Background/Objectives: Perineural (PnI), lymphatic (LI), and vascular invasion (VI) in tumor specimens are supposed to worsen the clinical course of oral squamous cell carcinoma (OSCC) and negatively influence survival outcomes. Despite this, these histologic features have not been implemented in the international staging recommendation for OSCC and their prognostic role remains questionable due to inconsistent findings in the related literature. Methods: To investigate the impact of PnI, LI, and VI on oral cancer-specific (OCSS), recurrence-free (RFS), and overall survival (OS), we hypothesized that these histologic features are independent risk factors for poor survival and therefore considered within a prospectively maintained single-center cohort of patients with OSCC. LI and VI were assessed together and reported as lymphovascular invasion (LVI). Results: This study included 439 patients with primary OSCC. Sixty-nine Patients (21.9%) had at least one of the two risk factors. Within the 5-year follow-up period, 61 of these patients (64%) died, and 30 patients (31%) developed locoregional recurrences. Both perineural and lymphovascular invasion were strongly correlated with the presence of lymph node metastasis. PnI and LVI were investigated separately using an adjusted Cox’s proportional hazards regression model. In addition to higher tumor size and the presence of nodal disease (higher stage) the presence of LVI was associated with poor OS, OCSS, and RFS on multivariate analysis, while PnI was associated with reduced OS. In stage III/IV postoperative radiotherapy improved survival in patients with PnI but not with LVI. Conclusions: We conclude that the evidence of LVI in tumor specimens should be considered a high-risk factor when planning adjuvant treatment and monitoring patients with OSCC.

Background: Recurrent oral squamous cell carcinoma (re-OSCC) poses a serious therapeutic challenge and is linked to poor survival outcomes. SOX2 and NANOG, key transcription factors in cancer stem cell biology, may drive tumor progression and therapy resistance. However, their prognostic value in re-OSCC and their relationship to adjuvant therapy remain unclear. Methods: We retrospectively analyzed a single-center cohort of 94 patients with re-OSCC treated with curative intent via (1) surgery alone, (2) surgery plus adjuvant radiotherapy (RT), or (3) surgery plus adjuvant radiochemotherapy (RCT). Tissue microarrays (TMAs) were constructed from matched primary and recurrent tumors and immunohistochemical (IHC) staining for SOX2, and NANOG was quantified using H-scores. Post-recurrence overall survival (prOS) and post-recurrence disease-free survival (prDFS) were evaluated using Kaplan–Meier analysis and Cox proportional hazards models. Results: SOX2 expression and survival: Elevated SOX2 expression (H-score > 14) in re-OSCC was significantly associated with improved prOS (p = 0.013) and prDFS (p = 0.026). Notably, patients who had received adjuvant therapy (particularly RCT) showed higher SOX2 levels in recurrent tumors compared to those treated with surgery alone. NANOG expression and therapy: NANOG expression declined markedly from primary to recurrent tumors (median H-score 42.2 vs. 8.7; p < 0.001). This decline was most pronounced in patients treated with surgery alone. Despite this dynamic change, NANOG expression did not correlate significantly with prOS or prDFS. Other prognostic factors include advanced tumor stage (rT2–rT4) and lymph node involvement (rN+/x)m which remained significant predictors of worse survival in the recurrent setting, regardless of adjuvant therapy. Conclusion: SOX2 overexpression in re-OSCC correlates with better survival, suggesting a unique prognostic role distinct from primary disease. Adjuvant therapy, especially RCT, appears to maintain or elevate SOX2 levels, potentially contributing to improved treatment response. In contrast, although NANOG expression decreases in recurrence, particularly in patients who undergo surgery alone, it does not significantly affect survival outcomes. These findings underscore the importance of context-specific biomarker assessments and provide a rationale for incorporating SOX2 status into personalized treatment strategies for re-OSCC.