Primary Analysis of a Phase II Study of Atezolizumab plus Bevacizumab for TACE-Unsuitable Patients with Tumor Burden beyond Up-To-Seven Criteria in Intermediate-Stage Hepatocellular Carcinoma: REPLACEMENT Study.

Introduction
Transcatheter arterial chemoembolization (TACE) is the standard of care for unresectable intermediate-stage hepatocellular carcinoma (HCC) according to global guidelines [1–5]. However, TACE is ineffective for certain patients due to tumor heterogeneity [6, 7]. The Japan Society of Hepatology defined TACE failure/refractoriness in 2011 [8] and 2014 [9]. Subsequently, the OPTIMIS trial demonstrated improved overall survival (OS) with early sorafenib versus continued TACE after TACE failure, although 20–25% of patients experienced hepatic function deterioration [10–13].
These findings led to the concept of TACE unsuitability, which takes into account the likelihood of refractoriness to TACE, the potential for deterioration in hepatic functional reserve, and the low probability of TACE benefit [6]. In support of the concept of TACE unsuitability, a recent study found that lenvatinib as an initial treatment improved progression-free survival (PFS) and OS compared to TACE in Child-Pugh A patients with HCC beyond up-to-seven criteria, while preserving hepatic functional reserve [14]. Current guidelines recommend systemic therapy over non-curative TACE to improve OS in patients with TACE-unsuitable HCC [15, 16].
The IMbrave150 trial demonstrated that atezolizumab plus bevacizumab improved PFS and OS compared with sorafenib in patients with unresectable HCC, including those with intermediate-stage disease [17, 18]. Based on this, Japanese guidelines currently recommend atezolizumab plus bevacizumab as first-line treatment for unresectable advanced-stage HCC that is not amenable to other locoregional therapies, including TACE [19].
The primary objective of this study was to prospectively evaluate the efficacy and safety of atezolizumab plus bevacizumab in TACE-naïve, unresectable, intermediate-stage HCC patients with tumor burden beyond up-to-seven criteria. In addition, this study aimed to compare, in an exploratory manner, the outcomes of atezolizumab plus bevacizumab to those of TACE using inverse probability weighting (IPW) in this patient population.

Methods

Study Design and Participants
A multicenter, open-label, single-arm phase II study of atezolizumab plus bevacizumab was conducted at 35 sites in Japan (Japan Registry of Clinical Trials, jRCTs071200051). We also conducted an exploratory comparison of the data for atezolizumab plus bevacizumab from this prospective phase II study with retrospectively collected data from patients who received TACE at the same sites participating in the phase II study (retrospective historical TACE data).
In the phase II study of atezolizumab plus bevacizumab, key inclusion criteria were unresectable intermediate-stage HCC beyond up-to-seven criteria (per Clinical Practice Manual Hepatocellular Carcinoma, fourth edition [20]), Child-Pugh A, no previous systemic therapy, no previous TACE treatment (TACE was permitted as part of a combination treatment with ablation with curative intent), and Eastern Cooperative Oncology Group Performance Status (ECOG PS) score of 0 or 1. Key exclusion criteria were current or recent use of full-dose anticoagulants or thrombolytic agents for therapeutic purposes and untreated or incompletely treated varices with bleeding or high risk for bleeding (assessed with esophagogastroduodenoscopy and treated according to local standard of care). In the retrospective historical TACE data, we included patients with unresectable HCC beyond up-to-seven criteria who had received their initial TACE treatment between January 2017 and December 2017 had no vascular invasion or extrahepatic metastasis, Child-Pugh A at the time of initial TACE, no prior systemic therapy for HCC (including sorafenib, lenvatinib, or immunotherapy), and no molecular targeted therapy administered between TACE sessions. Patients without baseline imaging within 28 days prior to TACE and insufficient baseline data were excluded.
This study was conducted in accordance with the Declaration of Helsinki, the Japanese Clinical Trials Act, the Ordinance for Enforcement of the Clinical Trials Act of the Ministry of Health, Labour and Welfare, and related notifications. Patients in the atezolizumab plus bevacizumab cohort were enrolled after giving consent to participate using the approved written informed consent form, while patients in the retrospective historical TACE cohort were collected using an opt-out approach for consent.

Study Treatment
Patients received atezolizumab 1,200 mg and bevacizumab 15 mg/kg every 3 weeks until unacceptable toxicity or disease progression, according to the respective labeling. Patients were allowed to continue treatment after radiographic progression if the investigators deemed the benefit-risk ratio favorable.

Endpoints
The primary endpoint was the 6-month PFS rate, according to the modified RECIST (mRECIST) by investigator assessment. Progression included both diagnostic radiographic progression according to mRECIST and clinical progression, as determined by the investigators, and may include deterioration of hepatic reserve function, worsening of performance status, or significant increase in tumor markers.
Secondary endpoints were the (i) PFS according to RECIST version 1.1 (RECIST v1.1) by investigator assessment; (ii) objective response rate (ORR), defined as the percentage of patients with best overall response of complete response (CR) or partial response (PR), according to mRECIST and RECIST v1.1 by investigator assessment, while ORR including curative intent therapy effect was defined as the percentage of patients with CR and PR, including responses from curative intent therapy performed after atezolizumab plus bevacizumab; (iii) duration of response (DoR) for patients with CR or PR as well as time to response according to mRECIST and RECIST v1.1 by investigator assessment; (iv) OS, defined as the period from the enrollment date to the date of death due to any cause; (v) safety in terms of types and grades of adverse events (AEs) as well as time courses of ALBI/mALBI scores [21] and Child-Pugh scores. Exploratory endpoints were the (i) change in tumor size from baseline in individual patients, including over time, calculated according to best overall response and plotted using waterfall plot and spider plot methods; (ii) comparison of the efficacy and safety data collected in this study with retrospective historical data of patients first treated with TACE from January 2017 to December 2017; (iii) conversion rate to a curative intent therapy after atezolizumab plus bevacizumab therapy. Curative intent therapy was defined as resection, RFA, or TACE with curative intent performed after a reduction in tumor burden from baseline following atezolizumab plus bevacizumab therapy, irrespective of outcome. TACE with curative intent was defined as superselective TACE for all intrahepatic nodules [15].

Statistical Methods
We assumed a threshold 6-month PFS rate of 55% based on previous studies reporting median PFS for TACE of 6.0 months (OPTIMIS trial [13]) and 6.9 months (TACTICS trial [22]). The expected 6-month PFS rate was set at 73% based on the subgroup results of IMbrave150 [23]. With a one-sided significance level of 0.05 and a power of 90%, the number of patients required for a binomial test was estimated to be 60 patients. Therefore, the planned number of enrolled patients was 70, allowing for censoring and exclusions. The treatment was considered effective if the lower limit of the 90% confidence interval (CI) for the 6-month PFS rate exceeded the pre-specified threshold of 55%.
The full analysis set consisted of all enrolled patients (intention-to-treat population) excluding ineligible patients and those who withdrew consent after enrollment or did not receive protocol treatment. The full analysis set was used as the primary analysis population for all efficacy endpoints.
PFS was evaluated by the Kaplan-Meier method, and the 90% CI for the 6-month PFS rate was calculated using Greenwood’s formula [24, 25]. The 95% CI for the 6-month PFS rate was calculated as the reference value. The 95% CI for median PFS was calculated using the Brookmeyer-Crowley method. Other efficacy endpoints, including DoR and OS, were analyzed similarly.
The safety analysis set (safety population) consisted of all enrolled patients who began protocol treatment and from whom at least some data were obtained. The frequency of AEs by severity grade was assessed according to CTCAE v5.0 in addition to calculation of the incidence of abnormal laboratory tests. The time course of ALBI and Child-Pugh scores were assessed by calculation of the means, standard deviation, median, minimum, maximum, and interquartile range in each treatment cycle.
An exploratory analysis was performed to compare the data of atezolizumab plus bevacizumab from this prospective phase II study with retrospective historical TACE data using the IPW method. The IPW method, based on the inverse probability of treatment assignment for each patient, was employed to control for biases and confounding factors [26]. Explanatory variables used for calculating propensity scores were Child-Pugh score, AFP value, PIVKA-II, diameter of tumor, and tumor number. We selected these as the most important explanatory variables for the analysis considering clinical significance and analytic stability of propensity scores. SAS® version 9.4 (SAS Institute Inc., Cary, NC, USA) was used for all statistical analyses.

Results
A total of 74 patients were enrolled from December 2020 to September 2021 (clinical cutoff: 15 August 2022; median follow-up of 15.1 months). This slight over-enrollment occurred because patients who had already provided informed consent were allowed to enroll as the study approached its target accrual. At baseline, the median (range) age was 74.0 (41–89) years with 81.1% and 41.9% of patients having a history of hypertension or diabetes, respectively (Table 1). Overall, 51 patients (68.9%) had 1–5 tumors and 23 patients (31.1%) had ≥6 tumors while the median (range) maximum tumor diameter was 4.8 (1.0, 13.0) cm. Further, 33.8% of patients met beyond up-to-11 criteria, signifying more severe disease. The etiology of HCC was predominantly non-B/non-C (58.1%), followed by HCV (31.1%) and HBV (10.8%).

Primary Endpoint
Regarding the primary endpoint, the 6-month PFS rate by mRECIST was 66.8% (90% CI: 56.8, 75.0), and the lower limit of the 90% CI for the 6-month PFS rate exceeded the pre-specified threshold of 55%, signifying that the primary endpoint was met (Fig. 1). The median PFS by mRECIST was 9.1 months (95% CI: 7.1, 10.2).

Secondary Endpoints
Result of median PFS by RECIST v1.1 was 8.5 months (95% CI: 6.5, 10.2) (online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000546899). Subgroup analysis of PFS by mRECIST found trends toward differences in subgroups related to sex, PS, and AFP level but not other baseline characteristics (online suppl. Fig. S2).
ORR by mRECIST was 40.5% and CR was 12.2% with corresponding values including curative intent therapy effect being 44.6% and 17.6%, respectively. ORR by RECIST v1.1 was 23.0% and CR was 0.0% with corresponding values including curative intent therapy effect being 29.7% and 6.8%, respectively (online suppl. Table S1). The PD rate by mRECIST was 6.8%, which was equivalent when assessed by mRECIST including curative intent therapy effect (online suppl. Table S1). During the atezolizumab plus bevacizumab treatment period, median DoR was 10.8 months and 9.8 months by mRECIST and RECIST v1.1, respectively, and median time to response was 2.3 months and 3.0 months by mRECIST and RECIST v1.1, respectively (online suppl. Table S1).
At the time of the clinical cutoff, 15 patients (20.3%) had OS events and the median OS was not reached (NR) months (95% CI: NR–NR). The 12-month OS rate was 84.6% (95% CI: 74.0, 91.2) (Fig. 2).
In terms of safety, AEs and treatment-related AEs of any grade were common and occurred in 97.3% and 90.5% of patients, respectively (Table 2). Grade ≥3 AEs occurred in 41.9% of patients and treatment-related grade ≥3 AEs occurred in 32.4% of patients while 13.5% of patients with any grade AEs required corticosteroids (Table 2). The most common AEs of all grades with an incidence of ≥10% were hypertension, proteinuria, and malaise (Table 3). Laboratory abnormalities were also common, especially increased aspartate aminotransferase/alanine aminotransferase (AST/ALT) (55.4%/36.5%), decreased platelet count, and increased creatinine (37.8% each). Both the mean Child-Pugh and ALBI scores remained relatively constant up until cycle 15, at which point the number of patients receiving atezolizumab plus bevacizumab remained in double digits (Fig. 3). More specifically, the mean ALBI score was −2.5 ± 0.4 at baseline and −2.3 ± 0.6 at cycle 15.

Exploratory Endpoints
In terms of change in tumor size, large improvements were seen among patients with CR and PR as shown by waterfall plot (Fig. 4). In addition, there were cases in which significant tumor shrinkage was obtained even in patients for whom such effects are difficult to obtain, such as those with hypertension or diabetes as well as those who met beyond up-to-11 criteria. In most cases in which CR and PR were obtained, the deep response continued over time as shown by the spider plot (Fig. 5). However, one case of acute exacerbation from CR to PD was observed.
Conversion to curative intent therapy was observed in 10 patients, 5 of whom had tumor burden beyond the up-to-11 criteria at baseline. Of these 10 patients, 5 patients achieved CR by RECIST v1.1 (one patient who initially achieved CR by mRECIST with atezolizumab plus bevacizumab subsequently underwent resection). Curative conversion consisted of TACE in 8 patients, resection in one patient, and RFA in one patient. Two patients who received resection/RFA were alive and recurrence free at the clinical cutoff date. Among the 8 patients who underwent curative TACE conversion, the median PFS by mRECIST was longer (14.5 months [95% CI: 10.1, not available, NA]) than in the overall population (9.1 months [95% CI: 7.1, 10.2]).
Exploratory analysis using IPW, which compared results from the present study with historical TACE data, showed that patient characteristics before and after IPW were relatively well matched (online suppl. Table S2). The TACE cohort was derived from an initial group of 449 patients who underwent TACE. After excluding ineligible patients, 429 patients remained. From these, 147 patients were confirmed to meet beyond up-to-seven criteria by CT examination prior to TACE. For the analysis, we further selected 81 patients who had all necessary baseline values, such as Child-Pugh score, AFP, and PIVKA-II, required for IPW analysis. Before IPW, there were some differences in baseline characteristics between the two cohorts. The TACE cohort had a slightly higher proportion of patients with Child-Pugh score A5 and a higher prevalence of patients with ≥10 tumors compared to the atezolizumab plus bevacizumab cohort. Further, in the atezolizumab plus bevacizumab group, there were two patients with missing baseline PIVKA-II values, so we conducted the IPW analysis with 72 patients after excluding these. PFS by mRECIST tended to be superior with atezolizumab plus bevacizumab (Fig. 6). According to the 5 explanatory variables selected for IPW analysis, PFS by mRECIST according to IPW was 9.2 months with atezolizumab plus bevacizumab compared with 5.7 months with TACE (hazard ratio 0.67, p = 0.029).

Discussion
This is the first study to evaluate the efficacy and safety of atezolizumab plus bevacizumab in patients with unresectable intermediate-stage HCC of tumor burden beyond up-to-seven criteria. The primary endpoint of the 6-month PFS rate was 66.8%, exceeding the lower threshold of the 90% CI (55%). The median PFS was 9.1 months, and ORR by mRECIST including curative treatment effect was 44.6%. These results are similar to those of IMbrave150, despite differences in patients’ characteristics [18]. Notably, approximately one-third of the patients in this study had tumor burden beyond up-to-11 criteria, and complete and partial responses were also observed in these cases. A total of 10 patients underwent curative intent therapy, and half of these patients had tumor burden beyond up-to-11 criteria.
In addition, atezolizumab plus bevacizumab maintained hepatic functional reserve, as measured by Child-Pugh scores and ALBI grades. These findings suggest that atezolizumab plus bevacizumab is effective in patients with high tumor burden, while preserving liver function. This has the potential to expand opportunities for curative treatment in patients previously considered ineligible due to advanced disease.
The safety profile in our study was consistent with that of the IMbrave150 trial. Grade ≥3 treatment-related AEs occurred in 32.4% of this study compared to 43.0% in IMbrave150 [17], with similar rates of steroid-requiring AEs (13.5% vs. 12.2%) [27]. These findings suggest that atezolizumab plus bevacizumab is tolerable even in the Japanese population, which tends to be older and has more comorbidities compared to the global cohort in IMbrave150 [28].
In an exploratory analysis, favorable trends were observed in atezolizumab plus bevacizumab compared to TACE by the IPW. The median PFS by mRECIST was 9.2 months for the combination therapy group versus 5.7 months for the TACE group (hazard ratio 0.67). This finding suggests that atezolizumab plus bevacizumab may be more effective than TACE in patients with tumor burden beyond up-to-seven criteria.
TACE has limited efficacy in cases of high tumor burden, although TACE effectively reduces tumor size by interrupting blood supply to the tumor and directly delivering chemotherapeutic agents to the target lesion. Also, repeated TACE procedures in patients with high tumor burden may result in deterioration of hepatic functional reserve [10, 12, 15, 29]. Hepatic functional reserve is one of the major prognostic factors of HCC, and it is important to choose a treatment that does not impair it. TACE has been reported to enhance immunotherapy by releasing tumor antigens [30, 31]. Furthermore, partial embolization techniques, instead of complete embolization, minimize the decline in hepatic functional reserve associated with TACE [32]. In recent years, several trials evaluating the efficacy of combination therapies involving immune-oncology (IO)-based therapy plus TACE [31, 33] are ongoing. Of them, some trials including the EMERALD-1 [34] and LEAP-012 trials have shown promising results [35]. Subgroup analyses in EMERALD-1 and LEAP-012 demonstrated a PFS benefit for IO-based therapy plus TACE in patients with high tumor burden. While these results are encouraging, our exploratory analysis also suggested that atezolizumab plus bevacizumab showed improved PFS compared to TACE. In patients with high tumor burden beyond up-to-seven criteria, better characterization is needed to understand which patients may benefit from adding TACE to IO-based systemic therapy. These findings highlight the need for further investigation and multidisciplinary approaches to optimize treatment strategies and improve outcomes for patients with unresectable HCC.

Limitations
This was a single-arm study with no comparator group and efficacy and safety were not compared with prospectively collected TACE treatment groups. The initial analysis was limited to a short follow-up period (15.1 months) and population size (74 patients). Additionally, although we performed an exploratory analysis using IPW to compare with historical TACE data, the IPW method may not have fully adjusted for all potential confounding factors. Finally, quality of life was not investigated in this study and, because of atezolizumab plus bevacizumab treatment, we could not evaluate the impact of treatment visits and AEs on quality of life.

Conclusions
This study suggests that it is preferable to use atezolizumab plus bevacizumab as first-line treatment for patients with unresectable HCC who are unsuitable for TACE due to large or multiple tumors. Future strategies combining systemic and locoregional therapies, such as conversion surgery/ablation or ongoing trials of TACE plus immunotherapy, may improve outcomes by utilizing multimodal approaches.

Acknowledgment
The authors thank Mark Snape, MB BS, of inScience Communications, Springer Healthcare, for providing medical writing assistance.

Statement of Ethics
The study protocol and written information for patients were approved by the Certified Review Board (CRB) of Kyushu University. This approval, including the informed consent form (ICF), was applicable to all participating sites. The full list of participating sites can be found at Japan Registry of Clinical Trial (jRCTs071200051). The study was conducted in accordance with the Declaration of Helsinki, the Japanese Clinical Trials Act, the Ordinance for Enforcement of the Clinical Trials Act of the Ministry of Health, Labour and Welfare, and related notifications. All participants provided written informed consent using the CRB-approved form.

Conflict of Interest Statement
M.K. has received grants for research from Otsuka Pharmaceutical Co., Ltd., Chugai Pharmaceutical Co., Ltd., Taiho Pharmaceutical Co., Ltd., G.E. Healthcare Japan Corporation, Eisai Co., Ltd., and AbbVie GK; consulting fees from Chugai Pharmaceutical Co., Ltd., F. Hoffmann-La Roche, Ltd., Eisai Co., Ltd., and AstraZeneca K.K.; and honoraria for lectures from Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., AstraZeneca K.K., Eli Lilly Japan K.K., and Takeda Pharmaceutical Co., Ltd., and is the Editorial-in-Chief of Liver Cancer. K.U. has received grants for research from Chugai Pharmaceutical Co., Ltd.; consulting fees from Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., AstraZeneca K.K., Pfizer, Takeda Pharmaceutical Co., Ltd., Eli Lilly Japan K.K., and Bayer Yakuhin, Ltd.; honoraria for lectures from Chugai Pharmaceutical Co., Ltd., Takeda Pharmaceutical Co., Ltd., Merck Sharp and Dohme, Sumitomo Pharma Co., Ltd., EA Pharma Co., Ltd., ASKA Pharmaceutical Co., Ltd., Nippon Kayaku Co., Ltd., Otsuka Pharmaceutical Co., Ltd., Eisai Co., Ltd., Eli Lilly Japan K.K., AstraZeneca K.K., Taiho Pharmaceutical Co., Ltd., AbbVie GK, Bayer Yakuhin, Ltd., Boston Scientific Japan K.K., and Kowa Company, Ltd.; and support for attending meetings and/or travel from Chugai Pharmaceutical Co., Ltd., and is an editorial board member of Liver Cancer. K.T. has received honoraria for lectures from Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., Takeda Pharmaceutical Co., Ltd., and AstraZeneca K.K. and is an editorial board member of Liver Cancer. T.Y. has received honoraria for lectures from Chugai Pharmaceutical Co., Ltd., AstraZeneca K.K., and Eisai Co., Ltd., and is an editorial board member of Liver Cancer. S.S. has received honoraria for lectures from AstraZeneca K.K., Chugai Pharmaceutical Co., Ltd., and Eisai Co., Ltd. S.I. has received honoraria for lectures from Chugai Pharmaceutical Co., Ltd., AstraZeneca K.K., and Eisai Co., Ltd. Y.T. has received scholarship donations from AbbVie GK and Otsuka Pharmaceutical Co., Ltd.; research funding from AbbVie GK, FUJIREBIO Inc., Sysmex Corp, GlaxoSmithKline PLC, Gilead Sciences, Inc., and Janssen Pharmaceutical K.K.; and lecture fees from AbbVie GK, Gilead Sciences, Inc., Chugai Pharmaceutical Co., Ltd., ASKA Pharmaceutical Holdings Co., Ltd., Otsuka Pharmaceutical Co., Ltd, Takeda Pharmaceutical Co., Ltd, GlaxoSmithKline PLC, AstraZeneca K.K., Eisai Co., Ltd., and HU frontier. H.K. has received honoraria for lectures from Eisai Co., Ltd., and Chugai Pharmaceutical Co., Ltd. H.A. has received honoraria for lectures from Chugai Pharmaceutical Co., Ltd. A.H. has received honoraria for lectures from Chugai Pharmaceutical Co., Ltd., Lilly, AstraZeneca K.K., Bayer, Eisai Co., Ltd., and Takeda Pharmaceutical Co., Ltd. M.M. has received consulting fees from Chugai Pharmaceutical Co., Ltd., and AstraZeneca K.K. and honoraria for lectures from Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., and AstraZeneca K.K. R.T. has received honoraria for lectures from Chugai Pharmaceutical Co., Ltd., AstraZeneca K.K., and Eisai Co., Ltd., and is an editorial board member of Liver Cancer. S.O. has received grants for research from Eisai Co., Ltd., Chugai Pharmaceutical Co., Ltd., Gilead Sciences, AstraZeneca K.K., Bayer, and Eli Lilly; consulting fees from MSD, AstraZeneca K.K., and Chugai Pharmaceutical Co., Ltd.; and honoraria for lectures from Eisai, Chugai Pharmaceutical Co., Ltd., Gilead Sciences, Takeda, AstraZeneca K.K., Bayer, Eli Lilly, and AbbVie and participated on advisory boards for MSD, AstraZeneca K.K., and Chugai Pharmaceutical Co., Ltd. K.Y. has received research funding from Otsuka Pharmaceutical Co., Ltd., Chugai Pharmaceutical Co., Ltd, Kyowa Kirin Co., Ltd., Euro Doctor Concierge, LLC, and Daiichi Sankyo Co., Ltd.; consulting fees from SYSMEX CORPORATION and Delta-Fly Pharma; and honoraria for lectures from FUJIFILM Toyama Chemical Co., Ltd., CMIC CO., Ltd, Daiichi Sankyo Co., Ltd., TME Therapeutics Inc., Takeda Pharmaceutical Co., Ltd., CM Plus Corporation, and AstraZeneca. M.I. has received research funding paid to his institution from AbbVie, AstraZeneca K.K., Bayer, Bristol Myers Squibb, Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., and MSD, also served in a consulting or advisory role for AbbVie, AstraZeneca K.K., Bayer, Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., Eli Lilly Japan, MSD, and Ono Pharmaceutical Co., Ltd.; and received honoraria from Abbott, AstraZeneca, Bristol Myers Squibb, Chugai Pharmaceutical Co., Ltd., Eisai Co., Ltd., Eli Lilly Japan, MSD, and Takeda Pharmaceutical Co., Ltd., and is an editorial board member of Liver Cancer. K.N. and Y.K. have no conflicts of interest to disclose.

Funding Sources
This study, including medical writing assistance, was funded by Chugai Pharmaceutical Co., Ltd.

Author Contributions
M.K., K.U., K.T., T.Y., H.A., A.H., M.M., R.T., S.O., K.Y., and M.I. were responsible for conception and design of the study. K.Y. performed data analysis. M.K., K.U., K.T., T.Y., S.S., K.N., Y.K., S.I., Y.T, H.K., H.A., A.H., M.M., R.T., S.O., K.Y., and M.I. participated in data interpretation, contributed to important intellectual content during manuscript drafting or revision, and approved the final manuscript.