Centre of Experimental Cancer Medicines
Bart's Cancer Institute
St Bartholomew Breast Cancer Centre
London, United Kingdom
Peter Schmid, MD, PhD, has disclosed that he has received consulting fees from AstraZeneca, Bayer, Boehringer Ingelheim, Celgene, Eisai, Merck, Novartis, Pfizer, Puma, and Roche and funds for research support to his institution from Genentech, Novartis, OncoGenex, and Roche.
Triple‑negative breast cancer (TNBC) is an aggressive subtype of breast cancer that more frequently affects younger women, with patients often having early disease recurrence. Chemotherapy has been the mainstay of treatment for TNBC both in the early and advanced settings, and until recently, there were no targeted therapy options for the treatment of this disease. This has changed recently with the approval of atezolizumab plus nab-paclitaxel for the treatment of patients with PD-L1–positive metastatic TNBC based on the phase III IMpassion 130 trial. Given the benefit of immune checkpoint inhibitor therapy in patients with metastatic TNBC, it has become a clear priority to determine whether immunotherapy can be used to improve outcomes for patients with early-stage disease.
Neoadjuvant Chemotherapy for Early-Stage TNBC and Pathologic CR
In early-stage TNBC, we have increasingly shifted from giving adjuvant to neoadjuvant chemotherapy. Giving chemotherapy before surgery allows us to identify a group of patients who achieve a pathologic CR (pCR)—the complete disappearance of all invasive cancer from the breast and the lymph nodes. These patients have a favorable outlook and low risk of recurrence, with 3‑year to 5‑year recurrence‑free survival rates of 90% to 95%. Patients who do not achieve pCR and have residual cancer at the time of surgery have a higher risk of recurrence, even if all tumor tissue is surgically removed, with 3-year to 5-year recurrence rates of 40% to 50%.
To improve outcomes and work toward increased cure rates, there is an urgent need to identify effective postsurgery strategies for patients who do not achieve pCR with neoadjuvant therapy as well as to increase the pCR rate for patients with early-stage TNBC in general. Several small phase II studies have consistently shown that the addition of immune checkpoint inhibitors to neoadjuvant chemotherapy improves pCR rates in patients with early-stage TNBC, providing the basis for the large, randomized phase III KEYNOTE-522 trial.
Phase III KEYNOTE‑522: Neoadjuvant Chemotherapy With or Without Pembrolizumab for Early-Stage TNBC
In KEYNOTE-522, 1174 patients with previously untreated stage II (≥ 2-cm tumor, node negative) or stage III (any size tumor, node positive) TNBC were randomized to receive either pembrolizumab or placebo in addition to neoadjuvant carboplatin/paclitaxel for 4 cycles and doxorubicin/cyclophosphamide or epirubicin/cyclophosphamide for 4 cycles followed by surgery and continuation of single-agent pembrolizumab or placebo in the adjuvant setting for approximately 1 year. Regarding trial design, we deliberately chose what we felt was the most effective chemotherapy regimen to see if the addition of immunotherapy could achieve a benefit on top of this. The primary endpoints were pCR and event-free survival (EFS), both by local review.
KEYNOTE-522 is a positive trial for the primary endpoint of pCR, showing a significant and, in my opinion, clinically meaningful pCR benefit with the addition of pembrolizumab to neoadjuvant therapy. For the first 602 patients randomized, the pCR rate with pembrolizumab was 64.8%, which is the highest pCR rate described in a phase III trial in early-stage TNBC to date, vs 51.2% with chemotherapy alone (Δ 13.6%; 95% CI: 5.4% to 21.8%; P = .00055). What I think is remarkable is that this benefit was seen above what is already probably the most effective conventional chemotherapy used in the treatment of early-stage TNBC. Furthermore, patients at higher risk appeared to benefit even more with immune checkpoint inhibition than those with lower risk (Δ 20.6% for node-positive disease and Δ 24.6% and Δ 25.6% for stage IIIA or IIIB disease, respectively).
Of note, patients who were PD-L1 negative appeared to derive a similar benefit to those who were PD-L1‑positive, which is in contrast to the advanced TNBC setting where benefit with immune checkpoint inhibitor therapy was limited to patients with PD-L1–positive tumors. The reasons for this difference are unclear, but my personal hypothesis is that tumors in the early disease setting have a substantially higher plasticity and that neoadjuvant chemotherapy is potentially upregulating PD-L1 expression in what was originally determined to be a PD-L1–negative tumor on pretreatment biopsy. However, it should be noted that different types of PD-L1 testing were used in each of these studies.
There was also a clear, emerging separation of the EFS Kaplan-Meier curves for the pembrolizumab vs placebo arms at the first interim analysis (HR: 0.63; 95% CI: 0.43-0.93), with an 18-month EFS rate of 91.3% with pembrolizumab and 85.3% with placebo. Although EFS has not reached the strict prespecified boundary for significance of .000051 (HR < .04), at this early time point with a median follow-up of just 15.5 months, these data are incredibly encouraging.
Of importance, no unexpected safety issues emerged with the use of immune checkpoint inhibition in this setting and the adverse event profile was dominated by chemotherapy-related toxicities.
Phase III NeoTRIPaPDL1: Neoadjuvant Chemotherapy With or Without Atezolizumab in High-Risk, Early-Stage TNBC
The phase III NeoTRIPaPDL1 trial randomized 280 patients with high-risk, early-stage TNBC to receive the immune checkpoint inhibitor atezolizumab in combination with neoadjuvant carboplatin/nab-paclitaxel chemotherapy vs neoadjuvant chemotherapy alone. The primary endpoint was EFS at 5 years and the key secondary endpoint was pCR rate.
NeoTRIPaPDL1 was a negative trial for the secondary endpoint of pCR rate, with a pCR rate of 43.5% with atezolizumab plus carboplatin/nab-paclitaxel and 40.8% with chemotherapy alone (Δ 2.63%; odds ratio: 1.11; 95% CI: 5.40.69-1.79; P = .66). The primary endpoint of EFS has not been reported.
This trial certainly raises questions and it is therefore important to look at how it differed as compared to KEYNOTE-522. A key concern is that NeoTRIPaPDL1 was a small study and we are looking at a secondary endpoint which was reported before the primary endpoint has been analyzed. It also seems that many patients in NeoTRIPaPDL1 had locally advanced disease. One could therefore wonder if the lack of a pCR benefit with addition of atezolizumab to neoadjuvant chemotherapy in NeoTRIPaPDL1 as compared with the pCR benefit observed in KEYNOTE-522 was because enrolled patients were of higher risk. However, KEYNOTE‑522 showed that patients with higher risk benefited even more from the addition of pembrolizumab to neoadjuvant chemotherapy than their lower-risk counterparts. Another possible explanation might be that the chosen chemotherapy regimen of carboplatin/nab-paclitaxel in the absence of an anthracycline component in NeoTRIPaPDL1 had a suboptimal interaction with the immunotherapy.
Clinical Implications—My Thoughts
When both the NeoTRIPaPDL1 and KEYNOTE‑522 studies were designed, we hypothesized that seeing a difference in short-term response, such as pCR rate, might be more difficult to demonstrate, but that differences would be more likely to emerge over a longer time frame, thereby providing the rationale for EFS as an endpoint.
In fact, the experience with checkpoint inhibitors across metastatic disease settings has shown that relatively small changes in response endpoints can be followed by substantial benefits in time endpoints such as OS, as seen in the IMpassion 130 trial. Thus, despite the negative outcome for pCR in the NeoTRIPaPDL1 study, it is possible that this trial will still be positive for EFS, which is what the trial was designed and powered to do.
This also emphasizes why I am so optimistic about the KEYNOTE-522 data. We saw a substantial benefit in the endpoint that we thought would be more difficult to achieve—pCR —as well as a very encouraging emerging signal for EFS that, although we can only speculate, I hope will eventually show a significant long-term benefit.
We very clearly have the first positive phase III trial of immunotherapy in early-stage TNBC with KEYNOTE-522, which establishes the addition of pembrolizumab to neoadjuvant chemotherapy as a promising treatment option that has the potential to change practice and become a new standard of care once approved. Furthermore, there are ongoing phase III trials that apply a similar design to KEYNOTE‑522 and will, hopefully, further substantiate these results.
Future studies will help us clarify our understanding of the impact of adjuvant therapy with pembrolizumab and the potential role of other chemotherapy backbones in the treatment of our patients with early-stage TNBC.
In your practice, do you discuss investigational immunotherapy options with your patients who have early-stage TNBC?I encourage you to answer the polling question and post your questions and thoughts in the discussion box below.
You are accessing CCO's educational content today as a Guest user.
If you would like to continue with free, full access to the CCO Web sites, including free CME/CE credits, please click the button below.