolic (SBP) and diastolic blood pressure (DBP; < 0.001). No positive correlation was detected between serum TnI levels and the presence of increased LV end-diastolic diameter, low ejection fraction or wall motion abnormalities. When introduced in a multiple regression algorithm, ECG-LVH occurrence and average SBP and DBP had independent contributions to serum TnI variance (P = 0.001, 0.02, and 0.007, respectively). Fig. 2 illustrates the linear regression relation between increasing TnI cutoff values and these three parameters. Individual TnI values correlated with correspondent individual SBP and DBP values as illustrated in Fig. 3 (P = 0.0004 and 0.006, respectively). Occurrence of ECG-LVH was significantly greater in five groups of patients with high TnI values in comparison to the correspondent groups of patients with TnI lower than the cutoffs analyzed (Table 3). Three patients of the four with serum TnI equal to 1.5 ng/ml or greater had a cardiac catheterization performed. Two of them did not present atherosclerotic lesions, and a third one had a 50% stenosis in a secondary obtuse marginal branch. A total of nine patients (23%) had wall motion abnormalities on echocardiogram, with five among the patients with serum TnI values of 0.1 ng/ml or greater (21%). We performed a second analysis to test the correlation between serum TnI values and the parameters investigated after excluding these nine patients, and highly significant correlations persisted with ECG-LVH occurrence and average SBP and DBP (P = 0.008, 0.004, and 0.002, respectively). 5. Discussion
Fig. 1. Scatter plot of peak serum troponin I values in 40 hospitalized CHF patients with low suspicion of myocardial infarction.
In the present study we investigated the possible existence of clinical or laboratory parameters that would corre-
198
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
Fig. 2. Linear regression correlation of serum TnI cutoff with occurrence of electrocardiographic left ventricular hypertrophy (ECG-LVH) (a), and average values of systolic (SBP) (b) and diastolic blood pressure (DBP) (c) in groups of patients defined by increasing TnI cutoff values.
late with elevated TnI values in hospitalized CHF patients with low clinical suspicion of MI. In order to perform a complete search we studied 20 variables of both numerical and categorical nature (Table 1). To be able to compare both types of parameters we took a step-wise analysis approach in the first part of our investigation, and identi-
fied which parameters contribute in an independent manner to the TnI variance. We subsequently confirmed the resulting parameters, either by correlating the numerical variables with the individual TnI values, or by comparing the occurrence of categorical parameters in groups of patients with TnI greater than the cutoffs analyzed versus the
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
199
Fig. 3. Linear regression correlation between individual TnI levels and correspondent individual values of systolic (SBP) (a) and diastolic blood pressure (DBP) (b). Table 3 Comparison of ECG-LVH occurrence between groups of patients with TnI levels greater than or equal to six TnI cutoffs and correspondent groups with TnI lower than the same values TnI cutoff (ng/ml)
1.5 0.6 0.4 0.3 0.2 0.1
Patients with TnI ≥ cutoff
Patients with TnI < cutoff
n
ECG-LVH occurrence
n
ECG-LVH occurrence
4 6 9 14 19 23
4 4 6 8 8 9
36 34 31 26 21 17
7 7 5 3 3 2
(100%) (67%) (67%) (57%) (42%) (39%)
(19%) (21%) (16%) (12%) (15%) (12%)
P-value
0.003 0.04 0.007 0.004 0.05 0.06
ECG-LVH, electrocardiographic left ventricular hypertrophy; n, number of patients; TnI, troponin I.
correspondent groups with TnI levels lower than the same values. Our results indicate that occurrence of ECG-LVH in CHF patients with low clinical suspicion of myocardial infarction independently correlated with increasing TnI cutoff values (Fig. 2). We also demonstrated that ECG-LVH occurrence was significantly higher in groups of patients with TnI
greater than or equal to 1.5, 0.6, 0.4, 0.3, and 0.2 ng/ml in comparison to the correspondent groups of patients with TnI lower than these values (Table 3). This comparison had a strong tendency for being significant at the TnI cutoff value of 0.1 ng/ml (Table 3). This findings are in agreement with two previous reports, which argued that left ventricular remodeling was a cause of
200
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201
elevated TnI in patients with severe non-ischemic CHF [6], and that elevated TnI was associated with left ventricular hypertrophy in outpatient hypertensive patients [7]. Several clinical and experimental studies could suggest a causal relationship between LVH and elevated serum TnI levels. Four clinical studies performed in patients without coronary artery disease showed that left ventricular mass correlated with the extent of coronary flow reserve impairment, and of ischemia as measured by exercise thallium perfusion imaging [11–14]. An animal study by Van Eyk et al. [15] showed that global ischemia can result in TnI breakdown and release. In light of these studies, the association between ECG-LVH and elevated TnI can be explained by a link between LVH, global myocardial ischemia, cardiac TnI breakdown, and resulting release of serum TnI. We also determined that average SBP and DBP in groups of patients defined by increasing TnI cutoffs, independently, correlated with the TnI cutoff values used (Fig. 2). At the same time individual SBP and DBP values recorded in each patient significantly correlated with the correspondent individual TnI levels (Fig. 3). An explanation for these results is suggested by an experimental study by Feng et al. [16], where elevated LV preload resulted in TnI degradation. It is reasonable to think that an increase in afterload induced by high blood pressure might have as end-result a similar process of TnI degradation. This inference makes intuitively logical the association between elevated values of SBP and DBP and increased TnI. There was no positive correlation between elevated serum TnI and the occurrence of low ejection fraction, increased LV end-diastolic diameter, or wall motion abnormalities (Table 2), although the small number of patients included in the study may have prevented these associations to occur. We made considerable efforts to insure that acute coronary events were not a cause of TnI elevation in our study. Patients were included in the study only in the absence of chest pain, ECG ischemic changes, and previous coronary artery interventions [9]. Three of the four patients with TnI greater than 1.5 ng/ml were subjected to cardiac catheterization, and no or trivial coronary artery disease was detected. All patients had a two-dimensional echocardiogram performed. There were a total of nine patients with wall motion abnormalities, with five among patients with serum TnI of 0.1 ng/ml or greater. Although numerous studies documented the presence of asynergic myocardial regions in patients with non-ischemic dilated cardiomyopathy [17], wall motion abnormalities are generally considered a sign of coronary artery disease. To insure that our results were not biased by the eventual presence of coronary artery disease, we excluded in a second analysis the nine patients with wall motion abnormalities. Highly significant correlations between TnI levels and the presence of ECG-LVH and values of SBP and DBP persisted even after these patients were excluded. Nonetheless, temporary intra-coronary thrombi caused by non-obstructive coronary plaques could be reason for short-lived release of serum TnI, and neither echocardiography nor angiography,
the methods employed in our study, could detect this type of atherosclerotic disease. Our study suggests that ECG-LVH occurrence, and SBP and DBP values independently correlate with increasing TnI cutoff levels in hospitalized patients with CHF symptoms and low clinical suspicion of myocardial infarction. Increased serum TnI is a highly specific and sensitive marker for myocardial damage [9], and early identification of its etiology could be paramount in adopting the proper inpatient medical management and therapy. Caution should nevertheless be exercised when interpreting these results, since a limited number of patients were included in the study, and the work directed towards ruling out coronary atherosclerosis was rather not exhaustive. Larger studies, eventually accompanied by a more invasive strategy of excluding coronary artery disease, should follow up.
Acknowledgements We gratefully acknowledge Drs. Burton C West, Hsien Wang, Michael Kalus and Robert E. Botti for valuable suggestions and technical support, and Mrs. Joyce Callahan for editorial assistance.
References [1] Antman EM, Tanasijevic MJ, Thompson B, et al. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. N Engl J Med 1996;335:1342–9. [2] D’Costa M, Fleming E, Patterson MC. Cardiac troponin I for the diagnosis of acute myocardial infarction in the emergency department. Am J Clin Pathol 1997;108:550–5. [3] Missov E, Calzolari C, Pau B. Circulating cardiac troponin I in severe congestive heart failure. Circulation 1997;96:2953–8. [4] Khan IA, Tun A, Wattanasauwan N, et al. Elevation of serum cardiac troponin I in noncardiac and cardiac diseases other than acute coronary syndromes. Am J Emerg Med 1999;17:225–9. [5] Horwich TB, Patel J, MacLellan WR, et al. Cardiac troponin I is associated with impaired hemodynamics, progressive left ventricular dysfunction, and increased mortality rates in advanced heart failure. Circulation 2003;108:833–8. [6] Logeart D, Beyne P, Cusson C, et al. Evidence of cardiac myolysis in severe nonischemic heart failure and the potential role of increased wall strain. Am Heart J 2001;141:247–53. [7] Siciliano M, Mettimano M, Dondolini-Poli A, et al. Troponin I serum concentration: a new marker of left ventricular hypertrophy in patients with essential hypertension. Ital Heart J 2000;1:532–5. [8] McKee PA, Castelli WP, McNamara PM, et al. The natural history of congestive heart failure: the Framingham study. N Engl J Med 1971;285:1441–6. [9] Alpert JS, Thygesen K, Antman E, et al. Myocardial infarction redefined--a consensus document of The Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol 2000;36:959– 69. [10] Sokolow M, Lyon TP. The ventricular complex in left ventricular hypertrophy as obtained by unipolar precordial and limit leads. Am Heart J 1949;37:752–8.
G.O. Angheloiu et al. / Resuscitation 63 (2004) 195–201 [11] Heupler S, Lauer M, Williams MJ, et al. Increased left ventricular cavity size, not wall thickness, potentiates myocardial ischemia. Am Heart J 1997;133:691–7. [12] Houghton L, Frank MJ, Carr AA, et al. Relations among impaired coronary flow reserve, left ventricular hypertrophy and thallium perfusion defects in hypertensive patients without obstructive coronary artery disease. J Am Coll Cardiol 1990;15:43– 51. [13] DePuey EG, Guertler-Krawczynska E, Perkins JV, et al. Alterations in myocardial thallium-201 distribution in patients with chronic systemic hypertension undergoing single-photon emission computed tomography. Am J Cardiol 1988;62:234–8. [14] Salcedo EE, Marwick TM, Korzick DH, et al. Left ventricular hypertrophy sensitizes the myocardium to the development of ischemia. Eur Heart J 1990;11:72–8.
201
[15] Van Eyk JE, Powers F, Law W, et al. Breakdown and release of myofilament proteins during ischemia and ischemia/reperfusion in rat hearts: identification of degradation products and effects on the pCa-force relation. Circ Res 1998;82:261–71. [16] Feng J, Schaus BJ, Fallavollita JA, et al. Preload induces troponin I degradation independently of myocardial ischemia. Circulation 2001;103:2035–7. [17] Waller TA, Hiser WL, Capehart JE, et al. Comparison of clinical and morphologic cardiac findings in patients having cardiac transplantation for ischemic cardiomyopathy, idiopathic dilated cardiomyopathy, and dilated hypertrophic cardiomyopathy. Am J Cardiol 1998;81:884–94.