Efficacy and Toxicity of Intravitreous Chemotherapy for Retinoblastoma: Four-Year Experience

Abstract

Purpose: To investigate the efficacy and toxicity of intravitreous melphalan for treatment of retinoblastoma, as a single agent or with concomitant topotecan.

Participants: A total of 130 eyes of 120 patients with retinoblastoma receiving 630 intravitreous (melphalan, topotecan) or topotecan periocular injections. A total of 83 (64%) of these eyes were treated with concomitant ophthalmic artery chemosurgery (OAC).

Design: Retrospective cohort study.

Methods: Indirect ophthalmoscopy and clinical imaging were used to evaluate clinical response. Ocular survival and disease-free survival were estimated using Kaplan-Meier methods in 130 eyes. Ocular toxicity was evaluated by clinical findings and electroretinography (ERG) on 244 evaluable injections in 63 patients using 30-Hz flicker responses. Analysis was performed using linear mixed effects models with a random intercept and slope for each patient and a fixed effect for number of injections, in addition to any other fixed effect of interest.

Main outcome measures: Ocular survival, disease-free survival, ERG: peak-to-peak ERG amplitudes in response to 30-Hz photopic flicker stimulation.

Results: There were no disease- or treatment-related deaths, and no patient developed externalization of tumor or metastatic disease. Two-year Kaplan-Meier estimates of ocular survival and disease-free survival were 94.2% (95% confidence interval, 89.2-99.4) and 86.2% (95% confidence interval, 78.7-94.5), respectively. There was a significant association between the number of injections and diminished ERG responses, such that on average each intravitreous melphalan injection was associated with a 5.3-μV decrease in ERG amplitude (P < 0.001). Concomitant intra-arterial chemotherapy (P = 0.01) and greater inherent ocular pigment also were significantly associated with a reduction in ERG (P = 0.045). Patient age and weight, new injection site location, addition of topotecan, concomitant focal treatment, and time interval between injections were not significantly associated with toxicity.

Conclusions: Intravitreous melphalan is an effective treatment for vitreous seeding in retinoblastoma, resulting in high rates of ocular survival and disease-free survival. However, in this study, each injection of melphalan was associated, on average, with a decrement in ERG response. The findings suggest increased toxicity (1) when OAC is given within 1 week of the intravitreous injection and (2) in more deeply pigmented eyes.

Figure 1

Kaplan Meier survival curves for (left) Ocular survival of all eyes, (right) Event free survival of all eyes.

Figure 1

Kaplan Meier survival curves for (left) Ocular survival of all eyes, (right) Event free survival of all eyes.

Figure 2

Waterfall plot demonstrating change in electroretinogram (ERG) response recorded after each intravitreal injection of melphalan

Figure 3

Individual patient trajectories of ERG over time, as measured by number of injections, (dotted lines) with a LOWESS smooth depicting the overall trend (solid line).

Figure 4

Representative cases of each seed classification (class 1, 2, and 3) and response to intravitreous melphalan. Representative eye with blue iris and class 2 (spheres) vitreous seeds (upper left) had a degradation of 3.6μV following 3 melphalan injections (upper right). Representative eye with dark brown iris and class 3 (cloud predominant) vitreous seeds (lower left) had comparatively more degradation (26.4μV) following 3 melphalan injections (lower right).