Residual cancer burden after neoadjuvant therapy and long-term survival outcomes in breast cancer
Prof William Symmans - MD Anderson Cancer Center, Houston, Texas
Approximately thirteen years ago we developed a system simply to try to measure the amount of residual disease after neoadjuvant chemotherapy. So after six months of chemotherapy you then have surgery and we’re just using from the general pathology, from the slides and organising the way we look at the specimen, this quantitative system called residual cancer burden. This study is a meta-analysis of twelve sites from Europe and the US where everybody pooled all their experience with this, and it’s 5,000 patients, to look at the outcomes and to see if this is generalizable and how well it does work. Also, importantly because there are four main types of breast cancer, it gives us big numbers to be able to look at each of the types to see where it works and what it means in the different types.
What were your methods?
The interesting thing is that after six months of chemotherapy it’s actually quite tricky to know where to look and to know how to report out what you find if you find any cancer. So obviously everybody is aiming for what we call a pathologic complete response, which is when we can’t find any cancer, having looked in the right place, in the breast or in the lymph nodes. That’s associated with an extremely good outcome. But then because six months of chemotherapy can alter the tumour, if there’s any left, in very substantive ways so it’s as much a method of standardising how we section, how we approach that sort of specimen, in order to have systematic standardised approaches to reporting back to the surgeon, oncologist and the patient about what we found and putting some meaning around it.
What were your results?
We found that it works and we found that each of these sites, many of them were retrospectively looking through their case records from their own respective studies and some of them were randomised clinical trials. There was a big Cancer Research UK trial that participated to this, as did individual sites in Edinburgh and in Cambridge. So there was a lot of UK representation in this group. So what we found was that it works but the most exciting thing about what we found when we had these numbers is it’s not just the category of how much there is but the actual numerical estimate on the scale from 0 to 5, say 1.37 or something like that. There’s a linear relationship between those numbers and the risk of the disease returning. It means we’re getting to a point where you’d be able to say, ‘Ah, the value that we found has this expectation in terms of risk,’ for an individual.
Did you look at event free survival?
Yes, event free survival and distant relapse free survival. They were basically the same result so we then for the rest of our studies we just chose one, we chose event free survival because that’s what the FDA had used when they looked at complete response.
How is the residual cancer burden index assessed?
When a surgery happens for breast surgery to remove what remains of the tumour and any lymph nodes that need to be sampled or removed, we look macroscopically to try to find. This is standard practice but this system says identify what you think is the largest cross-sectional area and then map it to the slides that will come out so that you can reconstruct from the slides the pattern, the extent and how much cancer there is and recreate it back to the picture. So that’s the foundation of what we’re doing here. It is simply just a way of providing some structure and organisation and standardisation around the way we approach the reporting of residual disease from the very moment we receive the specimen from surgery.
Do all cancer centres collect this sort of information?
No, they don’t. When we first reported on this at the ASCO meeting in 2006, soon followed with a paper, we set up a free website so that people can go to the website and there are educational videos and there are illustrations and protocols and descriptions and you can walk through the whole process and understand it. Most importantly the first thing you get to is a calculator so you can enter the data and create it, or the computer will automatically provide the index score and which of the four categories that index score falls into. That website currently continues to grow in its usage; so that website page was visited 16,000 times last month which is quite a lot. I think that it could be visited many times, people might just go there just to look so we don’t really know how often it’s actually used, how many patients it directly affects, but it’s quite widely used from that perspective. But for pathology around the world and even particularly in the US we’re really at the cusp of needing to have a standardised protocol that we say, ‘This is the way we’re going to do this for everybody.’ We’re not there yet but this will be an important study because it’s a big study and it provides some good foundational evidence to support certain elements of what might be in that kind of template. Because we don’t want to just make everybody collect information that’s not useful. So it’s very important but the implementation and adoption will continue to be things that we’re focussing on.
Is there anything you’d like to add?
Yes, the thing that I’d like to add overall is that this came together really very efficiently and quickly with a tremendous collaborative spirit. So the pathologists and oncologists and surgeons who have worked together to put this data together really took the time and trouble to get as much potentially useful information provided as well so that it can be a whole portfolio of analyses and studies to ask clinically relevant questions, several questions in the future, from this 5,000 patient meta-analysis.