It is entirely possible for there to be some numeric trend toward an increase in heart valve problems that the safety panel didn't think was worth stopping the trial.

FORBES: Magazine Article