Pembro + Chemo in Early TNBC: From the Trial to the Clinic

Liam Davenport

September 23, 2021

Even once a therapy has been approved for the clinic there are often a number of outstanding questions about how it should be applied in practice, and the addition of pembrolizumab (Merck) to chemotherapy in early stage triple negative breast cancer (TNBC) is no exception.

Earlier this year, the US Food and Drug Administration approved the neoadjuvant use of pembrolizumab in combination with chemotherapy for patients with high-risk early-stage TNBC, and as single-agent adjuvant treatment to be continued after surgery

This was based on findings from the randomised phase 3 KEYNOTE 522, which showed that the addition of pembrolizumab to chemotherapy in 1174 TNBC patients significantly prolonged event-free survival (EFS) versus chemotherapy plus placebo in previously untreated stage II or III TNBC.

The latest results from the trial, presented at a virtual plenary session of the European Society for Medical Oncology (ESMO) meeting in July and published in the Annals of Oncology, revealed that, after 37.8 months of follow-up, 15.7% of patients in the pembrolizumab group had disease progression versus 23.8% in the placebo group.

Although follow-up is ongoing, pembrolizumab had a favourable overall survival trend, with a hazard ratio of 0.72, and no significant increase in grade ≥3 adverse events, at 77.1% versus 73.3% in the placebo arm.

NICE Review

Now, while the National Institute for Health and Care Excellence examines the clinical and cost effectiveness of pembrolizumab for the neoadjuvant treatment of TNBC, the implications of the results for clinical practice were discussed at the ESMO Congress 2021 on September 17.

Crucially, lead investigator Dr Peter Schmid, from Barts Cancer Institute, Queen Mary University of London, also had the opportunity to respond.

Opening the discussion, Dr Lisa Carey, from the UNC Lineberger Comprehensive Cancer Center, Chapel Hill, North Carolina, USA, said that the study design was “fairly straightforward”.

She noted that “the patient population was generally fairly characteristic of a triple negative population”, with almost 75% having T1/T2-3 disease and approaching 50% being node negative.

“Not unexpectedly,” she continued, “you do get immune-mediated adverse events” when adding pembrolizumab, with the trial revealing a doubling of any grade immune events. Fifteen percent of patients given pembrolizumab experienced grade 3–5 toxicity, including two deaths.

She reminded the audience that, in July, Dr Schmid and the discussant at the time, Dr Giuseppe Curigliano, from the University of Milan, Italy, concluded that immune-related toxicities were “usually but not always low grade, and definitely needed to be managed”.

Management is typically via drug interruption, steroids and hormone replacement therapy, “but it’s worth noting that some will be irreversible or life-threatening”, Dr Carey said.

In addition, the “added expense” of giving patients pembrolizumab routinely is “going to be considerable [and] is going to be an obstacle for many countries or individuals”.

Consequently, “defining the optimal population to maximise benefit and minimise toxicity is crucial”.

Subgroup Analysis

The subgroup analysis in KEYNOTE 522 showed that the results favoured pembrolizumab in “virtually every” patient group.

The drug received accelerated approval in November 2020 for adjuvant use in locally recurrent unresectable or metastatic TNBC for patients whose tumours express programmed cell death ligand-1 (PD-L1). “It is also worth noting that PD-L1 itself was not predictive of benefit of pembrolizumab, unlike in the metastatic setting”, Dr Carey said.

Additional clues to which patients benefit the most from immunotherapy in this setting can be seen when stratifying the EFS results by pathologic complete response (pCR) status.

Dr Carey explained that patients who had a pCR had “very good outcomes, greater than 90% EFS, regardless of whether they received pembrolizumab or not”.

“Conversely, those with residual disease had quite a poor outcome, and quite a substantial improvement in [EFS] with the addition of pembrolizumab.”

Over on Twitter, Professor Aleix Prat, from Hospital Clínic de Barcelona, Barcelona, Spain, asked whether this means there are patients “destined” to have residual disease after chemotherapy, but the addition of pembrolizumab “allows them to achieve a pCR and an outstanding outcome”.

“Who are these patients at diagnosis?”

This point was picked up by Dr Bryan Vaca-Cartagena, an oncologist at Instituto Tecnológico Y De Estudios Superiores De Monterrey, Monterrey, México.

He asked the logical next question: If there is no pCR, “what is the best treatment”?

Dr Carey said that, taken together, the results show there are a number of outstanding questions over the use of pembrolizumab in early stage TNBC, the first being: Who benefits from immunotherapy, and who does not?

She underlined that the clinical variables studied in the trial did not identify a patient group “without benefit”, and so “we badly need predictive biomarkers”.

In terms of who does not benefit, she noted that T1N0 patients were not included in the trial, and there was “little benefit” of immunotherapy in patients with pCR.

However, the effectiveness of the regimen “may derive from the neoadjuvant combination” of chemotherapy and immunotherapy acting synergistically, and so it may not be possible to “salvage” patients without a pCR to chemotherapy by giving them immunotherapy only in the later, adjuvant setting.


The next outstanding question is who is likely to have toxicity and how best to ameliorate that.

“This is a very open-ended question and there is a lot of research going on,” Dr Carey said.

Other studies have suggested that baseline or on-treatment serum cytokines may predict toxicity, and if that were the case, there are treatments available to optimise management of immune-related toxicity.

“It’s also worth noting that toxicity comes from the chemotherapy regimen,” so the question is: “Do we need all four drugs, as was given in KEYNOTE 522?”

Following on from that: “Is the focus on minimising chemotherapy, or on minimising the immune checkpoint inhibitor? We may not be able to do both.”

The final big question is over optimising the adjuvant phase of treatment, Dr Carey said.

This includes whether adjuvant immunotherapy, as given in this trial, is needed at all, especially as adjuvant therapy on its own has, so far, achieved “inadequate” outcomes.

In response, Dr Schmid said the idea behind KEYNOTE 522 was to demonstrate whether an immune checkpoint inhibitor would add a benefit to “what we would consider the most effective chemotherapy backbone”.

He suggested that “less effective chemotherapy” might have resulted in lower pCR rates, even if the added benefit of immunotherapy was the same, and it is “unclear” whether the chemotherapy can be de-escalated.

Addressing the impact of pCR on EFS, Dr Schmid said it is “important to point out that there are different numbers of patients in each category”, with more patients achieving pCR with the addition of the immunotherapy.

“The comparison is, strictly speaking, not randomised.”

Nevertheless, Dr Carey’s observation that pCR was predictive of better outcomes, regardless of immunotherapy, is an “important” point, not just for this trial, but also for “future drug development”.

On the question of “who benefits most” from immunotherapy, however, Dr Schmid said there are “a large number of correlative studies ongoing…and I am sure we will learn a lot from them”.

But the question is: “Are there any patients who may not need immune checkpoint inhibitors?

“The question, if you phrase it differently, means: Is there a subgroup of patients, which we can reliably identify, who will achieve a pCR with chemotherapy alone?”

Dr Schmid said they “have been working on this question for the last 20 years and have not been able to identify that group reliably, but hopefully this trial and other trials will help us”.

In terms of safety, he highlighted that the immune-related adverse events were largely endocrinopathies and skin reactions, and had a very low incidence in the adjuvant phase of the trial.

He agreed that some immune events “may be irreversible” but other studies “support the long-term safety of pembrolizumab”, with no signal for late toxicities.

However, Dr Schmid said that additional follow up will help inform understanding of the regimens’ long-time safety.

KEYNOTE 522 was funded by Merck Sharpe & Dohme Corp.

Schmid declares relationships with AstraZeneca, Pfizer, Novartis, Roche, Meck, Boehringher Ingelheim, Bayer, Eisai, Puma, Celgene, Genetech, Oncogenex, Astellas.

Carey declares no relevant financial relationships.

ESMO Congress 2021: Abstract VP7-2021. Presented September 17.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: