Statistically significant improvement

Going back to the SIRFLOX results.
Sir Spheres + Folfox resulted in an increase in overall progression free survival by 0.5 of a month versus Folfox alone. But this was considered not be be statistically significant, hence the trial did not achieve the primary endpoint.

With the Sarah results coming up, has SRX stated how many months increase in OS would be needed to constitute a statistically significant result??