Identifying Tumor Microenvironment Biomarkers in Adherent and Cystic Vestibular Schwannomas

Abstract

Objective A subset of vestibular schwannomas (VSs), including cystic tumors, have higher postoperative morbidity because of the presence of adhesions between the tumor, facial nerve (FN), and brainstem. We identify tumor microenvironment (TME) biomarkers to better classify these tumors and predict the degree of tumor adherence. Study Design Retrospective case series. Setting Tertiary skull base referral center. Methods Adult patients with cystic and solid VS matched in tumor size who underwent surgical resection were included. Expressions of seven biomarkers of extracellular matrix remodeling and tumor immune response were quantified via immunohistochemistry. The distribution of CD45+ immune cells was evaluated in intratumoral and perivascular compartments. The degree of tumor adherence was categorized as none, adherent to FN, or adherent to both FN and brainstem. Results Twenty-eight patients were included. Cystic VSs were significantly more adherent than solid VSs (p = 0.02). Patients with adherent VS had shorter duration of symptoms and were more likely to undergo subtotal resection. In solid tumors, matrix metalloproteinase (MMP)-2 expression (p = 0.02) and CD163+ macrophage infiltration (p = 0.007) were correlated with tumor size. Linear discriminant analyses (LDAs) demonstrated MMP-2, MMP-14, CD80, CD163, and perivascular CD45 to be individually predictive of the degree of tumor adherence (all p < 0.05), with perivascular CD45 being the best independent predictor (p = 0.005). An LDA model including these biomarkers demonstrated 100% accurate discrimination of all three levels of tumor adherence (p = 0.04). Conclusions Adherent VS have a distinct proinflammatory TME characterized by elevated MMP expression, enrichment of tumor-associated macrophages, and perivascular immune cell infiltration.