The Urine Anion Gap: The Critical Clue to Resolve a Diagnostic Dilemma in a Patient with Ketoacidosis

Usually, ketoacidosis presents few if any diagnostic or therapeutic problems; in this article, we report a case where ketoacidosis was clinically occult and biochemically obscure. The patient presented with acute pancreatitis associated with a modest antecedent alcohol intake. Metabolic acidosis with a normal anion gap (10 meq/L) was observed together with moderate hyperglycemia and a 2 + (but not 4 +) test for serum ketones. None of the usual causes of metabolic acidosis with a normal anion gap was identified nor was there an obvious explanation for a reduction in unmeasured anion gap (e.g., hypoalbuminemia, dysproteinemia, or the presence of abnormal halides). Despite the initial normal anion gap, ketoacidosis was suspected clinically and this was confirmed by the elevated serum B-hydroxybutyrate of 8 mmol/L. We deduced that the serum unmeasured anions, which should have been increased by at least 8 meq/L, were being underestimated because of the effect of hypertriglyceridemia on the serum chloride determination. When the serum chloride was reestimated by a method not influenced by hyperlipidemia, the value was 102 mmol/L not 112 mmol/L and, when reevaluated, the anion gap was indeed appropriately elevated. In addition, the urine anion gap (Na + K − Cl) was 103 meq/L in the absence of renal disease. This indicated that the expected large quantity of urinary ammonium must have been masked by an even greater quantity of unmeasured anion; in this case proven by direct measurement to be B-hydroxybutyrate. Finally, metabolism of the alcohol ingested, which yields hepatic NADH, could explain, in part, the modest hyperglycemia and the absence of a 4 + test for serum ketones. With this information, appropriate treatment for ketoacidosis was instituted and within 36− 48 h the bicarbonate (23 mmol/L) and the anion gap (11 meq/L) returned to normal. Thus, knowledge of how the determination of the serum chloride can be influenced by hypertriglyceridemia and, at the same time, our utilization of the urine anion gap permitted us to make the diagnosis and effect the treatment of an obscure case of metabolic acidosis due to ketoacidosis.