MRI Evaluation of Optic Chiasm Compression and Its Relationship with Visual Field Defects in Pituitary Macroadenoma

Abstract

Background: Superior extension can be observed in macroadenomas, which compress the optic chiasm and cause visual field loss, often showing up as clinically silent until late. Estimating the correlation between magnetic resonance imaging (MRI) parameters and visual field defects may facilitate a prompt multidisciplinary consultation and decompression.  Aim: To investigate the optic chiasm compression related MRI findings and their correlation with visual field defects in patients with pituitary macroadenoma in Azadi Teaching Hospital, Kirkuk.  Methods: This retrospective cross-sectional analysis comprised 40 consecutive adults with pituitary macroadenoma, diagnosed by MRI, assessed from February 10, 2025 to March 20, 2026. Size of the tumor, vol-ume, suprasellar extension, grade of chiasm displacement/compression, cavernous sinus inva-sion, hemorrhage and optic pathway signal abnormality were evaluated by dedicated sellar MRI. Subtype and mean deviation (MD) of automated static perimetry were analysed. We analysed the associations using chi-square/Fisher exact tests, t tests or non-parametric equivalents, Spearman correlation, and multivariable logistic regression. Results: The mean age was 49.8 ± 12.4 years, and 22 (55.00%) patients were male. Visual field defects were identified in 29 of 40 patients (72.50%), with bitemporal hemianopia being the most common pattern, observed in 19 (47.50%). MRI showed mild optic chiasm compression in 11 (27.50%) patients, for this was moderate in 13 (32.50%), severe in 9 (22.50%) and 7 (17.50%) had chiasmal contact without apparent deformation. The frequency of visual field abnormalities show a progressive increase in severity of these groups, 28.57% of patients with contact only, up to 63.64%, 84.62%, and 100.00% of contacts of contacts with mild, moderate, and severe compression pattern (p for trend=0.002). Patients with defects had significantly larger craniocaudal tumor diameter on average than those without VF defects (29.8 ± 6.1 vs 20.9 ± 4.8 mm; p<0.001), suprasellar extension diameter (12.6 ± 4.7 vs 5.1 ± 3.0 mm; p<0.001), and median tumor volume (9.8 cm³ [interquartile range: 6.8–14.7] vs 4.2 cm³ [interquartile range: 2.8–6.0]; p<0.001). Higher grades of optic chiasm compression was highly correlated with more negative binocular MD (Spearman’s ρ=−0.72; p<0.001) for visual field loss. After controlling for age, symptom duration, and craniocaudal tumor diameter, moderate-to-severe optic chiasm compression was still independently predictive of the existence of a VF defect (adjusted OR=8.31; 95% CI: 1.35–51.20; p=0.022). Conclusion: Among them, severity of optic chiasm compression shown by MRI, vertical tumor size and suprasellar extension were most predictive for visual field defect in this study. Results of dedicated sellar MRI should be interpreted in conjunction with formal perimetric evaluation, even in patients with relatively intact central visual acuity.