Comparison of patient risk classifications by genomic assays for patients with early-stage breast cancer

DR LOVE: The main paper I want to ask you about was very recently published in The International Journal of Cancer, which looked at something that there’ve been many papers on before: “Summary of Head-to-Head Comparisons of Patient Risk Classifications by the 21-Gene Recurrence Score^® Assay and Other Genomic Assays for Early Breast Cancer.” Andy Seidman was on the paper, Zsuzsanna Varga. And then also there is an editorial, Joe, that you wrote in Nature Partner Journals, “Prognostic Gene Expression Assays in Breast Cancer: Are Two Better Than One?” Before we kind of get into how things are played out in practice, I just want to start with this generic question that’s raised by these 2 papers.

And Eric, the paper by Andy Seidman et al concludes that — they go through 5 genomic assays here, and they say, “Each risk-stratify the patients differently, carrying implications for the potential use of adjuvant chemo. And as such, these assays should not be used interchangeably.”

I’d like to ask you in general, Eric, what your take is on this question that’s been going around for a long time. We know that some have better clinical data. Of course, that’s where we follow, but in terms of what we know when you look at the same patient, different assays, what you see.

DR WINER: It’s a challenging question, because on the one hand, there are these multiple assays. Individually they’ve all been shown, with varying methodologies, to be prognostic. The prediction in terms of chemotherapy benefit has been somewhat more limited. And I ultimately think that there’s probably not a role for ordering multiple tests.

I don’t know what to do with multiple tests. And just in practice, I personally tend to order a test, and I order a test that I think is the best established and for which we have the most data, which is the Recurrence Score. But you’ll get arguments from different people about different tests.

DR LOVE: The one question that I haven’t been clear about is, at one point I had the impression that if you did the same assay on the same patient, you kind of saw the same thing. And then I started to see that maybe you don’t see this, in terms of even how they classify them, let alone the data. What do we know about this question?

DR WINER: I mean, I think that there are certainly some tumors that are consistently low-risk cancers when looked at with multiple assays. I think as is often the case, where you get the most discordance are those tumors that are in an intermediate zone. And I will just tell you a very brief anecdote.

I always say that I don’t think that there’s a role for 2 assays. That said, I recently saw a patient who had, by what I thought were clinical criteria, relatively low-risk disease, although she did have a lymph node involved. But she had a low-grade tumor that was strongly ER- and PR-positive, and she came back with a high score on one of the assays, specifically, the 70-gene assay. And if you looked at it carefully, she was on the very low end  of the high range. And it just didn’t make a lot of sense to me. And so we actually did do another assay. We did a Recurrence Score, which was somewhat lower, enough so that we decided together not to proceed with chemotherapy in this person.

DR LOVE: When you say somewhat lower, what was it?

DR WINER: The score was 18, and she was a woman who was 60.

DR LOVE: Eighteen. Okay, that’s low.

DR WINER: For me, that’s low. I mean, I’m respectful of different people’s definitions of these things, but 18 in a woman with a well-differentiated, strongly ER/PR-positive cancer who wanted to avoid chemotherapy was a slam dunk for me. And who was very clear that if the benefit were less than a couple of percent, she didn’t want it.

DR LOVE: Joe, this is a perfect case, illustrates the point exactly. I’m really curious, Joe, for your take on this question. Here you have a patient who one assay is reading low, really, particularly in a postmenopausal woman. Another one’s reading high. Is that what you see when you do the same assay? And what are your thoughts in general, Joe, about the interchangeability here?

DR SPARANO: Yes, these are important clinical issues, and the paper that you’re referring to clearly demonstrated that these assays don’t classify patients identically and are not interchangeable. I agree entirely. In fact, there’s only about a 50% concordance, so it’s almost like flipping a coin. They do agree, as Eric alluded to, at the ends of the spectrum, if the scores are very low or very high with the Recurrence Score. And the same is true for the 70-gene assay. It’s reported as a binary test, but it truly is a continuous test result if you look at the report. And I’d also agree with what Eric said: I think you need to choose the assay that you have the greatest comfort level with as it relates to the level of evidence supporting its use in the clinical context in which you’re using it.

Prognostic gene expression assays in breast cancer: Are 2 better than 1?

DR SPARANO: And lastly, regarding that editorial that I wrote about 2 gene expression assays better than 1, it was an editorial in relation to the results of the PROMISE trial, which was a study in which patients who had what was at that time defined as an intermediate Recurrence Score of 18 to 30 had a 70-gene assay performed. And the results were pretty predictable in that when you apply a binary test to a circumstance where you have an indeterminate result, you’re going to get a positive or a negative answer.

Of course, that trial didn’t report the clinical outcomes, so we don’t know whether or not the direction provided by the 70-gene assay in that context was an appropriate recommendation. In fact, the use of the assay in that setting led to more chemotherapy use in that particular population of patients.

The bottom line is, I think you have to pick an assay that you’re comfortable with, and you can’t ignore the clinical situation in which you’re ordering the assay. That’s critical. And those critical features include the age of the patient, the grade, the tumor size and, in circumstances where you’re using it in patients with positive lymph nodes, the nodal status. Having said that, generally most people would just use the assay if they’re going to use it in patients who have positive lymph nodes, in those who have 1 to 3 positive lymph nodes and in generally older, postmenopausal women.

Implications of gene expression assay results for therapeutic decision-making 

DR LOVE: I want to actually take this as a starting point to kind of go through — you mentioned the issue of what are some of the clinical factors, but another really important clinical factor, I think, is the issue of what the patient’s attitude is about treatment, particularly chemotherapy. And I’d really like to spend a lot of our time today kind of dissecting out how that really plays out in your own practices so our listeners and viewers can really get an idea of how you take all the science and really integrate it at a practical level.

Eric, it’s interesting, when you mention general medical oncologists, we’re used to being in situations where you have multiple options, theoretically. We have 6 checkpoint inhibitors approved right now — 6. But people follow the data. If you only see data in pembro, then you use pembro. Everybody kind of thinks that maybe they’re the same, but they go with the data. Really the first time I started to hear that, Eric, was about genomic assays. We did even programs, I remember, in the beginning, about what kind of data do you need to make a decision? Any thoughts on this, Eric?

DR WINER: In terms of what kind of data you need?

DR LOVE: And what it means. How the data drives what we do.

DR WINER: Yes, so look, I think that Joe’s study, the TAILORx study, provided a tremendous amount of data in terms of what we know and what we know more than from retrospective studies. What we know in a prospective setting with a large number of patients, thousands of patients, about the both prognostic and predictive value of the Recurrence Score. And what we don’t have at the moment are data with that same test prospectively in patients with positive lymph nodes.

But I think this is where — and we’ll get to this issue, I’m sure — in some of the cases, I think this is where sometimes you just have no choice but to begin to make some extrapolations. Because we are never going to have large prospective studies in absolutely every situation that we want it. We can’t afford it. It’s just not going to happen. But I do think that we need more than just retrospective series most of the time. And the stronger the data from 1 individual study, the more comfortable I am beginning to extrapolate from that study to other situations.

And I would liken this to what happened with HER2-positive breast cancer, where in 2005 we had the results from 2 large trials showing the unequivocal benefits of trastuzumab in HER2-positive, node-positive breast cancer. And it didn’t take very long for people then to begin to extrapolate to the node-negative setting. And it’s really, in many ways, quite analogous.

DR LOVE: And yes, I remember those days, because I think it was within 2 or 3 months, we were documenting a complete change in clinical practice. Like, instantly people were changing their practices.

Design and results of the Phase III TAILORx trial evaluating chemoendocrine therapy versus endocrine therapy alone for patients with ER-positive, HER2-negative, node-negative breast cancer and an intermediate 21-gene assay Recurrence Score^® (RS)

DR LOVE: Let’s maybe kind of shift over and talk a little bit about how this applies to practice. And before we actually get into the cases, because I want to start out with a couple of node-negative intermediate situations, Joe, which has really been the big question over the last few years, that now we’ve seen a lot of data with over the past year, really. Can you just summarize a little bit, Joe, about what the TAILORx study, the overall design of the study and what the key points have been along the last few years in terms of data being presented?

DR SPARANO: Sure. The TAILORx trial was designed to address that exact clinical scenario you mention: patients who were node-negative and, of course, had hormone receptor-positive, HER2-negative disease. Of course, we dialed down the intermediate range of Recurrence Score from 18 to 30 down to 11 to 25 for a variety of reasons. Number 1, the initial validations studies were done in patients who had HER2-positive disease. About 10% of those patients, both the B14 and the B20 trial, about 10% of the patients were HER2-positive. And almost all of those patients had a high Recurrence Score. If you excluded those patients — and there was recent publication on that from Chuck Geyer in PJ Breast Cancer. It demonstrated that you still saw that high Recurrence Score. Using a cut point of 26 or higher was predictive of chemotherapy benefit.

Secondly, the postmarketing data that was available that the score was being used in clinical practice indicated that physicians were using the test selectively in clinical practice, which was entirely appropriate. They were using it in situations where there was therapeutic equipoise — typically patients who had tumors between 1 and 2 centimeters and intermediate grade. And those patients generally tended to be younger than the average patient with an early breast cancer. And when you looked at the Recurrence Score distribution in that group, it was closer to the 11 to 25 range than it was the 18 to 30 range.

Thirdly, the trial was designed to show noninferiority in that group. In other words, we thought we were overtreating patients in that group. And we used data from the B-20 trial to test that hypothesis. We actually looked at the results and published this in 2008, showing that when you reanalyze the data from the B-20 trial at that time, including the HER2-positive population, and you used 11 to 25 as the cut points, you saw no chemotherapy benefit in that group. There was about a 5% distant recurrence rate irrespective of chemotherapy use, number 1. And number 2, the treatment effect associated with chemotherapy in the group that had a Recurrence Score of 26 or higher demonstrated a similar benefit than if you used a score of 31 or higher.

When the results came in from TAILORx, it turned out that the B-20 results were very much aligned with the results that we saw with TAILORx. Of course we did demonstrate noninferiority of endocrine therapy alone in the randomized population. However, there was a treatment interaction, this 3-way interaction between age, Recurrence Score and chemotherapy benefit, such that those patients who were 50 or younger who had a Recurrence Score in the range of 21 to 25 and possibly down to as low as 16 were having some chemo benefit.

And there was some criticism about the trial results and the analysis because of that exploratory analysis, but I thought it was and the study team thought it was important that if we were going to definitively walk away from potentially curative therapy in early-stage breast cancer, we wanted to make sure that there was no subpopulation of patients who were left behind.

To summarize, yes, the trial did meet its primary endpoint of showing noninferiority for endocrine therapy compared with chemoendocrine therapy in this particular group. And I think that’s helped inform practice, and it kind of solidified what I think the trend has been over the last decade since the Recurrence Score became available.

Clinical implications of the TAILORx trial for treatment decision-making for patients older than 50 with ER-positive early breast cancer (EBC)

DR LOVE: And then Eric, in terms of distilling it out into a practical message, because we saw some more data at ASCO looking at clinical factors, particularly higher risk factors such as tumor size/grade. Let me just give you my 30-second take on it, Eric. You can tell me if you agree or disagree, okay? Of all the stuff that came out the last couple years.

Part A, postmenopausal women, up to 25, node-negative, no benefit of chemotherapy, even in larger tumors. Eric, agree or disagree?

DR WINER: I agree, and actually I would say that in the range of 26 to 30 is still a pretty gray area for those postmenopausal women.

DR LOVE: I was just gonna say, and part 2 would be 26 to 30’s the new intermediate, but you just said that. Okay, so that’s part A. Joe, do you agree with this summary in terms of postmenopausal women?

DR SPARANO: Yes, I agree. That’s a fair comment. I think there are tools that are currently available that can help provide more refined or perhaps accurate estimates of recurrence risk  specifically in that population.

DR LOVE: And I want to get into that, because I noticed you both wrote your cases up differently. I was really curious about some of the things in Joe’s write-ups, which I think you just alluded to. Kind of getting into second-level clinical — that’s what I was kind of talking about.

Prediction of chemotherapy benefit by RS for patients 50 years or younger with ER-positive EBC

DR LOVE: But let’s get back to the last couple years. Again, 30-second take. Eric, women under 50, mostly premenopausal, almost all got just tamoxifen, not ovarian suppression. And there it looked like intermediate, maybe there was some benefit of chemotherapy. And again, the clinical thing being not so clearly not relevant, so questions about women under 50, but also the question of, Eric, is this because of chemo? Is this because of chemo-induced ovarian suppression, which maybe could be considered as an alternative? Eric, agree or disagree?

DR WINER: I just have to make a comment. I mean, I think I’d basically agree. I think for women whose tumors have scores of 21 to 25 who are premenopausal, I think that that benefit that we see in TAILORx may be coming from ovarian suppression. I don’t think we can say we know for sure that that’s the case. And so when I see a patient, I have a conversation about the overall trial results, the subset of women who are premenopausal, in the low 20s, and then the potential value of ovarian suppression versus chemotherapy.

DR LOVE: I think that still fits within the overall algorithm I just described, but Joe, you tell me, again, in the women under 50, presumably premenopausal, how are you thinking things through now? And are you more likely in general to do ovarian suppression anyhow to start with?

DR SPARANO: I will discuss the chemotherapy benefit that was seen in TAILORx. And I’ll explain that it’s possible that this benefit could’ve been due to the induction of early menopause from chemotherapy and that this patient can get similar benefit from ovarian suppression plus endocrine therapy. Given that choice, many people would choose the ovarian suppression plus endocrine therapy but are concerned about the at least 5-year commitment to it and the uncertainty about what happens at the end of the 5 years of ovarian suppression.

Others may prefer the chemotherapy because of its shorter course and also because of the uncertainty. Because the trial, in fact, did show a benefit from chemotherapy in this subset of patients. And we’re surmising but haven’t proven that more aggressive endocrine therapy could produce similar benefit.

DR WINER: Neil, can I make a comment?

DR LOVE: Sure.

DR WINER: I agree with what Joe said, but I think the other just important thing for everybody to keep in mind throughout this discussion is that there’s fundamentally a difference between a woman who has a 7-mm node-negative cancer and a woman who has a 3.2-cm node-negative cancer. And that difference, of course, is the level of risk that she faces. And while there may be benefits in both of those patients for various interventions that go beyond standard endocrine therapy, it’s going to be of very different magnitude. And I think it’s something that’s sometimes difficult for people who are practicing, but I really think it’s something people have to pay attention to.

DR SPARANO: Can I expand on that?

DR LOVE: Sure.

DR SPARANO: Because I think that’s a really, really critical issue. Because the factors that are associated with a recurrence, that drive recurrence risk, include what the underlying risk of recurrence is in the absence of any systemic therapy. Number 2, the sensitivity to endocrine therapy and the residual risk that’s present after an appropriate course of endocrine therapy. And then the third is, and the critical factor where I think the Recurrence Score provides clarity, is the sensitivity to chemotherapy. And we really need ways to factor in all of those 3 equations.

Benefit with the addition of ovarian suppression to endocrine therapy for premenopausal women with ER-positive breast cancer at high risk for recurrence

DR LOVE: It’s almost becoming a 3-dimensional puzzle, in my mind, as I listen to this. And Eric, one thing I’m kind of curious about, we’ve heard this term endocrine sensitivity. People talk about the fact that if you have a low Recurrence Score, for example, it’s not just that you don’t benefit from chemo. But maybe you have higher ER, you’d benefit from endocrine therapy. I can’t really figure out whether that has anything to do with choice of endocrine therapy. But how about if I just kind of flip it the other way, Eric, and say, can you describe for me a scenario right now where in a premenopausal woman whether there’s Recurrence Score involved or not, where you’d be very comfortable using tamoxifen and a situation where you’d be a lot more comfortable using ovarian suppression or ablation?

DR WINER: Sure. To begin with, I do basically buy into the results of SOFT, which would tell us that for a premenopausal woman who is thought to be at low enough risk that she doesn’t need chemotherapy — and that was not with any genomic assay, that was just in the eyes of the doctor — that the risk, at least in that study, was very, very low and tamoxifen was essentially as good as any other therapy. For a woman who’s at higher risk, either because of tumor burden or because of various biologic features like grade or HER2, that in that setting ovarian suppression with either tamoxifen or an aromatase inhibitor was clearly beneficial.

For me, and we usually do have benefit of a Recurrence Score in making these decisions, if I’m faced with a premenopausal woman who has a tumor that I wouldn’t consider chemotherapy for, and particularly if she has a Recurrence Score that’s in what I think of as the low range, which whether you want to call it below 16 or below 20, but somewhere in there, I’m pretty comfortable with tamoxifen alone.

Case (Dr Winer): A woman in her early 60s with Grade III ER-positive, HER2-negative, node-negative breast cancer has a 2.1-cm tumor and a RS of 25

DR LOVE: We have this series — I don’t think either of you have been involved with it — called Visiting Professors. We’ve only done it a few times. We send investigators to the practices of general oncologists, and they spend the day making rounds in their clinic. They see the patients and the family, and then we sit and talk about it. The reason it’s on my mind is, we just did one in lung cancer. We had Ben Levy from Hopkins. We flew him out to Springfield, Missouri.

Anyhow, he saw Ray Lobins and spent the whole day just seeing people with metastatic lung cancer who didn’t have tumor mutations. And it just got published. I was so excited, because Ray saw the thing, and he sent me back and email saying “Hey, guess what? That patient that Dr Levy saw with me who is on fifth-line therapy and he recommended ipi/nivo, even though it’s fifth line, has had a great response.” Of course, Hopkins is big on checkpoint therapy. But it really got me thinking a lot about the idea of really being able to see the patient and how much that helps when you have a case presented to you in an academic program like this one.

What I want you to do today is, I want you to take us into your clinics. I want to really get the feel, because we were talking about where the patient fits in these decisions. And as much as you can, I’d like you to tell me about the patient, whoever’s with the patient and really what they thought about chemo and kind of how you made this decision with them. And I’m going to be really curious, because, as I said, I see that you write the cases up a little differently.

We’ll start with you, Eric, because you write them up like most people do. And we’ve been talking about node-negative. You’ve got a postmenopausal patient who’s 63 years old, 2.1-cm, Grade 3, interestingly, node-negative cancer. ER 95%, PR 80%, HER2 1+, status post mastectomy, sentinel node biopsy, presumably negative, and has a Recurrence Score of 25. Eric, I’d like you to kind of set the stage and present the case to Joe about this patient and how these numbers worked in her discussion.

DR WINER: Yes. Okay. I don’t think that any more than any other patient there’s anything dramatically remarkable about this person. She was not totally opposed to the idea of chemotherapy, like most of our patients. She wasn’t totally excited about chemotherapy, either. And my general experience is that women who are a little older — and we all have to be careful about calling 63 very old these days — but are a little less inclined, oftentimes, to take a course of chemotherapy, and legitimately so.

DR LOVE: What was her life situation? And how much would chemotherapy interfere with it?

DR WINER: I think chemotherapy interferes with everybody’s life situation. And she was somebody who worked and had a pretty busy, active life. I mean, she was not impaired by any major medical problems. That said, we do know that some of the serious complications of chemotherapy are somewhat more common in older than in younger patients.

Risk of recurrence and benefit from chemotherapy for postmenopausal women with EBC

DR LOVE: I’m also curious, could you tell me also, like, her level of interest? Did she want numbers? Was she reading about it? Or was she more like, “Tell me what to do, doc”?

DR WINER: No, she was interested in a little bit more of a potentially quantitative discussion. And I often sit with patients and I describe the various outcomes, and I say, “Would you take chemotherapy if it improved your chance of having a recurrence from 15% to 13%?” I tend to be pretty specific with people, and I try to come up with the sweet spot for them. In my mind, this was somebody who had a Recurrence Score that was on the high end of TAILORx, although admittedly I think 26 to 30 is a gray zone for postmenopausal women.

DR LOVE: But she’s 25.

DR WINER: She was 25. I mean, of course we all realize that there’s no difference between 25 and 26, really.

DR LOVE: Just saying.

DR WINER: One point doesn’t — no. You gotta pick some point. She had a tumor that wasn’t particularly small but wasn’t huge. The Grade 3, I don’t actually make that much of, only because grade does not always predict Recurrence Score.

DR LOVE: Incidentally, was that grade by your pathologist at Dana-Farber?

DR WINER: That was graded by our pathologists, and I do accept the grade by our pathologists. And it makes me think a little differently. On the other hand, we used to not get Recurrence Scores in patients with high-grade tumors, and we found that a fairly high proportion of them turn out to have scores lower than we expected.

But for all these reasons, I think that chemotherapy is in the discussion, although she was somebody who was really not interested in chemotherapy unless the benefit were in the range of 3% to 5% or greater. And when I described the possibility that it’s at least conceivable that she could be 0 benefit from chemotherapy, she really wasn’t very interested.

DR LOVE: And I guess you hear people saying maybe there’s a 1% chance you’re going to cause a real big problem with chemotherapy. That may be on the other side.

Joe, based on the numbers that you hear, and you also hear a description of the patient, which kind of sounds like a lot of oncologists have that number, 3% to 5%, in their head, for some reason. But first of all, do you agree with Eric’s assessment of the numbers in this case and how he sees the numbers?

DR SPARANO: Yes, I agree entirely. I think the Recurrence Score can help identify those patients who clearly don’t benefit from chemo and that small, relatively small subpopulation of probably maybe 15% or 20% who derive a very large absolute benefit from chemotherapy. But there is this gray zone. And this patient definitely fits in this gray zone.

I will say this: I would agree with Eric’s estimate of what this patient’s absolute benefit from chemotherapy might be. And it’s really comparable to the use of chemotherapy in an unselected population. If you didn’t have the Recurrence Score, and counseling patients, you would normally be indicating that the absolute benefit would be in the range of about 3% to 5%.

The approach that I would have taken here would have been the same. I would’ve suggested that chemotherapy was a default recommendation because of the potential benefit associated with it. And I would’ve used the RSPC, as I alluded to before, to help provide a more refined estimate of what her recurrence risk would be, because the Recurrence Score alone may slightly underestimate what her true recurrence risk is.

DR LOVE: Yes, I want to get into that RSPC, because when I saw that, I was kind of like, “Oh, I didn’t realize people were still doing that.” Anyhow, Joe is. But your write-up doesn’t say what happened, but I’m assuming she got put on an AI, Eric?

DR WINER: She got put on an AI.

DR LOVE: And how long’s it been now?

DR WINER: In the range of about a year.

DR LOVE: And which AI did she get, and how’s she doing on it?

DR WINER: I’m a simple guy, and I like to put everybody on the same AI so when they call me up and I have to refill a prescription I don’t have to think about what somebody’s on. She’s actually on letrozole and doing fine, didn’t have a lot of side effects from it. Of course some people do, and that’s a whole different discussion about how we manage those side effects.

Significance of clinical risk category in prognosis and prediction of chemotherapy benefit by age and RS in the TAILORx trial

DR LOVE: And I guess Joe, the next time the patient comes in to see Eric he can say, “You remember how you were asking me about grade and the fact that you had Grade 3 and you were a little bit worried about it? Guess what? There’s some data that was just presented on that.” Because am I correct in saying that this lady, Joe, kind of fits into the high clinical risk thing that you just presented?

DR SPARANO: Yes, and our initial intention when we designed TAILORx was to use Adjuvant! Online and to develop models that integrated the Recurrence Score into Adjuvant! Online. Of course by the time the study was published, Adjuvant! Online was no longer available. But we did know from the MINDACT trial that they used Adjuvant! Online to assign a binary risk category of high or low that was calibrated to the Adjuvant! Online tool version 8.0. And in that classification, low risk was defined as a tumor of up to 1 centimeter high grade, up to 2 centimeters intermediate grade and up to 3 centimeters low grade. And it was prognostic. It did prove to be prognostic in MINDACT.

Not having the true Adjuvant! Online tool available, we prospectively applied it to TAILORx. And the results were, as one would have predicted, in that this binary stratification, this clinical risk stratification, added prognostic but not predictive information, mainly in patients who had an intermediate Recurrence Score of 11 to 25 or high, greater than 25. But it didn’t add predictive information from chemotherapy benefit. And that was the expectation going in, and that proved to be the case in the analysis.

DR LOVE: Just to clarify, because again, with all these studies flying by everybody, I try to just figure out what’s the main thing to take away from it. But the thing that I saw there, Joe, that stuck out, was tumors more than 3 centimeters and tumors that were Grade 2 to 3, particularly 3, as kind of sticking out of — oh, even if they have that and they’re postmenopausal, no benefit? Is that a correct take, Joe?

DR SPARANO: Yes, I mean, I would agree with Eric’s comment that a Grade 3 tumor doesn’t always guarantee a high Recurrence Score, and that I have likewise been pleasantly surprised to find a fair number of patients who have Grade 3 tumors can have a low or intermediate Recurrence Score.

DR WINER: Remember, the grade is actually — it’s divided as Grade 1, 2, 3. And I can’t tell you where she fell in the Grade 3, whether it was a Bloom-Richardson of 8 or 9. But there’s actually a little bit more of a continuous variable in grade than we realize sometimes.

Perspective on the use of Ki-67 as a biomarker in breast cancer

DR LOVE: All right, so I have to ask you. Joe, do you order Ki-67? And if you don’t and somebody else orders Ki-67, do you look at it? Or you don’t look at it? And does it depend where they got the Ki-67 done, Joe?

DR SPARANO: At our institution, we don’t do Ki-67 routinely. I do see cases referred from outside where it has been done. It generally doesn’t impact my decision — doesn’t add anything to the Recurrence Score. But having said that, it can impact the ordering of the Recurrence Score in the first place. So, for example, if you have a patient with a small tumor as a low grade, I’ve seen circumstances where a Ki-67 was ordered. It wasn’t really ordered, it’s just done at that institution, and it turned out to be high, and that prompted a 21-gene Recurrence Score, and that turned out to be high.

And conversely, you can see the circumstances where someone who maybe has a larger tumor or a higher grade, maybe 8 or so, has a Ki-67 done, and it may be low. And that may drive the use of a Recurrence Score in that situation.

The answer to your question is, no, I don’t use it. And I don’t use it in isolation, because we don’t do it at our institution, and when I do see it done, I generally see it done in the context of also having the information from having the Recurrence Score.

Clinical and genomic risk to guide the use of adjuvant therapy for breast cancer

DR LOVE: Eric, another question I have for you is, when you saw those data looking at clinical factors in node-negative patients, again, tumor size, grade, et cetera, did it strengthen, from your point of view, this concept of biology versus anatomy, which has been going on for a long time now? Did it make you more comfortable, for example, using a genomic assay in the node-positive situation, because this biology being more important than anatomy seems to be playing out in TAILORx?

DR WINER: I think they’re both incredibly important, and I think that was the lesson from MINDACT. I think that there’s a lesson there from TAILORx. And, of course, then there’s the other issue, which is that as tumors are more advanced in terms of stage, a higher proportion of them tend to be somewhat higher-grade tumors with higher Oncotype scores. But it’s by no means absolute, and there are plenty of patients who have node-positive disease and probably even multinode-positive disease who have relatively low scores.

I actually think that that’s the area that we’re really in a no-man’s land in terms of what is the optimal treatment for a patient who has multiple positive lymph nodes and a very low Oncotype and a strongly ER/PR-positive cancer? And maybe we’ll get to that later on. But I think at this point in time we all give chemotherapy in that situation, but when I do it, I give chemotherapy, and before I walk into the room I’m shaking my head asking myself what I’m doing.

DR LOVE: Are you asking yourself, “Am I going to be using endocrine therapy and a CDK inhibitor in a couple years?” in this situation?

DR WINER: I think there’s a reasonable chance that I will be, and I think that that may help with some of these decisions, but it won’t totally solve them, because many of those patients if not most of the patients who were enrolled in the various adjuvant CDK4/6 inhibitor trials also received chemotherapy.

Implications of the TAILORx trial results for the use of neoadjuvant endocrine therapy

DR LOVE: Joe, I gotta tell you, I had an incredible case from Matt Goetz from the Mayo Clinic. This patient presents with locally advanced ER-positive, HER2-negative disease, premenopausal woman — this dilemma. Are you going to give her chemo when you don’t think — but, unfortunately, they found she had a couple of bone mets. They give her a CDK inhibitor and endocrine therapy, and she had a complete clinical response. Now, 1 case. I don’t know whether you’ve ever seen anything like that. But I just wondered, Joe, do you think that our — not to get too far off the topic here — but do you think that our use of neoadjuvant endocrine therapy might increase use of genomic assays in the neoadjuvant setting if we get better endocrine therapy?

DR SPARANO: Yes, I agree entirely, but for the reverse reasons. Actually, I believe that the findings from TAILORx indicating really marginal or lack of benefit in the majority of patients who are older than 50 who have estrogen receptor-positive, HER2-negative tumors suggests to me that we’re overestimating the potential benefits from chemotherapy and underestimating the potential value of endocrine therapy.

And in circumstances like this, where patients who present with more locally advanced ER-positive disease, maybe a patient with a 2-cm tumor but a positive lymph node who would like to avoid chemotherapy but where the surgeon would like to cytoreduce the tumor, at our institution, we’re using more neoadjuvant endocrine therapy and then using the information from the Recurrence Score to guide the use of chemotherapy after surgery. In that circumstance, only about 15% to 20% of patients will actually have a pathologic complete response rate. Perhaps a slightly higher proportion would have a modified PEPI score of 0, meaning tumor less than 2 centimeters, negative nodes and a complete cell cycle arrest on Ki-67. Although again, we don’t do Ki-67 at our institution.

But I do think that the findings from TAILORx have provided some scientific rationale for exploring a greater use of neoadjuvant endocrine therapy, both in the clinical trial setting and I think also outside the context of a clinical trial.

DR WINER: I think Joe made a really important point, which is that in the patient who gets neoadjuvant endocrine therapy — and we treat a lot of patients with neoadjuvant endocrine therapy, both on and off trial — that it’s the Recurrence Score that you will obtain on the baseline tissue that will drive the decision about chemotherapy afterwards. We know that neoadjuvant endocrine therapy will do a better job probably than neoadjuvant chemotherapy in this situation in terms of shrinking the tumor and allowing for conservative surgery.

But the 1 thing you can’t do, unlike the triple-negative and HER2-positive setting, is probably use the pathologic result achieved from neoadjuvant endocrine therapy to then drive decisions about further chemotherapy. That’s where you need something like the Recurrence Score.

DR LOVE: I’m just saying, maybe naïvely, but if I’m in a situation where I really want to get a tumor response from endocrine therapy in that situation, I imagine it’s not very easy to do or maybe impossible to actually access it. I mean, wouldn’t you say if you add a CDK inhibitor you’re going to see a greater response rate?

DR WINER: In the case of Matt Goetz’s patient that you were referring to, she had metastatic disease, so it was pretty easy to justify using a CDK4/6 inhibitor.

DR LOVE: Yes, yes. Of course. But you wouldn’t have done it if she didn’t have mets.

DR WINER: Right. Outside of a trial, it would be hard to justify it. Again, we’ve had a couple of trials where we’ve done that in the neoadjuvant setting. But in truth, with neoadjuvant endocrine therapy alone, you can quite often convert someone from needing a mastectomy to needing less if that’s what they’re interested in.

DR LOVE: Yes, that’s a good point.

DR SPARANO: I will add an important point: There have been randomized trials looking at neoadjuvant endocrine therapy alone or in combination with CDK4/6 inhibitors. They generally have shown, as expected, that you get greater cell cycle suppression with the CDK4/6 inhibitor. That hasn’t necessarily translated into higher clinical or pathologic complete response rates. The other variable here is duration of endocrine therapy. We typically give it for 4 to 6 months, like chemotherapy, and that may not be the optimal duration of neoadjuvant endocrine therapy. You may need to give if for a longer period. How long? I don’t think we really know what the sweet spot is. Obviously you’d like to give it for a period of time that maximizes the potential for cytoreduction but minimizes the potential for developing endocrine therapy resistance.

DR LOVE: Yes, and I certainly am not in any way trying to encourage that outside of indication strategy, but I’m just kind of verbalizing what I think people who are listening to you all might want to say to you or ask you, and that was one of them.

Case (Dr Sparano): A woman in her early 60s with Stage IIA ER-positive, HER2-negative lobular breast cancer, a RS of 19 and a high ESR1 RNA score receives adjuvant docetaxel/cyclophosphamide followed by anastrozole

DR LOVE: I really want to find out about the workup here, Joe, because you’ve got this 60-year-old Asian lady with Stage IIA pT2 N0M0 lobular carcinoma. She has bilateral mastectomies, the other side for prophylaxis, ER/PR-positive, HER2-negative, Grade 6 of 9, which you might want to elaborate on, tumor size 4 centimeters and Recurrence Score of 19. So presumably she — yes, she’s postmenopausal, but then you have a bunch of other things here, Joe. ESR1 RNA score, and hers was high, 9.6. So why do you get that? And what does this mean?

DR SPARANO: Yes, so the ESR1 RNA score is one of the 4 results that you get as part of the Recurrence Score package. Of course, everyone focuses on the Recurrence Score, but the second I think really important piece of information one gets is the ESR1 score — RNA score — because it’s been shown that a high score identifies patients who are at higher risk of late recurrence after 5 years.

DR LOVE: This is just the regular ESR1 mutation?

DR SPARANO: It’s not a mutation. It’s the ESR1 RNA score, which contributes to the Recurrence Score. Of course, patients who tend to have low to intermediate Recurrence Scores are going to have higher ESR1 RNA scores. But it identifies a group of patients, I think, who might be more likely to derive benefit from longer-duration endocrine therapy. It hasn’t been confirmed in a prospective trial.

DR LOVE: This is the normal estrogen receptor level?

DR SPARANO: Yes. It’s not the receptor. It’s not the protein. It’s the message.

It’s the result by RT-PCR looking at the RNA message for production of the protein. And like I said, it’s baked into the Recurrence Score, but this provides a more granular look at it. If you do have a patient who has an intermediate or high Recurrence Score with a high ESR1 score, that identifies patients who could potentially be more likely to benefit from extended adjuvant therapy. Because it was tested mainly in the B-28 trial — that was a trial where patients, if they had node-positive disease they got tamoxifen concurrently with chemotherapy. They all got chemotherapy. Recurrence Score was shown to be prognostic in this setting, where all patients got chemotherapy. All patients got concurrent tamoxifen. But it was shown that in those who had a high ESR1 score that those who had an intermediate or high Recurrence Score that the Recurrence Score provided prognostic information beyond 5 years and the high ESR1 score was associated with a higher risk of recurrence, as was the high Recurrence Score.

DR LOVE: She also had Grade 6 out of 9. What is that?

DR SPARANO: Yes, this particular patient was an oncology nurse.

DR LOVE: Wow.

DR SPARANO: Certainly I wanted to avoid the use of chemotherapy in her — and she was diagnosed about 3 years ago. I really didn’t want to use chemotherapy, but the Recurrence Score wasn’t really low enough to make me comfortable sparing it. I guess I should have referred her to Eric.

And I think Eric probably would have said that she didn’t need chemotherapy, but unfortunately I think she just saw me, and I said, “Look. I can’t guarantee that you’re not going to have about a 5% benefit from chemotherapy.” And, in fact, when I plugged her information into RSPC, it was associated with about a 19% 10-year risk of distant recurrence. And if you use Recurrence Score alone without the RSPC, it was about 10%. I think now I probably wouldn’t treat this patient with chemotherapy. But at the time, with the information and the evidence level that we had, I thought chemotherapy was probably the best recommendation for her.

Tolerability and quality of life during treatment with chemotherapy

DR LOVE: And I see she got docetaxel/cyclophosphamide, a common choice in this situation. How did she tolerate it? Really curious about what it was like for her to get chemotherapy after being an infusion nurse.

DR SPARANO: She sailed through the TC. The TC wasn’t a problem. What’s been a problem for her has been the AIs.

DR LOVE: Wow.

DR SPARANO: She’s had to go through 2 different AIs because of arthralgias and pretty substantial musculoskeletal toxicities. We’re trying to find the sweet spot for her. As it relates to endocrine therapy, she could very well wind up on tamoxifen, which wouldn’t be my first choice, but it may be where we wind up.

Change in perspective for oncology professionals diagnosed with cancer

DR LOVE: What was your sense about what it was like to go through the chemotherapy?

DR SPARANO: I think certainly she was concerned about it, but she sailed through it really without having much in the way of short-term side effects or long-term side effects in terms of neuropathy. I did use the TC in combination with pegfilgrastim, which I think helped reduce the likelihood of her developing complications. Her hair did grow back after the docetaxel, so that was good. And so she really hasn’t had any of the long-term complications. And I don’t necessarily need to worry about the long-term risks of leukemia or myelodysplasia or cardiomyopathy, because we didn’t use an anthracycline.

DR LOVE: I once did a roundtable discussion of radiation oncologists, medical oncologists and urologists who had prostate cancer. It was pretty interesting how their view of taking care of patients was changed by this experience. Joe, do you have any sense that this experience changed her outlook on her own work?

DR SPARANO: Certainly having been an infusion nurse before, I think she understood the risks associated with chemotherapy. But on the other hand, she knew that the vast majority of patients who she treated with chemotherapy did well in this setting, in patients with early breast cancer.

In terms of how it’s impacted her care as a nurse, I think she said, like most healthcare professionals who experience a potentially serious illness and need to have some type of therapy, I think it makes them more attuned to the concerns and needs of the patient and it makes them a better clinician.

DR LOVE: I’ve got to ask, Eric, this topic of going through a difficult illness and how it affects how you view things is something you actually verbalized to a lot of people at the San Antonio meeting in the past. Any thoughts about this in your experience treating oncology professionals?

DR WINER: Yes, so I’m sure Joe and I both take care of a fair number of physicians and nurses. And it is a little different. And sometimes it’s easier, and sometimes it’s harder. But I do think that as with almost everyone, the experience of going through cancer treatment is somewhat life changing, not necessarily in a bad way, but people remember their life before the cancer treatment and remember their life after cancer treatment. And I think the one thing that people can take away from it that isn’t good is thinking that how they reacted is how everyone reacts.

But I don’t think most people do that, and I think that if you talk to people who have been through a cancer experience, they tell you that it’s given them a new view on how to take care of patients, and they feel, at least, that it’s made them that much more tuned in. Finally, I don’t think you have to have a cancer experience yourself, though, or some other health problem, to be tuned into taking care of patients.

And I just want to say that even for me, Joe, this would have been someone in a gray place 3 years ago. I suspect in the end I might have just given her endocrine therapy. I think it’s not so entirely clear cut. She has a relatively large tumor. She had a PR-weak tumor. You were in a place where you just don’t know what to do. And I often tell people that we can only make the best decisions we can at this moment in time, and you can’t go back and regret either doing or not doing something later.

Impact of the RS-pathology-clinical (RSPC) score on the assessment of recurrence risk for patients with breast cancer

DR LOVE: Joe, can you just talk a little bit about what the RSPC is and why you use it? A lot of people would have just looked at this case and go, “Oh, she’s got a 4-cm tumor. We know she’s got a higher risk.” This is exactly, again, the patient you all just presented data that relates to. But what exactly is the RSPC, and how much granularity does it add, Joe? Here you actually get a number, changes from 10% to 19%. That’s a big change.

DR SPARANO: Yes, it certainly is. It’s an online tool that’s available for physicians to order the test who are registered users. And it’s an algorithm that was developed from previous studies, previous data, including B-14 and ATAC. And of course it includes the Recurrence Score, but it also includes age, tumor size, grade and the type of endocrine therapy, whether it be tamoxifen or an aromatase inhibitor. Of course, that’s often driven by the patient’s age and menopausal status.

Currently, it applies only to patients who have node-negative disease. But in circumstances like this I think it could, as you see, could really drive the distant recurrence estimate up. This is a circumstance where relying on the Recurrence Score alone can actually underestimate what probably is the true risk of recurrence.

And I think it could also be helpful in the patient we discussed earlier, the 63-year-old with a Recurrence Score of 25. Because what that patient’s grade and tumor size is can really have a profound influence on what her underlying recurrence risk is, which ultimately may be that 1 piece of information that a clinician would use to make a chemotherapy recommendation or not.

DR LOVE: Very logically 3 years ago, even though this woman’s very reluctant to get chemotherapy, you recommended it? Because she had a 4-cm tumor, I guess, is the main reason, but the other reasons you talked about. Joe, if she came in today, right now, would you go, “You don’t need chemotherapy, period”?

DR SPARANO: Yes, I would agree that based on the results of TAILORx she would not need chemotherapy. She would fall in the clinical high-risk group, and we didn’t find that when we did the subset analysis. Clinical risk didn’t add predictive information. The other factor is that she had a lobular cancer.

DR LOVE: Right.

DR SPARANO: And some folks believe that lobular cancers may be less likely to benefit from adjuvant chemotherapy. And you will rarely see a high Recurrence Score in a lobular cancer, unless it’s a pleomorphic lobular. I have seen, actually, a couple recently. But it’s quite unusual. The Recurrence Score distribution is different in lobular cancer. It’s actually more shifted to the lower to low-intermediate range.

DR LOVE: Yes, I’ve heard a lot of intermediate lobular cases presented. I don’t know how common that is, but people present them a lot.

Case (Dr Winer): A woman in her late 30s with Grade II ER-positive, node-negative breast cancer and a RS of 20 receives treatment with ovarian suppression and tamoxifen

DR LOVE: Maybe we can just get a couple more maybe quicker cases just really focusing in on the numbers. Eric, you’ve got a 39-year-old premenopausal woman, so a young woman, 2.3-cm tumor, Grade 2, node-negative, ER 90%, PR 50%, status post lumpectomy and sentinel node biopsy, Recurrence Score of 20. I’m guessing you had that discussion you just referred to her. There may be some benefit for chemotherapy. It’s probably not very great. I don’t know when you saw this woman whether you already had the newer data from TAILORx, but what happened?

DR WINER: I had just that discussion, which is that there may be some benefit from chemotherapy. That that benefit may be able to be achieved through ovarian suppression, which by no means is necessarily an easy treatment, although if chemotherapy works partially through ovarian suppression, then it’s the same result, just easier than getting chemotherapy. And I think it’s very, very hard with someone to put any quantitation on the difference between ovarian suppression and chemotherapy. You can refer her to older studies done 2 decades ago that showed similarity in outcome between the two, but none of the studies were anything like what we would do in a modern era, and I think it’s hard to make too much of it.

My feeling here is that in somebody here who wants to take every possible precaution that it’s reasonable to take a course of chemotherapy. I would favor a nonanthracycline-containing regimen for such a patient just because of the potential for long-term risk with anthracyclines. And I think it’s reasonable — I would probably counsel, her, however, to take ovarian suppression and either tamoxifen or an AI. Probably tamoxifen.

DR LOVE: And when did you make this decision? And what did she do?

DR WINER: This was in the past several months. We had several long discussions, and she ultimately decided to get endocrine therapy.

DR LOVE: What kind of endocrine therapy?

DR WINER: Ovarian suppression and tamoxifen with the possibility that at some point in time we’ll switch to an AI.

I find that ovarian suppression plus an AI is a very difficult treatment for a premenopausal woman. I liken it to jumping off the high dive into menopause as opposed to wading into the water, which is supposed to happen when you go through a natural menopause.

DR LOVE: Joe, do you agree with that? And do you approach these patients more commonly with ovarian suppression and tamoxifen because of it?

DR SPARANO: I would have recommended chemotherapy as a default position. I think that’s reasonable with ovarian suppression plus endocrine therapy as an alternative, as Eric did. And I also agree with him that ovarian suppression plus an AI can be a very difficult treatment, especially for younger women. I would, though, generally recommend ovarian suppression plus the AI, at least to try it. And then if a patient is not able to tolerate it, then switch to tamoxifen. In the postmenopausal setting we do know that, from the BIG 1-98 trial, that that sequence resulted in outcomes that were comparable to patients who just received AI therapy alone. That would have probably been my recommendation.

DR WINER: Tam followed by an AI is about the same, too.

DR SPARANO: True.

DR LOVE: But you said chemo was the default position, I guess, technically, in terms of data and guidelines and stuff.

DR SPARANO: Sure.

DR LOVE: But in general, would a woman in this situation in your practice, Joe, end up likely on chemotherapy or not?

DR SPARANO: I would say it’s about 50/50.

DR LOVE: Really?

DR SPARANO: Yes. I mean, in terms of the data, in terms of the TAILORx and the more recent analysis that looked at clinical risk stratification for those women who were 50 or under who had a Recurrence Score of 21 to 25, there was somewhere in the range of a 6% to 8% absolute benefit — reduction in distant recurrence risk — in that range. Of course, this patient is 20, so she’s just below. She’s at the upper range. She’s 16 to 20.

DR LOVE: Wow.

DR SPARANO: In the 16 to 20 range, we found that about half of the patients overall had a Recurrence Score of 16 to 20 and low clinical risk. And those patients had no benefit from chemo. There was no chemo benefit. But if you were 16 to 20 and had high clinical risk, there was a benefit that was in the 4% to 5% range. Again, this may not necessarily be due to greater sensitivity to therapy but underlying recurrence risk, or it could be a combination of factors. She’s really just on the edge, and I think either decision would be appropriate.

Case (Dr Winer): A woman in her late 50s with ER-positive lobular breast cancer, a 5.2-cm tumor, 1 palpable lymph node and a RS of 8 receives neoadjuvant endocrine therapy and a CDK4/6 inhibitor on a clinical trial

DR LOVE: I want to finish out very briefly with a couple of neoadjuvant cases. We were talking about it before, and so rather than get into depth, I just want to cut to the bottom line. Eric, you have a 58-year-old postmenopausal woman who presents with a 5.2-cm lobular cancer, 1 palpable node, with an FNA-positive. She’s Grade 2, ER and PR strongly positive. And she would like breast preservation, which I’m guessing would require some tumor shrinkage. First of all, Joe, would you get an Oncotype in this situation? Because that’s what Eric did.

DR SPARANO: I usually do. I usually make the decision to proceed based on clinical grounds alone with neoadjuvant endocrine therapy and then order the Oncotype after surgery.

DR LOVE: On clinical decision alone, 5.2 centimeters, would like to have breast-conserving surgery, 1 positive node but needs tumor shrinkage. Joe, specifically, would you give her neoadjuvant therapy of some type?

DR SPARANO: Yes, I would feel comfortable offering her neoadjuvant endocrine therapy.

DR LOVE: Neoadjuvant endocrine therapy. So Eric, you elected to get a Recurrence Score, which gratifyingly came back as 8, maybe not too surprising. What did you do?

DR WINER: So we actually enrolled her in a trial of neoadjuvant endocrine therapy plus/minus a CDK4/6 inhibitor. And I think it’s fine to get the Oncotype afterwards. I mean, I think you could make a decision based on clinical grounds, particularly in a woman with a lobular cancer, where we know that preoperative chemotherapy almost never results in being able to do less than a mastectomy. But in my mind, whether you get it before or whether you get it after, it doesn’t really make a huge difference. Again, I didn’t need it to make this decision to do neoadjuvant endocrine therapy.

I will say that the Oncotype does influence my decision-making about what therapy she’ll receive afterwards. Now, the hard thing that’s going to come up is if after her neoadjuvant endocrine therapy she has her surgery and has 6 positive lymph nodes. Because as Joe mentioned earlier, many of us have started using Oncotype for quite some time in patients with 1 to 3 positive lymph nodes. We typically don’t do it in patients with a higher lymph node burden. On the other hand, you know that she has an Oncotype of 8, and so this is the situation I was describing, where if I give chemotherapy, I’m always saying to myself, “What am I doing?” But I think it’s the standard.

DR LOVE: Just to clarify, though, I don’t know if you know whether or not she’s getting a CDK inhibitor or if it’s blinded, but what’s happening right now? She’s getting the neoadjuvant therapy now?

DR WINER: She’s having a nice response to therapy. I don’t know how to use the results from neoadjuvant endocrine therapy, the pathologic results, in terms of making a decision about subsequent therapy. Even with a CDK4/6 inhibitor, you almost never see a path CR.

DR LOVE: It will be interesting to see what happens. You don’t know whether she’s getting CDK or not?

DR WINER: No, no, no. I actually do know that she was. It’s an open-label trial.

DR LOVE: She was getting a CDK inhibitor.

DR WINER: Yes, yes.

DR LOVE: Interesting.

Case (Dr Sparano): A woman in her early 60s with Stage IIA ER-positive, node-positive right breast cancer (RS 7) and Stage IA left breast cancer experiences a good response to adjuvant anastrozole alone

DR LOVE: Okay, why don’t we finish out — actually, we haven’t done any node-positive cases. Joe, let’s finish out, 62-year-old Hispanic woman, status post bilateral mastectomy for IIA, pT2 N0M0 right breast cancer, Stage IA left breast cancer, and she’s ended up on anastrozole because you had a Recurrence Score on the right one, which was 7. That’s good. ESR score 11.5 but high, but the left one didn’t have a Recurrence Score. Can you just talk a little bit about this case, Joe? I guess also she had 1 of 5 nodes that was positive on the right breast.

DR SPARANO: Right, so she was a clinical T2N0, and 1 positive node was unexpectedly found in the nonsentinel node, actually, because I believe the surgeon saw something or felt something and went for some additional nodes. And the Recurrence Score turned out to be very low. We didn’t do the Recurrence Score on the opposite side, because it was a small T1 lesion with a low-grade tumor. She received endocrine therapy alone and thus far, about 2 to 3 years in, she’s doing fine.

DR LOVE: Any final comments, Eric, on this case of this patient who has this positive node that’s found and yet has a Recurrence Score of 7? Agree with not using adjuvant chemo, Eric?

DR WINER: I totally agree, and I actually would just like to point out that in patients who have small tumors that are low grade, there probably isn’t a huge rationale for getting a Recurrence Score, just as Joe mentioned a few minutes ago.

DR LOVE: Small — this one was 1.3 millimeters. Eric, what range are you talking about?

DR WINER: I would say that in a postmenopausal woman with a tumor less than a centimeter that’s low grade and ER-positive or maybe even less than 2 centimeters if it’s with your pathologist and who you trust, and they tell you it’s a low-grade tumor and it’s strongly ER- and PR-positive, I’ll tell you, that tumor is not going to have a high Recurrence Score.

DR LOVE: Hmm. Joe, agree or disagree?

DR SPARANO: Yes, I think there are tools actually that have been developed to try and predict who will have a high Recurrence Score. I think that’s one of the critical questions in who to test. Clearly in TAILORx, not everyone was tested. There was selection bias that was introduced both by the selection criteria of the trial and the comfort level of the treating clinicians. That’s a critical issue, who to test.

Prognostic effect of the 21-gene assay RS for patients with high-risk ER-positive EBC and 1 to 3 positive nodes

DR SPARANO: The second issue is data in patients with positive lymph nodes. And the only prospective data that we have is from a nonrandomized trial with the 21-gene Recurrence Score, and that’s the PlanB trial. And this included patients who had high-risk, node-negative, generally tumors that were larger than 2 centimeters or had intermediate or high grade or had other high-risk features or who had 1 to 3 positive nodes. It was a small trial, somewhere in the range of about 400 patients. The vast majority of the patients who had positive nodes had just 1 positive node. There were very few patients who had 2 to 3 positive nodes.

They were treated with endocrine therapy alone, and they defined a low Recurrence Score as 11 or less, not 10 or less. But in any event, they found about a 94% to 95% 5-year disease-free survival with endocrine therapy alone. That’s the only prospective data that we have — truly prospective data.

There are some prospective/retrospective data from cohorts, and there’s also some real-world data from the SEER database, where the SEER database has been linked with the data from the 21-gene Recurrence Score, showing that a low Recurrence Score is associated with low 5-year breast cancer-specific mortality — not distant recurrence because of limitations of the SEER database, but breast cancer mortality.

DR WINER: And that’s actually without chemotherapy in those patients, too.

DR SPARANO: Correct.