Atezolizumab Combined With Cisplatin Plus Vinorelbine as Adjuvant Therapy for Completely Resected NSCLC With EGFR Mutation (West Japan Oncology Group 11719L: ADJUST Study)

Abstract

Introduction

The benefit of adjuvant EGFR tyrosine kinase inhibitors in resected EGFR-mutated NSCLC is limited by recurrence, and the role of immune checkpoint inhibitors remains unclear. We evaluated atezolizumab combined with cisplatin and vinorelbine as adjuvant therapy in completely resected EGFR-mutated NSCLC.

Methods

In a single-arm, multicenter phase II trial, patients with completely resected stage II to IIIA NSCLC harboring EGFR exon 19 deletion or L858R mutations received four cycles of atezolizumab with cisplatin and vinorelbine, followed by up to 12 months of atezolizumab maintenance. The primary end point was the 2-year disease-free survival (DFS) rate. Propensity score–matched patients from the phase III IMPACT trial (cisplatin + vinorelbine) served as an external control. Exploratory biomarkers included tumor mutational burden (TMB) and immune gene expression profiling.

Results

The 2-year DFS rate was 44.4% (lower limit of the one-sided 90% confidence interval: 29.4%). The median DFS was 21.1 months with atezolizumab in ADJUST versus 17.7 months in matched controls from IMPACT. Overall survival was not reached in either group. The median TMB was 5.1 mutations per megabase; 17% were TMB-high (≥10 mutations per megabase). RNA profiling suggested that immune-pathway enrichment may be associated with longer DFS. Immune-related adverse events were manageable, and no fatal events occurred.

Conclusions

Adjuvant atezolizumab with cisplatin and vinorelbine did not significantly improve DFS compared with patients with resected EGFR-mutated NSCLC. These results highlight the limited benefit of immune checkpoint inhibitors in this population and the need for predictive biomarkers to guide immunotherapy strategies for improving cure rates.

Introduction

Recent population-based analyses suggest that approximately half of the newly diagnosed lung cancers present as stage I to III disease without distant metastasis.^1,2 For decades, cisplatin-based adjuvant chemotherapy has been the standard of care for completely resected stage II and III NSCLC. However, its benefit in reducing recurrence remains limited, with only about 5% improvement in the 5-year overall survival (OS) rate.^3,4 Thus, more effective treatment strategies for resectable disease are needed.

Given the marked efficacy of EGFR–tyrosine kinase inhibitors (TKIs) in advanced EGFR-mutated NSCLC, their incorporation into adjuvant therapy for resectable disease has gained increasing attention. The ADAURA trial found that adjuvant osimertinib significantly prolonged disease-free survival (DFS) and OS,⁵ leading to approval by the U.S. Food and Drug Administration. However, consistent with previous trials of other EGFR-TKIs,^6,7 recurrence risk rebounds after completing oral therapy, raising concerns about the curative potential of EGFR-TKIs.^8,9

Immune checkpoint inhibitors (ICIs) have exhibited efficacy and a favorable safety profile in NSCLC. However, their benefit in advanced EGFR-mutated NSCLC has been limited,¹⁰ largely because of low tumor mutational burden (TMB) and an immunosuppressive tumor microenvironment.^11,12 Despite these limitations, subgroup analyses from the IMpower010 trial, phase III, open-label, randomized study to investigate the efficacy and safety of atezolizumab compared with best supportive care after adjuvant chemotherapy in patients with stage IB-IIIA NSCLC following complete resection, reported a DFS benefit with atezolizumab after adjuvant chemotherapy, even in completely resected EGFR-mutated NSCLC.¹³ This aligns with evidence that early-stage NSCLC exhibits a more immunologically active tumor microenvironment than advanced disease, which may enhance ICI efficacy in the adjuvant setting.¹⁴ This biological distinction provides a rationale for exploring the role of ICIs in resected EGFR-mutated NSCLC. This study, therefore, aimed to report the results of an open-label, multi-institutional, single-arm phase II trial assessing the feasibility and efficacy of atezolizumab combined with cisplatin and vinorelbine as adjuvant therapy in completely resected EGFR-mutated NSCLC.

Materials and Methods

Study Design and Patients

The ADJUST study was an open-label, multicenter, single-arm phase II trial. As an external control for cisplatin plus vinorelbine (CDDP+VNR), an integrated analysis was conducted using data from the previously reported phase III IMPACT trial, a randomized phase III trial of adjuvant gefitinib versus cisplatin and vinorelbine in completely resected NSCLC patients with mutated EGFR¹⁵ (Fig. 1).

Figure 1.

Study schema. CDDP, cisplatin; ECOG PS, Eastern Cooperative Oncology Group performance status; p, pathologic; VNR, vinorelbine.

Patients were eligible for ADJUST trial if they had undergone complete resection within 21 to 56 days before registration and had pathologically confirmed stage II to IIIA NSCLC (Union for International Cancer Control, UICC, version 8) harboring an EGFR exon 19 deletion or exon 21 L858R mutation; were aged 20 years or older; had an Eastern Cooperative Oncology Group performance status of 0 to 1; and had adequate organ function. Patients with autoimmune disease or previous corticosteroid or immunosuppressive use within 14 days before registration were excluded.

Treatment consisted of atezolizumab (1200 mg/body on day 1), vinorelbine (25 mg/m² on days 1 and 8), and cisplatin (80 mg/m² on day 1) every 3 weeks for four cycles, followed by atezolizumab (1200 mg on day 1) every 3 weeks. Atezolizumab was continued for up to 12 months or until relapse or unacceptable toxicity.

Baseline imaging included chest computed tomography (CT) and brain CT or magnetic resonance imaging. For recurrence assessment, chest-abdominal CT was performed every 6 months for the first 24 months and subsequently at 6-month intervals in principle. To evaluate the primary end point (2-year DFS rate), chest-abdominal CT and brain CT or magnetic resonance imaging were performed at 2 years postregistration. Adverse events (AEs) were graded according to the Common Terminology Criteria for Adverse Events, version 5.0.

Integrated Analysis

For the integrated analysis, the CDDP+VNR arm of the IMPACT trial was used as an external control. Details of the IMPACT trial have been published.¹⁵ In brief, gefitinib was compared with CDDP+VNR in patients with completely resected EGFR-mutated NSCLC in this trial, and its eligibility criteria closely matched those of ADJUST.

Propensity score matching was performed to minimize baseline imbalances between ADJUST and IMPACT when estimating the effect of adding ICI. The propensity score model was estimated using a logistic regression model adjusted for the patient characteristics, including sex, smoking status, pathologic stage, and EGFR mutation subtype.16, 17, 18, 19 Matching was performed without replacement on the logit of the propensity score using calipers of width equal to 0.2 of the SD of the logit of the propensity score.

End points and Sample Size Calculation

The primary end point was the 2-year DFS rate. Secondary end points were DFS, OS, and safety.

On the basis of pivotal phase III studies of atezolizumab combined with platinum-doublet chemotherapy in advanced NSCLC, the 2-year DFS rate was set at a threshold of 55% and the expected rate of 80%. Under this assumption, the required sample size to test the difference in population proportions was 16 (α = 0.10, one-sided; β = 0.20). Allowing for a 10% ineligibility rate, the final sample size for ADJUST was estimated at 18.

Statistical Analysis

DFS and OS were estimated using the Kaplan-Meier method, and median values were reported with 95% confidence intervals (CIs). DFS and OS were compared using the log-rank test, and hazard ratios (HRs) were calculated with Cox proportional hazards regression. Data analyses were performed using the Statistical Analysis System version 9.4 (SAS Institute Inc., Cary, NC) and Python version 3.10.1 (Python Software Foundation).

Biomarker Analyses

For exploratory genomic analyses, tumor tissue samples were collected at baseline. DNA and RNA were extracted from formalin-fixed paraffin-embedded sections using the AllPrep DNA/RNA formalin-fixed paraffin-embedded kit (cat. 80234, QIAGEN, Venlo, Netherlands) and quantified with Qubit assays (Thermo Fisher Scientific, Waltham, MA). Sequencing was performed on the Ion GeneStudio S5 System (Thermo Fisher Scientific, Waltham, MA). Variants with a variant allele frequency less than 5% or a minor allele frequency greater than or equal to 1% in population databases (Tohoku Medical Megabank Organization, Genome Aggregation Database, and Genome Medicine Database Japan) were excluded. ClinVar and OncoKB annotations were applied, and clinically relevant variants were retained. Gene expression was assessed using the NanoString nCounter PanCancer IO 360 Panel (Nanostring, Bruker Spatial Biology, Bothell, WA). Data were processed with nSolver and Advanced Analysis software (Bruker Spatial Biology), including normalization, differential expression with multiple-testing correction, and immune-related signature and cell type profiling. Full biomarker procedures are provided in the Supplementary Materials.

Ethical Considerations

The main study was approved by the Wakayama Medical University Institutional Review Board (approval date: July 2, 2019; approval number: 1-01008A) and registered in Japic Clinical Trials Information, Japan (JapicCTI-194849). Written informed consent was obtained from all participants. The integrated and biomarker analyses were separately approved by the Wakayama Medical University Institutional Review Board (approval dates: December 26, 2023 and June 2, 2023; approval numbers: 4021 and 3858).

Results

Baseline Characteristics

In the ADJUST study, 18 patients were enrolled between August 2019 and August 2021 (Supplementary Fig. 1). All patients were included in the safety and efficacy analyses. At data cutoff, the median follow-up was 34.4 months (range: 12.8–43.6).

The median age was 67 years (range: 48–78). Most patients were female (n = 11, 61%), never-smokers (n = 10, 56%), and had adenocarcinoma (n = 18, 100%). Patients with exon 19 deletion accounted for 56% (n = 10). Nearly all underwent lobectomy (n = 16, 89%). Baseline characteristics are summarized in Table 1.

Table 1.

Baseline Characteristics

Characteristic	CDDP+VNR+Atezo N (%)	CDDP+VNR N (%)
Patients	18	18
Age
Median (range)	67 (48–78)	62 (48–72)
Sex
Male/Female	7 (39)/11 (61)	10 (56)/8 (44)
ECOG PS
0/1	12 (67)/6 (33)	16 (89)/2 (11)
Smoking status
Never/Current or former smoker	10 (56)/8 (44)	10 (56)/8 (44)
EGFR mutation
exon 19 del/exon 21 L858R/both	10 (56)/8 (44)/0 (0)	10 (56)/8 (44)/0 (0)
PD-L1
<1%/1–49%/≥50%/NE	8 (44)/4 (22)/4 (22)/2 (11)
pT factor
T1/T2/T3/T4	7 (39)/9 (50)/2 (11)/0 (0)	4 (22)/10 (55)/3 (17)/1 (6)
pN factor
N0/N1/N2	1 (6)/6 (33)/11 (61)	3 (17)/3 (17)/12 (66)
Disease stage
II/III	6 (33)/12 (67)	5 (28)/13 (72)
Surgery
Lobectomy/Other	16 (89)/2 (11)	18 (100)/0 (0)

Atezo, atezolizumab, CDDP, cisplatin; ECOG PS, Eastern Cooperative Oncology Group performance status; N, number; NE, not evaluable; p, pathological; PD-L1, programmed death ligand 1; VNR, vinorelbine.

For the external control, 18 patients treated with CDDP+VNR in the IMPACT trial were matched to patients in the ADJUST study (Table 1).

Efficacy Outcomes

The primary end point, the 2-year DFS rate, was 44.4% (lower limit of one-sided 90% CI: 29.4%) in the atezolizumab plus cisplatin and vinorelbine (Atezo+CDDP+VNR) group, compared with 47.1% in the matched CDDP+VNR group. The median DFS was 21.1 months (95% CI: 12.5–not reached) in the Atezo+CDDP+VNR group and 17.7 months (95% CI: 11.0–not reached) in the matched CDDP+VNR group (HR, 0.97; 95% CI: 0.43–2.21; p = 0.95) (Fig. 2A). Subgroup analyses of DFS revealed no significant differences (Fig. 2B).

Figure 2.

Survival and subgroup analyses in the integrated analysis. (A) Kaplan-Meier curves of disease-free survival. (B) Forest plot including ECOG PS, sex, smoking status, EGFR mutation type, and age. Atezo, atezolizumab; CDDP, cisplatin; CI, confidence interval; ECOG PS, Eastern Cooperative Oncology Group performance status; HR, hazard ratio; VNR, vinorelbine.

The median OS was not reached in either group. The 2-year OS rates were 83.3% in the Atezo+CDDP+VNR group and 94.1% in the CDDP+VNR group (HR, 0.47; 95% CI: 0.09–2.56; p = 0.38) (Supplementary Fig. 2).

In this study, programmed death ligand 1 (PD-L1) expression was assessed at each institution using the Dako 22C3 pharmDx kit (Dako, Carpinteria, CA). Among 18 patients, the PD-L1 tumor proportion score was less than 1% in 8 (44%), 1% to 49% in 4 (22%), greater than or equal to 50% in 4 (22%), and not evaluable in 2 (11%).

Although the subgroup analysis was performed as an exploratory analysis on the basis of a small number of patients, the median DFS in subgroups defined by PD-L1 tumor proportion score was 17.8 months (95% CI: 6.0–not reached), 23.7 months (95% CI: 12.5–not reached), and 18.5 months (95% CI: 14.8–not reached) in the less than 1%, 1% to 49%, and greater than or equal to 50% groups, respectively.

At analysis, recurrence was observed in 12 of 18 patients (67%). Local recurrence occurred in five patients, and central nervous system recurrence occurred in 3 (25%). There were 12 patients who received subsequent therapy; five (42%) underwent local treatment such as radiation or surgery, and seven (58%) were treated with osimertinib. No patients with recurrence received immunotherapy as subsequent treatment.

Safety and Tolerability

AEs related to study treatment are summarized in Table 2. There were 19 occurrences of treatment-related grade 3 or 4 hematologic AEs, including neutropenia (72%), decreased white blood cell count (22%), anemia (5.6%), and decreased platelet count (5.6%). The following 14 treatment-related grade 3 or 4 nonhematologic AEs were observed: anorexia (17%), nausea (11%), constipation (5.6%), colitis (5.6%), adrenocorticotropic hormone deficiency (5.6%), type 1 diabetes (5.6%), febrile neutropenia (5.6%), hyperkalemia (5.6%), increased alanine aminotransferase (5.6%), vertigo (5.6%), and malaise (5.6%). Toxicity led to treatment discontinuation in seven patients. No treatment-related deaths were reported.

Table 2.

Treatment-Related Adverse Events (Any Grade ≥20% or Any Severe Case)

N (%)	Any Gr	Gr1	Gr2	Gr3	Gr4	Gr5
Any adverse events	18 (100)	15 (83)	16 (89)	13 (72)	7 (39)	0 (0)
WBC decreased	7 (39)	2 (11)	1 (6)	3 (17)	1 (6)	0 (0)
Neutropenia	14 (78)	1 (6)	0 (0)	7 (39)	6 (33)	0 (0)
Anemia	8 (44)	4 (22)	3 (17)	1 (6)	0 (0)	0 (0)
PLT count decreased	3 (17)	2 (11)	0 (0)	1 (6)	0 (0)	0 (0)
Nausea	12 (67)	2 (11)	8 (44)	2 (11)	0 (0)	0 (0)
Vomiting	4 (22)	0 (0)	4 (22)	0 (0)	0 (0)	0 (0)
Anorexia	11 (61)	6 (33)	2 (11)	3 (17)	0 (0)	0 (0)
Constipation	8 (44)	6 (33)	1 (6)	1 (6)	0 (0)	0 (0)
Diarrhea	4 (22)	3 (17)	1 (6)	0 (0)	0 (0)	0 (0)
Colitis	1 (6)	0 (0)	0 (0)	1 (6)	0 (0)	0 (0)
Hiccups	4 (22)	2 (11)	2 (11)	0 (0)	0 (0)	0 (0)
Mucositis oral	4 (22)	3 (17)	1 (6)	0 (0)	0 (0)	0 (0)
Alopecia	8 (44)	7 (39)	1 (6)	0 (0)	0 (0)	0 (0)
Hyperthyroidism	4 (22)	4 (22)	0 (0)	0 (0)	0 (0)	0 (0)
ACTH deficiency	1 (6)	0 (0)	0 (0)	1 (6)	0 (0)	0 (0)
Type 1 diabetes	1 (6)	0 (0)	0 (0)	1 (6)	0 (0)	0 (0)
Febrile neutropenia	1 (6)	0 (0)	0 (0)	1 (6)	0 (0)	0 (0)
Hyponatremia	5 (28)	3 (17)	2 (11)	0 (0)	0 (0)	0 (0)
Hyperkalemia	1 (6)	0 (0)	0 (0)	1 (6)	0 (0)	0 (0)
ALT increased	1 (6)	0 (0)	0 (0)	1 (6)	0 (0)	0 (0)
Creatinine increased	6 (33)	5 (28)	1 (6)	0 (0)	0 (0)	0 (0)
Vertigo	1 (6)	0 (0)	0 (0)	1 (6)	0 (0)	0 (0)
Dysgeusia	5 (28)	5 (28)	0 (0)	0 (0)	0 (0)	0 (0)
Malaise	3 (17)	1 (6)	1 (6)	1 (6)	0 (0)	0 (0)

ACTH, adrenocorticotropic hormone; ALT, alanine aminotransferase; Gr, grade; N, number; PLT, platelet; WBC, white blood cell.

Gene Alterations and Clinical Correlates

Figure 3 summarizes detected gene alterations, TMB, and clinical efficacy. Beyond EGFR, TP53 was the most frequent co-mutation (n = 7; 39%), followed by PIK3CA (n = 3; 17%), ARID2 (n = 2; 11%), SMAD4 (n = 2; 11%), and PTEN (n = 2; 11%). The median TMB was 5.1 mutations/megabase (mut/Mb) (range: 1.7–35.5), with three patients (17%) classified as TMB-high (≥10 mut/Mb). According to an exploratory analysis on the basis of a small number of patients, the median DFS was 35.1 months (95% CI: 14.8–35.1) in the TMB-high group and 18.5 months (range: 7.8–not reached) in the non–TMB-high group. RNA expression analysis was available for 15 patients. Among 25 predefined gene sets (770 immune-related genes), elevated expression in three pathways, lymphoid compartment (Fig. 4A), JAK-STAT signaling (Fig. 4B), and hypoxia (Fig. 4C) were associated with improved DFS, although the differences were not statistically significant.

Figure 3.

Oncoplot of gene alterations and tumor mutation burden. DFS, disease-free survival; TMB, tumor mutational burden.

Figure 4.

RNA expression analysis. (A) Lymphoid compartment. (B) JAK-STAT signaling. (C) Hypoxia.

Discussion

In phase III, double-blind, placebo-controlled, multi-center international study of neoadjuvant/adjuvant durvalumab for the treatment of patients with resectable stages II and III NSCLC (AEGEAN trial), which evaluated perioperative durvalumab added to neoadjuvant platinum-doublet chemotherapy, EGFR-mutated cases represented approximately 6% (n = 51) of enrolled patients but were excluded from the primary end point analysis.²⁰ In an exploratory subgroup analysis of the EGFR-mutated population, median event-free survival was 30.8 months in the durvalumab arm (n = 26) versus 19.6 months in the placebo arm (n = 25). However, the Kaplan-Meier curves crossed, and the HR was 0.86 (95% CI: 0.35–2.19).²¹ Similarly, in the phase III, randomized, double-blind trial of platinum doublet chemotherapy and pembrolizumab as neoadjuvant/adjuvant therapy for participants with resectable stage II, IIIA, and Resectable IIIB NSCLC (KEYNOTE-671 study), which investigated perioperative pembrolizumab with neoadjuvant platinum-doublet chemotherapy, EGFR-mutated patients accounted for only 3.5% (n = 14), and no clinical data have been reported for this subgroup.²² These findings highlight a clear evidence gap regarding the benefit of ICIs in early-stage EGFR-mutated NSCLC. In this context, our integrated analysis prospectively assessed the feasibility and antitumor activity of adding ICI postresection, providing the first dedicated evidence to inform clinical strategies for early-stage EGFR-mutated NSCLC.

A distinctive feature of ADJUST is its use of an integrated analysis with an external control cohort rather than a traditional randomized control arm. When phase III randomized trials are impractical because of small patient numbers or ethical constraints, well-matched historic or real-world controls can help estimate treatment effects.^23,24 This strategy has the following precedent: the U.S. Food and Drug Administration–approved blinatumomab for relapsed acute lymphoblastic leukemia on the basis of a single-arm phase II trial supported by a propensity-matched external control.²⁵ Our integrated analysis similarly approximates the rigor of a randomized comparison by matching key prognostic variables, including sex, smoking status, pathologic stage, and EGFR mutation type. Although this method cannot replace a phase III randomized trial, it enhances the credibility of our findings, supports planning for future phase III studies, and helps accelerate evidence generation. Overall, this study provides a pragmatic and scientifically sound framework for evaluating therapeutic efficacy under the real-world constraints of clinical research, including financial, logistical, and patient-related limitations.

Our study provides new insights into the evolving paradigm of adjuvant targeted therapy in EGFR-mutated NSCLC. The ADAURA trial advanced the field by suggesting that osimertinib, a third-generation EGFR-TKI, significantly reduced recurrence risk compared with placebo, ultimately exhibiting an OS benefit and reshaping the standard of care for early-stage EGFR-mutated NSCLC.⁵ However, even in pivotal trials such as ADAURA, DFS curves for adjuvant EGFR-TKIs decline rapidly after treatment completion, with long-term outcomes approaching those of the control arm.^6,15 This underscores the limitation of EGFR-TKIs in achieving a cure, a critical goal of adjuvant therapy. By contrast, perioperative immunotherapy, now standard for resectable NSCLC without driver mutations, has exhibited durable DFS benefits and sustained responses beyond therapy completion, suggesting its potential to induce lasting antitumor immunity and eradicate minimal residual disease.^13,20,22,26 These observations highlight the importance of exploring immunotherapeutic strategies in EGFR-mutated NSCLC to improve cure rates in early-stage disease.

Regarding safety, Atezo+CDDP+VNR exhibited a profile consistent with previously reported platinum-doublet plus ICI regimens.^20,22,27 Immune-related AEs, including type 1 diabetes (5.6%) and endocrine disorders (5.6%), were observed but were manageable, and no fatal events occurred. Grade 3 to 4 neutropenia was frequent (72%) but comparable to historical rates for CDDP+VNR regimens (approximately 80%).^4,28 Thus, this regimen does not seem to introduce new safety concerns and can be considered feasible for clinical implementation.

From a translational perspective, our findings underscore the challenges of applying immunotherapy in EGFR-mutated NSCLC. This is the first prospective study confirming that EGFR-mutant tumors generally have low TMB, with a median of 5.1 mut/Mb and only 17% of patients classified as TMB-high (≥10 mut/Mb), consistent with previous reports of their low immunogenicity.²⁹ Notably, consistent with previous prospective data suggesting that a subset of patients with EGFR mutations can benefit from ICIs,³⁰ a small subgroup in our study with more immune-active profiles (e.g., TMB-high or enriched inflammatory gene signatures) experienced prolonged DFS. This finding may suggest the existence of an ICI-responsive subgroup even within this traditionally ICI-resistant population. However, unlike EGFR wild-type NSCLC, in which perioperative immunotherapy has produced durable remissions and potential cures,^20,22,26 long-term benefits were not observed in our study. Although our molecular and subgroup findings are exploratory and underpowered, they highlight the need to determine whether specific subsets of EGFR-mutated patients might derive durable benefit from immunotherapy, potentially guided by predictive biomarkers or tailored treatment strategies.

This study has several limitations. As a phase II trial, statistical power and generalizability are restricted. In particular, although molecular profiling was performed, biomarker and subgroup analyses were underpowered and hypothesis-generating, thus limiting the ability to draw definitive conclusions. However, the integrative analysis (n = 36) was comparable in size to the EGFR-mutated subgroup in AEGEAN (n = 51), supporting the hypothesis that EGFR-mutated tumors are less susceptible to ICI plus chemotherapy, even in early-stage disease. Moreover, reliance on a propensity score-matched historical control cannot eliminate unmeasured confounding despite methodologic rigor. Finally, because adjuvant osimertinib has become the current standard therapy for this population, the contemporary clinical relevance of the specific regimen tested in ADJUST is necessarily limited. Nonetheless, our study applied robust methodology with careful matching, providing valuable insights into both the feasibility and the limitations of adjuvant chemoimmunotherapy for EGFR-mutated NSCLC.

To conclude, in patients with completely resected EGFR-mutated NSCLC, adjuvant Atezo+CDDP+VNR did not provide a clear clinical benefit. Nonetheless, exploring immunotherapeutic strategies in this setting may help improve cure rates in early-stage disease and underscores the need for further studies to identify predictive biomarkers to guide precision medicine.

Data Availability Statement

The corresponding author affirms full access to all study data and responsibility for the integrity and accuracy of the analysis. Data supporting the findings are available from the corresponding author on reasonable request. Data are not publicly available owing to privacy and ethical restrictions.

CRediT Authorship Contribution Statement

Ryota Shibaki: Conceptualization, Methodology, Enrollment, Writing - original draft, Writing - review & editing.

Hiroaki Akamatsu: Conceptualization, Methodology, Enrollment, Writing - original draft, Writing - review & editing.

Morihito Okada: Enrollment, Writing - review & editing.

Kenta Murotani: Statistics, Writing - review & editing.

Mitsuo Osuga: Genetic analysis, Writing - review & editing.

Kazumi Nishino: Enrollment, Writing - review & editing.

Shuji Murakami: Enrollment, Writing - review & editing.

Kazushige Wakuda: Enrollment, Writing - review & editing.

Tetsuya Mitsudomi: Enrollment, Writing - review & editing.

Hirohito Tada: Enrollment, Writing - review & editing.

Kenichi Yoshimura: Statistics, Writing - review & editing.

Yasuhiro Koh: Genetic analysis, Writing - review & editing.

Nobuyuki Yamamoto: Supervise, Writing - review & editing.

Declaration of Generative AI and AI-assisted Technologies in the Writing Process

During the preparation of this work, the authors used ChatGPT to refine the word count. After using this tool/service, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

Disclosure

Dr. Shibaki received personal fees from AstraZeneca K.K., Chugai Pharmaceutical Co. Ltd., MSD K.K., Taiho Pharmaceutical Co. Ltd., Ono Pharmaceutical Co. Ltd., Daiichi Sankyo Co. Ltd., and Boehringer Ingelheim Japan Inc. Dr. Akamatsu received honoraria from Amgen Inc., AstraZeneca K.K., Boehringer Ingelheim Japan Inc., Bristol Myers Squibb, Chugai Pharmaceutical Co. Ltd., Daiichi Sankyo Co. Ltd., Eli Lilly Japan K.K., Janssen Pharmaceutical K.K., Merck Sharp & Dohme K.K., Nippon Kayaku Co. Ltd., Novartis Pharma K.K., Ono Pharmaceutical Co. Ltd., Pfizer Inc., Takeda Pharmaceutical Co. Ltd., and Taiho Pharmaceutical Co. Ltd.; served in an advisory role for Amgen Inc., Janssen Pharmaceutical K.K., and Sandoz; and received research funding from Amgen Inc., Chugai Pharmaceutical Co. Ltd., and Merck Sharp & Dohme K.K. Mr. Murotani has received personal fees from AstraZeneca K.K., Chugai Pharmaceutical Co. Ltd., MSD K.K., Taiho Pharmaceutical Co. Ltd., Nippon Shinyaku, and Pfizer Inc. Dr. Nishino reports receiving grants or contracts to their institution from AbbVie, Amgen, AstraZeneca, Bayer Yakuhin, Ltd., Daiichi Sankyo Company, Limited, Eisai Co., Ltd., Eli Lilly Japan K.K., Janssen Pharmaceutical K.K., Merck Biopharma Co., Ltd., Merus N.V., MSD, Novartis, Ono Pharmaceutical Co., Ltd., Pfizer, Sanofi K.K., Taiho Pharmaceutical Co., Ltd., Delta-Fly Pharma, Inc., IQVIA Holdings Inc., Nippon Boehringer Ingelheim Co., Ltd., Parexel International Inc. and Takeda Pharmaceutical Company Limited; and payments or honoraria from Amgen, Asahi Kasei Pharma Corporation, AstraZeneca, Daiichi Sankyo Company, Limited, Limited, Eli Lilly Japan K.K., Janssen Pharmaceutical K.K., KYORIN Pharmaceutical Co., Ltd., Kyowa Kirin Co., Ltd., Merck Biopharma Co., Ltd., Mochida Pharmaceutical Co., Ltd., MSD, Nippon Boehringer Ingelheim Co., Ltd., Nippon Kayaku Co.,Ltd., Novartis Pharmaceuticals, Ono Pharmaceutical Co., Ltd., Pfizer, Taiho Pharmaceutical, Takeda Pharmaceutical Company Limited, Varian Medical Systems, Inc., and Yuyu Teijin Medicare Inc. Dr. Murakami reports receiving grants or contracts to their institution from Chugai Pharmaceutical Co., Ltd., Daiichi Sankyo, Janssen Pharmaceutical K.K., Ono Pharmaceutical Co.,Ltd., Merck Sharp & Dohme, Takeda Pharmaceutical Co. Ltd., and Sanofi; and received honoraria from AstraZeneca K.K., Chugai Pharmaceutical Co., Ltd., Eli Lilly Japan K.K., Merck Sharp & Dohme, Novartis Japan., Pfizer Japan., Taiho Pharmaceutical, Co., Ltd., and Takeda Pharmaceutical Co. Ltd. Dr. Wakuda reports receiving grants or contracts to their institution from AbbVie, Amgen, AstraZeneca, Chugai Pharmaceutical Co., Ltd., Daiichi Sankyo, MSD, Novartis, and Dizal Pharma; and payments or honoraria from Amgen, AstraZeneca, Boehringer Ingelheim, Chugai Pharmaceutical Co.,Ltd., Daiichi Sanky, Eli Lilly Japan K.K., Janssen Pharmaceutical K.K., MSD, Nihon Kayaku, Ono Pharmaceutical Co., Ltd., Taiho Pharmaceutical, Takeda Pharmaceutical Company Limited, Yuyu Teijin Medicare Inc. Dr. Mitsudomi received honoraria from Amgen Inc., AstraZeneca K.K., Boehringer Ingelheim Japan Inc., Chugai Pharmaceutical Co., Ltd., Taiho, Daiichi Sankyo Co. Ltd., Guardant, MSD, Novartis Pharma K.K., Nihon Kayaku, Novocure, Merck Biopharma, Janssen Pharmaceutical K.K., Ono Pharmaceutical Co. Ltd., Pfizer Inc., Takeda Pharmaceutical Co. Ltd., and Bayer.; and participated in the advisory boards of AstraZeneca K.K., Novartis Pharma K.K., Regeneron, BMS, Taiho, and Daiichi Sankyo Co. Ltd. Dr. Koh reports receiving grants or contracts from Chugai Pharmaceutical Co., Ltd.; and received honoraria from Chugai Pharmaceutical Co., Ltd. Dr. Yamamoto reports receiving grants or contracts to their institution from A2 Healthcare Corporation, AbbVie Inc., Amgen K.K., AztraZeneca K.K., Bristol Myers Squibb, Chugai Pharmaceutical Co., Ltd., Eli Lilly Japan K.K., EPS Co., Ltd., GlaxoSmithKline K.K., Kyowa Kirin Co., Ltd., Janssen Pharmaceutical K.K., Novartis Japan., Ono Pharmaceutical Co., Ltd., Medpace Japan K.K., Medical Market Vision Co., Ltd., Merck Sharp & Dohme, Pfizer R&D Japan, Taiho Pharmaceutical, Co., Ltd., IQVIA Solutions Japan G.K., Syneos Health ClinicaL K.K., Nippon Boehringer Ingelheim Co., Ltd., Nippon Chemical Industrial Co., Ltd., PRiME-R, Inc., and RPM Co.; and received honoraria from Accuray Japan K.K., Amgen K.K., AstraZeneca K.K., AbbVie Inc., Nippon Boehringer Ingelheim Co., Ltd., Chugai Pharmaceutical Co., Ltd., Daiichi Sankyo, Inc., Eli Lilly Japan K.K., Guardant Health Japan Corp., GlaxoSmithKline K.K., Kyowa Kirin Co., Ltd., Kyorin Pharmaceutical Co., Ltd., Merck Sharp & Dohme, Miyarisan Pharmaceutical Co., Ltd., Medpeer, Inc., Merck Biopharma Co., Ltd., Nippon Chemical Industrial Co., Ltd., Novartis Japan., Novocure Ltd., Otsuka Pharmaceutical Co., Ltd., Ono Pharmaceutical Co., Ltd., Pfizer Japan., Janssen Pharmaceutical K.K., Taiho Pharmaceutical, Co., Ltd., Terumo.co.jp., Tsumura & Co., Takeda Pharmaceutical Co. Ltd., USACO Corporation, and Taiho Pharmaceutical Co. Ltd.