Computer modeling helps docs determine need for genetic testing

A customized computer model is part of a lower-cost, prescreening method designed to help determine whether individuals have Lynch syndrome, a genetic defect that increases the likelihood an individual will develop colon, uterine, pancreatic and urological cancers, according to researchers at the Intermountain Healthcare Clinical Genetics Institute at LDS Hospital, Salt Lake City.

Lynch syndrome occurs in people who have an inherited genetic mutation of mismatch repair genes, which normally assist in everyday repair of an individual's damaged DNA, according to a news release e-mailed from Intermountain.

Individuals who have Lynch syndrome have a "substantially increased risk" of developing colon and other cancers, so "being able to identify people who carry a gene change is profoundly important because earlier and more-frequent screening—not just for colon cancer but also for other cancers—could save their lives," institute director and research team member Dr. Marc Williams said in the release.

Performing a full genome sequencing is one way to identify the mutation, but it is cost-prohibitive at $4,000 to $6,000 a person. The research team sought to develop an accurate, lower-cost method for prescreening patients using data gathered from existing tests to identify candidates for the more costly genomic sequencing.

Team members relied on Intermountain patient data, published literature and outside groups to come up with a prescreening approach. They found that results from two relatively inexpensive tests can help doctors determine whether full genome sequencing is warranted.

Of 272 colon cancer patients screened using the researchers' method, 261 were ruled out as carriers of the mutated genes.

"That left only 11 patients who we would recommend going forward with the full genome sequencing test," Williams said. The prescreening approach allows for "the wisest use of the expensive resource of full sequencing," he added.

"We think application of this type of modeling can help healthcare systems make better decisions about how to best treat patients," Williams said.