🇺🇸 ASCO Annual Meeting 2026 にて国立がん研究センターとの共同研究をポスター発表 🧬
米国シカゴで開催されたASCO Annual Meetingで、国立がん研究センター中央病院の井川由季子先生、堀之内秀仁先生らとの共同研究のポスター発表が行われました。
本演題は「Applying a large language model to detect immune-related adverse events in electronic health records: Effect of integrating prescription and laboratory data on accuracy.」と題し、肺がん症例において免疫チェックポイント阻害薬の副作用である「免疫関連有害事象(irAE)」を、電子カルテ情報からLLMで早期検出できるかを調べた研究です。これまで臨床医が手作業で電子カルテを見ていた作業を、生成AIを使い支援する取り組みです。電子カルテの診療録のみのシングルモーダル解析と、診療録に加えて全身性ステロイドの処方情報や検査データを加えたマルチモーダル解析を行いました。296症例を対象としたところ、マルチモーダル解析の方が精度が上がるとともに、約4日早くirAEの検出を出来る結果となりました。
テンクーの竹内、西村も共著として参加しています。
テンクーは今後も、AIなどのデジタル技術を通じて、ガイドライン支援を含むがん医療・研究の現場に広く貢献してまいります。
🇺🇸 Poster presentation of joint research with National Cancer Center Hospital, Japan, at the ASCO Annual Meeting 2026 🧬
We're excited to share that a collaborative research poster with Dr. Yukiko Igawa and Dr. Hidehito Horinouchi of the National Cancer Center Hospital Japan was presented at the ASCO Annual Meeting in Chicago, USA.
The study, titled "Applying a large language model to detect immune-related adverse events in electronic health records: Effect of integrating prescription and laboratory data on accuracy," investigated whether LLMs can be used to detect immune-related adverse events (irAEs) — side effects of immune checkpoint inhibitors — from electronic health records (EHRs) in lung cancer patients. This work aims to alleviate clinician workload by automating a task that has traditionally required a person's review of patient records.
The research compared a unimodal approach using clinical notes alone against a multimodal approach that also incorporated systemic steroid prescription data and lab results. Across 296 cases, the multimodal analysis not only achieved higher accuracy but also accomplished irAE detection approximately 4 days earlier.
Xcoo's Takeuchi and Nishimura participated as co-authors on this work.
We remain committed to contributing towards cancer care and research — including clinical guideline support — through AI and digital innovation.
Yukiko Shimoda Igawa, Hidehito Horinouchi, Toshiki Takeuchi, Kunihiro Nishimura, Masaaki Komatsu, Nobuji Kouno, Satoshi Takahashi, Masami Mukai, Tatsuya Yoshida, Yasushi Goto, Ryuji Hamamoto, Applying a large language model to detect immune-related adverse events in electronic health records: Effect of integrating prescription and laboratory data on accuracy. J Clin Oncol. 2026;44:12133-12133. doi: 10.1200/JCO.2026.44.16_suppl.12133.
https://t.co/e2AvUvkpgu
#ASCO2026 #肺がん #irAE #医療AI #LLM #テンクー #Xcoo #NCCJ
📄 共同研究成果が国際学術誌「ESMO Open」に掲載
このたび、テンクーが西日本がん研究機構(WJOG)らと共同研究として参画した研究成果が、欧州臨床腫瘍学会 (European Society for Medical Oncology:ESMO) が発行する国際査読付き学術誌 「ESMO Open」 に掲載されました。
本論文「Gene-Testing and Prognosis in Postoperative Recurrent Non-Small Cell Lung Cancer: Subgroup Analyses of WJOG15421L」は、 術後再発非小細胞肺がんにおける遺伝子検査と予後の関連を検討した研究であり、臨床的意義の高い知見を示すものです。
術後再発非小細胞肺がんの症例において、多遺伝子パネル検査は44%しか実施されておらず、十分な手術標本が入手可能であるにもかかわらず低いことが明らかになりました。周術期の個別化医療の進歩に伴い、多遺伝子パネル検査の早期実施は、切除された患者の予後改善に向けた最も重要な課題の一つです。
テンクーは今後も、医療機関および研究機関との連携を通じて、科学的エビデンスの創出と医療現場への実装を推進してまいります。
📄 Collaborative Research Published in the International Journal ESMO Open
We are pleased to announce that the results of a collaborative research project with WJOG in which Xcoo participated have been published in ESMO Open, a peer-reviewed international academic journal published by the European Society for Medical Oncology (ESMO).
The paper, titled “Gene-Testing and Prognosis in Postoperative Recurrent Non-Small Cell Lung Cancer: Subgroup Analyses of WJOG15421L,” investigates the relationship between genetic testing and prognosis in patients with postoperative recurrent non-small cell lung cancer and provides clinically significant insights.
This paper shows that multigene testing was conducted in only 44% of post-operative cases, highlighting its underutilization in post-operative recurrent NSCLC cases. Early multigene testing is crucial for advancing personalized perioperative care.
Xcoo will continue to promote the generation of scientific evidence and its implementation in clinical practice through collaboration with medical and research institutions.
Matsubara T, Sakamoto T, Takahama T, Yokoyama T, Yoshino I, Akamatsu H, Yamaguchi M, Baba J, Tokito T, Tachihara M, Sato Y, Takenaka T, Sugio K, Mori M, Takeuchi T, Okazuka K, Kenmotsu H, Fujimoto J, Shimokawa M, Yamamoto N, Nakagawa K. Gene testing and prognosis in post-operative recurrent non-small-cell lung cancer: subgroup analyses of WJOG15421L. ESMO Open. 2026 Mar 31;11(4):106890. https://t.co/zURHrhn67Q
#ESMOOpen #Xcoo #テンクー #ゲノム医療 #WJOG
FIRST-EVER WRC WIN! 🏆
History made in Kenya! Takamoto Katsuta and Aaron Johnston take their maiden WRC win at the demanding Safari Rally Kenya. A sensational drive through one of rally’s toughest challenges!
#WRC | #SafariRallyKenya 🇰🇪
📘「医療AIの臨床実装――医療現場で使用されるAI技術」にテンクーの竹内が寄稿しました
国立がん研究センター研究所の浜本隆二先生による本「医療AIの臨床実装――医療現場で使用されるAI技術」が発売されました。
臨床応用を目的とした医療AI研究開発パートの「自然言語処理を用いた電子カルテ情報の解析」を弊社の竹内、西村、そして国立がん研究センター中央病院の堀之内先生による共著として掲載されています。
ぜひお手に取ってご覧ください。
中外医学社の書籍ページ
https://t.co/ua0djTfIgQ
📘 Xcoo’s Takeuchi Contributes to the publication "Clinical Implementation of Medical AI—AI Technologies Used in Healthcare Settings"
The book "Clinical Implementation of Medical AI—AI Technologies Used in Healthcare Settings," authored by Dr. Ryuji Hamamoto of the National Cancer Center Research Institute, has just been published.
Within it, our colleagues Takeuchi and Nishimura co-authored the chapter “Analyzing Electronic Medical Record Data Using Natural Language Processing,” along with Dr. Horinouchi from the National Cancer Center Hospital East. The chapter is featured in the section on medical AI research and development aimed at real-world clinical applications.
We invite you to take a look!
Link to book
https://t.co/ua0djTfIgQ
#中外医学社 #Xcoo #テンクー #ゲノム医療 #PrecisionMedicine #LLM
🇩🇪 ESMO AI & Digital Oncology Congress 2025 にて国立がん研究センターとの共同研究をポスター発表🧬
ベルリンで開催された国際会議ESMO AI & Digital Oncology Congress 2025で、国立がん研究センター中央病院 先端医療科との共同研究の成果ポスターが採択され、現地で発表を行いました。
本演題は、今年の日本メディカルAI学会優秀一般演題賞を受賞した「肺癌診療ガイドライン作成における大規模言語モデルを用いた文献スクリーニング」の内容を発展させ、国際会議の場で共有したものです。ガイドライン作成における透明性・再現性を高めるため、文献の採否だけでなく除外理由を提示可能にしました。テンクーの竹内、青木、西村も引き続き共著として参加しています。
テンクーは今後も、AIなどのデジタル技術を通じて、ガイドライン支援を含むがん医療・研究の現場に広く貢献してまいります。
🇩🇪 The poster which summarizes the findings from our joint research with the National Cancer Center was presented at the ESMO AI & Digital Oncology Congress 2025 🧬
Our poster, summarizing the findings from our joint research with the Department of Advanced Medicine at the National Cancer Center Hospital, was accepted and presented at the international ESMO AI & Digital Oncology Congress 2025, held in Berlin.
This presentation builds on the work, “Literature Screening Using Large Language Models in the Development of Lung Cancer Clinical Practice Guidelines,” which received the Excellent Oral Presentation Award at this year’s Japan Medical AI Society meeting. For this event, we further expanded on how the LLM can filter which clinic research paper should be included, and furthermore provide reasoning for any exclusion of research papers made. Overall, this clarification enhances transparency and reproducibility in guideline development. For this presentation, Xcoo's Takeuchi, Aoki, and Nishimura also participated as co-authors.
Xcoo will continue its contribution in cancer care and research, including guideline development support, through AI and other digital technologies.
https://t.co/8OaIhEMcFg
Katsuya Y, Takeuchi T, Aoki T, Nishimura K, Koyama T. A Fine-Tuned Large Language Model for Literature Screening and Exclusion Reason Prediction in Lung Cancer Clinical Practice Guideline Development. ESMO AI & Digital Oncology Congress 2025. 2025 Nov.
#ESMOAI #ESMOAI25 #Xcoo #テンクー #LLM