Adjuvant I/O in TNBC

Transcript:

Sara M. Tolaney, MD, MPH: Very recently, we’ve seen some exciting data that came out from KEYNOTE-522, which looked at a preoperative trial of adding pembrolizumab to chemotherapy in patients with triple-negative breast cancer.

In this trial, patients received chemotherapy with anthracycline, Cytoxan, taxane, and platinum and then were randomized to receive that chemo [chemotherapy] backbone with or without pembrolizumab, and then patients went to surgery. And they found that patients experienced higher rates of pathologic complete response with the addition of pembrolizumab to that chemo backbone with an absolute difference between the 2 arms of about 14%, again favoring pembrolizumab in that case.

We do have some very early interim analyses from the event-free survival, suggesting a trend toward better long-term outcomes in patients who received pembrolizumab compared with those who did not. Although, again, at that point in time, with just 18 months of follow-up, that did not reach statistical significance. And so we are waiting on longer-term follow-up from KEYNOTE-522 to see if it meets the event-free survival end point, which I think most of us think will probably happen, given the very strong trend that we currently see.

I think if that occurs and pembrolizumab were to become a standard option in the preoperative setting, the challenge we’ll face is whether patients need that chemotherapy backbone that was administered in KEYNOTE-522, a very intensive chemotherapy regimen, which many of us don’t often give. Many of us will omit platinum, for example, and just give anthracycline, Cytoxan, and taxane. So I think we’ll need more work to see if we can de-escalate the chemo backbone. I think we’ll need to know if patients actually require receiving adjuvant pembrolizumab which was also done in KEYNOTE-522 where patients in the PEMBRO [pembrolizumab] arm also went on to receive 9 additional cycles of pembrolizumab in the adjuvant setting.

It did not seem like PD-L1 was a biomarker to predict benefit, at least in terms of pathologic complete response. The benefit was seen both in patients with PD-L1—positive disease and in those with PD-L1–negative disease. But we don’t yet have data to know if event-free survival would be dependent on PD-L1 status or not, so we do need those data to know if PD-L1 will have any predictive ability in the early disease setting. So I think these data are very encouraging and certainly lead us to think that checkpoint inhibition may have a role in the preoperative setting hopefully soon.

We had seen other data from the NeoTRIP trial, which had looked at omitting anthracycline in the preoperative setting, so the investigators just gave carboplatin and taxane with or without atezolizumab prior to surgery for 8 cycles. And the data for pathologic complete response were presented at San Antonio [2019 San Antonio Breast Cancer Symposium], and they really showed no difference from the addition of atezolizumab to the platinum and taxane backbone. We don’t yet have the event-free survival data to know if it will be better in patients receiving atezolizumab. But at least what we do know is that there was no difference in pCR [pathologic complete response], and there was no difference regardless of PD-L1 status. And so we don’t know why KEYNOTE-522 seemed to be positive and why NeoTRIP seemed to be negative. Is it because they used different checkpoint inhibitors? Maybe is PD-1 better than PD-L1? Is it not KEYNOTE-522 used an anthracycline and NeoTRIP did not? Is it that anthracycline works so much better synergistically with checkpoint inhibition and that’s why KEYNOTE-522 was positive and NeoTRIP was not? So I think there are many questions that still remain, but I think the initial signal from KEYNOTE-522 is very promising, and I think there likely will be a role for checkpoint inhibition in patients with early triple-negative breast cancer.

In the metastatic setting, I think 1 question that arises is, What’s the optimal chemotherapy backbone to give with checkpoint inhibition? Certainly, IMpassion130 looked at nab-paclitaxel. And I think nab-paclitaxel was selected because it does not require steroid premedication, and there is some thought that use of steroids may hinder benefit from checkpoint inhibition by suppressing the immune response. Nab-paclitaxel seemed like an ideal chemo backbone, but we truthfully don’t know if there would be a difference if we had given paclitaxel even if steroid premedication was required.

The investigators in IMpassion131 actually explored this question by giving paclitaxel with or without atezolizumab to see if a very similar benefit is seen in this setting, again with using paclitaxel instead of the nab-paclitaxel regimen. I think we’ll have to wait on those data to see if we can swap out the different taxanes if the benefit would be the same. But I think another challenge is that many patients with triple-negative breast cancer recur soon after their adjuvant taxane. And so paclitaxel doesn’t seem like the idea chemo backbone in these patients, so I think we’re also looking to see if other chemotherapy backbones may also have benefit with checkpoint inhibition.

IMpassion132 is also looking at this question in patients who have early relapse after their adjuvant taxane and looking at the CARBO [carboplatin]/GEM [gemcitabine] backbone or capecitabine with or without atezolizumab. I think that will also be important data, particularly again, when you don’t want to be giving a taxane. And certainly, KEYNOTE-355 allowed choice of chemo backbone with either carboplatin/gemcitabine/paclitaxel or nab-paclitaxel. Those data will also be very important to flush out differences in choice of chemo backbone and impact on efficacy.

While there is definitely a signal for benefit of checkpoint inhibition that has emerged from KEYNOTE-522 in the neoadjuvant setting, I think the question will also remain, Would benefit be seen in the adjuvant setting? So, if someone went to surgery first and then went on to get chemotherapy, would adding checkpoint inhibition to chemotherapy in this setting make a difference?

And I think there’s some biological differences between administering checkpoint inhibition in the neoadjuvant setting as opposed to the adjuvant setting where many of us wonder if there could be differences in efficacy. Because in the neoadjuvant setting, you still have your tumor in place, and so potentially as you’re killing tumor and you’re seeing more neoantigen expression and could potentially see a higher immune response. Whereas, in the adjuvant setting, the tumor is removed and you’re looking to kill micrometastatic disease. And so it’s unclear if you’d see the same sort of synergy between chemotherapy and immunotherapy in that setting.

There are trials exploring both of these settings, and so IMpassion030 is looking at adding checkpoint inhibition to adjuvant chemotherapy—taking ACT [anthracycline-cyclophosphamide and taxane] chemotherapy with or without atezolizumab. That trial will help us better understand if adding atezolizumab to standard ACT chemo will improve outcomes. And there’s also another preoperative trial that’s looking at ACT with or without atezolizumab. That trial has completed accrual and we’re waiting on data to also see if in the neoadjuvant setting with an anthracycline backbone, if adding atezolizumab is having benefit. So, again, I think there are many unanswered questions that remain, but large numbers of studies have looked to address these issues. And so a lot of data are pending.

I think in the early disease setting, I’m really hoping that we will see a statistically significant event-free survival benefit from KEYNOTE-522, and if this is the case, I think we’d also want to know whether event-free survival is different based on PD-L1 status. So hopefully, we’ll be able to better understand that and see if there are any biomarkers that can help us select patients for checkpoint inhibition in the early disease setting. And I hope in the future, we’ll be able to see if we can de-escalate the chemotherapy backbone for our patients and as we’re adding checkpoint inhibition, to potentially lessen toxicity.

Transcript Edited for Clarity