Association of diabetic ketoacidosis and acute pancreatitis: observations in 100 consecutive episodes of DKA

Abstract

OBJECTIVE: The aim of this study was to evaluate the incidence, pathogenesis, and prognosis of acute pancreatitis (AP) in diabetic ketoacidosis (DKA). DKA is associated with nonspecific increase in serum amylase levels. Autopsy studies, on the other hand, had previously raised the issue of pancreatic necrosis in patients with DKA. However, the incidence, pathogenesis and prognosis of AP in the setting of DKA has not been prospectively evaluated. METHODS: This is a prospective evaluation of 100 consecutive episodes of DKA during a period of 13 months starting in January 1998, in a university hospital in New York City. In addition to careful history, complete blood count, arterial blood gas estimation, and a comprehensive metabolic assay, serum amylase, lipase, and triglyceride levels were estimated on admission and 48 h later. All patients with abdominal pain or elevated serum levels of amylase or lipase (more than three times normal) or triglyceride levels >5.65 mmo/L (500 mg/dl) had a CT scan of the abdomen. The diagnosis of AP was confirmed when pancreatic enlargement or necrosis on contrast enhanced CT scan was seen. RESULTS: Eleven patients (11%) had AP. History of abdominal pain, not a feature on admission to include AP in the differential diagnosis, was elicited subsequently in eight patients. Abdominal pain was absent in two and one was comatose on admission. The etiology of AP was hypertriglyceridemia in four, alcohol in two, drug induced in one, and idiopathic in four patients. The hypertriglyceridemia was transient in four patients and resolved once the episode of DKA was corrected. Lipase elevation was noted in 29% and amylase elevation in 21% of all patients with DKA. Similar to increased amylase levels, serum lipase levels were also noted to be high in the absence of CT evidence of AP. CONCLUSIONS: DKA may mask coexisting AP, which occurs in at least 10-15% of cases. The pathogenesis of AP in DKA varies, but at least in some transient and profound hyperlipidemia is an identifiable factor. AP is more likely to be associated with a severe episode of DKA with marked acidosis and hyperglycemia. Ranson's prognostic criteria are not applicable to assess the severity of AP in DKA because they overestimate the severity. Severity index based on CT findings appears to better correlate with outcome. Elevation of serum lipase and amylase occur in DKA, and elevation of lipase levels appears to be less specific than amylase levels for the diagnosis of AP in the diagnosis of DKA. Although in this study AP in DKA appeared to be mild, a definite conclusion with regard to the severity should be based only on a much larger number of patients, as only 20% of patients with AP in general have serious disease.