Abstract
Ameloblastoma is a histologically benign odontogenic epithelial tumor with locally destructive growth and a clinically relevant risk of recurrence. Established molecular research has been largely focused on MAPK-pathway alterations, whereas markers capturing other dimensions of tumor biology remain limited. Preferentially expressed antigen in melanoma (PRAME) is a tumor-associated antigen implicated in retinoic-acid receptor signaling, altered differentiation, and antitumor immune recognition, but dedicated data in ameloblastoma remain limited. In this retrospective cohort, 37 histologically confirmed ameloblastomas were evaluated immunohistochemically for PRAME on tissue microarray sections by two independent pathologists. A four-tier extent score (0, 1+, 2+, 3+) and a four-tier intensity score (absent, weak, moderate, strong) were recorded. PRAME positivity was defined as nuclear staining in ≥25% of tumor cells for exploratory association analyses. Clinicopathologic variables included age, sex, anatomic site, side, mandibular subsite, histologic type, surgical approach, recurrence, and follow-up. Interobserver reproducibility was assessed using percentage agreement and Cohen’s κ statistics. Holm adjustment was applied as a post hoc sensitivity analysis across eight exploratory clinicopathological comparisons. PRAME was positive in 30 of 37 tumors (81.1%; exact 95% CI 64.8–92.0). Twenty-nine tumors (78.4%) received a 3+ extent score from both observers, seven (18.9%) were negative, and the remaining tumor was scored 1+ by one observer and 2+ by the other. Extent scoring agreed in 36 of 37 cases (97.3% agreement; Cohen kappa = 0.923). Intensity scoring agreed in 35 of 37 cases (94.6%); unweighted kappa was 0.905, with linear and quadratic weighted kappa values of 0.927 and 0.949. PRAME positivity was more frequent in women than in men (95.2% vs. 62.5%; unadjusted two-sided Fisher exact p = 0.029; Holm-adjusted p = 0.230). All 11 patients with documented recurrence had PRAME-positive tumors, compared with 19 of 26 patients without documented recurrence (unadjusted p = 0.080). PRAME positivity was also more frequent among conservatively than radically treated patients (91.7% vs. 61.5%; unadjusted p = 0.072). None of the eight clinicopathological comparisons met the 0.05 threshold after Holm adjustment. PRAME is frequently and extensively expressed in ameloblastoma and can be assessed with excellent interobserver reproducibility. The recurrence and treatment patterns are hypothesis-generating, but the small cohort, sparse PRAME-negative subgroup, and treatment imbalance preclude claims of independent prognostic value. Larger multicenter studies with standardized scoring, molecular validation, and time-to-event analysis are warranted.