Perioperative Pembrolizumab in Japanese Participants With Early-Stage NSCLC: Subgroup Analysis From KEYNOTE-671 Authors: Masahiro Tsuboi, Hisashi Saji, Morihito Okada, Norihiko Ikeda, Kazuya Takamochi, Kenji Sugio, Yoshitsugu Horio, Fumihiro Tanaka, Wataru Nishio, Satoshi Muto, Hidehito Horinouchi, Heather Wakelee, Shirong Han, Tomoko Mori, Steven M Keller, Terufumi Kato Published: 2026 ABSTRACT AbstractIntroductionThe global KEYNOTE-671 (clinicaltrials.gov: NCT03425643) study demonstrated significantly improved survival with perioperative pembrolizumab plus neoadjuvant chemotherapy versus neoadjuvant chemotherapy alone in early-stage non–small-cell lung cancer (NSCLC). We present findings from Japanese participants from KEYNOTE-671.MethodsEligible participants (≥18 y) with previously untreated stage II, IIIA, or IIIB (N2) NSCLC were randomized 1:1 to neoadjuvant pembrolizumab 200 mg or placebo plus cisplatin-based chemotherapy every 3 weeks for up to 4 cycles followed by surgery and adjuvant pembrolizumab or placebo for up to 13 cycles (approximately 9 mo). Primary endpoints were event-free survival (EFS) and overall survival (OS).ResultsAmong 82 participants enrolled in Japan, 39 were randomized to pembrolizumab and 43 to placebo. Median time from randomization to database cutoff (August 19, 2024) was 52.7 (range, 32.7–74.9) months. The 48-month EFS rates (95% confidence intervals [CIs]) were 60.3% (42.8%–74.0%) and 39.2% (24.8%–53.4%), respectively, and the EFS hazard ratio was 0.54 (95% CI: 0.29–1.02). For OS, 48-month rates (95% CIs) were 79.5% (63.1%–89.2%) and 68.9% (52.4%–80.7%), and the OS hazard ratio was 0.76 (95% CI: 0.33–1.78). Grade 3 or higher treatment-related adverse events occurred in 23 participants (59%) treated with pembrolizumab and 20 participants (47%) with placebo; there were no deaths due to treatment-related adverse events.ConclusionsPerioperative pembrolizumab plus neoadjuvant chemotherapy improved efficacy versus neoadjuvant chemotherapy alone and had manageable safety in Japanese participants enrolled in KEYNOTE-671, supporting the use of perioperative pembrolizumab in Japanese patients with early-stage NSCLC.Keywords: Immunotherapy, Japan, Pembrolizumab, Perioperative medicine, Non–small-cell lung cancer Introduction Several global phase 3 studies have demonstrated the clinical benefits of neoadjuvant, adjuvant, and perioperative immune checkpoint inhibition with anti–programmed cell death protein 1 (anti–PD-1) and anti–programmed cell death ligand 1 (anti–PD-L1) inhibitors for the treatment of resectable early-stage non–small-cell lung cancer (NSCLC).1, 2, 3, 4, 5, 6 The positive findings from these studies led to the approval of several (neo)adjuvant therapy regimens as treatment for resectable NSCLC.7, 8, 9, 10 One of these is perioperative pembrolizumab, an anti–PD-1 monoclonal antibody, which is approved in combination with neoadjuvant chemotherapy in several countries worldwide for the treatment of patients with resectable stage II, IIIA, or IIIB (N2) NSCLC.7 Approval of perioperative pembrolizumab was based on findings from the global, phase 3, randomized KEYNOTE-671 study, which evaluated neoadjuvant pembrolizumab plus chemotherapy followed by surgery and adjuvant pembrolizumab versus neoadjuvant chemotherapy and surgery alone in participants with resectable stage II, IIIA, or IIIB (N2) NSCLC.4 Perioperative pembrolizumab significantly improved event-free survival (EFS; hazard ratio [HR] = 0.58; 95% confidence interval [CI]: 0.46–0.72; p < 0.001), major pathological response (MPR; 30.2% versus 11.0%, respectively; difference: 19.2%; p < 0.0001), and pathological complete response (pCR; 18.1% versus 4.0%; difference: 14.2%; p < 0.0001) at the first interim analysis and overall survival (OS; HR = 0.72, 95% CI: 0.56–0.93; p = 0.0052) at the second interim analysis.4,11 Among participants who underwent surgery in KEYNOTE-671, more participants achieved a complete resection (R0; 92% versus 84%) and lymph node downstaging (39% versus 28%) after neoadjuvant pembrolizumab plus chemotherapy versus neoadjuvant chemotherapy alone.12 Biological and ethnic differences in response to treatment in patients with NSCLC have been observed, with evidence demonstrating prolonged survival in Asian versus non-Asian patients after systemic anticancer therapy, including with anti–PD-(L)1 inhibitors.13, 14, 15 In addition, toxicity with immunotherapy may differ in Asian populations, with a higher incidence of adverse events (AEs) reported in Japanese versus non-Japanese patients with NSCLC.16, 17, 18 As such, it is important to evaluate treatment strategies in Japanese patients. Subgroup analyses of global studies have reported clinical benefits with neoadjuvant or adjuvant PD-(L)1 inhibitors in Japanese participants with resectable early-stage NSCLC.17,18 We report outcomes for Japanese participants enrolled in KEYNOTE-671 to determine the efficacy and safety of perioperative pembrolizumab in this population. Materials and Methods Study Design and Participants The study design of KEYNOTE-671 (clinicaltrials.gov: NCT03425643; registered February 7, 2018) has been published previously.4 Briefly, eligible participants were 18 years of age or older with previously untreated, pathologically confirmed, stage II, IIIA, or IIIB (N2) NSCLC per the American Joint Committee on Cancer (AJCC), eighth edition, that was deemed resectable after surgical consultation and investigator assessment. Additional eligibility criteria included provision of a tumor sample for assessment of PD-L1 expression, an Eastern Cooperative Oncology Group (ECOG) performance status of 0 or 1, and adequate organ function per defined laboratory parameters. The study was conducted in accordance with the principles of Good Clinical Practice and was approved by the appropriate institutional review board and regulatory agencies. All participants provided written informed consent before enrolling. Treatment Eligible participants were randomized 1:1 to receive 4 cycles of neoadjuvant pembrolizumab 200 mg or placebo administered intravenously every 3 weeks plus cisplatin 75 mg/m2 every 3 weeks and gemcitabine 1000 mg/m2 on days 1 and 8 of each 3-week cycle for squamous histology or cisplatin 75 mg/m2 and pemetrexed 500 mg/m2 every 3 weeks for nonsquamous histology, followed by surgery and 13 cycles (approximately 9 mo) of adjuvant pembrolizumab or placebo every 3 weeks. Surgery was to be performed within 20 weeks after the first dose of neoadjuvant therapy if the participant received four doses of neoadjuvant therapy or 4 to 8 weeks after the last dose of neoadjuvant therapy if the participant received fewer than four doses. Adjuvant therapy was to be initiated within 4 to 12 weeks after surgery. Treatment continued until the maximum number of cycles, disease progression or recurrence, unacceptable toxicity, or investigator or participant decision. Radiotherapy was only permitted for participants who did not undergo planned surgery for any reason other than local progression or metastatic disease or for participants with microscopic positive margins, gross residual disease, or extracapsular nodal extension after surgery. Randomization was stratified by disease stage (II versus III), PD-L1 tumor proportion score (TPS; <50% versus ≥50%), tumor histology (squamous versus nonsquamous), and geographic region (East Asia versus not East Asia). Endpoints The dual primary endpoints were EFS, defined as the time from randomization to first occurrence of local progression precluding planned surgery, unresectable tumor at the time of surgery, progression or recurrence per Response Evaluation Criteria in Solid Tumors, version 1.1, by investigator assessment, or death from any cause; and OS, defined as the time from randomization to death from any cause. The secondary endpoints included MPR, defined as up to 10% viable tumor cells in the resected primary tumor and lymph nodes; pCR, defined as the absence of residual invasive cancer in the resected primary tumor and lymph nodes (ypT0/Tis ypN0) as assessed on the basis of blinded, central examination by a pathologist; and safety. Assessments Tumor imaging was performed at baseline and then at weeks 7 and 13 after the first dose of neoadjuvant treatment. After surgery, imaging was performed within 4 weeks before initiating adjuvant treatment and then every 16 weeks. AEs were monitored throughout the study until 30 days after the last study treatment (or 90 d for serious AEs in the absence of new anticancer therapy) and graded in severity according to National Cancer Institute Common Terminology Criteria for Adverse Events version 4.03. Immune-mediated AEs and infusion reactions were determined based on a prespecified list of preferred terms intended to capture known risks of pembrolizumab and were considered regardless of attribution to treatment by the investigator. Statistical Analysis Efficacy was analyzed in the intention-to-treat (ITT) population, consisting of all randomized participants; safety was analyzed in the as-treated population, consisting of all participants who received at least 1 dose of study treatment. EFS and OS were estimated using the Kaplan–Meier method, with between-group differences assessed by the stratified log-rank test. The magnitude of treatment differences (HRs and 95% CIs) was calculated using a stratified Cox regression model with Efron’s method of tie handling. Between-group comparisons of the percentage of participants with MPR and pCR were performed using the stratified Miettinen and Nurminen method with strata weighting by sample size. These analyses in Japanese participants from the third interim analysis of KEYNOTE-671 were post hoc and descriptive, and no α was assigned. Results Participants There were 797 participants enrolled in the global KEYNOTE-671 study,4 of whom 82 were enrolled at sites in Japan. Of these, 39 participants were randomized to neoadjuvant pembrolizumab plus chemotherapy followed by surgery and adjuvant pembrolizumab (pembrolizumab arm), and 43 participants were randomized to neoadjuvant placebo plus chemotherapy followed by surgery and adjuvant placebo (placebo arm; Supplementary Fig. 1). Demographics and baseline clinical characteristics were generally balanced across treatment arms (Table 1). Overall, the median age of participants was 68 (range, 35–83) years, and most participants were current or former smokers (85%). In addition, 63% of participants had nonsquamous histology, 55% had stage III disease, and 66% had PD-L1 TPS less than 50% at baseline. Table 1Baseline Demographic and Clinical Characteristics in the ITT PopulationCharacteristicPembrolizumab Plus Chemotherapy n = 39Placebo Plus Chemotherapy n = 43Age, median (range), y68 (40–83)67 (35–79)Male, n (%)28 (72)36 (84)ECOG PS, n (%) 035 (90)43 (100) 14 (10)0Smoking status, n (%) Current3 (8)3 (7) Former27 (69)37 (86) Never9 (23)3 (7)Clinical disease stage, n (%) IIA5 (13)4 (9) IIB12 (31)16 (37) IIIA20 (51)21 (49) IIIB2 (5)2 (5)Tumor stage, n (%) T17 (18)10 (23) T215 (38)16 (37) T310 (26)10 (23) T47 (18)7 (16)Clinical node stage, n (%) N015 (38)18 (42) N112 (31)8 (19) N212 (31)17 (40)Histological features, n (%) Nonsquamous22 (56)30 (70) Squamous17 (44)13 (30)PD-L1 TPS, n (%) ≥50%15 (38)13 (30) <50%24 (62)30 (70) 1%–49%9 (23)13 (30) <1%15 (38)17 (40)EGFR mutation status, n (%) Yes3 (8)2 (5) No19 (49)25 (58) Unknown17 (44)16 (37)ALK translocation status, n (%) Yes1 (3)2 (5) No15 (38)20 (47) Unknown23 (59)21 (49)ECOG PS, Eastern Cooperative Oncology Group performance status; ITT, intention-to-treat; PD-L1, programmed cell death ligand 1; TPS, tumor proportion score. Baseline Demographic and Clinical Characteristics in the ITT Population ECOG PS, Eastern Cooperative Oncology Group performance status; ITT, intention-to-treat; PD-L1, programmed cell death ligand 1; TPS, tumor proportion score. Among Japanese participants, median time from randomization to database cutoff (August 19, 2024) was 52.7 (range, 32.7–74.9) months. In the pembrolizumab arm, 23 of 39 participants (59%) received 4 cycles of neoadjuvant therapy, 33 (85%) underwent in-study surgery (0 participants received radiotherapy), 28 (72%) received at least 1 dose of adjuvant pembrolizumab, and 18 (46%) completed treatment (Fig. 1A and B, Supplementary Fig. 1). In the placebo arm, 26 of 43 participants (60%) received 4 cycles of neoadjuvant therapy, 41 (95%) underwent surgery and/or radiotherapy, 31 (72%) received at least 1 dose of adjuvant placebo, and 24 (56%) completed treatment. In the pembrolizumab arm, seven of 12 participants (58%) who experienced disease progression or recurrence received at least 1 subsequent therapy, and 21 of 26 participants (81%) in the placebo arm received at least 1 subsequent therapy (Supplementary Table 1). Of these, two participants (17%) and 15 participants (58%), respectively, received subsequent therapy with a PD-(L)1 inhibitor–based regimen. Duration of follow-up and PD-L1 status for all participants in the (A) pembrolizumab arm and (B) placebo arm. The bar lengths represent the duration of follow-up for each participant. Histology, clinical stage, and PD-L1 TPS are illustrated for each participant. NS, nonsquamous; PD-L1, programmed cell death ligand 1; S, squamous; TPS, tumor proportion score. Surgical Outcomes Among the 33 participants in the pembrolizumab arm who underwent surgery, none had a surgical delay (defined as >20 wk between the first neoadjuvant dose and surgery [4 cycles of neoadjuvant therapy] or >8 wk between the last neoadjuvant dose and surgery [1–3 cycles of neoadjuvant therapy]); two of 39 participants (5%) in the placebo arm had a surgical delay, both due to an AE (Table 2). In the pembrolizumab arm, six participants (15%) did not have in-study surgery due to an AE (n = 5) or disease progression (n = 1); in the placebo arm, four participants (9%) did not have in-study surgery due to participant refusal (n = 1), physician decision (n = 1), or disease progression (n = 2). The most common surgical procedure was lobectomy, with 29 of 33 participants (88%) in the pembrolizumab arm and 35 of 39 (90%) in the placebo arm undergoing the procedure. The median (range) duration of hospital stay across procedures was 8 (4–20) days and 8 (7–15) days, respectively. In the pembrolizumab arm, 32 of 33 participants (97%) had a complete (R0) resection, and one (3%) had an incomplete (R2) resection; in the placebo arm, 37 of 39 participants (95%) had a complete (R0) resection, and two (5%) had an incomplete (R1) resection. Table 2Surgical Outcomes and Type of SurgeryOutcomePembrolizumab Plus Chemotherapy n = 39Placebo Plus Chemotherapy n = 43Participants without in-study surgery, n (%)6 (15)4 (9) Due to an AE5 (13)0 Due to participant refusal01 (2) Due to physician decision01 (2) Due to progressive disease1 (3)2 (5)Participants who underwent in-study surgeryn = 33n = 39Participants with surgical delay,an (%)02 (5) Due to an AE02 (5)Time from randomization to surgery, median (range), wk All participants15.3 (6.7–19.1)14.6 (4.6–19.9) Participants with surgical delay–13.6 (13.1–14.1)Time from last neoadjuvant dose to surgery, median (range), wk All participants5.7 (4.1–9.4)5.3 (3.1–9.9) Participants with surgical delay–9.5 (9.0–9.9)Duration of hospital stay, median (range), d8 (4–20)8 (7–15)Surgery type, n (%) Bilobectomy2 (6)4 (10) Lobectomy29 (88)35 (90) Pneumonectomy2 (6)0Surgical completeness, n (%) R032 (97)37 (95) R102 (5) R21 (3)0AE, adverse event.aSurgical delay was defined as more than 20 weeks between the first dose of neoadjuvant therapy and surgery for participants who received 4 cycles of neoadjuvant therapy and more than 8 weeks for those who received 1 to 3 cycles of neoadjuvant therapy. Surgical Outcomes and Type of Surgery AE, adverse event. Surgical delay was defined as more than 20 weeks between the first dose of neoadjuvant therapy and surgery for participants who received 4 cycles of neoadjuvant therapy and more than 8 weeks for those who received 1 to 3 cycles of neoadjuvant therapy. Efficacy Among those who received surgery, 13 of 33 participants (39%) in the pembrolizumab arm and eight of 39 participants (21%) in the placebo arm experienced nodal downstaging, all of whom experienced downstaging to pN0 disease (Fig. 2A). The degree of pathological regression, defined as percentage residual viable tumor (%RVT) after resection, among participants who received surgery in both treatment arms is found in Figure 2B. The median (IQR) %RVT after resection was 33% (2%–55%) in the pembrolizumab arm and 52% (35%–72%) in the placebo arm. (A) Nodal status at baseline and after surgery in participants who underwent in-study surgery and (B) percentage residual viable tumor size after resection among participants who received surgery in both treatment arms. In panel B, each tick mark on the x-axis represents an individual participant. pCR, pathological complete response; Pts, participants. By data cutoff, 15 participants (38%) in the pembrolizumab arm and 27 (63%) in the placebo arm had an EFS event (Supplementary Table 2). Median EFS (95% CI) was not reached (NR; 25.6 mo–NR) in the pembrolizumab arm and was 25.1 (13.6–56.1) months in the placebo arm; the HR was 0.54 (95% CI: 0.29–1.02; Fig. 3A). The 48-month EFS rates (95% CIs) were 60.3% (42.8%–74.0%) and 39.2% (24.8%–53.4%), respectively. With nine deaths (23%) in the pembrolizumab arm and 13 (30%) in the placebo arm, median OS (95% CI) was NR (56.9 mo–NR) and NR (NR–NR), respectively (HR = 0.76, 95% CI: 0.33–1.78; Fig. 3B). The 48-month OS rate (95% CI) was 79.5% (63.1%–89.2%) in the pembrolizumab arm and 68.9% (52.4%–80.7%) in the placebo arm. In the pembrolizumab arm, 11 participants (28%) achieved an MPR versus three participants (7%) in the placebo arm (difference: 21.2%, 95% CI: 5.2%–38.1%). Five participants (13%) in the pembrolizumab arm and one participant (2%) in the placebo arm achieved a pCR (difference: 10.5%, 95% CI: −1.1% to 24.8%). Kaplan–Meier estimates of (A) EFS and (B) OS in the ITT population. EFS, event-free survival; HR, hazard ratio; ITT, intention-to-treat; NR, not reached; OS, overall survival; Pts, participants. Safety Participants received a median (range) of 10 (1–17) administrations of pembrolizumab in the pembrolizumab arm and 15 (2–17) administrations of placebo in the placebo arm (Supplementary Table 3). Across the combined treatment phases, treatment-related AEs were reported in 39 of 39 participants (100%) in the pembrolizumab arm and 43 of 43 participants (100%) in the placebo arm, 23 (59%) and 20 (47%) of which, respectively, experienced grade 3 or 4 events, with no treatment-related AE deaths (Table 3). Treatment-related AEs led to treatment discontinuation in 13 participants (33%) in the pembrolizumab arm and six participants (14%) in the placebo arm. The most common treatment-related AEs were decreased neutrophil count (67% and 44% in the pembrolizumab and placebo arms, respectively), decreased white blood cell count (62% and 42%), constipation (54% and 67%), and nausea (49% and 63%). Treatment-related AEs are summarized by treatment phase in Supplementary Table 4 (neoadjuvant and surgical phase) and Supplementary Table 5 (adjuvant phase). AEs of any cause during the surgical phase among those who underwent surgery occurred in 26 of 33 participants (79%) in the pembrolizumab arm and 36 of 39 participants (92%) in the placebo arm (Supplementary Table 6), the most common of which was procedural pain (39% and 56%, respectively). Table 3Summary of Treatment-Related AEs Across the Combined Treatment Phases in the As-Treated PopulationAdverse EventPembrolizumab Plus Chemotherapy n = 39Placebo Plus Chemotherapy n = 43Treatment-related AEs, n (%)39 (100)43 (100) Grade 3/423 (59)20 (47) Serious9 (23)8 (19) Led to treatment discontinuation13 (33)6 (14) Led to death00Treatment-related AEs in ≥10% of participants in either treatment arm, n (%)Any GradeGrade 3/4Any GradeGrade 3/4 Neutrophil count decreased26 (67)15 (38)19 (44)15 (35) White blood cell count decreased24 (62)5 (13)18 (42)5 (12) Constipation21 (54)029 (67)0 Nausea19 (49)1 (3)27 (63)0 Decreased appetite15 (38)5 (13)14 (33)0 Malaise12 (31)014 (33)1 (2) Anemia10 (26)1 (3)18 (42)1 (2) Platelet count decreased10 (26)2 (5)13 (30)4 (9) Alanine aminotransferase increased9 (23)2 (5)7 (16)1 (2) Blood creatinine increased8 (21)05 (12)0 Hiccups8 (21)017 (40)0 Fatigue7 (18)01 (2)0 Aspartate aminotransferase increased6 (15)2 (5)6 (14)0 Stomatitis6 (15)04 (9)0 Dysgeusia4 (10)08 (19)0 Eczema4 (10)000 Peripheral sensory neuropathy1 (3)06 (14)0AE, adverse event. Summary of Treatment-Related AEs Across the Combined Treatment Phases in the As-Treated Population AE, adverse event. Immune-mediated AEs and infusion reactions occurred in 17 participants (44%) in the pembrolizumab arm and three participants (7%) in the placebo arm across treatment phases, with grade 3 or 4 events occurring in seven participants (18%) and one participant (2%), respectively (Supplementary Table 7); there were no grade 5 immune-mediated AEs or infusion reactions. The most frequent immune-mediated AEs and infusion reactions were hypothyroidism (10% and 2% with pembrolizumab and placebo, respectively), pneumonitis (8% with pembrolizumab), and severe skin reactions (8% with pembrolizumab). Immune-mediated AEs and infusion reactions that occurred during the neoadjuvant and surgical phase are found in Supplementary Table 8 and during the adjuvant phase in Supplementary Table 9. Among participants who had in-study surgery, one of 33 participants in the pembrolizumab arm died within 90 days of surgery due to progression of NSCLC, and one of 39 participants in the placebo arm died within 30 days of surgery due to pneumonia. Discussion In this subgroup analysis of Japanese participants from the phase 3 KEYNOTE-671 study, perioperative pembrolizumab added to neoadjuvant chemotherapy numerically improved EFS, OS, MPR, and pCR compared with neoadjuvant chemotherapy alone, consistent with the global KEYNOTE-671 study.4,11 Neoadjuvant pembrolizumab improved surgical outcomes and did not adversely affect the type of surgical procedure, delay surgery, or prolong the hospital stay. In addition, safety was manageable and no new safety signals were identified in this population. Importantly, to our knowledge, this is the first analysis from a large phase 3 clinical trial demonstrating the benefits of perioperative immune checkpoint inhibition in Japanese patients with early stage, resectable NSCLC. After a median follow-up of more than 4 years, prolonged EFS was observed with perioperative pembrolizumab versus neoadjuvant chemotherapy and surgery alone in Japanese participants, with an HR of 0.54 (95% CI: 0.29–1.02) and higher 48-month EFS rates in the pembrolizumab versus placebo arm (60.3% versus 39.2%). HR for OS (0.76; 95% CI: 0.33–1.78) and 48-month OS rates (79.5% versus 68.9% for the pembrolizumab and placebo arms, respectively) suggest an improvement in OS with perioperative pembrolizumab, median OS was NR in both treatment groups due to the low number of events (nine in the pembrolizumab arm versus 13 in the placebo arm). MPR and pCR were also improved with pembrolizumab (MPR, 28% versus 7%; pCR, 13% versus 2%) compared with chemotherapy. Although these analyses were descriptive only, the benefits observed overall with perioperative pembrolizumab in Japanese participants are generally consistent with those reported previously from the global KEYNOTE-671 study, which revealed statistically significant improvements in EFS, MPR, pCR, and OS.4,11 This analysis also evaluated surgical outcomes in Japanese participants who received in-study surgery after neoadjuvant treatment. Overall, surgical outcomes after neoadjuvant pembrolizumab plus chemotherapy reported here were generally consistent with those from the global KEYNOTE-671 study, with no effects on surgical procedure choice, surgical delays, or hospital stay duration versus chemotherapy alone.12 Surgical completeness was similar regardless of treatment, with 97% of participants in the pembrolizumab arm and 95% of participants in the placebo arm achieving a complete R0 resection. Postsurgery mortality was infrequent and similar regardless of treatment, with one death reported in each treatment arm within 90 days of surgery. The benefits of neoadjuvant pembrolizumab were also demonstrated in this analysis, particularly with regard to lymph node downstaging in those who underwent surgery; 39% of participants achieved nodal downstaging to N0 with neoadjuvant pembrolizumab plus chemotherapy versus 21% with neoadjuvant chemotherapy alone. Moreover, neoadjuvant pembrolizumab was associated with greater pathological regression, based on %RVT after resection, compared with chemotherapy alone. Evidence regarding neoadjuvant or adjuvant immune checkpoint inhibition in Japanese populations is limited. In the phase 3 CheckMate 816 study of participants with stage IB to IIIA NSCLC (AJCC seventh edition), neoadjuvant nivolumab (anti–PD-1) plus chemotherapy demonstrated improved EFS (HR = 0.60, 95% CI: 0.30–1.24), MPR (39% versus 11%), and pCR (30% versus 6%) versus neoadjuvant chemotherapy alone in Japanese participants.18 In addition, the phase 3 IMpower010 study of participants with stage IB to IIIA NSCLC (AJCC seventh edition) revealed prolonged disease-free survival with adjuvant atezolizumab (anti–PD-L1) compared with best supportive care after resection (HR = 0.61, 95% CI: 0.34–1.10).17 The effects of perioperative treatment with PD-(L)1 inhibitors other than pembrolizumab have not been evaluated in Japanese populations specifically; however, Japanese participants have been enrolled in global studies, including the phase 3 CheckMate 77T study in stage IIA to IIIB NSCLC, which revealed improved EFS with perioperative nivolumab plus neoadjuvant chemotherapy compared with neoadjuvant chemotherapy alone (HR = 0.58, 95% CI: 0.42–0.81).5 In addition, in the global phase 3 AEGEAN study of stage II to IIIB NSCLC, perioperative durvalumab plus neoadjuvant chemotherapy improved EFS compared with neoadjuvant chemotherapy alone (HR = 0.68, 95% CI: 0.53–0.88).3 In both of these studies, as well as in the global KEYNOTE-671 study, improved EFS with perioperative immune checkpoint inhibition was generally maintained across subgroups by geographic regions including Asia.3, 4, 5 This is consistent with findings in the advanced and metastatic disease settings, in which global population outcomes in Japanese patients treated with immune checkpoint inhibitors are largely similar to those in the global study population.19, 20, 21 Taken together, these findings demonstrate benefits with perioperative immunotherapy in Japanese patients with early-stage resectable NSCLC. Notably, perioperative pembrolizumab has been approved in Japan and is recommended with neoadjuvant platinum-based chemotherapy for patients with stage II to IIIB NSCLC by the Japan Lung Cancer Society.22,23 The findings presented here support this recommendation. Safety with perioperative pembrolizumab plus neoadjuvant chemotherapy in Japanese participants was manageable, with no new safety signals reported. Grade 3 or higher treatment-related AEs were generally similar between treatment arms across the combined treatment phases and were reported in 59% and 47% of participants in the pembrolizumab and placebo arms, respectively. Similar to the global KEYNOTE-671 study,4,11 in this analysis, the most common treatment-related AEs reported in Japanese participants were decreased neutrophil count, decreased white blood cell count, constipation, and nausea. Although the grade 3 to 5 treatment-related AE rates are slightly higher than in the global KEYNOTE-671 population (45% and 38% in the pembrolizumab and placebo arms, respectively),11 comparisons between study populations should be made with caution considering the smaller sample size and longer follow-up of this analysis. Moderately worse toxicity with systemic anticancer therapies in Japanese participants with early-stage NSCLC is not an unexpected finding, as higher rates of grade 3 or higher treatment-related AEs were reported in Japanese versus non-Japanese participants in CheckMate 816 (59% versus 28%, respectively).18 Increased toxicity has also been observed with anti–PD-(L)1 inhibitors plus chemotherapy in Japanese populations with metastatic NSCLC.19,21,24 Overall, the safety findings observed here in Japanese participants were generally consistent with the global KEYNOTE-671 population. In conclusion, this analysis of Japanese participants with stage II, IIIA, or IIIB (N2) resectable NSCLC from KEYNOTE-671 demonstrated numerically improved EFS, MPR, and pCR with perioperative pembrolizumab plus neoadjuvant chemotherapy compared with neoadjuvant chemotherapy alone. Moreover, perioperative pembrolizumab numerically improved lymph node downstaging, particularly to N0, and resulted in greater pathological regression versus chemotherapy. Safety with perioperative pembrolizumab was manageable, and no new safety signals were observed. Overall, these findings support the use of perioperative pembrolizumab in a Japanese population with early-stage resectable NSCLC. Data-Sharing Statement Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA (MSD), is committed to providing qualified scientific researchers access to anonymized data and clinical study reports from the company’s clinical trials for the purpose of conducting legitimate scientific research. MSD is also obligated to protect the rights and privacy of trial participants and, as such, has a procedure in place for evaluating and fulfilling requests for sharing company clinical trial data with qualified external scientific researchers. The MSD data-sharing website (available at: https://externaldatasharing-msd.com/) outlines the process and requirements for submitting a data request. Applications will be promptly assessed for completeness and policy compliance. Feasible requests will be reviewed by a committee of MSD subject matter experts to assess the scientific validity of the request and the qualifications of the requestors. In line with data privacy legislation, submitters of approved requests must enter into a standard data-sharing agreement with MSD before data access is granted. Data will be made available for request after product approval in the United States and European Union or after product development is discontinued. There are circumstances that may prevent MSD from sharing requested data, including country- or region-specific regulations. If the request is declined, it will be communicated to the investigator. Access to genetic or exploratory biomarker data requires a detailed, hypothesis-driven statistical analysis plan that is collaboratively developed by the requestor and MSD subject matter experts; after approval of the statistical analysis plan and execution of a data-sharing agreement, MSD will either perform the proposed analyses and share the results with the requestor or will construct biomarker covariates and add them to a file with clinical data that are uploaded to an analysis portal so that the requestor can perform the proposed analyses. CRediT Authorship Contribution Statement Masahiro Tsuboi: Data curation, Investigation, Resources, Writing – review & editing. Hisashi Saji: Data curation, Investigation, Resources, Writing – review & editing. Morihito Okada: Data curation, Investigation, Resources, Writing – review & editing. Norihiko Ikeda: Data curation, Investigation, Resources, Writing – review & editing. Kazuya Takamochi: Data curation, Investigation, Resources, Writing – review & editing. Kenji Sugio: Data curation, Investigation, Resources, Writing – review & editing. Yoshitsugu Horio: Data curation, Investigation, Resources, Writing – review & editing. Fumihiro Tanaka: Data curation, Investigation, Resources, Writing – review & editing. Wataru Nishio: Data curation, Investigation, Resources, Writing – review & editing. Satoshi Muto: Data curation, Investigation, Resources, Writing – review & editing. Hidehito Horinouchi: Data curation, Investigation, Resources, Writing – review & editing. Heather Wakelee: Conceptualization, Methodology, Data curation, Investigation, Resources, Writing – review & editing. Shirong Han: Conceptualization, Methodology, Writing – review & editing. Tomoko Mori: Conceptualization, Methodology, Writing – review & editing. Steven M. Keller: Conceptualization, Methodology, Writing – review & editing. Terufumi Kato: Data curation, Investigation, Resources, Writing – review & editing. Disclosure All authors from participating clinical sites received research funding from Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA, for the conduct of this study. Dr. Tsuboi reports receiving funding to the institution from AstraZeneca, AstraZeneca KK, Ono Pharmaceutical Co., Ltd., Bristol Myers Squibb KK, Eli Lilly, Novartis, and MiRXES; serving as an advisor for AstraZeneca KK, MSD, Chugai Pharmaceutical Co., Ltd., Novartis, Ono Pharmaceutical Co., Ltd., Bristol Myers Squibb KK, and MiRXES; serving as a consultant for Chugai Pharmaceutical Co., Ltd.; serving as a speaker for Johnson & Johnson Japan (Ethicon), AstraZeneca KK, Eli Lilly Japan, Chugai Pharmaceutical Co., Ltd., Taiho Pharma, Medtronic Japan, Ono Pharmaceutical Co., Ltd., MSD, Bristol Myers Squibb KK, Daiichi Sankyo, and Amgen KK. Dr. Saji reports receiving research funding to the department from Chugai-Pharma, Taiho, Boehringer Ingelheim, Lilly, Ethicon, and Covidien and honoraria from MSD, Boehringer Ingelheim, Ethicon, Covidien, Chugai-Pharma, Astellas Pharma, Fujifilm Medical, Bristol Myers Squibb, Takeda, AstraZeneca, CSL Behring, and Taiho. Dr. Ikeda has received research funding to the department from AstraZeneca, Chugai Pharma, Boehringer Ingelheim, Taiho Pharma, Eli Lilly, Ono Pharma, BMS, MSD, Nihon Mediphysics, Teijin Pharma, Eisai, Daiichi Sankyo, Fuji film, and Johnson & Johnson; honoraria from AstraZeneca, Chugai Pharma, Boehringer Ingelheim, Taiho Pharma, Eli Lilly, Ono Pharma, BMS, Olympus, MSD, Johnson & Johnson, Nihon Mediphysics, Medtronics, and Teijin Pharma; and served as a past president of the Japan Surgical Society and a past president of the Japan Lung Cancer Society. Dr. Takamochi reports receiving honoraria from MSD, Ethicon, Covidien, Chugai-Pharma, and AstraZeneca. Dr. Sugio reports receiving research funding to the department from MSD, Eli Lilly, and Covidien Japan; and honoraria from Chugai Pharma, AstraZeneca, Eli Lilly, MSD, Taiho Pharma, BMS, and Boehringer Ingelheim. Dr. Tanaka reports receiving honoraria from AstraZeneca, Bristol Myers Squibb, Boehringer Ingelheim Japan, Chugai Pharmaceutical, ONO pharmaceutical, MSD, Johnson & Johnson, Covidien Japan, Taiho Pharmaceutical, Eli Lilly Japan, Kyowa-Kirin, Takeda Pharmaceutical, Pfizer, Olympus, Stryker, and Intuitive Japan; and research grants from AstraZeneca, Bristol Myers Squibb, Boehringer Ingelheim Japan, Ono Pharmaceutical, Taiho Pharmaceutical, Eli Lilly Japan, Chugai Pharmaceutical, and MSD. Dr. Nishio reports receiving research funding to the institution from MSD, AstraZeneca, and Chugai Pharmaceuticals. Dr. Horinouchi reports receiving grants or contracts from MSD, AbbVie, AstraZeneca, BMS, Ono, Daiichi Sankyo, Janssen, and Chugai/Roche; receiving honoraria from AstraZeneca, MSD, Ono, BMS, Chugai/Roche, Amgen, and AbbVie; and having advisory board participation for AstraZeneca, Chugai/Roche, Ono, BMS, and AbbVie. Dr. Wakelee has received research funding to the institution from Bayer, AstraZeneca, BMS, Genentech/Roche, Gilead (via IIT mechanism), MSD, Helsinn, SeaGen, and Xcovery; serving as a compensated advisory board member for IOBiotech, OncoC4, GSK, and BeiGene (now BeOne Medicines); serving as an uncompensated advisory board member for MSD, Genentech/Roche, BMS, and AstraZeneca; and received travel support for CME lectures from Chugai. Dr. Han and Ms. Mori are employees of MSD K.K., Tokyo, Japan, and hold stock options in Merck & Co., Inc., Rahway, NJ, USA. Dr. Keller is an employee of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA, and shareholder of Merck & Co., Inc., Rahway, NJ, USA. Dr. Kato reports receiving honoraria to self for IDMC membership, advisory board meetings, or serving as a speaker from AbbVie, AstraZeneca, Boehringer Ingelheim, Chugai, Daiichi Sankyo, Merck Biopharma, MSD, Johnson & Johnson, Novartis, Novocure, Pfizer, Taiho, and Takeda; receiving grants from commissioned/joint research to institution to support study conduct from Amgen, Arrivent, AstraZeneca, Bayer, BeiGene, BluePrint Medicines, Bristol Meyers Squibb, Chugai, Daiichi Sankyo, Eli Lilly, Gilead, GlaxoSmithKline, Johnson & Johnson, Merck Biopharma, MSD, Novocure, Pfizer, Regeneron, and Takeda. The remaining authors declare no conflict of interest. Acknowledgments The authors thank the participants and their families and caregivers for participating in this study, along with all investigators and site personnel. The authors also thank Akira Kurisu of MSD K.K., Tokyo, Japan, for statistical support and Kayo Mochizuki and Masaru Watanabe, both of MSD K.K., Tokyo, Japan, for publication support. Medical writing assistance was provided by Aisling Towell, PhD, of ICON plc (Blue Bell, PA, USA). This assistance was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. Funding for this research was provided by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. Footnotes Cite this article as: Tsuboi M, Saji H, Okada M, et al. Perioperative pembrolizumab in Japanese participants with early-stage non–small-cell lung cancer: subgroup analysis from KEYNOTE-671. JTO Clin Res Rep 2026;7:101009 This work is dedicated to the memory of the late Dr. Masahiro Tsuboi. Note: To access the supplementary material accompanying this article, visit the online version of the JTO Clinical and Research Reports at www.jtocrr.org and at https://doi.org/10.1016/j.jtocrr.2026.101009. Supplementary Data Kato_Supplementary Material References Associated Data Supplementary Materials Kato_Supplementary Material