[Debbie L.]

Argg, I just wrote a long reply and there was an error when I submitted it. It was probably too long, anyway (smile).

First of all, does it help to know that premenopausal women with ER+ HER2+ cancer routinely take Tamoxifen adjuvantly (unless they remove or suppress their ovaries)?

Second of all, I think your doc's conclusion that the issue was that the AI "didn't work" because it didn't prevent the second primary could be argued. Perhaps the addition of Herceptin TO the AI, rather than necessarily a switch to Tamoxifen, would make just as much sense, especially if you're feeling an AI is important.

If you're in the US, it might be a good time for a real (as opposed to online forum, smile) second (3rd, 4th) opinion, from an NCI-designated Comprehensive Cancer Center.

My take on this (just an opinion, and not an expert one, either) is that much of the discussion about HER2+'ivity and Tamoxifen comes from the early days when they were just beginning to tease out some of the peculiarities of HER2+ cancer behavior. They saw that ER+ HER2+ cancer tended to be less responsive to endocrine therapy -- and at that time, endocrine tx was, by default, Tamoxifen. By the time the AIs were coming to the forefront, so was Herceptin, and we now know that Herceptin (or some anti-HER2 agent) improves response to endocrine therapy. Probably any endocrine therapy. Most of the negative stuff about Tamoxifen and HER2 doesn't seem relevant today, as it's without Herceptin, and it's often Tamoxifen vs. nothing rather than Tamoxifen vs an AI (plus Herceptin) -- which is what we really want to know.

I see differing amounts of enthusiasm from the experts about the small improvements in efficacy that AIs seem to offer in most studies (generic use of AIs, not specific to HER2+). Some say it's not enough to make a decision on and one should go with what's best tolerated or most appropriate. Others say that for those at high risk of recurrence, even a small improvement in benefit is important and so AIs are a strong first choice, even going to the point of rendering someone artificially postmenopausal specifically so they can take an AI (an approach I don't think we have much evidence for).

Debbie Laxague