I included the link so that people can read the whole article and am glad that you did. In a few places the author noted where the results of both drugs v. letrozole and a placebo or letrozole alone "were not statistically significant." I assume that it is that, that points to a lack of clear benefit that led to the headline conclusion.

The authors cited the problem of missing data as you noted. But statistically-valid conclusions cannot be drawn by ignoring missing data in such a study. That would be cherry-picking data or biased outcome.

So they mentioned what they could deduce (which you excerpted) from the incomplete data but nonerheless concluded as stated in the headline. At least that's what I deduce from the seeming discrepancies.

It would be valuable to know the cause of the missing data. E.g., if people dropped out if the study and why. And why it was the missing data was "disproportionate"? One reason people drop our of studies is adverse effects. In such a case, had the people stayed in the study, several of the measure variables would have been negatively affected.

But here's one conclusion from the study and the link again for anyone interested.
******************
“Overall survival was numerically longer in the palbociclib-plus-letrozole arm, but the results were not statistically significant,” Dr. Finn announced. “The interpretation of overall survival is limited by the large and disproportionate percentage of patients with missing survival data [ie, lost to follow-up or censored] between the treatment arms.”