Efficacy and toxicity of second-course ophthalmic artery chemosurgery for retinoblastoma.

OBJECTIVE

Assess the usefulness of second-course ophthalmic artery chemosurgery (OAC) for patients with intraocular retinoblastoma that recurred after prior OAC. This study evaluated the efficacy and toxicity of second-course OAC.

DESIGN

Single-arm retrospective study of 29 eyes of 30 patients treated with second-course OAC at Memorial Sloan Kettering Cancer Center between May 2006 and July 2013, with a median follow-up of 25.9 months.

PARTICIPANTS

Retinoblastoma patients who underwent a course of OAC, with a minimum of 2 months of progression-free follow-up at monthly examinations, but who subsequently received additional OAC for recurrent tumor.

METHODS

To determine efficacy, Kaplan-Meier survival estimates were generated and the Mantel-Cox test was used to compare curves. To determine toxicity, electroretinography (ERG) amplitudes were measured in response to 30-Hz photopic flicker stimulation before and after OAC treatment; systemic adverse events were graded according to the Common Terminology Criteria for Adverse Events version 4.0 (CTCAE 4.0).

MAIN OUTCOME MEASURES

For efficacy, ocular progression-free survival, ocular event-free survival (e.g., enucleation, external-beam radiation, or intravitreal melphalan), and ocular survival. For toxicity, peak-to-peak comparisons between ERG studies before and after OAC treatment and CTCAE 4.0-graded systemic adverse events.

RESULTS

Fifty percent of all recurrences were within 4.4 months and 90% were within 16 months of completion of the first course of OAC. The 2-year Kaplan-Meier ocular survival, event-free survival, and progression-free survival estimates after second-course OAC were 82.8% (95% confidence interval [CI], 60.1%-93.2%), 57.3% (95% CI, 36.1%-73.7%), and 26.5% (95% CI, 11.0%-45.0%), respectively. All eyes without vitreous seeding were progression free, whereas eyes with vitreous seeding were associated significantly with worse ocular survival after second-course OAC (P = 0.03). After second-course OAC, 90% of eyes had stable or improved ERG responses. Of all evaluable cases, there was no increased risk of systemic toxicity during the second course compared with the initial course of OAC.

CONCLUSIONS

Retinoblastoma eyes requiring second-course OAC after initial OAC treatment have good salvage rates, and the treatment has an acceptable ocular and systemic toxicity profile. However, these eyes often require additional (third- or fourth-course) OAC or other treatment methods because of progression of disease after second-line OAC, particularly if vitreous seeds are present at the time of initial OAC failure.