Optic Nerve Infiltration in Retinoblastoma: Correlation of Histopathological and MRI Findings in Enucleated Eyes

Background and objective The accurate assessment of optic nerve invasion in retinoblastoma is critical for prognosis and treatment planning. Although MRI is routinely used for preoperative evaluation, it may fail to detect early microscopic invasion. Ensuring appropriate imaging interpretation and obtaining a sufficiently long optic nerve stump during enucleation are essential to avoid unnecessary morbidity related to adjuvant therapies. This study aimed to compare MRI findings with histopathological evaluation regarding optic nerve infiltration in enucleated eyes with group E retinoblastoma. Materials and methods This cross-sectional study included 45 patients with group E retinoblastoma who underwent upfront enucleation. Preoperative 1.5-T MRI was interpreted by an experienced neuroradiologist, and histopathology was performed according to the International Retinoblastoma Staging Working Group (IRSWG) guidelines. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and diagnostic accuracy of MRI for optic nerve invasion were calculated using histopathology as the reference standard. Results Histopathology detected post-laminar optic nerve invasion in 19 cases (42.2%), while MRI detected it in 17 cases (37.7%). Prelaminar invasion was identified in 12 eyes (26.7%) histologically, but was underestimated by MRI, which detected only five cases (11.1%). MRI showed a sensitivity of 76.0%, specificity of 72.7%, PPV of 80.9%, NPV of 66.7%, and overall accuracy of 75.0%. Conclusions MRI is a valuable noninvasive modality for assessing tumor extent and detecting post-laminar invasion; however, it consistently underestimates early (prelaminar) and microscopic optic nerve infiltration. Histopathology remains the gold standard for definitive evaluation and risk stratification. Regardless of MRI findings, obtaining an adequately long optic nerve stump during enucleation is crucial, as radiologic imaging may fail to detect microscopic invasion, and the cut-end status remains the most important prognostic determinant.