Diagnostic tests are the unsung heroes of cancer treatment. While there are more treatments for cancer coming out every year, prevention and early treatment remains the best option by far. The development of modern diagnostic tools usually centers around protein or genetic markers.
Epigenetics has become one of the leading paths for diagnostics. Epigenetics is essentially the study of how genes are regulated indirectly. DNA methylation is an example of this. DNA methylation suppresses the expression of the DNA that is being methylated. In cancer, there are several sites in the genome that are abnormally methylated. These could, in theory, be used as markers for cancer diagnostics. A team from Nagoya University in Japan hypothesized that they could use DNA methylation markers to create a method that could diagnose pancreatic cancer patients.
First, they needed to identify the DNA methylation markers in question. They began by identifying all highly methylated DNA sites in pancreatic cancer samples. Then they narrowed down the possibilities by logical reasoning, bringing the possible markers down to five from 5,575 initial hits.
Using The Cancer Genome Database (TCGA) as well as their own testing, the team set out to validate these markers for the diagnosis of pancreatic cancer. At least one of the five methylation markers were found in nearly all pancreatic cancers tested. In healthy tissues, none of the sites were methylated. The team also found that including the commonly mutated KRAS as a marker alongside the five DNA methylation markers, the accuracy of the test could reach 98% correct.
To implement these markers as a diagnostic tool, the team established an MBP-ddPCR method. The method needed something called cell-free DNA (cfDNA), which is found in blood or serum. Unfortunately, cfDNA is not very enriched in either. To get around this, they first collected methylated DNA using a methylation binding protein (MBP), then amplified it to see if the methylation marker sites were present. This method managed to produce reliable data on a tenth of the cfDNA.
Using this method, the team tested forty-seven pancreatic cancer patients alongside a control group of healthy volunteers. Alone, the five methylation markers could not reach a statistically significant accuracy. However, alongside KRAS mutation markers, it was nearly 70 percent accurate at diagnosing pancreatic cancer.
While not perfect, most diagnostic tools today are not the most accurate. The issue of accuracy is often overcome by conducting three or more tests and examining all three. This method also showed the ability to identify pancreatic cancer in earlier stages, before malignancy commonly sets in. Overall, this novel test shows promise for future studies beyond just pancreatic cancer. The group concludes, “Although further optimization may be required for practical clinical use, the combination of DNA methylation markers with highly sensitive MBD–ddPCR may provide a non-invasive effective diagnostic strategy for identifying devastating pancreatic cancers.”