LAUSR

CA 19-9 antigen has been identified in patients with colorectal, pancreatic, bile duct, hepatocellular, stomach, and esophageal cancers. Noncancerous conditions that may elevate CA 19-9 levels include cirrhosis, cholangitis, hepatitis, pancreatitis, and nonmalignant gastrointestinal diseases. CA 19-9 antigen levels may be used as an aid in monitoring response to therapy or disease progression in cancer patients. The objective of this study was to assess the analytical performance of the Access GI Monitor assay on Beckman Coulter UniCel DXI 800. CA 19-9 was quantitatively determined on the Beckman Coulter UniCel DXI 800 following CLSI guidelines. The performance was evaluated for linearity, sensitivity, reference range, and precision. The within-run and between-run precisions were assessed by analyzing QC material at low and high levels of concentrations. Accuracy was assessed by comparison with the previously established Siemens ADVIA Centaur XP. Antigen level values were classified as positive or negative based on the upper limit of the reference range to assess concordance between the two analyzers. The accuracy of each analyzer was also assessed by correlating the discordant antigen values with the patients’ clinical history. For the Beckman Coulter DXI assay, the analytic measurement range was determined to be linear between 0.8 and 2,000 U/mL with a slope of 1.018 and intercept of 4.13. The limit of blank was determined as 0.09 U/mL. The reference range was verified as 0 to 35 U/mL. The within-run CVs for CA 19-9 were 3.7% at both the low level of 23.54 U/mL and high level of 262.56 U/mL. The between-run CVs at low and high levels were 4.34% and 6.36%, respectively. A total of 327 patient samples were analyzed in the comparison of CA 19-9 levels on Beckman Coulter DXI and Siemens ADVIA Centaur XP. On Deming regression, the slope was 0.665 with an intercept of 136.4 and correlation coefficient of 0.8964; there was wide scatter between the two methods. The mean bias between the two analyzers was –291.1 (–22.8%). The correlation coefficient was 0.8931, and the bias was –770.7 (–26.1%) when both were positive. The correlation coefficient was 0.7949, and the bias was –0.8 (–6.1%) when both were negative. The agreement of the two methods was 91% (n = 298) and disagreement was 9% (n = 29). After clinical analysis of 18 discordant values, six data points correlated better with the Siemens Centaur XP and the remaining 12 correlated better with the Beckman Coulter DXI. Our data demonstrate that CA 19-9 on Beckman Coulter DXI has good linearity and precision. There is poor to fair correlation between the two methods. Agreement based on clinical classification of positive and negative results is good. The Beckman Coulter DXI correlates with the patient clinical history better than the Siemens Centaur XP when comparing discordant values between the analyzers.