Akademik Veri Yönetim Sistemi
Endocrine, cilt.91, sa.1, 2026 (SCI-Expanded, Scopus)
Purpose: This study aimed to investigate the association between histopathological proliferation markers and microscopic sellar floor invasion in pituitary adenomas treated via endoscopic endonasal transsphenoidal surgery. Specifically, we assessed the predictive value of the Ki-67 proliferation index, Modified Hardy’s Classification, and tumor phenotypes in identifying microscopic invasion. Materials and methods: Sixty patients who underwent endoscopic endonasal transsphenoidal surgery between January 2018 and January 2023 were retrospectively analyzed. Data included Ki-67 index, p53 status, ATRX expression, hormonal subtype, MHC grade, and histopathological confirmation of sellar floor invasion. Receiver Operating Characteristic (ROC) curve analysis and multivariate logistic regression were used to determine independent predictors. Results: Microscopic sellar floor invasion was present in 46.7% of cases. The mean Ki-67 index was 4.2%. Invasion was most common in somatotroph adenomas (72.7%). Multivariate analysis identified Ki-67 > 3%, MHC grade > 2, and somatotroph phenotype as independent predictors. Ki-67 and MHC demonstrated moderate diagnostic accuracy (AUC:0.73 and 0.70, respectively), with Ki-67 showing superior sensitivity (73%) and specificity (69%). Importantly, tumors with low MHC grades exhibited invasion when Ki-67 was elevated. No significant associations were found for p53 nuclear staining, ATRX loss of expression, patient age, or gender. Discussion: Ki-67 index > 3%, MHC grade > 2, and somatotroph phenotype were independently associated with histological sellar floor invasion. Ki-67 outperformed MHC in predictive strength and diagnostic reliability. The elevated Ki-67 underscores the value of molecular assessment even in radiologically non-invasive cases. No significant associations were observed for p53, ATRX, age, or gender. To our knowledge, this study uniquely integrates radiological classification and histopathological confirmation with multivariate modeling and ROC-based thresholds.