Beyond Lymphovascular Invasion: Site-Dependent Patterns of Lymphatic, Venous and Perineural Involvement in Laryngo-Hypopharyngeal Carcinoma.

BACKGROUND: Building on our recent characterization of a layer-specific neurovascular microenvironment in the normal human vocal fold, we examined how that unit behaves in laryngo-hypopharyngeal carcinoma, where lymphatic and venous involvement are conventionally reported together. METHODS: We retrospectively analyzed 48 laryngo-hypopharyngeal squamous cell carcinomas (2017-2025) with complete TNM vascular reporting, confirmed by CD31 and neuron-specific enolase (NSE) immunohistochemistry. Lymphatic (L) versus venous (V) assignment rested on morphological criteria, since CD31 cannot discriminate lymphatic from blood vessels and D2-40 was unavailable. L, V and perineural (Pn) status were related to pT stage and topography and combined into a composite score (NV = L + V + Pn), reassessed after adjustment for pT. RESULTS: Concordant patterns predominated (L0V0 60.4%, L1V1 27.1%); 12.5% were dissociated. Vascular invasion correlated with pT (ρ = 0.393, p = 0.006) and, unadjusted, with Pn, but not after adjustment for pT. NV tracked pT more strongly than any single axis (KR-20 = 0.725; ρ = 0.534, p < 0.001). Glottic-only carcinoma showed the lowest involvement on all axes. CONCLUSIONS: Reporting L, V and Pn separately-parameters already in routine TNM notation-may capture tumor aggressiveness more completely than LVI as a single entity. This series cannot establish stage-independent co-variation in these axes, validate the L/V distinction molecularly, or, lacking follow-up, demonstrate prognostic value.