Design and results of the Phase III KATHERINE study investigating adjuvant T-DM1 versus trastuzumab for patients with residual invasive HER2-positive early breast cancer

DR LOVE: Welcome to Oncology Today: Breast Cancer Edition, a special audio program focused on recently approved and emerging strategies in breast cancer. This is medical oncologist Dr Neil Love.

For this program, I met jointly with Drs Sara Tolaney and Charles Geyer. And to begin Dr Geyer talked about the practice-changing KATHERINE trial first presented at the December 2018 San Antonio Breast Cancer Symposium, evaluating the use of T-DM1 versus trastuzumab in the adjuvant HER2-positive setting after patients received neoadjuvant treatment with chemotherapy and anti-HER therapy.

The results from this trial, for which Dr Geyer was the senior author, were simultaneously published in The New England Journal of Medicine.

DR GEYER: KATHERINE was a Phase III study that simply looked at the high-risk patients defined by having residual disease after receiving standard neoadjuvant trastuzumab with or without a second targeted agent and chemotherapy. We had pretty broad descriptive criteria for what was an allowable regimen. Just any standard neoadjuvant regimen would qualify a patient. And if they had any residual disease in the breast or found in their axillary lymph nodes, they were eligible for KATHERINE.

The randomization was to complete a standard course of postoperative trastuzumab or T-DM1, for 14 doses, basically. It was a bit beyond a year, but we did that so the KATHERINE therapy was consistent across all regimens received initially.

DR LOVE: Can you talk a little bit about the data that were presented in San Antonio and then published in terms of roughly the number of patients and what was seen?

DR GEYER: Sure. We had projected a sample size of just under 1,500 patients, and the study was powered to detect about a 25% to 30% reduction in hazard for invasive disease-free survival, which was the primary endpoint. We did plan an interim analysis when we hit about two thirds of the 384 events that were required for the definitive analysis. And it was that data that widely crossed the reporting boundary, because we had a hazard ratio of 0.50 for IDFS, highly statistically significant. The 3-year IDFS improved from 77% up to 88.3% with the substitution of the T-DM1, so a very, very positive early result of the trial.

DR LOVE: So, of course, one of the issues with the trial, Chuck, was, I guess, very few of these people got pertuzumab.

DR GEYER: About 20% of the patients received a second HER2-targeted agent, and the vast majority of those were pertuzumab. And that was largely from the accrual in the United States at the time. So we did have a good representation of patients who received dual HER2-targeted neoadjuvant therapy in the trial.

DR LOVE: So Sara, any comments about your reaction when you saw these data?

DR TOLANEY: I think it was very impressive. It showed that giving T-DM1 for 14 cycles compared to trastuzumab really results in decreasing the invasive disease-free survival events by half. So it had a hazard ratio of 0.5, an absolute difference of 11%. So I think this is incredibly meaningful and a practice-changing study. So I do think that for our patients who fail to achieve a pathologic complete response to standard HER2 preoperative therapy, they should get 14 cycles of adjuvant T-DM1.

Side effects and tolerability of T-DM1 compared to trastuzumab

DR LOVE: Chuck, any thoughts or speculations that you have or that you’ve heard about why there was such a dramatic advantage? I mean, these patients had gotten chemotherapy already and for that matter anti-HER therapy, to be such a great advantage.

DR GEYER: I have not heard a specific mechanistic reason. I mean, I think part of the appeal of the KATHERINE design was we were using it in a situation where we had seen its greatest activity, namely patients who had had prior exposure to trastuzumab, to chemotherapy and still had progressive disease. So we were seeing it there.

I don’t know the reason why it seems to be more effective almost if tumors had been primed by prior therapy, what mechanism that might exist for that. I have no idea, and I’ve really not heard anybody come up with one. I mean, the idea has always been with the immunoconjugates that you might have so-called bystander effect, things like that that might be part of the efficacy, but why there would be this very marked hazard of 0.5 — I mean, obviously the activity was beyond what we were, in a sense, hoping we might see.

DR LOVE: Chuck, could you also talk a little bit about what was observed in terms of toxicity in the 2 arms?

DR GEYER: Of course there’s different ways to look at toxicity. The most simplest is to say, what was the percentage of patients who received all planned 14 cycles? That was 80% with trastuzumab, 70% with T-DM1. We had 18% of patients who discontinued the T-DM1 early, did so due to adverse events. When you look at the adverse events, though, it was really an array of different toxicities — thrombocytopenia, LFT elevations were the most common, but the thrombocytopenia was 4% of the total.

So it really was, I think, more of the fact that patients had been on therapy for a while. I do think T-DM1 is chemotherapy. I know some people call it nonchemotherapy. I think it’s chemotherapy. And these people had been on therapy for a while. So the difference between the 2 arms to me is really not surprising given the nature of the drug and how well tolerated we all know trastuzumab is generally when you get out to that point.

DR LOVE: Sara, you’ve had experience using T-DM1 in the early disease setting. You have your — I think it’s the ATTEMPT trial. Maybe you can talk a little bit about that. And also what your experience is using T-DM1 in the adjuvant setting.

DR TOLANEY: Sure. So we do have a trial that is also exploring T-DM1 in the adjuvant setting, which, as you noted, is the ATTEMPT trial. This is a trial, however, that is a very different population than KATHERINE. This is a Stage I HER2-positive population. And so in this trial, patients were randomized to receive a year of T-DM1 — so in this case it was 17 cycles — or to receive the TH regimen as was administered on the APT trial, so that’s 12 weeks of paclitaxel and trastuzumab followed by 9 months of trastuzumab.

The trial has completed accrual, and we’re actually hoping that the data will be mature enough to present at San Antonio this year. But certainly from that trial I certainly have had a lot of experience treating patients, and I think, unlike KATHERINE, these are patients who didn’t have prior chemotherapy exposure. And I’ll say anecdotally, my experience is very similar to what Chuck was describing, that T-DM1 is still chemotherapy. Even though it’s a very well-tolerated drug, you do see patients who develop some peripheral neuropathy. And patients do get LFT elevations, and they do have a little bit of fatigue. So it does have potential side effects, even in an untreated — previously untreated population.

DR LOVE: What about alopecia, Chuck? Is that seen with T-DM1 in this situation?

DR GEYER: Generally, the patients have lost their hair from their initial —

DR LOVE: Right.

DR GEYER: — chemotherapy, so it’s not — wasn’t an issue on KATHERINE. I assume we probably could have tracked it and saw that hair came back a little bit later if you were getting the T-DM1 relative to trastuzumab, but…

DR LOVE: Right. I didn’t think about that.

Clinical implications of the KATHERINE trial

DR LOVE: So Chuck, it seems like people saw these data and of course wanted to apply it immediately. Has that been your thought as you see people outside a trial setting? Any residual disease, and you’re thinking T-DM1? Path CR, no? Is that your approach?

DR GEYER: A very appropriate question, but right away was, “What about the patients with minimal residual disease? Did you do RCB1 scoring?” So forth and so on. And we did not attempt RCB1 scoring, since this was a global study, but we did have a substantial number of patients with really minimal disease — 1 centimeter or less, negative nodes. We actually were concerned as we accrued our patients — “Gee, are we going to have enough events in the patient population coming in?” But for me, and equally striking to the hazard ratio, is how consistent that hazard ratio is across all the subgroups, even the patients with very minimal amounts of residual disease, tumors 1 centimeter or less, with negative nodes, at pathology still had a hazard ratio of 0.6, an absolute improvement in 5%.

We also went and looked at — if you looked at the clinical staging prior to their neoadjuvant therapy, and we had a substantial number of those patients that were found to have residual disease that were in the trial, same thing — we just could not find any of the groups not deriving substantial benefit. So the patients on trastuzumab do really pretty well, but they do even better with T-DM1.

So for me, if you have residual disease it’s certainly worth a trial of therapy to see what the patient’s tolerance is to it and then adjust accordingly, which is what I think medical oncologists do routinely.

DR LOVE: So I don’t know, Chuck, whether anybody ever brings this up, but what about the issue of using in the patients with path CRs. What is your usual estimate of risk for recurrence after path CR, and would you expect it to be cut in half also?

DR GEYER: I think it’s a very interesting question. And the magnitude of the improvement does suggest that there could be patients with PCR who might have further risk reduction, but we have no data on that. I guess 1 group where I would think about it would be a patient who really came in with inflammatory HER2-positive breast cancer that yes, we’re real happy we got the PCR, but we know that those patients are at risk. I guess I have to admit, I might think about possibly converting over to give them T-DM1 based on this data.

DR LOVE: In general, Chuck, before I ask Sara for her thoughts about this, what rough estimate do you have in your own mind as the risk for relapse after path CR? The range.

DR GEYER: I think it varies a lot on where the patient started. And I think we’re in need of really looking at our data sets and seeing how much the stage of disease at presentation impacts on that. I think the patients with smaller node-negative tumors do extremely well. But as you have positive nodes, increasing nodal burden, it drops off. And it’s probably substantial in these larger tumors. It may still be as much as 30%, 35% in patients with larger tumors.

DR LOVE: So Sara, what’s your take on this also, in terms of application outside a trial setting? Are you thinking about T-DM1 in general for all patients with any residual disease? And again, any patients with path CR where you’d want to use it?

DR TOLANEY: I think for now I will be restricting my use for preop adjuvant T-DM1 to the patients who failed to achieve a path CR.

If we look at data from NeoSphere and NeoALTTO, the patients who achieve a path CR have about a 14%, 15% rate of recurrence, but as Chuck very rightly noted, the range is there, depending on what you start out with. I think if you have a very early Stage II patient who achieves a path CR, I actually think the question is, can you deescalate therapy?

So I think what was so ingenious about this trial is that it used personalized treatment. It looked at your response to therapy to guide your adjuvant treatment, and it told us that if you have residual disease, we should escalate therapy and give T-DM1. But on the flip side, what if you achieve a path CR, should we deescalate therapy?

So there is this trial that’s ongoing, the DAPHNe study, which is a pilot study taking patients who have predominately Stage II disease who get preop paclitaxel, trastuzumab and pertuzumab, and if they achieve a path CR they just go on to get antibody-based therapy. So they’re not getting any further chemotherapy. Really trying to show, can you stop giving additional chemotherapy to those patients who do well? So I think we also need to address that question as well is, who can get deescalation and not just escalation?

Case (Dr Geyer): A 53-year-old woman with a locally advanced ER/PR-negative, HER2-positive, ulcerated infiltrating ductal carcinoma (IDC) experiences neuropathy after 6 cycles of neoadjuvant TCHP (docetaxel/carboplatin/trastuzumab/pertuzumab)

DR LOVE: So both of you submitted cases of patients who had residual disease in the pre-KATHERINE era, and I thought we could maybe briefly go through those 2 cases. And maybe also as we go through the case, get a little bit of insight in terms of kind of the numeracy of these cases. And I don’t know whether these particular patients asked you for any numbers, but I’d like to know what you would estimate.

And also, I want to ask you, I notice there’s an interesting difference in the type of chemotherapy that was used in the neoadjuvant setting with Chuck taking, I think, the most common approach with TCHP but with Sara using THP, which is something I hear a lot from people at Dana-Farber. But Chuck, could we just briefly hear about your 53-year-old lady?

DR GEYER: Sure. This is a 53-year-old woman who presented with a locally advanced T4N0 ER/PR-negative, HER2-positive infiltrating ductal carcinoma of the right breast. She had history of poorly controlled hypertension, substance abuse, so there was a decision to avoid anthracyclines and see how she tolerated TCHP. She did really quite well with it, but when she came back for cycle 6, she had developed pretty substantial neuropathy. Given her amount of disease, we wanted to keep going for a little bit longer, so just switched her over to CMF — this is sad, because we were afraid of the anthracyclines — and then went ahead with surgery.

She had a 1-mm focus of residual cancer in a lymphatic channel within the breast and also had a lymph node with some residual cancer in a subcapsular sinus. So clearly had a very marked response to therapy. So she received radiation therapy and completed a year of the dual antibodies. But now I would have given her T-DM1, but she’s still doing well so far.

DR LOVE: How did she do with the trastuzumab/pertuzumab?

DR GEYER: Kept it going to complete a total year of therapy.

DR LOVE: So a couple of questions about her case, Chuck. She had this ulcerated large lesion. Do you think that was from delay in coming in for treatment or just because of rapid natural history?

DR GEYER: I think there were contributions from both sides on this one. I mean, I do think it had developed fairly quickly from her history, but it still took her several months to present for care. So a little bit of both. I don’t think it was totally one way or the other.

Management of the axilla in patients with positive sentinel lymph nodes after neoadjuvant chemotherapy

DR LOVE: And then another issue is, I see that she had radiation therapy to the axilla. Can you talk a little bit about how you manage patients with positive sentinel nodes after neoadjuvant therapy and in what situations you send patients for axillary dissection or your surgeons do axillary dissection?

DR GEYER: That’s something that is always a great deal of discussion at our multidisciplinary boards. I think generally if there is minimal nodal disease, I don’t think this patient had a sentinel node. I think this patient went straight to an axillary dissection. So with a more advanced cancer like this, we will tend to go ahead and do an axillary dissection, not so much the sentinel nodes. But our surgical group is pretty much evidence based. They’ll follow where the trials take them.

We do, though, a lot of times, if it’s a gray area, we default to not operate and radiate as a group. I have noticed that. So I think there’s more of a tendency to try to lessen the surgery and maybe use radiation more in my institution.

DR LOVE: Yes, I saw that she had the 1 lymph node, and I assumed it was sentinel node, but now you tell me she actually had an axillary —

DR GEYER: Yes, it was one of the…

DR LOVE: Right.

So Sara, again, I’m curious about the approach of you and your group in terms of the axilla when there’s involvement after neoadjuvant therapy.

DR TOLANEY: If someone does have a positive sentinel node at the time of surgery after preop therapy, most of our surgeons are moving to axillary dissection. There is an ongoing randomized trial actually trying to address this question about whether or not one should do radiation versus dissection, and that trial is still enrolling. So I think we just don’t have the final answer on the best approach.

DR LOVE: I’m kind of curious, Chuck, and I don’t know whether this patient asked you this, but at the completion of her treatment, given the fact that she had the trastuzumab/pertuzumab, the pertuzumab with chemo, trastuzumab as neoadjuvant therapy, but she had this very advanced presentation, what are you currently estimating, or what did you estimate at the time when she completed therapy were her long-term risk of relapse?

DR GEYER: Oh, I think she’s still going to be 25%, 30%, because she did not PCR, so she is at substantial risk. But that’s a conversation that I don’t have many patients ask me for, and I really don’t think we have data that makes those conversations, I guess, comfortable for me. Because I think that’s — you know in a broad sense, but when there’s so much variability on these locally advanced cancers, we know that as a group they are at much higher risk, but getting more specific, there may be about a 1 in 3, 1 in 4. That’s about as much as I’ll share with the patient.

DR LOVE: I don’t know if this comes up in questions that you get about this, Chuck, but what about the patient who’s halfway through, let’s say, adjuvant trastuzumab/pertuzumab. Would you switch them over? If they completed the year, would you add T-DM1 as opposed to, say, use neratinib? I don’t know if you get these kinds of questions, but any thoughts about it? It seems like it would be a natural question that would come up.

DR GEYER: Yes, I mean, time is passing, but if that was a question immediately there at San Antonio, if somebody is only a month or two into their adjuvant, would you switch? I would. Yes, I would switch. I mean, and at what point would I say it’s too far? I think for me, if it’s only a month or 2 left, I wouldn’t switch it. I sure wouldn’t ask the patient to start another full 14 cycles of therapy. I just think there’s a limit to what people can go through. So I would think to me, it would be a reasonable thing to say, “Okay, if we’ve got 4 months left, let’s switch to T-DM1 for those 4 months,” not “We’re going to give you 14 cycles of T-DM1.”

Case (Dr Tolaney): A 56-year-old woman with a 3.7-cm ER/PR-negative, HER2-positive IDC receives neoadjuvant THP (paclitaxel/trastuzumab/pertuzumab) and adjuvant AC followed by trastuzumab/pertuzumab

DR LOVE: So Sara, let’s hear about your 56-year-old lady. What happened there?

DR TOLANEY: This is a 56-year-old woman who had had an abnormal screening mammogram and was found to have a 3.7-cm mass in her left breast. She did have a palpable mass when examined but did not have any enlarged axillary nodes either by exam or by ultrasound imaging. So she did have a biopsy of the breast mass. It did come back as a high-grade ER-negative, HER2-positive cancer.

She did get staging studies and did not have any metastatic disease. And we did decide to go ahead and give her preoperative therapy. And I had given her weekly paclitaxel with trastuzumab and pertuzumab for 12 weeks, and then she underwent a lumpectomy and sentinel node evaluation.

She did have some residual cancer left, so 6 millimeters of disease. And on sentinel node evaluation did not have any nodal involvement. And because she still had some residual disease, I elected to complete a standard regimen and give her 4 cycles of AC and then complete out the year with trastuzumab and pertuzumab-based therapy.

DR LOVE: So where is she right now? And she’s completed all the therapy?

DR TOLANEY: She has, and she’s doing well and didn’t have any significant toxicities from treatment. No issues with cardiac toxicity or anything like that.

DR LOVE: And again, looking at her clinical situation, what would you estimate at the end of her treatment her risk of relapse is, and presumably that would be cut in half by T-DM1?

DR TOLANEY: So in her case, she only got THP up front, and so looking at, like, the NeoSphere data, which that would represent this patient population, there was about a 14% chance of recurrence even in those patients who had a path CR. Certainly she didachieve a path CR, although she was near. And so my guess is she’s somewhere in that range of 15%, 16% kind of chance of recurrence. She did go on to complete AC and HP. So she may be a little bit better than that. But it’s probably at this point 10% range or something like that.

Selection of (neo)adjuvant treatment for patients with HER2-positive breast cancer

DR LOVE: What’s your thinking, and some of your colleagues also at Dana-Farber’s thinking, in terms of using the paclitaxel/trastuzumab/pertuzumab in the neoadjuvant setting as opposed to TCHP?

DR TOLANEY: I think it really comes from our sense of tolerability. I think docetaxel tends to be a challenging drug to give TCHP. And we have found THP is certainly much more tolerable.

We don’t have any comparative data for THP versus TCHP by any means, but it has been a standard regimen used as part of APHINITY, and so we felt comfortable using it in the preop setting.

I will say, it does make it a little challenging using this regimen now than with making decisions about what to do post-THP if someone fails to achieve a path CR. Certainly one would say, “Okay, we should give them KATHERINE-based therapy,” but one has to also realize that this patient didn’t get a complete standard regimen, and 75% of patients in KATERHINE had received anthracycline-based treatment in the preoperative setting. And if we had just given THP, she wouldn’t have met the general KATHERINE population. The other patients got more TCHP.

So I think it does make it challenging. It does make us question what is the role for anthracycline, particularly now in the era of T-DM1. And so I will say that these are unknown questions. But I just find it very tolerable and achieves almost a 50% rate of path CR, and patients do very well with it.

DR LOVE: So Chuck, I’m curious about your take on this alternate path in the neoadjuvant/adjuvant setting. From your point of view, and if you see somebody as a second opinion who’s had that recommended as a first opinion, is that equivalent, or do you maybe think it’s got some liabilities?

DR GEYER: Oh, I think it’s a reasonable approach to take. For us doing KATHERINE and doing the trials, one of the challenging things is, you really have to decide what you’re studying, and if you want to look at a new therapy, you have to give your patient standard regimens. And so there was really not an option trying it, because in this case we were wanting to identify our patients as having resistant disease to standard regimen, so that’s who we had to use for our trial.

I think in the clinic, things become different, and you can use different approaches. I guess I personally think that the deescalation and shortening the duration of the antibodies is an interesting question, but I have a strong concern, I guess, that that is very much a research question, and I worry that we might be having problems with losing some of our benefit by shortening our antibodies.

I think it’s far more interesting to continue to work to develop our targeted therapies, our nonchemotherapeutic agents, get to a point where we really are being very effective in the large majority of patients and then try and back out the chemotherapy. Because I think the chemotherapy at this point, it probably is excessive and this is what the idea of trying the taxane first is getting at: Let’s see if we can get the patients to a PCR with minimal chemotherapy and then just go with the antibody.

So I think it’s an approach that makes a lot of sense, but I don’t think it’s a standard approach. And I think it would be a challenge to use that approach probably if you were trying to look at a new therapy, such as, okay, if we want to do a trial, maybe looking at tucatinib to address — for me, the thing that the KATHERINE had in it that we still need to look at are the brain met problems. That was the one area where T-DM1 did not perform better than trastuzumab. I don’t think either agent is really helping there. So we still have an unmet need there. So I think that if one wants to think in terms of addressing that, then it’s a little bit more of adding on therapy as a next stage.

But I think both are very appropriate avenues of research. We clearly want to get better at tailoring our therapy and trying to identify the people who really need all of it and those who don’t need it. I think the challenge we have is, we just don’t have really good patient-specific biomarkers that can tell us when we had delivered sufficient therapy, which is what we would ultimately like to have.

DR LOVE: So both of your patients were ER-negative/HER2-positive. Chuck, what about the benefits of T-DM1 in KATHERINE based on ER status?

DR GEYER: Yes, we of course had a greater number of ER-positive patients in KATHERINE, because patients with ER-negative disease have a higher PCR rate, so we had about 72% of our patients were hormone receptor-positive. But they derived the same degree of benefit from T-DM1 as did the ER-negatives in terms of a relative risk reduction. They started out in a better place. When you go through the tables, you can see that with trastuzumab, the ER-positive patients are doing about 10% better than the ER-negatives, but both groups move up to the same degree by switching over to receive T-DM1.

Optimal treatment for early-stage HER2-positive breast cancer

DR LOVE: So Sara, I see that you have this great review article in Cancer, December 1^st, I thought that was interesting, a review article on “Optimal Treatment of Early-Stage HER2-Positive Breast Cancer,” and you had this beautiful algorithm in there —

DR TOLANEY: It’s now destroyed!

DR LOVE: Five days later, it gets blown up.

DR GEYER: It’s just modified!

DR TOLANEY: Totally true!

DR LOVE: I wanted to ask you, because that is one interesting outcome of this, at least it seems like it. It seems like a few years ago when the only way you could get pertuzumab neoadjuvantly, everybody was trying to give it neoadjuvantly, and they had the FDA thing with the 2 centimeters and all that stuff. And then when people saw the APT trial, I saw people pulling back from smaller lesions in terms of neoadjuvant therapy, hoping they’d be able to have negative nodes. Now I’m hearing people say they’re going to be more aggressive, because they want to have the opportunity to give T-DM1. So what is your algorithm going to look like when you send an addendum to that paper?

DR TOLANEY: I do think that the availability of the KATERHINE data certainly does change our algorithm. I do think it also sways us more towards giving preoperative therapy, because we do get that information back about response to therapy to guide our adjuvant treatment. But I also think we have to remember that patients with small Stage I cancers do really well with TH, and we may be overtreating those patients by giving them THP in the preoperative getting or TCHP or ACTHP. And so I think we have to be cautious.

I think we need to take patients who have clinically very small cancers that are node-negative and potentially take those patients to surgery first. That patient who had the 5-mm abnormality on mammogram that came back HER2-positive, whereas that patient who has closer to a 2-cm cancer or greater and has no nodal involvement up front, those patients should get preoperative therapy up front. So I think it should be standard to approach all our patients with 2 centimeters or more of disease or nodal positivity with preop treatment, but I think some consideration should be made to the small HER2-positive cancers to get them to surgery first.

DR GEYER: What about the 1 or 2 centimeters?

I guess just something that I’ve wondered about is — just thinking about this — is the patients with the T1C tumors, clinical Stage T1CN0, what about just doing a sentinel node? And even if it’s node-negative, going ahead and giving them neoadjuvant therapy to see if they really do have the PCR. Because again, we had, like, 180 patients with those small tumors before their neoadjuvant therapy come into KATHERINE who did not have a PCR who came into KATHERINE who benefited. And the thing we don’t have from KATHERINE is information on how many patients had a PCR for all these different types. We’re seeing the patients who didn’t do as well on KATHERINE, so to speak, with the neoadjuvant therapy. Because it still seems like KATHERINE is saying that there are some of those patients with small tumors who might benefit if you can identify them.

DR TOLANEY: I mean, I think it’s a good point. I think, though, if you go back to the APT data, just so few patients recurred. We only had 4 distant metastases with 7 years of follow-up in Stage I disease. And 50% of those patients did have tumors that were over a centimeter.

And so it just does make you wonder, can you do better than a 97.5% recurrence-free interval? And so I do think it’s just hard to know. I do think those larger, 1.7-, 1.8-cm cancers, I think it’s reasonable to give them preop treatment.

Role of the recently FDA-approved subcutaneous formulation of trastuzumab alone or in combination with chemotherapy for patients with HER2-positive BC

DR LOVE: So 1 final thing before we move on to PARP inhibitors. I came across my computer that FDA approval of subQ trastuzumab — we’ve been talking about subQ rituximab for a while with lymphomas and subQ daratumumab with myeloma that’s been out there. I haven’t heard about subQ trastuzumab. Sara, do you know anything about it? Have you used it? Are you going to use it?

DR TOLANEY: I have not used it. They have been using it in Europe for quite some time. So they had done some trials where they looked at using subQ trastuzumab versus IV trastuzumab and found similar efficacy, similar rates of path CRs, similar safety profiles, similar pharmacokinetics. And they’ve also done almost 2,000 patients in a combined adjuvant preop analysis also looking at safety, which looks similar to IV trastuzumab. And I think what’s really interesting is, it seems like patients do prefer subQ trastuzumab. It makes sense — it’s only a couple minutes to get a subQ dose versus a 30-minute trastuzumab infusion. And so even when they did a very nice patient preference trial, they found that almost 90% of patients preferred to get the subQ version compared to IV when they had been exposed to both.

So I do think it’s a really nice addition. Again, I have not used it, though. I think we do have to see, since now the majority of our early-stage patients are getting pertuzumab as well as trastuzumab. I know there is also a formulation of combined subQ trastuzumab/pertuzumab that’s in Phase III testing, and I think that would really be great to be able to have a combined drug to really save patients time.

DR LOVE: Chuck, any comments on this? I was thinking, particularly in the metastatic setting, where patients can be getting trastuzumab for so long, and maybe they’ll be on something oral, et cetera. Any thoughts about whether or not you’d want to use this, Chuck?

DR GEYER: Oh, I’d certainly use it, particularly if I’m not giving it with pertuzumab.
I tend to give the two together a lot, both early and metastatic. But when you’re only doing trastuzumab, it’d be much better for the patient.

I guess a real busy clinic, you could improve your efficiency by giving the trastuzumab subQ and getting more patients treated faster in your office. I think it’ll have uptake, assuming there’s not a reimbursement disincentive for patients in terms of copays or physicians. I mean, it always boils down to how easy is it to get? How much will it be relative to the other drug, those granular clinic things that come into play.

Final results from the OlympiAD trial of olaparib versus standard chemotherapy; results from the EMBRACA trial comparing talazoparib to standard therapy for patients with germline BRCA mutations

DR LOVE: So I’m going to move on and talk a little bit about PARP inhibitors. In a second, Sara, I’m going to ask you to present your 59-year-old woman. But before we do, I listed here a couple of key recent papers related to the approved PARP inhibitors, olaparib and talazoparib. There was an update of the OlympiAD trial published recently in the Annals of Oncology, January 2, 2019. And then the big talazoparib paper was published in the New England Journal actually last summer. Sara, can you summarize from your point of view the key takeaways from these 2 trials?

DR TOLANEY: Sure. So the OlympiAD trial compared olaparib to standard treatment of physician’s choice chemotherapy in patients with metastatic breast cancer who had germline BRCA mutations.

The design was very similar to the EMBRACA trial, which was really comparing talazoparib, another oral PARP inhibitor, also to treatment of physician’s choice, in a very similar patient population. And both of these studies did show statistically significant improvements in progression-free survival with an absolute difference in both trials. It was very similar, of about 3 months. And the OlympiAD trial is not powered to look at survival but has presented the data, and there is a trend towards survival benefit in the first-line setting but not in the later-line setting. And we need more mature data from EMBRACA to see the survival analyses.

So I think in general what we found is giving an oral agent was associated with longer PFS and, I think, very importantly, fewer adverse events and better quality of life for patients getting the PARP compared to chemotherapy.

I will say, I was a little disappointed with the 3-month absolute PFS difference. I was certainly hoping we’d see a little bit more. And I think the other concerns that we’ve had with the trials is that the treatment of physician’s choice isn’t really completely treatment of physician’s choice — it did not allow for platinum as a comparator. So we don’t really know how PARP compares to platinum.

And when you looked at the subgroup analyses for those patients who had previously been exposed to platinum, even though they weren’t allowed to progress on it in the metastatic setting, they could have been exposed to it. And in OlympiAD, for example, I think it was about 14% of patients had platinum in the metastatic setting before going on, their benefit does seem to be a little bit less than the overall population.

So I still think there are lots of questions: Can we do better than this? And do we need to add other agents to PARP to make outcomes better? How will it perform postplatinum? Is there a right order to give? PARP first then platinum, or platinum then PARP? So I think there are lots of remaining questions, but it’s wonderful to have a targeted oral agent available for our BRCA-mutant patients.

DR LOVE: Chuck, can you separate out these 2 agents in these 2 trials? Or from your point of view, is it a coin flip? Anything that you’re able to pick up, either from the data or clinical experience, in terms of toxicity? I know there’s a slightly different mechanism of action, but is it a coin flip in your mind?

DR GEYER: Yes, in terms of the efficacy, yes, I don’t see, really, that either stands out. They had really, to me, very similar toxicities. If I was going to pick one, I would tend to pick talazoparib simply because it’s a once-a-day dosing as opposed to twice-a-day dosing.

But that would be the basis of where I would start, but I would certainly want to try to find out which one would be less out of pocket for the patient. I don’t see that in the metastatic setting that there appears to be really any difference between the 2 agents. So yes, it would be a coin flip.

DR LOVE: And again, Sara, from your point of view clinically, any way to separate out the two? How do you decide which one to use?

DR TOLANEY: This is all coming back to, like, the CDK4/6 inhibitor story.

DR LOVE: Yes, absolutely.

DR TOLANEY: Very similar. I think, just as Chuck said, we can’t see any difference.  Hazard ratios are very similar across the studies. The subgroup analyses are very similar, so it’s hard to tease out differences from an efficacy standpoint.

I do think if you look at the toxicity standpoint — again, this is cross-trial comparison, very hard to know — but it did seem like rates of alopecia were a little bit higher with talazoparib relative to olaparib, which is certainly a concern for patients. So that’s something to keep in mind.

I do think mechanistically, the drugs, as Chuck alluded to, are a little different, with talazoparib having more potent PARP trapping than olaparib. But again, we can’t really see that in terms of efficacy just looking across these studies. So I think it kind of is a coin flip, and I think, just as Chuck said, I think what’s going to have the lower copay and be easier access to the patient.

Case (Dr Geyer): A 31-year-old woman with metastatic ER/PR-positive, HER2-negative breast cancer receives ovarian suppression, fulvestrant and abemaciclib

DR LOVE: So I want to talk a little bit about clinically how this works into your practice and maybe talk a little bit about your cases. And actually Chuck, I wanted to ask you about your patient, the 31-year-old woman.

The reason I was going to ask you about this 31-year-old woman was, it looks like she has a PALB2 germline mutation. And I’m curious, Chuck, are you using a PARP inhibitor for other germline mutations like PALB2? And also, what about somatic mutations?

DR GEYER: Yes, I mean, honestly, my practice is small enough, since I’m in clinic 1 day a week, that I don’t have a lot of patients with BRCA mutations that I follow. So I can answer things, but for her, I mean, I had definitely planned on pursuing those — I guess in her specific case there seems to be some very interesting data about combining these PARP inhibitors with immunotherapies with checkpoint inhibitors. And I know there’s a number of early trials going on that they seem to have an effect to heat up the tumor, so to speak, so you might get a little more punch out of an immunotherapy. She’s ER-positive, so she’s not somebody I would normally be thinking about that, but being 31, I guess I’d been, in my mind, thinking of watching the data and see what evolves there specifically to her situation.

I think right now the challenge would be getting reimbursement for it if you tried to prescribe it, quite frankly, and I’m always hesitant to bring up therapies that I think I’m going to run into a wall and not be able to provide. Fortunately, she’s had a great response to the CDK4/6 inhibitor and her estrogen therapy. So we’re not at a point now where we’re really even discussing this in the clinic with her.

Use of PARP inhibitors for patients with metastatic breast cancer and BRCA mutations

DR LOVE: So Sara, again, let’s maybe put aside reimbursement issues, or maybe you can comment on your reimbursement experience, but are there germline mutations for which you would like to use a PARP inhibitor right now? And are there somatic changes or mutations for which you’d like to use it?

DR TOLANEY: So I think the somatic mutations, we’ve been wondering somatic BRCA, we’re now finding a lot now that we do more next-generation sequencing on our tumors in the metastatic setting. And there is a trial ongoing, the olaparib extended trial that is looking at these questions, though it does have a cohort that’s analyzing the somatic BRCAs and then has a cohort that’s actually looking at all these other mutations — PALB2, ATM, ATR, et cetera — and so we don’t have that data available at this time. But I do think it’s not unreasonable, particularly in a somatic BRCA, to consider an oral PARP as an option for your patient. Again, provided it could be reimbursed.

DR LOVE: Chuck, one of the things that Sara was commenting on is the issue about platinums and, obviously, their efficacy in patients with BRCA germline mutations. Obviously in ovarian cancer, the clinical trials had a completely different pathway in that they gave these patients platinum and then followed up with PARP maintenance as opposed to the breast cancer clinical research strategy looking at PARP monotherapy. For practical purposes, Chuck, do you consider platinum, at least from an efficacy point of view, a platinum agent equivalent to a PARP inhibitor or maybe not as effective in patients with BRCA germline mutations?

DR GEYER: I guess it would depend on the population that we’re talking about. I guess I tend to use carboplatin as neoadjuvant therapy when I do have patients with germline mutations for DNA repair. So I guess I bought into the idea of including it in the neoadjuvant therapy. So it would be patients who are relapsing.

I have had some patients who come in who present with de novo metastatic triple-negative with locally advanced disease. I tend to treat those patients with standard neoadjuvant regimens just because when they come in, they have such a heavy disease burden and it’s such an aggressive cancer.

So I, Neil, I tend to use carboplatin really very fairly early, much sooner than I’m really thinking about a platinum versus carboplatin. So when I’m thinking about using a PARP inhibitor for metastatic disease, I’ve already done carboplatin probably in their neoadjuvant therapy.

So I tend to still use chemotherapy first line for metastatic disease and then I would go to the PARP. But I say that not having done it yet.

They haven’t been available that long, but that would be my general approach. I have a number of triple-negative patients now that have had good responses with the first-line regimens I’m giving them. So that’s kind of my plan, would be to use that. But again, these are in a very limited subset of patients who have germline mutations, so I just don’t have a large practice experience with that subcohort.

Case (Dr Tolaney): A 59-year-old woman with ER/PR-positive, HER2-negative metastatic breast cancer and a BRCA2 mutation receives olaparib as second-line therapy

DR LOVE: So Sara, I’d like to tease out from you where PARP inhibitors would fit in in your management of patients with metastatic disease based on their ER and HER2 status. Your 59-year-old woman obviously fits into the largest status of being ER-positive/HER2-negative. Maybe you can briefly overview her course, and then we can talk about ideally right now where you see PARP inhibitors fitting in, for example, ER-positive/HER2-negative metastatic disease.

DR TOLANEY: So in my particular patient, she was a 59-year-old woman with a strong family history of breast cancer who had presented actually with de novo metastatic disease and what appeared to be an inflammatory breast cancer presentation. So I had started her on hormonal therapy, so letrozole in combination with abemaciclib. And she actually had a very nice response with resolution of the breast erythema and decrease in her liver metastases.

She had been found to have a BRCA2 mutation very shortly after I had started her on endocrine therapy. And at time of progression on her endocrine treatment, I then had gone on to a PARP inhibitor at that time, so she had started on olaparib. And had had a nice response, but it only lasted unfortunately for about 6 months. And so now she’s on to platinum. She’s getting cisplatin treatment.

So I think in this particular case I had started endocrine treatment first for an ER-positive BRCA-mutant patient, and I think generally speaking, I still would have if I had known her BRCA2 status up front. I generally would like to see someone go through endocrine therapy and then once they’ve progressed on endocrine therapy to then offer them PARP inhibition. Usually I’d probably use that prior to exposure to chemotherapy in a hormone receptor-positive patient.

I think it’s a little different in a triple-negative patient obviously, where we don’t have the endocrine option, and so then you’re really looking at choices that involve chemotherapy versus PARP. Obviously now we think we’re soon to have availability potentially of immunotherapy and chemotherapy in the triple-negative setting if someone’s PD-L1-positive. I think then we’re going to come up with this dilemma of “what if you have a BRA-mutant PD-L1-positive patient, are you going to offer them PARP, or are you going to offer them nab paclitaxel and atezolizumab?”

It’s hard, because at this time, we don’t have a survival benefit seen with PARP. We have a trend in the first-line setting from the OlympiAD data that first-line patients may have a survival benefit. The IMpassion data suggest there may be a survival benefit — again, it was a trend towards survival. And so I think it depends a little bit on the patient, the disease burden and how, in terms of what they want from a quality-of-life standpoint, because PARP is so much easier on a patient to get an oral agent. But at the same time, I would probably want them to get exposure to immunotherapy in the first-line setting if they were PD-L1-positive.

DR LOVE: How is she doing on the cisplatin now? Is she responding?

DR TOLANEY: She just started on it. She does seem to be responding. She’s been in a lot of discomfort because her axillary disease had grown. And so it seems like it’s coming down a little bit. So hopefully she’ll continue to have a response.

DR LOVE: And how did she do on the olaparib? Any issues there?

DR TOLANEY: No. Her tolerability was really good. A lot of patients do get that up-front few weeks of low intermittent nausea, and then it tends to dissipate. And she had that and never needed any further antiemetics after that.

DR LOVE: Sara, right now, where, if at all, do PARP inhibitors fit into patients with HER2-positive disease?

DR TOLANEY: So it’s just not as common in HER2-positive patients to have germline BRCA mutations, so we don't see this very often at all. And I think that’s also a good question. I personally haven’t had this experience with the metastatic HER2-positive patient that was germline BRCA mutant. So I haven’t had to make that decision. But I think again, it is also a question of quality of life. I think just like we know in the HER2-positive setting we have the CLEOPATRA data with this 16-month OS benefit in the first-line setting, it’s hard to deprive someone of that therapy up front. But I do think probably in a later-line setting I would integrate the PARP whether it’s second, third line, somewhere in there, but probably not first.

Sequencing of PARP inhibitors for patients with metastatic triple-negative breast cancer (TNBC) and BRCA mutations

DR LOVE: So any comments, Chuck, in terms of sequencing PARP inhibitors, particularly this new issue, which is a great issue to have, which now in triple-negative metastatic disease we have the option now of atezolizumab with nab paclitaxel showing such promising results, but yet also the option of a PARP inhibitor. How do you see that working out?

DR GEYER: I mean, I think the thing that strikes you about the checkpoint inhibitors are that when patients respond to it, yet they respond well and for a long time. I mean, that’s the tail of the curve with those drugs. It is so appealing. And there is really very consistent evidence across a number of tumor types that if you’re going to use it, use it earlier. I mean, that makes sense. It isn’t surprising, since you’re really depending on the person’s immune system to do the work. You’re just trying to take off some of the restrictions.

And I think the PARP inhibitors in the metastatic setting are going to be a palliative drug. So it’s just about sequencing, where people are, what are their cumulative toxicities from prior therapies, how they’re feeling, what events in their life. I don’t think there’s really an efficacy-based algorithm for sequencing these agents. I think there’s so much variability that it’s really just tools in a toolbox that you decide how to use.

Really, the big question is, what do I lead with? How intensely do I treat this patient when they first develop their metastatic disease? What are my regimens? To me, that’s the critical point, and as you get further into it, it’s just much more about tolerability and persistent side effects in my mind, Neil.

Ongoing PARTNER/PARTNERING studies evaluating olaparib with platinum-based neoadjuvant chemotherapy or as a part of novel combination approaches; investigation of neoadjuvant talazoparib for patients with TNBC and a germline BRCA mutation

DR LOVE: So let’s spend a minute talking about current and new trials that are looking at PARP inhibitors, Sara. I listed a few trials that were presented at San Antonio in terms of ongoing trials. I was particularly interested in this partner, I guess, initiative, looking at combinations. Also this other paper looking at neoadjuvant PARP. Sara, any comments on some of these new strategies?

DR TOLANEY: So I think it does come back to this question of “Can we do better by adding chemotherapy to PARP inhibition?” And I’ll say it’s been a challenge, because our prior work that’s been done looking at combining particularly platinum and with PARP inhibition, has been challenging due to hematologic toxicities. These agents do cause enhanced neutropenia and anemia, and so dosing has been a challenge, at least if you were trying to maintain full-dose continuous PARP with platinum. So that hasn’t been well tolerated. So I think this strategy now has been, can we use intermittent PARP and still get the synergistic benefits?

And so one of the trials that you brought up was looking at whether or not we could combine a standard preoperative regimen, such as the regimen that was used, for example, in the CALGB study, the ACT carbo-based regimen and add on olaparib to it, looking at optimal dosing on an intermittent schedule and then, assuming that that works in a safe fashion, then enrolling a large, randomized study.

I think an also very provocative question, though, is, can you just do as well with PARP alone in the preoperative setting? And so Jennifer Litton has done a lot of work in this area, where she had initially piloted giving talazoparib just for a couple of months and found that she could get tumors to reduce just by ultrasound imaging, almost 90% reductions in tumors. And so then she expanded that to longer exposure to PARP inhibitors with up to 6 months of exposure and showed in a very nice presentation, even though it’s just 20 patients, that there was a 50% path CR rate and 60% of patients had RCB01.

So I think that was very impressive. And I think that leads to the question is, (1) can we confirm this in a larger patient population, which is the study that she’s now doing? And if so, how can we then better modify our treatment? How can we deescalate therapy for germline BRCA patients? Can they actually get away with maybe PARP alone? And so I think that’s a direction we’re heading to.

Ongoing Phase III OlympiA trial evaluating olaparib as adjuvant treatment for high-risk, HER2-negative breast cancer with a germline BRCA mutation

DR LOVE: Chuck, any comments about ongoing trials? There’s also, I guess, the OlympiA trial out there, your adjuvant trial.

DR GEYER: Yes, OlympiA is — we’re reaching the end of the accrual. We’ll have 1,800 patients that were higher-risk triple-negative breast cancer or ER-positive breast cancer with an associated germline mutation. That is the maintenance use of the drug, single agent, to get around the problem of coadministration with chemotherapy. So I think we’re really pleased to have that study finishing accrual. That’s been quite an effort to get that study done. But I think it should be very interesting data.

DR LOVE: So this is all BRCA germline?

DR GEYER: Right, all BRCA germline. And that was part of the challenge. If women are aware they have a germline mutation, they’re getting very intensive screening, so when they develop their breast cancers, if they’re in major centers, they’re going to be lower-risk triple-negatives, so they wouldn’t meet the eligibility criteria.

So we really have had to get out into community practices with this trial where patients are diagnosed with more advanced disease, higher-risk disease, and then it’s learned as part of their evaluation that they have a germline mutation. So that’s been a challenge on that trial.

DR LOVE: A challenge. I’d say it’s unbelievable. You’ve got 1,800 patients on that study. It’s, like, amazing. Amazing. Congratulations.

DR GEYER: Yes, we’re excited. We’ll be doing an interim analysis sometime this year, we believe. We’ll be reaching the interim analysis point for it. Stay tuned.

DR LOVE: That was like when we were waiting for SOLO1 in ovarian cancer, like, literally the hazard rate was exactly the same as metastatic disease. So I’m not going to tell you where my money is for your OlympiA trial right now, but I think it’d be a pretty —

DR GEYER: But I do agree. I think Jennifer Litton’s data on that neoadjuvant activity of talazoparib as a single agent is just — I mean, that is stunning. That is amazing how active that therapy is. And that’s what — you start thinking, “Wow, oh, yes. Let’s put that with the checkpoint inhibitor. Maybe that’s all you need.” I think it’s very exciting for these women who develop these cancers related to these germline mutations.

Efficacy and safety of atezolizumab with nab paclitaxel for advanced TNBC in the Phase III IMpassion130 trial

DR LOVE: So speaking of checkpoint inhibitors, our next topic is something we’ve already been talking about, which is triple-negative disease. And Sara, I listed here the New England Journal paper from a few months ago with those data that were, I think, presented in Europe first.

Can you talk about the study? How they did it? What they found? And then we’re going to go through some of these other newer studies looking at this strategy. But what about this trial?

DR TOLANEY: So this was the IMpassion130 trial, which took patients who had metastatic triple-negative breast cancer who were getting their very first treatment in the metastatic setting and randomized them to receive nab paclitaxel alone or nab paclitaxel with atezolizumab. And the study had a very complex primary endpoint and design. So it had, in essence, 4 coprimary endpoints. It was looking at progression-free survival and overall survival in the intent-to-treat population and in the PD-L1-positive population. Again, it was a hierarchical design, so you had to meet the endpoint in the intention-to-treat population before you could analyze the PD-L1-positive subgroup.

And so what they found was that in the overall population, there was a statistically significant improvement in progression-free survival in the intention-to-treat population and in the PD-L1-positive population. And so I will say that the improvement wasn’t as large as I would have hoped. It was about 1.7 months in the ITT population and about 2 and a half months in the PD-L1-positive population. And then when they looked at overall survival in the intention-to-treat population, there was no difference seen. So they weren’t able to statistically analyze the PD-L1-positive population but did find in just looking at the numbers, there was almost a 10-month difference in survival, which is quite startling and really very impressive.

So I think overall what we’re seeing is that the benefit is really restricted to PD-L1-positive. So even though the PFS is technically positive in the intention-to-treat population, it’s really all being driven by the PD-L1-positive patients. If you look at the PD-L1-negative subgroup, there really is no PFS benefit. So we’re seeing this impressive benefit really in terms of, I think, a clinically relevant survival benefit.

I do think it’s important to realize this is an interim analysis for survival. And so we will have to wait to see what the final survival data is. But this is a disease where the median survival is 13, 14 months. And so seeing a 10-month OS difference, even at the interim, I think is impressive enough to make us want to start using this treatment for our PD-L1-positive patients.

Choice of chemotherapy for patients with metastatic TNBC

DR LOVE: I’m just curious, Sara, what’s your prior use been of nab paclitaxel in breast cancer? And are you using nab paclitaxel in this situation?

DR TOLANEY: So generally I do not use nab paclitaxel. I will say I usually use paclitaxel. And certainly this trial was designed because they were trying to avoid steroid use in combination with immunotherapy and wanted to use the nab paclitaxel for that reason. I think it brings up a very good point: Would the utility be the same if paclitaxel had been used? There is an ongoing trial actually that we don’t have data for yet that substituted paclitaxel instead of nab paclitaxel with atezo. So we’ll have to see what that data shows. My guess is that it probably will show similar results.

And then I think the next question is, what about other chemotherapy choices? A lot of patients recur very soon after adjuvant taxane. This is a trial for patients who are more than 12 months out from their adjuvant taxane. We don’t want to really reexpose them to taxane again if they’re recurring within 12 months. And so this isn’t an optimal choice for those patients. And so we don’t really know, would benefits of immunotherapy be the same in someone who recurs sooner? And would it work in someone who’s paired with a different chemo partner or a chemo partner that utilizes steroids? And I think the other question I have in my mind is, the PD-L1 testing was also done with a very specific antibody, the SP142 antibody, which is testing PD-L1 on the immune cells. Which is different than a lot of hospitals currently do, where they’re using standard antibodies used, for example, in lung cancer.

And so I think the question in my mind, just from a practical question is, can I trust any antibody and what PD-L1 has been done on and say yes, I would have seen the same benefit in that patient? Also, our own hospital has taken the stance that we did want to adopt using SP142 for our breast cancer patients until we get data from other trials like KEYNOTE 355, which looked at pembro. So we’re waiting on that data to know. But until then, we’ve taken the stance that if someone is PD-L1-positive with the SP142 antibody, we will give them atezolizumab.

Yes, I probably would use nab paclitaxel to be consistent with the data if they were greater than 12 months out from their adjuvant therapy. If someone recurred sooner, I think I’d still try to give them another chemo choice with atezolizumab, maybe platinum and atezolizumab if I could get it covered by insurance, even though I don’t have the data for that at this time.

DR LOVE: So Chuck, Sara just brought up about 15 different points. Let me just see if you could cover a few of them. First, again, I seem to recall that you tended to use more nab paclitaxel even before this, but maybe I’m not remembering it. But I am curious what your thoughts are about choice of taxane. And is this generally your standard of care right now?

DR GEYER: Yes. I’ve tended to use nab paclitaxel if it will be covered by insurance for metastatic disease. I have, I guess, a biased personal opinion, you seem to be able to go a bit longer with it before you have to stop for neuropathy. And so in metastatic disease, I’m viewing it as that’s an important feature of things. Where if — obviously if people have allergic reactions to paclitaxel…so I have used it in some. But the coverage is intermittent. The metastatic studies basically did show that there really didn’t seem to be a lot of difference between paclitaxel, nab paclitaxel, in terms of efficacy. CALGB had done that trial.

I guess you tend to be influenced by personal experience.

DR LOVE: But what about the other point that Sara was bringing up of a patient who relapses not long after having a taxane? Would you be comfortable trying to access atezolizumab with another type of chemotherapy?

DR GEYER: Yes. Yes, I would. I think the checkpoint inhibitors, all of them have been studied in other tumor types with these same chemotherapy agents. And to be the issue of adding things where there is no experience is, you have no information on tolerability. But we see these patients with lung cancer get these drugs in combination with cisplatin. So I would not be uncomfortable about using chemotherapy drugs that made sense to me in combination with my checkpoint inhibitor, because really in breast cancer it seems that we need our chemotherapy to do some cancer killing, to present antigens in hoping that the immune system can pick up on that and get the desired effect.

Perspective on overall survival for patients with advanced TNBC and PD-L1-positive tumors treated with atezolizumab and nab paclitaxel

DR GEYER: I think the interesting thing about the IMpassion130 data still to me is, there seems to be this binary effect that if 1% of the immune cells are positive, they benefit. And it’s just surprising to me, because they did look to see, do patients with higher levels, higher percentage, do better? And they really didn’t seem to, looking at the forest plots. It’s benefit, benefit and then it’s gone if they don’t have any staining, which to me is still very interesting data. Biologically that doesn’t make sense to me, but it’s what the data shows.

DR LOVE: Any comments, Chuck, about this big survival bump that you get? It seems like every now and then you see trials where you — even CLEOPATRA — where, I don’t know, the survival benefit is much more than you would expect based on, say, progression-free survival. Any thoughts about that here? And is that a big motivating factor for you in terms of using the therapy?

DR GEYER: Oh yes, definitely so. I mean, I think one of the consistent patterns of the checkpoint inhibitors, particularly the PD-1/PD-L1 agents, is that you seem to get a bigger effect when you get out to survival over the PFS, and the only way that makes sense is if that means something has changed during the exposure that is making the immune system work better. You’ve recharged it so it’s able to carry the fight. I know that’s very teleological, but the data is so consistent that to me, again, using it first, using it early, it all makes sense to me.

So I really think there is something to that. It’s beyond now what you can say might be a statistical quirk. It’s consistent.

I think that’s what CLEOPATRA is telling us, that the 2 HER2 antibodies are working as immune therapies in CLEOPATRA because of that pattern. I think that there’s something going on there, and that’s why it is important to get the antibodies in early in patients who have metastatic disease to try to get that benefit for them.

DR LOVE: So the pessimist me always thinks, “Maybe there’s some other reason related to how the trial was done or post-trial therapy.” But I like your theory a lot better.

Case (Dr Tolaney): A 52-year-old woman develops pneumonitis while receiving pembrolizumab and eribulin for metastatic TNBC

DR LOVE: I want to also get into the issue of toxicity, particularly autoimmune toxicity of checkpoint inhibitors. Of course we talk about that all the time when we do programs on lung cancer, bladder cancer, et cetera. But now breast cancer is there.

And Sara, you actually had a patient — without going through all the details, but it was interesting — who developed a pneumonitis. This patient was on a trial of eribulin and pembrolizumab. And then you were part of this paper I was asking you about. I was amazed. It was Endocrine Reviews, February 2019, “Endocrine Toxicity of Cancer Immunotherapy.” I guess you somehow got involved in looking at toxicity of checkpoint inhibitors at Dana-Farber. But can you talk about, first of all, what happened with your patient with the pneumonitis? And globally, what are some of the issues that are coming up, particularly in this scenario of breast cancer?

DR TOLANEY: So my particular patient had been on a combination of chemotherapy and pembrolizumab, so she got eribulin and pembrolizumab and did develop very acute presentation with cough, shortness of breath and hypoxia and actually required hospitalization and IV steroids to treat her.

So I do think we are now seeing these toxicities occur. Generally, these immune-related toxicities do occur early, usually within the first 3 months of initiation of checkpoint inhibition. And the biggest rate of fatalities due to PD-1 inhibition is pneumonitis. So it is something to be very aware of even though it only occurs in about 2% to 3% of patients. It is a large cause of death. And so we are very cognizant of this issue. And we do make sure we get on top of initiating steroids, holding immunotherapy and watching them carefully.

So she has now been on a prolonged steroid taper. We usually do taper very slowly with pneumonitis. And she’s doing well. But, obviously, we always get cautious once they discontinue steroids and make sure they don’t recur.

I was not going to reexpose this particular patient to immunotherapy, given just the severity of her initial presentation and with pretty significant hypoxia. And so she’s just continued on eribulin alone. She did achieve a very impressive response early on and has continued to maintain that.

Endocrine toxicities associated with immune checkpoint inhibitors

DR TOLANEY: I think, going to your other point about the endocrine toxicities, they’re the most common immune-related toxicities that we actually do see, with thyroid toxicity being the most common. Interestingly, if you look at the IMpassion130 data, it’s almost 20% of patients that do develop thyroid toxicity.

So I think it does really make us need to be aware of the fact that we need to check thyroid function testing. The guidelines suggest at baseline, then every cycle of therapy for the first 5 cycles or the first 3 months and then every other cycle thereafter. So it is important to screen for it.

DR LOVE: I’m kind of curious, how many patients clinically are hyperthyroid before they’re hypothyroid, Sara?

DR TOLANEY: So the vast majority actually present with thyrotoxicosis. So it’s about 80% of patients. But 80% of the people who present with thyrotoxicosis actually convert to hypothyroidism. And so that’s actually something people really need to be careful of, that once you see the thyrotoxicosis, you really should be rechecking their thyroid function every couple weeks to see when they convert to hypothyroidism.

Ongoing Phase III NSABP-B-59/GBG-96-GeparDouze trial evaluating neoadjuvant chemotherapy with atezolizumab followed by adjuvant atezolizumab

DR LOVE: So Chuck, any comments about new directions and using checkpoint inhibitors? I particularly wanted to ask you, you had a poster at San Antonio, ongoing trial poster looking at neoadjuvant chemotherapy with atezolizumab in triple-negative — this is NSABP-B-59. We’ve seen some really cool things with neoadjuvant checkpoint inhibitors in lung cancer and also bladder cancer, probably other cancers I don’t know about. How about this trial and the idea of incorporating checkpoint inhibitors into the neoadjuvant setting?

DR GEYER: Yes. I mean, it is aggressive to move it into the neoadjuvant situation, but we felt like the promise of the drugs was sufficient that it warranted really evaluating it in a trial. SWOG is doing a study looking at pembrolizumab versus no therapy in patients with residual disease, using that residual disease as an indicator. I guess part of the thinking here of the neoadjuvant therapy is that you’re exposing the patient to the checkpoint inhibitor during the time that you are cytoreducing, killing the tumor, presenting the antigens, and so there might be an enhanced response to the therapy in the curative setting. And our study is actually the second study — there’s a global study that’s already finished accrual, their neoadjuvant study.

There’s a follow-up on the GeparNuevo data from GBG as well. So there’s going to be 3 of these large neoadjuvant studies in triple-negative breast cancer proceeding. So we’re excited about it. We’ve got about 130 patients in. And, obviously, the IMpassion130 data is a bit of a challenge to us, because we did not — starting the study, we did not stratify for PD-L1 status. We’re doing that. We are waiting for some — they have a smaller neoadjuvant Phase III study, that’s IMpassion030, that we’re kind of waiting and watching to see what that reports in terms of possible PCR rates based on PD-L1 status. So we will be probably modifying our study, but we’re proceeding.

Ongoing Phase III ALEXANDRA/IMpassion030 study evaluating standard adjuvant chemotherapy with or without atezolizumab for early TNBC

DR LOVE: So Sara, I was just flashing on the fact that I was working with your colleague Geoff Oxnard from your lung cancer department, and we were talking about the use of durvalumab after chemoradiation therapy that’s now standard of care. It’s really adjuvant therapy. A big reduction in recurrence rates. So starting to see these agents come into the adjuvant setting.

And I see there’s actually an adjuvant trial with atezolizumab, Phase III study going on in triple-negative disease. Any comments, Sara, about these new trials looking at checkpoint inhibitors earlier on?

DR TOLANEY: So I think, to Chuck’s point, that I think the excitement really comes in the preoperative setting. Most of our triple-negative patients actually do get preoperative therapy. I usually give preop treatment to anyone with a tumor over 1 and a half centimeters in size with triple-negative disease. And so I will say it’s a little challenging to accrue a large adjuvant Phase III trial, because most of our patients aren’t getting adjuvant treatment instead. They’re getting preop therapy. And I do think this idea of exposure with the tumor in place may be beneficial and this concurrent nature with the chemotherapy.

And so I do think it is an exciting field, and we’ll have to see where it goes. I think the bar is certainly higher in the adjuvant space and where we actually do achieve a path CR rate of about 50% in the triple-negative setting. And I do think preoperative trials really need to be focused on long-term outcome with event-free survival rather than judging benefit purely by PCR in the triple-negative population.

Emerging data with the HER2-selective tyrosine kinase inhibitor tucatinib for advanced HER2-positive metastatic breast cancer

DR LOVE: So I want to finish out talking about some new agents in development. And we were talking about obviously HER2 disease, and tucatinib was mentioned. So Chuck, can you talk a little bit about this TKI tucatinib, what we know about it and what’s being looked into right now in terms of ongoing clinical research?

DR GEYER: Sure. It’s a small-molecule TKI that is specific for HER2. So it seems to, not surprisingly, be associated with significantly less toxicity than the small molecules that inhibit other members of the HER family.

So it’s had a very favorable toxicity profile. And it’s been studied, of course, primarily in more advanced stages of metastatic breast cancer, some studies looking at basically patient populations who had had T-DM1 — so fairly heavily pretreated — looking at tucatinib with trastuzumab, with capecitabine. And there have been really very impressive activity of the 3-drug combination. I don’t think anyone has really viewed it as an important single agent. But its low toxicity profile, ability to be a second drug targeting HER2, being a small molecule that can get into the CNS.

I think a lot of the excitement I have for the drug is, it seems a leading candidate for maybe finally doing something about the problem of brain metastases. And to me, it makes sense to move it aggressively into the earlier stages to prevent it.

There’s Phase II studies, Phase I studies, they have included patients with brain metastases, and there’s been impressive 35%, 40% response rates in patients with brain metastases that had been presumably developed with lots of prior therapy. So it does have activity on established disease. But none of them have a complete response, and it comes back.

So to me, the excitement of the drug is bringing it forward into earlier stages of disease. It’s being looked at now in the HER2CLIMB study. That’s going to be the hopeful indication study basically for patient populations who had had T-DM1. And everybody gets trastuzumab/capecitabine, and then the randomization is tucatinib or placebo. So it’s a triple-drug regimen.

There’sa study looking at T-DM1 with and without tucatinib in patients who are with metastatic disease who are appropriate for T-DM1. So it is proceeding now to Phase III studies that, hopefully, might lead to approval if the early promises are confirmed in these larger, more definitive trials.

DR LOVE: I’m curious, Chuck. How often do you use trastuzumab/capecitabine outside a trial setting in the management of metastatic disease?

DR GEYER: I mean, for period of time, I had stopped using it, because I kept getting denials for it, and I had stopped. But my colleagues in my clinic who see more patients say now that they have no trouble with it and they notice that for patients with brain metastases, it seems to be something that they’re using a bit more, just feeling like it’s a way of dealing with the CNS disease without the toxicity of capecitabine/lapatinib combination.

I think it’s a reasonable combination of drugs, again, in that advanced setting where you’ve been through a lot of therapy and you’re just trying to find things that folks will tolerate well. Because trastuzumab is a very well-tolerated drug, and I kind of think once people have metastatic disease, there’s benefit for having that in the background as you’re changing the therapy around it.

Comparison of the activity and tolerability of tucatinib to that of lapatinib and neratinib

DR LOVE: So Sara, I’m curious about your thoughts about tucatinib. Nancy Lin from your group has had a particular interest in brain mets. I’m not sure to what extent she —  and you’ve worked with her as it relates to this. Can you comment a little bit about what we know about tucatinib and what experience, if any, you have with it?

DR TOLANEY: So as Chuck alluded to, tucatinib is an oral novel tyrosine kinase inhibitor that more potently hits HER2 than HER1, so it’s actually 1,000-fold more potent against HER2 than HER1, which is a very unique ratio relative to the other TKIs, like lapatinib and neratinib, which are more equally potent with HER1. And so what the benefit here is that not only are you hitting HER2 harder, but you’re not getting the EGFR-related toxicities like those other drugs are, such as the diarrhea and the rash. And so it’s a much better-tolerated agent. And then has this very nice property of CNS penetration. And we certainly know that about 50% of our patients with HER2-positive disease end up with brain metastases and, unfortunately, do die of their brain metastasis.

So Nancy, as you know, in our group, has done a lot of work with tucatinib. And we had some of the initial work that had been done just with tucatinib and trastuzumab and then have participated in HER2CLIMB, and I do think it’s a very active drug that appears to be well tolerated.

So I do really like the way HER2CLIMB was also designed, because it’s unique and that it not only is for patients who have systemic progression with their metastatic disease but also allows for patients with CNS progression onto the same trial. And so it is really a very clever design to really help address both populations, which are areas that we struggle with HER2-positive brain mets.

DR LOVE: You talked about the specificity biochemically, Sara. What about clinically? How do you compare the toxicity of tucatinib to, say, lapatinib and also neratinib?

DR TOLANEY: So I do think it is better tolerated. You don’t see the same level of diarrhea that we do see with lapatinib or neratinib. And, obviously, currently in the HER2CLIMB study it’s given with capecitabine, which obviously does cause diarrhea. But it’s not the same level of diarrhea that, for example, we see with capecitabine and neratinib, which does require significant loperamide prophylaxis up front. And so I do think it is a well-tolerated regimen.

Efficacy and safety of the novel antibody-drug conjugate trastuzumab deruxtecan for HER2-positive breast cancer

DR LOVE: So Chuck, I’m curious about another anti-HER agent. I’ve heard of this described as, I don’t know, the son or daughter of T-DM1, but trastuzumab deruxtecan. Can you talk about what this antibody-drug conjugate is and what we know about it?

DR GEYER: Yes. I mean, again, it’s trastuzumab linked to a camptothecin derivative. So it’s a TOPO-1 inhibitor. So I think a nice feature of that drug is that you’re not using a chemotherapeutic agent with potential neurotoxicity.

The very interesting thing about the drug, though, is that it has really remarkable activity when you look at the Phase I studies that they elected to include patients with so-called HER2-low disease based on their preclinical work that suggested that you might not need the full HER3 level of protein expression.

So they went into their development plan intentionally looking at the spectrum of IHC intensity, so to speak. And the waterfall plots from their early-phase trials are really very impressive. The drug has substantial activity in the 3+ patient population. These were heavily pretreated, prior trastuzumab/pertuzumab, T-DM1. They’ve got a 50% response rate. And most patients have some regression, even if they don’t respond when looking at the drug. But really the waterfall plots though for the 1+/2+ patients were really remarkably similar.

So I think it’s a very interesting immunoconjugate. There were some early problems with pulmonary toxicity with the drug. And so they have had to work through figuring out management algorithms and also deciding on what dose they were going to use. Because the toxicities with the drug tended to show up later, so they had escalated on past — they had a couple of doses, and they were seeing late toxicities with the higher doses. So they had some nice posters at San Antonio that basically showed the work they had done to more or less pick a dose.

And they’re taking it into metastatic HER2-positive patient populations with what they’re calling their DESTINY studies, directly comparing it with T-DM1 or in patient populations appropriate for T-DM1 or if they’ve had prior T-DM1, then it’s capecitabine with either trastuzumab or capecitabine, based on investigator choice.

So the drug is moving forward, and we’re going to have direct comparisons available with T-DM1 in the metastatic setting ahead. So I think it’ll be interesting, which, in and of itself, is very interesting, but it will also be really intriguing to me to see if this does — our B-47 effort to look at trastuzumab in the HER2-low and early breast cancer was a stunning negative study, when you looked at how on top of each other the curves were. And it will be interesting to see how development goes in this 1+/2+ population, which, of course, is a much bigger cohort of patients than the HER2-positive subset. So I think it’s a very interesting drug that will be fun to watch.

DR LOVE: That’s really interesting about this 1+ and 2+. I was flashing on all the discussions that have gone on about brentuximab vedotin in lymphoma based on CD30 levels. And for a while, they were seeing maybe it works without CD30. But kind of another antibody-drug conjugate. I’m just kind of curious, Chuck, do we know whether or not T-DM1 works in IHC or low HER2?

DR GEYER: I don’t know that they looked. They took it there based on preclinical work. Do you know, Sara?

DR TOLANEY: Actually, I was looking for that when I had to do a discussion last year on the new HER2 ADCs, and I actually didn’t see any data. And my understanding is, the thought is it likely wouldn’t work, because it is a noncleavable linker, that they don’t have bystander effect with T-DM1. And I think part of the thought of activity in the low HER2 has also been not only is it very potent against HER2 but also does have bystander effect and can work in heterogeneous tumors. So I think that has been part of the benefit of DS-8201A.

DR GEYER: And it carries more per trastuzumab molecule too, as well.

DR TOLANEY: Yes.

DR LOVE: Interesting. So interesting.

Results of the Phase III SOLAR-1 trial evaluating the PI3 kinase inhibitor alpelisib with fulvestrant for ER-positive advanced breast cancer

DR LOVE: So one final topic. Amazing, here we talked for all this time and think about all the things we haven’t talked about in breast cancer. We just talked about little pieces. But there’s so much to talk about nowadays. But I’ve got to hear a little bit about PI3 kinase inhibitors. And particularly the big presentation, I guess it started — the first time I heard about it was at ESMO with alpelisib, Sara, the SOLAR-1 trial. I think they presented some more data at San Antonio. What is alpelisib, and what do we know about it? Is it going to be part of our care in the near future?

DR TOLANEY: So the SOLAR-1 data were very interesting. We have a whole history where we’ve looked at PI3 kinase inhibitors in breast cancer. Initially, we looked at pan inhibitors with endocrine therapy and saw no benefit. And then we looked at beta-sparing inhibitors with taselisib and saw no benefit. And now we’re seeing an alpha-specific inhibitor, alpelisib, in SOLAR-1, where patients were randomized to receive either fulvestrant or fulvestrant with alpelisib. And they found in the patients who did have a PI3 kinase mutation that there was a significant improvement in progression-free survival from about 5.7 to 11 months, which is, I think, a pretty impressive benefit that was seen.

There were toxicities and I do think are important to note. Almost 35% of patients did have Grade 3 hyperglycemia. About 10% of patients had Grade 3 rash. So there are certainly some toxicities with this drug that can be managed but are important to be aware of.

Some of the interesting data I thought that was also presented at San Antonio was looking at PI3 kinase status by circulating tumor DNA rather than just the initial analysis, which was based on tissue analysis. And they found even more profound benefits in terms of the hazard ratio and the ctDNA-positive patients, so the patients who were PIK3CA mutant on circulating tumor.

So I think it is going to be a drug that I will consider using in my PI3 kinase-mutant patients. I think I will generally approach patients who are post-CDK4/6 progression. So, for example, a patient who progressed on an AI/CDK4/6 inhibitor and that has a PI3 kinase mutation, I would likely offer them fulvestrant and alpelisib in the second-line setting. And again, being cognizant of toxicities and monitoring them.

I think it does call into play, how are we also going to diagnose these patients with the PIK3CA mutation? Is it going to be tissue or ctDNA? And I know they’re in the works of coming up with a companion diagnostic on circulating tumor DNA, which would be great and certainly make practice easier.

DR LOVE: Just to clarify, in terms of your theoretical algorithm if the drug becomes approved. So I would guess that if you normally, after CDK/hormones, you’re either going to with everolimus plus something? Is that your usual, or fulvestrant?

DR TOLANEY: Yes. No, I think usually my options are either fulvestrant, fulvestrant/everolimus or exemestane/everolimus.

I will say that the data from the PrECOG study looking at fulvestrant/everolimus looked just as good as BOLERO-2. With exemestane/everolimus, the hazard ratios are very similar. And so it’s actually a very similar hazard ratio to also SOLAR-1. And so I do think that all of these options are very reasonable. I think in the PI3K-mutant patient though, I probably would pick alpelisib over everolimus-based therapy in the second-line setting.

DR LOVE: Chuck, any comments on alpelisib? Sara sounds like she would use it if it were approved. Would you?

DR GEYER: I think I would in this patient subset, because the other thing that was there at San Antonio is, they did have a small number of patients in the SOLAR-1 who had had prior CDK4/6, and it seemed to work in that patient population kind of to the same degree. Because that’s also the question that comes up when these studies are done now without that prior exposure, since we really all make — I mean, you’re going to go to those agents first, because of the marked improvement we saw. But I think it makes sense to use it. I haven’t used it yet, but I have used the everolimus. And I do think that it’s great that we don’t have the mucositis problem, but the fatigue is still there. It’s a problem that I haven’t figured out. So I think I would look for it and try to use this more selected drug if I had a patient that had the appropriate mutation.

NEO-ORB: A randomized Phase II study of neoadjuvant letrozole with alpelisib for HR-positive, HER2-negative breast cancer

DR LOVE: So one final question, and forgive me for my lack of memory, but Sara, did you do the neoMONARCH study?

DR TOLANEY: No, I didn’t participate in it. Actually, Sara Hurvitz —

DR LOVE: Okay, Sara Hurvitz. Okay.

DR TOLANEY: Different Sara.

DR GEYER: Different Sara.

DR LOVE: Okay. The reason I’m asking is because one of the papers we pulled was by Ingrid Mayer, who looked at neoadjuvant letrozole plus alpelisib. Okay? And from what I can see, it did not work. It didn’t increase the response rate. But if I understand it correctly, in the neoMONARCH —

DR GEYER: Neil, I think it dropped the Ki-67, it’s just that letrozole drops the Ki-67 pretty well on its own. It didn’t add that much.

DR LOVE: It didn’t add much to letrozole.

DR GEYER: Yes.

DR LOVE: But I know it’s — I mean, I guess the thing I was — when I was looking at the neoMONARCH study, the thing that — what struck me was (A) that it worked. Okay. Right? It dropped the Ki-67 more than endocrine therapy alone. I think it increased the response rate. Right?

DR TOLANEY: I mean, I think the problem with both of these trials is that they really aren’t seeing path CRs on any study, right? They’re very rare to see a path CR. But I think, certainly to your point, they do see tumor reduction.

DR LOVE: Yes. I was just thinking about the — here’s the real question I have. The only thing that I thought really struck me was, if I understood the neoMONARCH study, they saw the same thing with abema alone. Is that right?

DR TOLANEY: It was more profound with the combination. So if you looked at the AI, AI/abema and abema, I think the combination was the most profound. Obviously, just as Chuck said, even the AI alone had a very profound drop in Ki-67, but it was more robust with the combination.

And it is interesting, in that they’re not seeing this with the PI3 kinase inhibitor in the preop setting, though it did seem like they did see more benefit in a biomarker group, I think it was phospho-AKT group that seemed to have the most benefit from alpelisib.

I think it’s hard, because in the preoperative setting we’re really looking for a biologic predictor of benefit, because we know that path CR in ER-positive disease isn’t going to predict long-term outcome. So it’s hard to judge, are these drugs going to have long-term efficacy just based on response at time of surgery? You’re more looking at these biomarker changes, and it’s hard to know what is the optimal biomarker to judge the PI3 kinase efficacy in this preop population. So it’s a tough outcome.

DR LOVE: Subsequent to this interview, on March 8, 2019, based on the results of the IMpassion130 trial, the FDA granted accelerated approval to the combination of atezolizumab plus nab paclitaxel as first-line treatment for patients with unresectable, locally advanced or metastatic triple-negative breast cancer whose tumors are PD-L1 positive.

This concludes our program. Special thanks to Drs Tolaney and Geyer, and thank you for listening. This is Dr Neil Love for Oncology Today: Breast Cancer Edition.