Biology of triple-negative breast cancer (TNBC); implications for therapeutic decision-making

DR LOVE: I just want to begin with an academic-like question. We’re focusing on triple-negative disease, and I guess my question to you today is, at this point, with a lot more translational biologic knowledge about this disease, how you view triple-negative breast cancer? Do you even view it as the third disease? And is it just the absence of ER and HER2, or is there something in your mind that characterizes the biology of triple-negative breast cancer?

DR RUGO: I think what we’ve realized, it’s kind of funny. I always used to think of triple-negative breast cancer as non-A, non-B hepatitis from training, where you really characterize something by the absence of things that we use for treatment. And having been practicing really since the discovery of HER2 and then development of targeting HER2, it’s been an interesting transition to understand as we had treatment for HER2 that there was a separate group where people had very aggressive disease and often fared very poorly.

But I think since that time, which now is 2 decades old, we have really in clinical practice, what we recognize is that this is a very heterogeneous disease, and in some ways we haven’t exactly figured out how to change that heterogeneity for the entire group of patients. How to address that heterogeneity, more aptly put.

You have patients, the majority of patients who have very proliferative high-grade triple-negative disease. Some have a little percent of ER, but they still act very much like triple-negative disease, maybe 2% or less than 10% — hard to know where that transition is. And then some people have the ER of 0, very aggressive, rapidly growing disease, and then still others, a smaller subset, have this kind of indolent triple-negative disease that almost looks like it’s hormone receptor-positive, but it has no receptors.

When you see a patient who has triple-negative breast cancer, first you have to categorize in your mind, where does this cancer fit? Obviously our approach to early-stage in metastatic disease is going to be different as well, because in early-stage disease, if you fall into the majority group where you have a more rapidly growing hormone receptor-negative disease, HER2-negative, falls into the triple-negative category or close enough, the next differential for us really is, how do you respond to chemotherapy? How does the cancer respond to chemotherapy? Because if you see a rapid shrinkage of the cancer, that falls into a category of very chemotherapy-responsive disease. And then some of the patients have sort of a sluggish response, and then the cancer starts growing back again. That’s our worst category of triple-negative breast cancer.

In some ways you’re using the clinical differentiation and then response to therapy in the early-stage neoadjuvant setting to try and differentiate out different subgroups biologically. Now, how that translates into our expression groups, the gene expression groupings that have been created with intrinsic subtyping or with further differentiation in the Pietenpol classification of triple-negative disease.

It’s not entirely clear, because we don’t use those expression subgroups, gene expression subgroups, in clinical practice. I’m guessing that there’s some overlap, that it’s not all — for example, if you look at a basal-like triple-negative breast cancer, some of those cancers do very well, and then there’s a subgroup that don’t do very well, and we don’t really understand why some of those groups are genetically plastic and mutate as you’re treating them and become more and more resistant. And that’s a huge challenge for us in the treatment of triple-negative disease.

Median overall survival estimates for patients with metastatic TNBC

DR LOVE: Let me ask you kind of another related question, which is — because you refer several times to the idea of triple-negative disease being aggressive — so let me ask you for these numbers.

I’m just kind of curious what you say if one of your fellows said, “In general, if you break it down again into these 3 subsets: ER-positive, HER2-positive and triple-negative, and you look with our current therapies and you look at survival from diagnosis of metastatic disease” — and of course yes, there’s all the variables — but just in general, how much difference is there? What’s the number you would give? What’s the global median survival of a patient with metastatic triple-negative, HER2-positive, ER-positive with our current therapies?

DR RUGO: I think that we have recent data that gives us an idea about the median survival for triple-negative disease in the metastatic first-line setting from IMpassion130. In that trial, we enrolled patients who were at least a year from their adjuvant or neoadjuvant taxane chemotherapy. And that actually differentiates a group of patients who are going to have relatively better outcome than the patients who are just progressing and developing metastatic disease within weeks or months from their early-stage diagnosis.

We’re already looking at a group that’s a little bit better, and the median survival was about 18 months, regardless of whether you had PD-L1 expression or not in the tumor immune cells. It does go along with what I think we see in clinical practice. Sometimes we see a more chemotherapy-responsive disease and the median survival’s a little bit longer. Sometimes we see very resistant disease with rapid progression to brain or other critical organs with organ dysfunction, and those patients live appreciably less than 18 months.

Survival outcomes for patients with metastatic ER-positive and HER2-positive breast cancer

DR RUGO: In contrast, in hormone receptor-positive disease the median survival for most patients, again, depending on the clinical factors like disease-free interval and how a patient presents, the median survival is more like 5 years. I say that keeping in mind the fact that I have patients who’ve I been treating for much longer than 5 years and patients who die within 2 years of diagnosis of metastatic disease. Medians are really just that.

If you present with diffuse visceral disease with hormone receptor-positive breast cancer, even if you’re 8 or 10 years out, your prognosis is going to be much worse than a patient who presents with bone-only disease. And, of course, if you present within a year of your initial diagnosis, you’re going to have more hormone-resistant disease.

For HER2-positive breast cancer, it’s interesting. If you have de novo HER2-positive metastatic breast cancer, a subset of those patients will be long-term survivors. We saw that in the recent update of the CLEOPATRA trial, where there’s now 8-year follow-up of patients that was presented at ASCO. And it was interesting. If you look at the patients who are progression free, there’s a subset of patients with 8 years have never progressed after first-line therapy with a taxane, pertuzumab and trastuzumab.

And so I think that we know there’s that subgroup. They have HER2 enriched disease. We don’t totally understand why there is that particular subgroup and it’s not everyone. If you had great HER2-targeted therapy in the neoadjuvant and adjuvant setting and you develop metastatic breast cancer, your median survival is going to be shorter, even with our novel agents. Those patients are going to have a median survival in the 2- to 3-year range. You still have more treatment options in triple-negative breast cancer even today. The median survival is going to be a little bit longer, but it’s shorter in general than patients with hormone receptor-positive disease. There are variations in HER2-positive breast cancer based on ER.

DR LOVE: Let me just pin you down a little bit. Because that number that you just gave I think is pretty interesting: 18 months, 5 years. Big difference. Roughly, globally, all comers, what are you looking at in terms of median survival with metastatic HER2-positive disease?

DR RUGO: Again, I think there’s a huge range for HER2-positive breast cancer, and it’s so dependent on where you come from: de novo, prior treatment, disease-free interval. The median survivals can be as short as 18 months and as long as 10-plus years, depending on the subset of patients you come from. Probably the median for HER2-positive disease is in the 3- to 4-year range if you take that very resistant HER2-positive disease and mix it with the exquisitely sensitive patients.

Case: A woman in her early 30s with localized TNBC receives neoadjuvant chemotherapy on the control arm of the I-SPY 2 trial

DR LOVE: Certainly the experience as an oncologist taking care of patients with breast cancer is very different for a woman with metastatic triple-negative and for localized triple-negative. And I want to kind of spend most of our time today getting into kind of how it works in your own practice and particularly how you integrate the individual preferences and issues with each patient that you see in crafting a decision with them.

We asked you to come up with a patient with localized disease so we could get into the issue of triple-negative disease. And, of course, as we guessed, you pulled somebody who was involved with I-SPY, because almost all your patients were involved with I-SPY. It’s great, because I want to see what’s going on with I-SPY. But anyhow, what happened with this 33-year-old physician?

DR RUGO: This is actually a physician who’s in postresidency training who presented, actually, kind of overnight with a big mass in her left breast. And she’s lived a busy life, working many long hours and with long call. And so she noticed a sort of swelling in her breast, and that rapidly became quite large, painful and distorted the appearance of her breast with some nipple inversion.

And so she came in for evaluation and was found on imaging to have a big mass. It’s interesting. These very young women, often the tumors are not well seen on any kind of mammography, standard mammography. And having just seen a woman who was in some other local places where she was referred and people kept saying, “Oh, you’re too young. It’s a fibroadenoma.” And I think that’s a really dangerous thing, and it doesn’t hit us as oncologists. But we see those patients, and it’s just so important for that education to hit people in practice, that even though it’s uncommon, we do see breast cancers in young women.

She came in, had this ultrasound that showed a tumor that was a little under 2 centimeters, although on exam the tumor was much larger. And on exam we could also find an axillary node. A core biopsy showed a triple-negative breast cancer with a Ki-67 of 70%. And although we could feel the node, we actually did fine-needle aspiration several times of her axillary nodes, and they were benign. That was kind of encouraging.

We did an MRI then, which we do in young women with these dense breasts where you really can’t see things well on the mammogram, actually, which we see a lot in these triple-negative cancers — central hypodensity, which was consistent with central necrosis of this large mass. And overall, the total size of the mass was just a little under 4 centimeters, at 3.8 centimeters, another thing we see in these young women where the ultrasound really often undersizes the tumor. The tumor’s actually quite a bit larger on MRI. Although again, MRI, it’s very sensitive and not as specific. You have to be cautious in what you recommend for surgery in those women.

DR LOVE: Could I just ask, I’m just kind of curious what the MRI showed of the nodes that were negative on FNA.

DR RUGO: Right. She has these prominent axillary nodes — one of the problems we always get into as oncologists is, we see people after they’ve been poked a lot. We have to be a little cautious. And we have to be cautious in our exams, too. You don’t want to recommend a mastectomy for a young woman without germline mutations based on something that’s not really the extent of her tumor, particularly if you could shrink it with initial therapy, because this is a life-long issue for these women who may be cured of their disease.

She actually had an MRI that was done after she’d had a core biopsy of her breast and fine-needle aspirations of her axillary nodes twice, so it did show some prominence of the axillary nodes. I’ve seen this in some of the young women where we get discordance. We find nodes that feel like they should have cancer. They look a little funny on MRI and ultrasound, but we do the needle aspirations and they don’t have cancer in them. Obviously we sample at the time of surgery.

And for a big triple-negative cancer, having a positive node or not is not going to alter our treatment preoperatively. And I think the key thing is that most of these women do not need extensive axillary node sampling. If you do a sentinel node and any enlarged nodes right around it and those are negative, you can feel pretty confident. These women, also, they present with big cancers. They’re young. They’re all going to need radiation. We don’t need to do big axillary dissections in these women even though we’re not 100% sure whether or not originally their node had cancer in it.

Genetic testing for patients with TNBC

DR LOVE: This is a shocking situation for anybody, particularly and including somebody in the medical profession. But I see I guess you all took a little bit of a breath and thought about the issue of fertility.

DR RUGO: Yes, we thought about 2 big issues, and I think that’s really important. When you diagnose these cancers, you want to take 1 step back and say, what are the 2 key issues? One is that you want to start thinking about germline genomic testing early. This is a patient who clearly fits into all the guidelines for germline genomic testing. Although she didn’t have a family history, we know that a certain percentage of these young women, not a small percentage, might have a germline mutation. That’s something you have to talk to the patient about and send off early, because we know it’s going to impact her decisions about surgery in terms of bilateral mastectomy or a lumpectomy, unilateral.

The second thing you have to think about right away is fertility issues, because the chemotherapy that we give in the neoadjuvant setting may cause ovarian suppression that’s more a long duration. We know it causes temporary ovarian suppression. What we don’t know is what chemotherapy does for ovarian reserve. Most young women, depending on the chemo you give, will recover their ovarian function. But the problem is that we don’t know what it does for their ovarian reserve.

It’s a huge issue for ER-positive cancer — we have to give hormone therapy afterwards — but for triple-negative disease in a 33-year-old woman, we already know that your ovarian reserve is starting to decline, sadly, so you want to not close any of those doors. This woman is single. She definitely had not closed the door on having children someday. And you always have the time to harvest eggs.

I would say that it takes 2 weeks to harvest eggs. And now there’s huge success with using frozen mature oocytes. It’s really, really important to give this option to our patients.

Fertility preservation in young women undergoing therapy for breast cancer

DR LOVE: Just kind of curious about the logistics of how you make this happen. You’ve got this woman who’s got this big triple-negative tumor and these kind of concerning nodes. I’m sure you want to get therapy started right away. You say 2 weeks. It sounds like the day they come to clinic you get them started?

DR RUGO: The day they come to clinic we call the — we actually have an inside line — and you just have to have that set up. People who are working in the fertility field are very much oriented towards this approach. They understand the timing. And you just do not want to ever close the door on a woman for fertility issues.

I’ve seen that happen where somebody said, “You have to start treatment. We don’t have time to harvest eggs.” And then the door is closed forever. And it’s really tragic, because it’s not the case. We’ve never, ever shown that 2 weeks or even 2 and a half weeks or 3 weeks makes a difference in the timing to start therapy.

What we do is, we send a message to our fertility group. They get the patient in right away, and they’re able to start medications right away. We actually have also the communications to foundations like the Livestrong Foundation that will provide some additional financial support to help women who don’t have the money or their insurance doesn’t cover the medications you need. And if you start it right away, you can actually harvest the eggs within 2 weeks, and you can start treatment the same day you harvest eggs, if you really feel pushed. I usually try and wait a couple of days, in case they get a little bit of bloating and ascites. If they’re very young, they can get a little bit of that with the harvest.

DR LOVE: I’m thinking about the word quality. We once did a daylong symposium on just quality, and I really wonder how many women in the United States right now are getting this procedure done in the first 2 weeks, that kind of setup, which seems like it really ought to be standard.

Perspective on the use of GnRH agonists for premenopausal women with TNBC

DR LOVE: The other thing about fertility that we’ve heard about, and I’m curious whether you do this strategy, is using LHRH agonist, so to suppress the ovaries, obviously not related to the tumor. That is a strategy. It’s been described. What do you think about it? Do you use it? Did you offer it to this lady?

DR RUGO: Yes, I do. I mean, it’s so funny, because I really have 4 things that I make sure I’ve covered. The germline testing, which often has been done before I see a patient. Fertility, in terms of egg harvesting, the use of GnRH agonists and then scalp cooling for hair loss.

The GnRH agonists, it’s an interesting area. I think it makes theoretic sense to us that if you shut down the activity of the ovaries that you might reduce the amount of toxicity from chemotherapy, in some ways similar to what we with scalp cooling, where you actually use something that may reduce metabolism. I know that some fertility experts don’t believe that, but others very strongly believe in it. We have the POEMS study. But the problem with all these studies is, no matter what you do — and that was really a herculean effort — is that you always have smaller numbers than you intended, which was true of POEMS. And then people relapse with cancer. They decide not to get pregnant. They have other issues with their partner. You can’t just, like, randomize to having babies or not.

There is some data that suggests it doesn’t do anything and other data that suggests that if you give the GnRH agonist, you can help preserve ovarian function. The downside of using these drugs — they’re general approved by insurance. In the US, it’s not a financial issue for our patients. The downside is that you do have the immediate menopause to deal with along with the psychological trauma of breast cancer, starting chemotherapy, whereas the ovarian suppression that occurs from chemotherapy is slower, and it’s a little bit easier to adjust to.

When I tell women that this is an option for treatment — and I have to say, everybody who’s young says they want to use GnRH agonists, with very few exceptions — I do tell them that the side effects, the menopause side effects, are going to be a little stronger than if they didn’t use the GnRH agonist. But I do use them routinely. That’s another thing we put in for authorization.

Importance of providing support to patients coping with a diagnosis of TNBC

DR LOVE: It’s amazing, when you think about it, what happens, what transpires between the oncology team and a young patient like this with triple-negative disease trying to cope with the cancer and at the same time having the courage to say, “I’m thinking maybe I’ll be okay. Let’s see if I can do something about fertility.” What an amazing situation to see people go through. How did she do with all this?

DR RUGO: And I have to say that at the same time, for a young woman who’s working, they’re worried about work. Are they going to keep their job? Are they going to be relevant? Somebody’s going to get rid of them because they’re not there? If they know they have cancer, do they think they won’t be there in 2 years, so they’re not going to — people worry about that. They also worry about the appearance of, again, not having hair. That’s one issue.

And then another issue which comes up with the young women, because of reading on the Internet, is, they worry about maybe they’re going to have so-called chemo brain and they won’t be able to think as well or do their job as well. And I really reassured women about that, that a lot of it is the premedications and rescue medications for nausea. And with our current chemotherapy treatments for young women, we really don’t see appreciable chemotherapy brain, so to speak, that’s long lasting.

I think anytime you’re not sleeping well, you’re taking lorazepam and feeling nauseated or coming in all the time and worrying, not sleeping well because of hot flashes, you’re going to have some word-finding difficulties and things like that. But this isn’t a long-lasting issue.

The other thing, just to mention about adjustment, is that if you have access to psychology or some kind of supportive mechanism, it’s really helpful. I find support groups aren’t so useful if they’re all cancer, because you have people in all stages and all ages. But there are some blogs online of young women’s groups with breast cancer, and we have a local group here that’s been incredibly supportive and helpful for women. And I think it’s really worthwhile to refer these young women for some of those supportive groups, because it can be tremendously helpful at a time when psychologically they’re feeling so vulnerable.

I-SPY 2 trial: Neoadjuvant and personalized adaptive novel agents to treat breast cancer

DR LOVE: How did this woman do? I guess she ended up on the I-SPY trial. What was the randomization that she encountered?

DR RUGO: Our I-SPY program we talk to women about neoadjuvant therapy and in this case what I feel is the really incredible importance of giving neoadjuvant therapy in order to try and understand the response to therapy, and then I modify treatment to try and improve outcome based on response to therapy, either in the postneoadjuvant setting, where we have CREATE-X with capecitabine, or even before surgery, where we do try and modify treatment based on response and are learning more and more about that.

We talked about I-SPY, which is an adaptively randomized Phase II trial where we compare a standard chemotherapy backbone to novel agents that are generally added to the paclitaxel component, which occurs first, but sometimes are extended through the anthracycline component as well and in certain situations have actually replaced the paclitaxel component of chemotherapy.

When you enroll onto I-SPY, you actually are screened. And, of course, triple-negative breast cancer is always going to be eligible for chemotherapy, but we screen patients with ER-positive disease as well as all comers, getting the 70-gene MammaPrint^® assay, and patients who are high risk go on I-SPY. All patients with triple-negative and HER2-positive disease are also eligible. But this allows us to get a number of different genomic markers as well. Patients are then randomized, after they’ve had their staging scans, their MRIs, the biopsy and their MammaPrint comes back, to one of the arms of the trial.

You actually are consented again, because you have an opportunity not to participate in the trial if you’re not happy with the arm that you’re randomized to, of course. And we go through the specific toxicity of the randomized arm. I don’t think for any patient that you can really take in 5 different arms of a trial and all the different toxicities. It’s much better, I think, to be able to randomize for screening and then randomize again once in you’re in the arm. This patient was randomized to our control arm, which is weekly paclitaxel for 12 weeks followed by doxorubicin and cyclophosphamide, generally given dose dense for 4 weeks, and then surgery.

In our I-SPY trial, just like any trial, the patient is the important person here, and so we actually see the patients weekly. And if we feel like the response is not very good, we get MRIs periodically as well, for the study. Then we can always change gears. For example, you could add carboplatin at some point if you felt that response was not great.

I like the idea of not giving carboplatin to everyone, although we may be evolving as a field to giving carboplatin to more patients with triple-negative disease. My main concerns are that it does increase toxicity and bone marrow suppression makes it a little harder and also that it is more toxic to the ovaries for patients who are young.

Novel agents under evaluation in the I-SPY 2 trial

DR LOVE: I’m just kind of curious, this lady I guess is actually an ongoing patient, and she got randomized to the control arm, but what were some of the novel agents that she might have been eligible to receive? What are you looking at in I-SPY in triple-negative right now?

DR RUGO: We’ve been looking at checkpoint inhibitors. We had a pembrolizumab arm that included pembrolizumab with paclitaxel for 4 doses, and that one we’ve, of course, presented and hopefully will be published in the not-too-distant future, that showed a marked improvement in PCR, not just in triple-negative disease but in the high-risk hormone receptor-positive cohort of patients enrolled on I-SPY. And, actually, that was the first hormone receptor-positive high-risk group to graduate in I-SPY with the addition of a novel agent. And I think that’s a really interesting area moving forward. But we extended it to include pembrolizumab through AC and pembrolizumab after paclitaxel. So that was one area.

We’ve also studied an antibody-drug conjugate called LIV1A. We’ve studied the combination of a checkpoint inhibitor durvalumab and a very low dose of the PARP inhibitor olaparib. And then we have a number of other areas that we’re also studying. One is to try and improve the response to checkpoint inhibitors.

We know from the metastatic setting that only 40% of patients had the tumor immune cells that express PD-L1 and seem to benefit from the addition of atezolizumab in IMpassion130. And the whole idea is to have more patients benefit.

One way to have them benefit is to treat them earlier in the course of disease when they have more tumor-infiltrating lymphocytes and more PD-L1 expression. But the other is to use some kind of immune agonist approach. And we have a TLR9 agonist called SD-101 that’s actually injected directly into the tumor to try and enhance or recruit immune cells from the host, and then that’s given with pembrolizumab and paclitaxel together before we go onto the AC. That’s another arm. And then we’re always, as arms close we have new arms that opening all the time.

Investigation of immune checkpoint inhibitor combination approaches

DR LOVE: Yes, I want to spend really most of the time talking about some of the newer agents that are coming out in triple-negative disease. But I have to ask you one thing. I’m curious whether you heard about it. We already have seen some pretty interesting data in renal, metastatic renal, checkpoint inhibitor with TKI, VEGF TKI, for example, axitinib, and there’s another one. And there was a poster at ASCO that I’m wondering if you were exposed to — it was just a poster. And it was in colon and gastric cancer that was MSS stable. And they saw a whole bunch of responses to nivolumab plus regorafenib.

DR RUGO: Oh, yes, I did see that.

DR LOVE: Do you know anything about that? And is it possible that maybe we are going to find the combinations? There were objective responses. Any thoughts about that?

DR RUGO: One of the problems is, when you use an active agent along with another and you have a small study, you don’t really know what the novel agent is adding in. But I have to say, we have some intriguing data with using a checkpoint inhibitor and adding in something that in preclinical settings may be an immune agonist, essentially. And maybe it will be a TKI. Maybe it’ll be another antibody. We don’t really know.

I mean, there’s data with bevacizumab and checkpoint inhibitors. There’s data with MEK inhibitors, immune agonists. And there was a fascinating study just published that we’ve seen presented a couple of times called the TONIC trial, where patients received, actually, a really short induction of standard treatments, a little cisplatin, a little oral cyclophosphamide, a dose of radiation, a couple of doses of doxorubicin, low dose. And then they went on. They didn’t continue that. That was just a 2-week induction. Then they went on and received just a single checkpoint inhibitor. And these were patients with triple-negative disease, a study done in the Netherlands.

And what they saw was fascinating — was that there were very good responses. It’s a tiny study, and they found the doxorubicin and cisplatin might be the best immune agonist. But they also did all these laboratory studies looking at gene expression, induction of CD8-positive T-cells, TILs, and they really saw very, very encouraging and exciting results. They now have a Phase II trial that’s larger, looking specifically at the immune agonists that they thought were most effective. But I think that this is what the future is right now, is looking at ways to try and enhance the effect of this checkpoint inhibition and immune agonist in patients who have breast cancer and in all subtypes.

Cooling caps to reduce chemotherapy-related hair loss

DR LOVE: Let’s just finish up with your patient, because I want to get into new agents, new directions and maybe the beginning of moving out the survival curve with triple-negative disease that you were talking about earlier. Let’s just finish out, though, with this lady who got the standard arm of paclitaxel followed by AC. How did she actually do? Did she do the cooling cap? Do most of your patients do the cooling cap?

DR RUGO: Many do, but I have to say patients who are receiving AC, we don’t. And also, remember that if you get weekly paclitaxel plus AC, that’s actually 16 different chemotherapy infusions, and it’s a lot of time to spend.

DR LOVE: Yes.

DR RUGO: It’s an extra hour and a half after paclitaxel and 2 hours after AC. There’s a new cap design for one of the automated cooling systems that actually individually can be shaped to the head. And I’m very hopeful that this is going to work even better with AC, because now the success rate is maybe 30%. With this new cap, maybe we’ll see better responses. She chose not to use the cooing cap, in part because of the fact that insurance doesn’t cover it, and secondly because she was interested in continuing to work and continue with her training program and not take the extra time with the scalp cooling.

She actually — it’s interesting. She started with paclitaxel. She’s really, really worried about neuropathy, because — felt that it would really affect her work. And after 2, 3 weeks, she actually had quite a bit of tingling in her fingertips and toes.

One of the things I did, which is nonstandard but allowed within I-SPY, and I do it quite frequently, is, I changed her to nab paclitaxel, because first of all, there is data from the German group that suggests that nab paclitaxel may have some advantage in triple-negative disease. It’s not definitive data, but it’s interesting, recently published. And secondly, nab paclitaxel at 100 mg per m² in the metastatic setting had much less neuropathy than we see with paclitaxel, and these were in patients in the BLUM study that had previous paclitaxel, so they were at higher risk for peripheral neuropathy.

I did change her to nab paclitaxel. And although this is just an individual patient, she did very well with it.

DR LOVE: Just curious. I’ve heard people anecdotally talk about having patients hold bags of frozen peas and stuff like that. Do you ever do that?

DR RUGO: I’m not big on the frozen peas. And Charles Loprinzi at Mayo is really interested in using these cold, like, bags of ice, that people would hold. I don’t think that the cooling, and he doesn’t either now, think that the cooling is efficient enough. You can actually buy gloves that you freeze. And it turns out that probably for this to work, you have to have a frozen glove that you change if it warms up, and you can wear frozen booties, and it has to be closely applied. It actually has to have some pressure. And we use that with scalp cooling. We actually put the cap on, and you want it to really closely be opposed to the scalp. Same with the hands.

Use of frozen gloves and socks to lessen chemotherapy-induced peripheral neuropathy

DR RUGO: There’s actually a group in Japan that showed that it might indeed reduce neuropathy. They put on one hand the cold glove and bootie, and then the other hand they used as the control. Patients were their own control. They did see really, I thought, impressive results. And in Singapore, they’ve worked on these gloves that have some pressure applied to them. Actually, I’m very interested in this as a potential — something to study as we get better and better cooling systems and maybe have one of these automated units where you have a glove.

But now, actually, an increasing number of my patients, because of online discussions and blogs, et cetera, will purchase these cold gloves and freeze them and bring them in and use them to try and reduce the risk of neuropathy. I mean, the problem with neuropathy is, we know a group of patients based on their pharmacogenomics are at higher risk, but we don’t know how to identify them. And so it’s hard to know who to advise.

You hate to see those patients who 2 years out still have enough neuropathy they can’t wear their shoes. These patients are worried about it. She didn’t use the frozen gloves, but we did switch to nab paclitaxel.

Coping with the side effects of chemotherapy

DR RUGO: What was interesting about her course on the chemotherapy, which I found in a lot of professional women, is that — and not to say that it doesn’t happen to everybody — but I think people think of themselves in general as being invincible and that there’s a lot of mind over matter. And for these young women, of course, the matter overtakes the mind eventually. You feel tired and a little nauseated. And your hair falls out, and you get a little depressed. They’re in menopause. There’s a lot of issues that go on. And that, of course, was the case for this woman as well.

The other thing I see with the young women is that they sometimes actually have a lot of toxicity with AC. I would love it if we knew who needed AC or didn’t in the neoadjuvant setting. And there’s some fascinating work going on in that area as well, using docetaxel/carboplatin backbone that our colleague Priyanka Sharma has been working on in Kansas. And I think that you do trade some toxicity in terms of nausea throughout, but if you could avoid the AC, that would be nice, for a number of reasons. Anyway, she has had a fabulous clinical response and imaging response on MRI but has really struggled, I think, physically and psychologically with the treatment.

DR LOVE: I’m sure. And I wonder, I’m sure it must be pretty interesting to go to work as a — she’s working as a clinician?

DR RUGO: Eventually. I would say to people, let’s take it week by week. She did work in her capacity for the first probably 12 weeks during the paclitaxel, although she found it increasingly hard. And I’ve said to women, if you burn the candle at both ends, it can be harder and harder to get through the chemo. And by the time she got to AC, she was willing to take off on short-term disability, which was really critical.

DR LOVE: Yes. It’s an interesting experience I’m sure, particularly for somebody who is going to be seeing patients. You wonder how it affects their perception. We once did a roundtable of 12 radiation oncologists, urologists and medical oncologists who had prostate cancer.

DR RUGO: Oh, fascinating.

DR LOVE: And it was really interesting. Some of these people had very low-grade, 100% curable and were completely convinced they were going to die for sure. You wonder, when people go through it, when they go back to practice, how it affects the way they see their patients.

DR RUGO: One thing that’s intriguing, I think, and I don’t have — this is all anecdotal data, but in general, my physician patients, and sometimes others, but a lot of time physician patients who are just working incredibly long hours and seeing sick patients and — they push and push in the beginning to work through the chemotherapy. They get to the surgery, and they’re, like, done. They need time to recover physically and psychologically and get back on track.

A number of my patients have really needed to take time off, 6 months, sometimes 12 months, after they finish in order to really get back on track and adjust to the fact that you are a survivor, or as some people say, thriver, where you really are on the other end of something that was really, really terrifying, and you’re still getting through that and surviving, whether or not you’re on hormone therapy or you’re done with chemo and hoping that the cancer doesn’t ever come back.

DR LOVE: Yes. And I really can imagine how much that would affect how people the way people see their patients in their own practice and how they relate to them.

We started out here. This lady sounds like she’s on her way in a positive direction. It’ll be interesting to see what happens 5 or 10 years from now, maybe she’ll have kids and be successful in practice. That would be really, really inspiring and great. And I’m sure it happens a lot.

Case: A woman in her early 60s with metastatic TNBC is treated with atezolizumab and nab paclitaxel

DR LOVE: Let’s hear about your 61-year-old lady, because I want to spend most of our time now talking about metastatic disease and new developments. What happened with this lady?

DR RUGO: This woman actually was in another part of the United States when she originally was diagnosed with a breast cancer. She felt a mass in her breast a few years ago and then had a 2.1-cm high-grade cancer, negative nodes and triple-negative. She received standard adjuvant chemotherapy with dose-dense AC followed by paclitaxel and then radiation therapy, because she had breast-conserving surgery, appropriately so.

It’s interesting. Would we now give that patient neoadjuvant therapy? I have to say, for triple-negative disease, I really like to give neoadjuvant therapy, just like HER2-positive breast cancer, for any tumor that’s 2 centimeters or greater and maybe in some of the T1cs. Because you could maybe use more of the nonanthracycline-based regimen if they have a great response.

Two-thousand-fifteen, maybe a different approach. Now if I could see a patient like that, even though they have surgically accessible disease, I would really consider doing neoadjuvant therapy versus adjuvant.

DR LOVE: I think there’s been already a major trend/shift towards — we did a big project looking at this in terms of HER2-positive and triple-negative, using neoadjuvant therapy — the surgeons like it, because it gives them a better chance to get good margins. You can maybe avoid axillary dissection. A lot of advantages. But now you have a new model out there in terms of the KATHERINE trial in HER2-positive disease really pushing people to use neoadjuvant, because now it gives you the postop ammunition — a different ammunition in T-DM1. And maybe, do you think that could be the way we’ll end up evolving in triple-negative disease as these new agents come into practice?

DR RUGO: I hope so. I think that we have 2 ways of trying to modify treatment based on response, and that’s our only chance. It really is. And look at the huge data from KATHERINE, so impressive, really changed practice right away in the US. But we also have CREATE-X for triple-negative disease, where the subset of patients who had triple-negative disease had the improved overall survival. And although I think most of us think of capecitabine as not being our favorite drug in triple-negative breast cancer, the data has really changed the practice, standard of care, around the world, because the patients live longer. You can’t argue with survival. I think that we already have data in triple-negative disease.

The other thing is that it may be that we can alter therapy, although we need to prove that still in the neoadjuvant setting based on response. If you have a sluggish response early, maybe going on and adding in different treatments will allow us to better individualize therapy for patients and not give everything to everybody. I do think it’s critical, and it should already be incorporated into practice, but I think it will be even more so as we move forward.

Importance of accurately determining tumor of origin in patients who present with metastatic disease after diagnosis of multiple primary cancers

DR LOVE: Let’s talk about your 61-year-old lady diagnosed with triple-negative breast cancer in 2015. What happened then?

DR RUGO: Then she actually had a routine colonoscopy 3 years later and had an adenocarcinoma of the colon that eventually was resected and cured by resection. But at that time — and she doesn’t have any germline mutations, interestingly — but then she had a chest x-ray in preparation for moving forward with definitive treatment of her colon cancer and had a concern for lung nodules. Then she had staging scans with the CT of chest and abdomen, and it showed lung nodules and mediastinal adenopathy.

They basically went ahead and resected her colon mass because that was already planned, like, 2 days after her CT scan, and found a colon cancer with negative nodes. But then she went on and had a PET-CT after her surgery, which showed these lung nodules and some mediastinal adenopathy that was all FDG avid. And on biopsy, that was a triple-negative breast cancer, similar to her original cancer.

DR LOVE: Could I just ask how comfortable you were that this was her breast cancer, not her colon cancer or something else?

DR RUGO: Yes. I mean, it had markers that looked just like breast cancer. It wasn’t a colon cancer.

You can actually go and do a lot of markers in the biopsy sample and tell the cell of origin, unless a tumor is highly dedifferentiated. I mean, sometimes these carcinomas of unknown primary, you just can’t tell where it comes from. Sometimes you can rule out a few organs. But in this situation, this was adenocarcinoma of the breast in her mediastinal nodes. It’s a very good point and an important part of her staging. The colon cancer needed to be resected.

DR LOVE: If she were ER-positive or HER2-positive it would made it a little bit more straightforward, I would have thought, although you can see HER2-positive colon cancer too.

DR RUGO: Yes. And you can sometimes see expression of ER. You really need the other markers that tell you cytokeratin and stuff, et cetera, that will give you some information about this. And they did do markers to show that this was a breast cancer origin.

DR LOVE: It is interesting when you think about the path that a patient takes, triple-negative breast cancer versus, let’s say, metastatic colon cancer. Very different treatments. Not sure the long-term outcome is that much different, but just interesting how different the therapies are.

DR RUGO: Yes, the therapies are different. I mean, some colon cancers, like breast cancers, patients can live a long time with incurable disease, and it’s probably a little less common in poorly responsive triple-negative breast cancer. And there are a number of targeted agents now, so…

Approach to genetic testing for women with metastatic TNBC

DR LOVE: I was going to ask you before with your other patient, the younger patient, but I’ll bring it up here because I think it’s relevant too, which is, when you see a patient who now is diagnosed with metastatic triple-negative disease, 61 years old, I think no family history, what kind of germline or somatic mutation workup do you think is appropriate at that point?

DR RUGO: In this patient, she actually was under the age of 60 when she was diagnosed, so she had germline testing then, which showed no mutations. But at 61, under the age of 60 there are a number of different series that have looked at patients who were tested, and if you have no family history in triple-negative disease over the age of 60, the incidence of germline mutations, certainly in BRCA1 and 2, is very low, but I think you have to individualize it based on the patient. And in general, the oncologic community has leaned more towards testing patients who have metastatic disease, because we have the option of giving PARP inhibitors to patients who have germline mutations in BRCA1 and 2. Even patients who are over the age of 60.

That’s not in our guidelines, mainly because the incidence of the mutations is so low. But in a patient who has children, et cetera, I think at some point you really have to think about whether or not it’s worthwhile doing some sort of testing looking for also other germline mutations that may have a lower incidence of breast cancer associated with the mutations but still would be important for counseling.

DR LOVE: Let me just clarify this, because I think it is a very important practical point in the older patient, for example, and in an age where oncologists are looking for NTRK fusions and all kinds of other stuff that’s not too common, because we have good therapies or at least somewhat effective therapies. Obviously you mentioned PARP inhibitors. For sure BRCA germline. I’d say BRCA somatic. And then you get into the whole arena of, like, what I hear people calling different things, but DNA repair deficiency, which sometimes I hear people say, “Oh, they have that. Let’s try a PARP inhibitor.”

But what would you say about — forget 61. Seventy-year-old woman, no family history, if you do panel germline testing and, we’ll say, NGS or something, what’s the chance that you’re going to find something you might want to try a PARP inhibitor in for metastatic disease? I’m talking about somatic mutations that people group under the DNA repair, all kinds of things I’ve never heard about. But there’s a whole bunch of things like that.

DR RUGO: It’s been easier with the use of NGS and some drugs that are disease agnostic based on NGS. Because if you get NGS, of course you’re getting the germline too, and so if you see a mutation, if you haven’t gotten germline testing, you can go back and actually just look at the germline to see whether that’s germline or somatic. That actually gives us a way of doing testing, even though it falls outside of our general guidelines for just germline diffuse testing. You will get any mutation there.

The chance in a 70-year-old of finding something is relatively small. I think that we’re certainly in the less-than-10% category and, probably, in a 70-year-old with triple-negative breast cancer, probably in the even less than that. I don’t know that we have full data on metastatic disease looking at somatic mutations in triple-negative breast cancer. They’re fairly uncommon. You’re really looking at a very small number.

We also don’t know what the efficacy is of these PARP inhibitors in patients who have somatic mutations or alternative germline DNA repair defect mutations. And that’s being studied, actually, in a trial that we have within our Translational Breast Cancer Research Consortium that’s being run by Nadine Tung in Boston. And I think really fascinating, because even for a PALB, which involves BRCA, we don’t have approval for PARP inhibitors. Hopefully we’ll learn a lot more from that trial as well as others that are going on.

Use of genetic test results in clinical decision-making

DR LOVE: We did a meeting at AACR this year where we had a breast cancer person, an ovarian cancer person, a prostate cancer person and a pancreatic person, and guess what they were all talking about? PARP inhibitors. And then when you start asking them, like, which somatic mutation… It is certainly an evolving area that is affecting a lot of different cancers.

But in any event, this lady it sounds like didn’t have anything in that direction. But she did have a number of assays done that were interesting.

First, Mayo Clinic PD-L1 SP142 testing, 3% positive, IHC. Foundation Medicine testing, TP53 splice site, low TMB, MS stable. How do you incorporate these pieces of data into thinking through her options?

DR RUGO: It’s interesting. Although there are reports of patients having MS unstable and higher tumor mutational burden, it’s extremely uncommon in breast cancer, and studies looking at patients who have intermediate or high tumor mutational burden really need to have multiple centers, because it is so uncommon.

There is some data suggesting it might be more common than we thought, but I just don’t see it in my patients really. Maybe 1 ever. P53 is a common mutation that’s seen loss of the tumor suppressor gene. We can’t target it, and it’s associated with more resistant and aggressive cancers. We see it in all different types. But in triple-negative disease it’s not uncommon.

Improvement in overall survival with atezolizumab and nab paclitaxel for patients with PD-L1-positive TNBC

DR RUGO: The PD-L1 testing is interesting, because this really comes from the FDA approval of atezolizumab and nab paclitaxel from the IMpassion130 data, where the patients who benefited from the addition of atezolizumab were those who had tumor immune cells, so not the tumor cells but tumor immune cells that were positive for PD-L1 using this VENTANA SP142 antibody. And the patients who were defined as positive were 1% or greater. And so far in the analysis of the patients on those trials, we haven’t seen a difference based on 3+ versus 2+ versus 1+ or higher percentage of positivity.

One percent or more was categorized as positive. And in those patients we saw an improved survival, as you know, with the addition of atezolizumab.

This testing was done because the patient actually was more than a year from her adjuvant paclitaxel treatment. And so she would have been eligible for IMpassion130. And doing PD-L1 testing tells you whether or not a patient would be eligible to receive atezolizumab.

DR LOVE: I want you to go through those data and what you think it means. But before we get to that, just again to take a step back looking at this patient, I’m gathering she was asymptomatic, no pulmonary symptoms. But this imaging sounds a little bit scary. Like, were you concerned that maybe it was going to blow up into lymphangitic disease or something that would be a real problem?

DR RUGO: Lymphangitic spread obviously we hate, right? This is a terrible consequence of metastatic disease. But having lung nodules and mediastinal adenopathy doesn’t make me worry too much about lymphangitic spread. I think mainly we were worried about the nodes. And, in fact, her lung nodules weren’t that big. I mean, the biggest nodule was, on the PET-CT, was 1.3 centimeters. You can see a lot of little nodules, and people can do pretty well with it for quite a while. Sometimes the adenopathy is a bigger issue because it causes compression of the bronchia or blood vessel compression or clot.

She really was asymptomatic, so I wasn’t too worried about it right at that moment, although we don’t — and we know we’re going to give chemo, because we don’t have another option first line. I wasn’t going to give checkpoint inhibitor alone, especially with the IMpassion130 data. Yes, I think the chances she would respond to chemotherapy alone was pretty good. The issue here is duration of response.

DR LOVE: And we didn’t mention, but when she was initially diagnosed in 2015, she had adjuvant dose-dense AC followed by paclitaxel. Did she have any residual neuropathy that was evident at the time that she had the metastatic disease?

DR RUGO: She really doesn’t at all. I mean, it’s really fortunate. She did not have neuropathy. I think that what she — because I didn’t give her the original adjuvant chemotherapy, she said that during the paclitaxel she had some tingling in her fingertips and toes, but it had completely resolved weeks after the end of her treatment. And it only came back when we gave her a number of cycles of nab paclitaxel.

Response and tolerability of atezolizumab with nab paclitaxel on the IMpassion130 study

DR LOVE: And I want to just start getting into the IMpassion data, because she actually got treated with nab paclitaxel and atezolizumab, and you mentioned her 3% level. I’d like you to go through that in terms of what the study looked at. But what actually happened with her when she was treated? What happened to the tumor? What happened to her?

DR RUGO: She started with nab paclitaxel and atezolizumab per the IMpassion130 trial. Actually, she started getting it even before the FDA approval of the regimen. It was nice, because she was able to get atezolizumab based on the strength of the data that had been presented and published.

And what she had, she had a follow-up scan after receiving a couple of months of treatment and had a marked reduction in her disease. Her lung nodule, which had been 1.3 centimeters, was 4 millimeters. The mediastinal adenopathy had completed resolved. There were a few nodes you could see, but none of them had any standardized uptake value or SUV on the PET-CT that she was getting. Then she continued along with her chemotherapy. And then I think one of the questions that came up then was, “How long do we continue therapy for? Should we at some point drop the nab paclitaxel and continue atezolizumab alone?” We don’t really know the answer to that yet. I hope that we do.

It’s interesting. With lung and node disease, some data had suggested that there are more TILs and greater maybe PD-L1 expression in lung and node versus, say, liver, where patients don’t do as well. That may also correlate with a poorer outcome in patients who have very high LDH, that they tend to have a lot of liver disease. Hard to know.

She had a follow-up PET-CT, again, that showed essentially no residual hypermetabolic foci and just nothing left. The idea is to maybe give her another couple of cycles and then at some point drop the nab paclitaxel.

She had, after receiving 6 cycles, some peripheral neuropathy. And really significant fatigue and was feeling quite depressed about this, that this quality of life was kind of miserable for her. We talked about a couple of different ways to make it better for her, either giving 2 weeks on/1 week of versus 3 weeks on/1 week off or dropping her dose of nab paclitaxel. We actually dropped her dose of nab paclitaxel to 80 mg/m² from 100, and she’s actually feeling much better and feels like her neuropathy is a little bit better.

Again, I think that at some point we’re going to have to give her a chemo holiday and continue atezolizumab alone. But because she has a PCR now, having started in the beginning of February of this year, and I’m very encouraged by her response and hoping that it will be durable. Interestingly, she’s not had any immune toxicity at all.

Thyroid dysfunction associated with immune checkpoint inhibitors

DR LOVE: And I was going to ask you that. Nothing whatsoever? I’m guessing you’re following her thyroid function?

DR RUGO: We check thyroid-stimulating hormone on a regular basis. I think most of the trials did it every 6 weeks. If people are fatigued, we end up getting it a little more frequently. It’s every 8 weeks on that cycle. Sometimes we get it every 4 weeks and particularly in patients who’ve had some abnormality of their TSH.

Sometimes we see them become hyperthyroid first and then hypothyroid. You really want to catch that abnormality. And then I have a very low threshold for doing additional testing, for example, if a patient develops a cough or diarrhea or other electrolyte abnormalities, et cetera, creatinine problems. But in this patient, she’s had zero.

DR LOVE: Of course we hear that story and that type of story in many different kinds of cancers nowadays. It’s becoming kind of a little bit of a checkpoint world. But globally, when you look at patients with breast cancer, maybe they’re a little bit younger. As you look out over, let’s say, a year, for example, in the metastatic setting, what’s your usual estimate for the likelihood they’re going to have a thyroid abnormality? The likelihood they’re going to have something else?

DR RUGO: I would say somewhere between 5% to 10% of our patients have a thyroid abnormality. It’s interesting. It’ll be interesting to look at time frames in terms of safety in our studies going on with checkpoint inhibitors in breast cancer. I have a tendency to believe that if you’re going to get an immune toxicity, you tend to get it early. There’s been some time courses in the articles published over the last couple of years, kind of interesting, where you can actually see late immune toxicity with colitis and pneumonitis and other kinds of things like that. And I’m sure that’s true for thyroid, it’s just that most of the cases we tend to uncover relatively early.

It’s interesting. I’ve seen patients who’ve developed relatively severe colitis and relatively severe rash with exposure to a single dose of checkpoint inhibition. It doesn’t seem to be so related to having lots and lots of exposure to the checkpoint inhibitor, although you do accumulate more cases over time.

Those are rare, but if you treat a lot of patients with checkpoint inhibitors on clinical trials, you’re going to see more of this. And you have to have a fairly high suspicion.

I had a patient also who had cytopenias. One patient. You can count them all, right, because these are unusual toxicities. I’ve never seen renal toxicity, for example. I’ve never seen cardiac toxicity.

DR LOVE: I’ve heard all of that.

DR RUGO: Or pulmonary toxicity.

DR LOVE: I’ve heard it all. I’ve heard everything.

DR RUGO: Yes.

DR LOVE: Sitting in this chair, everything: cardiomyopathy.

DR RUGO: I bet.

DR LOVE: But I was trying to think, what triggered me to ask you about thyroid? And now I know what it was. When we first started hearing about checkpoint inhibitors in Hodgkin lymphoma, and there, again, you’re seeing younger people, and they were saying, “Oh, yes, we’ve got to really keep an eye on the thyroid.” Maybe, I don’t know, maybe it had something to do with age or whatever.

DR RUGO: I don’t think that — so far, we really haven’t seen any variation in age where we see really — I would say overall, the population I treat might be a little bit younger than the global population of breast cancer just because of the nature of my practice. But we see it across the board.

Biologic rationale for the investigation of immune checkpoint inhibitors in breast cancer

DR LOVE: And of course this issue of atezolizumab in the triple-negative is a huge story. Can you talk a little bit about what we know about the biology and the immunology of breast cancer, what we learned initially about monotherapy and how we ended up here, like now, a lot of cancers: lung cancer, head & neck cancer, bladder maybe, chemo plus checkpoint inhibitor. Can you kind of go through that story?

DR RUGO: It’s actually fascinating, because when I was training and I was a fellow a long, long time ago now, we were giving interleukin-2 to patients, and they would go in the ICU and be so sick. And the renal cell and melanoma patients were being treated. And at that time, we really believed that breast cancer was immunologically dead, that the breast cells looked so much like itself that you just didn’t react to them, and of course that is true for the most frequent form of breast cancer in older women. ER-positive, lower-grade breast cancers have very little immune activity.

But for the more aggressive subtypes that we’ve identified subsequent to that time, triple-negative disease, HER2-positive breast cancer and very proliferative ER-positive disease. This very high-risk ER-positive disease, very proliferative. What we do see in those patients is, we see tumor-infiltrating lymphocytes. And in the early-stage setting, the presence of tumor-infiltrating lymphocytes — and Sherene Loi has done some beautiful work on this in triple-negative disease, as well as others — have correlated with outcome, where patients who have more of these tumor-infiltrating lymphocytes have a better outcome.

We also have found that PD-L1 expression is increased on these more aggressive cancers and that both tumor-infiltrating lymphocytes and expression of PD-L1 seems to decrease as tumors go from the early- to late-stage setting. And as they go throughout the late-stage setting, they become less and less immune sensitive.

Activity of anti-PD-1/PD-L1 antibodies in patients with breast cancer

DR RUGO: What we actually learned — we kind of went backwards in this a little bit — we learned first that there was some activity of immune checkpoint inhibitors in breast cancer, and this was astonishing to us. We’re, like, “Wow, it’s so exciting.” These were basket trials. We saw response rates of 18 and a half percent to 19% on some of those patients. I mean, who cares about a response rate under 20%? It’s because the responders had this great outcome, where their overall survival was huge. I mean, it was just — you look at those curves, and you know that you’re really onto something very exciting. And the treatment was well tolerated, recognizing now that we have to, of course, be on top of the immune toxicity.

What happened then was, these expansion trials were developed with pembrolizumab and atezolizumab in patients who had triple-negative disease, including first and greater lines of therapy. And what we found from both of these trials was that the response rates in patients who are in the second line or greater — of course there were many patients who were farther along the road than second line — the response rates were quite low in the probably less than 5% or 5% range, so very, very low. But if you treated patients in the first-line setting you got response rates, combining the 2 studies, in the 25% or so range, and those responses could be quite durable.

Interestingly, in the study where they looked at second- or greater-line therapy, we didn’t select for PD-L1 positivity, but in the first-line trial we did select for PD-L1 positivity in the pembrolizumab trial. And with atezolizumab, there was a suggestion of potentially better responses in PD-L1-positive disease. Again, small numbers there looking at that subset.

That was really exciting. It made us go back and do some of those studies, which I mentioned, to have shown that you actually start turning down your immune response and decreasing PD-L1 expression as the cancer progresses. We’ve also learned that where the cancer metastasizes to may make a difference as well. Hard to know.

That, actually, if you go back to the laboratory and animal models and in humans see that if you give chemotherapy you can enhance the immune infiltration of tumors in some cases, that led to the IMpassion130 trial, which was the idea of giving chemotherapy and a checkpoint inhibitor to see if you can improve outcome, looking at progression-free survival and overall survival.

At the same time, in parallel, the idea was maybe if we treated even earlier in the course of the disease we could change outcome. A bunch of neoadjuvant studies started up that extended the checkpoint inhibitor through the adjuvant setting and were really stimulated by these PCR marked improvement we saw in the I-SPY 2 Phase II adaptively randomized trial.

Design, eligibility and outcomes of the IMpassion130 study evaluating atezolizumab with nab paclitaxel as first-line therapy for advanced TNBC

DR RUGO: IMpassion130 is a fascinating trial, now published and updated and led to FDA approval of the combination. It enrolled patients who had triple-negative breast cancer that were at least a year out from their adjuvant or neoadjuvant taxane chemotherapy. Patients were randomized to receive nab paclitaxel with placebo or nab paclitaxel with atezolizumab. We used 100 mg/m² 3 weeks on/1 week off, and I really feel that was a critical decision, because when we made that decision as a group, there was a lot of interest in using higher doses of nab paclitaxel, 125 mg/m². The problem with that is that you get a lot of neuropathy. And patients fall off. They can’t continue the treatment. I’m really pleased that we chose the 100 mg/m² for this metastatic trial.

The endpoints are interesting, because there was a lot of interest in understanding whether PD-L1 positivity would make a difference in response to atezolizumab. The primary endpoint was progression-free survival in the intent-to-treat arm, and if that was positive looking at the PD-L1 subset population, along with overall survival in intent to treat, and then if that was positive, looking at the PD-L1-positive population.

DR LOVE: Could I just clarify one thing in terms of the percent of patients globally in this situation who have greater than 1%?

DR RUGO: We didn’t know. I mean, that was the thing. Who knew? And we didn’t have exactly that trial population.

DR LOVE: But what do we know now?

DR RUGO: We know now, and it was 41% of the trial population that had immune cells that were at last 1% positive for PD-L1 using the VENTANA SP142 assay.

DR LOVE: Most patients did not.

DR RUGO: Most patients did not. Sixty percent. Which is why it’s so incredibly important that we’re pursuing these combinations of immune agonists with checkpoint inhibitors. But not only that, that we’re moving the treatment earlier in the course of therapy, because it may not actually matter as much if you have PD-L1-positve disease in the neoadjuvant setting or adjuvant setting. And we’re going to learn a lot more about that in the next, actually, very short time period, based on a press release suggesting that patients had a better PCR from a Phase III trial using pembrolizumab in the neoadjuvant setting regardless of PD-L1 positivity, using a different assay, I’ll point out. An assay, which marks most tumors as PD-L1-positive.

DR LOVE: But just to clarify, though, in terms of the actual overall results of the study as well as the results within the 1% population, in terms of hazard ratios and absolute differences, progression-free survival, response rate, survival.

DR RUGO: I think that in this trial probably the absolute differences are more important to us than the hazard ratios. What we saw was something fascinating, to me. Had been seen in other diseases, actually, with immune therapy, and was part of the reason why patients could not be unblinded in the trial. Even after their disease progressed, they couldn’t be unblinded, because the hypothesis was that immune therapy might impact survival more than it impacted progression-free survival.

It’s actually fascinating that in the CLEOPATRA trial we saw that. Big PFS difference. But, actually, almost 3 times greater difference in overall survival. And it’s again, we know HER2-targeted therapy has an immune effect as well.

The idea is that you change the tumor microenvironment. And so subsequent therapies work better and you live longer, even though potentially you don’t have a much longer disease control rate with the first-line therapy. We saw that in IMpassion. Fascinating.

The difference in progression-free survival — I remember when we first saw that data — the difference in progression-free survival in the intent-to-treat population was less than 2 months. Significant, but really, using an expensive drug, potential toxicities, is it worth it? Then we looked at the PD-L1-positive population, and that was more impressive with this difference that was in the 2 and a half-month range, so maybe more worthwhile treating patients.

But then when we looked at overall survival, no significant difference in the whole population, intent-to-treat. By the original design, then you wouldn’t look at the PD-L1-positive population, but of course you have that data, so you look. And it was 10-month difference. Striking. Something we don’t see by chance. And the FDA actually agreed with that, and that data led to the approval for patients who have PD-L1-positive disease.

In longer follow-up, that was just the first interim analysis. At the second interim analysis that was recently presented, we saw a difference of 7 months in overall survival. There is no other therapy that has that impact in triple-negative metastatic breast cancer. Unfortunately, it was only seen in the PD-L1-positive population. It’s fortunate in that it’s more precision medicine. We can identify the group that benefits. Unfortunate in that there’s 60% of patients who don’t have PD-L1-positive disease, where we need to keep looking for ways of improving response to immune therapy.

Therapeutic approach for patients with TNBC and BRCA mutations

DR LOVE: I want to weave in the next topic, which is PARP inhibitors in triple-negative disease, and I want to hear about your patient, your 43-year-old woman who ended up getting a PARP inhibitor. And, of course, I want to hear about the data and what your take is on it. But just to introduce it, just out of curiosity, I want you to think about this same patient you just discussed with the lung nodules that you ended up giving the nab and atezolizumab to. And I guess my question is, what would you have done if you knew she had a BRCA germline mutation?

DR RUGO: I would have treated her the same way. I think there’s a lot of interest in giving PARP inhibitors with checkpoint inhibitors, because there’s some data that you can goose up the immune system via the STING pathway through interferon, mediated cell inflammation. And so the data from 2 Phase II studies, not randomized, looks pretty intriguing. And there’s now randomized studies going on and also subgroup studies going on. But there’s a lot of interest in using — I mean, this is not in any way the approval for PARP inhibitors — but there’s interest in using that combination as maintenance therapy after chemo.

I was talking about this patient — I mean, boy, you get chemo fatigue. I mean, you’re looking, oh, my median survival is 18 — I have to say, 7-month improvement in survival means your median survival is a little over 2 years. It’s not, like, great. This patient, I think, will live appreciably longer, but she’s depressed. She doesn’t want to stay on chemo forever. You can’t travel as far.

DR LOVE: Yes. But, I mean, that’s about the chemo. You could always keep the atezolizumab going.

DR RUGO: Right.

DR LOVE: I’m just talking practical right now. Practical.

DR RUGO: Right. This is practical. You drop the chemo you could actually — so in this patient, if she had a BRCA mutation, the idea would be — you’d still start with a treatment that’s been demonstrated to improve survival, but as she’s saying, complaining about her chemo fatigue, et cetera, you could say, “We could always drop the chemo, continue atezolizumab,” and we’re studying adding in PARP inhibitors in this setting.

Let’s say you didn’t want to take that exploratory approach, which is being studied in trials. You could say, “We’ll drop the chemo, continue atezolizumab and if your cancer progresses we could either add back in chemo or we could give PARP inhibitor, oral therapy, pretty well tolerated. And this would be a great option for you,” and it’s an option for her after the atezolizumab/chemotherapy combination stops working as well. And I think most of us would now use it as a second-line approach, given the fact that we have survival data in the first line.

DR LOVE: It’s so interesting, because when we ask oncologists — we present cases of metastatic breast cancer and BRCA germline, and a lot of them talk about using chemo followed by PARP maintenance, because they’re used to that from ovary. It makes complete sense to them. And then we go, “Yes, but there’s no data in breast cancer on it,” but yet it’s just so appealing. It makes so much sense. And as you’re saying, clinically it makes a lot of sense. But let’s bring in the data, because obviously it sounds like you’re more impressed with the bump that the PD-L1-positive patients get in survival, and that makes a lot of sense.

Case: A woman in her early 40s with a germline BRCA1 mutation receives chemotherapy followed by maintenance olaparib for metastatic TNBC

DR LOVE: But I do want to find out how you approach the issue of particularly genetic testing and integration of PARP inhibitors and which one, a common question for oncology. Do you know there are 6 checkpoint inhibitors that approved nowadays? Anyhow, 3 CDK inhibitors, et cetera. But we’ve got 2 PARP inhibitors approved in breast cancer. Anyhow, let’s hear about your 43-year-old lady. What happened with her?

DR RUGO: This is a really tragic case, as many of our patients are who have metastatic triple-negative breast cancer. A 43-year-old woman who had a triple-negative breast cancer — she actually had surgery first, unfortunately. I don’t like that approach so much. And she had a 4.3-cm triple-negative breast cancer with LVI and 4 positive nodes with extracapsular extension, so Stage III disease. Really unfortunate.

She actually had a family history with a paternal grandmother who had breast cancer at age 50 and a paternal cousin, and this is actually a real key you should think BRCA mutations. This paternal cousin had both breast and ovarian cancer. In this case the paternal cousin was known to have a BRCA mutation, but even they’d not, those 2 cancers together or any ovarian cancer is really a key in male breast cancer.

This patient unfortunately did have a germline BRCA1 mutation. And she received standard adjuvant chemotherapy. She received dose-dense AC times 4 followed by weekly paclitaxel. But, actually, as she got to her tenth cycle, and this is our worst of the worst in terms of triple-negative breast cancer that presents with early stage, is cancer progression during neoadjuvant or adjuvant chemotherapy.

While she was getting her paclitaxel, around cycle 10 and probably a little bit earlier, she started noticing redness on the left breast that spread over her reconstructed left breast and had a skin punch biopsy around cycle 10 of paclitaxel that showed metastatic carcinoma, triple-negative. She went on to receive additional chemotherapy, but she got very intensive chemotherapy at another location, causing a lot of thrombocytopenia. She was really limited in her treatment due to thrombocytopenia at that time.

The cancer actually responded, even though with a limitation, and the redness actually resolved, initially. Then she received radiation therapy, stopping the chemotherapy. Of course the radiation therapy caused her to continue to be thrombocytopenic, which was unfortunate, because really, it hurts for us to treat low platelets. And then at the end of her radiation, she started on a PARP inhibitor.

It’s kind of interesting. This a patient who has metastatic disease, recurrent inflammatory disease in skin, on her chest wall, nonresectable disease. Gets an induction chemotherapy and really was using olaparib in a maintenance setting in this situation, knowing that she had really a terrible prognosis with growth on adjuvant paclitaxel.

She received olaparib and had her ovaries out. She had her ovaries out at her local institution. I wouldn’t have been really pushed to take her ovaries out at that time, but the plan had been made originally to take her ovaries out, so they went ahead, actually, with the surgery, and she felt better about doing that. She stayed on her olaparib for about 5 and a half months. During that time, we struggled a little bit with her platelets. Remember, this is a patient who has an individual risk for thrombocytopenia. She didn’t have any anemia, which is the most common cytopenia we see with PARP inhibitors, but her platelets kind of up and down, always under 100,000.

Then she started having a recurrence on the contralateral breast after about 5 months of therapy. And it’s really interesting with these patients with chest wall disease. At that point, she had no disease systemically, but the disease was increasing on her chest wall. And she then became more neutropenic. Her platelets were 50,000, so we stopped the olaparib in order to try and treat the progressing disease on her chest wall.

Efficacy of olaparib and talazoparib for patients with advanced breast cancer and a germline BRCA mutation

DR LOVE: Maybe we can just take a breath here, and this case obviously is very complicated, but I did want to get your take on the data with PARP inhibitors. Bottom line, we have 2 approved. What are the key Phase III data that fed into that approval? And what’s it mean in terms of how you approach patients and also how you test patients?

DR RUGO: It’s interesting. I mean, we participated in the Phase III clinical trials with talazoparib, but I’ve also given quite a lot of olaparib. And in terms of the trials themselves, 2 Phase III trials with some differences, but not huge, treating patients who had germline BRCA1 or 2 mutations in the metastatic setting with either the PARP inhibitor or treatment of physician’s choice with the selection of chemotherapy agents, both showing a doubling, essentially, of response rates and improved progression-free survival.

The doubling of response you really see in clinic. I mean, you give these patients who don’t have platinum-resistant disease, that was key for both of these trials. But you treat the patients, and they respond rapidly. I mean, it’s very, very quick response, symptomatic improvement. And it lasts a reasonably long time compared to chemotherapy for triple-negative disease, but not long enough. I mean, so the median duration in the 7-month-ish range — the patients’ tumors develop resistance to these PARP inhibitors. They were pretty well tolerated, with cytopenias being the most common toxicity. Some nausea, which we see with PARP inhibitors and is really important to control. Dose reductions, treating with nausea medication. I use sometimes the antipsychotic drug olanzapine for that kind of nausea, because you don’t want to be nauseated continuously on an oral medication. Dose reduction for cytopenias, and then anemia sometimes can be a big issue.

It’s difficult to compare the 2 agents in terms of toxicities because they were 2 parallel trials. You can’t really do cross-trial comparison, so we don’t know very well how to compare the 2 drugs.

But I think the one thing we’ve learned from the OlympiAD trial with olaparib from a subset analysis, which we hope to see an update for — it was recently published, actually — was looking at overall survival where, somewhat disappointingly, there was no difference in overall survival comparing the PARP inhibitor to chemotherapy. But if you looked at the patients who were treated in the first-line setting, in a subset analysis it looked like there was a bigger difference in overall survival. And that’s actually resonated a lot with me, because if you look at mechanisms of resistance in patients who have germline mutations to PARP inhibitors, where you see these acquired acquisition of a function of DNA repair, you actually correct the deficiency of DNA repair as a mutation. And then there are other mutations as well.

We think about triple-negative disease, where we see this plasticity and more and more resistance as you go down lines of therapy. It makes perfect sense that that would apply for PARP inhibitors, so that using the drugs earlier makes a lot of sense in the first-line setting, depending on what your options are, or second-line setting, and that’s where the interest in maintenance is. Of course we’re waiting for results from the OlympiA trial, which is using the PARP inhibitor olaparib in the adjuvant setting. And then we saw this really intriguing data with talazoparib in the neoadjuvant setting in only 19 patients, from Jennifer Litton from MD Anderson, where the pathologic CR rates were really astonishingly high from just getting a PARP inhibitor as a single agent if you had a germline BRCA mutation.

I think where that puts me is that I want to test patients who have metastatic triple-negative disease for sure. Most of the patients have already been tested, because we have NGS as we have talked about. We get the germline testing, in some ways an early look. Younger patients are generally tested in the early-stage setting or right away when they come in with metastatic disease. We know if we can use these agents. We’re exploring greater use across other germline mutations and somatic mutations, although we don’t have data yet. Even looking at ctDNA or plasma to look for somatic mutations is a fascinating new area of exploration.

Tolerability profile of olaparib versus talazoparib

DR RUGO: The choice of what PARP inhibitor to use and when to give it is a big question. If you have horrible early-stage breast cancer with tons of residual disease, that’s even a setting where I would consider using a PARP inhibitor off label, because those patients have such a poor prognosis and we really want to alter their outcome.

Whether you use olaparib or talazoparib, in some ways it takes us back to the aromatase inhibitors and the CDK4/6 inhibitors. I just alternate, because I don’t know. And I think by alternating, I get the greatest experience using the drugs. If a patient doesn’t tolerate one, I’ll switch to another. And I don’t feel like I need to use one over the other in this situation.

DR LOVE: The fact that you do alternate them obviously suggests that you can’t distinguish benefit from the two drugs. You mentioned tolerability. I think I’ve heard people talk about alopecia with talazoparib?

DR RUGO: I haven’t seen a lot of alopecia with either talazoparib or olaparib and certainly not as much as I’ve seen with the CDK4/6 inhibitors, for example.

DR LOVE: Is there anything with talazoparib, is there anything that distinguishes —

DR RUGO: People are concerned about more anemia.

DR LOVE: Yes, but, I mean, they both cause anemia, right? I mean, can you tell the difference?

DR RUGO: A little more anemia. Maybe some variations in nausea between the two drugs. Those are the two main side effects that —

DR LOVE: But nothing qualitative?

DR RUGO: No. Not really. I mean, if you have nausea all the time, it’s a big qualitative effect.

DR LOVE: Yes. No, but what I mean is, different, new kinds of side effects? It’s just the same kind?

DR RUGO: No. It’s interesting. I thought a lot about the anemia and have had people talk about how they managed it. Really didn’t have patients with anemia. The patient with thrombocytopenia, I’ve had some patients with that before. But I recently had a patient who had BRCA2, and, interestingly — actually, she has a BRCA1 mutation and ER-positive metastatic disease. Very young woman. And she had transfusion-requiring anemia on talazoparib. I mean, she needed blood every 2 weeks. We actually dose reduced her and eventually changed to olaparib, and her transfusion requirement went away.

I don’t know if it was the change, because we didn’t wait long enough with the dose reduction on talazoparib, or whether or not there was something else going on. But it was really quite dramatic from when we started until now when she has no transfusion requirement and a normal hemoglobin.

Emerging data with the novel antibody-drug conjugates sacituzumab govitecan, trastuzumab deruxtecan and ladiratuzumab vedotin for advanced breast cancer

DR LOVE: There’s so much more to talk about, but I just want to end here with a couple of thoughts about antibody-drug conjugates. Which ones are out there, and are any of them likely to come into clinical practice?

DR RUGO: I do think they will come into clinical practice, and I hope very much in the near future. There are two, I think probably our two most exciting ADCs in terms of the data available are sacituzumab govitecan, which is the Trop-2 targeted ADC that’s been studied in triple-negative disease. The toxin is SN-38, the active metabolite of irinotecan. That data is actually quite encouraging, with a 30-plus-percent response rate in heavily pretreated metastatic triple-negative breast cancer that appears to be reasonably durable. Side effects of what you would expect with chemotherapy, but able to be controlled.

Some hair loss and nausea and cytopenias. But we’ve actually figured out how to control all of those fairly easily.

DR LOVE: I have to say, though, that when you think about the other antibody-drug conjugates that are out there, T-DM1 and brentuximab vedotin in lymphomas, you don’t hear that much — I mean, peripheral neuropathy and all — but you don’t hear much about chemotherapy side effects. You’re saying this agent actually causes chemotherapy side effects?

DR RUGO: It causes chemotherapy side effects as do the two other ones that I’ll talk about briefly. There is some difference. And what that difference is due to we’re not sure. II don’t think it’s due to the toxin. I think it has to do with the number of drugs per antibody. You can load up now with new linkers and new technologies more of the toxin per antibody. And that may have something to do with it. And it may have something to do with the linker and the metabolism of the linker itself.

Sacituzumab actually has been studied in a Phase III trial called the ASCENT trial, which has mostly completed accrual. There’s a little bit of PK work going on, but it completed the main study accrual, randomizing to sacituzumab versus treatment of physician’s choice in patients who received at least 2 lines of prior chemotherapy for metastatic disease. A really bad situation for our patients. And we’re really excited to see that done I hope next year, from this ASCENT trial.

Sacituzumab is also being tested now in the TROPICS trial in hormone receptor-positive disease, where there was some very encouraging data as well, recently presented. That’s one ADC I think has very encouraging data in triple-negative disease.

The second ADC actually has been tested in HER2-positive disease, trastuzumab DS-8201. That has a novel toxin, exatecan derivative, and linked to a trastuzumab, essentially, biosimilar. And that has shown some really remarkable responses in patients who have resistant HER2-positive disease. But in patients who had a new category of cancer that’s been defined really by this agent called HER2-low disease, 1+ and 2+ expression, no FISH amplification, those patients also had a very impressive response that seemed to be quite durable. A subset of patients with triple-negative disease will have 1+ expression of HER2 by IHC and could be eligible for that agent. There’s now a Phase III trial going on comparing DS-8201 to treatment of physician’s choice in patients who have HER2-low metastatic breast cancer.

Again, limits on line of therapy. It’s incredibly important if you’re thinking about these trials for your patients to explore them early, because there are always going to be limits on line of therapy. If you use up all your therapy, they won’t be eligible for the trials.

We hope to see accelerated approval of sacituzumab and potentially trastuzumab DS-8201 sometime in the not terribly distant future based on the efficacy data we’ve already seen presented and, in some cases, published.

There’s a third ADC that’s actually being tested in TNBC called ladiratuzumab. And ladiratuzumab is an ADC to LIV1A. And its toxin is MMAE, an auristatin analog. That agent has been tested. It has some efficacy as a single agent. A little bit of neuropathy and cytopenias. Some other toxicities associated with the auristatin analog.

There’s been some very interesting data in combination with checkpoint inhibitors. Of course now the next wave is combine all of these ADCs with checkpoint inhibitors and see if we can further enhance the activity, since the antibodies have an immune effect. You can enhance the immune effect and give the chemo to goose up the immune response. There are a lot of reasons why those combinations might be appealing.

Case: A woman in her early 50s with de novo metastatic TNBC is enrolled on the IMpassion130 trial evaluating atezolizumab with nab paclitaxel

DR TELLI: This was a lady who had, over a few months’ period of time, started to notice a fullness in her upper left breast. Ultimately she came in to get this evaluated. She had diagnostic mammography ultrasound. It showed a very large, nearly 8-cm mass in her left upper breast, multiple enlarged left axillary lymph nodes and she had a core needle biopsy that showed a high-grade invasive ductal cancer that was triple negative with a high Ki-67 of about 70%.

She then went on and had routine staging studies, because she, at presentation, clinically, was Stage III. And this unfortunately showed a lesion in the spine, a solitary liver mass and then multiple hilar and mediastinal lymph nodes.

DR LOVE: As you looked at her, how concerned were you about local control in the breast?

DR TELLI: Her tumor was very large. The entire breast was enlarged compared to the opposite side, and she was having some discomfort at the site. She didn’t have any inflammatory appearance to the skin, but it was a very locally advanced presentation. But certainly the CT scan findings showing the liver mass that was highly suspicious. It was positive on an FDG PET scan. And so she went forward and did have a biopsy to confirm the suspicion of metastatic disease. She had one of the mediastinal lymph nodes biopsied, and this did, unfortunately, show metastatic carcinoma. She was Stage IV at presentation, which is a less common presentation of breast cancer here in the United States.

DR LOVE: What was her life situation, and do you have any sense whether she delayed coming in or not?

DR TELLI: I think that she was married, newly married, and worked full time in IT. I think that she initially just kind of felt that there was something going on in the breast but that it was more vague, and it wasn’t until it progressed over a period of a few months that she became more concerned that it was a mass there.

Initial workup and genetic testing for patients with metastatic TNBC

DR LOVE: Fifty-three years old. I see paternal grandmother with breast cancer diagnosed in the 80s. But in general, what would you consider a minimum workup for a patient in this situation?

DR TELLI: Certainly the systemic staging, given the Stage III presentation. Biopsy of the metastatic site. And then for any patient with triple-negative breast cancer, regardless of whether they have a family history of breast cancer or not, if they’re age 60 or less at diagnosis, NCCN Guidelines would endorse germline BRCA1 and BRCA2 mutation testing. And so she did have this done, and it did not show a mutation in BRCA1 or BRCA2.

DR LOVE: You said the NCCN Guidelines, but now we’re talking about metastatic disease. Does that apply, or does every patient with metastatic breast cancer need BRCA testing?

DR TELLI: Yes, so it’s been a moving target. As of last year, in 2018, we had the first approvals of PARP inhibitors for advanced BRCA-associated breast cancer. And so the guidelines have shifted over time, that for patients with metastatic disease, if they have HER2-negative metastatic breast cancer that they should have germline BRCA1/BRCA2 mutation testing, because this targeted therapy is now available. She would have met the criteria regardless.

DR LOVE: Also, you have the question of panel testing germline, but you also have the testing of multiplex somatic testing, such as NGS. Again, what do you consider standard at this point?

DR TELLI: I think, again, that has certainly been a moving target in breast cancer. Increasingly we have more and more reasons to do something like a next-generation sequencing assay of a metastatic site, because we are increasingly having more targeted therapies. This year we had the approval of alpelisib for hormone receptor-positive breast cancer for patients with the PIK3CA mutation in their tumor. We have the PARP inhibitors for those with BRCA mutations, and so increasingly there are more and more reasons to be getting it. And so I think that certainly where we are today, we’re seeing an uptake in NGS early on in metastatic disease.

Risk of recurrence for patients with localized TNBC; patterns of metastasis

DR LOVE: In terms of the workup, I want to ask you, in terms of immune-related workup, particularly PD-L1 testing. But before we get into that, just a couple of basic points about triple-negative metastatic disease. Can you talk about the issue of the hazard rate for distant mets with triple-negative disease compared to the rest of breast cancer?

DR TELLI: Yes, so we’ve known for a long time that for patients who are diagnosed with localized breast cancer, if the tumor’s triple-negative, really a lot of the risk that they face is early. There’s a much higher risk hazard for recurrence early, particularly in the first 3 years, and this really tapers down, particularly after the 5-year mark, approaches zero at that point. But for nontriple-negative breast cancer, we know that there’s a persistent hazard for recurrence over time. It’s certainly a very different pattern in terms of the timing of recurrence for patients initially diagnosed early stage.

DR LOVE: What about patients presenting with metastatic disease, as this woman presented? Do you see a higher fraction that are triple-negative? I feel like I hear more cases like that.

DR TELLI: Yes. Certainly I think the more virulent breast cancer subtypes are the ones that are more likely to be diagnosed with de novo metastatic disease, HER2-positive, triple-negative. There are, however, ER-positive, HER2-negative cases that present in this way too, often, though, with isolated bone metastasis.

If you look at the pattern of metastasis for patients with triple-negative breast cancer, what is really striking is that there is a much higher likelihood that they’ll have visceral organ involvement, liver/lung metastasis, even brain metastasis, as a site of first recurrence. And bone, for example, as a site of first recurrence, while very common in hormone receptor-positive disease, is quite uncommon for triple-negative disease.

DR LOVE: Yes, that really struck me on that curve. Not that we hadn’t heard it, but just to see the number, bone as primary — first met, 10%, versus 40% for nontriple-negative, so that’s really interesting.

Approach to PD-L1 testing for patients with TNBC

DR LOVE: Getting back to this lady, obviously one of the issues is going to be immunotherapy in this situation. Can you talk a little about how you approach patients like this and how you approached her, particularly as it relates to PD-L1 testing?

DR TELLI: Yes, so of course we had the approval of atezolizumab earlier this year, in March, given in combination with nab paclitaxel for patients with metastatic PD-L1-positive, triple-negative breast cancer. And in the IMpassion130 clinical trial, they selected patients as being PD-L1 positive using the VENTANA SP142 antibody.

There are a number of PD-L1 antibodies out there, and there is certainly discordance if you look at the same tumor across these different antibodies, but my practice has been, since as of today the drug we have approved is atezolizumab, my practice has been to use SP142 as the antibody of choice. If you look at the approval from the FDA of atezolizumab, or if you look at the update to the NCCN Guideline following this data, the language there is that any FDA-approved PD-L1 test could be considered. I personally have stuck with ordering the test that was the actual antibody used in the trial.

DR LOVE: As you try to make clinical decisions and think about other options, the trial, I think it was greater than 1%, but do you feel more likely to utilize atezolizumab/nab in somebody with a higher PD-L1 level, 20, 30, et cetera? Lung cancer, we have kind of a 50% division line. What about in breast cancer?

DR TELLI: Yes, I mean, it’s really interesting in breast cancer. If you look at that data from IMpassion130, most of the PD-L1 expression is at the really low level. The vast majority of the patients had staining between 1% to 4% in that clinical trial. And there were fewer patients that had 5% or greater. We just don’t see the really high PD-L1 expression in triple-negative breast cancer like you can see, for example, in non-small cell lung cancer. And, of course, this is the, again, the SP142 antibody, and the way that it is reported, it’s looking specifically at PD-L1 expression on the immune cell compartment, not the tumor compartment. But it’s really striking that most of the patients, when you order this test, they come back 1%, 5%. It’s very unusual to see, in my experience, something greater than 10%.

Results of the Phase III IMpassion130 study of atezolizumab with nab paclitaxel for previously untreated advanced TNBC

DR LOVE: Can you talk a little bit more about some of the findings in the trial, both in terms of the benefit that was observed as well as toxicity?

DR TELLI: Yes, so this was, of course, a really important trial. The first PD-L1 inhibitor approved in breast cancer. And so this was a Phase III trial, first-line, triple-negative breast cancer. And patients all received nab paclitaxel dosed weekly days 1, 8 and 15 on a 28-day schedule. They used a little bit lower dose of nab paclitaxel, 100 mg/m². And then patients were randomized to atezolizumab dosed at 840 mg on days 1 and 15 of the 28-day schedule versus placebo. And patients were then followed for the coprimary endpoint of the trial, which was both progression-free as well as overall survival in the overall population but also specifically in the PD-L1-positive population.

And so when you get into the nitty gritty, the statistical design of the study, because of these coprimary endpoints and because of the desire to look at both the overall population and specifically just the PD-L1-positive population, it was a very complicated statistical design. And I think that’s where there’s still a lot of question out there regarding the true benefit of this therapy. In terms of progression-free survival, they hit the mark, the primary endpoint for both the overall group of patients and the 41% of patients who were PD-L1 positive.

The differences in progression-free survival were pretty modest, so about 1.7-month advantage overall, 2.5-month advantage in progression-free survival in the PD-L1-positive. If you look at the overall survival, that’s really what got a lot of the press, but this was the part that was somewhat more controversial. And that’s because the statistical plan was that they had to first demonstrate overall survival advantage in the overall intent-to-treat population, and if positive, only then could they test overall survival in the PD-L1-positive.

And so the kicker was that in the overall group of patients, in the intent-to-treat population, overall survival was not improved. But they nonetheless looked at overall survival in the PD-L1-positive subset, and they found about a 9.5-month improvement in median overall survival in that group. But it could not be formally tested. And so this was, I think, some of the complexity of the trial, so it was positive for progression-free survival, no questions about that. More modest benefit. The big benefit looked like overall survival for the PD-L1-positive, but this is where that p-value and formal statistical testing could not be performed, given the statistical plan. And I think this is really why the FDA only gave it an accelerated approval and that additional data would be required.

There was, at ASCO, an update, so the second interim. At the first reporting, this was just an interim analysis for overall survival. The second interim, the results are pretty much the same. The difference now in benefit in the PD-L1-positive is slightly lower at 7 months, but it’s still not the final overall survival analysis from this study. And so I think that there are, of course, questions out there regarding how to make sense of this trial, given the complexity of the statistical design and that the major finding could not be formally tested.

Clinical implications of the IMpassion130 study results

DR LOVE: Do you think that investigators, and you specifically, have accepted this as, at least for the moment, standard of care?

DR TELLI: Yes. I think so. And certainly from a patient perspective, patients are very motivated, if they’re PD-L1-positive, to access this therapy. Absolutely.

DR LOVE: But yet the thing that’s motivating people the most is survival, which is really a nice bump. It kind of reminds me of CLEOPATRA. It was, like, whoa, where did that come from? But yet technically you have this issue with the statistics. Can you talk about how you view that and why it is that you turn around and go, okay, even though it wasn’t the statistical design, I’m treating?

DR TELLI: I mean, it was a large study. And it was very well conducted. The PD-L1-positive subgroup was 41%. I think that it’s just the way that they set up the statistical design, because it was this hierarchical design. If they had flipped it and tested in the PD-L1 population first, and then if positive there to test in the overall population, it would have been different. I think for most of us, we think that the signal is real.

Certainly there are going to be more data coming out in the next 12 to 18 months. There is another trial, IMpassion131, which is looking at paclitaxel/atezolizumab, that will provide some additional data. And then we’re all really curious to see the KEYNOTE-355, which is first-line triple-negative breast cancer study for pembrolizumab with various chemotherapy combinations. That is a huge study at this moment in time that’s outstanding. And so I think as we get these data in, things will become a lot clearer in this space.

Perspective on progression-free and overall survival outcomes with atezolizumab/nab paclitaxel

DR LOVE: Yes, I want to hear a little bit also about what else we’re going to be hearing from in terms of pembrolizumab. But just sticking with this first, any thoughts — occasionally in oncology you see, I mean, CLEOPATRA was an example where you see a much greater benefit in survival than progression-free survival. People always come up with theories why. Any hypotheses in this particular situation, if it’s really true, why you see such a disproportionately positive effect on survival?

DR TELLI: Yes, I mean, I think that this has certainly been an observation in other tumor types with immuno-oncology drugs that some of the big impact — there may not be much difference at all with the progression-free survival but that there can be this tail on the curve, where there’s a group of patients who really sustain long-term disease control. And I think we were all hoping in breast cancer that when we looked at those curves initially from IMpassion130 that we were going to see a nice tail on the curve. Unfortunately, we didn’t see the shape of curve that we would typically have come to recognize with these agents in melanoma and lung cancer and things like that. But that’s really been the idea that there’s going to be a group of patients that really have much more sustained long-term benefit.

DR LOVE: And certainly in other tumors, for example, non-small cell, a big part of the motivation is that the patient’s going to end up on that tail of the curve. They’re going to have a prolonged benefit, maybe unmaintained in some situations. Of course here it’s a little bit complicated because you have chemo onboard also, but, I mean, do you see patients — were patients seen in the trial — was there enough follow-up to know in terms of what happens 2 or 3 years later? Are there more people alive and without progression in the checkpoint arm?

DR TELLI: I think that the numbers are small.

Clinical experience with atezolizumab and nab paclitaxel as first-line therapy for advanced TNBC

DR TELLI: In my experience, I do have 1 patient who was on IMpassion130 who did extremely well. Once she was unblinded — when the results were available, we found out she did, in fact, receive atezolizumab. And she has actually been off all therapy for the last year.

DR LOVE: Wow.

DR TELLI: I saw her at her 12-month period off after treatment a couple months ago. She’s now beyond that. And so it’s those types of patients, right? We wish there were more of them, but I think there is a group of patients like her who can get sustained long-term benefit.

DR LOVE: Just out of curiosity, where was her disease? And did she have a complete clinical response?

DR TELLI: She did. She had a very early, great clinical response. She had metastatic disease in the lung and lymph nodes. And certainly as we are learning more, there are definitely patients, based on their pattern of metastatic spread, that have been associated with having better response to this therapy. Patients with lymph node-only disease, lung disease, liver disease, for example, doing worse with these agents. And so I think that she did kind of fit that picture of a patient in the first line with a high likelihood of benefit.

DR LOVE: I assume she was tested for MSI and was negative? Or stable?

DR TELLI: Yes.

DR LOVE: Have you seen any patients with MSI-high disease, incidentally?

DR TELLI: There are a few here and there that we see in breast cancer. Very uncommon. Not a common indication for us to give pembrolizumab for that in breast cancer.

DR LOVE: Getting back to your patient who had the 5% PD-L1 staining, you described her clinical situation. She goes on paclitaxel/atezo — actually, it looks like she was on trial here. She was on the trial. I don’t know, do you know what she got?

DR TELLI: As of this point, we don’t. But I think that this is the type of patient — there wasn’t anything about her toxicity that would have suggested she did, in fact, receive atezolizumab. But I think she’s the type of patient — even with atezolizumab, there are certain factors about her that would predict for greater benefit.

In IMpassion130, patients who had never received prior therapy, which was actually quite a significant chunk of patients enrolled to that trial — about 37% — those patients definitely did better. Patients with no prior chemotherapy, so the de novo patient. But patients with liver metastasis, for example, did worse. They had less benefit with atezolizumab.

I think for her, she has some favorable prognostic factors. Her tumor was PD-L1-positive. She was de novo. But certainly the adverse predictor was the liver metastasis. She also had an elevated LDH. It’s also been associated with less benefit with these agents.

DR LOVE: And I see that even though she was stable for a while, then she had disease progression. What’s her current situation?

DR TELLI: Yes, so she had continued to have disease progression with subsequent lines of therapy. And so she, unfortunately, recently passed away.

DR LOVE: What happened to her breast? You said she had very locally advanced disease. Did it get worse? Did it get better? What happened?

DR TELLI: Yes, so as she on paclitaxel and either atezolizumab or placebo, she had very minimal regression of all of her lesions. She continued to have a very large and palpable breast tumor. It never really became, until the very end, a locally advanced problem. I mean, ultimately it did start to break through the skin. That was after a few additional lines of treatment. But she really never had any objective response. She had stable disease as best response and then progressed. And that was, unfortunately, a similar story with her subsequent lines of treatment. There’s certainly patients like this that they just have very refractory disease, both to chemotherapy and other novel agents.

Tolerability of atezolizumab in the IMpassion130 trial

DR LOVE: You mention toxicity and autoimmune toxicity that you see with checkpoint inhibitors. What was seen in the IMpassion study? And what do we know in general about checkpoint inhibitors in terms of toxicity in women with breast cancer?

DR TELLI: Yes, I think with IMpassion130 we were all pretty relieved to see that overall, the immunologic toxicity seemed to be similar to what would have been expected. No unexpected signals. Primarily thyroid dysfunction. There are the rare cases of more significant toxicities, hepatitis, pneumonitis, but those were really quite uncommon. And there was a question about adrenal insufficiency, because some earlier studies in the early-stage setting with pembrolizumab suggested in breast cancer that there was maybe a little bit more adrenal insufficiency. But that was not observed here.

I think that from a toxicity standpoint, there were really no new signals or reasons for concern. Most patients are able to — even with these toxicities — are able to maintain therapy. Certainly some of them that are more severe, higher grade, discontinuation is recommended. But I think overall patients have done well with atezolizumab with breast cancer.

Activity of single-agent pembrolizumab in patients with TNBC

DR LOVE: You talked about the various studies looking at atezolizumab, referred to some things going on with pembro. Right now what do we know about other checkpoint inhibitors, including pembro, in metastatic breast cancer?

DR TELLI: Yes, so I think the big study that we’re waiting for is the KEYNOTE-355. Because that is a similar patient population to IMpassion130, first-line triple-negative breast cancer. It’s a large study. Patients were randomized to pembrolizumab or placebo, but the investigator could select the chemotherapy backbone. It was either paclitaxel, nab paclitaxel or gemcitabine/carboplatin. I think that’s going to be an important trial to give us a little more data looking at some of these other chemotherapy backbones.

We will see at the ESMO meeting the results, the topline results, from the other big Phase III in metastatic triple-negative breast cancer with pembrolizumab, and that was specifically for the second- and third-line patient population. That’s the KEYNOTE-119 trial looking at pembrolizumab as monotherapy versus chemotherapy of physician’s choice. And so there was a press release that came out a few months ago saying that the primary endpoint was not met from KEYNOTE-119, but I think we’ll see the full data very soon to get a better sense of how monotherapy performed in a later-line pretreated triple-negative breast cancer setting.

DR LOVE: Although I’m not aware of any superexciting data in general in monotherapy with checkpoint inhibitors in breast cancer.

DR TELLI: Yes.

DR LOVE: I guess it’s not a big surprise.

DR TELLI: Yes, I mean, I think at the time this study was designed, monotherapy in a later-line setting was thought to be potentially viable. I think as more and more data came in in the Phase I and II setting, it became increasingly clear that probably monotherapy was never going to make it. I don’t think we’re really surprised by the results of this study. But it will be interesting to see the data, see if there are any signals. That’s one question out there: Is there a group of patients where potentially monotherapy is viable? And so we’re curious to see the results presented in full for that one.

Emerging data with pembrolizumab in combination with chemotherapy for TNBC in the (neo)adjuvant setting

DR LOVE: I know also there’s a big neoadjuvant study that’s going to be presented at ESMO. What do we know about that? And what do we know in general right now? I think it’s been looked at in I-SPY about checkpoint inhibitors in the neoadjuvant and adjuvant setting.

DR TELLI: Yes, so it’s an incredibly kind of exciting space. Early on, the first study that we saw was a Phase II arm of the I-SPY 2 trial that looked at weekly paclitaxel and pembrolizumab followed by AC given without the pembrolizumab followed by surgery. And what they reported in that — it was a small group of patients, but it was a really significant finding — that the addition of pembrolizumab given concurrently with the taxane phase of treatment led to a tripling of the estimated pathologic complete response rate. And what was also incredibly interesting with that study is that a similar trend was seen in high-risk hormone receptor-positive patients as well.

And so KEYNOTE-522 was the follow-on Phase III study looking at pembrolizumab in an early triple-negative breast cancer setting. There is a parallel study ongoing now, KEYNOTE-756, looking at that chemotherapy with or without pembrolizumab in hormone receptor-positive disease. That’s going to be a while until we get those results.

But what we’re really excited about is the 522 result that will be reported at ESMO. This is the first report of a Phase III trial looking at a checkpoint inhibitor in triple-negative early-stage breast cancer. The press release did come out saying that it met its primary endpoint. This was a large study, over 1,100 patients, and all patients receive a chemotherapy backbone of paclitaxel and carboplatin followed by AC. Somewhat of an interesting choice. They decided to go with the 4-drug backbone including carboplatin.

And then patients were randomized to receive pembrolizumab or placebo, and the pembrolizumab was given all the way through the neoadjuvant therapy, so through the taxane/carboplatin portion, through the AC portion and then, even after surgery, it is continued to complete a full year of treatment. There’s the neoadjuvant and then the adjuvant phase of the study.

And so it’s really exciting that it met the primary endpoint for pathologic complete response rate improvement. It was powered, as well, for event-free survival. That’s really important. And so those data will be coming subsequently. But I think we’re all really excited to look at this.

Evaluation of anti-PD-L1 antibodies in the neoadjuvant setting

DR TELLI: In terms of other things that have been reported, we were all excited from the I-SPY pembrolizumab data, but then the German Breast Group, they reported some Phase II data from GeparNuevo trial. This was a Phase II, about 175 patients, and these patients received durvalumab with nab paclitaxel, followed by EC versus placebo with paclitaxel/EC. And in that trial, they only found with durvalumab overall a 9% improvement in the pathologic complete response rate with durvalumab, the PD-L1 inhibitor, but it was not significant.

I think we’re very enthusiastic about the potential in early-stage disease, somewhat disappointed by that study with durvalumab. But it looks like KEYNOTE-522 hit the mark, and so be very excited to get those data out there.

DR LOVE: It’s going to be really interesting, though, because if it does show a significant improvement in path CR, I guess people might think about it. Maybe it’ll get approved. But then are you going to give the year of maintenance?

DR TELLI: Yes.

DR LOVE: Because you don’t really know what the impact of that’s going to be.

DR TELLI: Not at all. The trials, they do have a large neoadjuvant Phase III, a couple of them ongoing. There’s the NSABP-B-59 study, and then the one that will report sooner will be IMpassion031. But again, it is looking at treatment given in the neoadjuvant phase, as well as an adjuvant phase after surgery. And so yes, at this point in time, we don’t know what the impact of that will be. But that is the current design of these early trials.

DR LOVE: That’s really wild. What’s the NSABP trial design?

DR TELLI: It’s very similar to the IMpassion031. It is a taxane chemotherapy followed by AC, and it is with or without atezolizumab.

DR LOVE: Hmm.

Case: A woman in her late 20s with a family history of breast cancer and a germline BRCA mutation receives talazoparib and avelumab on a clinical trial for metastatic TNBC

DR LOVE: Let’s move on now and talk about a big story in triple-negative — in breast cancer in general — which is PARP inhibitors and the whole issues with DNA repair findings. And maybe to get into it we can hear about your 28-year-old lady.

DR TELLI: Yes. This is really a heartbreaking case. A 28-year-old young woman who initially presented with a very locally advanced left breast cancer. She did have a family history of breast cancer and did have appropriate genetic testing that showed a deleterious germline BRCA1 mutation. She was treated with neoadjuvant chemotherapy. She received standard AC, dose dense, followed by dose-dense paclitaxel.

She then went to surgery and unfortunately had significant residual disease remaining in the breast and in the lymph nodes. She had about 13, 14 lymph nodes involved with cancer following AC/paclitaxel. And so she had mastectomy, was recommended for additional adjuvant therapy. She went on and received a platinum doublet. In this case, she received a somewhat less commonly used regimen of eribulin with carboplatin, based on some data from Virginia Kaklamani’s Phase II trial, and completed her postmastectomy chest wall and nodal radiation. And then, unfortunately, about a year and a half later presented just on a routine exam with a palpable supraclavicular lymph node.

DR LOVE: Just to go back for one second to her first therapy — and I don’t know if you were the one who delivered it or you saw her afterwards.

DR TELLI: I saw her after.

Efficacy and tolerability of platinum regimens for patients with TNBC

DR LOVE: I’m kind of curious in general how you would approach today, for example, a patient with a known BRCA germline mutation with locally advanced triple-negative disease? Would you be more likely to use a platinum? Would you even think about a PARP inhibitor?

DR TELLI: Yes, so, I mean, I think over the last several years, there’s been a lot of data trying to really give us an answer for what is the role of platinum in early-stage triple-negative breast cancer? We had a couple of the early randomized Phase II, the German Breast Group GeparSixto trial and then the CALGB-40603. They both showed an improvement in pathologic complete response rate with carboplatin, somewhere in and around the 15% range. They weren’t powered for long-term outcomes, but with GeparSixto they did show an improvement of about 10% in event-free survival at 3 years. That was not observed in the CALGB study. There was a lot of question around there.

We then, this past year, had published a really important trial, which was BrighTNess. This was a Phase III neoadjuvant trial, and it looked at 3 different arms for early-stage triple-negative breast cancer. Paclitaxel with carboplatin and with the PARP inhibitor veliparib, that was the first arm. Paclitaxel/carboplatin, and then the third arm was paclitaxel alone. Patients all received AC prior to their surgery. And what the BrighTNess study showed, it was really negative for the primary endpoint. Veliparib/carboplatin did not improve the PCR rate significantly compared to paclitaxel/carboplatin. But it was a really strong study to support platinum, because in both of the platinum-containing arms, the PCR rates were greater than 20% increased for those patients who received platinum.

We are still waiting for the long-term event-free survival data there. There is a big adjuvant study ongoing, again through NRG, that’s NRG-BR003, looking at standard taxane/AC with or without carboplatin. That continues to enroll slowly but continues to enroll.

And so at this point in time, we don’t yet know who are the patients that should get carboplatin. Certainly it adds to toxicity, particularly hematologic toxicities. There is less ability to maintain dose schedule and intensity, dose reductions, dose delays, things like that. And so it’s been a complicated area. I think for me, my practice currently is that I typically start with AC. And I assess how the patient, after their 4 cycles of AC, what kind of clinical response they’ve had. And if I feel that there is certainly residual tumor there, at that point in time, I add carboplatin to the paclitaxel.

There are some groups that are giving carboplatin to all patients, but I think that it’s a question. Who are the patients? There’re plenty of patients who are going to do well with standard therapy. Who are the patients that we need to escalate the therapy in? And should we be doing it across the board? Should we be doing it selectively? I think talking to a lot of peers, I think a lot of us are giving AC, assessing where the patient is in terms of tumor response and then making a decision at that point whether to add platinum or not.

DR LOVE: Now again, are you talking about people with BRCA germline also?

DR TELLI: Yes, so that’s what I was going to get into. I think initially we all thought that yes, the addition of platinum was definitely going to benefit the BRCA carriers the most. But the data just didn’t turn out to show that. It was first reported in the GeparSixto trial, because they did do BRCA testing on all the patients in that trial, and they found that if you looked at the BRCA mutation carriers, that group specifically, the patients who just had the anthracycline/taxane backbone chemotherapy with bevacizumab compared to paclitaxel/anthracycline and carboplatin with the bevacizumab, that there was really no improvement with platinum in the BRCA mutation-positive subset. It was all in the BRCA-negative, which was really very interesting.

And then that was further corroborated in BrighTNess. BrighTNess was a larger Phase III trial. All patients had to have germline BRCA mutation testing up front. The randomization was stratified on that basis. And so there was a modest improvement in pathologic complete response rate in the BRCA carriers with platinum, but it was not significant. And so it really appears at this time that, which is very interesting, that the patients getting the most benefit from platinum in the neoadjuvant setting are the BRCA-negative patients.

DR LOVE: That is really interesting.

DR TELLI: Yes.

DR LOVE: Getting back to your patient, when she presented 18 months later, can you talk about what was found in terms of her workup, particularly her PD-L1 assay?

DR TELLI: Yes, so she just came in for a routine visit, had a very palpable left supraclavicular lymph node. And so this led to systemic restaging. She had a CT/PET, and this unfortunately revealed metastasis, nodal. She had bone metastasis. She had lung metastasis at the time. She did go ahead and have a biopsy of one of the lymph nodes. She had a contralateral supraclavicular lymph node as well that was biopsied. Again, it was triple-negative breast cancer confirmed, and she was PD-L1-positive, using, again, the VENTANA SP142 antibody assay. And it was positive at 1%.

DR LOVE: And again, just from the point of view of PD-L1, theoretically for you, 1% is kind of the same as 5%?

DR TELLI: Yes, so, I mean, in the study, that was the cutoff used, 1% in the immune cell compartment. And the majority of the patients, like we talked about earlier, fell into the 1% to 4% range.

DR LOVE: Right.

Results from the OlympiAD (olaparib) and EMBRACA (talazoparib) trials for patients with HER2-negative advanced breast cancer and a germline BRCA mutation

DR LOVE: This patient really presents an interesting dilemma. I know she ended up going on a trial, and I want to hear about that trial, because it was really interesting, but outside a trial setting, how do you approach a patient like this who’s got a PD-L1 that’s positive but also has a BRCA germline mutation?

DR TELLI: Yes, so I think this is a real question in the field right now. What is the optimal first-line therapy for a patient like this? If you look at the Phase III clinical trials of the PARP inhibitors, we have OlympiAD, which evaluated olaparib, and, importantly, that was compared to treatment of physician’s choice chemotherapy in the first- through third-line setting. And then we have data from EMBRACA, which evaluated talazoparib monotherapy versus treatment of physician’s choice in a first- through fourth-line population.

We don’t have any pure first-line PARP inhibitor data. What we do have, though, is some data looking at overall survival from OlympiAD. So this was a subsequent presentation and publication. Overall in that trial, there was about a 3-month improvement in median progression-free survival favoring olaparib compared to chemotherapy single agent. No difference in overall survival at the end of the day, but in a subset analysis, if you looked at just the first-line patients, the hazard ratio for survival there was 0.51. It certainly appeared to be that if you looked only at the first-line patients with a BRCA mutation that they did appear to have an overall survival advantage.

We have that, which is a subset analysis, and then we have this overall survival data right now from IMpassion130, which is complicated for all of the reasons we discussed earlier, because it couldn’t be formally statistically tested. And so I think, from my perspective, either approach is very viable. And a lot of it is always a conversation with the patient. PARP inhibitor therapy, the good thing about it is that it does result, as a monotherapy, in incredibly high response rate. Both EMBRACA and OlympiAD showed response rates of about 60% objective response rates, and that was double what was seen with single-agent chemotherapy.

She had visceral disease. I think that a PARP inhibitor alone would have a very high likelihood of buying her a good likelihood of objective response. The issue is that the progression-free survival is not — it’s from the 2 studies, the medians, again, in that first- through third- and first- through fourth-line population, 7 to 8.6 months. And so certainly we know resistance can develop with PARP inhibitors, and certainly there are patients who can be on a PARP inhibitor very long term and a monotherapy and have good disease control. But there are certainly more commonly patients who have a great initial benefit and then go on and have progression of disease.

It’s a complicated thing, and I think that PARP inhibitor monotherapy, it’s oral, from a toxicity/quality-of-life standpoint, much superior compared to atezolizumab/nab paclitaxel. I think for many patients it’s a discussion at this point in time until we have more data.

DR LOVE: They say you’re not supposed to compare indirectly between trials, but everybody does it.

DR TELLI: Yes.

Activity and side-effect profile of olaparib versus talazoparib

DR LOVE: What’s your perspective on talazoparib versus olaparib in terms of the data we have right now, both in terms of efficacy and also as well as tolerability?

DR TELLI: Yes, so I think we’re still waiting for the final overall survival results from EMBRACA. At this point in time, OlympiAD, the first report was the final progression-free survival data. The subsequent report was the final overall survival. We don’t yet have overall survival data for talazoparib. At this point in time, if you look at the 2 trials, both of them showed a median of about a 3-month improvement in progression-free survival. And so I think at this point in time either of them, from an efficacy standpoint, they’re very reasonable choices. I think they’re quite comparable. If, of course, EMBRACA comes out with an overall survival advantage, that will change things. But at this point in time I think that they’re pretty comparable.

Certainly from a toxicity standpoint, there is more hematologic toxicity that was reported with talazoparib in EMBRACA compared to what was reported with olaparib in the OlympiAD trial. Certainly higher rates of Grade 3/4 anemia, thrombocytopenia and neutropenia compared to what was reported with olaparib. I think that typically these are well controlled by just using dose reductions, but there is more hematologic toxicity with the talazoparib 1-mg dosing.

DR LOVE: I’m trying to remember whether I’ve heard about — is talazoparib — has alopecia been seen with that? Am I remembering that correctly?

DR TELLI: Yes, and that’s another thing. Typically Grade 1 alopecia, so just hair thinning. But that was reported in EMBRACA in about 25% of the patients. And so that has not really been seen, or at least reported, with a lot of the other PARP inhibitors. That is a real thing. It tends to be pretty mild. Patients notice it. But it’s a hair thinning. We don’t really see complete alopecia, Grade 2.

DR LOVE: Clinically for you, is it a coin flip? Or do you lean towards one versus the other?

DR TELLI: I think that there are always different issues. Talazoparib is just once a day, 1 pill. Olaparib’s 2 pills twice a day. Sometimes there are insurance issues, right, but I think overall they’re pretty comparable drugs.

Ongoing investigation of the combination of anti-PD-1/PD-L1 antibodies and PARP inhibitors

DR LOVE: Actually, this lady ended up going on a trial that combines the best of both worlds, so to speak. We were talking about PD-L1 levels, et cetera. She goes on a trial of talazoparib and avelumab.

DR TELLI: Yes. I got an easy out with that one.

DR LOVE: Yes.

DR TELLI: I didn’t have to make a decision. But there are a lot of studies ongoing right now looking at PARP inhibitors and IO combinations, not only in breast cancer but in many different tumor types. I think really the hope is that potentially adding a PD-1 or PD-L1 inhibitor could help to sustain the durability of response. Because like we talked about earlier, what we have right now in breast cancer is progression-free survivals on the order of 7 to 8.5 months. And so the question is, if we add a checkpoint inhibitor to a PARP inhibitor, can we potentially see an improvement in that durability and the length of response? And so I think that there are a number of studies ongoing. TOPACIO was a Phase II trial that looked at niraparib/pembrolizumab, recently reported, but there are a number of others that are currently ongoing.

DR LOVE: It’s interesting too, thinking about getting these two classes of drugs that are fairly new to us and trying to sort out side effects. What happened with this lady?

DR TELLI: Yes, overall she’s done really great. She remains on treatment. At her first reassessment she had an objective response. I think probably a lot of that is being driven by the PARP inhibitor, just given the very high objective response rates we see with PARP inhibitor therapy in this kind of patient. She did have to be dose reduced for a Grade 3 anemia from 1 mg of talazoparib down to 0.75, and she continues to do really well.

DR LOVE: And just doing that dose adjustment was all you needed to do in terms of the anemia?

DR TELLI: Mm-hmm.

DR LOVE: What about GI problems? Do you hear about that with PARP inhibitors? I don’t know if it’s more with olaparib, but did she have any GI problems? And what have you seen in terms of that?

DR TELLI: Yes, I think we definitely do see it, and we always warn the patients up front that nausea is possible. We give them antiemetics. What’s really interesting about PARP inhibitors, and I think looking at multiple different drugs, it’s a similar phenomena. If the patient has nausea, it tends to be very early, and then it just self resolves with time. I think the typical patient would report maybe some low-grade nausea for a week or 2 when they first start the drug that then dissipates. I’ve had a rare patient who has reported it up to 2 months, for the first 2 months, and then it resolves. I think when patients get it it’s just very transient, and then typically they don’t need any antiemetic therapy or anything moving forward.

Sequencing PARP inhibitors for patients with PD-L1-negative breast cancer

DR LOVE: What do we know about PARP inhibitors, specifically in terms of triple-negative breast cancer in a patient whose PD-L1 is negative, so we kind of move checkpoint inhibitors off the grid for a second. Ideally, when would you like to use a PARP inhibitor in metastatic triple-negative disease?

DR TELLI: Yes, I will, if the immuno-oncology agents are not an option because PD-L1 is negative, then I typically like to use it up front. I think when you think about the overall combination of efficacy, high response rate, tolerability, it’s easy. It’s oral. Patients, I think, find that it’s a difficult transition, always, moving into therapy for newly metastatic disease. And I find it’s a much smoother entry into that with an oral drug compared to going back, for many of them, to IV chemotherapy. I tend to use it early.

DR LOVE: How about metastatic ER-positive, HER2-negative disease? When are you starting to think about a PARP inhibitor in a patient with a BRCA germline mutation, for example?

DR TELLI: Yes, so I think the appropriate time there is once you’ve really exhausted all of your endocrine therapy options, including CDK4/6, alpelisib if the patient is a candidate, when you’re starting to think about chemotherapy, that is the point in time to consider it. I wouldn’t, in an ER-positive, HER2-negative patient, with a new diagnosis of metastatic breast cancer to bone, I would certainly lead with endocrine-based therapy, and I wouldn’t go there, to a PARP inhibitor, in the first line.

DR LOVE: And what about in HER2-positive disease? It’s really not approved, but do you use it? Are we going to see some data on that?

DR TELLI: Yes. I mean, I think if you look at germline BRCA1 and 2 mutations, it’s very interesting, but we almost never see HER2-positive breast cancer associated with them. I have, over the years, seen a few patients. They’re out there. But the indication is in HER2-negative disease. I would have absolutely no issue at all giving it to a HER2-positive breast cancer patient, and I would advocate for that if the patient happened to be HER2-positive.

Use of PARP inhibitors for patients with breast cancer and mutations in DNA damage repair genes beyond BRCA

DR LOVE: It seems pretty clear cut. Your algorithms are pretty clear cut in terms of patients with BRCA germline. What about other germline mutations? And what about somatic mutations, including BRCA?

DR TELLI: Yes, and so I think the field in general is starting to amass more and more data around some of the beyond BRCA genes. There are a number of other genes beyond BRCA1 and 2 that are important in DNA repair, specifically homologous recombination. And the big question is, if the patient has a germline mutation in one of these genes, would they similarly benefit from a PARP inhibitor? And I think all of these genes are not the same.

But certainly there is very strong reason to believe that there would be benefit in that group, and certainly emerging clinical data suggesting that some of these patients with PALB2 mutations, RAD50, RAD51, ATM potentially, mutations, might benefit from PARP monotherapy. This has been an area of interest of mine. We reported some data at ASCO looking at talazoparib beyond BRCA, showing really the strongest signal in patients with germline PALB2. But potentially these could also be options for patients with somatic mutations.

And so there are a number of studies ongoing trying to look into this. Many of them actually outside of breast cancer, so in prostate and pancreas, for example, a lot of action right now in that space. But I think that with the somatic BRCA1 or 2 mutation, my experience has been that they behave pretty similarly to germline BRCA and that certainly the data would suggest that as well. In ovarian cancer we have some data suggesting platinum sensitivity from the TNT trial for patients with nongermline but somatic BRCA mutation. I would absolutely be in favor of treatment with a PARP inhibitor in a somatic BRCA1 or 2 mutation-containing tumor.

DR LOVE: You participated in a CME meeting we did at AACR where we actually had you, we had an ovarian cancer investigator, prostate and pancreas, which you never would imagine would all be on the same stage. And obviously PARP inhibitors, as you mention, is really in play for all those tumors at this point. One of the things we were talking about is, is PARP or DNA repair damage deficiency the new MSI? It seems like maybe it’s not going to work that way, because, like, ovary is very different than breast. But do you see eventually this coming down into a nontumor-specific algorithm?

DR TELLI: I absolutely do. I think where we are at this point in time, there are a number of putative genes that you could consider including on a DNA damage response gene panel, for example. The issue is that ATM, for example, is not going to be the same as PALB2. And so these genes are going to — there are many different panels. All the different trials are looking at different numbers and types of genes that they’re calling HRR or DDR gene panels.

And so I think with time it will become clearer. Certainly I think there’s a more consistent signal across tumor type that PALB2 mutations, tumors associated with that mutation are highly susceptible to PARP inhibition. And there initially was a lot of excitement about ATM, but increasingly data coming out suggesting that PARP inhibitor monotherapy in that group of patients may not be as clearly beneficial as initially thought.

Tolerability of PARP inhibitors for patients with germline BRCA mutations

DR LOVE: One final question. I was just flashing back to that AACR symposium, and one thing that Katie Moore, who was there in terms of representing ovary, brought up, that I hadn’t really thought about, but she said she got a lot of questions about something I hadn’t thought about, which is, do people with germline mutations have more toxicity from PARP inhibitors? And I know the answer is no, but can you explain the biology of that?

DR TELLI: Yes, I mean, I think that there was a lot of question about looking at chemotherapy, for example. Would BRCA mutation carriers potentially be more susceptible to myelosuppression, for example, with chemotherapy? And would the same potentially be the case with PARP inhibitors?

I think that if you look at the germline BRCA1 or 2 mutation situation, of course, these patients are born with 1 faulty copy of, for example, BRCA1, and 1 normal copy. And so in all of their normal cells, including normal hematopoietic cells, they should have a functional BRCA1 gene. And it’s only in the tumor when they have loss of the normal copy, and then they have a tumor that has really deficient BRCA function. That’s when that whole metastatic transformation and progression takes off. And so from a toxicity standpoint, because the hematopoietic cells really do have intact functional BRCA, they are not particularly exquisitely susceptible to some of these toxicities from platinums, PARP inhibitors, other DNA-damaging agents.

DR LOVE: Subsequent to this interview, a number of relevant research results were reported at the 2019 ESMO meeting. The KEYNOTE-522 trial demonstrated the addition of pembrolizumab to platinum-containing neoadjuvant chemotherapy resulted in a significant improvement in pathCR in patients with previously untreated triple-negative breast cancer.

In addition, findings from the KEYNOTE-119 trial showed that while pembrolizumab monotherapy as second- or third-line treatment for metastatic triple-negative breast cancer did not improve overall survival versus chemotherapy in the primary analysis population, there was a trend for improvement in patients with PD-L1-positive tumors.

Finally, intriguing data from a Phase II study were reported showing a significant improvement in overall survival on a novel CDK4/6 inhibitor, trilaciclib, when added to gemcitabine/carboplatin compared to gemcitabine/carboplatin alone for patients with metastatic triple-negative breast cancer.