Login to view PhD Thesis

Enter your username and password here in order to log in on the website:

Forgot your password?

Neurohormones and left ventricular dysfunction

Kirsten Vilian Mikkelsen


            Background: Early manifestations of mild heart failure (HF) due to left ventricular dysfunction (LVD) often present with an abnormal filling pattern which are often described as isolated diastolic dysfunction (IDD) or as HF with preserved function. As of today brain natriuretic peptide (BNP) and its precursor-derived peptide (NT-proBNP) seem to be the most sensitive plasma (p-) markers of mild HF. Focus has been related predominantly to p-BNP and p-NT-proBNP as markers of increased cardiac wall tension, but also cardiac hypoxia induces BNP expression in cardiac myocytes.       
            Objectives: The main objectives were by investigation of patients with suspected, incident, mild, untreated HF to assess the activation of the brain natriuretic peptide system (BNP and NT-proBNP) relative to the renin-angiotensin-aldosteron system (RAAS) in plasma and to estimate diagnostic accuracy of these hormones at baseline and at re-evaluations after six and 12 months. In patients with HF, we sought to assess the relation of plasma NT-proBNP with Tei index, assess serial changes in indices, and to assess the value of indices to predict functional status and quality of life (QOL) during one year of follow-up.
            Methods: Doppler echocardiography and plasma measurements of BNP, NT-proBNP, RAAS components and noradrenaline were performed in 150 consecutive patients. Overall LV function was assessed with Tei index. Functional status was assessed with NYHA functional class. QOL was registered from the “Minnesota-Living with Heart Failure questionnaire”. Hormonal analyses were blinded until termination of follow-up. A post hoc threshold for an abnormal test at a baseline sensitivity level of  >90%  was selected.
            Results: Systolic dysfunction (ejection fraction ≤0.45) was present in 22 patients (group1), and abnormalities in LV filling alone were present in 58 (group 2). Patients in group 1 were older and more symptomatic. Ischemic heart disease was very frequent in group 1 (64%). Hypertension was most frequent (69%) in group 2.
             Median p-BNP were 307 pg/ml (90; 1179, 10th and 90th percentiles) and 37 pg/ml (13; 74) while median p-NT-proBNPRoche were 2285 pg/ml (595; 6395) and 199 pg/ml (92; 500) in the two groups, respectively (p=0.0001). Patients with no HF (group 3, n=70) had significantly lower BNP and NT-proBNPRoche concentrations (8 pg/ml (1; 18) and 55 pg/ml (24; 122), respectively, p=0.0001). P-concentrations of renin, angiotensin II, and noradrenaline were not different in the three groups and demonstrated no diagnostic potential (areas under ROC curves (AUC) <0.60). Baseline AUC were 0.93 (95% CI 0.90-0.98) for BNP and 0.95 (0.91-0.99) for NT-proBNPRoche.  In 5-10% of patients with LVD, p-BNP and p-NT-proBNPRoche were under the selected thresholds. Using a one-site antibody NT-proBNP assay (Biomedica) AUC was significantly lower (0.77, 0.70-0.85, p=0.0001).
            Based on reassessment of LVD at six and 12 months AUC of BNP was 0.81 (0.74-0.99) and 0.83 (0.76-0.89) respectively, and AUC of NT-proBNPRoche was 0.84 (0.77-0.91) and 0.87 (0.81-0.93), respectively. Using baseline thresholds (p-BNP=10 pg/ml, p-NT-proBNPRoche=87 pg/ml, achieving 95% sensitivity) reduced sensitivity of BNP from 95% (88-99) to 90% (82-97) at both six and 12 months and specificity from 60% (48-72) to 45% (33-59) and 49% (36-62), respectively. Sensitivity of NT-proBNPRoche was reduced from 95% (88-99) to 85% (76-92) and specificity from 76% (64-85) to 61% (47-73) and 65% (52-77), respectively.
            Tei index correlated with p-NT-proBNP (r=0.75, p<0.0001) and changes in Tei index in patients with LVD (n=80) correlated with the changes in p-NT-proBNP (r= 0.36, p=0.001). No improvement in NYHA class (n=47) was related to a median increase in Tei index (-0.2, -0.16; 0.09); an improvement (n=31) to a median reduction (0.06; -0.19; 0.35), p=0.02. The fall in p-NT-proBNP was greater in relation to improvement compared with no functional improvement (median 384 pg/ml (-173; 5885) vs. 64 pg/ml (-107; 360), p=0.002). The fraction of patients with ≥30% reduction in P-NT-proBNP was also significantly higher in the group with functional improvement (p<0.01). P-NT-proBNP and Tei index provided incremental prognostic information to identify patients with no functional improvement during follow-up. An improvement in log-likelihood X2 from 9.32 to 20.18 (p=0.001) and 20.67 (p=0.001) was seen by addition of NT-proBNP or Tei index to a model of clinical variables.
            Conclusion: patients with incident, untreated HF demonstrated considerable activation of the BNP system. RAAS hyperactivity could not be demonstrated. P-BNP and p-NT-proBNPRoche demonstrated high diagnostic accuracy for the diagnosis of HF with preserved or reduced systolic function at baseline, but diagnostic accuracy was sensitive to introduction of therapeutic intervention and the duration of LVD. NT-proBNP correlated well with overall LV function. The fraction of patients with at least 30% reduction in p-NT-proBNP during follow up was significantly higher in patients with functional improvement, and they had a decrease in Tei index. A poor baseline LV-function and a high NT-proBNP level indicated a potential of improvement in QOL and functional class, but only NT-proBNP and Tei index were predictors of functional class at 12 months. Thus, the BNP system demonstrated a diagnostic and prognostic potential, but further investigation is necessary before definite conclusions can be made about the potential added value of BNP or NT-proBNP measurements compared to clinical information alone in patients with HF and predominantly preserved systolic function.