High plasma N-terminal pro-brain natriuretic peptide level in patients with diabetes with myocardial infarction and cardiogenic shock

Cardiology Review® Online, June 2006, Volume 23, Issue 6

We examined the association between plasma N-terminal pro-brain natriuretic peptide (Nt-pro-BNP) levels and the occurrence of cardiac events, including cardiogenic shock and mortality, among hospitalized diabetic patients with acute myocardial infarction. Nt-pro-BNP level was shown to be a reliable predictor of outcome in this group of patients.

The pro-brain natriuretic peptide (pro-BNP) is a prohormone secreted from cardiomyocytes in reaction to intensified stress on the myocardial wall. It is cleaved via enzymes into the BNP and the N-terminal part of pro-BNP (Nt-pro-BNP).1 For patients with acute myocardial infarction (MI), left ventricular dysfunction can be assessed using plasma concentrations of Nt-pro-BNP and BNP.2 For patients with acute MI, Nt-pro-BNP may be a better indicator of left ventricular dysfunction than BNP because its increase is greater and its ability to identify early cardiac dysfunction is higher than BNP.3 In recent years, plasma Nt-pro-BNP measurement has been useful in clinical practice among patients with acute MI because it has been reported to provide important predictive data on mortality.4

Because of high vascular risk, patients with diabetes mellitus are frequently referred to intensive care cardiac units following acute MI. Furthermore, they have a markedly greater risk of cardiogenic shock and heart failure after acute MI.5 In patients with diabetes without obvious cardiovascular disease, elevated plasma levels of Nt-pro-BNP have been reported to indicate a greater incidence of asymptomatic left ventricular dysfunction.6 The prognostic value of plasma Nt-pro-BNP following MI, however, has not been studied in patients with diabetes. In a prospective study, we analyzed the association between plasma Nt-pro-BNP measurements and short-term outcomes after MI, including cardiogenic shock and mortality, in patients with diabetes.

Patients and methods

We evaluated 560 patients from the French regional RICO survey, a registry of patients hospitalized for acute MI in eastern France. Acute MI was confirmed by an elevated serum troponin I level. Nt-pro-BNP levels were determined in each patient from admission laboratory values. Patients with a mean fasting blood glucose level ≥7.0 mmol/L on the fourth and fifth hospital days or who had been diagnosed with diabetes prior to admission were considered to have diabetes mellitus. All adverse events during the post-MI hospital stay were recorded, including death, ventricular arrhythmia, recurrent MI, and cardiogenic shock. The diagnosis of cardiogenic shock was made according to the criteria of Hasdai and colleagues: a systolic blood pressure < 90 mm Hg lasting longer than 1 hour even with fluid challenge and clinical indications of hypoperfusion.7

Results

Of the 560 selected patients, 199 (35%) had diabetes. Myocardial infarction location, mean plasma creatine kinase MB level, mean plasma troponin level, and the number of revascularization procedures (coronary artery bypass graft surgery and primary percutaneous coronary intervention) were similar for patients with and without diabetes. As shown in the Table, patients with diabetes were more likely to have had a prior MI. Compared with patients with diabetes, patients without diabetes had higher creatinine clearance and left ventricular ejection fraction (LVEF) measurements and lower glycosylated hemoglobin and fasting blood glucose levels.

Table 1 - Patient characteristics

P

P

P

P

P

P

P

Patients without diabetes had lower plasma Nt-pro-BNP levels than patients with diabetes (130 [range, 49-199] pmol/L compared with 245 [range, 81-777] pmol/L, respectively; < .001). In multivariate analysis, a positive association was shown between Nt-pro-BNP levels and plasma peak troponin level ( < .001), age (= .002), diabetes ( = .005), and female sex ( = .01). A negative association existed between Nt-pro-BNP levels and LVEF ( < .001) and creatinine clearance ( < .001).

P

P

P

P

P

P

Patients without diabetes had a lower in-hospital mortality rate than diabetic patients (3.3% compared with 15.6%; < .001). The incidence of cardiogenic shock was 2.2 times greater among patients with diabetes (17.6%) than among patients without diabetes (7.7%; < .001). No significant differences were noted between patients with and without diabetes with regard to recurrent MI (12.1% vs 8.3%; = .15) or ventricular arrhythmia (12.1% vs 8.9%; = .23). Patients who died during their hospital stay had higher Nt-pro-BNP levels (800 [range, 147-3915] pmol/L compared with 143 [range, 55-357] pmol/L, respectively; < .001). Patients who experienced in-hospital cardiogenic shock also had higher plasma Nt-pro-BNP levels than those who did not (680 [range, 164-1577] pmol/L compared with 137 [range, 53-336] pmol/L, respectively; < .001).

P

P

P

did not

P

P

P

P

When Nt-pro-BNP levels were not included in the model, multivariate analysis showed that cardiogenic shock correlated with diabetes (odds ratio [OR] = 1.45; 95% confidence interval [CI], 1.22-1.72; = .04), systolic blood pressure (OR = 0.96; 95% CI, 0.95-0.97; < .001), and creatinine clearance < 60 mL/min (OR = 1.54; 95% CI, 1.30-1.82; = .01). With Nt-pro-BNP included in the model, Nt-pro-BNP levels correlate with diabetes but the relationship with systolic blood pressure (OR = 0.96; 95% CI, 0.95-0.97; < .001) and cardiogenic shock (OR = 2.22; 95% CI, 1.92-2.58; < .001) remained. When Nt-pro-BNP was not included in the model, multivariable analysis showed that diabetes was an independent factor for mortality ( = .006), but it was less significantly so when Nt-pro-BNP was included in the model ( = .01).

Discussion

Little is known about the relationship between Nt-pro-BNP and patients with diabetes. Patients with diabetes with no obvious cardiovascular disease6,8 and acute coronary syndrome9 have been shown to have elevated levels of plasma Nt-pro-BNP in some studies. These studies did not take into account LVEF, however, which significantly affects plasma Nt-pro-BNP. After adjusting for anterior wall location, plasma troponin level, body mass index, hypertension, creatinine clearance, LVEF, sex, and age, the results of our study showed a marked elevation of plasma Nt-pro-BNP level in diabetic patients. This suggests that, following MI, diabetes is a significant independent factor for plasma Nt-pro-BNP.

It has been reported previously that diabetic patients with acute MI have greater in-hospital mortality and a greater occurrence of cardiogenic shock.5 A significant correlation between Nt-pro-BNP levels and both cardiogenic shock and mortality following MI was indicated by the results of our study. Diabetes is a known independent risk factor for cardiogenic shock, but is no more associated with cardiogenic shock when Nt-pro-BNP is taken into account. Because diabetes is a known independent risk factor for cardiogenic shock, this strongly suggests that increased plasma Nt-pro-BNP may be one of the connections between diabetes and the increased risk of cardiogenic shock after MI.

Our data showed that the higher risk of mortality and cardiogenic shock among hospitalized diabetic patients following MI is tied to increased Nt-pro-BNP levels. The degree of ischemic territory as well as the size of myocardial necrosis are indicated by the level of plasma Nt-pro-BNP.4,10 Systolic impairment, hypertension, and renal dysfunction, which are associated with unfavorable results after MI, are also correlated with increased plasma Nt-pro-BNP levels.2,4,8,9 This suggests that plasma Nt-BNP level is a useful indication of the total effect of risk markers for adverse outcomes after MI.

A number of factors may be involved in the rise in Nt-BNP levels among patients with diabetes following MI, including reduced diastolic function, reduced left ventricular compliance, and greater myocardial stiffness, which occur in diabetic patients with or without cardiovascular symptoms.11 Because of a deficiency of intracellular glucose, patients with diabetes have decreased adenosine 5'-triphosphate (ATP) production, resulting in Na+/K+-ATPase pump and Ca2+-ATPase pump dysfunction and, thus, decreased myocardial relaxation.12 This may be responsible for the greater release of Nt-pro-BNP. Moreover, the higher collagen content seen in the hearts of patients with diabetes is responsible for increased myocardial stiffness and may participate in the elevation of plasma Nt-pro-BNP levels in diabetes.13 In addition, the higher plasma Nt-pro-BNP levels found in patients with diabetes after MI may also be explained by ischemia that is more significant than that of patients without diabetes, even when the size of the infarct is the same.

Conclusion

Among in-hospital patients with MI, diabetes is a significant independent factor for elevated plasma Nt-pro-BNP levels. Higher Nt-pro-BNP levels are powerfully correlated with the higher occurrence of in-hospital cardiogenic shock and death seen in diabetic patients following MI. This shows that plasma Nt-pro-BNP level is a reliable predictor of outcome in hospitalized diabetic patients after MI.