2026 Poster Presentations
P278: NEAR UNIVERSAL VISION IMPROVEMENT FOLLOWING EEA FOR CRANIOPHARYNGIOMA IRRESPECTIVE OF OPTIC PATHWAY INVOLVEMENT
Brock Gjesdal; Sharath Kumar Anand, MD; Robert Dambrino, MD; Anthony Tang; Garrett Choby; Eric Wang, MD; Susan Stefko, MD; Carl Snyderman, MD, MBA; Georgios Zenonos, MD; Paul Gardner, MD; UPMC
Introduction: Visual dysfunction represents a major presenting symptom in craniopharyngioma patients, with tumors frequently compressing or invading the optic apparatus. While endoscopic endonasal approaches (EEA) offer direct visualization and decompression of visual structures, the prognostic value of preoperative anatomic involvement patterns is ill-described. Specifically, whether differential involvement of the optic chiasm, optic tract, or combined structures as determined by preoperative MRI impacts visual recovery in patients undergoing EEA for tumor resection is still being established as historical case volume grows. We investigated whether specific optic pathway involvement patterns serve as prognostic indicators for visual improvement following EEA for craniopharyngioma.
Methods: This is a single-institution retrospective cohort study of consecutive adult and pediatric patients who underwent EEA for craniopharyngioma resection from 2008 to 2024. Preoperative visual disturbance was defined as decreased acuity or field defects identified on formal ophthalmologic evaluation. Postoperative visual outcomes were defined based on either subjective patient reported outcome (normalized, partially improved, unchanged or worsened), or on formal postoperative ophthalmologic evaluation. Patients with or without visual disturbance were compared for baseline characteristics. Preoperative MRIs were reviewed, and involvement of the optic apparatus was categorized as optic chiasm alone, optic tract alone, involvement of both,or neither. For anatomic predictor analyses, patients with both preoperative MRI data (location determined by pre-operative MRI sequence) and documented visual outcomes were compared using Kruskal-Wallis and chi-squared tests as appropriate. Other variables tested included age, tumor volume, extent of resection (GTR vs STR+NTR), symptom duration, optic pathway involvement pattern, radiographic hypothalamic invasion, and third ventricle involvement.
Results: Of 114 patients, 84 (73.7%) presented with preoperative visual disturbance and 75 had documented post-operative subjective visual outcomes data at time of first follow-up. Patients presenting with visual disturbance were significantly older (44.9 vs 28.7 years, p=0.005) and more likely to have optic chiasm involvement (86.7% vs 57.9%, p=0.012). 53 (70.7%) patients had normalization of vision, 10 (13.3%) had partial improvement, 8 (10.7%) remained unchanged, and 4 (5.3%) worsened, yielding an overall improvement rate of 84%. In patients with MRI location data, improvement rates remained consistently high across optic pathway patterns: chiasm-only 93.8%, neither chiasm nor tract 83.3%, combined involvement 82.5%, and tract-only 66.7% (p=0.662). There were no factors that independently predicted improved visual outcomes. Among 3 patients who had worsened postoperative vision for whom we had complete MRI data, all had both optic chiasm and tract involvement as well as third ventricular invasion; 2/3 (66.7%) had hypothalamic invasion. 75% (3/4) of these patients had recurrent tumors (p=0.284) with median tumor volume size of 14.8 cm³ (IQR 9.3-16.8, p=0.093).
Conclusions: Visual symptoms significantly improved following EEA for craniopharyngioma resection irrespective of preoperative optic pathway involvement patterns. This study further underscores the benefit of direct visualization of the entire optic apparatus afforded by EEA for craniopharyngioma resection. These findings support aggressive visual decompression for all craniopharyngioma patients with preoperative deficits and enable confident counseling regarding visual prognosis.