Tumors, No Longer a Guessing Game

What makes one tumor different from another? The word ‘cancer’ can be used to describe hundreds of maladies, from a misshapen mole, to sickle shaped blood cells, to massive tumors in the heart, lungs, or anywhere else. But what makes a heart tumor different from a lung tumor? And why are there four different types of lung cancer? The answer lies in the DNA of each cancer cell.

Currently, doctors diagnose different types of cancer based primarily on where they are, what they look like in scans, and what molecules they are made of when samples are taken in biopsies. A new study introduces a novel method, looking at the methylation of cancer DNA, a process that inhibits or expresses specific functions in a cell that can characterize different disorders. By identifying the specific DNA methylations in each tumor, they can be attributed to a specific diagnosis; this is done simply by removing a small piece of the tumor, and analyzing it in the lab.

This study looked specifically at tumors on the central nervous system (CNS) and performed the methylated cancer DNA diagnostic test for 1104 patients, each diagnosed with one of 64 distinct cancers resulting in CNS tumors. In 76% of cancers, the test matched the existing diagnosis. In 12% of cases, the original diagnosis was revised as a result of this test. In the remaining instances, the test could not match to a known methylation class.

These results are extremely encouraging, and point to the end of subjective cancer diagnoses, where doctors must act as detectives, gathering clues to point to the most likely diagnosis. But this new test is a fingerprint; every cancer has a different one, and once they can be told apart, they need only consult a database stocked with thousands of cancer prints.

And this database is already live and free the world over, the data for every cancer case tested in this manner will be available and shared. Doctors will be able to readily compare the DNA in the cancers they encounter with other cases, which will likely lead to the discovery of new, unique cancers, and dramatically improve the accuracy of diagnoses overall.



Capper D, Jones DTW, Sill M, Hovestadt V, Schrimpf D, et al., (2018) DNA methylation-based classification of central nervous system tumours. Nature. doi:10.1038/nature26000

2 Replies to “Tumors, No Longer a Guessing Game”

  1. Super interesting! I really like your fingerprint metaphor to explain the cancer cell-methylation database. It’s pretty incredible how technology like this is taking human analysis out of the equation and using computers and online databases to diagnose diseases, like cancers. In a way, it concerns me that medical treatment is becoming more and more computerized and reliant on technology, but it is also making diagnoses faster and simpler.

Leave a Reply

Your email address will not be published. Required fields are marked *