CT Texture Patterns Reflect HPV Status but Not Histological Differentiation in Oropharyngeal Squamous Cell Carcinoma

BACKGROUND: Texture analysis (TA) has shown promise in characterizing intratumoral heterogeneity from imaging data. We add to the literature that shows its capability to differentiate oropharyngeal cancers based on HPV status. METHODS: Multislice CT analysis was done in 120 patients with confirmed OP SCC: a single 5 mm region of interest was placed on three consecutive homogeneous CT slices per patient. Texture features were extracted by using gray-level co-occurrence matrices averaged per patient. HPV status (via p16 IHC and molecular confirmation) and differentiation grade (i.e., good, moderate, and poor) were recorded. Non-parametric statistical tests assessed differences between subgroups. RESULTS: Seven texture parameters (i.e., angular second moment, contrast, sum of squares, sum entropy, entropy, inverse difference moment, and difference variance) differed significantly between HPV+ and HPV- tumors (all <i>p</i> < 0.05). HPV+ tumors exhibited increased heterogeneity and complexity on CT imaging. No texture feature correlated with histological grade. CONCLUSIONS: This study adds to the growing evidence that CT-based TA can assess HPV status in OP SCC. TA may be promising, though it requires further validation as an adjunctive method integrating into radiomics workflows to develop predictive models for diagnosis, prognosis, and treatment planning.