Optic Nerve Invasion in Retinoblastoma: Impact of Eye Salvage and Adjuvant Chemotherapy

Abstract

Purpose: Optic nerve invasion (ONI) is an important risk factor for extraocular metastasis in retinoblastoma. This study evaluates the impact of pre-enucleation chemotherapy, delayed enucleation, and adjuvant chemotherapy on survival in patients with varying degree of ONI on histopathology.

Methods: A retrospective review of consecutive enucleated eyes with any degree of ONI from 29 Chinese treatment centers 2012-2017. Children with other high-risk histopathological features, extraocular disease at diagnosis or bilateral enucleation were excluded.

Results: Among 2500 enucleated eyes, 386 eyes with isolated ONI (one eye per child) met the inclusion criteria: prelaminar (n = 204), intralaminar (n = 70), retrolaminar without tumor at transected optic nerve (n = 96), or retrolaminar with tumor at transected end (n = 16). Primary enucleation was performed in 59% of cases, whereas 41% underwent secondary enucleation after eye salvage therapies. Delayed enucleation beyond six months from diagnosis significantly increased the likelihood of tumor at optic nerve transection (21% vs. 2%; P < 0.001). The five-year cause-specific survival (CSS) was 96.7% overall: prelaminar (100%), intralaminar (98%), retrolaminar without transected end involvement (94%), or tumor at transected optic nerve (58%). Secondarily enucleated eyes with retrolaminar ONI without transected end involvement had lower CSS than those primarily enucleated (85.6% vs. 100%; P = 0.022). The five-year CSS was higher but not statistically significant, for eyes treated with or without adjuvant chemotherapy: intralaminar ONI (100% vs. 95.7%; P = 0.193), retrolaminar ONI without tumor at transected end (100% vs. 93.5; P = 0.413), and tumor at the transected end (80.0% vs. 45.0%; P = 0.192).

Conclusions: Timely enucleation reduces the risk of tumor involvement at optic nerve transection. Delay in enucleation by pre-enucleation chemotherapy may reduce the effectiveness of subsequent adjuvant chemotherapy.

Figure 1.

Boxplot showing time from diagnosis to enucleation (months) based on the degree of ONI.

Figure 2.

Kaplan-Meier curves of CSS of children with various degree of ONI.

Figure 3.

Kaplan-Meier curves of CSS of children with primary and secondary enucleation. (A) Intralaminar ONI. (B) Retrolaminar optic nerve invasion without tumor at transected end. Enu, enucleation. *P < 0.05.