First-line pembrolizumab + chemotherapy in Japanese patients with advanced/metastatic esophageal cancer from KEYNOTE-590

Background The phase 3 KEYNOTE-590 (NCT03189719) study showed first-line pembrolizumab plus chemotherapy significantly prolonged overall survival and progression-free survival versus placebo plus chemotherapy in patients with advanced unresectable or metastatic adenocarcinoma or squamous cell carcinoma of the esophagus or advanced/metastatic Siewert type I adenocarcinoma of the esophagogastric junction. We describe a subgroup analysis of Japanese patients from KEYNOTE-590. Methods Eligible patients were randomly assigned 1:1 to pembrolizumab 200 mg intravenously every 3 weeks or placebo plus chemotherapy (cisplatin 80 mg/m2 and 5-fluorouracil 800 mg/m2/day). Efficacy was evaluated in all Japanese patients and those with esophageal squamous cell carcinoma and programmed death ligand 1 combined positive score ≥ 10. Dual primary endpoints were overall survival and progression-free survival per RECIST v1.1 by investigator. Secondary endpoints included objective response rate per RECIST v1.1 by investigator and safety and tolerability. Results At data cutoff (July 2, 2020), 141 Japanese patients were randomly assigned (pembrolizumab plus chemotherapy, 74; placebo plus chemotherapy, 67). In all Japanese patients, median overall survival was 17.6 months with pembrolizumab plus chemotherapy versus 11.7 months with chemotherapy (hazard ratio, 0.71; 95% confidence interval, 0.47–1.09), median progression-free survival was 6.3 versus 6.0 months (hazard ratio, 0.58; 95% confidence interval, 0.40–0.84), and objective response rate was 56.8% versus 38.8%. Grade 3–5 treatment-related adverse events were 74.3% and 61.2%. Conclusion First-line pembrolizumab plus chemotherapy demonstrated improvement in overall survival and progression-free survival compared with placebo plus chemotherapy in Japanese patients with advanced/metastatic esophageal cancer; safety was comparable between treatment groups. Clinical trial registry ClinicalTrials.gov, NCT03189719. Supplementary Information The online version contains supplementary material available at 10.1007/s10388-022-00920-x.


Introduction
In 2018, 20,000 cases and 12,000 deaths from esophageal cancer (EC) were reported in Japan [1]. The two primary subgroups of EC are esophageal squamous cell carcinoma (ESCC) and adenocarcinoma, and they vary in etiology and geographic distribution. ESCC is the predominant type of EC in East Asia, including Japan, where the ratio of ESCC to adenocarcinoma is 26:1 [2].
Practice guidelines in Japan recommend cisplatin plus 5-fluoruracil (5-FU) as first-line therapy for 1 3 unresectable advanced or recurrent EC [3]. The median duration of survival for Japanese patients with advanced EC receiving chemotherapy is < 8.1 months, highlighting the unmet need for these patients [4]. Pembrolizumab is approved in Japan for the treatment of patients with radically unresectable advanced or recurrent ESCC of the esophagus whose tumors express PD-L1 (combined positive score [CPS] ≥ 10) with disease progression after ≥ 1 line of chemotherapy based on data from the phase 3 KEYNOTE-181 study [5][6][7].
Data from KEYNOTE-590 in patients with untreated advanced esophageal and esophagogastric junction (EGJ) cancer demonstrated that first-line pembrolizumab-chemotherapy is a new standard of care. Herein, we present the results in the Japanese population of the KEYNOTE-590 study.

Study design, treatment, and participants
The study design of the randomized, double-blind, phase 3 KEYNOTE-590 trial has been published [7][8][9]. Briefly, eligible patients had treatment-naive, histologically or cytologically confirmed, locally advanced unresectable or metastatic esophageal adenocarcinoma, ESCC, or locally advanced or metastatic Siewert type I adenocarcinoma of the EGJ. Previous treatment with curative intent, including neoadjuvant or adjuvant treatment, was permissible if disease progression occurred > 6 months after cessation of treatment. Patients were randomly assigned 1:1 to receive intravenous (IV) pembrolizumab 200 mg or placebo (normal saline) every 3 weeks (Q3W) for up to 35 cycles (~ 2 years) plus chemotherapy (cisplatin 80 mg/m 2 IV Q3W for ≤ 6 doses and 5-FU 800 mg/m 2 /day continuous IV infusion on days 1-5 Q3W per local standard) until disease progression, unacceptable toxicity, or withdrawal of consent. Randomization was stratified by geographic region (Asia vs. non-Asia), histology (adenocarcinoma vs. ESCC), and Eastern Cooperative Oncology Group (ECOG) performance status (0 vs. 1).

Outcomes and assessments
Assessments of primary efficacy and safety outcomes have been described [8,9]. In the current analysis, efficacy and safety endpoints were assessed in patients enrolled at Japanese sites. Dual primary endpoints were OS and PFS per RECIST v1.1 by investigator assessment. Secondary endpoints included ORR and duration of response (DOR) per RECIST v1.1 by investigator assessment, safety and tolerability, and health-related quality of life.
PD-L1 expression was assessed in archival or newly collected tumor samples using PD-L1 IHC 22C3 pharmDx (Agilent) and measured using CPS (defined as the number of PD-L1-staining cells [tumor cells, lymphocytes, macrophages] divided by the total number of viable tumor cells, multiplied by 100).

Statistical analysis
In the Japanese population, efficacy was evaluated in the intention-to-treat population and in the ESCC, PD-L1 CPS ≥ 10, and ESCC PD-L1 CPS ≥ 10 subgroups as specified in the protocol. OS and PFS were estimated using the nonparametric Kaplan-Meier method, and treatment differences were assessed using a Cox proportional hazards model with Efron's method of tie handling to estimate the magnitude of the treatment difference (HR). The Japanese subgroup analysis was not controlled for multiplicity, and no alpha was allocated to the comparisons. The estimated sample size of the Japanese population was calculated to guarantee that there is >80% probability of consistency between overall and Japanese populations on the primary endpoint of OS. Consistency was defined as the probability that the estimated HRs for the overall and Japanese populations are both < 1. The necessary sample size to achieve >80% probability of consistency in the subgroup of patients with ESCC PD-L1 CPS ≥ 10 was 55 and 45 for patients with ESCC. With 141 Japanese patients enrolled into the study, the probabilities increased to 91.6% for patients with ESCC PD-L1 CPS ≥ 10 and 93.6% for patients with ESCC.
Data cutoff for protocol-specified interim OS and final PFS analyses was July 2, 2020. This trial is registered with ClinicalTrials.gov (NCT03189719).

Response
In

Discussion
Poor prognosis and limited treatment options highlight the unmet need in Japanese patients with advanced EC.
In this subgroup analysis of Japanese patients enrolled in KEYNOTE-590, first-line pembrolizumab-chemotherapy prolonged OS and PFS and improved response rates over placebo-chemotherapy. The safety profile was comparable between treatment groups, and no new safety signals were detected. These findings in Japanese patients are consistent with the data reported in the total population of KEYNOTE-590 [9]. Use of a two-drug cytotoxic chemotherapy regimen for first-line treatment in Japanese patients with advanced EC has shown response rates of 20-60% [3]. However, median survival times remain low, as described in a retrospective analysis of cisplatin plus 5-FU treatment in which the median OS was 10.4 months [10]. Three-drug chemotherapy regimens have demonstrated improved response rates compared with two-drug regimens; however, they are associated with higher toxicities, and the impact on survival is unknown [3,11,12]. Targeted therapy with trastuzumab has improved clinical outcomes in HER2-positive gastric or gastroesophageal cancer when added to first-line chemotherapy and is included in National Comprehensive Cancer Network and Pan-Asian European Society for Medical Oncology guidelines [11,13,14]. Other targeted therapies have not been successful in improving clinical outcomes as first-line treatment when combined with chemotherapy for advanced EC [15,16].
Data with first-line anti-PD-1/PD-L1 treatment in patients with EC are limited but promising. Pembrolizumab and nivolumab have demonstrated significant improvements in survival with second-line treatment in advanced EC [6,17]. In KEYNOTE-181, pembrolizumab prolonged OS compared with chemotherapy in patients with ESCC PD-L1 CPS ≥ 10 (HR, 0.64; 95% CI 0.46-0.90) [6,18]. Pembrolizumab is approved in the United States and Japan for second-line treatment of patients with ESCC PD-L1 CPS ≥ 10 [5,18]. In ATT RAC TION-3, nivolumab significantly improved OS compared with chemotherapy in patients with ESCC (HR, 0.77; 95% CI 0.62-0.96; P = 0.019) [17]. Based on these data, nivolumab is approved in the United States and Japan for second-line treatment of patients with ESCC [19]. Recent results from the KEYNOTE-590 primary analysis (ESCC and adenocarcinoma) and CheckMate-649 (gastric cancer and esophageal adenocarcinoma) studies have demonstrated that first-line treatment with immune checkpoint inhibitors has the potential to be standard of care for patients with EC. In KEYNOTE-590, pembrolizumab-chemotherapy was superior to placebo-chemotherapy for OS (HR, 0.73; P < 0.0001) and PFS (HR, 0.65; P < 0.0001) in patients with unresectable locally advanced metastatic esophageal adenocarcinoma or ESCC or Siewert type I EGJ adenocarcinoma [9]. In CheckMate-649, nivolumab plus chemotherapy was compared with chemotherapy as first-line treatment of patients with unresectable or metastatic gastric cancer, gastroesophageal junction cancer, or esophageal adenocarcinoma; patients with ESCC were not included in this study [20]. In data presented thus far, the benefit of nivolumab plus chemotherapy for patients with esophageal cancer, regardless of histology, is unclear.
In this analysis from KEYNOTE-590, pembrolizumab-chemotherapy improved clinical outcomes compared with placebo-chemotherapy in the Japanese population. Although subpopulation sample sizes were small in the Japanese population, clinical outcomes were further improved in patients with ESCC, PD-L1 CPS ≥ 10, and ESCC PD-L1 CPS ≥ 10 compared with the overall Japanese population. Median OS in the pembrolizumab-chemotherapy group was longer in the Japanese population than in the total population [9]. Median OS and PFS were 17.6 and 6.3 months in the Japanese population and 12.4 and 6.3 months in the total population; the higher percentage of patients with ESCC and PD-L1 CPS ≥ 10 receiving subsequent systemic therapy in the Japanese population may explain this result. Most patients in the pembrolizumab-chemotherapy (68.8%) and the placebo-chemotherapy (75.4%) groups received subsequent systemic therapy; nonetheless, pembrolizumab-chemo therapy-treated patients experienced clinically meaningful improvement in OS compared with placebo-chemotherapy-treated patients. A higher percentage of patients in the Japanese population than in the total population received subsequent therapy (pembrolizumab-chemotherapy, 68.8% vs. 43.5%; placebo-chemotherapy, 75.4% vs. 47.8%), potentially contributing to improved outcomes in the pembrolizumab group in the Japanese population. Although the study was double-blind, patients received immune checkpoint inhibitor therapy at a lower rate in the pembrolizumab-chemotherapy group (10.9%) than in the placebo-chemotherapy group (16.9%), but this did not appear to have an impact on clinical outcomes. Notably, a higher proportion of Japanese patients had better ECOG performance status at baseline in both the pembrolizumab-chemotherapy group and the placebo-chemotherapy group (ECOG performance status 0, 64.9% and 79.1%, respectively) compared to the same groups in the total population (ECOG performance status 0, 40% and 40%, respectively) [7], which could have contributed to the higher proportion of Japanese patients who received subsequent therapy. Responses to pembrolizumab-chemotherapy were durable in Japanese patients and in line with the total population (median DOR, 8.3 months).
The safety profile of pembrolizumab-chemotherapy was consistent between Japanese patients and the total population [9]. Treatment-related AEs were reported in 98.6% patients in the Japanese population versus 98.4% in the total population; grade 3-5 treatment-related AEs were reported in 74.3% versus 71.9% patients, respectively.
A limitation of this study is the evaluation of a subgroup of patients from a larger clinical trial in which Japanese patients represented approximately 20% of the total population. Given that 90% of the Japanese population had ESCC, evaluation of clinical outcomes in the overall Japanese population may not be directly comparable to those in the total population (73% ESCC). Small sample sizes in the ESCC, PD-L1 CPS ≥ 10, and ESCC PD-L1 CPS ≥ 10 Japanese subgroups limit conclusions about the improved outcomes with pembrolizumab-chemotherapy versus placebo-chemotherapy.
Pembrolizumab-chemotherapy improved clinical outcomes, including OS, PFS, and ORR, compared with placebo-chemotherapy in Japanese patients with treatment-naive advanced EC. The safety profile was comparable between the two treatment groups, and there were no new safety signals for pembrolizumab in the Japanese population. These data suggest that pembrolizumab-chemotherapy should be considered a new first-line treatment option for all Japanese patients with unresectable recurrent or advanced EC.

Acknowledgements
The authors thank the patients and their families and caregivers and all primary investigators and site personnel for participating in the study. Medical writing and/or editorial assistance was provided by Kathleen Richards, PhD, and Holly C. Cappelli, PhD, CMPP, of ApotheCom (Yardley, PA, USA). This assistance was funded by Merck Sharp and Dohme Corp, a subsidiary of Merck and Co., Inc., Kenilworth, NJ, USA.
Author contributions TK conceptualization, data acquisition, writing-reviewing and editing. HH; data acquisition, data interpretation, writing-reviewing and editing. AT data acquisition, data interpretation, writing-drafting, reviewing and editing. HY data acquisition, writing-reviewing and editing. KM data acquisition, writing-reviewing and editing. TS data acquisition, data interpretation, writing-reviewing and editing. TO data acquisition, writing-reviewing and editing. RI data acquisition, writing-reviewing and editing. MG data acquisition, writing-reviewing and editing. HB conceptualization, data acquisition, writing-reviewing and editing. TN data acquisition, data interpretation, writing-reviewing and editing. SH formal analysis, writing-reviewing and editing. TS formal analysis, writing-reviewing and editing. NY data interpretation, writing-reviewing and editing. TD conceptualization, data acquisition, data interpretation, writing-drafting, reviewing and editing. KK conceptualization, data acquisition, formal analysis, data interpretation, writing-reviewing and editing.
Funding Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. The study was funded and designed under the responsibility of Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA who also provided the study drug(s) for this trial. The study funder participated in study design, data interpretation, and writing of this report. The sponsor maintained the study database. All authors had full access to the study data, were involved in writing or reviewing and editing drafts of the manuscript, and approved the decision to submit for publication.
Data availability Data collection was provided by each clinical trial site and data were analyzed by the authors. All authors had full access to the data and take responsibility for the integrity of the data and the accuracy of the data analysis. Medical writing and/or editorial assistance was funded by Merck Sharp & Dohme LLC, a subsidiary of Merck & Co., Inc., Rahway, NJ, USA. and Bayer; and served on advisory boards of AstraZeneca and Ono Pharmaceutical Co., Ltd. Taroh Satoh has received grants from MSD, Ono Pharmaceutical Co., Ltd., Yakult Honsha, Chugai Pharmaceutical Co. Ltd., Gilead Sciences, BeiGene, Eli Lily and Company, and Taiho Pharmaceutical Co., Ltd; and consulting fees from Takara-Bio Inc. Hideo Baba has received a grant from MSD K.K. Tomohiro Nishina has received a grant from MSD and honoraria from Ono Pharmaceutical Co., Ltd., Taiho, Takeda, Chugai Pharmaceutical Co. Ltd., and Daiichi Sankyo. Toshihiko Doi has received grants from Lilly, MSD, Daiichi Sankyo, Sumitomo Dainippon, Taiho, Novartis, Merck Serono, Janssen Pharma, Boehringer Ingelheim, Pfizer, Bristol Myers Squibb, AbbVie, IQVIA, and Eisai; consulting fees from Sumitomo Dainippon, Taiho, Takeda, Chugai Pharmaceutical Co. Ltd. Pharma, AbbVie, Bayer, Rakuten Medical, and Otuska Pharma; honoraria from Bristol Myers Squibb, Rakuten Medical, Ono Pharma, Oncolys BioPharma, and Taiho; and advises on the boards of MSD, Daiichi Sankyo, Amgen, Novartis, Boehringer Ingelheim, Janssen Pharma, AbbVie, Bayer, and Astellas Pharma. Ken Kato has received grants from MSD, Ono Pharmaceutical Co., Ltd., Bristol Myers Squibb, BeiGene, Shionogi, Merck Biopharma, Oncolys Biopharma, Chugai Pharmaceutical Co. Ltd., Bayer; honoraria from Ono and Bristol Myers Squibb; and advises on the boards of Bristol Myers Squibb, BeiGene, and Ono Pharmaceutical Co., Ltd. Masahiro Goto, Shirong Han, Ryu Ishihara, Takashi Ogata, Tomoko Sakata, and Naoyoshi Yatsuzuka report no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.