The Diagnostic and Prognostic Value of Hyponatremia in Acute Pancreatitis

Abstract

Introduction: Hypertriglyceridemia-induced acute pancreatitis (HTG-AP) represents approximately 1-4% of all acute pancreatitis (AP) cases, and is known to have a more severe clinical presentation. A prompt diagnosis is vital while managing HTG-AP, as the timely initiation of triglyceride (TG)-lowering treatment could rapidly resolve the etiology and change the clinical course. This study aims at exploring the diagnostic and prognostic value of hyponatremia in in this patient population. Methods: We retrospectively compared 140 patients with HTG-AP and 266 patients of AP of other etiologies. We performed receiver-operating characteristics (ROC) analysis for initial serum sodium level on basic metabolic panel to predict the diagnosis of HTG-AP, and further evaluated the correlations between hyponatremia and various clinical end-points in both HTG-AP and AP cohorts. Results: Presenting hyponatremia had an area under curve (AUC) of 0.926 for predicting the diagnosis of HTG-AP. The optimal cut-off value was determined to be 130mEq/dl, correlating to a sensitivity of 82.9% and specificity of 90.6%. After dichotomization, patients of AP with hyponatremia had a higher prevalence of acute kidney injury (AKI) (61.9% vs 38.1%, p=0.036) Conclusion: HTG-AP tends to have a more severe clinical course as compared to AP of other etiologies, with higher prevalence of pancreatic necrosis, organ failure, and higher rates of in-hospital mortality and re-admission. Considering the universality of pseudohyponatremia in HTG-AP, we explored its diagnostic value and proved that it’s highly efficient in differentiating HTG from other etiologies. We postulate that when using the optimal cut-off value of 130mg/dl, a presenting hyponatremia could serve as a sensitive and specific sign to diagnose HTG-AP, and confirming the diagnosis with a complete lipid panel should be prioritized in patients with AP when hyponatremia is present on the initial BMP. The prognostic value of hyponatremia to predict worse clinical outcomes has been extensively explored in multiple clinical entities. We explored the prognostic value of presenting serum sodium and found that hyponatremia is significantly associated with higher prevalence of AKI, SIRS; and patients with hyponatremia had higher BISAP and Ranson Criteria score than those without. In comparison with NLR and LDH, hyponatremia proved superior in the prediction of AKI and equally effective as NLR in the prediction of SIRS.60_A Figure 1. Receiver Operating Characteristic Curve for Sodium to predict HTG-AP60_B Figure 2. Characteristics of patients with acute pancreatitis by whether hypertriglyceridemia was the etiology60_C Figure 3. Clinical course of acute pancreatitis by level of serum sodium on presentation