Background: Optic nerve invasion is one of the predictors for retinoblastoma metastases. This study was designed to investigate the risk of optic nerve invasion and clinical features that may identify those children with optic nerve invasion.
Methods: We reviewed the charts of 289 children with retinoblastoma treated initially with enucleation. Logistic regression analysis was performed to assess the risk for metastases from varying degrees of optic nerve invasion and to assess the clinical and histopathologic predictors of optic nerve invasion.
Results: There were 84 eyes (29%) with optic nerve invasion. The invasion was prelamina cribrosa in 44 cases (15%), up to but not posterior to the lamina cribrosa in 21 cases (7%), posterior to the lamina cribrosa but not to the cut end of the optic nerve in 17 cases (6%), and to the site of optic nerve transection in 2 cases (1%). Patients with optic nerve invasion were more likely to develop metastasis (P = 0.0016), particularly those with invasion to the postlaminar and cut section of the optic nerve (P = 0.0001). Development of metastasis was not statistically associated with laminar or prelaminar involvement. If those patients with choroidal invasion simultaneous with optic nerve invasion were excluded from evaluation, the presence of optic nerve invasion alone was not significant for development of metastasis. The clinical factors found to be predictive for optic nerve invasion from a univariate analysis included exophytic growth pattern (P = 0.011), elevated intraocular pressure (> 22 mm Hg) (P = 0.02), and tumor thickness greater than or equal to 15 mm (P = 0.03). The histopathologic factor significantly associated with optic nerve invasion (univariate analysis) was simultaneous choroidal invasion (P = 0.001). A trend toward an association with optic nerve invasion was found with vitreous hemorrhage (P = 0.06), iris neovascularization (P = 0.10), and poorly differentiated retinoblastoma (P = 0.07). A multivariate analysis showed the most significant clinical factors to be exophytic growth pattern (P = 0.002), tumor thickness greater than or equal to 15 mm (P = 0.01), and vitreous hemorrhage (P = 0.05).
Conclusions: Optic nerve invasion of retinoblastoma beyond the lamina cribrosa is associated with a greater metastatic risk. Large exophytic retinoblastoma with secondary glaucoma is at highest risk for optic nerve invasion.