Giuseppe Curigliano:
So ladies and gentlemen, good morning, good afternoon and good evening. It’s really my pleasure to introduce this outstanding post-St. Gallen with Lisa Carey from North Carolina Medical School in the United States, Hope Rugo, University of California, San Francisco, and Fatima Cardoso from Lisbon, Europe. My name is Giuseppe Curigliano, and I work in Milan.

So I believe we can start soon in discussing the first topic that is the role of pCR in the post-neoadjuvant setting. So you know very well that FDA stated that you need a pCR in order to approve a drug in the neoadjuvant setting. So the question for the panelists was “is pCR alone a valid surrogate to confirm the role in the neoadjuvant setting or do we need the data on event-free survival and overall survival?”. And the panelists answered that in order for pCR to be validated as a good biomarker, we need the data on event-free survival and overall survival for 83% of the panelists. So we just don’t need to have higher pCR to approve an agent, maybe. We need more data on event-free survival and overall survival. So let’s start with Lisa. What is your opinion on the results of the panel?

Lisa Carey:
I agree with the results of the panel and the question was very appropriate which is, is that enough? Is pCR enough to get registration and approval for a specific drug, a new drug? And I do think that you have to get to clinically meaningful endpoints of event-free survival and overall survival for approval of a new drug. But there’s some nuance here, right? Because pCR itself has some… The local endpoints are irrelevant, right? pCR overall generally also means pCR in the axilla and there are surgical management strategies that pCR is a intermediate endpoint essentially or a proxy for things that are meaningful from the standpoint of the extent of breast surgery both in the breast and the axilla. Moreover, pCR at the moment tailors our adjuvant therapy for both triple-negative and HER2-positive breast cancer. So I think we should also acknowledge that there is some value to it. But I think for introducing a brand new drug, I think we have to have event-free and overall survival endpoints.

Giuseppe Curigliano:
Hope, what is your opinion?

Hope Rugo:
I think that I fall in the middle. I do think that for final approval of a novel agent that we need something about event-free survival, we need to understand this. But we have to keep in mind that the trials that had been done in the neoadjuvant setting until very recently have been very small. So they’re only powered to really look at pathologic complete response as an endpoint and they aren’t powered to look at event-free survival. So we’re taking these small nimble trials that give us a lot of information about tumor biology and tumor response to specific agents, and we’re asking the impossible.

So in some ways by making this, you have to see overall survival benefit. You’re going to take away the benefit that we’ve come to believe in in the neoadjuvant setting. So I think we have to really think about this and maybe take it back one step. In addition to that, we know that post-neoadjuvant therapy, which is impossible to control, significantly impacts outcome. And for example, the reason that this came up and will be addressed again for a number of drugs at the end of April by the US FDA is that if you have really big trials that improve pCR and you continue the novel therapy in the post-neoadjuvant setting, you may further impact event-free survival even for the patients who didn’t get a pCR.

So what should we do with this? And I think that if you have a drug that’s already approved in the metastatic setting and you know what the toxicity is and the potential benefits in the metastatic setting, that you could consider pCR an approval path for accelerated approval so that you enable patients to receive a drug, balancing risk and benefit while you wait for the event-free survival endpoint. But then you have to require that those trials are powered in a way that can give us an EFS endpoint and you have to balance risk versus toxicity, which was the issue in the most recent discussion of immunotherapy.

Giuseppe Curigliano:
Fatima, what do you believe?

Fatima Cardoso:
So I’ve been fighting for this for many years because I always said that I was not happy with pCR alone. I think that pCR is a very good initial endpoint and I think there is a smart way of developing trials which is more, I think, what Hope was trying to say and what we see for example with the immunotherapy trials in the early setting, why starting neoadjuvantly, having a first readout at the pCR level. I guess that if we don’t have a pCR benefit that is meaningful, for the majority of types of drugs, not for immunotherapy but for the majority of drugs, we could even stop the trial and not continue to go with long-term follow-up.

For those where you have a meaningful change, whatever we define as meaningful change in the pCR, then you could extend the trial and follow for event-free survival. I know that the worry is that trials become more expensive if you have to follow the patients for the event-free survival. So I think for example if you use pCR as also a triage method like the I-SPY program, I think it is very good. You can triage the drugs that you should not waste resources in large phase III trials, but you cannot just approve a drug based on pCR alone. That was, I think, the smart way of putting the question and it will push, I believe, it will push to a better drug development because you can still, I strongly believe, you can still start in neoadjuvant and have a first readout and then decide what to do in terms of spending more money following the patients until survival or not.

Lisa Carey:
And Giuseppe, let me make one additional comment because I think what you’re hearing from all three of us is pCR alone, like a cheap pCR, here’s your delta, and you’re done with drug development is not enough. I mean, you need to continue. But I think it’s also worth noting and I think Fatima was alluding to this as was Hope, it’s a great place to start, right? It is a very good place to begin with because it gives you a readout of activity. It also gives you the opportunity to look for biomarkers that are related to prediction of response and outcome and it’s really hard to do that in the adjuvant setting. So the neoadjuvant setting actually gives us a lot of information about the disease, not just the drug, right? Because in many cases, response to therapy has as much to do with the underlying heterogeneity within our tumors as it does with the drugs that we’re using and you really need the neoadjuvant setting to address that.

Hope Rugo:
It also brings up one extra question which is interesting which is what is the delta in pCR that means that you want to move forward? And I think that when you look for example at chemo, it was really interesting to come up with immunotherapy that if you plan an endpoint and then you continue to follow patients along and your pCR and you end up with more patients than you intended for your endpoint like the patients you needed for EFS are now being looked at for pCR, which wasn’t your plan. And so then you have a narrower pCR but that pCR difference is still very big for say node-positive disease. How do you interpret that is a whole another question. Does it have to be the whole group? Or is there something special about patients whose tumors fit into specific groups they metastasize to nodes? This may be particularly important for certain drugs like immunotherapy and not as important for others, but maybe the delta is greater where we have higher risk disease.

Giuseppe Curigliano:
But since you mentioned immunotherapy, I’d like to remember to you that we asked the specific question. So if you remembered the question was, should we integrate immunotherapy in the neoadjuvant setting for triple-negative breast cancer since there is a large delta of increase of pCR and the answer from 90% of the panelists was no. So let’s comment. Lisa, you start with the comment. Why do you believe that the panelists say no?

Lisa Carey:
I think the panelists said no for several reasons. The first is there is a pCR delta, but even in the KEYNOTE trial that is the largest and is in fact powered for event-free survival, that delta actually has, as I clean the data, has come down a bit. Number two, we have immature data on event-free survival in a trial that’s powered to give us mature data on event-free survival, so we will have information on this. And there’s a hint of moving in the direction of favorability for it, but we don’t really know whether it’s statistically significant yet. And there is clearly toxicity of this kind of drug that’s unique and can be long lasting. And so I think it’s incumbent on us as providers and investigators to be particularly careful with these drugs in the early breast cancer setting.

Giuseppe Curigliano:
What do you think, Hope?

Hope Rugo:
I think that it’s really interesting what we saw with the IMpassion trial and the KEYNOTE trial in the neoadjuvant setting where pCR was increased particularly in node-positive disease which we talked about earlier. I think that that particular point was hard because that wasn’t an endpoint of the trial. Maybe we didn’t know. But we also saw, for example, a number of other important endpoints that we may be able to incorporate into treatment. Not exactly the question you asked but still really important which is that PD-L1 positivity seemed to correlate with a better response regardless of what treatment you got.

So it may actually be that what we’re learning from these neoadjuvant trials is who needs more therapy versus who needs less and what specific therapy they need. For example, maybe node-positive disease and it’s possible that PD-L1 positive disease with poor responses benefits more from immunotherapy. It’s hard to know. But I guess my thought on this is that as an overall approach, which is what the trial was designed for, node-negative, node-positive disease, immunotherapy in the neoadjuvant setting and continuing for a year after treatment.

The data wasn’t strong enough to allow accelerated approval of that approach. But I think for example if you gave immunotherapy in the neoadjuvant setting for patients who have a low chance of pCR and these are node-positive patients, the difference I think is clinically very relevant. So it brings up the question. I mean, I agree that it’s early to make a decision about the use of immunotherapy in all patients and that when you balance toxicity versus risk, this is a question we still need more results on. So I agree with Lisa completely.

But I also think that there’s a group of patients that we are not serving well now and that immunotherapy may serve those patients better. How we select at those patients right now while we’re waiting for event-free survival, that is difficult. Because on the individual patient level, we hate to fail our patients and when the tumor biology is quite aggressive. So you could take the approach that there may be select patients where immunotherapy now is an appropriate treatment in the neoadjuvant setting, but it’s better if we can do that still in a clinical trial setting and wait for the event-free survival data.

Giuseppe Curigliano:
And how do you comment the ODAC analysis of FDA?

Hope Rugo:
I think that’s what ODAC was. I was at ODAC, and I think what the panel was really saying, is that unless you had organized the data to say, and this is all about trial design in the end. You do a trial design, you’re looking at a certain group of patients, and you’re looking at pCR in half of the patients and event-free survival and overall survival in the entire group. So then you say, “Okay, we improved the pCR rate in half of the patients, so should we approve the drug?” But then as you went on their secondary analysis wasn’t a little less than the full patient population.

There was still a big difference in pCR, but when they went to the full, almost 1200 patients, the pCR difference narrowed to just 7% but with still substantially over 10% in node-positive disease. But that wasn’t the question that was being asked. The question was, should we approve immunotherapy for everybody who has a greater than two-centimeter tumor in the neoadjuvant setting continuing for one year of therapy? And the answer to that had to be that we have to delay for EFS data because we’re giving people a lot of therapy, and some patients died from therapy-related complications thought to be due to immunotherapy and some of these are hard to pick up.

Patients who developed adrenal insufficiency that wasn’t picked up before patients went to surgery and people have other medical problems that could be exacerbated. So I think that the ODAC answered the question from the FDA in the only way that it could be answered. Does that mean that tomorrow, if you see a patient who’s not responding to AC for example or EC, that immunotherapy won’t provide any benefit? We don’t know the answer to that question and it’s really a challenge at the present time. But I don’t think everyone with triple-negative disease should receive neoadjuvant immunotherapy.

Giuseppe Curigliano:
And Fatima, what do you believe?

Fatima Cardoso:
I think Hope touched a very important point. It’s all about trial design and we should always remember that trials are experiments and that they have an objective and a method. And so the conclusion is determined by their design and their statistical plan. And sometimes, we try to extract conclusions from trials that they were not designed to answer to certain questions and we need to be careful with that because like we always sometimes joke that we torture the data until the data shows us what we want to see and that cannot be done.

So I think that there is a fundamental piece of information lacking for us to be able to use pCR for drug approval, and that is, what is the delta? So when you are comparing two treatments, what is the delta that tells you that in the long term that difference is substantially and meaningful? Because for patients, except for surgical approach like Lisa mentioned, except for that, for improving the surgical approach, pCR has no meaning if it does not translate into a long-term benefit, right? So we need to find out, maybe we will in the future, for specific mechanisms of action, specific subtypes, what is the delta and then we can use pCR.

I think until then we need to see if that short-term endpoint translates into a long-term endpoint. And that’s why we said, well, I said, and I voted that we still don’t have enough knowledge to prescribe immunotherapy to everybody. And I just wanted to make one small note. Another thing that worried me a lot about the pCR-based approvals is that we’re often said to be temporary. So let’s think about pertuzumab approval in the neoadjuvant setting. It’s says, “It’s pending the APHINITY trials and then it will be re-evaluated.” The fact is that it was not.

And so now you have, in some countries, the fact that you have access to neoadjuvant is hindering and not allowing us to fight with the regulators of our own countries to have access in adjuvant. Because they tell us, “Oh you can do it in just the neoadjuvant.” And that is, for me, is I’m overtreating some patients that do not need pertuzumab, and I do not have access to those who need one-year pertuzumab. So if we do it temporarily then it’s mandatory that the regulators go back and re-evaluate. They cannot just say it’s temporary and then never go back to that evaluation.

Giuseppe Curigliano:
So, thank you. I believe we have different point of views. Please, Lisa, go on.

Lisa Carey:
I have to say there’s something that’s come up that I think has bubbled up and has to be attended to which is the importance of the nature of the immune activation status of the underlying tumor. Hope mentioned this but I think it’s worth noting that in all four of the larger neoadjuvant trials testing immune checkpoint inhibitors, it is true that you see an augmented pathological complete response rate to either the immune checkpoint inhibitor based therapy or just to chemotherapy, right? So PD-L1 status isn’t a proxy for immune activation in a tumor. This has been seen in chemo trials and triple-negative breast cancer using immune activation signatures. And it is a possible way towards tailoring therapy because it’s also prognostic in the absence of therapy in the small studies we have. So I think we have to keep in mind that this evidence of immune activation within the underlying tumor is hugely important for all the therapies that we give and we ignore it at our peril.

Hope Rugo:
Yeah, because you showed really nice data, Lisa, in the HER2-positive group-

Lisa Carey:
Thank you.

Hope Rugo:
… that the immune signature played a very big role and there’s data looking at TILs, for example, as a marker of immune response that showed that even patients who didn’t get chemotherapy for example had a better outcome. But I always look at that data and I think, “Well, we could have made it even better because that’s a great group of patients. No one should recur.” But maybe that is all, we’ll come back to our understanding of HER2-positive disease in the end that immune activation is key and how to. But how to manipulate that is tough and we don’t know the right answer to that yet. But I predict, I mean, it’d be interesting to know what everybody predicts, but I do predict that there is going to be a role for immunotherapy in early stage triple-negative breast cancer and that I’m very hopeful that we will see an event-free survival difference than patients in KEYNOTE-522.

Giuseppe Curigliano:
So let’s move from immunotherapy to carboplatin, the neoadjuvant setting for triple-negative breast cancer. We had the data of the BrighTNess trial. It seems that additional carboplatin may increase pathological complete response rate. We had a beautiful editorial of Hope discussing with Rebecca Dent on the role of carboplatin. But I remember the editorial of Lisa together with Eric Winer on the same topic. So the question was, should we add carboplatin to neoadjuvant chemotherapy in triple-negative, and the answer was for 40% yes, for 60% no. So which is your comment, Lisa?

Lisa Carey:
You let me go first all the time, so I have to frame this.

Giuseppe Curigliano:
Okay, I will change.

Lisa Carey:
No, no, no, it’s all good. No, I have to say so I think there’s nuance here and I think so it does increase pathologic complete response. It is a drug that we’re very familiar with. We know the toxicity of it. In a node-positive patient, you theoretically can eliminate axillary dissection requirements in quite a few patients with this. So I think you can, in patients with larger tumors, node-positive tumors, the local regional impact can be quite real of incorporating carboplatin. What we don’t have is what we started this conversation about which is do we know that it affects event-free survival? Do we know that it affects the likelihood of relapse and the data there are very mixed.

In 40603, the CALGB trial using a pretty traditional chemo backbone, it does not. Underpowered, as Hope commented, but the hazard ratio was one, it’s not trending. In the GBG trial, the triple-negative subset, it looked like it was. So I think there are trials that are answering this but right now I consider it an open question which is why I don’t think it should be routine.

Giuseppe Curigliano:
And for you, Hope?

Hope Rugo:
Yeah. I mean, I think we’re excited to see in the, I hope, not too distant future the event-free survival data from the BrighTNess trial which is we have really lack to that. I mean, we keep talking about how should we use pCR for approval of novel drugs. But here, we’re using an established chemotherapy drug that doesn’t really need to be approved because we can give it freely relatively. But we’re saying that pCR should be the primary reason for using this across the board. There is the downside. Patients who get carboplatin have thrombocytopenia, more nausea, more count suppression, use of growth factors, and in fact in some cases might even have delay in surgery depending on how this is used. We know that when drugs go outside of the clinical trial setting that we see more toxicity overall.

So I think we have to really carefully evaluate, is pCR enough? Again, I think that there may be a difference in patients who have more locally advanced disease as Lisa was mentioning versus those that do not. And so our approach has been to start with paclitaxel and add carboplatin in the more slow-response setting. I don’t know if that’s the right way to do it but it certainly avoids carboplatin in those patients who have a dramatic and rapid response to chemotherapy.

I think it’s a really important part of our neoadjuvant regimen for triple-negative breast cancer. We just don’t know yet who absolutely needs to receive a platinum as Lisa mentioned, and we need to see the event-free survival data to incorporate this as a general standard for all patients with triple-negative disease. Does that mean that some patients could benefit from carboplatin? Absolutely. And so I think that we need to consider each patient individually.

Giuseppe Curigliano:
So, Fatima, which is your perspective on this one?

Fatima Cardoso:
So I’ve always used carboplatin in selected cases. So I usually like to start with anthracyclines. And if there is no excellent response to anthracyclines, then do the taxane with the platinum. Where there are cases where there’s an excellent response, then you can withhold the platinum because of what Hope was saying, increased toxicity sometimes making it difficult to even give the taxane. And for example, for young women, also affecting fertility. So carboplatin affects particularly fertility. So I think I agree and I voted against the use in everybody. I think there are selected cases where it is necessary and we should try to individualize those cases either through response or perhaps in the future with some other form of biomarker that we could find.

Giuseppe Curigliano:
So, Fatima, another question for you. We asked the panelists-

Fatima Cardoso:
Ok I’ll go first.

Giuseppe Curigliano:
Yes, yes, Fatima. We asked to the panelists based on the tolerability of olaparib and waiting for the data of the OlympiA trial which was finally the benefit in terms of invasive disease-free survival at three years to use the olaparib in the adjuvant setting. And so the absolute benefit that we asked was more than 10%, more than 5%, and more than 2%. The majority of the panelists answered that you need the more than 5% in order to use olaparib. So which is your impression in this vote?

Fatima Cardoso:
So I actually voted that way also. Well, I was, as usual, thinking if I should demand overall survival benefit. But then I was considering that in terms of balance, efficacy, and toxicity, and considering also the type of tumors that were included in the OlympiA trials. So high-risk situations where unfortunately the rates of relapse are substantial. I think that to decrease the invasive disease-free survival in a way that is meaningful because 2% wouldn’t be meaningful, so 5 to 10% is already important. Ideally, we should see the same as we saw with T-DM1, but I’m not sure if we will get there so high. But I think it is a good compromise considering also that it’s relatively well-tolerated.

Giuseppe Curigliano:
So I know there’s escalation trial, Hope, olaparib in the post-neoadjuvant. Do you agree with Fatima impressions?

Hope Rugo:
I think that OlympiA took forever to do. I mean, it really was. I think we’ve been waiting for these results forever. As we all know, most acutely, there was a big delay even in studying PARP inhibitors in breast cancer simply because it was difficult to show a benefit and originally the idea was that patients would benefit who had specific phenotypes or genomics that didn’t pan out. So far, at least we’ve only seen the benefit in patients who have true germline mutations and now some data on somatic mutations as well.

I think that doing another trial in the post-neoadjuvant setting isn’t going to be practical or feasible. I mean, why would anybody, if they saw a DFS benefit in the adjuvant setting and had a BRCA1 mutation, why would they not take a PARP inhibitor? So you’re not going to be able to randomize easily. The one question is going to be, and I don’t know that we’re ever going to be answer that either is whether or not you would have been done just as well with few cycles of platinum as a DNA damaging agent that you need a PARP inhibitor for a year.

In general, the drugs are very well tolerated. But in a patient I had with locally recurrent disease that did not melt away with neoadjuvant therapy and gave her a PARP inhibitor, her tolerance, I mean, some patients really have a lot of nausea, anemia, other symptoms that make their quality of life in that year after completed their main therapy quite difficult. So I think that one doing another trial isn’t going to work. I do hope, we all hope, that the difference that’s seen in OlympiA isn’t just statistically significant but it’s also clinically relevant for our patients.

But I do predict that every patient who has a BRCA mutation high-risk features will take a PARP inhibitor. Based on the press release alone, they’re going to want to be taking a PARP inhibitor. So there’s a lot of desperation in these patients to have a marker of risk than their young and have bad disease. So I don’t know how we’re going to get around that. It’s going to be critical for us to follow the risks of leukemia. So far we haven’t seen an explosion of this leukemia and myelodysplasia and ovarian cancer, but there is a defined tiny risk so we need to balance that as well.

Giuseppe Curigliano:
So, Lisa, I would like to complicate the question for you. BRCA mutated, triple-negative, those then say carboplatin in the neoadjuvant setting and despite this residual disease. So who do you proposed, olaparib or capecitabine in these patients?

Lisa Carey:
It is a very unfair question so you let me go last. But then you made it a very unfair question because we haven’t seen the OlympiA data yet, and as Hope says, there’s statistically significant and there’s clinically meaningful, and those are not always the same. If we assume that it is clinically meaningful and you see a 5 or 10% difference in relapse rate, I have to say I think you’re going to end up trying to thread the needle. Personally, if it’s going to be similar to the subset analysis of CREATE-X in a large properly powered trial, I’m probably going to reach for olaparib, but I can promise you that there will be efforts to try and figure out how to do both in young, high-risk patients and that’s probably something we should acknowledge and be proactive about.

I think the extent to which OlympiA informs what we actually do is going to be really dependent on what Fatima was alluding to which is in general, these sorts of studies in different risk groups have proportional benefits that are very similar across different strata of risk. But the absolute difference is, of course, entirely dependent on who the patient is and who the study population is. When we’ve asked our patient advocates typically they say the same thing that we do, that they start to feel that things are clinically meaningful when you get above about 3%. So I think in the voting, I think you’re seeing people reflecting what in fact our patients considered to be true also.

Giuseppe Curigliano:
So, Fatima, now we move to ER-positive/HER2-negative patients. Let’s start with the first statement following the results of RxPONDER, MINDACT and TAILORx. So are there postmenopausal patients with clinical presentations meeting the criteria of MINDACT, TAILORx and RxPONDER with low risk signature and low recurrence score less than 25 who should receive chemotherapy and the answer was no for 79% of the panelists. So do you agree in this statement and do you believe at classical clinical pathological features like Ki-67 may have seen a role determining the risk?

Fatima Cardoso:
So I agree with the statement. I think now we have three large trials that have consistent results for postmenopausal women with low or intermediate results in a genomic profile, the 70G profile MammaPrint or the Oncotype. So I think now we have distant results and I think we can be on the safe side by saying that we can withhold chemotherapy for postmenopausal women who meet these criteria. And just one short reminder that clinical pathological factors were included in the design of MINDACT, not exactly in the design of the TAILORx and RxPONDER but there were analysis based on the criteria used on MINDACT to also help particularly for TAILORx to help interpret a little bit better. So I think, yes, if I have someone postmenopausal with clinical pathological features of high risk but a low risk genomic profile or an intermediate risk genomic profile we can withhold chemotherapy.

Giuseppe Curigliano:
What do you think, Hope?

Hope Rugo:
I think that medicine is never concrete. It’s just difficult. I mean, we’re always going to be faced with situations where we’re not comfortable with a decision either way. For example, I’ll give you an alternative example. You have a recurrence score of 28 and a patient who has locally advanced breast cancer who’s 76. Are you going to start with chemotherapy? This is ER/PR strongly positive. Or would you potentially give endocrine therapy first to see if you get a response?

And in some of those patients and we all have our own clinical examples, we’ve seen great responses to endocrine therapy and we know pCR rates are low in that group of patients. There’s a lot of consideration. So what do you do with the opposite situation which is really what we’re addressing here. So you have a patient who’s postmenopausal but has a high-grade tumor with lymphovascular invasion and three positive nodes with extranodal extension. I think that we do in that situation. I agree with Fatima that we don’t want to take the clinical pathologic characteristics as our primary driver but we can’t ignore them entirely. No trial is going to be able to tell us in that group of patients that there’s no benefit from chemo. So, of course, I voted them and this is what our information has voted the same way. But I think we do still need to take the individual patient into account much as we’ve been discussing with the other issues here. So there may be a patient who falls into an eligibility criteria for these trials who might still benefit from chemotherapy.

Giuseppe Curigliano:
So, Lisa, please, yes.

Fatima Cardoso:
Giuseppe, just let me say that what I said is meeting the criteria of high risk within MINDACT, so locally advanced were not included. So again, I think we need to be careful when we say so it’s not for locally advanced not even stage three breast cancer.

Hope Rugo:
Of course.

Fatima Cardoso:
All right. So following those other characteristics included in the trial up to three positive nodes, no more than that. And of course there will be patients where you are doubtful. I think the biggest clinical pathological criteria that I still don’t know what to do with it in this circumstances is grade because in MINDACT grade was not independent of the signature. But for example in the ADAPT trial it was. And so I still don’t know if you have a grade three by itself alone if you should always go for chemotherapy and that’s perhaps a question for the next St. Gallen.

Giuseppe Curigliano:
So, Lisa, to you, the conclusions regarding postmenopausal, you have been also the discussant for the TAILORx panel, I remember this. So please, go ahead.

Lisa Carey:
Well, I think what you’re hearing from Fatima and from Hope is also I think the consensus opinion which is in postmenopausal women with up to three involved lymph nodes, and I have to admit, we should acknowledge there weren’t very many three involved lymph node patients in those trials. But just generally meeting the criteria, I think there’s general consensus that you can use the genomic assays to defend omitting chemotherapy in a rational way. I think the devil being in the details, we have to be very careful that this shouldn’t be extrapolated to patients who didn’t belong in the trials, who were ineligible for the trials and we shouldn’t assume that no benefit in a lower clinical risk situation means there’s no benefit in very high clinical risk situations.

Giuseppe Curigliano:
Very good. Now we move to premenopausal women, ER-positive/HER2-negative. I would like to start with Hope with the first question we asked to the panelists. So for premenopausal women with node-negative breast cancer and recurrence score between 16 and 25, the recommended treatment was tamoxifen alone, ovarian function suppression with tamoxifen or AI, and chemotherapy plus endocrine therapy. And the surprise was that 53% of the panelists answered ovarian function suppression plus AI. So there was someone proposing not to give chemotherapy to those patients. Which is your opinion?

Hope Rugo:
It’s complicated and I think that because we didn’t answer that question. I mean, we didn’t look specifically is ovarian function suppression and an aromatase inhibitor or possibly tamoxifen adequate to treat patients who have a score between 16 and 25 who in TAILORx seemed to benefit a little bit from chemotherapy although I think the benefit was much greater in scores from 21 and above. And we know that there may be some impact of individual clinical pathologic characteristics here in young women. And we also know from TEXT and SOFT that the benefit of ovarian function suppression seems to be significant based on clinical pathologic characteristics where we didn’t get recurrence scores. So what do we do with that group of patients?

So I think for the majority of patients who have node-negative hormone receptor positive disease and scores in that range, that indeed ovarian function suppression and an aromatase inhibitor is a very reasonable option. And that chemotherapy as a treatment should be limited to patients who have really a very high, I think, risk sense themselves who want to do everything for a small benefit and for patients potentially who have other high risk features even though we’re trying not to look at clinical pathologic features too much. I mean, it may be that very high grade disease and other risk features play a role. So here again, I think we need to individualize treatment. But for the majority of young women, I think that ovarian function suppression and an AI is reasonable. That said, that’s tougher than doing a few cycles of chemo. There’s no question about that. Doing five years of ovarian function suppression and an AI for a young woman is tough in terms of toxicity. But indeed we think that’s really better in terms of therapy for this disease than chemotherapy.

Giuseppe Curigliano:
Lisa, do you agree?

Lisa Carey:
I agree with a slightly different spin. I’m pretty conservative in this arena, and I think to Fatima’s point earlier, trials only answer the questions that they ask. And so I think we wade into extrapolation a little bit cautiously here. I think at some point in the risk continuum in a premenopausal patient, in the node-negative setting, we did see, again, in a post hoc subset analysis that we have to be cautious about, that it looks like the recurrence score did start to behave as one might have a priori thought it should where at the higher risk scores you start to see a benefit of chemotherapy.

What I’m not clear about is whether ovarian suppression supplants that or just is numerically equivalent. Because I’m not sure that the ones benefiting from chemo are the same as the ones who benefit from ovarian suppression. So if you just treat it as a numbers game, I think the majority of patients with a low to intermediate recurrence score in node-negative for premenopausal likely can do well with just ovarian function suppression and endocrine therapy. As you start getting up into the mid-20s or higher, I actually use chemotherapy.

Hope Rugo:
And I agree with you. I was really impressed with the difference in data looking at when you’re looking at 16 to 20 and 21 to 25. And so I use the same approach and it’s really splitting hairs to some degree because I don’t know that was retrospective but I take the same approach.

Giuseppe Curigliano:
So, Fatima, you recently updated the data of MINDACT and published it already. Can you comment on the use of chemotherapy on premenopausal women?

Fatima Cardoso:
So what I take from these observation and remember like Lisa was saying, this was a post hoc unplanned subgroup analysis done in TAILORx that was then also done in MINDACT and then also in the RxPONDER. But despite that, the results are consistent so there’s something here telling us that premenopausal women chemotherapy does something. If what it does is just the induction of a menopause and that’s the effect of chemotherapy or if there is a direct cytotoxic benefit of chemotherapy, we do not know.

So how it has affected my practice? Perhaps the best way to say is that I used less genomic profile now in premenopausal women because I do not know how to interpret the results so well. And also a word of hope is that there is an ongoing trial and ADAPT trial that is comparing both approaches, the ovarian function suppression and chemotherapy, so we will know.

Giuseppe Curigliano:
So, Lisa, now we move to the last topic, that is the use of CDK4/6 inhibitor in the adjuvant setting. So the question for the panelists was should patients with ER-positive/HER2-negative high-risk breast cancer receive adjuvant CDK4/6 inhibitor therapy with abemaciclib and the answer was yes for 54% of the panelists and no for 46%. So can you comment on these results?

Fatima Cardoso:
So a clear consensus.

Lisa Carey:
And I think the fact that it was so split I think reflects the fact that we’re in a zone of emerging data, right? So monarchE in a high-risk population at an early look showed a benefit. PALLAS and PENELOPE-B did not using a different a drug. I think many people in a high-risk patient with multi-node positive breast cancer if you can get abemaciclib would add it at this point awaiting further data to come out with the maturity of monarchE. But I think in the conventional sense, I think many people for conventional adjuvant therapy probably would not at this point.

Giuseppe Curigliano:
Well, do you believe, Hope? Any comment on this?

Hope Rugo:
Yeah. I mean, I think that we struggle so much with the treatment of high-risk hormone receptor positive disease with the exception of patients who have extraordinarily high-risk disease, more basal like. Most patients don’t have a pathologic complete response. We see relapses, we see multiple positive nodes in these patients and it’s quite difficult to know how we can really change the risk of distant recurrence in the long term. So seeing this data was very exciting with abemaciclib in monarchE. How are we going to parse this out? Of course, we need longer follow-up. But at the moment, it’s very hard to not think about using these drugs in patients who have very high-risk disease. So four more positive nodes, we know those patients have such a high risk of recurrence and the risk of recurrence based on the HICAM data is actually earlier. I mean, we know the risk continues out particularly in indolent disease. But they do have a high risk in that first five years where we already know there’s a benefit from CDK4/6 inhibition based on the monarchE data.

So I think that in that group of patients, we could consider using abemaciclib now based on the data we have. We’re not creating cancers, people didn’t die of toxicity, and we understand the toxicity very well with lot of experience. What to do with one to three positive nodes and Ki-67? I don’t think we have the answers there yet until we need more data to treat those patients with a drug that albeit is very tolerable and doesn’t cause mortality in almost all patients still has toxicity and impacts quality of life. I just saw a number of my patients came in for their two-year end of therapy on monarchE and they were very happy. I tell you. Okay, no more diarrhea, no more fatigue, my hair might get thicker. There’s a few different things that people really struggled with during their two years albeit everyone would take it again. So that’s my approach right now.

Giuseppe Curigliano:
So, Fatima, your comment?

Fatima Cardoso:
So I think there is one piece of data that bothers me a little bit. So if you look at the curves of the trials with the palbo and if you try to say, “Okay if they had reported more or less at the same time that monarchE reported, the difference was more or less also the same.” So what will happen to the curves on monarchE with longer followup will help me take a better decision. I am still indecisive for the moment and I still would like to see more data.

Having said that, particularly for those who have very, very high risk, I think I would like to optimize endocrine based therapy and one way of optimizing is clearly adding the CDK4/6 because we see that in the metastatic setting. So I don’t have access to it. I cannot do it. It’s not approved or accessible for us to do it. But I understand that in some circumstances, it should be okay. But I would cautious that it’s not for every single patient with ER-positive/HER2-negative disease and I’m afraid that because it’s freely available that it might be an exaggeration of the use for every single patient and I think also here we have a very important question to be answered in the future is, is there a circumstance where this optimization can replace chemotherapy? And I think it will be very interesting to run trials into that perspective.

Giuseppe Curigliano:
Thank you so much. Optimizing the local and systemic treatment for the early breast cancer was the [inaudible] of St. Gallen 2021. So thanks a lot to Lisa Carey, Hope Rugo, Fatima Cardoso. This is Giuseppe Curigliano from Milan and see you to the next VJ Oncology meeting.

Fatima Cardoso:

Personal financial interest in form of consultancy role for:
Amgen, Astellas/Medivation,
AstraZeneca, Celgene, Daiichi Sankyo, Eisai, GE Oncology, Genentech, GlaxoSmithKline,
Macrogenics, Medscape, Merck-Sharp, Merus BV, Mylan, Mundipharma, Novartis, Pfizer,
Pierre-Fabre, prlME Oncology, Roche, Samsung Bioepis, Sanofi, Seagen, Teva.

Institutional financial support for clinical trials from:
Amgen, Astra-Zeneca, BoehringerIngelheim, Bristol-Myers-Squibb, Dalichi, Eisai, Fresenius GmbH, Genentech, GlaxoSmithKline,
Ipsen, Incyte, Nektar Therapeutics, Nerviano, Novartis, Macrogenics, Medigene, Medimmune,
Merck, Merus, Millenium, Pfizer, Philips, Pierre-Fabre, Roche, Sanofi-Aventis, Sonus, Tesaro,
Tigris, Varian, Wilex, Wyeth.

Non-Financial disclosures:
Chair ABC Global Alliance and ABC Consensus Conference and Guidelines. Member/Committee
Member of ESMO, ESO, EORTC-BCG, IBCSG, SOLTI, ASCO, AACR, EACR, SIS, ASPIC.

Lisa Carey:

Dr. Carey has no conflicts of interests to disclose, but wishes to disclose the following relationships. Dr. Carey does not personally receive any compensation from the following pharmaceutical companies and any monies are to a University of North Carolina Lineberger Comprehensive Cancer Center account, of which she does not have any signatory authority:

Sanofi-Aventis, Novartis, G1 Therapeutics, Genentech / Roche, GSK, AstraZeneca / Daiichi Sanyo, Aptitude Health

Giuseppe Curigliano:

Prof. Curigliano reports personal fees from Roche, Pfizer, Novartis, Lilly, BMS, Samsung, Astra Zeneca, Daichii Sankyo, GSK, Seagen, grants from Merck, other from Ellipsis.

Hope Rugo:

Research support for clinical trials through the University of California from:
Pfizer, Merck, Novartis, Lilly, Genentech, Odonate, Daiichi, Seattle Genetics, Eisai, Macrogenics, Sermonix, Boehringer Ingelheim, Polyphor, Astra Zeneca and Immunomedics.

Other:
Has received travel support from: Daiichi, Mylan, Pfizer, Merck, AstraZeneca, Novartis, and Puma. Honoraria: Puma, Mylan, Samsung.

Consulting/advisory:
PUMA, Samsung