Even though there are guidelines about giving surgery to older women who have got early breast cancer, if you look at the statistics you find that the rates of surgery decrease with age, particularly over the age of 80. A previous Cochrane meta-analysis didn’t find a significant difference in survival between giving immediate surgery or waiting until local progression occurs to give surgery. [??] three randomised trials of immediate versus deferred surgery using updated follow-up data from these three trials in just over 1,000 women.

What was the study design?

In order to answer the question we have acquired individual patient data, so data from every woman in three randomised trials that were run of immediate versus deferred surgery. These are trials of immediate surgery with tamoxifen or just tamoxifen and then surgery in the case of local recurrence or progression.

What were the results of this study?

As expected, there is very good improvement with surgery in local control so that the local recurrence or progression rate is reduced by about three-quarters. That benefit becomes apparent immediately. The big question is whether that benefit on local control extends to breast cancer mortality, distant recurrence and, indeed, overall survival. What we have found is that there is a smaller but still highly significant improvement in distant recurrence rates with immediate surgery and that this translates to a similar reduction of about a third in deaths from breast cancer and a significant benefit on overall survival as well. But that benefit only becomes apparent after a few years of follow-up. There seems to be little effect on distant recurrence on the first two years after surgery.

What do you think is the clinical significance of these results, and what is next for this study?

There are two important take aways from this study. The first one is that immediate surgery does not just give improved local control but that it reduces the rates of distant recurrence and it reduces deaths from breast cancer. The results of the updated analyses of the BNS NSABP B-06 study will be important as well to contextualise these results.

The second important finding is about how we run trials of differences in local control or screening. What it shows is that even when you’ve got a very big difference like surgery versus no surgery, the differences in breast cancer mortality don’t appear immediately. So if we are looking at lesser differences we again will need to have sufficient patients and sufficiently long follow-up to be able to give reliable results.