At one point several years back, we at Pancreatica looked at the number of studies in pancreatic cancer (ductal adenocarcinoma of the pancreas) that dealt with early detection versus those aimed at treatment. We found that less than 3% were aimed at the detection of pancreatic cancer. This rate has since increased, but when interesting studies related to the early detection of pancreatic adenocarcinoma are published, we try to highlight them.
Thus, we note with interest the work of researchers from the United kingdom and Spain which include a number of those well known to the pancreatic cancer literature including John Neoptolemos and Nick Lemoine who published a paper in the August 2015 edition of Clinical Cancer Research (the official journal of the American Association for Cancer Research) that presented an intriguing method aimed at the early detection of pancreatic cancer. The authors developed an inexpensive three-protein panel array of biomarkers which appears to identify early stage pancreatic adenocarcinoma via a non-invasive urine test.
After preliminary work, in-depth proteomic urine analysis was performed on 18 individuals (nine women and nine men), six each with adenocarcinoma of the pancreas, chronic pancreatitis, and healthy volunteers. Approximately 1,500 proteins were identified. The protein array in men and women was not identical, and only 481 were found to be common to both. From this group, three proteins were selected on careful examination as expressing discrimination for pancreatic cancer. Two of these, REG1A and TFF1, had already been shown to be associated with adenocarcinoma of the pancreas; LYVE1 had not.
On subsequent testing, this three-panel urine protein array demonstrated a sensitivity for the diagnosis of adenocarcinoma of the pancreas of greater than 75%, and a specificity of greater than 85%. The panel showed increased concentration of the biomarker in early stage pancreatic cancer, as well as late stage. It discriminated between pancreatic cancer and chronic pancreatitis. In stage I-II pancreatic adenocarcinoma, the 3-panel biomarker showed a higher AUC of 0.97 (95% CI, 0.94-0.99) than that of serum CA19.9. And though adding the serum 19.9 results to the urine panel increased the AUC to 99%, it did not improve upon the diagnostic identification by the 3-protein urine panel of early stage pancreatic cancer patients as compared with that of the urine in healthy controls.
The authors conclude that this three-protein urine panel is a tool that can detect stage I and II pancreatic cancer with over 90% accuracy.
This research is an impressive study which deserves further confirmation, explication and possibly practical application. We applaud the ingenuity and care of the authors of this impressive and grounded systematic investigation.
Dale O’Brien, MD