Objective: This prospective study examines the diagnostic and prognostic use of tumor-M2-pyruvate kinase (Tu-M2-PK) used in conjunction with carbohydrate antigen (CA) 19-9 in patients with subsequently histologically confirmed periampullary malignancy.
Methods: Plasma Tu-M2-PK and serum CA 19-9 levels were measured at admission in a cohort of patients with suspected pancreatic cancer. Values for Tu-M2-PK and serum CA 19-9 were compared with a control group comprising jaundiced patients in whom malignancy was excluded by endoscopic retrograde cholangiopancreatography and nonjaundiced individuals undergoing laparoscopic cholecystectomy.
Results: The mean (SD) plasma Tu-M2-PK level for patients with histologically proven malignancy was 40.5 (26.4) U/mL and for noncancer patients, 29.9 (20.9) U/mL (Mann-Whitney U = 1163, P = 0.006). Tumor-M2-pyruvate kinase had an area under the curve of 0.623 on receiver operating characteristic curve analysis, and at optimal cutoff of 27 U/mL, sensitivity is 66%, and specificity is 58%.However, on multivariate Cox regression modeling, elevated Tu-M2-PK (>27 U/mL) was strongly correlated with the subsequent finding of poorly differentiated cancer and/or metastatic disease and strongly predicted survival on Kaplan-Meier analysis.
Conclusion: An elevated Tu-M2-PK more than 27 U/mL measured on admission in suspected periampullary cancer is a predictor of adverse prognosis in periampullary cancer.