During a case-based roundtable event, Kevin Kalinsky, MD, MS, discussed clinical typing of patients with breast cancer and the results of trials of therapies for patients with HR-positive, HER2-negative breast cancer.
KHAN: These cases have become a challenge in clinical practice. I just saw a patient last week who had negative lymph nodes and her tumor was 1.2 cm, grade 2. I don’t know about Ki-67, but her Oncotype DX [Breast Recurrence Score result] was 51. I think we are moving toward genotyping [rather than clinical features] and to me that would be a much more high-risk patient than your patient with 2 positive lymph nodes.
KALINSKY: When you’re individualizing risk, there’s the consideration of standard clinical and pathologic features, but there’s something about such a high Oncotype DX result that would make it [worth considering].
KHAN: It has become a dilemma, when you have medicines like abemaciclib [Verzenio], to decide whether to give that drug to a patient whose [lymph nodes are] negative but [who] has an Oncotype DX score of 51.
VERMA: I agree with what Dr Khan is saying, but sticking with the patient in this case, she has a high risk of recurrence because of the lymph node–positive status and the Ki-67 being high.
KALINSKY: At your institutions, is Ki-67 something that you generally had been checking or have started checking with the approval of abemaciclib, or are you not yet checking Ki-67 in breast tumors?
JONAS: For a while, our pathologists were not allowed to mandatorily check Ki-67. It was part of their guidelines that we had to specifically request it, otherwise it was not getting reimbursed. So we’ve had to go back to them with that. I guess that was some national guideline for them as pathologists.
KALINSKY: Interesting. What about at other institutions? Is that something that’s standard of care? If not, do you have to go back and request nonspecific specimens?
KUMAR: We are checking Ki-67 in all the patients and TP53 as well. This patient has high Ki-67, grade 3 disease, positive lymph nodes, and T2 lesions, which are all features that increase recurrence risk.
JONAS: For early recurrence, you always think of high-risk patients, so within 2 to 3 years, just like other diseases. Late recurrences for me used to be 5 to 10 years, but now people are doing 10 years of endocrine therapy, so it would seem to be after that period. I’m not letting them go all the way to 10 years. I stop it at around 7 to 8 years and they are recurring at 10 years. I don’t think if they stayed on it for 2 more years it would have made a difference, but those are the late recurrences I’m thinking of.
KALINSKY: I think what Dr Jonas is touching on is the biology of this subtype, where you can see recurrences 15 to 20 years later. There are data in the New England Journal of Medicine where they looked at the early breast cancer trialist group and they see these late recurrences even for those patients who have node-negative disease; their risk is higher with the nodes that are positive, but the recurrence increases as years go on.1
KUMAR: Do we need [to use] Oncotype DX for these patients? We already know they are at the highest risk for recurrence and most of them are going to get chemotherapy anyway.
KALINSKY: I’m going to pose that to the group. So, our patient is 52 years old and postmenopausal [and] has a high Ki-67 and 2 lymph node metastases—would you reflexively check Oncotype? Are there any circumstances when someone wouldn’t check an Oncotype for this postmenopausal patient?
KHANWANI: In a postmenopausal patient, I would check, but in a premenopausal one, not necessarily.
KALINSKY: Is there anybody who, in this postmenopausal patient, wouldn’t check it? Would you generally check it for postmenopausal patients?
DAILY: I would check it in this patient. I think the only patient that’s postmenopausal and [in whom] I might not check is a patient with 3 [positive lymph] nodes. I would discuss with the patient [the fact] that the data are more limited and it would be reasonable to do chemotherapy, but it could go either way. But I think for 1 to 2 nodes like in this patient, I would get an Oncotype DX [test].
KALINSKY: I think that from the data we have—the grade and the Ki-67—there’s already something suggesting a concerning biology with it. The pretest probability is this patient’s going to have a high Oncotype DX score. But I do think it’s worth checking just based upon [findings from] the RxPONDER study [NCT01272037]. I also hear the point about if a patient has 3 positive lymph nodes. Whenever we’ve interpreted data, ultimately there were less than 10% of patients who had 3 nodes involved. Most patients had 1 node involved. It’s always important to consider that factor.
KHAN: This patient probably just had a sentinel node [biopsy] done, so how sure are you that this patient doesn’t have 3 to 4 positive lymph nodes?
KALINSKY: In the RxPONDER trial, about one-third of patients had sentinel lymph node biopsy alone. It does come up in conversation and based upon RxPONDER, the general take-home was that those patients who have N1 disease and are postmenopausal don’t benefit from the addition of chemotherapy, but it’s always in the details. It’s not always such a broad stroke, and different features may sway you one way or the other.
If this patient’s premenopausal with 2 nodes involved, would anybody check for Oncotype?
KHANWANI: I wouldn’t.
KHAN: I would only if the patient wants to know how much benefit she’s going to get from therapy. If she has a small benefit then I might not, because if she is premenopausal, positive, and the Oncotype is low, the benefit might not be as high as if she has a high Oncotype score.
KALINSKY: I agree with that statement. I think we know that, prognostically, recurrence score can be informative. The RxPONDER data were published online on December 1, 2021.2 In the continuum for recurrence score from 0 to 25, the absolute benefit was higher if the recurrence score was 22 compared with 7. Sometimes, prognostically, it can just help inform the absolute benefit, but this is also an individualized conversation with the patient.
KHAN: The absolute benefit is quite small if your Oncotype DX score is, let’s say, 11.
KALINSKY: That’s right. The benefit was linear. Depending upon whether it is age or postmenopausal status. If the age was less than 50 it was about 7% and then got wider from there.
JIANG: I have a case I want to share with you. Several years ago, I had a postmenopausal patient and the Oncotype score was 8 or 9. She had a lumpectomy, radiation, and endocrine therapy, but within 1 year she had widespread lung metastases from the ER-positive breast cancer. I was just wondering if most Oncotype scores are accurate predictors of what’s going on. From this example, I’m not sure if the Oncotype score was wrong or sometimes we see the rare situation where Oncotype score is low but the cancer is aggressive.
KALINSKY: Does anybody have a comment about that? I’m happy to chime in, but again, this was a patient with early-stage breast cancer, low Oncotype, who had an early recurrence and…was [adherent] with taking her endocrine therapy.
JONAS: With breast cancer, there’s no 100% with what we do. One can never tell somebody you’re curing them with guarantees. I mean, bad things happen with cancer. It does what it wants to do. You’re just playing the odds an awful lot of the time and it’s not clear that even giving her chemotherapy would have made a difference there. It’s just the nature of the disease unfortunately. An old partner of mine used to say it’s just the biology of the disease. You can’t control that.
KALINSKY: Right. Nobody’s risk is 0%.
KALINSKY: I’m going to presume that everybody would consider all these factors when making this decision. Is there anything else that anybody would add?
VERMA: I don’t know if the tumor size is included in the clinical risk.
KALINSKY: I know we’re talking about node-positive disease, but in node-negative disease there is a calculator tool on the genomic health website where you can put in size, grade, age, and recurrence score and it gives you an individualized risk for recurrence, and not just the recurrence score by itself, but there are other features. I don’t know if anybody else commonly uses that calculator tool, but I find it helpful for node-negative patients to help individualize the risk.
KHAN: I think patient characteristics [are] like…the comorbidities; if somebody has a problem, you want to avoid enthesitis and things like that when selecting adjuvant therapy for a patient.
VERMA: I’ve used abemaciclib for 1 patient. She’s been on it for a couple of weeks now and she told me when she got the medication from the pharmacy, it came with antidiarrheal medication that was mailed to her. So they are shipping the medicine with the antidiarrheal medications. But she’s been tolerating it well so far.
DAILY: I’ve had experiences with abemaciclib. I haven’t had patients who’ve had to come off it. Some have required dose reductions, but I haven’t had anyone discontinue entirely.
KALINSKY: Great. What about those who have not used it yet? I recognize that the approval only came relatively recently.3 Have you not had patients who’ve been reflective of the monarchE study [NCT03155997] population or have you had concerns about the data or have you not had a patient who would be a good candidate?
KHANWANI: [I have not had a] patient yet, but I don’t think I have any doubt on the data….
XIA: I ordered one, but insurance denied it.
KALINSKY: Did you have the Ki-67 score and did they fit the criteria?
XIA: They fit it. It was a young patient, premenopausal, and very high risk. [Insurance] still denied it.
KALINSKY: Interesting. Have others had rejections from insurance companies?
BHIMANI: I have a couple of questions about this indication as well. The NCCN [National Comprehensive Cancer Network] is not very straightforward on strongly recommending CDK4 and CDK6 inhibitor therapy for patients who are high risk. Unless I read it wrong, they’re still recommending endocrine treatment for 8 or 10 years.
In the adjuvant setting, [deciding on whether to give] 1 vs 2 years of treatment, and diarrhea can be an issue. A 16% discontinuation rate is high and there is no overall survival [OS] benefit. That’s bothersome mostly in the adjuvant setting when you are truly trying to cure the cases, rather than just trying to delay recurrence. But I have not seen a patient [who fits the criteria for abemaciclib] yet.
KALINSKY: The guidelines are constantly evolving and ASCO [American Society of Clinical Oncology] has released guidelines that are less narrow and not based just on Ki-67. I would think about what the study was powered for. I think that to achieve an OS benefit with this subtype of breast cancer requires some maturity because we’ll see some late recurrences. Also, is this sort of study powered to answer that question? The OS was a secondary end point.
KHAN: Why wasn’t it powered for OS?
KALINSKY: I think it’s because it would have required many more patients and more time. I think you raise a good point about the duration, 1 year vs 2 years. The NATALEE study [NCT03701334], which will hopefully read out soon, was 3 years….The drug itself may just cause some senescence. The question is what the optimal duration of these agents is. I don’t think anybody knows, just given that the study design is different across the various trials.
KUMAR: Any difference if you add leuprorelin [Lupron] to it or any ovarian ablation therapy in premenopausal or postmenopausal patients?
KALINSKY: This is a higher-risk patient population. These are patients who would have been candidates for the SOFT [NCT00066690] and TEXT [NCT00066703] trials, for instance. I’m curious what other people do in the premenopausal population. These are patients [for whom] I generally think about using ovarian function suppression and an AI. I also have this concern about tamoxifen [Nolvadex] and abemaciclib.
KHAN: Another focus in the study was the Ki-67. How valid is it from a laboratory in a community setting? I have seen a lot of Ki-67 scores that are sometimes 98%, and the grade is grade 1, and I always wonder what Ki-67 98% means in a grade 1 tumor. Should it not be standardized? [Also], to me this looks like going backwards, rather than going forwards. We’ve moved away from clinical features, and we’ve gone on to the biology of the disease with Oncotype DX. I pointed out those 4 positive lymph nodes because currently, in my clinic, I don’t see many patients with so many positive lymph nodes. Usually they are at sentinel 1, 2, or 3. We are focusing more on Oncotype DX results. How reliable is Ki-67? Can you use Oncotype DX to make the decision?
KALINSKY: I think when you have such a big disparity between grade and Ki-67 [score], it’s always worth repeating the pathology review. It is quite strange to be so proliferative and then have a grade 1 or 2 tumor. There’s a real disconnect there. They are now going back and doing additional evaluations with the tumors that they collected, where they’re going to look at RNA-based signatures, and maybe we’ll get some information just about scoring with different signatures, including Oncotype DX.
For 3 decades there’s been an issue with Ki-67 and all the preanalytic and analytic considerations that haven’t been perfected. In the monarchE study, it was based upon central testing, but in our clinics some are using Ki-67, others aren’t; some use 1 assay, some use a different one. It’s a little all over the place and it can sometimes be challenging.
KHAN: That’s exactly my point. I’m dealing in a clinic with these things in the back of my mind and then I have a very discrete, reliable Oncotype DX score. So should I even look at those things? Most of the time, you are making decisions based on Oncotype DX results, rather than these clinical features. That’s where the science is moving, even for patients with 1 to 3 positive lymph nodes because of the RxPONDER study. I think that’s a challenge for me.
KALINSKY: I hear you.
BHIMANI: The FDA approval is only based on the Ki-67. If we have 4 positive nodes and a 5-cm tumor or 1 to 3 positive lymph nodes with the Ki-67 of less than 20, then that wouldn’t be approved based on the FDA indication. Is that right?
KALINSKY: I think you could use the ASCO [American Society of Clinical Oncology] guidelines, which are a little bit more open, if you were facing issues with insurance. It ultimately comes down to the patient and their individualized risk. If I have a patient who has a Ki-67 of 19% but has 6 lymph nodes involved and is 40 years of age, then they may be a patient that I would have that discussion with.
BHIMANI: The primary end point is a little bit touchy. We have data in lung cancer with the same disease-free survival in the adjuvant setting with immunotherapy, as well as with a tyrosine kinase inhibitor. We are wrestling with this question in early lung cancer. The original idea of adjuvant treatment is to improve the survival and we are curing more patients than we thought. Two years seems to be longer. Maybe in the demographics where I’m practicing, a lot of patients do not want to get anything. With COVID-19, patients have developed some sort of mistrust in the health care. To keep on adding these medications and having these significant adverse effects [AEs] is becoming challenging.
I have a very young patient. She’s a banker and does not want to even take ovarian suppression because she does not want to come every 28 days to get the goserelin [Zoladex] treatment. I’m just making a point here. Should it be a standard of care? That’s subjective. As you mentioned with the patient discussion and how the physicians perceive these data, I can’t speak on it. But if the OS is not there and it was not powered, we may not get an answer. Even if there is no OS benefit, then you can always go back and say this was not powered for it. It’s difficult to make it a new standard of care.
KUMAR: I don’t know what, in total, the data may mean. Is it too early to make [abemaciclib] a standard of care?
KHAN: I have not seen any of the recent breast cancer drugs showing OS. When we talk about HER2, [and] immunotherapy in triple-negative breast cancer, it looks like the FDA is changing their parameters and instead of OS, they’re looking at disease-free survival and things like that. It looks like all the recent breast cancer drugs that have been approved in the adjuvant setting; none of them, I think, have a clear-cut OS.
KHANWANI: I would say all those points are valid including what Dr Bhimani mentioned. But even though there’s no OS, a lot of these patients with recurrence are miserable. Even if they may not live longer, if the disease comes back, for a lot of people it’s unacceptable. I see patients of a young age who have the disease back. They’re miserable hearing that news.
With these data, I still won’t be comfortable not offering [abemaciclib] in the adjuvant setting in someone I see with a Ki-67 of more than 2 or 3 positive lymph nodes. I will offer it. Now of course, patients, after listening to the AEs, cost, and everything, may refuse. But as a physician, I think my role is to still offer it as a standard of care.
JONAS: I keep it simple. We give chemotherapy for less benefit than this seems to be offering and [abemaciclib is] a whole lot easier than chemotherapy. I say yes [to it being standard of care].
KALINSKY: It’s interesting to get everybody’s viewpoint on this. There’s definitely, especially for these high-risk patients, an absolute benefit that’s seen early on. I think one of the things that we’ll see over time is when patients stop taking the medicine and what the curves will continue to look like. Are they going to continue to separate? Are they going to widen? Are they going to get closer together? I think we’ll see.
1. Pan H, Gray R, Braybrooke J, et al; EBCTCG. 20-year risks of breast-cancer recurrence after stopping endocrine therapy at 5 years. N Engl J Med. 2017;377(19):1836- 1846. doi:10.1056/NEJMoa1701830
2. Kalinsky K, Barlow WE, Gralow JR, et al. 21-gene assay to inform chemotherapy benefit in node-positive breast cancer. N Engl J Med. 2021;385(25):2336-2347. doi:10.1056/NEJMoa2108873
3. FDA approves abemaciclib with endocrine therapy for early breast cancer. FDA. Updated October 13, 2021. Accessed March 10, 2022. https://bit.ly/3J3ch0S