Suprachoroidal Injection of Topotecan for Retinoblastoma: A Preclinical Study

Abstract

Purpose

To assess whether levels of topotecan that are expected to be therapeutic against retinoblastoma tumors can be achieved within the retina and choroid by suprachoroidal injection (SCI) and to assess toxicity and safety in vivo.

Design

Pharmacokinetics and dose escalation toxicity study.

Subjects

New Zealand white rabbits.

Methods

In a pharmacokinetic study (N=18), aqueous, vitreous, retina, choroid, and plasma were separated and harvested serially (15, 30, 60, 120, and 360 minutes) following SCI. Topotecan (lactone) levels were measured and pharmacokinetic parameters were calculated. In the dose escalation toxicity study (N=8), toxicity was evaluated by ocular examination, fundus photography, OCT, full-field electroretinography (ffERG), and tissue histology. A single SCI of 50 μg/0.05 mL or two consecutive SCI totaling 100 μg/0.1 mL (N=4 rabbits per group) were administered.

Main Outcome Measures

Topotecan (lactone) tissue levels and ocular toxicity (25% reduction in ffERG).

Results

Following a single SCI of 50 μg topotecan, high levels of topotecan were achieved rapidly in both the retina and choroid. Retinal levels peaked by 15 minutes (12400±7336 ng/gm) followed by rapid decline to 2899±1361 ng/gm by 30 minutes, and then slower progressive decline that reached lowest levels at 360 minutes (469 ng/gm). Half-life (T_1/2) in the retina was 24.8 minutes. Choroidal levels were 3.3-fold higher than retina with a similar rapid decline pattern. Vitreous level was highest at 15 min (278 ng/mL) with a slow progressive decline until 360 min (16.9 ng/ml). Plasma (mean 4.3±2.6 ng/ml) and aqueous (peak at 120 min, mean 87 ng/ml) levels remained low throughout the study. There were no signs of ocular toxicity or other adverse ocular events on either clinical examination, serial imaging studies, ffERG, or histology following sacrifice at 28 days.

Conclusions

A single SCI of topotecan (50 μg/0.05 ml) achieved selective tissue distribution of its lactone moiety (retina/plasma, 1377.8) that was 23-fold higher than that reported with intraarterial chemotherapy (58.9) and more than 1000-fold higher than intravenous chemotherapy (1.32). These retinal levels were nontoxic and were 885-fold higher than the known topotecan IC50 for human retinoblastoma cells (IC50 14 ng/gm). Our findings support potential benefit of SCI of topotecan for patients with retinoblastoma.

Financial Disclosure(s)

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Treatment of retinoblastoma (RB) has undergone a paradigm shift with the use of chemotherapy.¹ In high-income countries, intraarterial chemotherapy (IAC) is preferred because of its greater efficacy for advanced cases of intraocular RB, with the goal of vision preservation.² Although in low- and middle-income countries, intravenous chemotherapy (IVC) is widely used as the first-line treatment because of the complex drug delivery approach and expensive infrastructure required for IAC delivery. In India, with the largest number of RB incident cases per year,³ <10% of all children with RB have access to IAC because of the high cost.⁴

Intraarterial chemotherapy has the advantages of delivering high doses of chemotherapy to the target tissue while simultaneously minimizing systemic toxicity,^5,6 which are well known with the standard IVC regimens for RB (neurotoxicity from vincristine,⁷ nephrotoxicity,⁸ and ototoxicity from carboplatin,⁹ and possibility of myelogenous leukemia with etoposide).¹⁰ However, IAC with current melphalan-based regimens is associated with ocular toxicity,¹¹ including vision-threatening serious adverse events such as retinal or choroidal vascular occlusion, vitreous hemorrhage, retinal detachment, and melphalan-related retinotoxicity (retinal pigment epithelial and choroidal atrophy).^12,13 Patient-related factors such as young age (<3 months), anatomic variations of ophthalmic artery, and bilateral advanced disease are few situations that further limit the use of IAC that requires ≥3 cannulations in an infant with smaller arteries,¹⁴ a challenge even for centers with highest level of surgical expertise.¹⁵

Intravitreal chemotherapy (IVitC) provides a direct approach to drug delivery, but its effect is largely limited to the treatment of vitreous seeds and subretinal seeds as an adjunct to IVC and IAC, with minimal effect on the retinal tumors (the source of the seeds).16, 17, 18, 19, 20, 21

In pursuit of delivering drugs close to the target area, suprachoroidal delivery of chemotherapy may be an ideal route to the target tissue (retina),²² bypassing the ophthalmic artery to access the choroidal circulation, thus avoiding some of the major risks of IAC.²³ Pharmacodynamic studies have shown uniform and rapid distribution of fluorescein through the choroid and retina into the vitreous after injection of fluorescein in the suprachoroidal space.^24,25 Prior studies employing the suprachoroidal route, including phase III trials, have indicated that triamcinolone acetonide via this delivery route is safe,^26,27 wherein the drug rapidly travels circumferentially across the entire choroid with some barrier effect from the posterior ciliary artery²⁸ and relative sparing of the macula—a desirable property when using chemotherapy.

We explored the suprachoroidal route of delivery for intraocular chemotherapy specifically for treatment of RB with an aim to assess whether therapeutic drug levels could be achieved within the retina and choroid by suprachoroidal injection (SCI) of topotecan and to assess its in vivo safety.

Methods

Animals and Ethics Statement

All experiments were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. The protocol was approved by the Institutional Animal Care and Use Committee of the Cleveland Clinic (Protocol Number: 2018-1924) and Pharmaron (Protocol Number: 23IA-OC-03). All experiments were performed in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. New Zealand white rabbits, aged 4 to 5 months and weighing 3.0 to 4.0 kg, were used, with experiments performed at Cole Eye Institute and Pharmaron Lab Services, LLC.

Animal Preparation

Buprenorphine (0.01–0.05 mg/kg SC) was given before injection for systemic analgesia. Rabbit eyes were dilated with 1% tropicamide (Somerset) and 10% phenylephrine (Lifestar) and anesthetized with an intramuscular injection of ketamine hydrochloride (up to approximately 50 mg/kg) and xylazine (up to approximately 10 mg/kg). Post anesthesia, eyes were prepared for the SCI by applying povidone iodine (betadine, Alcon) for at least 5 minutes, after which the eyes were rinsed with balanced salt solution (Alcon) and the surrounding area was wiped with gauze. Before placing the animals on the surgical table for injections, 1 to 2 drops of topical proparacaine hydrochloride anesthetic (10%) (Somerset) were then applied to the eye. The eye was draped by taking precautions for sterility, and a sterile wire speculum was placed to retract the eyelids. Animals were placed on a thermal heating pad to help maintain body temperature. Vital signs were monitored every ∼5 minutes during the surgery and ∼10 minutes throughout recovery.²⁹

Suprachoroidal Delivery of Topotecan

Topotecan 1 mg/mL (NDC# 0409-0302-01, Hospira, Inc) was used for the experiments and stored at 4 °C protected from light throughout the injection procedure. For pharmacokinetic studies, a single SCI was administered (50 μg topotecan/0.05 mL) into both eyes of each animal (Fig 1A).³⁰ For toxicity study, a single SCI of topotecan (50 μg/0.05 mL, superior temporal quadrant) or 2 SCI (50 μg/0.05 mL each, one in each of the superotemporal and inferotemporal quadrants) were given in the right eye (Fig 1B). Contralateral eyes received saline (0.05 mL single or 2 injections) (Hospira, Inc) as vehicle control. In brief, the bevel needle (30G, 0.70 mm) (Pricon, Iscon Surgical Ltd) attached to a 1 cc syringe was inserted through the sclera into the suprachoroidal space approximately 4 to 5 mm posterior to the limbus in the superotemporal or inferotemporal quadrant at a 90° angle to the eye surface. The needle was advanced to its full length, placing the needle tip into the SCS. Insertion of the needle into the suprachoroidal space was confirmed by applying gentle pressure on the injector’s plunger to check for resistance. If any resistance was encountered, the position of the needle was adjusted until no more resistance was felt. Once placement of the needle tip in the SCS was confirmed, topotecan or saline was injected over 5 to 10 seconds. After the injection, the needle was kept in place in the eye for approximately 30 to 60 seconds before being withdrawn. On slow withdrawal of the needle, a sterile cotton-tipped applicator was placed over the injection site for approximately 10 seconds to prevent drug efflux.

Figure 1.

The study was performed in 2 phases. A, Phase I: pharmacokinetic study, and B, phase II was a dose escalation toxicity study. In the first cohort (N = 4), topotecan 50 μg/0.05 mL was delivered to the right eye (OD) in a single suprachoroidal injection (SCI). In the absence of toxicity in cohort 1, the second cohort (N = 4) received 2 SCI of topotecan 50 μg/0.05 mL, 15 minutes apart and at a site 90° to 180° away from the first injection (∗). Both injections were given to the OD, with the left eye (OS) serving as the control.

Ophthalmic Examinations

Clinical examinations (slit-lamp biomicroscopy and indirect ophthalmoscopy) were performed (McDonald and Shadduck,³¹ 1977) on injected eyes of all animals at baseline and then post injection before euthanasia. Anterior segment photographs were taken using a blue filter on all animals preinjection and immediately postinjection. Blue-Light Autofluorescence imaging (Heidelberg Spectralis, Heidelberg Engineering) was performed to evaluate the distribution of the topotecan by capturing the natural fluorescence property of topotecan immediately post-SCI. In addition, OCT using a 55° lens on Heidelberg Spectralis instrument and color fundus photographs were performed on both eyes of all study animals immediately post-SCI to confirm the quality of the SCI and evaluate the retinal structure before euthanasia on days 28.²⁹

Full-Field Electroretinography

After pupillary dilation with topical application of 10% phenylephrine and 1% tropicamide, animals were dark-adapted for at least 2 hours before electroretinography (ERG) responses were recorded on both eyes before treatment (baseline), and again on day 28 before the animals were euthanized. Animals were anesthetized and maintained as described previously. Electroretinograms were recorded during full-field light stimulation (Diagnosys LLC) with the electrodes on the surface of the corneas. The International Society for Clinical Electrophysiology of Vision standard for rabbits was followed,^32,33 with clinical and statistical significance as described previously.34, 35, 36 Scotopic responses were obtained under dark conditions by following a 5-step intensity series from 0.0002 to 39.8 cd∗s/m². Photopic responses were obtained after 5 minutes of light adaptation by following a 3-step intensity series from 0.0084 to 39.8 cd∗s/m². For scotopic and photopic recording with increasing intensity, the interstimulus interval between the sweeps increased. Responses of 3 to 10 flashes were averaged to generate ERG a and b waveforms at each flash intensity.

Blood and Ocular Tissue Collection

Two milliliter of entire blood was collected in K₂EDTA vials from animals immediately before euthanasia at 15 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, and 6 hours postdose (N = 3–6/time point). After euthanasia, the choroid, retina, vitreous, and aqueous were collected. The choroid and retina were dissected separately; the retina was further sampled as proximal and distal to the injection site for each eye (N = 3/time point).

Ocular Tissue Sample Processing and Topotecan Lactone Assay

The samples obtained from study animals (plasma, vitreous, aqueous, retina, and choroid) were processed as described previously37, 38, 39, 40 to determine the total topotecan lactone concentrations using high-performance liquid chromatography according to a method modified from Warner and Burke^41,42 as we have described previously.¹⁸ Briefly, the chromatographic system consisted of a high-performance liquid chromatography pump (Waters 515; Waters) and a fluorometric detector (FL-45A; Bioanalytical Systems) with an excitation wavelength of 368 nm and an emission wavelength of 592 nm. We used a reverse-phase 3-μm, 3 × 150-mm column (C18; Phenomenex Co). Samples (20 μL) were injected at a flow rate of 0.4 mL/min at room temperature. Retention times of carboxylate and lactone topotecan were 3.5 and 8.4 minutes, respectively. For the preparation of topotecan standards, stock solutions of 1 mg/mL topotecan hydrochloride were prepared in methanol and stored at –20 °C. Topotecan lactone and carboxylate working solutions of 500 μg/mL were obtained by mixing equal volumes of the topotecan stock solution with pH 3 or pH 10 phosphate buffer, respectively. These solutions were maintained for 30 minutes at room temperature before further processing to ensure conversion to the pH-dependent forms of the drug.

Data Display

Topotecan lactone concentrations from each source (plasma, choroid, retina, vitreous, and aqueous) was averaged at each time point (N = 3 rabbits per time point). The liquid samples (plasma, vitreous, and aqueous) were pooled from both sets of rabbits were averaged, and the concentration was converted from micromoles/mL to ng/mL by multiplying by 0.458 (molar weight = 458 g/mole). For solid samples, the choroid and retina were dissected separately, and the retina was further sampled as proximal and distal to the injection site. Because the density of retina is essentially identical to the density of water,⁴³ the levels can be interchangeably depicted as ng/g or ng/mL for comparison with liquid levels, as has been done previously.⁴⁰

Pharmacokinetic Experiments

For 18 rabbits, CI of 50 μg/0.05 mL topotecan was performed as above, under general anesthesia, and all experiments were terminal. Animals were euthanized at various time points (15 minutes, 30 minutes, 1 hour, 2 hours, 4 hours, and 6 hours; N = 3 rabbits per timepoint). Ocular tissues were harvested and processed, and topotecan levels were analyzed, as described above.

The resulting mean time–concentration data were used to determine pharmacokinetic parameters (maximum concentration [Cmax] values), T_max, T_1/2, and area under the curve (AUC) from 0 to ∞ (AUC0–i∞) for choroid and retina. The time in the data set began at 15 minutes; it was assumed that at 0 minutes, the concentration would be 0 ng/g. Cmax and T_max calculations were conducted using the calc.ctmax() function in R.⁴⁴ The AUC was obtained using the trapezoid method⁴⁵ and calculated using the trapz() function in R.⁴⁶

Toxicity Experiments

Suprachoroidal injection procedures, ophthalmic examination and imaging, and ERGs were all performed as described above. Eight rabbits were included. In the first cohort (N = 4), suprachoroidal space injection of topotecan 50 μg/0.05 mL was delivered to the right eye, and the left eye acted as a control with SCI of 0.05 mL of saline. In the absence of toxicity, the second cohort (N = 4) received 2 SCIs of topotecan 50 μg/0.05 mL, 15 minutes apart and at a site 90° to 180° degree away from the first injection, with the left eye being control (SCI of 0.05 mL saline).

Histopathology

For toxicity experiments, after euthanasia, entire eye globes from each animal were fixed for 24 hours in modified Davidson’s fixative at room temperature After fixation, eye globes were processed and trimmed at 3 levels (nasal, central, and temporal), embedded in paraffin, sectioned at ∼5 μm thickness, and stained with hematoxylin and eosin, cover slipped for the imaging.^47,48

Results

The correct localization of topotecan within the suprachoroidal space could be demonstrated by the presence of a hyporeflective cleft on OCT in every case as well as by the natural fluorescent property of topotecan (captured by blue-light photography),⁴⁹ which showed circumferential and posterior spread externally and by fundus autofluorescence (Fig 2). Both methods demonstrated that the injected drug extended toward the posterior pole of the globe, as well as circumferentially, even reaching the opposite side of the globe within the first minute after SCI. In addition, OCT/fundus and ocular examinations showed no evidence of vitreous hemorrhage or retinal tear during the injection procedure, further confirming successful delivery of topotecan in the suprachoroidal space.

Figure 2.

Distribution of topotecan in the suprachoroidal space. A, External photograph shows distribution of topotecan fluorescence around the injection site (black arrow) in suprachoroidal space imaged using blue light. B, Green fluorescence signal from topotecan detected in the superior and inferior suprachoroidal space in all retinal quadrants. Yellow arrows represent signal in the suprachoroidal space, blue arrows represent choroidal vessels that block the underlying fluorescent signal, and the white star represents the optic nerve. Spectral domain OCT immediately postinjection reveals opening of the suprachoroidal space (C, red arrows).

After a single SCI of 50 μg topotecan, very high levels of topotecan lactone were achieved rapidly in both the retina and the choroid. Because the retina has a specific gravity of ∼1.0 mg/mL, 1 ng/g equals 1 ng/mL. Thus, the tissue concentration of the drug is expressed in ng/g, allowing for a direct comparison to vitreous, aqueous, and plasma levels in ng/mL. Retinal levels peaked by 15 minutes (the earliest timepoint assessed), with a mean retinal concentration of 12 400 ± 7336 ng/g that had rapid subsequent decline, achieving a lower tissue mean concentration of 2899 ± 1361 ng/g by 30 minutes and slower progressive decline that reached lowest levels at 360 minutes (469 ng/g). Unsurprisingly, choroidal levels were likewise very high, with a peak at the earliest postinjection timepoint (15 minutes) of 41 411 ± 21 959 ng/g (Fig 3). There was a rapid decline by 30 minutes to 19 322 ± 9334 ng/g, which continued at a slower rate, reaching the lowest levels by 360 minutes (2876 ng/g). The peak concentrations of topotecan in choroid and retina were both observed at the earliest recorded timepoint (15 minutes), with choroidal levels being 3.3 times that of retina. Because topotecan tissue levels in the retina were similar on both the side of the treatment and the opposite side of the globe at all time points, the values were combined and averaged for all analyses (Fig 4).

Figure 3.

Tissue drug concentrations over time after suprachoroidal injection of topotecan (50 μg/0.1 mL). The peak concentrations of topotecan in the choroid (CHOR) and retina were both observed at 15 minutes. The plasma levels were low at all time points. AQ = aqueous; VIT = vitreous.

Figure 4.

Topotecan tissue levels in the retina. Levels proximal (P) and distal (D) to the injection site were similar at all time points.

In contrast, plasma, aqueous, and vitreous levels remained low throughout the study. The plasma levels were low at all time points (mean, 4.3 ± 2.6 ng/mL), peaking at 60 minutes (8.7 ng/mL) and lowest at 360 minutes (1.4 ng/mL). The aqueous level peaked at 120 minutes (mean, 87 ng/mL) with reduction at 240 minutes (19.7 ng/mL) and 360 minutes (16.9 ng/mL). In contrast, the vitreous level was highest at 15 minutes (278 ng/mL) with a slow progressive decline until 360 minutes (16.9 ng/mL).

Peak concentrations in the retina were thus 1426-fold higher than peak plasma levels, 143-fold higher than peak aqueous levels, and 45-fold greater than peak vitreous levels. Peak concentrations in the choroid were even higher than peak plasma, aqueous, and vitreous levels (4760-fold, 476-fold, and 149-fold, respectively).

Levels of peak concentration (choroid > retina > vitreous > aqueous) and time to peak concentration (choroid = retina = vitreous [15 minutes] and aqueous = 120 minutes) followed the expected gradient and flow of drug injected in the suprachoroidal space.²²

The Cmax in the choroid was 41 411 ng/g with T_max of 15 minutes, and the AUC obtained using the trapezoid method was 3 834 915 ng.min. The Cmax in the retina was 12 400 ng/g with T_max of 15 minutes, and the AUC for retina obtained using the trapezoid method was 713 167.5 ng.min. Half-life (T_1/2) of topotecan within the retina was calculated as 24.8 minutes, very similar to the half-life of topotecan in the choroid (29.1 minutes).

Electroretinography demonstrated no reduction in retinal function at 28 days, compared with baseline before topotecan injection (Fig 5). Similarly, there were no signs of ocular toxicity or other adverse ocular events on either clinical examination, serial imaging studies (Figure 6, Figure 7), or histology after sacrifice at 28 days (Fig 8). Absence of toxicity was demonstrated both in the cohort #1 receiving a single dose (Fig 6), (50 μg topotecan/0.05 mL) as well as in the cohort #2 receiving 100 μg topotecan (administered as 2 separate SCIs of 50 μg topotecan/0.05 mL delivered 90°–180° apart from each other, 15–30 minutes apart) (Fig 7).

Figure 5.

Electroretinography demonstrated the absence of reduction in retinal function at 28 days, compared with the baseline (before topotecan injection). SD = standard deviation.

Figure 6.

Lack of toxicity. Cohort #1. Single dose, 50 μg topotecan/0.05 mL. Color retinal fundus, A, and spectral domain OCT, B, on day 28 post suprachoroidal administration show normal retinal/choroidal/vitreous structure.

Figure 7.

Lack of toxicity. Cohort #2 (double dose, 2 × 50 μg topotecan/0.05 mL). Color retinal fundus, A, and spectral domain OCT, B, on day 28 post suprachoroidal administration show normal retinal/choroidal/vitreous structure.

Figure 8.

Lack of toxicity. Normal histology after sacrifice on day 28 post suprachoroidal administration in both cohorts (cohort #1: single dose 50 μg topotecan/0.05 mL; cohort #2: double dose, 2 × 50 μg topotecan/0.05 mL). Photomicrograph depicting normal structure (black arrows) of the anterior segment, optic nerve (ON) head, and posterior segment in the animals treated with saline and topotecan (50 and 100 μg/eye). Top 2 panels, left, stain, hematoxylin and eosin; original magnification, ×2; and bottom panel, stain, hematoxylin and eosin; original magnification, ×20. AC = anterior chamber; C = cornea; CB = ciliary body; I = iris leaflet; INL = inner nuclear layer; L = lens; ONL = outer nuclear layer; PR = photoreceptor layer; R = retina; S = sclera; SCS/C = suprachoroidal space/choroid; V = vitreous.

Discussion

Although local administration of chemotherapy to the eye via IVitC has been used as an adjunct to treat vitreous and subretinal seeds, treatment of the retinal tumors themselves has usually required either IVC or IAC,²¹ because periocular chemotherapy has largely been abandoned because of complications associated with this therapeutic approach.⁵⁰ Even with high doses of topotecan (90 μg/0.18 cc) standalone effect on retinal tumor has been limited (1 of 3 tumors⁵¹ and 3 of 13 recurrent tumors²¹) with remarkable lack of retinal toxicity.⁵¹

Intraocular delivery of chemotherapy via SCI gives selective access to the choroid and retina, without the potential complications associated with catheterizing the internal carotid artery to reach the ophthalmic artery during IAC. Other limitations of IAC related to anatomic variation of ophthalmic artery (origin, angulation, and size),⁵² dependence on expertise (intervention neuroradiologist),⁵³ and equipment can potentially be avoided in cases where SCI could be employed.

Although the suprachoroidal space is considered a potential space,²⁵ it can be visualized using enhanced-depth imaging OCT in about 45% of healthy subjects, particularly in hyperopes.⁵⁴ Suprachoroidal injection preferentially delivers drugs to the posterior segment as shown by studies in rats²⁴ and rabbits²² that have low molecular weight (300 Da–250 KDa) and are water soluble, such as sodium fluorescein (C₂₀H₁₀Na₂O₅) with a molecular weight of 376 Da. Topotecan (C₂₃H₂₃N₃O₅) is a water-soluble (1 mg/mL)⁵⁵ topoisomerase inhibitor chemotherapeutic agent with similar chemical properties to fluorescein (molecular weight 421 Da). Therefore, it can be expected that topotecan will also have preferential distribution in the posterior segment. In vivo studies have shown that the rabbit eye can accommodate SCI of up to 200 μL sodium hyaluronate.⁵⁶ Therefore, it is likely that the larger human eye can potentially tolerate larger SCI volumes of up to 250 μL.⁵⁷

SCI is Safe

Suprachoroidal injection is safe in adults with well-established techniques⁵⁸ and extensive data derived from multiple trials that used proprietary suspension of triamcinolone acetonide (CLS-TA) injected using a microinjector (Clearside) are summarized in a report by the American Academy of Ophthalmology.²⁷ Data sets from 8 clinical trials of 621 adult patients with noninfectious uveitis, diabetic macular edema, and retinal vein occlusion were reviewed.⁵⁹ Procedure-related ocular serious adverse events that occurred in any patient were deemed not related to treatment by a masked investigator. There were no serious adverse events involving lens injury, suprachoroidal hemorrhage, endophthalmitis, or retinal tear. Treatment emergent adverse events were limited to eye pain in 6% compared with 1.6% in controls (undergoing intravitreal injection).⁵⁹

SCI of Topotecan is Nontoxic

Historically, topotecan by various other administration routes seems to be safe, with clinical usage for last 20 years in children with relapsed/refractory metastatic and intraocular RB because initial reports of intravenous delivery (2 mg/m²/d × 5)⁶⁰ and in a clinical trial (3 mg/m²/d × 5).⁶¹ Since then, topotecan toxicity has been evaluated in animal models (pig^62,63 and rabbit^18,64,65) with ERG and histopathology^18,65 delivered by episcleral implant,⁶⁴ peribulbar injection⁶² IVitC,^18,65 or by IAC.^62,63

These animal data and observations of topotecan safety have been supplemented by clinical observations after IAC^53,66,67 and IVitC.^18,22 (Table 1).^18,51,53,62, 63, 64^,66,68 After IAC (0.4 mg in 30 mL), Francis et al⁶⁶ reported a reduction in ERG of 4.6 mV/0.1 mg increase in topotecan dose. With IVitC injection of topotecan (30 μg in 0.1 mL in the rabbit eye, equivalent to 60 μg in the larger human eye), Bogan et al¹⁸ did not observe any changes in ERG. However, with an even higher dose (90 μg in 0.18 mL), Francis et al⁶⁶ reported ERG reduction, considered as indicative of retinal toxicity, similarly found in 1 of 3 eyes by Abramson et al.⁵¹ Considering human vitreous volume of 3.0 mL,⁶⁹ effective safe concentration of topotecan in vitreous can be calculated to be approximately 7.3 μg/mL. The vitreous Cmax after SCI of only 0.278 μg/mL is not expected to be toxic. Lack of toxicity was confirmed by clinical examination, color fundus imaging, OCT, ERG, and, finally, histopathology (harvested at 28 days). Absence of toxicity was demonstrated in both the cohort receiving a single dose (50 μg topotecan/0.05 mL) and in the cohort receiving a double dose (100 μg topotecan/0.1 mL total). We defined 100 μg topotecan as the NOAEL (“no observed adverse effect level”), meaning that it was the maximum tolerated dose, not because there were dose-limiting toxicities but, rather, because this was the highest dose tested and there were no dose-limiting toxicities at this dose.

Table 1.

Topotecan (Dose, Volume) Safety Profile in Ocular Tissues by Route of Delivery in Animal Models and Clinical Use

Animal/Clinical	Author, Yr	Route, Dose, Volume	Nontoxic Tissue Effect/Levels
			ERG/Pathology	Choroid pg/mg	Retina pg/mg	Vitreous μg/mL
Rabbit	Carcaboso et al,64 2010	Episcleral implant	Pathology	600	0.012	300
		0.3/2.3 mg
	Bogan et al,18 2022	IVitC	No ERG reduction/ pathology			21.4∗ (calculated)
		30 μg
		0.1ml
	Del Sole et al,68 2022	IVitC	No ERG reduction/ pathology			7.1–35.7∗ (calculated)
		10–50 μg
		0.15 mL
	This study	SCI	No ERG reduction/pathology	41 411†	12 400†	0.608
		50 μg
		0.05 mL × 2
Pig	Schaiquevich et al,62 2012	Peribulbar				0.015
		1 mg
		1 mL
	Schaiquevich et al,62 2012	IAC				0.138
		1 mg
		30 mL
	Taich et al,63 2016	IAC			11 434	1.371
		4 mg
		30 mL
Clinical	Gobin et al,53 2011	IAC	Reduction at 1.5 mg
		0.4 mg
		30 mL
	Francis et al,66 2014	IAC	Reduction of 4.6 mV/0.1 mg increase in dose
		0.4 mg
		30 mL
	Bogan et al,18 2022	IVitC	No ERG reduction
		30 μg
		0.1 ml
	Abramson and Francis,51 2023	IVitC	Reduction of 5 μV in 1 patient
		90 μg
		0.18 mL

ERG = electroretinography; IAC = intraarterial chemotherapy; IVitC = intravitreal chemotherapy; SCI = suprachoroidal injection.

^∗Vitreous concentration calculated by dose (either 30 μg, or 10–50 μg) divided by the vitreous volume of the rabbit (1.4 mL).

^†pg/mg = ng/g.

SCI of Topotecan Offers Selective Tissue Distribution

Suprachoroidal injection offers several potential advantages as compared with intravenous and intraarterial routes because it is expected to lack systemic toxicity because of low plasma levels while simultaneously providing levels of drugs in the choroid and retina that are unachievable even with intravitreal injection. Accounting for variations in drug dose, concentration, volume, and route of administration, and animal model, the tissue levels of topotecan can be best depicted as a ratio to the level in the plasma, which emphasizes selective tissue distribution. The ratio of choroid/plasma of (Cmax) 4601.2 with SCI was incomparable with other routes of delivery because such data do not exist. However, the ratio of the plasma to the retina, which was more clinically meaningful, and was 1377.8 for SCI, was greater than that reported with IAC (58.9) and IVC (1.32).^62,63 High vitreous/plasma ratio can be achieved with direct IVitC (254) than with SCI (30.9) (Table 2).^43,61,62,68 The vitreous and aqueous levels are not critical for clinical consideration for evaluation of SCI because these can be supplemented by direct, safe, and easy access to these compartments.^18,51,70, 71, 72

Table 2.

Total Topotecan Levels (Cmax, ng/g⁴³) in Ocular Tissues by Route of Delivery in Animal Models^61,62,68

Ratio	Route of Delivery
	SCI (This Study)	IAC	IVC	IVitC	PeribulbarC	ICamC
Choroid/plasma	4601.2	–	–	–	–	–
Retina/plasma	13 77.8	58.9	1.32	–	–	–
Vitreous/plasma	30.9	17.0	0.07–0.04∗	254 I	0.11, 1.4†	-
Aqueous/plasma	9.7	–	–	–	–	–

Cmax = maximum concentration; IAC = intraarterial chemotherapy; IVC = intravenous chemotherapy; IVitC = intravitreal chemotherapy; SCI = suprachoroidal injection.

^∗0.07 (Rabbit) and 0.04 (pig).

^†0.11 (Rabbit) and 1.4 (pig).

SCI of Topotecan Reaches Therapeutic Tissue Levels

Clinical efficacy of topotecan for intraocular RB delivered by all routes (IVC,^61,73 IAC,^53,66,74,75 and IVitC^18,51,70) except peribulbar^42,76 and injected into both ocular compartments (vitreous cavity^18,51,70 and anterior chamber⁷²) has been established. Topotecan undergoes a reversible pH-dependent hydrolysis into the lactone form that is pharmacologically active, with acidic pH favoring lactone formation.⁷⁷ All the topotecan levels reported herein represent lactone moiety levels.

Retinal levels are most critical as RB arises in the retina and then extends to the vitreous cavity internally and the choroid externally as it advances through its stages. The choroidal levels of the drug drive the retinal levels as the drug diffuses through the retina into the vitreous and aqueous in a posterior-to-anterior gradient.^24,25 After SCI, the mean retinal concentration of 3573 (range, 469–12 400) ng/g was comparable with that obtained in a pig with IAC of 4286 (range, 3552–11 434) ng/g, levels higher than that achieved with IVC 64.7 (range, 30.0–84.7) ng/g. After SCI, the peak retinal levels of 12 400 (standard deviation, 7336) ng/g rapidly declined by 30 minutes (mean, 2899 ng/g; standard deviation, 1361) with a slower progressive decline that continued to decrease until 360 minutes (469 ng/g). In contrast, retinal levels of melphalan in rabbit models of IAC reach Cmax (T_max) in 30 minutes, with rapid decline by 120 minutes, with a longer half-life (69.6 minutes).⁴⁰ However, for the duration of the study (360 minutes), the retinal topotecan levels were approximately 34-fold to 885-fold higher than the concentration required to inhibit 50% of the RB cell cultures (IC₅₀, 14 ng/mL)⁷⁸ and 3.8-fold to 98-fold higher than the concentration required to inhibit 90% of RB cells (IC₉₀, 126 ng/mL),¹⁸ based on in vitro data.

SCI of Topotecan: Study Limitations

Tissue levels of topotecan in non–disease-bearing rabbit eyes are not representative of tissue levels in the human eye with RB. Hypoxic⁷⁹ and acidic⁸⁰ microenvironment within the RB tumor is expected to enhance the efficacy of topotecan by favoring its conversion to the pharmacologically active lactone moiety. Although the data seem promising, ongoing efficacy studies in the rabbit RB model^40,81 need to be demonstrated before initiating a phase I/II clinical study.

SCI of Topotecan: Potential Limitations in Children

No such data exist for usage in children who have smaller eyes and thinner sclera. The scleral thickness in children (up to 10 years of age) is measured as 500 ± 36 μm,⁸² with a lack of formal measurements in the younger age group that would be expected to have RB. We are currently performing studies using ultrasound biomicroscopy to measure scleral thickness in treatment-naïve children with RB. Initial clinical studies would necessitate that SCI be done after ultrasound biomicroscopy confirmation of scleral thickness and with demonstration of underlying choroid to be free of tumor infiltration (that could hamper drug distribution, and which might lead to extrascleral/extraocular tumor dissemination) at the injection site.

Total topotecan levels measured between the hemisected half of the retina near the injection location, and the half of the retina opposite the injection location (n = 18 eyes) were similar at all time points (Fig 4) suggesting that rapid distribution of topotecan occurs within retina both near to and distant from the SCI site. Importantly, this suggests that in clinical practice, SCI could potentially be performed in patients on the side of the eye opposite to where tumors are located, thus increasing the safety of the procedure, likely without sacrificing tumor drug levels or efficacy. However, there may with some limitation in quadrantic distribution in the human eye above and below the horizontal because of the long posterior ciliary artery,²⁸ which may be overcome by injecting in the quadrant of the tumor or injecting in 2 quadrants.⁸³ Similarly, the short posterior ciliary arteries may prevent circumferential spread toward the optic disc, limiting therapeutic levels for juxtapapillary tumors. On the other hand, relative sparing of the macula (due to short posterior ciliary arteries) may be advantageous by reducing the risk of potential toxicity.²⁸

Conclusions

A single SCI of topotecan (50 μg/0.05 mL) achieved selective tissue distribution of its lactone moiety (retina/plasma, 1377.8) that was 23-fold higher than that reported with IAC (58.9) and >1000-fold higher than that with IVC (1.32). These retinal levels were nontoxic and achieved drug levels up to 885-fold higher than the known topotecan IC₅₀ for human RB cells (IC₅₀ 14 ng/g). Even doubling this dose to 100 μg topotecan remained nontoxic, suggesting a wide therapeutic window. These findings support the potential benefit of suprachoroidal topotecan in clinical use for patients with RB.

Manuscript no. XOPS-D-25-00154R1.