Science News is reporting on a rare cardiac side-effect of the tyrosine kinase inhibitor Gleevac, which was identified for its ability to inhibit the bcr-abl cancer mutation in CML (and cure CML with high efficacy).

This is an interesting science story, because this inhibitor was lauded for its specificity for this fusion protein created in a specific type of cancer, and its stunning ability to cure CML without serious side effects. But now that the drug has been studied in so many patients they're identifying a harmful side effect - heart failure in a rare subset of patients. I'm not pointing this out in a kneejerk attack against a drug, because this drug, despite this side-effect, is still about a billion times better than any chemotherapy regimen alternative. Instead I think this represents an opportunity for pharmacogenomics to enter into the picture of treatment of CML. Here you have a drug that is incredibly beneficial to a large number of patients who would otherwise have to undergo chemotherapy and dangerous treatment regimens, except for a rare, but potentially lethal side-effect.

This is the future of medicine. There is some genetic predisposition to cardiac failure on Gleevac, let's identify it, prospectively identify at-risk patients, and make sure each patient gets the ideal treatment based on their genomic profile.

There is no question Gleevac is an excellent drug. It was difficult initially to get it studied because it serves the relatively small population of CML sufferers (about 5k people get diagnosed with it per year) so it isn't necessarily a highly profitable drug. But it is so effective against a cancer that was essentially untreatable before this drug was discovered. We could continue to prescribe the drug universally for CML, but it would be better to identify the protein this supposedly specific inhibitor is interfering with, and figure out how to identify which patients should either be subjected to a lower dose, shorter time course of treatment, or possibly, alternate therapy to reduce the risk from this side-effect.

Granted there are many drugs that would benefit from pharmacogenomics determining exactly which people benefit from treatment. In particular, drugs that a larger number of people take would ideally be tailored to specific genotypes, but that will take an incredible amount of time, money and work. Here, with Gleevec, I see a great opportunity to nail a specific negative side-effect, tailor treatment to individual patients, and prove pharmacogenomics is a great solution to potentially dangerous side-effects in a small, easily-screened population.