Neoadjuvant chemotherapy is widely accepted as the standard of care for early-stage TNBC and is critical for improving patient outcomes.¹ Neoadjuvant chemotherapy helps downstage tumors, which can allow for smaller surgeries and provides additional time for germline mutational testing. Neoadjuvant treatment also allows healthcare professionals (HCPs) to get a general sense of the responsiveness of the tumor to various chemotherapeutics and can be used in various approaches to increase the likelihood of a pCR. Of importance, for patients who do not achieve a pCR, neoadjuvant chemotherapy is an opportunity to reduce the risk of recurrence. 

Some landmark clinical trials that established residual disease treatment options for TNBC include the CREATE-X trial and the OlympiA trial. The CREATE-X trial demonstrated the benefit of giving adjuvant capecitabine to improve disease-free survival and overall survival (OS) in patients with residual disease after neoadjuvant chemotherapy.² Likewise, the OlympiA trial illustrated that adjuvant olaparib significantly improved OS in patients with TNBC and a BRCA mutation.³ 

Like the CREATE-X and OlympiA trials, the KEYNOTE-522 trial was also a milestone study, supporting the FDA approval of a new perioperative regimen.

KEYNOTE-522 was a randomized, double-blind phase III trial comparing neoadjuvant pembrolizumab vs placebo with chemotherapy, followed by adjuvant pembrolizumab vs placebo for early TNBC.^4,5

Patients eligible for this study had newly diagnosed, previously untreated, nonmetastatic TNBC (tumor stage T1c and nodal stage N1-2, or tumor stage T2-4 and nodal stage N0-2, according to the American Joint Committee on Cancer staging criteria, 7th edition). Patients received the same chemotherapy backbone on each arm, consisting of carboplatin plus paclitaxel followed by an anthracycline with cyclophosphamide. Patients on the investigational arm received neoadjuvant pembrolizumab every 3 weeks for 24 weeks and adjuvant pembrolizumab every 3 weeks for 27 weeks following surgery. Patients on the comparator arm received doses of placebo on the same schedule. Patients were stratified by nodal status, tumor size, and carboplatin schedule. The primary endpoint was pCR, and secondary endpoints included OS, EFS, adverse events (AEs), and pCR by subgroups.

The baseline characteristics of the KEYNOTE-522 study population were quite well matched between both arms.⁵ Of note, this study enrolled patients regardless of their PD-L1 status, with both the pembrolizumab and placebo groups enrolling approximately the same proportion of PD-L1–positive patients: 83.7% and 81.3% on each arm, respectively.

With respect to the primary endpoint of pCR, at a median follow‑up of 15.5 months (interim analysis 1), patients who received pembrolizumab exhibited a higher rate of pCR vs patients who received placebo. The rate of pCR with pembrolizumab was 64.8% vs 51.2% with placebo, for a difference of approximately 14% in favor of pembrolizumab.⁵

Of importance, this difference was observed regardless of whether patients’ PD‑L1 status was positive or negative. Among PD‑L1‒positive patients, the difference in pCR between the pembrolizumab group and the placebo group was approximately 14%. Strikingly, in PD‑L1‒negative patients, the difference in pCR between the pembrolizumab group and the placebo group was approximately 18%.

The data from interim analysis 6, at a median follow-up of 63.1 months, also demonstrated improved EFS with pembrolizumab⁶: 5-year EFS was 81.3% among those who received pembrolizumab vs 72.3% among those who received placebo (HR: 0.63; 95% CI: 0.49-0.81). Furthermore, pembrolizumab was associated with a 7.6% increase in distant recurrence‒free survival (DRFS): 5-year DRFS was 84.4% among those who received pembrolizumab vs 76.8% among those who received placebo (HR: 0.64; 95% CI: 0.49-0.84)​.

Next, examining EFS by pCR reveals that patients who achieved a pCR exhibited improved EFS compared with patients who did not achieve a pCR, regardless of whether they received pembrolizumab.⁶ However, among the subgroups of patients with a pCR and patients without a pCR, pembrolizumab was associated with better EFS compared with placebo. Specifically, 5-year EFS among those with a pCR who received pembrolizumab was 92.2% vs 88.2% among those who received placebo (HR: 0.65; 95% CI: 0.39-1.08). Similarly, among those who did not achieve a pCR, 5-year EFS was 62.6% with pembrolizumab vs 52.3% with placebo (HR: 0.72; 95% CI: 0.54-0.96)​.

The pCR and EFS data to date are highly promising, but the OS data remain immature. Investigators reported an estimated 36-month OS of 89.7% for patients receiving pembrolizumab and 86.9% for patients receiving placebo, with the median OS not yet reached on either arm.⁴

The EFS benefit associated with pembrolizumab persisted through the subgroup analysis of this study. Pembrolizumab was favored across all subgroups, regardless of nodal status or tumor size.

Of particular importance, 5-year EFS was higher with pembrolizumab irrespective of PD‑L1 status, with an HR of 0.64 (95% CI: 0.48-0.85) for patients with PD-L1‒positive tumors and an HR of 0.57 (95% CI: 0.33-0.98) for patients with PD-L1‒negative tumors.⁶

Immune-mediated AEs can occur at any time while receiving immunotherapy and even can occur after discontinuing immunotherapy. Across both the neoadjuvant and adjuvant phase in KEYNOTE-522, AEs of any grade occurred in just under 44% of patients receiving pembrolizumab, and grade 3-5 AEs occurred in 14.9% of patients. In comparison, the rates of any-grade AEs and grade 3-5 AEs for patients receiving placebo were 21.9% and 2.1%, respectively.⁷

Across both phases, the most common immune-mediated AEs were infusion reactions, hypothyroidism, skin reactions, hyperthyroidism, adrenal insufficiency, pneumonitis, thyroiditis, hypophysitis, colitis, and hepatitis.⁴

In just the adjuvant phase, patients receiving pembrolizumab experienced immune-mediated AEs of any grade at a rate of 10.2% vs 6.0% for patients receiving placebo. Grade 3-5 AEs occurred in 2.9% of patients receiving pembrolizumab vs 0.3% of patients receiving placebo.

The most common immune-mediated AEs in the adjuvant phase were hypothyroidism, infusion reactions, skin reactions, pneumonitis, hyperthyroidism, adrenal insufficiency, colitis, and myocarditis.

Ultimately, across both phases, AEs that led to discontinuation of any study drug occurred in 10.9% of patients receiving pembrolizumab vs 2.6% of patients receiving placebo. In the adjuvant phase, AEs leading to discontinuation of treatment occurred in 1.4% of patients receiving pembrolizumab vs 0.3% of patients receiving placebo.

Now, having extensively covered the landmark study that led to FDA approval of neoadjuvant pembrolizumab for TNBC, I would like to highlight several important studies of neoadjuvant pembrolizumab in other settings. The phase II NeoPACT study sought to determine whether anthracycline is necessary in treating early‑stage TNBC or whether a deintensified anthracycline‑free neoadjuvant regimen also could be robustly effective.⁹ Although anthracyclines are highly active for treatment of TNBC, they are associated with increased risk for serious complications—namely, secondary leukemia/myelodysplastic syndrome and cardiomyopathy. The severity of these potential AEs prompted investigators to explore the possibility of developing an anthracycline-free neoadjuvant therapy with pembrolizumab.

In this study, patients received pembrolizumab plus chemotherapy consisting of docetaxel and carboplatin once every 21 days for 6 cycles prior to surgery for stage I, II, or III TNBC. So far, the results are promising. The 3‑year EFS in patients who achieved a pCR was 98%, and the 3‑year EFS in patients who did not achieve a pCR was 68%. Similarly, investigators reported a robust estimated 3-year OS rate of 100% in patients with a pCR and 79% in patients without a pCR.

As noted in our case study, it is unclear how much benefit adjuvant pembrolizumab confers for patients who achieve a pCR with neoadjuvant chemotherapy plus pembrolizumab. The phase III OptimICE‑PCR trial seeks to determine whether adjuvant pembrolizumab, given as 9 doses in the current KEYNOTE‑522 regimen, confers additional benefit for patients who achieved a pCR following neoadjuvant chemotherapy with pembrolizumab (NCT05812807).

In the KEYNOTE-522 study, patients who achieved a pCR after neoadjuvant therapy experienced superb outcomes, regardless of whether they received pembrolizumab during the adjuvant phase. Although patients receiving pembrolizumab still had a higher 5-year EFS vs placebo (92.2% vs 88.2%), the data raise the question of whether achieving a pCR during the neoadjuvant phase is sufficient to set patients up for success or whether adjuvant pembrolizumab conveys any additional benefit after a pCR.⁶

The OptimICE‑PCR study is currently enrolling adult patients with nonmetastatic early-stage TNBC (tumor stage T1c and nodal stage N1-2, or tumor stage T2-4 and nodal stage N0-2, according to the American Joint Committee on Cancer staging criteria, 7th edition) who have no residual disease in either the breast or lymph nodes following completion of neoadjuvant therapy. Patients are required to have completed at least 6 cycles of neoadjuvant chemotherapy with pembrolizumab and must be HER2 negative, with estrogen and progesterone ≤10%. Participants are randomized to receive pembrolizumab or undergo observation within 12 weeks of surgery. The primary endpoint of this study is relapse-free survival, with secondary endpoints including safety, OS, and locoregional recurrence. This study is ongoing, with an estimated primary completion date in 2033.

Next, I would like to discuss 2 phase III clinical trials of neoadjuvant ICI-based regimens for high risk estrogen receptor (ER)‒positive/HER2-negative EBC.

The first, KEYNOTE-756, is a phase III trial that also utilizes pembrolizumab, this time comparing neoadjuvant chemotherapy plus pembrolizumab or placebo followed by adjuvant endocrine therapy plus pembrolizumab or placebo.¹⁰ The patients enrolled on this study have ER-positive/HER2-negative stage II/III invasive ductal breast carcinoma and are randomized to receive either neoadjuvant pembrolizumab or placebo plus paclitaxel for 4 cycles, followed by 4 cycles of pembrolizumab or placebo plus cyclophosphamide with either doxorubicin or epirubicin. After surgery, patients were treated with adjuvant pembrolizumab or placebo for 6 months, with endocrine therapy for up to 10 years. The coprimary endpoints of this study are the rate of pCR and EFS. Although the EFS data remain immature, the rate of pCR was significantly higher with pembrolizumab vs placebo at 24.3% vs 15.6%, respectively (difference: 8.5%; 95% CI: 4.2%-12.8%; P = .00005). So far, no new safety signals were observed for pembrolizumab in this setting. This study is ongoing, but positive final EFS data would provide strong evidence for pembrolizumab in the setting of ER-positive/HER2-negative EBC.

The next study, CheckMate 7FL, is a randomized phase III clinical study assessing neoadjuvant nivolumab for high-risk ER-positive/HER2-negative EBC (NCT04109066).¹¹ To be eligible for this study, patients must have stage II/III EBC (grade 3 with ER expression ≥1% or grade 2 with ER expression 1%-10%). In the neoadjuvant phase, patients received 4 cycles of nivolumab or placebo plus paclitaxel, followed by 4 cycles of nivolumab or placebo with cyclophosphamide plus either doxorubicin or epirubicin. After surgery, patients received 7 cycles of nivolumab or placebo plus standard-of-care endocrine therapy. Originally, the coprimary endpoints of this study were pCR and EFS. However, following approval of adjuvant abemaciclib in this setting, investigators anticipated that patients would withdraw to access abemaciclib, which cannot be coadministered with nivolumab. Considerable patient withdrawals would jeopardize the EFS analysis, so the study design was altered to focus solely on pCR. Following this change, the study met its primary endpoint, with patients receiving nivolumab demonstrating a 24.5% pCR rate vs 13.8% of patients receiving placebo. These results corroborate the findings of the KEYNOTE-756 trial, demonstrating the benefit of adding an anti–PD-1 agent to standard neoadjuvant therapy for high-risk ER-positive/HER2-negative EBC. No new safety signals were reported.

Of note, subgroup analyses from both trials showed that patients with high PD-L1 expression benefited the most from ICI therapy.

In spite of the clinical benefits associated with ICIs, these agents also are associated with a spectrum of irAEs.¹² Potential AEs can range from more common endocrinopathies, such as hypothyroidism, to cardiotoxicity, pneumonitis, ocular toxicity, colitis, and less common endocrinopathies such as hypophysitis and autoimmune diabetes.

AEs related to immunotherapy can occur at any time, including after just 1 dose of immunotherapy, but there are consistent patterns for when certain immunotoxicities are more likely to occur.¹³ For example, cutaneous immunotherapy toxicity tends to occur earlier, whereas renal immunotherapy toxicity tends to occur later. Knowing these patterns can help HCPs implement strategies to prevent or mitigate these treatment-related AEs.

Many of the more common irAEs already have established workup strategies. For example, if a patient—such as the one in our patient case—seems to have immunotherapy colitis, the National Comprehensive Cancer Network guidelines recommend obtaining an abdominal/pelvic CT with contrast, performing stool evaluation to evaluate for infectious etiologies, and considering a gastroenterology consultation.¹⁴

If there is a concern for hepatitis, guidelines advise performing an abdominal ultrasound; evaluating for potential viral etiologies, cancer-related etiologies, or other potential drug-related etiologies; and stopping any other potentially hepatotoxic medications or exposures.

If the patient has potential nephritis, imaging is advised to rule out potential obstructive uropathies, as is urine testing with the involvement of a nephrologist.

In instances where there is a concern for pneumonitis, the guidelines suggest a chest CT with contrast and selectively performing infectious workup. If symptoms are more significant (ie, grade ≥3), a bronchoscopy should be part of the evaluation, in collaboration with HCPs specializing in pulmonology.

When dealing with rarer and potentially more dangerous immune-related toxicities, it is important to consider them in the differential for patients with a history of ICI exposure. For example, hypophysitis is rare, but if there is concern for that, brain MRI should be performed.¹⁴ 

Myasthenia gravis is also rare but should be considered if a patient has facial, ocular, or bulbar weakness. A brain MRI should be done to rule out central nervous system disease, along with consulting a neurologist.

Myocarditis also is quite rare and can be evaluated with a cardiac MRI. The guidelines also advise involving cardio-oncology colleagues. In many cases, patients experiencing myocarditis need to be hospitalized for close monitoring on telemetry and echocardiogram and may sometimes require cardiac catheterization and/or myocardial biopsy. 

Lastly, myositis can be evaluated using MRI, along with blood workup, including creatinine kinase. Patients suspected of having myositis may need to be further evaluated using electromyography, muscle biopsy, and myositis antibodies.

There are also established general guidelines for management of irAEs. In general, if a patient has grade 1 toxicity related to immunotherapy with asymptomatic to mild symptoms, the recommendation is to observe and monitor while continuing immunotherapy.^14,15

If a patient has moderate grade 2 symptoms, the guidelines recommend considering holding immunotherapy and reintroducing it if the AE resolves expeditiously to grade ≤1 toxicity. HCPs also may consider low-dose steroids.

Grade 3 toxicity is medically significant but not immediately life-threatening, so the recommendation is to hold immunotherapy and start high-dose steroids with a slow taper once symptoms have resolved to grade ≤1. If symptoms do not improve within 1-3 days, HCPs should increase immunosuppression.

Finally, grade 4 toxicity is considered life-threatening, where urgent intervention is indicated and immunotherapy should be permanently discontinued, perhaps except for select cases of endocrinopathies.

In instances of steroid-refractory AEs, the guidelines advise adding infliximab or mycophenolate mofetil.

In addition to steroids and mycophenolate, other immunomodulatory agents are used to manage AEs from immunotherapy.¹⁵ In particular, biologic agents—such as abatacept, rituximab, infliximab, tocilizumab, and vedolizumab—may be useful.

Finally, I will summarize approaches to the treatment of early‑stage TNBC depending on the type of disease present at treatment initiation.¹⁶ For patients with tumor stage T1c and nodal stage N0 breast cancer, therapy typically is approached with neoadjuvant treatment, often composed of a taxane plus a platinum agent or docetaxel, doxorubicin, and cyclophosphamide vs docetaxel and cyclophosphamide, followed by surgery. If the patient has a pCR in that instance, further treatment is not required. If the patient does not have a pCR and does not have any germline mutations, then the guidelines advise adjuvant therapy with capecitabine, per evidence from the CREATE‑X trial.² Currently, there is no clear guidance on whether pembrolizumab should be used in the adjuvant setting in such cases if the patient did not receive neoadjuvant pembrolizumab. If the patient in this circumstance does not have a pCR but has a germline BRCA mutation, then olaparib would be the preferred treatment.

Now, if the patient has a tumor that is at least stage T2 or any lymph node involvement, the first course of action should be neoadjuvant treatment using the KEYNOTE‑522 regimen (carboplatin with a taxane followed by an anthracycline with cyclophosphamide, plus pembrolizumab), followed by surgery. If the patient achieves a pCR, the next step is to complete 1 year of adjuvant pembrolizumab therapy, per the KEYNOTE‑522 regimen.⁶ If the patient does not have a pCR and does not have a BRCA mutation, the recommended adjuvant treatment is capecitabine in conjunction with pembrolizumab. If there is a BRCA mutation, then the adjuvant treatment should consist of pembrolizumab and olaparib.

HCPs may wonder, what should one do for a patient with residual disease after completing the KEYNOTE‑522 regimen? In the adjuvant setting, based on standard of care, I would treat this patient with adjuvant pembrolizumab based on the results of the KEYNOTE‑522 study. In addition, I would treat this patient with capecitabine based on the CREATE‑X study, and I would feel comfortable using those 2 agents together, although that was not done in the KEYNOTE‑522 study. Based on the results of the CREATE-X study, capecitabine is associated with an OS benefit in patients who have residual disease, and thus I feel it is important to include it in treatment for patients who have residual disease who do not have BRCA mutations. If a patient had a BRCA mutation and residual disease, then instead of capecitabine, I would favor the use of olaparib based on the OlympiA data.³

What if a patient is treated with the KEYNOTE‑522 regimen but later develops metastatic disease? Could they receive immunotherapy again? There are no robust data in this area given that the KEYNOTE‑522 regimen was incorporated into standard of care practices only recently. However, based on the KEYNOTE‑355 study, HCPs currently treat patients with metastatic disease if they are PD‑L1 positive, defined as a combined positive score ≥10.¹⁷ So, if a patient was PD‑L1 positive and developed metastatic disease, and an appreciable amount of time had elapsed from when they were treated in the curative setting with the KEYNOTE‑522 regimen—that is, they had not recently received immunotherapy—then I would selectively consider using immunotherapy in conjunction with chemotherapy in the metastatic setting. However, there currently is no clear guideline on what that time frame would be. I certainly would consider it for a patient if they experienced a relapse more than 1 year after completing treatment and would selectively consider it for patients less than 1 year out from treatment completion.