2023 Fiscal Year Final Research Report
Development of AI for the drug therapy of hepatocellular carcinoma using medical image and immunogenomics
| Project/Area Number |
21K07184
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 50020:Tumor diagnostics and therapeutics-related
|
|---|
| Research Institution | Kindai University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
目加田 慶人 中京大学, 工学部, 教授 (00282377)
|
|---|
| Project Period (FY) |
2021-04-01 – 2024-03-31
|
| Keywords | 肝細胞癌 / 薬物療法 / 人工知能 / 医用画像 / 腫瘍免疫微小環境 / 免疫チェックポイント阻害剤 / ドライバー変異 / 遺伝子変異 |
|---|
| Outline of Final Research Achievements |
We developed an AI to predict the efficacy of immune checkpoint inhibitors (ICIs) for hepatocellular carcinoma (HCC). 154 HCC tissues were used to classify tumors based on mutation, histopathology and imaging information. Using scores generated by CD8+ lymphocytes, PD-L1, and β-catenin activation, progression-free survival for PD-1 antibody treatment was successfully stratified in test cohort.Another model was also developed to predict disease control by lenvatinib using imaging information. Radiomics features were extracted from CT, and compared to the LightGBM model, the neural network (NN) model showed better performance with the average correct response rate was 0.58. To further improve accuracy, we used the mutual information as a measure of the relevance of features to create a better model with fewer features. As a result, the GLSZM (Gray Level Size Zone Matrix), a radiomics feature, was particularly useful for discrimination with better performance.
|
| Free Research Field |
消化器内科
|
| Academic Significance and Societal Importance of the Research Achievements |
肝細胞癌（以下肝癌）は本邦のがんの部位別死亡率で5位を占め、その治療法の確立は社会的に重要な問題である。現在、肝癌の1次治療薬として、チロシンキナーゼ阻害剤が2種、複合免疫療法が2種承認されている。しかし、これらの薬物療法においては有効なバイオマーカーがなく、治療薬は医師の経験に基づき選択されている。本研究成果は肝癌の治療における反応性を病理、遺伝子、画像から予測するのものであり、これにより、無駄な治療を避け、患者の生活の質を向上させることが期待できる。さらに治療前にその効果の予測ができれば、肝癌治療の個別化医療の実現に向けた一歩となり得る。
|
