Unregistered.

I am not sure it is known. A PET or PET/CT shows whether an area has increased metabolism of glucose--thus it is useful to show whether something is a tumor and using lots of glucose to actively divide or just inactive scar whch doesn't need glucose for activity left over from the killing of tumor cells. (other things can cause the PET to light up such as infection, fracture or tumors of the bone itself)
But noone knows how many tumor cells could still be left within the scar without lighting up on a PET or PET/CT. With lung scans I believe the nodule has to be more than 3-5mm before it will show up on PET/CT.
Another question is whether an old-fashioned bone scan would be more sensitive--a question I would have to bring up with a friend who is a nuclear medicine specialist.
But it is clear that even with that scan, noone knows what minimum population of cells is necessary to light it up--just that it is much less than those needed to see an abnormality on Xray.

If the bone is one that has marrow and is Axial vs peripheral (part of the central skeleton like the spine, pelvis and rib cage vs middle parts of the arm bones, leg bones, wrists, ankles hands and feet ) there is always the question of whether there are dormant slowly dividing "stem cells" left. That used to be determined by a bone marrow aspiration-- a test which seems to have fallen out of fashion.

As I understand it, what one sees as bone mets on an Xray has mostly to do with chemicals that the tumor secretes and its effect on the marrow cells or matrix, causing the cells which destroy bone normally to become more active than the cells that build bone normally thus "eating away at the bone" That is why drugs like Zometa work by making the osteoclasts(bone dissoving cells) less active. I think they are actively researching how many tumor cells it takes to secrete these chemicals to have this effect and they seem to think the Zometa has a second effect besides the effect on osteoclasts--it actually slows the cancer's growth. I am not sure they know the mechanism of that, but I know it is an area of active research

I am sorry if my answer seems very nit-picky and academic and not very practical.

But the good new is that her scans are not getting worse which would be worrisome that her present or previous treatment is/was not working and that certainly does not seem to be the case.

Not being able to say with CERTAINTY that there are no microscopic tumor cells left is nothing new for the participants on this forum. Having her2neu as a marker may be an adverse prognostic marker, but at least there is something to follow (vs those who are her2-ER-andPR-) -- the serum her2 test elevation and depression seem to occur with metastasis and response to herceptin, respectively.

I am obviously not an oncologist but hope my speculation answered some questions at least, or at least didn't confuse you more.

Now here is a topic for discussion:
Is it better to know what/that they don't know or to get a dogmatic answer which you later find is wrong?