Tool matches cancer genetics to approved treatments
When treating cancer, it's difficult toknow whether a chosen treatment is proving effective. While new breakthroughs may give doctors faster indications of whether drugsare working, by the time a new treatment is attempted, it might betoo late to achieve a positive outcome. A new tool, developed byresearchers at the University of Colorado, could have a big impact onwhich treatment doctors decide to use, using data from whole exomesequencing to pick out drugs likely to prove effective at tacklingtumors on a case-by-case basis.
The tool, known as IntegratingMolecular Profiles with Actionable Therapeutics (IMPACT), is designedto take data from whole exome sequencing of tumors, then mapping it onto the humangenome to create segments that correspond to each of the body'sroughly 20,000 genes. It then looks at those results against normalgene patterns, identifying variations that might play a factor in cancer occurrence.
With that list of candidate genes,IMPACT then uses publicly-available data to determine which alreadyFDA-approved drugs target those specific genes, and might thereforebe effective therapies for tackling the specific cancer occurrence.
To test whether the tool works asintended, the researchers took whole exome sequencing data from TheCancer Genome Atlas for patients with EGFR-mutated non-small celllung cancer. The tool took the data, and was able to successfullyidentify the gene EGFR as a major factor in the cancer, recommendingthat approved EGFR inhibitors be used as a treatment.
The tool was also used to analyze exomesequences from melanoma patients, successfully identifying theresponsible gene, and once again recommending an approved drug thattargets it. Furthermore, when one patient's tumor relapsed two years later,resequencing it revealed a new mutation, leading to the administeringof a combination of drugs to tackle both genetic issues.
The treatment controlled the tumor fora further two years, at which point it relapsed once more. The thirdanalysis revealed a loss of the gene CDKN2a, which is a known tumorsuppressor. No drug is currently FDA-approved to tackle theproblem in melanoma patients, but a medication called palbocicilibwas recently approved for breast cancer treatment. The researchersbelieve that that drug, in combination with the existing treatments,could suppress the tumor once more.
Such personalized treatment planswouldn't be possible without tools such as IMPACT. Overall, theresearchers believe that it could be an important step towards a morepersonal approach to cancer treatment.
"We hope that IMPACT proves to be animportant tool in empowering the shift towards precision medicine,"said senior paper author Aik Choon Tan, PhD.
The findings of the work were publishedin the Journal of American Medical Informatics Association.
Source: University of Colorado