Changes in B-type natriuretic peptide levels in hemodialysis and the effect of depressed left ventricular function.

B-type natriuretic peptide (BNP) is a cardiac neurohormone specifically secreted by the cardiac ventricles in response to volume expansion, pressure overload, and resultant increased wall tension. Previous research has shown elevated BNP levels in patients with volume overload caused by end-stage renal disease (ESRD). This pilot study was designed to describe the relative reduction in BNP that occurs as a result of hemodialysis (HD) in relation to baseline left ventricular function. Hemodialysis patients (n = 34) with ESRD were enrolled in a prospective manner. All patients had blinded serum BNP levels measured at the initiation and termination of HD. Levels of BNP were also measured in the dialysate and any residual urine, if available. In addition, monthly urea reduction ratio (URR) and Kt/V were obtained. The most recent measurement of left ventricular ejection fraction (LVEF) by any method was obtained from chart review. The BNP reduction ratio (BNPRR) was calculated by the following expression: pre-BNP-post-BNP/pre-BNP. The mean age was 50.8 years, 50.0% were male, and 55.9% were African American. A mean of 3239.4 mL of fluid was removed during HD. The mean pre-BNP, post-BNP, and change in BNP were 556.3 +/- 451.5 pg/mL, 538.6 +/- 488.3 pg/mL, and -17.6 +/-147.0 pg/mL. Of the 27 patients who had both pre-BNP and post-BNP values in the measurable range (< 1,300 pg/mL), BNP rose in 9 (33.3%) and fell in 18 (66.7%). The BNPRR had the following correlations: volume removed, r = -0.33, P = .07; Kt/V, r = -0.51, P = .01; URR, r = -0.34, P = .09; and change in body weight, r = -0.33, P = .07. The BNPRR was not correlated with time on dialysis or change in blood pressure. A total of 20 patients had LVEF recorded and post-BNP levels in the measurable range. For this group, the BNPRR values stratified by lowest to highest LVEF group were 4.6%, 19.1%, and 21.8%; P = .95 for trend. The BNP values were elevated in ESRD patients and decreased slightly during HD. This change was more pronounced in patients with normal or mildly impaired LVEF. The BNPRR correlated with the volume removed, change in body weight, and Kt/V. Future research with the BNPRR as a potential marker of the adequacy of volume removal in HD is warranted.