Correlation of serum brain natriuretic peptide with hyponatremia and delayed ischemic neurological deficits after subarachnoid hemorrhage

Abstract

Objective: Serum brain natriuretic peptide (BNP) is elevated after subarachnoid hemorrhage (SAH), causes diuresis and natriuresis (cerebral salt wasting), and may exacerbate delayed ischemic neurological deficits. We examined the temporal relationship between serum BNP elevation, hyponatremia, and the onset of delayed ischemic neurological deficits and determined whether serum BNP levels correlated with the 2-week outcome after SAH.

Methods: Serum BNP and sodium were measured prospectively every 12 hours for 14 days in 40 consecutive patients admitted with SAH. All patients remained euvolemic, underwent transcranial Doppler assessment every 48 hours, and underwent angiography at the onset of delayed neurological deficits. New-onset neurological deficits were attributed to vasospasm only in the absence of other causes and when supported by transcranial Doppler or cerebral angiography.

Results: Sixteen patients (40%) experienced symptomatic cerebral vasospasm after SAH. A more than threefold increase in admission serum BNP was associated with the onset of hyponatremia (P < 0.05). Mean BNP levels were similar between vasospasm and nonvasospasm patients fewer than 3 days after SAH (126 +/- 39 pg/ml versus 154 +/- 40 pg/ml; P = 0.61) but were elevated in the vasospasm cohort 4 to 6 days after SAH (285 +/- 67 pg/ml versus 116 +/- 30 pg/ml; P < 0.01), 7 to 9 days after SAH (278 +/- 72 pg/ml versus 166 +/- 45 pg/ml; P < 0.01), and 9 to 12 days after SAH (297 +/- 83 pg/ml versus 106 +/- 30 pg/ml; P < 0.01). BNP level remained independently associated with vasospasm adjusting for Fisher grade and Hunt and Hess grade (odds ratio, 1.28; 95% confidence interval, 1.1-1.6). In patients in whom vasospasm developed, mean serum BNP increased 5.4-fold within 24 hours after vasospasm onset and 11.2-fold the first 3 days after vasospasm onset. Patients with increasing BNP levels from admission demonstrated no change (0 +/- 3) in Glasgow Coma Scale score 2 weeks after SAH versus a 3.0 +/- 2 (P < 0.05) improvement in Glasgow Coma Scale score in patients without increasing serum BNP levels.

Conclusion: Increasing serum BNP levels independently were associated with hyponatremia, significantly increased the first 24 hours after onset of delayed ischemic neurological deficits, and predicted the 2-week Glasgow Coma Scale score.