I think the reason why I was asked to give a talk at this presentation is I’m pushing the envelope in how we can try and tailor both treatment and screening in an individualised way. What I’m trying to do is move the changes in precision medicine further up the food chain so meaning earlier in the time course of development of disease to the early breast cancer setting, the early treatment setting, not the metastatic setting, and, more importantly, even further upstream which is to prevention and screening.
What is the WISDOM study?
I’m going to talk a little bit about the WISDOM study which is a big personalised breast cancer screening study. As well there’s a companion study in Europe rolling out called the MyPEBS, the personalised breast cancer screening trial. All of us together will work on combining our data to really try and sort this out and do better about targeting our resources to the people who will benefit most from them and hopefully in the future even find better ways of identifying people who are at risk for the most aggressive cancers.
That study is open and enrolling, we’ve actually accrued 14,000 women already. We are now going to be rolling it throughout California and in the Midwest and Northern states and we’re about to roll it across the country. Very importantly we actually, in the United States, are collaborating with some of the insurers – Blue Cross Blue Shield in particular – and we’re rolling this out under what I call coverage with evidence progression or what Trent Haywood from Blue Cross Blue Shield calls it. This is really important. What we’re trying to do is bring all the stakeholders together up front to discuss all these issues and we are starting to meet and project what the different kinds of results might be so that we can spend our time while we’re waiting for results to argue about how we’re going to use those results. Why should we start that whole process in five years or four years when the results come out? When the results come out we should be ready to change that day. So, as I point out also in the talk I’m giving later today, data often doesn’t change people’s minds. People have strongly held beliefs, they are sometimes not going to be dissuaded in those beliefs by data. So it’s really important to address all those issues up front and make people confront their biases and work it out ahead of time so that we don’t have these ten and fifteen year knowledge changes or knowledge cycles. That doesn’t benefit patients at all.
It’s like we all believe that personalised medicine is going to work but you can’t just know that because we have to prove it. I’m not rolling out personalised screening because I know it’s going to work; I’m testing it because I think it deserves its day in court and that we should find out is the personalised model that we’ve built something that actually is as safe? Is it as effective, is it more effective? If it’s not then we’ve got to go back to the drawing board and do something else. Everyone has to be willing to change when the data comes out. But we’ve tried to build the trial in a very flexible way such that every six months when the data on risk models comes out and changes that we can change the risk model that we’re using. People say, ‘Oh, you can’t do that, you have to fix it in time,’ but why should we do that? That’s exactly what’s led to the idea of running trials that by the time the results come in are obsolete because the field has moved on. We’re testing not a specific model of risk space screening, we’re trying to test the idea of risk space screening and personalised screening.
For example, I think trials need to look more like care and care needs to look more like trials. If someone comes in to see me and in the last year their twin sister developed breast cancer I wouldn’t say to them, ‘Oh well, I’m sorry, I can’t incorporate that into my assessment of your risk because I assessed you last year and that wasn’t there.’ No, of course we wouldn’t do that so why should we do that in trials? That makes no sense either. So you created a flexible model so that you can continue to learn.
That’s actually something we’ve done in the iSPY trial. That’s the other part of the talk which is I fundamentally think that if you want to make a big difference in curing women from breast cancer, the place to focus all your energy is not on metastatic disease but in the early high risk setting. So the people most at risk to die of their disease, to try to figure out how to get information about their response to treatment right up front, right at the time of diagnosis. As a surgeon I can say one of the most important things for surgeons to do is to not operate on people when they come in the door. If you have a big cancer and you’re at high risk for developing metastatic disease surgery is not going to change your outcome unless you delay that surgery until you can get an assessment of the response to whatever systemic therapy you’re giving them. That’s our big chance to change the outcome and we have shown in the iSPY trial, and we’re now on our 15th and 16th drug, is that if you get a complete response to therapy, no matter how you get there, what drug you use, what subtype you have, you have a great outcome. And if you don’t your outcome is not so good. What that tells me about the future is that our pivot should be to get everybody to a complete response at the time they’re diagnosed, in that first six months. That’s the goal. So we’re changing, we’re modifying the iSPY trial to allow us to do just that – introducing sequential multiple assignment randomised trial technology into the trial so that we can continue to give everybody. So we’re in the process of transitioning the trial to one where we now have demonstrated that pCR, complete response, is the right endpoint. Now we want to figure out how to do our best to get everybody there and to do that with the least toxicity possible. An incredibly exciting opportunity, that’s the sea change.
What does this mean in light of the TAILORx trial?
TAILORx actually confirms what was published in 2004 in the New England Journal. This is confirmatory but it’s not new, so we’ve known this. This is for hormone positive, node negative women and it says that everyone doesn’t benefit from chemotherapy. This is the change from the stage was king to now the biology is king. So it helps us say who doesn’t need chemotherapy. Again, when we think about this another huge sea change was the MINDACT trial which again something similar but included not just hormone positive women and node negative women but node positive women, up to three positive nodes, up to 5cm tumours, all receptor subtypes, to say that if you were biologically low risk by the 70-gene study you didn’t benefit from chemotherapy either. So it’s not that you’re not at risk, it’s just that you’re not at risk for early recurrence.
We, in fact, made that gamble nine years ago when we started iSPY and basically excluded all 70-gene low risk patients because we felt that, a) they weren’t going to benefit from chemotherapy and, two, they weren’t at risk for early recurrence and so our endpoint of complete response wasn’t going to be a good predictor. So a lot about precision medicine is trying to figure out how to get the denominator right, what population of patients do you want to focus on? So iSPY focusses on those women who are at early risk for recurrence, people who have more proliferative tumours where chemotherapy or those kinds of agents would benefit. But now we’re starting to find out what other new drugs are out there in the pipeline. Thousands of drugs and combinations are ready. We have to be ready for them.
I also think the PERSEPHONE trial is an interesting example of where I think we need change. So this is a trial looking at 12 months versus 6 months of adjuvant Herceptin and this is a trial with AC, Taxol and Herceptin followed by more Herceptin. The 6 month versus 12 month absolutely indistinguishable and yet the comment from the platform was, ‘This is not practice changing.’ So you can ask yourself why is that – is it because people have beliefs and they’re not to be dissuaded by data? Or is it because the trial tested a concept that has now changed? So people with early stage disease aren’t getting Taxol, Herceptin and AC anymore, they’re getting Taxol and Herceptin alone. So now people say, ‘Well I don’t know if this data applies to them.’ So why bother to do a trial with 5,000 patients if by the time you get those results they’re obsolete. To me that says we have to rethink the way we do everything.
But I would say, clearly, someone who has had a great response to therapy probably doesn’t need much else and probably doesn’t need AC. In fact that’s what we’re doing in the iSPY trial – we’re dropping AC if someone has got a complete response. Why? Because additional adjuvant therapy once you’ve gotten to a pCR, once you’ve gotten to that great response doesn’t add much. There’s going to be new data coming out on that. That’s so powerful.
What this tells us is if people say ‘Oh, I can’t make these judgements because I don’t know if it applies to my patient,’ what that tells me is we’ve failed to get the right information. Because we operated on people, took out the biomarker and are basically treating everyone the same. That’s not precision medicine. Nobody should go to the OR without understanding what the response to therapy is. People say, ‘Well, I have to operate on them to know exactly what to give them.’ No, we have enough tools to figure out how to stage patients. Give them what you think is the best therapy and if they have a great response you’re done. Then you can do targeted de-escalation. But if they don’t have a great response then you’re going to give them something more and the time to do that is before surgery. So you have to figure it out first.
The reason why we’re stuck with a lot of rules in adjuvant therapy that we don’t know how to apply is because we spent decades just treating people without any evidence of what someone’s individual response is. That’s the sea change. Everything has to be conditional on someone’s individual response. What’s exciting about that is the main concept of precision medicine is if you have higher risk or a poor response then you need to escalate your therapy. But if you have lower risk or a great response then you can de-escalate therapy. That’s what so exciting and that’s where I believe the future lies.
