HPV-associated macrophage states and expression-based scores in head and neck squamous cell carcinoma
Human papillomavirus (HPV) status defines contrasting immune ecosystems in head and neck squamous cell carcinoma (HNSCC). We investigated how tumor-associated macrophage (TAM) states and their regulatory programs contribute to this diversity and whether these features can be captured in patient-level expression signatures. We integrated 229,196 single cells from 59 patients with HNSCC (26 HPV-positive and 33 HPV-negative), defined eight TAM states, and combined functional and trajectory analyses with regulon inference, virtual perturbation, and CellChat. A 12-gene Weighted Inflammatory Score (WIS) and a TAM-related score (TRS) were evaluated in independent transcriptomic cohorts. Multiplex immunofluorescence provided representative tissue-level observations. HPV-positive tumors showed a lymphoid-enriched landscape, whereas HPV-negative tumors had greater TAM and fibroblast representation. Patient-level comparisons identified lower Inflammatory TAM and Pro-tumor/Secretory TAM fractions in HPV-positive patients (Benjamini–Hochberg [BH]-adjusted p = 0.0172 and 0.0486, respectively). Regulon analysis and virtual SNAI1 perturbation converged on inflammatory, leukocyte migration, and chemotaxis programs. Targeted CellChat analysis identified a candidate epithelial-source HLA-DRA–CD4 interaction involving seven TAM recipient states in HPV-positive tumors compared with two in HPV-negative tumors. Fixed-coefficient WIS achieved an area under the receiver operating characteristic curve (AUC) of 0.938 (95% confidence interval [CI] 0.863–1.000) across 36 independent patients. TRS provided complementary HPV-status discrimination, with a cross-validated AUC of 0.783 in The Cancer Genome Atlas (TCGA) and an external AUC of 0.649 in GSE40774. HPV-associated immune differences in HNSCC involve distinct macrophage states, candidate SNAI1-linked inflammatory and migratory regulatory programs, and inferred epithelial–TAM communication patterns. WIS and TRS connect selected single-cell observations with patient-level transcriptomic features and may guide future mechanistic and validation studies.
Authors
- Yanchen Li
- Zihui Huang
- Anxun Wang
Institutions
- The First Affiliated Hospital, Sun Yat-sen University (CN)
Publication Details
- Journal
- Biomolecules and Biomedicine
- Published
- 2026-11-19
- DOI
- https://doi.org/10.17305/bb.2026.14972
- Primary Topic
- Immune cells in cancer
- Type
- article
- Field-Weighted Citation Impact
- 0.00