INTRODUCTION

Topotecan, a topoisomerase I inhibitor of DNA replication that mainly acts during the S phase of the cell cycle, has shown effective antitumor activity in the treatment of paediatric tumours including retinoblastoma.^{1 2} Several routes of topotecan administration have been used for the treatment of intraocular retinoblastomas including intravenous, intra-arterial, intravitreal, intracameral and periocular.^{1 3}

Intravitreal chemotherapy, which is used as an adjunct to other treatments, is primarily used for eyes with refractory and persistent vitreous seeds.^{2 4 5} The advantage of intravitreal administration is that it can achieve a high vitreous concentration of chemotherapy without systemic side effects. However, in animals, higher vitreous concentrations of certain chemotherapeutic agents including melphalan lead to ocular toxicity measured with electroretinograms (ERGs).^{1 6} The systemic administration of topotecan can lead to side effects including pancytopenia, diarrhoea, rash or fever.⁷ Studies on the intravitreal delivery of topotecan have shown low levels of the drug in plasma, and no cytopenias have been reported from intravitreal use.^{1 8} ERGs were introduced by our group for assessment of retinal function after both intra-arterial chemotherapy⁹ and intravitreal melphalan⁶ and have subsequently been used following intravitreal therapy with topotecan.¹⁰ Melphalan is the most commonly used agent for intravitreal injection, and ocular salvage rates exceeding 80% have been widely reported^{10 11}; however, it has well-defined retinal toxicity as measured by ERGs and a clinical grading system.^{6 10 12}

The aim of our retrospective study was to evaluate the effect of intravitreal topotecan injections alone on retinal function in patients with retinoblastoma by measuring interval changes in ERG amplitude of 30 Hz.

MATERIALS AND METHODS

Memorial Sloan Kettering Cancer Center Institutional Review Board approval was obtained.

We retrospectively reviewed eyes with retinoblastoma treated with intravitreal topotecan at Memorial Sloan Kettering Cancer from December 2014 to May 2019. The flicker ERG recordings at 30 Hz were obtained under anaesthesia immediately prior to each injection and at the next visitor at approximately one-month follow-up prior to administering additional injections. Injections were excluded from the analysis if there was not a baseline ERG recorded on the date of injection or if the ERG amplitude was not sufficient to allow ERG change over the injection course (amplitude less than 5 μV). ERG measurements above 100 μV were normalised to a value of 100 μV.

The primary outcome measured was the change in ERG amplitude. Demographic information collected included age at diagnosis, sex, initial classification at diagnosis (International Classification/Children’s Oncology Group Version and Reese-Ellsworth), prior cancer-directed therapies, family history and genetic history. Patients were included who received intravitreal topotecan alone and those who received concurrent injections of intravitreal melphalan.

A paired two-tailed t-test was performed to compare ERG recordings immediately prior to injection and at the follow-up visit. We explored trends in the data through a line plot of each patient’s trajectory of ERG over injection, with a linear regression line showing the overall trend in the data. All statistical analyses were conducted using Graph Pad Prism Version 8 and a p value of less than 0.05 was considered significant.

RESULTS

This study included 42 eyes of 41 patients (16 with unilateral disease; 25 with bilateral disease) including 22 males and 19 females. Five (12.2%) patients had a family history of retinoblastoma. Thirty-five (83.3%) eyes had very advanced retinoblastoma with the International Classification Groups ‘D’ or ‘E’ on initial evaluation. Nineteen (63.3%) out of 30 patients who were tested had a positive germline genetic mutation. Median age at diagnosis was 17 months. Twenty (48.8%) patients received prior systemic chemotherapy, 5 (12.2%) patients received intravitreal chemotherapy at an outside institution and all patients received prior intra-arterial chemotherapy.

Among the 41 patients who were treated with intravitreal topotecan during the study period, 108 intravitreal injections of topotecan were administered. Twenty-seven injections were excluded because they did not have an ERG recorded both on the date of the injection and at the next follow-upprior to administering additional injections. Thirty-one injections were excluded as the baseline ERG was less than 5 μV.

There were 50 topotecan injections administered to 28 eyes that were included in the final analysis. Twenty-two (44.0%) of those injections were performed with concurrent intravitreal melphalan. Among the 12 eyes that received concomitant melphalan, 10 (83.3%) had very advanced disease at diagnosis. Eleven (68.8%) of the 16 eyes that received intravitreal topotecan without melphalan had very advanced disease at diagnosis. The median time to ERG measurement following an injection was 27 days.

By using a paired t-test, intravitreal topotecan combined with melphalan (n=22) at a dose of 25 μg or 30 μg was associated with a significant decrease in ERG amplitude at follow-up (p=0.046, 95% CI −20.4 μV to −0.2 μV). Among eyes that only received topotecan (n=28) at doses of 20 μg or 30 μg, there was not a significant difference in ERG amplitude at follow-up (p=0.85, 95% CI −7.0 μV to 5.8 μV).

Seven (14.0%) topotecan injections were administered within 7± 3 days of the administration of intra-arterial chemotherapy; in this cohort, there was no significant decrease in ERG amplitude at follow-up (p=0.16, 95% CI −26.9 μV to 5.6 μV). Forty-five (90.0%) injections were administered with concomitant focal treatment including laser and/or cryotherapy.

Individual patient trajectories of ERG values over time are plotted in figure 1. For patients who received topotecan only, there was no significant degradation of the flicker ERG response at 30 Hz with a slope of −1.6 μV/injection (p=0.72, 95% CI −10.5 μV/injection to 7.4 μV/injection) (figure 1A). The fitted line in figure 1B shows an overall downward trend in ERG amplitude over time among eyes that received concurrent melphalan with a median of two combined injection observations per eye. For each combined topotecan and melphalan injection observation, the slope of the regression line shows a decrease in ERG response of 6.0 μV per injection (p=0.13, 95% CI −13.8 μV/injection to 1.8 μV/injection).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Individual patient trajectories of ERG over time, as measured by number of injection events (individual points) with a fitted line demonstrating the overall trend (A). Intravitreal topotecan injections alone with a linear regression line fitted to the data with a slope of −1.6 μV/injection (p=0.72, 95% CI −10.5 μV/injection to 7.4 μV/injection). (B) Concurrent intravitreal topotecan and melphalan injection events with a linear regression line fitted to the data with a slope of −6.0 μV/injection (p=0.13, 95% CI −13.8 μV/injection to 1.8 μV/injection). ERG, electroretinogram.

DISCUSSION

The majority of retinoblastoma eyes at presentation have vitreous and/or subretinal seeding, and until the introduction of ophthalmic artery chemosurgery and intravitreal chemotherapy, seeding was the main reason for enucleation.¹³ In advanced stages of retinoblastoma, systemic chemotherapy has limited efficacy in cases with vitreous seeding.¹² Intravenously administered chemotherapy does not reach a sufficient concentration to treat vitreous seeds due to its low permeability across the blood–retinal barrier and the avascular nature of the vitreous humour.⁸

Intravitreal topotecan has been reported to be efficacious in causing the regression of vitreous seeds in patients with advanced stages of retinoblastoma. Rao et al showed that in their cohort of 17 patients, they achieved total resolution of vitreous seeds after a mean of 3 intravitreal topotecan injections.¹⁴ Ghassemi et al studied the clinical use of intravitreal topotecan combined with melphalan, achieving control of vitreous seeds in all 9 eyes following a mean of 1.9 injections but they did not report on success with intravitreal topotecan alone. They did not observe any changes in a- or b-waves of bright-flash ERGs.²

While melphalan has documented efficacy to treat vitreous seeds in advanced retinoblastoma, it has established toxicity. Francis et al showed that the time to regression of vitreous seeds and the response to intravitreal melphalan were dependent upon the seed classification with an overall 98.5% ocular event-free survival from intravitreal melphalan.⁵ Munier et al showed that for 23 patients with active vitreous seeding treated with intravitreal melphalan combined with other routes of chemotherapy, globe retention occurred in 87% of cases; 100% of the retained eyes achieved remission. They noted a localised peripheral salt-and-pepper retinopathy in 10 (43%) of eyes at the site of injection.¹¹ Francis et al showed that intravitreal melphalan was associated with high rates of ocular survival and disease-free survival of 94.2% and 86.2%, respectively. They also showed that intravitreal melphalan was associated with an average decrease in ERG response of 5.3 μV for each injection. They found that the extent of salt-and-pepper retinopathy and the degree of eye pigmentation were associated with a significant reduction in ERG response. The use of concomitant intravitreal topotecan was not associated with toxicity.¹⁰ Xue et al showed that vitreous seeds were controlled in 26 (86.7%) out of 30 eyes that received intravitreal melphalan.¹⁵

To our knowledge, this is the first study to date to investigate retinal toxicity (as measured by ERG responses at 30 Hz) of intravitreal topotecan injections alone in human retinoblastoma.

The limitations of the study include its retrospective nature and that several patients also received other cancer-directed therapies including intravitreal melphalan throughout their treatment course. The eyes that received intravitreal topotecan combined with melphalan had a very advanced or recalcitrant disease.

Overall, this report builds upon the previously noted studies that showed that intravitreal topotecan combined with other therapies leads to regression of vitreous seeds in advanced retinoblastoma. The significance of our work is that it demonstrates that intravitreal topotecan injections are not associated with retinal toxicity (ie, decrements in ERG) in patients with retinoblastoma.