First-line pembrolizumab ± chemotherapy for recurrent/metastatic head and neck cancer: Japanese subgroup of KEYNOTE-048

Background Here, we report the results of the Japanese subgroup of the phase 3 KEYNOTE-048 study of pembrolizumab alone, pembrolizumab plus platinum and 5-fluorouracil (pembrolizumab–chemotherapy), or cetuximab plus platinum and 5-fluorouracil (EXTREME) in previously untreated recurrent/metastatic (R/M) head and neck squamous cell carcinoma (HNSCC). Methods Primary end points were overall survival (OS) and progression-free survival (PFS). Efficacy was evaluated in patients with PD-L1 combined positive score (CPS) ≥ 20 and ≥ 1 and the total Japanese subgroup (n = 67). Results At data cutoff (25 February 2019), pembrolizumab led to longer OS versus EXTREME in the PD-L1 CPS ≥ 20 subgroup (median, 28.2 vs. 13.3 months; HR, 0.29 [95% CI 0.09–0.89]) and to similar OS in the total Japanese (23.4 vs. 13.6 months; HR, 0.51 [95% CI 0.25–1.05]) and CPS ≥ 1 subgroups (22.6 vs. 15.8 months; HR, 0.66 [95% CI 0.31–1.41]). Pembrolizumab–chemotherapy led to similar OS versus EXTREME in the PD-L1 CPS ≥ 20 (median, 18.1 vs. 15.8 months; HR, 0.72 [95% CI 0.23–2.19]), CPS ≥ 1 (12.6 vs. 15.8 months; HR, 1.19 [95% CI 0.55–2.58]), and total Japanese subgroups (12.6 vs. 13.3 months; unadjusted HR, 1.10 [95% CI 0.55–2.22]). Median PFS was similar for pembrolizumab and pembrolizumab–chemotherapy versus EXTREME in all subgroups. Grades 3–5 treatment-related adverse events occurred in 5 (22%), 19 (76%), and 17 (89%) patients receiving pembrolizumab, pembrolizumab–chemotherapy, and EXTREME, respectively. One patient receiving pembrolizumab–chemotherapy died because of treatment-related pneumonitis. Conclusion These results support the use of first-line pembrolizumab and pembrolizumab–chemotherapy for Japanese patients with R/M HNSCC. Clinical trial registry ClinicalTrials.gov, NCT02358031. Supplementary Information The online version contains supplementary material available at 10.1007/s10147-022-02233-6.


Introduction
Head and neck squamous cell carcinomas (HNSCC) are anatomically heterogenous, often aggressive malignancies commonly associated with tobacco use, alcohol consumption, and human papillomavirus (HPV) infection [1,2]. The incidence of head and neck cancers varies significantly by country and are particularly common in Japan [3,4]. The incidence of lip, oral cavity, and pharynx cancers in Japanese men in 2014 was 21.6 per 100,000 population compared with a world incidence of 9.4 per 100,000, and the incidence of larynx cancer was 7.8 per 100,000 compared with 2.9 per 100,000, respectively [4]. Although the incidence is lower in women, it is still significantly more common in Japanese women than in the global population (lip, oral cavity, pharynx cancer: 8.4 per 100,000 vs. 3.2 per 100,000; larynx cancer: 0.5 per 100,000 vs. 0.2 per 100,000).
Despite improvements in management and diagnostics, more than 65% of patients with HNSCC develop recurrent or metastatic (R/M) disease, which has a poor prognosis [2]. Until recently, the standard of care for R/M disease in the USA and Japan was cetuximab with a platinum-based agent and 5-fluorouracil-the EXTREME regimen [5][6][7][8]. However, lately, immune checkpoint inhibitors as first-and second-line treatments have demonstrated significant survival benefits in HNSCC [9][10][11].
KEYNOTE-048 was a phase 3 study evaluating pembrolizumab alone and in combination with chemotherapy in previously untreated R/M HNSCC [11]. In KEYNOTE-048, pembrolizumab monotherapy significantly prolonged overall survival (OS) in patients with a PD-L1 combined positive score (CPS) of ≥ 20 and CPS ≥ 1 and had noninferior OS in the total population compared with EXTREME. Safety of pembrolizumab was favorable compared with EXTREME. Pembrolizumab plus platinum and 5-fluorouracil (pembrolizumab-chemotherapy) significantly prolonged OS in patients with PD-L1 CPS ≥ 20, CPS ≥ 1, and in the total population compared with EXTREME. Safety of pembrolizumab-chemotherapy was comparable with that of EXTREME. Based on these results, pembrolizumab is approved in Japan as a first-line treatment option as monotherapy and in combination with platinum and 5-fluorouracil for all patients with R/M HNSCC, regardless of PD-L1 CPS [12].
Given the high incidence of HNSCC in Japan [4], it is important to investigate the efficacy and safety of pembrolizumab in patients of Japanese ethnicity. Here, we report the final analysis of the Japanese subgroup of KEYNOTE-048.
The study protocol and all amendments were conducted in accordance with Good Clinical Practice Guidelines and approved by the appropriate ethics committee at each center. All patients provided written informed consent.
Imaging was performed at baseline, week 9, and then Q6W until year 1, and Q9W thereafter. Response assessments used RECIST v1.1, with confirmation by blinded independent central review. Safety was assessed throughout the study and for 30 days after completion (90 days for serious adverse events [AEs]). AEs were graded per the National Cancer Institute Common Terminology Criteria for Adverse Events, version 4.0.

Outcomes
Primary end points were OS and progression-free survival (PFS). Secondary end points included safety and tolerability, PFS rates at 6 and 12 months, and objective response rate (ORR). Duration of response (DOR) was an exploratory endpoint. Efficacy was evaluated in patients with PD-L1 CPS ≥ 20, with PD-L1 CPS ≥ 1, and in the total Japanese subgroup.

Statistical analyses
Efficacy was assessed in the intention-to-treat (ITT) population of all patients randomly allocated to treatment. Safety was assessed in all patients who received ≥ 1 dose of study treatment. OS, PFS, and DOR were estimated using the Kaplan-Meier method. Additional methods on the Cox proportional hazards model are provided in the Supplement.

Discussion
These results from the Japanese subgroup analysis of previously untreated R/M HNSCC in KEYNOTE-048 were generally consistent with those of the global population, particularly with respect to pembrolizumab monotherapy [11]. In the global KEYNOTE-048 population, pembrolizumab monotherapy significantly prolonged OS in the PD-L1 CPS ≥ 20 and CPS ≥ 1 populations and had noninferior OS in the total population versus EXTREME, and pembrolizumab-chemotherapy significantly prolonged OS in all populations. No improvement in PFS or ORR was observed in either pembrolizumab arm. In the current analysis, pembrolizumab monotherapy showed an OS benefit versus EXTREME in the PD-L1 CPS ≥ 20 subgroup, and similar OS in the total Japanese and PD-L1 CPS ≥ 1 subgroups. PFS was similar between the pembrolizumab monotherapy and EXTREME arms. OS and PFS were similar between pembrolizumab-chemotherapy and EXTREME in all subgroups. Responses were durable with both pembrolizumab and pembrolizumab-chemotherapy, as expected, given the known ability of anti-PD-1 therapies to produce durable responses [13].
Because this post hoc analysis was based on a small subgroup of patients, it was not powered to show differences in efficacy for pembrolizumab monotherapy or pembrolizumab-chemotherapy versus EXTREME. The current analysis is limited by the low number of patients enrolled in Japan, and consequently the small sizes of the PD-L1 CPS subgroups, which resulted in wide 95% confidence intervals for survival estimates. In addition, differences in baseline characteristics between treatment arms may have influenced the results. More patients in the pembrolizumab Fig. 3 Kaplan-Meier estimates of progression-free survival in the Japanese subgroup. Pembrolizumab monotherapy versus EXTREME in the a PD-L1 CPS ≥ 20, b PD-L1 CPS ≥ 1, and c total Japanese subgroups. a From product-limit (Kaplan-Meier) method for censored data. b Based on Cox regression model with Efron's method of tie handling, with treatment as a covariate. CI, confidence interval; CPS, combined positive score; EXTREME, cetuximab plus platinum and 5-fluorouracil; HR, hazard ratio; NR, not reached; PD-L1, programmed death ligand 1; PFS, progression-free survival 6-months monotherapy arm versus the EXTREME arm had an ECOG PS of 0 (74% vs. 42%), whereas the proportion was similar in the pembrolizumab-chemotherapy and EXTREME arms (48% vs. 44%). This may have contributed to the improved survival seen in the pembrolizumab monotherapy arm in the Japanese subgroup. In contrast, ECOG PS was balanced between treatment arms in the overall study population as it was a stratification factor for randomization. The adjusted hazard ratio for OS suggested a treatment benefit for pembrolizumab-chemotherapy versus EXTREME in the total Japanese subgroup, although with a wide confidence interval.
The larger proportion of Japanese patients (61%) who received subsequent anticancer therapy versus the total population (48%) may also have impacted the results. As expected, more patients in the Japanese subgroup who received pembrolizumab monotherapy (57%) and pembrolizumab-chemotherapy (32%) received a subsequent epidermal growth factor receptor inhibitor than those in the EXTREME arm (11%), and more patients in the EXTREME arm (47%) received subsequent anti-PD-1/anti-PD-L1 therapy than those in the pembrolizumab monotherapy (9%) or pembrolizumab-chemotherapy (0%) arms. The higher proportion of patients in the EXTREME arm who received subsequent immunotherapy may have led to OS being higher than expected. Thus, the survival benefit from subsequent immunotherapy in the EXTREME arm of the Japanese subgroup may have confounded the OS results in the current analysis.
Limited data are available regarding the efficacy of targeted therapies and immunotherapies in Japanese patients with R/M HNSCC. A phase 2 study of EXTREME in Japan reported a median OS of 14.1 months, median PFS of 4.1 months, and an ORR per RECIST of 45% (N = 33) [14]. These results are consistent with those of the EXTREME arm in the current report (median OS, 13.6 months; median PFS, 3.9 months; ORR, 37%), noting that the sample size in both studies was relatively small. Similar results were reported from a retrospective study of first-line EXTREME in Japan, showing median OS of 11 months and a median PFS of 5 months [7]. A subanalysis of the Asia-Pacific region, including Japan, of the phase 1b KEYNOTE-012 study of pembrolizumab in R/M HNSCC reported an ORR of 19%, a median OS of 11.6 months, and a median PFS of 2.1 months [15]. Although this ORR was similar to that observed in the pembrolizumab monotherapy arm (ORR, 17%), the Fig. 4 Kaplan-Meier estimates of progression-free survival in the Japanese subgroup. Pembrolizumab-chemotherapy versus EXTREME in the a PD-L1 CPS ≥ 20, b PD-L1 CPS ≥ 1, and c total Japanese subgroups. a From product-limit (Kaplan-Meier) method for censored data. b Based on Cox regression model with Efron's method of tie handling with treatment as a covariate. CI, confidence interval; CPS, combined positive score; EXTREME, cetuximab plus platinum and 5-fluorouracil; HR, hazard ratio; NR, not reached; PD-L1, programmed death ligand 1; PFS, progression-free survival ◂ Fig. 5 Treatment exposure and response duration in the Japanese subgroup receiving pembrolizumab monotherapy, pembrolizumab-chemotherapy, and EXTREME. a Patients in the EXTREME arm who were not included in the comparison analysis of pembrolizumab-chemotherapy versus EXTREME because they were enrolled in the EXTREME arm while enrollment in the pembrolizumab-chemotherapy arm was temporarily halted. CR, complete response; EXTREME, cetuximab plus platinum and 5-fluorouracil; PD, progressive disease; PR, partial response Fig. 6 Change from baseline in target lesion size in the Japanese Subgroup receiving a pembrolizumab monotherapy (n = 23), b pembrolizumab with chemotherapy (n = 24), a and c EXTREME (n = 19). a Target lesion size data with confirmation by blinded independent central review were not available for one patient in the pembrolizumab-chemotherapy arm. EXTREME, cetuximab plus platinum and 5-fluorouracil   [16].
Although there were differences in the patient populations-29.5% of patients had received first-line nivolumab and 16.1% had non-SCC cancers-these results are generally similar to those of the pembrolizumab monotherapy arm in the current analysis. An Asian subanalysis of the CheckMate 141 study of nivolumab has been reported; however, most patients had platinum-refractory HNSCC, and results are therefore not comparable to the current analysis [17]. The safety of pembrolizumab monotherapy and pembrolizumab-chemotherapy was similar in the Japanese subgroup and the global population [11]. Any-grade and grade 3-5 TRAEs were less frequent for pembrolizumab monotherapy compared with pembrolizumab-chemotherapy and EXTREME. The incidence of grade 3-5 TRAEs in the EXTREME arm (89%) was similar to that observed in the phase 2 study of cetuximab combined with cisplatin and 5-fluorouracil in Japanese patients with HNSCC (97%) [14]. The incidence of imAEs was similar in all treatment arms, with hypothyroidism being most common. Overall, pembrolizumab had favorable safety and pembrolizumab-chemotherapy had comparable safety versus EXTREME.
In this analysis, pembrolizumab monotherapy and pembrolizumab plus chemotherapy demonstrated efficacy and manageable safety in Japanese patients with previously untreated HNSCC. These results support the use of pembrolizumab or pembrolizumab-chemotherapy as firstline therapy for Japanese patients with R/M HNSCC.  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/.