At this year’s conference, I was invited to present the results of ABCSG-16, which is a trial in the field of extended adjuvant endocrine treatment in post-menopausal women. What we essentially did was comparing an additional two years of an aromatase inhibitor – anastrozole, in the case of this trial – to an additional five years of anastrozole. The background of this is that we know that after five years of tamoxifen, extension of treatment duration is beneficial. This has been demonstrated by a number of trials. It’s a little bit less clear if patients had already received aromatase inhibitors in the first five years, but what was unknown until this presentation is what is the optimal duration of that extended adjuvant therapy.
In ABCSG-16 we recruited 3,484 post-menopausal patients with hormone-receptor-positive breast cancer, and after their initial five years of treatment, they were randomised to receive either an additional two years of anastrozole, or an additional five years of anastrozle. In terms of patient characteristics, the two trial groups were completely comparable, as you would expect in such a large cohort. We have a median follow-up of almost nine years after randomisation, which means that we on average follow these patients for almost fourteen years after the initial diagnosis and surgery. Disease-free survival was the primary endpoint, and essentially we did not find a difference between the two trial arms, so additional five years was not better in terms of disease-free survival compared to additional two years. There was neither a difference in secondary trial endpoints such as overall survival, contralateral breast cancer or new primary cancers. There was, however, a difference in clinical fractures. Additional five years will in result in a more fractures than additional two years, which is also a plausible result.
In summary, while this is scientifically a negative trial I think it has actually some good news for clinical practice in there. When you use extended adjuvant aromatase inhibitors, additional two years is good enough. This can help in clinical practice to reduce side effects, to eliminate fractures and to improve quality of life of patients.
From the slides this morning and the press session you did note the treatment adherence between both groups.
In clinical practice, treatment adherence is a relevant issue, and we took the opportunity of this large trial to also look into this aspect. What we see is, and that’s very comparable to other reports, about 20% of patients drop off the treatment in the first two years, and about 40% within five years. Based on tolerability issues, side effects, but probably also based on the perception that “I’m cured”, and for a variety of reasons. Even when we looked at the comparison within the trial, this is an exploratory analysis, of those patients who were perfectly adherent, presumably on a very favourable tolerability profile, we didn’t find a difference, which I think generally supports the overall conclusion of this trial.
If two years is no worse than five are there any plans to find out whether one year is no worse than two, or if anyone is even best suited for this?
You’re right. I think two years should now be considered the standard of care. Whether we could further de-escalate to, let’s say, one year, or whether even some patients might not need extension of treatment beyond five years at all needs to be determined in future trials.