I really don't see reduced osteoclastic activity in late-stage patients without known bone mets as concerning at all. It's cerainly plausible these patients have occult bone mets which haven't yet shown up on a scint scan.
In non-PC patients (I.e where Bone mets much less common)? Of the almost exactly the same distributions and magnitude. I'd suggest that borders on the incredible. And if the same effect exists for anemia (which I missed) then that too begs questions IMO - although not as obviously incredible IMO.