1
|
Adachi M, Taki T, Sakamoto N, Kojima M, Hirao A, Matsuura K, Hayashi R, Tabuchi K, Ishikawa S, Ishii G, Sakashita S. Extracting interpretable features for pathologists using weakly supervised learning to predict p16 expression in oropharyngeal cancer. Sci Rep 2024; 14:4506. [PMID: 38402356 PMCID: PMC10894206 DOI: 10.1038/s41598-024-55288-y] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/04/2023] [Accepted: 02/22/2024] [Indexed: 02/26/2024] Open
Abstract
One drawback of existing artificial intelligence (AI)-based histopathological prediction models is the lack of interpretability. The objective of this study is to extract p16-positive oropharyngeal squamous cell carcinoma (OPSCC) features in a form that can be interpreted by pathologists using AI model. We constructed a model for predicting p16 expression using a dataset of whole-slide images from 114 OPSCC biopsy cases. We used the clustering-constrained attention-based multiple-instance learning (CLAM) model, a weakly supervised learning approach. To improve performance, we incorporated tumor annotation into the model (Annot-CLAM) and achieved the mean area under the receiver operating characteristic curve of 0.905. Utilizing the image patches on which the model focused, we examined the features of model interest via histopathologic morphological analysis and cycle-consistent adversarial network (CycleGAN) image translation. The histopathologic morphological analysis evaluated the histopathological characteristics of image patches, revealing significant differences in the numbers of nuclei, the perimeters of the nuclei, and the intercellular bridges between p16-negative and p16-positive image patches. By using the CycleGAN-converted images, we confirmed that the sizes and densities of nuclei are significantly converted. This novel approach improves interpretability in histopathological morphology-based AI models and contributes to the advancement of clinically valuable histopathological morphological features.
Collapse
Affiliation(s)
- Masahiro Adachi
- Department of Pathology and Clinical Laboratories, National Cancer Center Hospital East, Kashiwa, Japan
- Department of Otolaryngology, Head and Neck Surgery, University of Tsukuba, Tsukuba, Japan
| | - Tetsuro Taki
- Department of Pathology and Clinical Laboratories, National Cancer Center Hospital East, Kashiwa, Japan
| | - Naoya Sakamoto
- Department of Pathology and Clinical Laboratories, National Cancer Center Hospital East, Kashiwa, Japan
- Division of Pathology, National Cancer Center Exploratory Oncology Research and Clinical Trial Center, 6-5-1, Kashiwanoha, Kashiwa, Chiba, 277-8577, Japan
| | - Motohiro Kojima
- Department of Pathology and Clinical Laboratories, National Cancer Center Hospital East, Kashiwa, Japan
- Division of Pathology, National Cancer Center Exploratory Oncology Research and Clinical Trial Center, 6-5-1, Kashiwanoha, Kashiwa, Chiba, 277-8577, Japan
| | - Akihiko Hirao
- Division of Pathology, National Cancer Center Exploratory Oncology Research and Clinical Trial Center, 6-5-1, Kashiwanoha, Kashiwa, Chiba, 277-8577, Japan
| | - Kazuto Matsuura
- Department of Head and Neck Surgery, National Cancer Center Hospital East, Kashiwa, Japan
| | - Ryuichi Hayashi
- Department of Head and Neck Surgery, National Cancer Center Hospital East, Kashiwa, Japan
| | - Keiji Tabuchi
- Department of Otolaryngology, Head and Neck Surgery, University of Tsukuba, Tsukuba, Japan
| | - Shumpei Ishikawa
- Division of Pathology, National Cancer Center Exploratory Oncology Research and Clinical Trial Center, 6-5-1, Kashiwanoha, Kashiwa, Chiba, 277-8577, Japan
- Department of Preventive Medicine, Graduate School of Medicine, The University of Tokyo, Tokyo, Japan
| | - Genichiro Ishii
- Department of Pathology and Clinical Laboratories, National Cancer Center Hospital East, Kashiwa, Japan
- Division of Innovative Pathology and Laboratory Medicine, National Cancer Center Exploratory Oncology Research and Clinical Trial Center, Kashiwa, Japan
| | - Shingo Sakashita
- Department of Pathology and Clinical Laboratories, National Cancer Center Hospital East, Kashiwa, Japan.
- Division of Pathology, National Cancer Center Exploratory Oncology Research and Clinical Trial Center, 6-5-1, Kashiwanoha, Kashiwa, Chiba, 277-8577, Japan.
| |
Collapse
|
2
|
Etzel BM, Gerth M, Chen Y, Wünsche E, Facklam T, Beck JF, Guntinas-Lichius O, Petersen I. Mutation analysis of tumor necrosis factor alpha-induced protein 3 gene in Hodgkin lymphoma. Pathol Res Pract 2016; 213:256-260. [PMID: 28189285 DOI: 10.1016/j.prp.2016.11.001] [Citation(s) in RCA: 9] [Impact Index Per Article: 1.1] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 09/20/2016] [Revised: 11/01/2016] [Accepted: 11/03/2016] [Indexed: 12/01/2022]
Abstract
AIMS Survival and proliferation of Hodgkin and Reed-Sternberg (HRS) cells, the malignant cells of classical Hodgkin lymphoma (CHL), are dependent on constitutive activation of nuclear factor kB (NF-κB). A20, encoded by TNF alpha-induced protein 3 (TNFAIP3), one of the inhibitors of NF-kB, was found to be inactivated by deletions and/or point mutations in CHL. METHODS TNFAIP3 mutations were examined in 37 patients with CHL by using PCR and direct sequencing. In addition, protein expression of A20 was evaluated by immunohistochemistry. Epstein-Barr virus (EBV) status of HL samples was determined by EBV EBER chromogenic in situ hybridization (ISH). RESULTS We identified 8 mutation positive cases in a collective of 37 investigated cases (22%). Mutations were most frequent in the nodular sclerosis subtype. Our results revealed the tendency that cases harboring A20 mutations were negative for A20 staining. None of A20 mutation-positive CHL cases showed EBV infection. CONCLUSIONS Our study confirms the involvement of the TNFAIP3 tumor suppressor gene in CHL. A20 may represent a suppressor of human lymphoma and provide a critical molecular link between chronic inflammation and cancer. None of A20 mutation-positive CHL cases showed EBV infection. This fact suggests complementing functions of TNFAIP3 inactivation and EBV infection in CHL pathogenesis and may represent an interesting point of further investigations.
Collapse
Affiliation(s)
| | - Melanie Gerth
- Institute of Pathology, Jena University Hospital, Jena, Germany
| | - Yuan Chen
- Institute of Pathology, Jena University Hospital, Jena, Germany
| | - Elisa Wünsche
- Institute of Pathology, Jena University Hospital, Jena, Germany
| | - Tina Facklam
- Institute of Pathology, Jena University Hospital, Jena, Germany
| | - James F Beck
- Children's Clinic, Department of Pediatric Hematology and Oncology, Jena University Hospital, Jena, Germany
| | | | - Iver Petersen
- Institute of Pathology, Jena University Hospital, Jena, Germany.
| |
Collapse
|