Association Between PET/CT Texture Features and Tumor Immune Microenvironment in Oral Squamous Cell Carcinoma: A Preliminary Study
Background: Noninvasive characterization of the tumor microenvironment may support the development of imaging biomarkers in oral squamous cell carcinoma. This retrospective study investigated associations between PET/CT-derived texture features and immune cell densities. Methods: 73 patients who underwent preoperative PET/CT and surgical resection, 22 were included after excluding those who received preoperative chemotherapy or whose specimens required decalcification. Maximum standardized uptake value (SUVmax) and eight texture features were extracted from PET/CT images. Immune cell populations and programmed death-ligand 1 (PD-L1)-expressing cell subsets were quantified in surgical specimens using multifluorescence immunohistochemistry. Associations were evaluated using Spearman’s rank correlation with false discovery rate correction. Results: Long-run emphasis and gray-level non-uniformity from the gray-level run length matrix, and long-zone emphasis from the gray-level zone length matrix, were positively correlated with PD-L1+CD68+ macrophage density (all adjusted p = 0.048). SUVmax was not significantly associated with any evaluated microenvironmental marker.
Authors
- Susumu Suzuki (ORCID: https://orcid.org/0000-0002-1466-8824)
- Sanako Nakaya (ORCID: https://orcid.org/0009-0007-5329-4400)
- Ikuko Okubo
- Mitsuo Goto (ORCID: https://orcid.org/0000-0001-8019-0438)
- Shoya Ono (ORCID: https://orcid.org/0000-0002-4433-1667)
- Masahiro Umemura
- Masashi Kimura (ORCID: https://orcid.org/0000-0003-3425-5004)
- Y. Kondo (ORCID: https://orcid.org/0000-0002-7558-2944)
Institutions
- Aichi Medical University (JP)
- Aichi Gakuin University (JP)
- Gifu Prefectural Tajimi Hospital (JP)
- Toyokawa City Hospital (JP)
- Ogaki Municipal Hospital (JP)
Publication Details
- Journal
- Clinics and Practice
- Published
- 2026-11-17
- DOI
- https://doi.org/10.3390/clinpract16110215
- Primary Topic
- Radiomics and Machine Learning in Medical Imaging
- Type
- article
- Field-Weighted Citation Impact
- 0.00
Funders
- Japan Society for the Promotion of Science