Metavariables Resuming Host Immune Features and Nodal Involvement Are Associated with Oncological Outcomes in Oral Cavity Squamous Cell Carcinoma

Abstract

Oral cavity squamous cell carcinoma (OSCC) is a common head and neck cancer characterized by a poor prognosis associated with locoregional or distant failure. Among the predictors of prognosis, a dense infiltration of adaptive immune cells is protective and associated with improved clinical outcomes. However, few tools are available to integrate immune contexture variables into clinical settings. By using digital microscopy analysis of a large retrospective OSCC cohort (n = 182), we explored the clinical significance of tumor-infiltrating CD8+ T-cells. To this end, CD8+ T-cells counts were combined with well-established clinical variables and peripheral blood immune cell parameters. Through variable clustering, five metavariables (MV) were obtained and included descriptors of nodal (NODALMV) and primary tumor (TUMORMV) involvement, the frequency of myeloid (MYELOIDMV) or lymphoid (LYMPHOIDMV) peripheral blood immune cell populations, and the density of tumor-infiltrating CD8+ T-cells (TI-CD8MV). The clinical relevance of the MV was evaluated in the multivariable survival models. The NODALMV was significantly associated with all tested outcomes (p MV showed a significant association with the overall, disease-specific and distant recurrence-free survival (p MV with the locoregional control only (p MV was associated with distant recurrence-free survival (p = 0.029). Notably, the performance in terms of survival prediction of the combined effect of NODALMV and immune metavariables (LYMPHOIDMV, MYELOIDMV and TI-CD8MV) was superior to the TNM stage for most of the outcomes analyzed. These findings indicate that the analysis of the baseline host immune features are promising tools to complement clinical features, in stratifying the risk of recurrences.

Keywords

• oral neoplasms
• lymphocyte
• CD8
• biomarker
• survival modeling
• data reduction