ABSTRACT
PURPOSE: Identifying relationships between craniopharyngiomas (CPs) and contiguous structures, and tumor origin are crucial for treatments. This study attempted to explore the relationships and tumor origin. METHODS: CPs that underwent endoscopic surgeries were enrolled. The interfacial specimens of CPs attaching the hypothalamus, pituitary stalk (PS), pituitary grand (PG), optic chiasma (OC) and brain tissue (BT) were pathologically examined. Boundaries between CPs and these structures were observed during operations. Expression of ß-catenin and stem cell markers were analyzed to explore the tumor origin. Outcomes of patients were assessed. RESULTS: A total of 34 CPs were categorized into two groups based on the locations of finger-like protrusions (FP). Group A comprised 18 CPs with FP only present in the specimens attaching to hypothalamus. The surface of these CPs was fused with hypothalamus under endoscopic videos. However, the specimens attaching to the PS, PG, OC, and BT showed no FP. Clear boundaries was observed between these CPs and these structures. Group B comprised 16 CPs with FP only present in the specimens attaching to PS. The tumor surface was fused with PS. Specimens attaching to the hypothalamus, PG, OC and BT showed no FP. Clear boundary was observed among these CPs with these structures. These results implied CPs only invaded a certain part of hypothalamic-pituitary axis. ß-catenin and stem cells markers mainly distributed in the FP tissues of both groups. Patients in group B achieved better outcomes than group A. CONCLUSIONS: CPs only invade the hypothalamic-pituitary axis with FP and the FP would be the tumor origin.
Subject(s)
Craniopharyngioma , Pituitary Neoplasms , Humans , Hypothalamus , Pituitary Gland , beta CateninABSTRACT
PURPOSE: Accurate prediction of topographical correlation between craniopharyngiomas (CPs) and hypothalamus is important for treatment. This study sought to develop a predicting tool based on preoperative-MRI through radiological-surgical-pathological-outcome analysis. MATERIALS AND METHODS: Third ventricle floor (TVF), mammillary bodies and cerebral peduncle were evaluated through preoperative-MRI. An eagle-head-like sign named "eagle sign" was observed. Normal TVF on sagittal-MRI was defined as the baseline. Variants of the sign were analyzed by comparing with the baseline and corresponding correlations of CPs with hypothalamus were verified using intraoperative records, histopathology and outcome evaluation. RESULTS: A total of 146 CPs patients, who undergone endoscopic endonasal procedure were divided into four groups based on the variants of "eagle sign". Group A: 24 patients with the upward sign; group B: 81 with the downward sign; group C: 21 with the anterior TVF upward sign and group D: 20 with the unidentifiable sign. Surgical-pathological analysis showed significant correlations between 95.8% CPs in group A and 95.2% in group C with tumor topography and tumor adherence to the hypothalamus. These CPs had their origins beneath the hypothalamus. In contrast, groups B and D, with hypothalamic origin, showed hypothalamic infiltration by tumor in 97.5% and 95% of cases in groups B and D, respectively. Outcomes of groups A and C were relatively better than groups B and D. Predictive sensitivity and specificity of "eagle sign" were more than 90%. CONCLUSION: "Eagle sign" is an accurate tool for predicting topographic correlations between CPs and hypothalamus with high sensitivity and specificity.