53 Validating a Functional Test for Chronic Pancreatitis

• Published in: American journal of clinical pathology Vol. 149; no. suppl_1; pp. S191 - S192
• Main Authors: Lloyd, Kathryn, Misciasci, Anthony, Fox, Wendy, Ashwood, Edward
• Format: Journal Article
• Language: English
• Published: US Oxford University Press 11.01.2018
• Subjects: Bicarbonates; Diabetes mellitus; Endoscopy; Fluids; Laboratories; Morbidity; NADH; Oxidation; Pancreatic cancer; Pancreatitis; pH effects; Quality control; Secretin
• ISSN: 0002-9173; 1943-7722
• DOI: 10.1093/ajcp/aqx149.422

Abstract We conducted a validation for pancreatic fluid bicarbonate using an automated chemistry analyzer, the Beckman Coulter AU5800. This analyzer utilizes a multistep enzymatic reaction in which the conversion of phosphoenolpyruvate and HCO3– to malate is coupled to NADH oxidation. The reaction is monitored 380/410 nm. This assay is useful as an aid in the detection of chronic pancreatitis following endoscopy and secretin injection. The gastroenterologist collects stimulated pancreatic fluid (up to 300 mL). Pancreatic fluid bicarbonate concentrations are used as an indicator of pancreatic function, with expected concentrations up to 150 mmol/L. Despite increasing requests for this assay from clinicians, a test for pancreatic fluid bicarbonate is currently only offered on the menu of one US clinical laboratory. In our validation, the AU5800 bicarbonate of three patient samples was compared to results from the reference laboratory. The results demonstrated statistically different means that are clinically insignificant: bicarbonate results were (AU, reference): 96, 91; 117, 110; and 120, 110 mmol/L (mean difference 6.7%, P = .04, paired t-test). The pH of the fluids (using the Corning 150 Ion Analyzer) was 8.54, 8.50, and 8.27. Osmolality (using 2020 Advanced Instruments Osmometer) was 301, 296, and 293 mOsmol/kg. Three serial 1:2 dilutions demonstrated excellent linearity: Sample #1: 96, 46, 23, 12 mmol/L; Sample #2: 117, 56, 28, 14 mmol/L; Sample #3: 120, 57, 28, 14 mmol/L. The regression slopes were 1.004–1.013 with Sy/x of 1.16–1.40 mmol/L confirming linearity and the absence of clinically significant matrix bias. Spike and recovery studies produced slightly less robust results with a mean recovery of 88%. This poor recovery is likely due to significant rounding error and the relatively low concentration of our spiking material (high quality control, 30 mmol/L) compared to our patient sample range (96–120 mmol/L). The high pH values indicate that bicarbonate will be stable in pancreatic fluid. The osmolality results indicate that pancreatic fluid is isotonic. Given that routine proficiency testing for body fluid bicarbonate is not available, alternative proficiency testing is required by both the Centers for Medicare and Medicaid Services and the College of American Pathologists. We plan to perform biannual pathologist chart review at the time of testing. Chronic pancreatitis is a serious disease associated with significant morbidity, from diabetes to constitutional genetic conditions to pancreatic cancer. It is also a diagnostic challenge as indicators commonly seen in acute pancreatitis (eg, severe abdominal pain and elevated pancreatic serum enzymes) are often absent. In-house performance of this test will undoubtedly improve the quality of patient care through improved accuracy and turnaround time for this crucial diagnostic test. Furthermore, in-house testing will reduce testing costs by greater than $2,000 per procedure, an advantage that may become increasingly important under the currently proposed healthcare reform.