論文

PURPOSETreatment options are limited for patients with previously treated metastatic colorectal cancer (mCRC). In the LEAP-017 study, we evaluate whether lenvatinib in combination with pembrolizumab improves outcomes compared with standard of care (SOC) in previously treated mismatch repair proficient or not microsatellite instability high (pMMR or not MSI-H) mCRC.METHODSIn this international, multicenter, randomized, controlled, open-label, phase III study, eligible patients age 18 years and older with unresectable, pMMR or not MSI-H mCRC, that had progressed on or after, or could not tolerate, standard treatment, were randomly assigned 1:1 to lenvatinib 20 mg orally once daily plus pembrolizumab 400 mg intravenously once every 6 weeks or investigator's choice of regorafenib or trifluridine/tipiracil (SOC). Randomization was stratified by presence or absence of liver metastases. The primary end point was overall survival (OS). LEAP-017 is registered at ClinicalTrials.gov (NCT04776148), and has completed recruitment.RESULTSBetween April 8, 2021, and December 21, 2021, 480 patients were randomly assigned to lenvatinib plus pembrolizumab (n = 241) or SOC (n = 239). At final analysis (median follow-up of 18.6 months [IQR, 3.9]), median OS with lenvatinib plus pembrolizumab versus SOC was 9.8 versus 9.3 months (hazard ratio [HR], 0.83 [95% CI, 0.68 to 1.02]; P =.0379; prespecified threshold P =.0214). Grade ≥3 treatment-related adverse events occurred in 58.4% (lenvatinib plus pembrolizumab) versus 42.1% (SOC) of patients. Two participants died due to treatment-related adverse events, both in the lenvatinib plus pembrolizumab arm.CONCLUSIONIn patients with pMMR or not MSI-H mCRC that had progressed on previous therapy, there was no statistically significant improvement in OS after lenvatinib plus pembrolizumab treatment versus SOC. No new safety signals were observed.

This paper introduces LLM-jp, a cross-organizational project for the research and development of Japanese large language models (LLMs). LLM-jp aims to develop open-source and strong Japanese LLMs, and as of this writing, more than 1,500 participants from academia and industry are working together for this purpose. This paper presents the background of the establishment of LLM-jp, summaries of its activities, and technical reports on the LLMs developed by LLM-jp. For the latest activities, visit https://llm-jp.nii.ac.jp/en/.

Background: The presence of circulating tumor DNA (ctDNA) can identify patients (pts) at higher risk of recurrence. Longitudinal surveillance of ctDNA may enable early identification of pts who are likely to relapse and presents a window of opportunity for early interventions to improve outcomes. A tumor-naïve plasma-only approach for minimal residual disease (MRD) assessment accelerates turnaround time, enabling rapid adjuvant chemotherapy (ACT) treatment decisions and ongoing molecular surveillance. Methods: Samples were collected from GALAXY in CIRCULATE-Japan and analyzed for postsurgical ctDNA detection at a landmark time point (LMT), 4 weeks after surgery in pathological stage II or III colorectal cancer (CRC). Initially, 80 pts meeting pre-specified eligibility criteria were randomly selected, with enrichment for recurrence to 50% while maintaining the stage II:III recurrent (R)/nonrecurrent (NR) ratio observed in GALAXY. Of those, 23 pts also had longitudinal plasma samples analyzed to date (including 17 false negative (FN) at LMT). Residual plasma samples were analyzed with the Tempus xM MRD assay (xM), a tumor-naïve MRD ctDNA assay that integrates methylation and genomic variant data to deliver a binary MRD call blinded to clinical outcomes. The methylation workflow detects fragments with CRC methylation signatures in differentially methylated regions. The variant classifier detects highly prevalent CRC variants. Longitudinal MRD+ status was defined as any MRD+ call after surgery or ACT; lead time as the time from first MRD+ call after definitive therapy to clinical recurrence. Results: Of the 80 pts, 70 [Stage II n= 29 (41%), Stage III n= 41 (59%)] were evaluable (36 R and 34 NR). At the LMT and with at least 1-year follow-up, 18/36 R pts had detectable ctDNA (MRD+) and 30/34 NR pts had undetectable ctDNA (MRD-), providing clinical sensitivity (CS) of 50% and specificity of 88%. Additional sensitivity analysis included 2 false positive pts who cleared ctDNA with ACT and provided CS 53% and specificity 94% at LMT. Longitudinal analysis of all FNs and ACT treated true positive (TP) pts improved CS from 50% at LMT to 90.9% (stage II 91.7%, stage III 90.5%) on surveillance. The mean lead time was 4.72 mos overall, 5.29 mos for pts treated with surgery alone. Additionally, xM MRD status at LMT strongly correlated with disease-free survival (DFS) with a hazard ratio (HR) of 5.09, superior to 12 week carcinoembryonic antigen (CEA) correlation with DFS, HR 2.63. Conclusions: xM is a novel tumor-naïve MRD assay that demonstrated remarkable clinical longitudinal performance, and can accurately predict clinical recurrence on surveillance. xM ctDNA status was a strong prognostic biomarker to DFS and superior to standard of care CEA. Research Sponsor: Tempus AI, Inc.

Background: Evaluation for activating mutations in KRAS, NRAS, and BRAF in colorectal cancer (CRC) and in KRAS in pancreatic ductal adenocarcinoma (PDAC) is essential for clinical care. Plasma cell-free DNA (cfDNA) next-generation sequencing (NGS) allows convenient assessment of a tumor’s molecular profile, however low tumor DNA shedding limits sensitivity. We investigated mutant allele frequency (MAF) of other oncogenic dominant genes to identify a threshold for accurate detection of KRAS, NRAS, and BRAF (RAS/RAF) mutations in cfDNA. Methods: Molecular and clinical data were obtained from the Duke Molecular Registry of Tumors and the SCRUM-Japan GOZILA study. Patients with CRC or PDAC and a KRAS, NRAS, or BRAF activating single nucleotide variant (SNV) present on tissue NGS and with available cfDNA assays were included. Recursive partitioning and Wilcoxon-rank statistics methods identified potential cut-points for discriminative MAF values. Results: One hundred and thirty-five CRC and 30 PDAC cases with 198 total cfDNA assays met criteria. Greatest non-RAS/RAF dominant gene MAF of 0.34% provided maximum discrimination for predicting RAS/RAF SNV detection. Sensitivity for RAS/RAF SNVs increased with dominant gene MAF, with MAF ≥1% predicting sensitivity >98%, MAF between 0.34 and 1% predicting sensitivity of 84.0%, and MAF ≤0.34% predicting sensitivity of 50%. For 43 cfDNA assays that did not detect RAS/RAF SNVs, 18 assays detected 34 other oncogenic variants, of which 80.6% were not also detected on tissue. Conclusions: Non-RAS/RAF dominant oncogenic mutation MAF ≥1% on cfDNA NGS predicts high sensitivity to detect RAS/RAF oncogenic SNVs in CRC and PDAC. MAF ≤0.34% indicates an assay may not reliably detect RAS/RAF SNVs, despite detection on tissue testing. Most variants from assays that did not detect RAS/RAF had MAF <1% and were not detected on tissue, suggesting potential confounding. These data suggest a practical approach to determining cfDNA assay adequacy, with implications for guiding clinical decisions in CRC and PDAC.

Background: Previous clinical trials have demonstrated the potential efficacy of poly (ADP-ribose) polymerase (PARP) inhibitors (PARPis) in patients with cancer involving homologous recombination repair (HRR) gene-mutation. Moreover, HRR gene-mutated cancers are effectively treated with immune checkpoint inhibitors (ICIs) with the increase in tumor mutation burden. We have proposed to conduct a multicenter, single-arm phase II trial (IMAGENE trial) for evaluating the efficacy and safety of niraparib (PARPi) plus programmed cell death-1 inhibitor combination therapy in patients with HRR gene-mutated cancers who are refractory to ICIs therapy using a next generation sequencing-based circulating tumor DNA (ctDNA) and tumor tissue analysis. Methods: Key eligibility criteria for this trial includes HRR gene-mutated tumor determined by any cancer gene tests; progression after previous ICI treatment; and Eastern Cooperative Oncology Group Performance Status ≤ 1. The primary endpoint is the confirmed objective response rate (ORR) in all patients. The secondary endpoints include the confirmed ORR in patients with HRR gene-mutation of ctDNA using the Caris Assure (CARIS, USA). The target sample size of the IMAGENE trial is 57 patients. Biomarker analyses will be performed in parallel using the Caris Assure, proteome analysis, and T cell repertoire analysis to reveal tumor immunosurveillance in peripheral blood. Expected outcome: Our trial aims to confirm the clinical benefit of PARPi plus ICI combination therapy in ICI-resistant patients. Furthermore, through translational research, our trial will shed light on which patients would benefit from the targeted combination therapy for patients with HRR gene-mutated tumor even after the failure of ICIs. Trial registration: The IMAGENE trial: jRCT, Clinical trial no.: jRCT2051210120, Registered date: November 9, 2021.

PURPOSE Earlier detection of cancer recurrence using circulating tumor DNA (ctDNA) to detect molecular residual disease (MRD) has the potential to dramatically affect cancer management. We review evidence supporting the use of ctDNA as a biomarker for detection of MRD and highlight the potential impact that ctDNA testing could have on the conduct of clinical trials. METHODS We searched the literature using MEDLINE (via PubMed) for articles from January 1, 2000, focusing on studies that assessed ctDNA as a predictor of cancer recurrence. Broadly focused searches on ctDNA and cancer were also performed to provide additional background information. www.clinialtrials.gov was searched to identify trials that incorporate ctDNA testing. RESULTS Numerous studies across different cancer types indicate that ctDNA-based MRD detection predicts recurrence with high sensitivity and specificity, and with lead times that precede standard imaging by up to 12 months. Recently, ctDNA testing has started being used to enroll MRD-positive patients at high risk of recurrence into trials, promising gains in statistical power that allow clinical utility to be demonstrated with smaller cohorts. Trials where ctDNA testing based-MRD detection is used to stratify patients into low or high-risk categories for treatment assignment are also ongoing. In addition, there is increasing evidence supporting the use of ctDNA dynamics or clearance as a surrogate end point, which could significantly reduce trial duration. CONCLUSION ctDNA-based trial enrichment across many cancers seems likely to become increasingly common for cost- and time-reduction benefits. Trial efficiency could also benefit from using ctDNA as a surrogate end point, leading to accelerated approval of new therapeutics. A clear demonstration of efficacy from trials that use ctDNA-based MRD detection to assign treatment could transform clinical practice.

Background: Human epidermal growth factor receptor 2 (HER2) and claudin 18.2 (CLDN18.2) are key therapeutic targets in metastatic gastric and gastroesophageal junction cancer (mGC/GEJC). Trastuzumab deruxtecan (T-DXd) is standard care for previously treated HER2-positive mGC/GEJC, but the prognostic impact of CLDN18.2 remains unclear. Patients and methods: We retrospectively reviewed patients with HER2-positive mGC/GEJC treated with T-DXd at National Cancer Center Hospital East until February 2025. T-DXd outcomes were compared between CLDN18.2-positive (≥2+ in ≥75% of tumor cells) and -negative patients in two cohorts: those with HER2 positivity before first-line therapy (overall cohort) and those maintaining HER2 positivity immediately before T-DXd (remaining HER2 cohort). CLDN18.2 expression was evaluated at any time before T-DXd administration. Results: Eighty-seven patients were included in the overall cohort and 30 in the remaining HER2 cohort. In the overall cohort, progression-free survival (PFS) was shorter in CLDN18.2-positive patients [3.9 versus 5.3 months; hazard ratio (HR) 1.87, 95% confidence interval (CI) 1.04-3.34, P = 0.036], with an adjusted HR of 1.82 (95% CI 1.02-3.28, P = 0.047). Overall survival (OS) showed a shorter trend (8.6 versus 11.2 months, HR 1.36, 95% CI 0.76-2.45, P = 0.30). In the remaining HER2 cohort, CLDN18.2-positive patients had shorter PFS (3.2 versus 8.0 months, HR 3.40, 95% CI 1.38-8.40, P = 0.008) and OS (7.1 versus 12.9 months, HR 2.44, 95% CI 1.04-5.74, P = 0.041). Conclusions: CLDN18.2 positivity may attenuate the efficacy of T-DXd in HER2-positive mGC/GEJC, supporting the rationale for dual blockade of CLDN18.2 and HER2.

Background: The effect of claudin 18.2 (CLDN18.2) expression on the outcomes of a first-line immune checkpoint inhibitor (ICI)-containing chemotherapy (ICI-chemo) in patients with human epidermal growth factor receptor 2 (HER2)-negative, proficient mismatch repair (pMMR), and programmed death-ligand 1 (PD-L1)-expressing metastatic or recurrent gastric or gastroesophageal junction cancers (mGC/GEJC) remains unclear. Objectives: We assessed the effects of CLDN18.2 expression on the outcomes of first-line ICI-containing chemotherapy in patients with HER2-negative, pMMR, and PD-L1-expressing mGC/GEJC. Patients and methods: Medical records of patients with HER2-negative, pMMR, and PD-L1 combined positive score (CPS) ⩾1 unresectable/metastatic or recurrent GC/GEJC who received first-line ICI-chemo or chemotherapy alone (chemo-alone) between January 2016 and August 2024 were retrospectively analyzed. The impact of CLDN18.2 status on the clinical outcomes was evaluated in patients receiving ICI-chemo and chemo-alone to assess the prognostic significance of CLDN18.2 in the absence of ICI. Results: A total of 150 patients were treated with ICI-chemo, whereas 313 patients received chemo-alone. CLDN18.2 positivity (⩾2+ in ⩾75% tumor cells) was identified in 42 patients (28.0%) in the ICI-chemo group and 94 patients (30.0%) in the chemo-alone group. There were no significant differences in the objective response rate (ORR; 68.3% vs 70.3%, p = 0.842), disease control rate (DCR; 92.7% vs 94.1%, p = 0.718), progression-free survival (PFS; hazard ratio (HR) 1.01, p = 0.955), or overall survival (OS; HR 1.12, p = 0.615) between CLDN18.2-positive and CLDN18.2-negative patients in the ICI-chemo group. Similarly, the DCR, ORR, PFS, and OS outcomes were comparable between the CLDN18.2-positive and -negative patients in the chemo-alone group. Conclusion: CLDN18.2 expression exerted no impact on outcomes of first-line ICI-chemo and chemo-alone in patients with HER2-negative, pMMR, and PD-L1 CPS ⩾1 mGC/GEJC.

Background: Benchmarking international cancer survival differences is necessary to evaluate and improve healthcare systems. Our aim was to assess the potential regional differences in outcomes among patients with metastatic colorectal cancer (mCRC) participating in international randomized clinical trials (RCTs). Design: Countries were grouped into 11 regions according to the World Health Organization and the EUROCARE model. Meta-analyses based on individual patient data were used to synthesize data across studies and regions and to conduct comparisons for outcomes in a two-stage random-effects model after adjusting for age, sex, performance status, and time period. We used mCRC patients enrolled in the first-line RCTs from the ARCAD database, which provided enrolling country information. There were 21,509 patients in 27 RCTs included across the 11 regions. Results: Main outcomes were overall survival (OS) and progression-free survival (PFS). Compared with other regions, patients from the United Kingdom (UK) and Ireland were proportionaly over-represented, older, with higher performance status, more frequently male, and more commonly not treated with biological therapies. Cohorts from central Europe and the United States (USA) had significantly longer OS compared with those from UK and Ireland ( p = 0.0034 and p < 0.001, respectively), with median difference of 3–4 months. The survival deficits in the UK and Ireland cohorts were, at most, 15% at 1 year. No evidence of a regional disparity was observed for PFS. Among those treated without biological therapies, patients from the UK and Ireland had shorter OS than central Europe patients ( p < 0.001). Conclusions: Significant international disparities in the OS of cohorts of mCRC patients enrolled in RCTs were found. Survival of mCRC patients included in RCTs was consistently lower in the UK and Ireland regions than in central Europe, southern Europe, and the USA, potentially attributed to greater overall population representation, delayed diagnosis, and reduced availability of therapies.

PURPOSE – Human epidermal growth factor receptor 2 (HER2) amplification/overexpression (HER2-pos) is detected in 5% of RAS/BRAF wild-type metastatic colorectal cancers (mCRCs). Its prognostic/predictive role in terms of benefit from anti-EGFR/bevacizumab (bev) is debated. Similarly, the role of activating HER2 mutations (mut) is unclear.METHODS – We collected individual data of 1, 604 patients with proficient mismatch repair (pMMR)/microsatellite stable (MSS) RAS/BRAF wild-type untreated mCRC with HER2 amplification/expression status available enrolled in eight randomized clinical trials (RCT; TRIBE2, TRIPLETE, VALENTINO, ATEZOTRIBE, PANDA, PANAMA, PARADIGM, and CALGB/SWOG80405). Objective response rate (ORR), progression-free survival (PFS), and overall survival (OS) were assessed with respect to HER2 amplification/expression and HER2 mutational status and according to biologics (anti-EGFR/bev).RESULTS – Patients with HER2-pos were 81 (5%). HER2-pos patients experienced shorter PFS (median PFS [mPFS]: 9.8 v 12.2 months, hazard ratio [HR], 1.31, P =.02) and OS (median OS [mOS]: 28.0 v 34.9 months, HR, 1.37, P =.01), also after adjustment for covariates (PadjPFS =.02, PadjOS =.048). ORR was similar between HER2-pos and HER2-negative (HER2-neg) tumors (75% v 72%, odds ratio [OR], 1.21, P =.47). We found no interaction between HER2 amplification/expression status and biologics' effect in terms of PFS (Pint =.76), OS (Pint =.76), and ORR (Pint =.64). In left-sided HER2-pos tumors, outcomes were similar with chemotherapy plus bev/anti-EGFRs in terms of PFS (9.8 v 9.3 months, HR, 0.73, P =.29), OS (29.8 v 28.0 months, HR, 1.29, P =.40), and ORR (59% v 79%, OR, 0.39, P =.10). HER2-mutant tumors (2% of patients with HER2-neg tumors) showed shorter OS than HER2 wild-type ones (mOS: 23.7 v 34.4 months, HR, 1.56, P =.04) with no differential effect of biologics (PintORR =.81; PintPFS =.95; PintOS =.92).CONCLUSION – To our knowledge, this is the largest analysis of HER2 status in patients with untreated mCRC enrolled in RCT. Waiting for targeted approaches, HER2-pos and mut do not predict benefit from bev/anti-EGFRs and should be regarded as negative prognostic factors in pMMR/MSS RAS/BRAF wild-type mCRC.

Background: In metastatic colorectal cancer (mCRC), KRAS mutations are often associated with poorer survival; however, the prognostic impact of specific point mutations is unclear. In the phase III SUNLIGHT trial, trifluridine/tipiracil (FTD/TPI) plus bevacizumab significantly improved overall survival (OS) versus FTD/TPI alone. We assessed the impact of KRASG12 mutational status on OS in SUNLIGHT. Patients and methods: In the global, open-label, randomized, phase III SUNLIGHT trial, adults with mCRC who had received no more than two prior chemotherapy regimens were randomized 1 : 1 to receive FTD/TPI alone or FTD/TPI plus bevacizumab. In this post hoc analysis, OS was assessed according to the presence or absence of a KRASG12 mutation in the overall population and in patients with RAS-mutated tumors. Results: Overall, 450 patients were analyzed, including 302 patients in the RAS mutation subgroup (214 with a KRASG12 mutation and 88 with a non-KRASG12 RAS mutation). In the overall population, similar OS outcomes were observed in patients with and without a KRASG12 mutation [median 8.3 and 9.2 months, respectively; hazard ratio (HR) 1.09, 95% confidence interval (CI) 0.87-1.4]. Similar OS outcomes were also observed in the subgroup analysis of patients with a KRASG12 mutation versus those with a non-KRASG12 RAS mutation (HR 1.03, 95% CI 0.76-1.4). FTD/TPI plus bevacizumab improved OS compared with FTD/TPI alone irrespective of KRASG12 mutational status. Among patients with a KRASG12 mutation, the median OS was 9.4 months with FTD/TPI plus bevacizumab versus 7.2 months with FTD/TPI alone (HR 0.67, 95% CI 0.48-0.93), and in patients without a KRASG12 mutation, the median OS was 11.3 versus 7.1 months, respectively (HR 0.59, 95% CI 0.43-0.81). Conclusions: The presence of a KRASG12 mutation had no detrimental effect on OS among patients treated in SUNLIGHT. The benefit of FTD/TPI plus bevacizumab over FTD/TPI alone was confirmed independently of KRASG12 status.