Jeff

Thanks John and Christine for your additional input.

I do see that this seems to be a bit of a journal to journal spat. What is weird about it to me is how dishonest the Lancet piece is. The data they claim is missing from the NEJM piece is--unless I'm missing something--widely available all over the net. And was even partially presented at ASCO.

There is so much going on here that I don't understand--about the British health care system, US marketing of new drugs, etc.

But I think Christine is right: Overall Survival is the magic term in these trials. It's funny though because nobody at the Lancet kicked (as far as I can tell) about the publication of the ATAC trial data heralding the advance of Arimidex over Tamoxifen, even though the overall survival data is not clear here either.
The herceptin data from two trials was combined to give the results more statistical power--that is clear. But not, I think, out of any devious drug company machinations. Instead, I think, the pressure came from clinicians/investigators who were predicting--some more than a year ago (see Debu Tripathy's comments in breastcancerupdate, for instance) that there was absolutely no way that the results would be anything but strongly positive.

And hence, back to the ethical dilemma of what to do about randomization when the treatment arm is clearly getting benefit...

All the best,
Jeff