The addition of pembrolizumab to standard-of-care chemotherapy for neoadjuvant and adjuvant treatment of early triple-negative breast cancer lead to higher rates of pathologic complete responses in the KEYNOTE-522 trial, especially in those patients with stage III and/or node-positive disease.
Peter Schmid, MD, PhD
The addition of pembrolizumab (Keytruda) to standard-of-care chemotherapy for neoadjuvant and adjuvant treatment of early triple-negative breast cancer (TNBC) lead to higher rates of pathologic complete responses (pCR) in the KEYNOTE-522 trial, especially in those patients with stage III and/or node-positive disease.1
“We know that if you give [a patient] neoadjuvant chemotherapy of an anthracycline and a taxane, the pCR rates are about 40%, and if you add in platinum, it is about 50% to 55%,” said Peter Schmid, MD, PhD, professor of cancer medicine and lead of the Centre of Experimental Cancer Medicine at Barts Cancer Institute in London, during a presentation of the data at the San Antonio Breast Cancer Symposium in Texas. “Prior to this study, there were small trials demonstrating [that the addition of] pembrolizumab or other checkpoint inhibitors to neoadjuvant chemotherapy was safe and shows promising early activity in TNBC.”
In the phase III trial, patients with newly diagnosed, operable, stage II or III TNBC received both neoadjuvant and adjuvant systemic therapy. During the neoadjuvant phase of therapy, patients were randomized in a 2:1 ratio to received pembrolizumab at 200 mg every 3 weeks or placebo. All patients were administered carboplatin plus paclitaxel for 4 cycles followed by doxorubicin or epirubicin plus cyclophosphamide for an additional 4 cycles leading up to surgical resection. Following surgery, patients were treated with pembrolizumab or placebo for up to 1 year.
According the data presented during the European Society for Medical Oncology 2019 Congress, baselines characteristics were well balanced between the 2 groups. Median patient age was 49 years, about 75% had T1 or T2 tumors, and slightly more than half had node-positive disease. More than 80% of patients were considered to be PD-L1 positive based on a combined positive score ≥1 by PD-L1 IHC 22C3 pharmDx assay.2
Regardless of PD-L1 expression, patients in the pembrolizumab arm had a greater rate of pCR at 64.8% versus 51.2%, representing an absolute difference of 13.6% (95% CI, 5.4%-21.8%;P= .00055).1,2
“At this point in time, we have mature and definitive data for the first primary end point [of the trial], pCR, or the complete disappearance of all cancer,” Schmid said. “If you look at the definitive data based on 602 patients, there is a statistically significant and meaningful difference in pCR rates between the 2 groups.”
At the first planned interim analysis, the second primary end point of event-free survival (EFS) based on 1174 patients was 91.3% in the pembrolizumab groups versus 85.3% with placebo at 18 months (0.63; 95% CI, 0.43-0.93).2
“At a short follow-up for adjuvant [treatment], 15.5 months, we see a separation of the [Kaplan-Meier curves for EFS] but it is not statistically significant at this point, especially considering the precalculatedP-value boundary for significance of 0.000051.”
When patients were stratified by disease stage, reduction in pCR with pembrolizumab was observed consistently. However, the greatest magnitude of benefit to the addition of the PD-1 inhibitor was seen in patients with stage IIIB disease, at 48.6% versus 23.1% in those who received placebo, representing an absolute difference of 25.6% (95% CI, -6.1% to 48.9%). Corresponding rates of pCR in patients with stage IIIA disease were 66.7% versus 42.1%, representing an absolute difference of 24.6% (95% CI, 4.3%-43.1%); in stage IIB, 56.2% versus 48.4%, for an absolute difference of 7.8% (95% CI, -7.4% to 22.8%); and for stage IIA, 73.1% versus 62.1%, for a difference of 11.0% (95% CI, -0.7 to 23.2%).
Similarly, patients who were positive for lymph-node involvement were observed to have a greater pCR benefit with pembrolizumab versus placebo, at 64.8% and 44.1%, for an absolute difference of 20.6% (95% CI, 8.9%-31.9%). Patients who were negative for nodal involvement did receive benefit with the addition of pembrolizumab, but the absolute difference was 6.3% (64.9% vs 58.6%; 95% CI, -5.3% to 18.2%).
Positive PD-L1 status was not necessary for a tumor response. “In advanced disease, we see a clear difference in patients with PD-L1positive tumors and PD-L1–negative tumors in terms of the benefit from immune therapy,” Schmid said. “We don’t see this in early disease…my personal theory is that the tumor plasticity and the dynamics within the tumor biology is high in early disease.”
In patients with PD-L1 expression of <1 CPS, patients who were in the pembrolizumab group had a pCR rate of 45.3% versus 30.3% in the chemotherapy-alone group for an absolute difference of 18.3% (95% CI, -3.3% to 36.8%). For CPS ≥1, corresponding pCR rates were 68.9% versus 54.9% with a difference of 14.2% (95% CI, 5.3%-23.1%); in patients with CPS ≥10, rates were 77.9% versus 59.8% for a difference of 17.5% (95% CI, 6.2%-29.1%); and in patients with CPS ≥20, rates were 81.7% versus 62.5% for a difference of 18.5% (95% CI, 5.0%-32.7%).
“The absolute difference between chemotherapy and the combination was consistent whether you used a cutoff of CPS 1, 10, or 20,” Schmid said. “Of note, if you look at the patients with CPS ≥10 inflamed tumor we see…a pCR rate that we have never seen in such an aggressive disease group.”
Regardless of exposure to chemotherapy, patients treated with pembrolizumab had greater benefit as compared with those receiving placebo. Full exposure to chemotherapy plus pembrolizumab induced a pCR of 69.7% versus 55.3% in those treated with full chemotherapy and placebo, with an absolute difference of 14.4% (95% CI, 5.1%-23.6%). In those who received less than the full chemotherapy dose, pembrolizumab induced a pCR of 51.5% versus 35.7% with placebo, for a difference of 15.4% (95% CI, -3.0 to 32.1%).
Schmid concluded by saying that in the future, additional biomarker data, including tumor infiltrating lymphocytes andBRCA-mutational status, will be helpful in determining which patients receive the greatest benefit with this regimen.