How does renal insufficiency affect the performance characteristics of brain natriuretic peptide (BNP) and N-terminal prohormone of brain natriuretic peptide (NT-proBNP) assays?

Updated: Jan 08, 2018
  • Author: Donald Schreiber, MD, CM; Chief Editor: Erik D Schraga, MD  more...
  • Print


Given the renal clearance of NT-proBNP and, to a lesser degree, BNP, several studies have been completed to examine the effects of renal insufficiency on the performance characteristics of natriuretic peptide testing.

The authors of the Breathing Not Properly study suggested an increased rule-out CHF cutoff of less than 200pg/mL for patients with an estimated glomerular filtration rate (GFR) of less than 60mL/min. The investigators examined a subset of their 1452 patients for whom baseline BNP and estimated GFR were available. (Patients receiving dialysis were excluded from the original study.) BNP levels were inversely correlated with estimated GFRs. [22]

In a similar study, the effect of renal function on NT-proBNP and BNP levels was evaluated in 381 ED patients presenting with dyspnea. [23] Acute decompensated CHF was diagnosed in only 30%, but a similar inverse relationship was found between estimated GFR and NT-proBNP and BNP levels. The authors suggested rule-out CHF cutoffs of 290 and 515pg/mL for patients with estimated GFR of 60-89 and 15-29mL/min, respectively. For NT-proBNP assays, optimal cutoffs were 1360 and 6550pg/mL, respectively.

However, large studies are needed to validate these estimated GFR-dependent cutoff values. Furthermore, it is unknown if adjustment of NT-proBNP cutoffs beyond those recommended for age are needed to correct for impaired estimated GFR or if an estimated GFR–based stratification of NT-proBNP cutoffs is superior to the currently recommended age-dependent thresholds.

In contrast to renal impairment, the inverse relationship between obesity and BNP levels may adversely affect the utility of BNP levels to exclude CHF in obese patients. [24, 25] In the Breathing Not Properly study, body mass indexes (BMIs) were calculated from self-reported heights and weights in 86% of participants. Mean BNP values were 3 times higher for lean patients with CHF than for obese patients with CHF (517 vs 176mg/mL), and BNP was significantly and inversely correlated with BMI. [24]

In addition, in a study of 316 patients with CHF, those with BMIs of less than 25, 25-29.9, or 30kg/m2 or greater had median BNP levels of 747, 380, and 332pg/mL, respectively. [26] Obese patients with BMIs greater than 25 kg/m2 had median BNP levels below the recommended rule-in threshold of 500pg/mL. Additional study is required to further evaluate the cut-off values for BNP and NT-proBNP in obese patients.

Atrial fibrillation (AF) confounds the utility of BNP assay for diagnosing acute CHF exacerbation. [27, 28] A study found that in acute CHF, AF had no detectible effect on BNP levels, while in patients without CHF, AF was correlated with increased BNP levels. In this investigation, 1431 patients presenting with acute dyspnea, BNP levels were drawn, and patients were prospectively classified as those without AF and those with a history of paroxysmal or permanent AF (and, therefore, 1139 without).

Areas under the curve (AUCs) for CHF diagnosis were 0.84 and 0.91 for patients with and those without AF, respectively. With the previously established rule-out CHF cutoff of 100pg/mL, specificity was only 40% in patients with AF, compared with 79% in the others. The authors concluded that AF was associated with increased BNP levels in the absence of acute CHF. A high cutoff value to exclude CHF is required, but further study is needed to establish accurate thresholds.

Overall, the confounding variables of age, obesity, and AF may affect the clinical utility of BNP assessment in certain subsets of patients. Additional studies are needed to further evaluate these variables and their effect on the clinical performance of BNP and NT-proBNP tests. After the effects of these variables are elucidated, the accuracy of the natriuretic peptide assays will improve, but the complexity of their performance characteristics may also increase.

Did this answer your question?
Additional feedback? (Optional)
Thank you for your feedback!