The practice of personalizing cancer treatment for each patient is increasingly popular … and seems quite promising.

Genetic tests reveal abnormalities that doctors believe are associated with cancer, and a treatment plan is designed to respond specifically to those mutations.

But a newly published study shows these kinds of cancer treatments are sometimes built around erroneous assumptions. If doctors analyze only the genetics of the tumor and not any of the patient’s healthy cells, it’s tough to tell which genetic abnormalities are dangerous and which are harmless quirks.

Innocent mutations that do not cause disease are very common in human D-N-A, so the key is to focus on mutations that do pose a risk. Zooming in on the wrong mutations could lead a doctor to prescribe a useless and potentially damaging treatment plan.

This study analyzed the genetic makeup both of tumor cells and of healthy tissue from each of 815 participating patients. About two-thirds of the genetic anomalies in the tumor cells were just a regular part of a patient’s D-N-A, and not specific to the tumor.

Further study revealed something else: Nearly half the mutations appearing in genes for which a cancer-treatment plan or drug already existed were false positives. That is, these mutations were not actually linked to the patient’s cancer, although they initially appeared to be.

As genetically based cancer care becomes more common, it is imperative that doctors and patients understand the tricky business of identifying dangerous genetic anomalies.

The study’s researcher urges physicians to test the genes of healthy cells as well as the D-N-A of the cancerous cells. It’s the only way to be sure cancer care based on a patient’s genes is on target.