Invasiveness, SOX2 Expression, and Resection Outcomes in TF-Defined PitNETs After Extra-Pseudocapsular Resection
摘要
To evaluate the associations between transcription factor (TF)-defined molecular lineages and the clinical characteristics of pituitary neuroendocrine tumors (PitNETs).
MethodsA retrospective cohort analysis was performed in 274 patients who underwent extra-pseudocapsular transsphenoidal resection. Tumors were classified by TF-defined lineage and invasion grade (0–3) and Knosp score. Group differences were tested, and predictors of gross total resection (GTR) were examined using hierarchical multivariable logistic regression.
ResultsTF-defined lineage classification improved diagnostic precision. Among tumors with an immunonegative hormone profile, 74.7% received a definitive lineage assignment. The pituitary transcription factor 1–growth hormone/prolactin (PIT1–GH/PRL) subgroup had the highest incidence and grade of invasion; SRY-box transcription factor 2 (SOX2) positivity was most frequent in this subgroup (41.7%), which also exhibited the lowest GTR rate (58.3%). SOX2-positive tumors were associated with significantly higher preoperative adrenocorticotropic hormone (ACTH) (median: 38.80 vs. 31.60 pg/mL, P = 0.032) and GH (median: 1.12 vs. 0.43 ng/mL, P = 0.046) levels. SOX2 positivity was 20.5% in multilineage tumors and 20.0% in the recurrent cohort (n = 20). Compared with Trouillas grade alone, the combination of invasion grade and intraoperative features (capsule status and vascularity) significantly improved the ability to predict GTR (AUC, 0.866 vs. 0.795; ΔAUC = 0.071; DeLong 95% CI, 0.022–0.120; P = 0.005).
ConclusionsThe 2022 WHO TF-defined lineage system improves diagnostic precision and facilitates the interpretation of PitNET differentiation and molecular pathology. Tumors in the PIT1–GH/PRL subgroup demonstrate more aggressive invasive features. SOX2 positivity was relatively frequent in multilineage tumors, whereas the observation in recurrent cases was limited by the small exploratory cohort and requires validation in larger longitudinal studies.