Brain natriuretic peptide and other cardiac markers predicting left ventricular remodeling and function two years after myocardial infarction

Summary. Background. Left ventricular remodeling is a complex pathologic process of progressive left ventricular dilatation, leading to dysfunction and heart failure in patients after myocardial infarction. Objective. To evaluate biochemical markers, reflecting cardiac remodeling process after first myocardial infarction and compare those markers with clinical characteristics of left ventricular remodeling. Material and methods. Brain natriuretic peptide, troponin I, creatine kinase, creatine kinase MB mass, lactate dehydrogenase levels were measured in 30 patients with acute myocardial infarction on days 1, 2, 3–7 . Brain natriuretic peptide was measured at 3 months, 6 months, and 2 years after myocardial infarction. Echocardiographic parameters of left ventricular remodeling were determined in acute phase (day 1–3), at 3 months, 6 months, and 2 years after MI. Results. In acute phase, brain natriuretic peptide level progressively increased according to worsening of left ventricular geometry: in normal left ventricle geometry group, brain natriuretic peptide level was 84.1 (58.7–121) pg/mL, in concentric remodeling group – 125 (69.2–165) pg/mL, in concentric hypertrophy group – 128 (74–368) pg/mL, and in eccentric hypertrophy group – 470 (459–494) pg/mL, P=0.02. Patients who had increased left ventricular end diastolic diameter index during 2-year period had higher brain natriuretic peptide level in the acute phase (584 (249–865) pg/mL vs. 120 (67–202) pg/mL, P=0.04) and also higher peak lactate dehydrogenase and troponin I levels. Conclusions. Brain natriuretic peptide level in acute phase of myocardial infarction is strongly associated with the markers of myocardial injury and related to left ventricular geometry changes and remodeling. Brain natriuretic peptide together with troponin I levels in acute phase of myocardial infarction might be useful in predicting subsequent cardiac function.


Introduction
Left ventricular (LV) remodeling is a complex of pathological processes leading to progressive LV dilatation, dysfunction, and heart failure in patients with myocardial infarction (MI). Remodeling changes begin in the acute period of MI and are progressive and occasionally may continue long after (1). Therefore, early detection of LV remodeling is important in treatment of patients with acute MI. Biochemical markers, which directly reflect or indirectly impact the remodeling process, could be used along with the clinical characteristics for the risk of LV remodeling stratification.
Brain natriuretic peptide (BNP) is a cardiac neurohormone that is synthesized in ventricular myocardium and released in response to increased LV wall stress.
Its diverse actions include natriuresis, vasodilatation, inhibition of the rennin-angiotensin-aldosterone system, and inhibition of sympathetic nervous system activity (2). Recently, BNP has been used as a marker of LV dysfunction (3). The level of plasma BNP also increases in an early phase of acute MI and might be used as a predictor of the prognosis after MI (4). In addition, recent data suggest that BNP is not only indicator of impaired left ventricular function; it also might be a predictor of LV remodeling (5).
The aim of this study was to determine and evaluate biochemical markers of myocardial injury that have the highest impact on LV remodeling process and function as long as 2 years after the first MI and to compare those markers with clinical characteristics of LV function and remodeling.

Methods
Patient population. We studied 30 patients with acute coronary syndrome admitted to the Department of Cardiology, Kaunas University of Medicine Hospital. Only the patients with the first suspected MI and without previous percutaneous transluminal coronary angioplasty (PTCA), coronary stenting, coronary artery bypass graft operation, or electrocardiostimulation were included in the study. The diagnosis of MI was established according to WHO recommended criteria. All patients underwent clinical and ECG examination.
The investigation conforms to the principles outlined in the Declaration of Helsinki. Kaunas Regional Ethics Committee for biomedical research confirmed the study protocol. Informed consent was obtained from all subjects participating in the present study.
Study design and measurements. Concentrations of troponin I (TnI), BNP and creatine kinase isoenzyme MB mass (CK-MB) and as well as total activity of serum lactate dehydrogenase (LDH) and creatine kinase (CPK) were measured in acute phase on days 1, 2, 3, and 7. Patients were followed up over a period of 2 years. BNP concentration was measured at 3 and 6 months and 2 years after MI.
Blood samples for the measurement of BNP were obtained by direct venipuncture of antecubital vein in vacutainers containing EDTA, after patient's rest for at least 30 min. Immediately, the whole blood specimens were measured by immunofluorescence BNP assay (Triage BNP, Biosite Diagnostics), which can measure BNP from 5 pg/mL to 1300 pg/mL. The upper limit of normal range, suggested by manufacturer, is 50 pg/ml, although a concentration higher than 100 pg/mL suggests the diagnosis of heart failure (3). Tn I and CK-MB levels were also measured in whole blood specimens by immunofluorescence assay (Triage Cardiac, Biosite Diagnostics). The upper limits of normal ranges suggested by manufacturer were 1.0 ng/mL for Tn I and 4.3 ng/mL for CK-MB.
Echocardiography was performed in all patients in acute phase (1-3 day) as well as 3 and 6 months and 2 years after MI using Hewlett-Packard echocardiograph "Sonos 5500." All estimations were performed according to the criteria of the American Society of Echocardiography. The level of LV remodeling was assessed for all patients. LV geometry was considered as normal when the LV mass index and the relative wall thickness were within the normal rates, the concentric remodeling -when the LV mass index was normal, but the relative wall thickness exceeded 0.45, the concentric hypertrophy -when both the LV mass index and the relative wall thickness were increased, and the eccentric hypertrophy -when the LV mass index and the size of LV chamber were increased, but the relative wall thickness remained normal (6). The LV wall motion score index (WMI) was estimated dividing the LV segmental contraction score sum by the number of the segments. Each segment was assigned a score, based on its contractility as assessed visually: normal -1; hypokinesis -2; akinesis -3; dyskinesis -4; and aneurysm -5. LV ejection fraction (EF) was obtained from the apical four chamber and two chamber views and evaluated according to Simpson's method. Systolic dysfunction was defined when LV ejection fraction was less than 40%.
Statistical analysis. Continuous variables were described as mean±SD, and plasma concentrations of BNP were described as the median and interquartile range (IQR). Logarithmic transformation was performed to achieve normal distribution, because BNP concentration data were skewed. Pearson's correlation coefficient between logarithmically transformed BNP and variable parameters was calculated. Group comparisons were made by use of Student's t tests for independent samples and, if data were not distributed normally, using Kruskal-Wallis rank-sum test. The diagnostic utility of BNP was compared with the echocardiographic probability of LV dysfunction using the receiver-operating characteristic (ROC) curves. Multiple regression analysis with a stepwise procedure was used to model BNP as a function of covariables. Logistic regression was used in a univariate and a multivariate approaches for evaluating the ability of cardiac markers to identify LV dysfunction at 2 years after MI. A probability value of <0.05 was taken as the level of statistical significance. Statistical analysis was facilitated by the use of Statistica (Statsoft) software package.

Results
Basic clinical characteristics of the patients are summarized in Table 1. PTCA was successful in 19 patients, and in 11 patients, it was unsuccessful or not performed. All the patients received ACE inhibitors; other drugs were prescribed according to the individual indications. During the 2-year follow-up period, two patients have died (one patient due to sudden cardiac death and one of lung cancer), and three patients were lost because of noncompliance.
BNP and LV remodeling characteristics BNP concentration during acute phase of MI was found being related to LV geometry. BNP concentration on day 7 correlated best with LV geometry to compare with BNP on day 1 and day 2. Plasma BNP concentration on day 7 progressively increased with worsening of LV geometry: patients with normal LV geometry had the lowest BNP concentration -84.1 (58.7 -121) pg/mL and in the patients with concentric remodeling BNP concentration was 125 (69.2-165) pg/mL, with concentric hypertrophy -128 (74-368) pg/ml, and with eccentric hypertrophy -470 (459-494) pg/mL (Fig. 1). LV geometry at 3 months after MI was related to BNP concentration determined at this time, but not to BNP level in acute phase. In patients with normal LV BNP -brain natriuretic peptide; IQR -interquartile range; LV -left ventricular; MI -myocardial infarction; -PTCA -percutaneous transluminal coronary angioplasty; TnI -troponin I. geometry, BNP concentration was 84.6 (34.8-132) pg/mL, with concentric remodeling -37.9 (17.3-64.7) pg/mL, with concentric hypertrophy -84.5 (27-117) pg/mL, and with eccentric hypertrophy -134 (71.7-1130) pg/mL. BNP concentration was significantly higher only in eccentric hypertrophy group to compare with concentric remodeling group (P=0.04). In later period, 6 months and 2 years after MI, LV geometry was not associated with BNP concentration determined in acute phase of MI or 6 months or 2 years after MI.

LV dysfunction and cardiac markers
We tested the strength of the link between BNP concentration and LVEF in acute MI and in later period after MI. At all time points, BNP levels correlated strongly with LVEF ( Table 3). The logarithmically transformed maximal BNP levels in acute MI were associated significantly not only with LV ejection fraction in acute phase, but also with LVEF at 2 years after MI (Fig. 2). The relation between the common

BNP and markers of myocardial injury
We evaluated BNP links with other cardiac markers of myocardial injury. In acute MI, BNP levels correlated with LDH activity, TnI, and CK-MB concentrations on day 2 as well as on day 7. Meanwhile, BNP levels measured later after MI were associated more strongly with cardiac marker TnI levels measured on day 7 than on day 2 ( Table 4).
The results of multiple linear regression model showed that BNP concentration on admission, TnI concentration on day 7, and peak concentration of CK-MB had the strongest influence on BNP concentration at 6 months after MI (F=28.0, R 2 =0.97, P<0.01). In another model, where echocardiographic parameters were included, the variables with adverse remodeling characteristics had the strongest influence on BNP concentration at 6 months after MI WMI in the acute phase of MI, LVEDD after 6 months, WMI after 6 months, and LV geometry after 6 months (F=10.5, R 2 =0.7, P<0.001). When biochemical and echocardiographic parameters were included in the model, BNP concentration on day 1 (or on day 7) and TnI concentration on day 7 revealed the strongest association with BNP concentration after 6 months.

Discussion
The level of plasma BNP depends on the equilibrium between myocardial secretion as compensatory response to injury or wall stress and an amount and activity of expressed guanylyl cyclase-type BNP receptors, BNP clearance receptors, and also peripheral degradation rate of BNP through neutral endopeptidases. According to the data of multiple studies, serial factors such as age, gender, rheumatic valve disease, anemia, renal dysfunction, thyroid dysfunction, fasting and weight loss, some medicine could influence the elevation of plasma BNP levels (7). Though, it is established that BNP levels are associated with severity of heart failure (HF) and prognosis, but little is known about the release of BNP into plasma in early stages of LV remodeling and HF development after myocardial infarction. Considering this aspect, the causative The surrogate markers -LV ejection fraction, ventricular volumes, and mass -are often used for evaluation of the LV remodeling. The development and progression of HF is associated with LV remodeling, which manifests as gradual increase in LVED and LVES volumes, LV mass, wall thinning, changes in chamber geometry to a more spherical and less elongated shape (8).
The data from our study show that BNP concentration is increased during acute MI and correlates with LV remodeling, occurring after the first MI. Higher BNP concentrations in the acute phase of myocardial infarction as well as late BNP concentrations at 3 months and at 2 years after MI are linked to the increase of heart chamber size during 2-year period after MI.
BNP release appeared to be directly proportional to ventricular volume expansion and pressure overload. It is known that BNP is over expressed in hypertrophied myocardium. However, Takeuchi et al. (2005) have shown that LV overload sustaining hemodynamic status, which could influence the ventricular wall stress in cases of hypertrophic obstructive cardiomyopathy, was not associated with BNP levels (9).
Left ventricular dilatation as one of the characteristics of LV remodeling is important in predicting reduced survival after myocardial infarction. Therefore, early detection of LV remodeling is substantial for the treatment of patients after acute MI. Hence, the assessment of BNP concentration in the acute phase of myocardial infarction would be useful in providing essential information regarding prospective LV remodeling (10,11). LV remodeling after MI, associated with the changes which occur early (during the first days) and later (after few months), are not fully investigated.
Our findings are consistent with the results of other previously published studies about relation of BNP and LV remodeling (10)(11)(12). Nagaya et al. have found that BNP concentration on day 7 correlated with the increase in LV diastolic volume index through the first 30 days of acute MI, and that patients in LV remodeling group had higher BNP levels from day 2 to day 90 than patients in non-remodeling group (10). Meanwhile, Yoshitomi et al. have shown that BNP concentrations after 1 month are increased in patients, who are likely to develop a progressive LV dilatation in the first 3 months after MI (11). In our study, we have shown that LV remodeling is linked with both early and late BNP levels. However, in the study population, the LV geometry changes were not so prominent, and therefore correlation between BNP concentration and LV remodeling parameters was weaker.
In the present study, we have shown that BNP levels are relative to cardiac function subsequent later after MI. Higher BNP levels in acute phase of myocardial infarction were associated with worse EF after 6 months and 2 years. We have also found a significant relation between BNP concentration in acute phase of myocardial infarction and WMI after 3 months and 6 months as well as 2 years after MI. Crilley et al. have reported a strong association between BNP concentration and WMI as a marker of deterioration of left ventricular function after acute myocardial infarction (13). Recently, Mutlu et al. have shown that predischarge troponin T (TnT) levels have prognostic importance in acute myocardial infarction (14). Patients with higher TnT levels on day 7 had worse diastolic and systolic LV function after one month. The results from our study show that TnI concentration measured on day 7 is correlated with LV ejection fraction later after MI: at 3 months, 6 months, and 2 years after MI. Meanwhile,  (9) neither TnI concentration on admission nor TnI peak concentration correlated with the LV function after two years.
In the previous work, we have shown that BNP concentrations during acute myocardial infarction are strongly related to the markers of myocardial necrosis, reflecting the extent of injured myocardium, and to the degree of acute heart failure (15). Now we have confirmed these findings and also demonstrated that BNP together with other cardiac markers have additional information on predicting left ventricular remodeling. During acute MI, BNP levels correlated strongly not only with cardiospecific Tn I, but more surprisingly, also with less specific marker of myocardial injury -total activity of LDH. Recently, it has been shown that namely LDH cumulative release strongly predicts left ventricular function after one year after acute myocardial infarction (16).
It is worth to notice that unlike markers of myocardial necrosis, BNP is not considered to be released from necrotizing myocardium. Experimental studies have shown that BNP secretion and BNP mRNA expression are increased mainly in the borderline region between the infarcted and noninfarcted regions (17). The stimulus for this appears to be increased wall stress directly related to the infarction. However, less severe clinical ischemia, insufficient to result in extensive necrosis, is also associated with release of BNP (18). Recent experimental studies show that ischemia itself, rather than changes in wall stress secondary to ischemia, might promote BNP release (19).
Thus, the major finding of our work is that BNP level following MI appears to be associated not only with LVEF impairment but also with the extent of myocardial damage as assessed by cardiac markers, supporting the opinion that elevation of BNP is a general indicator of reduced cardiac performance.

Study limitations
We had possibilities to study only a small number of patients, thus there was insufficient statistical power in some analysis, and our findings should be confirmed in larger studies. Some limitations of this study, in addition to those discussed above, merit consideration. We did not examine changes due to various medications. However, variation of cardiac markers under therapy may dilute its predictive power. In our work, the evaluation of microvascular flow and myocardial perfusion was not performed. Also echocardiographic parameters of LV remodeling and geometry changes are subjected to a greater measurement error in comparison to BNP as a marker of cardiac function.

Conclusions
Brain natriuretic peptide level in acute phase of myocardial infarction is strongly associated with the markers of myocardial injury and related to left ventricular geometry changes and remodeling. Brain natriuretic peptide together with troponin I levels in acute phase of myocardial infarction might be useful in predicting subsequent cardiac function.