Perioperative pembrolizumab in lung cancer: KEYNOTE-671

A 62-year-old bus driver, a former smoker, undergoes a chest CT for mild hemoptysis. The exam shows a 5 cm mass in the right upper lobe with enlarged lymph nodes in the mediastinum on the same side. Bronchoscopic biopsy confirms squamous cell carcinoma, and puncture of a mediastinal lymph node is positive. Staging shows no distant metastases. In the multidisciplinary meeting, the surgeon thinks the tumor can be removed with lobectomy. The oncologist recalls that the patient takes methotrexate for rheumatoid arthritis and that the molecular profile has not yet returned. The scene is fictional and was written for teaching.

This text uses the case as a clinical reasoning scenario for students and health professionals. The topic is perioperative pembrolizumab in resectable non-small cell lung cancer, tested in the KEYNOTE-671 trial. The focus is integrating stage, resectability, molecular profile and contraindications before choosing immunotherapy. It does not replace your institution's protocol or the multidisciplinary team's decision. If you play the case of the day in NEXO, you will recognize the challenge: deciding with clues that arrive in sequence.

Key points

• The KEYNOTE-671 regimen combines 4 cycles of pembrolizumab with chemotherapy before surgery and up to 13 cycles of pembrolizumab after.
• Event-free survival improved, with a hazard ratio of 0.58 in the first interim analysis.
• In the second interim analysis, overall survival also improved, with a hazard ratio of 0.72 and 71 percent versus 64 percent alive at 36 months.
• The trial excluded people with active autoimmune disease that required systemic treatment in the previous 2 years and those who needed immunosuppression.
• Molecular testing was not mandatory in the trial, and few patients with altered EGFR or ALK were enrolled.
• The FDA approved the indication on 10/16/2023, and the EMA committee gave a favorable opinion on 02/22/2024.

Why this is a reasoning scenario

Until recently, patients with operable stage II or III lung cancer received surgery and chemotherapy, before or after the operation. Immunotherapy entered this scenario in two ways. One was immunotherapy only after surgery, as in the adjuvant atezolizumab and pembrolizumab trials (Wakelee et al., 2023). The other was immunotherapy with chemotherapy only before surgery, as in the CheckMate 816 trial with nivolumab (Clementi et al., 2025). KEYNOTE-671 tested the combination of both phases, before and after the operation (Wakelee et al., 2023) (Deboever and Zhang, 2023).

The choice among these strategies depends on many clues at the same time: stage, real chance of complete resection, molecular profile, histology, the patient's diseases and their preference. The post on signs, symptoms and history helps not to forget the driver's rheumatoid arthritis, and the post on clinical decision-making reminds us that the molecular profile must be ready before the decision.

What the scene teaches

The scene trains six habits. The first is confirming the stage, including the mediastinal lymph nodes. The second is deciding resectability as a team. The third is waiting for the molecular profile before starting immunotherapy. The fourth is looking for contraindications, such as autoimmune disease and immunosuppression. The fifth is planning the entire sequence, with dates for surgery and the adjuvant phase. The sixth is monitoring immune-mediated adverse effects in all phases.

What is not at stake

The text does not cover in detail adjuvant alectinib in ALK-positive cancer, which was the topic of the first text in this series. It also does not compare in detail all available perioperative regimens. The post on diagnostic uncertainty helps when the molecular profile is delayed and the team must decide whether to wait.

Who is a candidate: stage, resectability and molecular profile

Stage

KEYNOTE-671 included adults with previously untreated non-small cell lung cancer, stage II, IIIA or IIIB with N2, by the eighth edition of the staging system, considered resectable after surgical evaluation (Wakelee et al., 2023). The FDA label indicates the regimen for resectable tumors of 4 cm or more or with positive lymph node (Keytruda label of 04/2026). The European indication speaks of stages II, IIIA or IIIB (T3-4N2) with high risk of recurrence (EMA assessment report, 2024). The driver in the scene, with a 5 cm tumor and a positive mediastinal lymph node on the same side, fits both texts.

Resectability

The decision about what is resectable belongs to the team, not to one person alone. A 2026 review recalls that a phase 2 study provides initial evidence for the multidisciplinary team to guide neoadjuvant chemoimmunotherapy in disease at the limit of resectability (Patel et al., 2026). The same review recalls that the ninth edition of staging changes the classification of lymph nodes and should influence this decision (Patel et al., 2026). Another review highlights multidisciplinary collaboration and the individual plan to balance efficacy, feasibility of surgery, toxicity and cost (Kim and Lee, 2026). The post about interprofessional clinical reasoning shows how this conversation works in practice.

Molecular profile

In KEYNOTE-671, EGFR and ALK were tested locally at the investigator's discretion (Wakelee et al., 2023). Because testing was not mandatory, few patients with these alterations were enrolled, and the study does not allow conclusions in this group (Wakelee et al., 2023). Induction immunotherapy studies excluded tumors with altered EGFR or ALK (Lavaud et al., 2024). Therefore, the molecular profile should be ready before the decision. The Italian panel of experts speaks of perioperative chemoimmunotherapy in tumors without oncogene dependency (de Marinis et al., 2025). The post about differential diagnosis helps keep the options open until the result arrives.

Contraindications

The study excluded people with active autoimmune disease that required systemic treatment in the previous 2 years and people with a condition that required immunosuppression (Keytruda label of 04/2026). The label does not list absolute contraindications, but warns of severe and fatal immune-mediated reactions (Keytruda label of 04/2026). The driver uses methotrexate for rheumatoid arthritis, which would leave him out of the study. This does not forbid treatment, but it requires a conversation with the rheumatologist and with the patient about the risk of worsening of the disease and about the fact that there are no study data for someone like him. The post about diagnostic pause helps make this stop before the first infusion.

What KEYNOTE-671 tested

Design

KEYNOTE-671 was a phase 3, randomized, double-blind, placebo-controlled study (Spicer et al., 2024). Between April 2018 and December 2021, 1364 people were assessed and 797 were randomized, 397 to pembrolizumab and 400 to placebo (Wakelee et al., 2023). Randomization was stratified, among other factors, by region, East Asia or not (Wakelee et al., 2023).

Neoadjuvant phase

The phase before surgery had 4 cycles of intravenous pembrolizumab 200 mg or placebo every 3 weeks (Spicer et al., 2024). Together, everyone received cisplatin 75 mg/m² every 3 weeks (Spicer et al., 2024). Those with squamous histology received gemcitabine 1000 mg/m² on days 1 and 8 of 3-week cycles, and those with nonsquamous histology received pemetrexed 500 mg/m² every 3 weeks (Spicer et al., 2024).

Surgery

The operation, lobectomy, bilobectomy, pneumonectomy or sleeve resection, had to be done up to 20 weeks after the first neoadjuvant dose (Spicer et al., 2024). Imaging was done after the neoadjuvant phase, but the tumors were not formally restaged before surgery (Wakelee et al., 2023). Radiotherapy was used in selected situations (Wakelee et al., 2023).

Adjuvant phase

The phase after surgery began between 4 and 12 weeks after the operation, with pembrolizumab 200 mg or placebo every 3 weeks for up to 13 cycles (Wakelee et al., 2023). The primary endpoints were event-free survival and overall survival (Wakelee et al., 2023). Major pathological response, with up to 10 percent viable tumor cells, and complete pathological response were key secondary endpoints (Wakelee et al., 2023).

What KEYNOTE-671 showed

Surgery and resection

In the pembrolizumab group, 79.4 percent underwent the study surgery and 66.9 percent received at least one adjuvant dose (Wakelee et al., 2023). Among those operated on, complete resection was 92.0 percent with pembrolizumab and 84.2 percent with placebo (Wakelee et al., 2023). The label states that 6 percent of those treated with pembrolizumab and 4.3 percent of those treated with placebo did not have surgery due to adverse reactions (Keytruda label of 04/2026).

First interim analysis

At 24 months, 62.4 percent of the pembrolizumab group were alive and event-free, versus 40.6 percent of the placebo group (Wakelee et al., 2023). The hazard ratio for event-free survival was 0.58, with a 95 percent confidence interval of 0.46 to 0.72 (Wakelee et al., 2023). Major pathological response was 30.2 percent versus 11.0 percent, and complete pathological response was 18.1 percent versus 4.0 percent (Wakelee et al., 2023). In this analysis, overall survival had not yet crossed the significance boundary (Wakelee et al., 2023).

Second interim analysis

With a median follow-up of about 3 years, pembrolizumab improved overall survival, with a hazard ratio of 0.72 and an interval of 0.56 to 0.93 (Spicer et al., 2024). At 36 months, 71 percent were alive in the pembrolizumab group and 64 percent in the placebo group (Spicer et al., 2024). Median event-free survival was 47.2 months versus 18.3 months, with a hazard ratio of 0.59 (Spicer et al., 2024). The survival curves began to separate near month 16 (Spicer et al., 2024).

The table summarizes the main numbers.

Measure	Pembrolizumab	Placebo	Source
Event-free survival at 24 months	62.4 percent	40.6 percent	(Wakelee et al., 2023)
Complete pathological response	18.1 percent	4.0 percent	(Wakelee et al., 2023)
Complete resection among those operated on	92.0 percent	84.2 percent	(Wakelee et al., 2023)
Alive at 36 months	71 percent	64 percent	(Spicer et al., 2024)
Hazard ratio for overall survival	0.72 (0.56 to 0.93)	Reference	(Spicer et al., 2024)

Subgroups

The relative benefit grew with PD-L1 expression, with a hazard ratio for event-free survival of 0.42 with a score of 50 percent or more, 0.51 from 1 to 49 percent and 0.77 below 1 percent (Wakelee et al., 2023). In all cases the ratio favored pembrolizumab, but the subgroups were small (Wakelee et al., 2023). In an exploratory analysis of the 239 patients with stage II disease, 29 of 118 died with pembrolizumab and 42 of 121 with chemotherapy alone, a hazard ratio of 0.69 with an interval of 0.43 to 1.11 (Tsuboi et al., 2026). That interval includes the value 1, so the stage II result is uncertain. There is also an analysis of the Japanese participants (Tsuboi et al., 2026, Japanese participants).

The question the study does not answer

Because the two arms differed before and after surgery, the study does not say how much of the benefit comes from the adjuvant phase (Deboever and Zhang, 2023). There is still doubt about whether combining the neoadjuvant and adjuvant phases brings benefit beyond each one alone (Deboever and Zhang, 2023). A review recalls that CheckMate 816, neoadjuvant only, also showed an overall survival gain at 5 years (Clementi and colleagues, 2025).

Pathologic response and biomarkers

An exploratory analysis showed an event-free survival benefit with pembrolizumab both in those who had a major pathologic response and in those who did not, and the same for complete pathologic response (Wakelee and colleagues, 2023). This means that the absence of complete response in the specimen is not enough, on its own, to conclude that the treatment failed. Even so, the response in the specimen is a useful clue for talking with the patient about the risk of recurrence. Circulating tumor DNA is another candidate for guiding the decision after surgery (Deboever and Zhang, 2023). In CheckMate 816, no patient with circulating tumor DNA still positive after the neoadjuvant phase had a complete pathologic response (Deboever and Zhang, 2023). For the authors of another review, stage, tumor biology, pathologic response and biomarkers such as circulating tumor DNA should increasingly guide the choice among the strategies, which form a continuum and do not compete with each other (Patel and colleagues, 2026). These biomarkers are not yet part of the label indication.

How to use perioperative pembrolizumab

Dose and duration

The label indicates, for lung cancer, 200 mg every 3 weeks or 400 mg every 6 weeks (Keytruda label of 04/2026). There is no dose reduction for pembrolizumab (Keytruda label of 04/2026). Patients with disease progression or unacceptable toxicity linked to pembrolizumab in the neoadjuvant phase should not receive the adjuvant phase (Keytruda label of 04/2026).

Toxicity

In the first analysis, grade 3 or higher adverse events linked to treatment were seen in 44.9 percent of the pembrolizumab group and in 37.3 percent of the placebo group (Wakelee and colleagues, 2023). Events possibly immune-mediated and infusion reactions were seen in 25.3 percent and 10.5 percent (Wakelee and colleagues, 2023). By any cause, up to 30 days after surgery, 6 patients died in the pembrolizumab group, 1.8 percent, and 2 in the placebo group, 0.6 percent (Wakelee and colleagues, 2023). In the neoadjuvant phase, fatal adverse reactions occurred in 1.3 percent, and permanent discontinuation of any medication due to an adverse reaction occurred in 18 percent (Keytruda label of 04/2026). In the adjuvant phase, 14 percent had serious adverse reactions and 12 percent discontinued pembrolizumab due to an adverse reaction (Keytruda label of 04/2026).

What to watch for

Monitoring follows the label warnings. The points below summarize the main ones.

• Immune-mediated pneumonitis, more frequent in those who received chest radiotherapy.
• Hypothyroidism and other thyroid changes, with hormone replacement when indicated.
• Diarrhea and colitis, hepatitis and changes in kidney function.
• Interruption for serious grade 3 reactions and permanent stop for grade 4 or recurrent ones.
• Permanent stop if prednisone cannot be tapered to 10 mg per day or less within 12 weeks.
• Contraception, because the medication can cause fetal harm.

The items come from the label (Keytruda label of 04/2026). Immune-mediated pneumonitis appeared in 3.4 percent of patients in the safety base, and hypothyroidism in 8 percent (Keytruda label of 04/2026). The post about clinical deterioration in the ward helps to recognize the worsening of a hospitalized patient during treatment.

In real life

A national cohort study in the United States showed a benefit of chemoimmunotherapy before or after surgery, but fewer than 30 percent of potentially eligible patients received it in 2023 (Desai et al., 2025). Reviews on the role of immunotherapy in resectable disease also discuss the safety of surgery after immunotherapy (Dunne et al., 2024). The discussion about chemotherapy before or after surgery is old and still has open points (Murakami, 2024).

Regulatory status: FDA and EMA

FDA

Keytruda has BLA 125514 at the FDA, with original approval on 09/04/2014 (Drugs@FDA, BLA 125514). Efficacy supplement S-139, which added the perioperative use in resectable lung cancer, was approved on 10/16/2023 (Drugs@FDA, BLA 125514) (FDA approval letter of 10/16/2023). The current label was revised in 04/2026 (Keytruda label of 04/2026).

EMA

At the EMA, Keytruda received a positive opinion on 05/20/2015 and marketing authorization on 07/17/2015 (Keytruda page at the EMA). The indication extension for the KEYNOTE-671 regimen was assessed in procedure II/0134, with an opinion from the committee for medicinal products for human use on 02/22/2024 (EMA assessment report, 2024). The European indication restricts use to resectable cancer with high risk of recurrence, with the selection criteria in section 5.1 of the summary of product characteristics (EMA assessment report, 2024).

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 searching for the ICD-10 code. Perioperative immunotherapy is a good exercise because the right decision depends on putting together clues from several people on the team. The post about case-based learning explains why this format fixes knowledge.

Three exercises with cases

Change the target. Say what changes if the molecular profile showed an ALK rearrangement.

Change the autoimmune disease. Say what you would ask if the patient had active Crohn's disease on a biologic.

Change the surgery. Say what you would do if the resection were incomplete, with a positive margin.

The post about simulating a consultation helps train the conversation with the patient about benefit, toxicity and uncertainty.

The limits on use of the game are described on the NEXO economy page. If you research teaching of clinical reasoning, the page for researchers explains how to get in touch. Professionals who write or review cases find information under curation and under creators, and institutions can learn about the proposal on the partnerships page.

Frequently asked questions

Does perioperative pembrolizumab improve survival?

Yes. In the second interim analysis, the hazard ratio for overall survival was 0.72 (Spicer et al., 2024).

What is the dose?

200 mg intravenously every 3 weeks, 4 cycles with chemotherapy before surgery and up to 13 cycles after (Wakelee et al., 2023). The label also allows 400 mg every 6 weeks (Keytruda label of 04/2026).

Do I need the EGFR and ALK result before starting?

Yes. The study barely included tumors with these alterations, and induction immunotherapy studies excluded them (Wakelee et al., 2023) (Lavaud et al., 2024).

Can people with autoimmune disease receive it?

These people were excluded from the study when the disease was active with systemic treatment in the previous 2 years (Keytruda label of 04/2026). The decision must be individual.

Does negative PD-L1 prevent treatment?

Not according to the label. The benefit was smaller below 1 percent, but the hazard ratio still favored pembrolizumab (Wakelee et al., 2023).

Final checklist

• I confirmed stage and mediastinal lymph nodes.
• I discussed resectability at the multidisciplinary meeting.
• I waited for EGFR, ALK and PD-L1 before starting.
• I looked for autoimmune disease and immunosuppression.
• I prescribed 4 cycles with cisplatin and the partner for the histology.
• I scheduled surgery within the regimen window.
• I planned the adjuvant phase of up to 13 cycles.
• I gave guidance on immune-mediated symptoms and contraception.

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

References

1. Clementi TD, Colonese F, Canova S, et al. Perioperative Chemo/Immunotherapies in Lung Cancer: A Critical Review on the Value of Perioperative Sequences. Curr Oncol. 2025. PMID 40710207.
2. Deboever N, Zhang J. The opportunities and challenges of perioperative therapy of localized non-small cell lung cancer-thoughts from the KEYNOTE-671 trial. Transl Lung Cancer Res. 2023. PMID 38090518.
3. Desai A, Schwed K, Kalesinskas L, et al. Clinical Outcomes of Perioperative Immunotherapy in Resectable Non-Small Cell Lung Cancer. JAMA Netw Open. 2025. PMID 40587130.
4. Dunne EG, Fick CN, Isbell JM, et al. The Emerging Role of Immunotherapy in Resectable Non-Small Cell Lung Cancer. Ann Thorac Surg. 2024. PMID 38316378.
5. European Medicines Agency. Assessment report. Keytruda. Procedure No. EMEA/H/C/003820/II/0134. 22/02/2024. Official source.
6. European Medicines Agency. Keytruda | European Medicines Agency (EMA). Accessed on 24/10/2026. Official source.
7. Food and Drug Administration. BLA 125514/S-139. SUPPLEMENT APPROVAL. Approval letter dated 16/10/2023. Official source.
8. Food and Drug Administration. Drugs@FDA: FDA-Approved Drugs. BLA 125514, Keytruda. Accessed on 24/10/2026. Official source.
9. Kim DK, Lee JE. Challenges and Controversies in Perioperative Immunotherapy in Early-Stage Non-small Cell Lung Cancer: An Optimistic Perspective. Tuberc Respir Dis (Seoul). 2026. PMID 41713848.
10. Lavaud P, Bortolot M, Zullo L, et al. Early-Stage Non-Small Cell Lung Cancer: New Challenges with Immune Checkpoint Blockers and Targeted Therapies. Cancers (Basel). 2024. PMID 39199552.
11. Merck Sharp & Dohme. KEYTRUDA® (pembrolizumab) injection, for intravenous use. Prescribing information revised in 04/2026. Official source.
12. Murakami S. Will perioperative pembrolizumab treatment change perioperative treatment strategies for resectable non-small cell lung cancer? Transl Lung Cancer Res. 2024. PMID 38404997.
13. Patel A, Wakeam E, Josephides E, et al. Neoadjuvant chemoimmunotherapy in non-small cell lung cancer: evolving resectability criteria, biomarker-driven postoperative management, and emerging therapeutic strategies. J Immunother Cancer. 2026. PMID 41638870.
14. Spicer JD, Garassino MC, Wakelee H, et al. Neoadjuvant pembrolizumab plus chemotherapy followed by adjuvant pembrolizumab compared with neoadjuvant chemotherapy alone in patients with early-stage non-small-cell lung cancer (KEYNOTE-671): a randomised, double-blind, placebo-controlled, phase 3 trial. Lancet. 2024. PMID 39288781.
15. Tsuboi M, Saji H, Okada M, et al. Perioperative Pembrolizumab in Japanese Participants With Early-Stage NSCLC: Subgroup Analysis From KEYNOTE-671. JTO Clin Res Rep. 2026. PMID 42774586.
16. Tsuboi M, Wakelee H, Garassino MC, et al. A Subgroup Analysis of Perioperative Pembrolizumab in Clinical Stage II Non-Small-Cell Lung Cancer from the Randomized KEYNOTE-671 Study. Eur J Cardiothorac Surg. 2026. PMID 41875364.
17. Wakelee H, Liberman M, Kato T, et al. Perioperative Pembrolizumab for Early-Stage Non-Small-Cell Lung Cancer. N Engl J Med. 2023. PMID 37272513.
18. de Marinis F, Ardizzoni A, Attili I, et al. Perioperative Chemo-Immunotherapy in Non-Oncogene-Addicted Resectable Non-Small Cell Lung Cancer (NSCLC): Italian Expert Panel Meeting. Curr Oncol. 2025. PMID 39996910.