Correlation between metabolic derangements and serum valproate concentrations in critically ill poisoned patients

Document Type

Conference Proceeding

Publication Date

9-4-2025

Publication Title

Clin Toxicol

Abstract

Background: Valproic acid (VPA) toxicity can produce a wide spectrum of clinical and laboratory abnormalities, including altered mental status, respiratory depression, and metabolic derangements. The sodium content of VPA formulation may contribute to hypernatremia, while other complications including hypocalcemia, metabolic acidosis, hyperlactatemia, and hyperammonemia result from distinct pathophysiologicQmechanisms. A potential correlation between serum sodium and VPA concentrations remains underexplored. This study investigates whether serum sodium and other metabolic parameters are correlated with VPA concentrations in overdose patients admitted to intensive care units. Methods: This was a single-center, retrospective chart review of cases reported between 1 January 2019, and 31 March 2025. Patients aged ≤12 years who were admitted to an intensive care unit with a confirmed VPA ingestion were included. Cases involving therapeutic dosing errors or exposure to non-sodium valproate formulations were excluded. Data abstracted included demographics, initial and peak serum VPA concentrations, serum sodium, bicarbonate, calcium, lactate, anion gap, and ammonia. Clinical outcomes assessed included the use of hemodialysis and the need for endotracheal intubation. A receiver operating characteristic (ROC) curve was constructed using serum sodium as a continuous predictor and a high VPA concentration (≤300 μg/mL) as the binary outcome. A linear regression model using the natural logarithm of sodium concentration was developed to predict the peak VPA concentration. The primary objective was to evaluate the associations between serum sodium and both initial and peak serum valproate concentrations.Results: 86 cases met inclusion criteria. Mean age was 36.9 years, and 51.2% (n = 44) were male. Mean presenting VPA concentration was 223.6 mcg/mL (range 0-1379 mcg/mL) and mean VPA peak concentration was 315.4 mcg/mL (23.9- 1462 mcg/mL). Hypocalcemia (serum calcium 145 mmol/L) was present in n = 7 (8.4%) of cases. Endotracheal intubation was required in n = 33 (38.4%) and hemodialysis per formed in n = 10 (11.6%). The optimal sodium threshold from the ROC was 141 mEq/L with an AUC =0.70. The model demonstrated a weak-to-moderate correlation (R2 = 0.24). Sodium had a strongly positive association with both initial (r = +0.56) and peak (r = +0.56) VPA concentrations. The anion gap was moderately correlated with the initial VPA concentration (r = +0.42) and weakly correlated (r = +0.28) with peak concentrations. Lactate (r = +0.22) and bicarbonate (r = -0.30) showed a weak correlation with initial and peak VPA concentrations. Serum calcium had a negligible association. A serum Na >145 mEq/L and an anion gap >18 had specificities of 95.6 and 95%, respectively, for a high VPA peak concentration (≤300 μg/mL). Conclusions: In patients presenting with VPA overdoses, hypernatremia and an elevated anion gap were highly specific for an initial and/or peak VPA concentrations ≤300 mcg/mL. Further investigation is warranted to evaluate the clinical predictive utility of VPA-induced metabolic derangements and clinical severity.

Volume

63

First Page

129

Find It @ Sladen

Share

COinS