These data have been discussed by Kate Murphy. However I wanted to follow up with the significance of the data. To increase the risk of recurrence from 12% to 22% is not in any way or form helpful in the clinic, particularly because this outcome is independent of treatment effect.

In other words the technology used by Genomic Health did not result in any clinically meaningful markers which are helpful to decide whether chemotherapy should be given or not.

I was surprised that the data was presented like these are positive data because they have no impact on daily practices. These data need to be improved to make a difference so that they can be used to identify the patients who are at significantly higher risk.

Patients with stage II colon cancer have on average a 15 percent chance of cancer recurring. This is a big challenge! Should we treat everyone and treat many patients with no benefit to make sure we treat everyone who really is at higher risk.?

If we had a genetic marker set which could isolate these patients who are at higher risk, we could spare the ones with very low risk  from chemotherapy for 6 months.

The Genomic Health approach was not successful partly because they did not take advantage of the whole genetic make up. The technology is able to measure 40-50 thousand genes in one test and to figure out what signature would predict recurrence would be the solution.

In breast cancer they have developed a signature to predict recurrence risk and whether chemotherapy should be given.