Early triple-negative: pembrolizumab before surgery

A 41-year-old teacher notices a hard lump in her left breast while showering. Mammography and ultrasound show a 3-centimeter lesion and an enlarged axillary lymph node. Core needle biopsy confirms invasive carcinoma of no special type, grade 3, with negative estrogen and progesterone receptors and negative HER2. Fine needle aspiration of the lymph node confirms metastasis. Staging tests show no distant disease. She asks whether she can have surgery right away and leave the rest for later, because she is afraid of "giving the tumor time to grow". The scene is fictional and was written for teaching.

This text uses the case as a setting for clinical reasoning for students and health professionals. The question comes from Dr. Alexei's list for the oncology series: who should receive systemic treatment before surgery in early triple-negative breast cancer, and what did the KEYNOTE-522 trial change in that decision? The trial tested pembrolizumab with chemotherapy before surgery and alone after it. The text does not replace your country's guideline or your institution's protocol. If you play the case of the day on NEXO, you will recognize the challenge: the order of the steps matters as much as the choice of each one.

Key points

• In stage II or III triple-negative disease, chemotherapy before surgery is the preferred path, because it treats early and shows the response in the specimen.
• KEYNOTE-522 added pembrolizumab to a regimen with carboplatin, paclitaxel and anthracycline and increased pathologic complete response.
• The trial also showed a gain in event-free survival and overall survival.
• PD-L1 was not required for entry into the trial, and the test does not decide who receives the regimen.
• Immune-mediated events, especially endocrine ones, can be permanent and require active monitoring.
• Those who end with residual disease have a worse prognosis and need an adjuvant plan discussed case by case.

Why this is a reasoning scenario

Triple-negative breast cancer is defined by what is missing. There is no hormone receptor to block and no HER2 to attack with an antibody. For years, chemotherapy was the only systemic weapon, and the disease gained a reputation for being aggressive, with relapses concentrated in the first years after diagnosis. This biology explains why the question "do I operate right away or treat first?" is so frequent in the consultation.

The instinct of the patient in the scene is understandable. She imagines that every week with the tumor in her body is a week of risk. Clinical reasoning needs to show the opposite: chemotherapy before surgery is already treatment, it reaches micrometastases from the first cycle and it allows you to see, in the surgical specimen, whether the tumor responded. That information changes what comes next.

Pathologic complete response, defined as absence of invasive disease in the breast and lymph nodes, is the central marker of this reasoning. In a meta-analysis with 27 studies and 9,460 patients, triple-negative disease had a chance of complete response about three times greater than the other subtypes, with an odds ratio of 3.02 (Wu et al., 2014). In the same analysis, the complete response rate in triple-negative disease was around 29 in every 100 patients with the chemotherapy of that time (Wu et al., 2014). KEYNOTE-522 started from that point and asked whether immunotherapy would push that number higher.

The mistake the scenario aims to prevent

The most common mistake is not in the choice of regimen. It is in operating on a patient with stage II or III triple-negative disease without first discussing the case in a multidisciplinary meeting. After surgery, the opportunity to measure the response disappears, and the adjuvant plan is then decided with less information. The post on who decides the treatment sequence in breast cancer shows how to organize that conversation.

Who should receive treatment before surgery

Dr. Alexei's question calls for a clear criterion. In KEYNOTE-522, patients were enrolled with untreated triple-negative disease, with a T1c tumor and N1 or N2 lymph nodes, or with a T2 to T4 tumor and any N from 0 to 2 (Takahashi et al., 2026). In office language, that means a tumor larger than 2 centimeters or any tumor with involved lymph nodes.

Criteria that favor neoadjuvant treatment

• Tumor larger than 2 centimeters, even with a clinically negative axilla.
• Axillary lymph node involvement confirmed by fine-needle aspiration or biopsy.
• Desire to preserve the breast when the tumor size would make breast-conserving surgery difficult.
• Need for prognostic information to guide treatment after surgery.
• Clinical condition that allows intensive chemotherapy for about six months.

When surgery first still makes sense

Very small tumors, without involved lymph nodes, were left out of the study, and the decision in these cases depends on the local guideline. Patients with active autoimmune disease requiring systemic treatment were also left out, a condition that contraindicates immunotherapy and must be investigated at the first visit (Takahashi et al., 2026). Advanced age, frailty and cardiac comorbidities weigh on the decision, because the regimen includes anthracycline and carboplatin.

Tumor size and lymph node status need to be well documented before the first cycle. Marking the involved lymph node, with a clip or another marker, makes axillary surgery after the response easier. This care is part of planning, not a technical detail.

What KEYNOTE-522 tested

KEYNOTE-522 was a phase III, double-blind study with 2-to-1 randomization between pembrolizumab and placebo, both added to the same chemotherapy (Takahashi et al., 2026). The design was long and requires attention to doses, because the full regimen lasts about a year.

Doses and sequence

In the neoadjuvant phase, patients received pembrolizumab 200 mg every 3 weeks, with four cycles of paclitaxel 80 mg/m² weekly and carboplatin AUC 5 every 3 weeks or AUC 1.5 weekly (Takahashi et al., 2026). Then came four cycles of doxorubicin 60 mg/m² or epirubicin 90 mg/m², with cyclophosphamide 600 mg/m², every 3 weeks, keeping pembrolizumab or placebo (Takahashi et al., 2026). After surgery, pembrolizumab or placebo continued for up to 9 cycles (Takahashi et al., 2026). Check the package insert in your country before prescribing.

The chosen outcomes

The study had two primary outcomes. The first was pathological complete response, defined as ypT0 or Tis with ypN0 at surgery. The second was event-free survival, which counts progression that prevents surgery, recurrence, a second primary cancer and death (Takahashi et al., 2026). Overall survival was an important secondary outcome.

This choice deserves comment. Complete response is a fast marker, but it is not enough to approve a treatment with permanent toxicity. It was event-free survival, and then overall survival, that confirmed the gain in the specimen translated into more patients alive and without disease.

Time point	Outcome	Pembrolizumab	Placebo	Source
First analysis	Pathological complete response	64.8 in every 100	51.2 in every 100	Takahashi et al., 2026
Median follow-up of 39.1 months	Event-free survival at 36 months	84.3 in every 100	76.2 in every 100	Huang et al., 2023
Event analysis	Hazard ratio for events	0.63 (95% CI 0.48 to 0.82)		Takahashi et al., 2026
Overall survival analysis	Hazard ratio for death	0.66 (95% CI 0.50 to 0.87)		Takahashi et al., 2026

The results in three stages

KEYNOTE-522 was published in stages, and each one answered a different question. Reading the study in that order helps explain to the patient why the regimen became standard.

First stage: the response in the specimen

Pathologic complete response was 64.8 per 100 patients with pembrolizumab and 51.2 per 100 with placebo, a difference of 13.6 points (Takahashi and colleagues, 2026). The confidence interval for this difference was 5.4 to 21.8 points, with statistical significance (Takahashi and colleagues, 2026). The study included patients with any PD-L1 value, and in the Japanese subgroup the response findings were similar regardless of this marker (Takahashi and colleagues, 2026). Therefore, the test is not used to select who receives the regimen.

Second act: fewer recurrences

With a median follow-up of 39.1 months, event-free survival at 36 months was 84.3 per 100 patients in the pembrolizumab group and 76.2 per 100 in the placebo group, with a hazard ratio of 0.63 (Huang and colleagues, 2023). This means that, for every 100 patients treated, about 8 more were alive and event-free at three years.

Third act: more patients alive

The overall survival analysis showed a hazard ratio of 0.66, with an interval of 0.50 to 0.87 (Takahashi and colleagues, 2026). Based on these results, the regimen was approved in several countries for high-risk early triple-negative disease (Takahashi and colleagues, 2026). In the third interim analysis, overall survival at 36 months was numerically higher, 89.7 versus 86.9 per 100, still without significance at that point (Huang and colleagues, 2023). The difference became clear with longer follow-up, which reinforces the importance of waiting for mature data.

Quality of life

The analysis of patient-reported outcomes, using the EORTC QLQ-C30 and QLQ-BR23 questionnaires, showed no substantial difference between the groups (Dent and colleagues, 2024). In the neoadjuvant phase, quality of life fell in both groups, as expected with intensive chemotherapy, and pembrolizumab did not worsen this picture in an important way (Dent and colleagues, 2024). This finding helps in the conversation with the patient, who often fears that "one more drug" means much more suffering.

Cost and access

The price of pembrolizumab weighs on public and private systems. A cost-effectiveness analysis using data from the United States estimated a gain of 2.90 quality-adjusted life-years, with an incremental cost of about 27 thousand dollars per adjusted year (Huang and colleagues, 2023). A Swiss analysis estimated 14,114 Swiss francs per adjusted year, below the threshold of 100 thousand francs used in the country (Favre-Bulle and colleagues, 2024). Each system reaches its own conclusion, according to the negotiated price and the threshold adopted. Clinical reasoning needs to know these barriers so as not to promise a regimen that the service cannot offer.

Immune-mediated toxicity in practice

Pembrolizumab adds a set of immune-mediated effects to the toxicity of chemotherapy. Some are mild and pass. Others are permanent, such as hypothyroidism and adrenal insufficiency, and stay with the patient for the rest of her life.

What the studies show

In the Japanese subgroup of KEYNOTE-522, immune-mediated events and infusion reactions of any grade occurred in 24 of 45 patients with pembrolizumab and in 10 of 30 with placebo (Takahashi and colleagues, 2026). Grade 3 or 4 events of this type appeared in 9 patients in the pembrolizumab group and in 1 in the placebo group (Takahashi and colleagues, 2026). In a Brazilian cohort of 368 patients from ten centers, 31 per 100 had some immune-mediated event, and most of them arose in the neoadjuvant phase (Andrade and colleagues, 2025).

In the same Brazilian cohort, thyroiditis was the most frequent event, in 9.2 per 100 patients, followed by skin events, hepatitis and pneumonitis (Andrade and colleagues, 2025). Primary adrenal insufficiency occurred in 1.9 per 100 and hypophysitis in 1.4 per 100 (Andrade and colleagues, 2025). In a Korean cohort of 174 patients, hypothyroidism and adrenal insufficiency were the most common immune-mediated events (Lee and colleagues, 2026).

Signs that call for immediate attention

• Disproportionate fatigue, nausea, dizziness on standing and a drop in blood pressure, which may indicate adrenal insufficiency.
• Weight gain, cold intolerance and slowness, suggestive of hypothyroidism.
• Persistent diarrhea or blood in the stool, which may indicate colitis.
• New cough and shortness of breath, which may indicate pneumonitis.
• Intense thirst and a lot of urine, signs of sudden-onset diabetes.
• Skin with blisters or extensive peeling, which requires assessment the same day.

A practical lesson from this setting is that adrenal insufficiency can be mistaken for chemotherapy fatigue. That is why many services order routine cortisol and TSH during treatment. After surgery, the patient remains exposed to pembrolizumab, and events can appear in the adjuvant phase, when visits tend to become more spaced out (Andrade and colleagues, 2025).

Immune-mediated event and response

Some small studies suggested that those who have an immune-mediated event respond better. In an Austrian series of 35 patients, complete response was 72.2 in 100 with an immune-mediated event and 30.8 in 100 without an event (Marhold and colleagues, 2023). In the larger Brazilian cohort, the complete response rate was 66.1 in 100 with an event and 60.8 in 100 without an event, with no significant difference (Andrade and colleagues, 2025). The adverse event should not be read as a sign of success nor as a reason to insist on a dangerous treatment.

What clinical practice showed after the study

Studies outside the controlled setting help show whether the KEYNOTE-522 numbers repeat. Results vary, and that variation is part of the reasoning.

Complete response outside the study

In the Korean cohort of 174 patients, pathological complete response was 62.1 in 100, and fell to 47.4 in 100 in the subgroup with N3 disease (Lee and colleagues, 2026). High Ki-67 on biopsy was independently associated with complete response, with an odds ratio of 2.84 (Lee and colleagues, 2026). In a United States cohort of 118 patients, complete response was 63.6 in 100, and completing at least 8 cycles of pembrolizumab was associated with more response (LeVee and colleagues, 2024).

In a Belgian cohort, complete response was 58 in 100, but 35 in 100 patients stopped treatment because of adverse events in the neoadjuvant phase (Rached and colleagues, 2024). In that same group, 61 in 100 had at least one immune-mediated event, and there was one death from myocarditis (Rached and colleagues, 2024). These numbers are a reminder that toxicity can be greater outside a study with strict selection criteria.

Who responds less

Lymphocytes infiltrating the tumor on biopsy were associated with more response. In a diverse United States cohort, complete response was 70 in 100 when these lymphocytes were present and 48 in 100 when they were not (Albert and colleagues, 2026). Another study identified a high-risk group for non-response, with more metaplastic carcinoma, involved axilla and necrosis, in which complete response was 2.9 in 100 with chemotherapy alone and 20 in 100 with immunotherapy (Khoury and colleagues, 2025).

• N3 disease at presentation.
• Metaplastic carcinoma on biopsy.
• Few lymphocytes infiltrating the tumor.
• Lower Ki-67.
• Early discontinuation of pembrolizumab due to toxicity.

These factors do not exclude anyone from treatment. They serve to prepare the conversation about the chance of residual disease and to plan in advance what to do if it appears.

When disease remains in the specimen

Some patients finish the neoadjuvant phase with residual tumor. This group has a worse prognosis and is the focus of the hardest decisions. In KEYNOTE-522, all patients continued with adjuvant pembrolizumab, with or without complete response (Takahashi and colleagues, 2026).

Capecitabine and olaparib

Before immunotherapy, adjuvant capecitabine was the option for triple-negative disease with residual disease, based on the CREATE-X study, which used 1,250 mg/m² twice a day, on days 1 to 14, every 3 weeks, for 8 cycles (Di Lisa et al., 2023). For carriers of a germline mutation in BRCA1 or BRCA2, adjuvant olaparib, based on the OlympiA study, also showed a survival gain (Trédan et al., 2026). For this reason, germline BRCA testing should be requested early, ideally before surgery.

An Italian cohort of 270 patients showed good tolerance to capecitabine, with discontinuation due to toxicity in 10.4 out of 100, but lower efficacy than CREATE-X (Di Lisa et al., 2023). Completing six cycles or more was associated with better outcomes in this cohort (Di Lisa et al., 2023).

What is still unknown

There is no phase III study that defines how to combine pembrolizumab with capecitabine or olaparib after the KEYNOTE-522 regimen. A phase II study tested nivolumab plus ipilimumab against capecitabine in residual disease and showed no gain in disease-free survival, with a hazard ratio of 0.84 (Trédan et al., 2026). In the dual immunotherapy group, 38 out of 100 patients stopped treatment due to adverse events, versus 14 out of 100 with capecitabine (Trédan et al., 2026). The decision in these cases depends on a multidisciplinary meeting and an open discussion with the patient.

How NEXO helps train this reasoning

In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues, and you propose the diagnosis by looking for the ICD-10 code. Early triple-negative is a good exercise because the clues arrive in layers: first the size and the axilla, then the receptors, then the response in the specimen and, finally, the germline test result. The post on differential diagnosis with clues shows how to train with progressive clues.

Three exercises with cases

Change the size. Say what changes if the tumor is 1 centimeter and the axilla is negative.

Change the course. Say what to do if the patient arrives at the fourth cycle with low blood pressure, nausea and low sodium.

Change the pathology report. Say what to discuss if the specimen shows residual tumor of 1.5 centimeters and the germline test shows a BRCA1 mutation.

The post on residual disease in HER2-positive breast cancer shows the same logic in a different subtype. The limits of use of the game are on the NEXO Economy page. If you research teaching of clinical reasoning, the page for Researcher explains how to get in touch.

Frequently asked questions

What is the pembrolizumab dose in KEYNOTE-522?

Pembrolizumab 200 mg every 3 weeks, for 8 cycles before surgery and up to 9 after it (Takahashi et al., 2026). Check the package insert in your country before prescribing.

Do I need to measure PD-L1 before indicating the regimen?

Not to select the patient. The study included patients with any PD-L1 value, and in the Japanese subgroup the findings were similar regardless of the marker (Takahashi et al., 2026).

Does the patient with a complete response need to continue pembrolizumab?

In the study, all of them continued for up to 9 cycles after surgery (Takahashi et al., 2026). Whether it is possible to stop earlier in those who responded well is a question still under study.

Is operating right away always wrong?

No. In small tumors without involved lymph nodes, surgery first remains an option, because this group was not included in the study (Takahashi et al., 2026).

Does an immune-mediated event mean the treatment is working?

Not reliably. In the largest available cohort, there was no significant association between an immune-mediated event and complete response (Andrade et al., 2025).

Final checklist

• I confirmed hormone receptors and HER2 on the biopsy.
• I documented tumor size and axillary status, with marking of the involved lymph node.
• I took the case to the meeting before scheduling surgery.
• I requested germline testing of BRCA1 and BRCA2.
• I reviewed autoimmune disease, cardiac function and corticosteroid use.
• I requested baseline TSH, cortisol and glucose and planned the repeat testing.
• I explained to the patient the signs of an immune-mediated event and whom to call.
• I defined the plan for residual disease before surgery.

Want to practice reasoning with progressive clues? Play the case of the day.

References

1. Albert R, Thomas J, Sadeghi N, et al. Tumor infiltrating lymphocytes (TILs) as a predictive marker of pathological complete response (pCR) in a diverse patient population with early triple negative breast cancer (TNBC) treated with neoadjuvant real-world KEYNOTE-522 regimen. Breast Cancer Res Treat. 2026. PMID 42360368.
2. Andrade MO, Gutierres IG, Tavares MC, et al. Immune-related adverse events among patients with early-stage triple-negative breast cancer treated with pembrolizumab plus chemotherapy: real-world data from the Neo-Real/GBECAM 0123 study. Breast. 2025. PMID 40240201.
3. Dent R, Cortés J, Pusztai L, et al. Neoadjuvant pembrolizumab plus chemotherapy/adjuvant pembrolizumab for early-stage triple-negative breast cancer: quality-of-life results from the randomized KEYNOTE-522 study. J Natl Cancer Inst. 2024. PMID 38913881.
4. Di Lisa FS, Krasniqi E, Pizzuti L, et al. Adjuvant capecitabine in triple negative breast cancer patients with residual disease after neoadjuvant treatment: real-world evidence from CaRe, a multicentric, observational study. Front Oncol. 2023. PMID 37260975.
5. Favre-Bulle A, Huang M, Haiderali A, et al. Cost-Effectiveness of Neoadjuvant Pembrolizumab plus Chemotherapy Followed by Adjuvant Pembrolizumab in Patients with High-Risk, Early-Stage, Triple-Negative Breast Cancer in Switzerland. Pharmacoecon Open. 2024. PMID 37999854.
6. Huang M, A Fasching P, Haiderali A, et al. Cost-Effectiveness of Neoadjuvant Pembrolizumab Plus Chemotherapy Followed by Adjuvant Single-Agent Pembrolizumab for High-Risk Early-Stage Triple-Negative Breast Cancer in the United States. Adv Ther. 2023. PMID 36648737.
7. Khoury T, Gandhi S, Tozbikian G, et al. Clinicopathological evaluation of triple-negative breast cancer treated with keynote-522 regimen. Oncologist. 2025. PMID 40704692.
8. Lee S, Jeong H, Shin Y, et al. A Real-World Efficacy and Safety of KEYNOTE-522 Regimen in Patients With Early Triple-Negative Breast Cancer. J Breast Cancer. 2026. PMID 41918183.
9. LeVee A, Wong M, Flores S, et al. Impact of neoadjuvant pembrolizumab adherence on pathologic complete response in triple-negative breast cancer: a real-world analysis. Oncologist. 2024. PMID 38656345.
10. Marhold M, Udovica S, Halstead A, et al. Emergence of immune-related adverse events correlates with pathological complete response in patients receiving pembrolizumab for early triple-negative breast cancer. Oncoimmunology. 2023. PMID 38025838.
11. Rached L, Peyre-Pradat F, Spotti M, et al. Real-world safety and effectiveness of neoadjuvant chemotherapy combination with pembrolizumab in triple-negative breast cancer. ESMO Real World Data Digit Oncol. 2024. PMID 41648651.
12. Takahashi M, Mukai H, Takano T, et al. Pembrolizumab for Early-Stage Triple-Negative Breast Cancer: KEYNOTE-522 Japan Subgroup Analysis. Cancer Sci. 2026. PMID 41536071.
13. Trédan O, Loirat D, Chabaud S, et al. Nivolumab in combination with ipilimumab versus capecitabine as post-operative treatment for triple negative breast cancer patients with residual disease after neoadjuvant chemotherapy: a multicentre, randomized, open-label phase II trial - BREASTIMMUNE-03. Breast. 2026. PMID 41314029.
14. Wu K, Yang Q, Liu Y, et al. Meta-analysis on the association between pathologic complete response and triple-negative breast cancer after neoadjuvant chemotherapy. World J Surg Oncol. 2014. PMID 24731479.