Optical coherence tomography angiography reveals microvascular changes and predicts visual outcomes after decompression surgery for pituitary adenoma.

Lai, Ivan Pochou; Huang, Ching-Wen; Yang, Shih-Hung; Huang, Abel Po-Hao; Wang, Huan-Chih; Lai, Dar-Ming; Wang, Kuo-Chuan; Chen, Chang-Mu et al. · Br J Ophthalmol · 2026

prospective_cohort · Level II

Where this comes from

Abstract

To evaluate retinal microvascular changes using optical coherence tomography angiography (OCTA) and their prognostic value in patients undergoing transsphenoidal surgery for pituitary adenoma. In this prospective study of 90 patients, we assessed best-corrected visual acuity (BCVA), mean deviation (MD) on visual field (VF), and OCTA parameters at the macular region and peripapillary area. Patients with severe VF defects (VFD; MD <-10 dB) showed greater reductions in vessel densities of superficial capillary plexus (SCP) and radial peripapillary capillaries (RPC) than those with minimal VFD (MD >-3 dB) (p=0.021, 0.047). Higher SCP density, RPC density, peripapillary retinal nerve fibre layer (pRNFL) thickness and ganglion cell complex (GCC) thickness were associated with better postoperative visual acuity (r=-0.33, -0.44, -0.24, -0.30; all p<0.05) and MD (r=0.26, 0.46, 0.45, 0.45; all p<0.05). In patients with mild to moderate VFD (MD between -3 and -10 dB), foveal avascular zone area, foveal SCP and deep capillary plexus densities correlated strongly with BCVA outcome (r=0.61,-0.60, -0.63; all p<0.01). Area under the curves for predicting >50% MD improvement were 0.816 (SCP density), 0.784 (RPC density), 0.887 (pRNFL thickness), 0.816 (GCC thickness) and 0.905 (combined; all p≤0.001). After decompression surgery, retinal vessel densities may further decline at 3 months postoperatively, particularly in patients with severe baseline VFD. Foveal parameters predicted visual acuity in those with mild to moderate VFD, while in severe cases, higher SCP density, RPC density, pRNFL thickness and GCC thickness may serve as prognostic biomarkers for predicting better VF improvement.