A majority of cases of pancreatic cancer are diagnosed in advanced stage. The main reason is that early stage disease tends to have no symptoms, or vague symptoms. And there is no good screening test for early pancreatic cancer, or a simple diagnostic indicator. The two biomarkers most often used with pancreatic cancer are CA19-9 and carcinoembryonic antigen (CEA). However these assays are not very specific or sensitive in the early stages of the disease progression.
There have been a number of recent forays in terms of diagnostic biomarkers for adenocarcinoma of the pancreas based on specific gene mutations that tend to be seen in pancreatic cancer. And this has extended to the realm of MicroRNAs (or MiRNAs), small non-coding RNA strands that regulate the expression of specific referent genes. There have been a number of recent published studies demonstrating that certain members of the family of MiRNAs are affected by or affect many aspects of the natural history pancreatic cancer.
Now comes a study led by Danish researcher Johansen of Herlev Hospital near to Copenhagen (and her colleagues) which evaluates MiRNA panels as potential diagnostic markers for pancreatic cancer. The study, published in the Journal of the American Medical Association on January 22, 2014, is a case control study including 409 patients diagnosed with pancreatic cancer from six Danish hospitals from 2008 until 2012. The researchers looked at MiRNA in the whole blood of the patients with pancreatic cancer, in 25 with pancreatitis, and in 312 healthy controls. They evaluated 754 MiRNas, discovering 38 that appeared to identify pancreatic cancer. 19 of these MiRNAs were validated by a different method. Then two diagnostic panels (indices) consisting of these miRNAs were developed.
The area under the curve (AUC) for both MiRNa indices were higher than the AUC for CA19-9 (except in the validation cohort). This is an indication that these panels held improved diagnostic ability over CA19-9. Also, and perhaps importantly, including the CA19-9 together with the MiRNA indices gave better results than using CA19-9 alone.
There are serious limitations of this study. First, the differences between the AUC of the MiRNA indices and that of CA19-9 were quite small – and may not be clinically significant. Also, the age of the healthy control subjects was younger than those with pancreatic cancer.
Nevertheless, this is an intriguing finding in a clever study that requires validation and additional scientific investigation.
Dale O’Brien, MD