Your friend may be basing his negative opinion on historical failures of prior attempts to develop similar vaccines (using different strategies). That opinion may not be unjustified, but it is very unlikely he knows the specifics of the blended data publicly presented in peer review form. This is an under the radar intervention that most expect will fail, but the OS numbers observed to date surely indicate something is working. Nothing is ever guaranteed, but the likelhood that the trial results will represent a total failure is very low. There is a signal in the noise, and they are working hard to show it despite challenges with the "old fashioned" trial design they selected over 12 years ago.