ABSTRACT
PURPOSE: To determine the safety and toxicity profile of intravitreal carboplatin as salvage treatment for retinoblastoma with vitreous disease. DESIGN: Single-institution, interventional prospective clinical trial. PARTICIPANTS: Patients with progressive or recurrent vitreous seeds after completion of primary treatment for intraocular retinoblastoma. METHODS: Eligible eyes received an intravitreal injection of carboplatin every 14 to 21 days with simultaneous focal therapy (laser, thermotherapy, and brachytherapy) provided at the discretion of the ocular oncologist. The evaluation with examination under anesthesia, ultrasound biomicroscopy, and electroretinography (ERG) were performed before each injection to assess for tumor response and drug-related toxicity. A serious adverse event resulted in dose recalculation and ultimately early closure of the study. MAIN OUTCOME MEASURES: Regression pattern of vitreous disease and incidence of dose-limiting toxicities. RESULTS: Four patients were enrolled at an initial dose of 0.3 mg. Complete regression of vitreous seeds was noted in all patients after 5, 2, 2, and 1 injections (respectively). Two patients developed recurrent vitreous disease at 3 and 25 months after complete regression and ultimately required enucleation. A serious adverse event occurred in 1 patient who developed acute vision loss with extinguished ERG response 72 hours after the second injection; ultimately, this eye developed a cataract and required enucleation. After temporary suspension and dose modification, 3 patients were enrolled at an injection dose of 3 µg and treated with a total of 5, 2, and 1 injections, respectively. Complete regression of vitreous disease was not achieved in any patient though ERG amplitudes remained stable. After removal from protocol, all 3 patients had a complete response to intravitreal melphalan. Concern for dose escalation and further toxicity in the setting of an effective and safe alternative (melphalan) led to the termination of the study. CONCLUSIONS: Intravitreal carboplatin may be effective in treating progressive vitreous seeding at higher doses, but permanent retinal toxicity was observed. Other alternative agents should be considered. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found after the references.
Subject(s)
Retinal Neoplasms , Retinoblastoma , Humans , Retinoblastoma/drug therapy , Retinal Neoplasms/drug therapy , Carboplatin , Melphalan , Salvage Therapy , Prospective Studies , Vitreous Body/pathologyABSTRACT
BACKGROUND: A total of 5-10% of patients with retinoblastoma (RB) harbor deletion of the long arm (q) chromosome 13 (13q-). The treatment-related toxicities in this population have not been described. METHODS: Sixty-eight RB patients on a single institutional protocol (RET5) from 2005 to 2010 were reviewed. Genetic screening identified 11 patients (seven female) with 13q-. Patients with early (Reese-Ellsworth [R-E] group I-III) disease (6/23 with 13q-) received eight courses of vincristine/carboplatin (VC). Patients with advanced (R-E group IV-V) bilateral disease (2/27 with 13q-) received two courses of vincristine/topotecan (VT) followed by nine courses of alternating VT/VC. Patients undergoing upfront enucleation received histopathology-based chemotherapy: intermediate risk (2/8 with 13q-) or high risk (1/10 with 13q-). Dose reductions were mandated for >7 day delay in two consecutive courses following hematologic toxicity. Grades 3 and 4 hematologic, infectious, and gastrointestinal toxicities were compared between RET5 patients with and without 13q-. RESULTS: Demographics were similar between groups. When present, prolonged neutropenia (median 7 days, range 0-14 days) delayed chemotherapy and resulted in more frequent dose reductions among 13q- patients (5/11) than non-13q- patients (4/57) (P < 0.01). GI toxicity was similar between groups (5/11 13q- vs. 13/57 non-13q-; P = 0.14), but halted chemotherapy in one 13q- patient. Infectious complications and disease outcomes were similar between groups. At follow-up, all patients are alive (median 6.1 years, range 7.6 months-9.5 years). CONCLUSIONS: 13q- RB patients had a higher incidence of neutropenia requiring chemotherapy dose reductions, but did not have increased treatment failure.