In three studies, markedly elevated BNP levels, but very weak correlations (r = 0.4) or even no significant correlations, between PCWP and BNP have been found in ICU patients requiring invasive monitoring, of whom a considerable percentage had sepsis (Table 3). Of particular interest are the findings by Jefic and coworkers, who found that BNP level could not differentiate high vs low PCWP respiratory failure. Whereas BNP level was not related to PCWP, a weak inverse correlation between BNP and LVSWI (r = -0.48) has been found. However, there was no significant difference in BNP levels between patients with an LVSWI of > 35 g/m/m2 and those with an LVSWI of < 35 g/m/m2. It is important to recognize that the degree of BNP elevation in patients with sepsis can be considerably high, even though a cardiac disorder is not obvious. A small retrospective analysis revealed that BNP levels in patients with sepsis and preserved systolic left ventricular function can be as high as that in patients admitted to the hospital because of CHF due to severely impaired systolic left ventricular function (sepsis, six of eight patients with a BNP level of > 1,000 pg/mL; CHF, five of eight patients with BNP of > 1,000 pg/mL) [Fig 1]. read only
Mechanism of BNP Release: BNP is obviously not a reliable predictor of cardiac performance expressed as LVEF in patients with sepsis, which is not surprising, however. Data from the Breathing Not Properly Multinational study suggest that BNP has only a modest discriminatory value in differentiating patients with preserved LVEF from those with impaired LVEF in the emergency department setting, and thus LVEF cannot be estimated relying on a patient’s BNP level. In addition, as systemic vascular resistance can be markedly lowered in patients with sepsis leading to unloading of the left ventricle, intrinsic myocardial depression, as measured by LVSWI, can be present even if the echocardio-graphically assessed LVEF is preserved. In contrast to systolic left ventricular function, a clear correlation between the degree of left ventricular diastolic dysfunction (ie, impaired relaxation, pseudonormal pattern, and restrictive pattern) and BNP levels has been identified previously. As left ventricular diastolic dysfunction has been found in patients with sepsis, it might contribute to the raised BNP levels in patients with sepsis and preserved LVEF. However, the relationship between parameters of left ventricular diastolic function and BNP has not been evaluated systematically in previous studies in this setting.
Figure 1. Comparison of BNP levels in eight patients with CHF due to severely impaired LVEF (systolic CHF [S-CHF]) and eight patients with sepsis, severe sepsis, or septic shock. S-CHF group: mean LVEF, 23 ± 3.9%. Sepsis group: mean LVEF, 64.4 ± 4.9%; p < 0.0001. Note the very similar BNP levels in the two groups despite a highly significant difference in LVEF. The data are from Maeder et al.