Evaluation of corneal subbasal nerve plexus as a predictor of radiosurgery outcome in trigeminal neuralgia.

Gamma Knife radiosurgery (GKRS) is an established treatment for trigeminal neuralgia (TGN); however, no validated biomarkers exist to predict outcomes. In vivo confocal microscopy (IVCM) allows noninvasive quantification of corneal subbasal nerve plexus (SBP) and dendritic cell (DC) density, potentially reflecting trigeminal nerve integrity. This study investigated whether IVCM-derived corneal parameters correlate with and predict pain relief after GKRS in medically refractory TGN. 27 patients with medically refractory idiopathic TGN and 40 healthy controls underwent ophthalmologic examination and IVCM imaging. Pre-treatment parameters included corneal nerve fiber density (CNFD), branch density (CNBD), fiber length (CNFL), total branch density (CTBD), fiber area (CNFA), fiber width (CNFW), fractal dimension (CNFrD), and DC density. GKRS was performed with a single 4-mm collimator shot targeting the retrogasserian portion of the trigeminal nerve (80-90 Gy). Clinical outcomes were assessed using the Barrow Neurological Institute (BNI) Pain Intensity Scale at 12 months. Associations between IVCM parameters and change in BNI score (ΔBNI) were evaluated using correlation and regression analyses, with Benjamini-Hochberg correction for multiple comparisons. A complementary responder-based analysis defined favorable response as BNI I-IIIb at 12 months. Compared with controls, TGN eyes exhibited significantly lower CNFD, CNBD, CNFL, CNFA, and CNFrD (all p < 0.05). Following GKRS, CNFD further decreased (p = 0.034), while DC density increased (p = 0.025). ΔBNI correlated negatively with baseline CNFD (r = -0.526, unadjusted p = 0.003), CNBD (r = -0.500, unadjusted p = 0.006), CNFL (r = -0.436, unadjusted p = 0.018), and CTBD (r = -0.393, unadjusted p = 0.035). After false discovery rate correction, the associations with CNFD, CNBD, and CNFL remained statistically significant. In multivariate analysis, CNFD showed the strongest association with ΔBNI (β = -0.498, p = 0.006; R² = 0.258). Twenty-three patients (85.2%) achieved a favorable response. Baseline ipsilateral CNFD demonstrated moderate but statistically non-significant discrimination for favorable response, with an area under the curve of 0.707 (95% CI, 0.513-0.900; p = 0.195). Lower baseline CNFD was associated with greater pain improvement after GKRS and showed modest discriminatory performance for favorable response. CNFD may represent a candidate noninvasive biomarker of trigeminal nerve dysfunction and radiosurgical response; however, the findings are exploratory and require validation in larger multicenter cohorts.