Species differences in ocular pharmacokinetics and pharmacological activities of regorafenib and pazopanib eye‐drops among rats, rabbits and monkeys

Abstract Age‐related macular degeneration (AMD) is the leading cause of severe vision impairment in patients over the age of 60 years. Choroidal neovascularization (CNV) is the hallmark of neovascular AMD and vascular endothelial growth factor (VEGF) plays a causal role in the formation of CNV. Although regorafenib and pazopanib, small molecule VEGF receptor (VEGFR) inhibitors, were developed as eye‐drops, their efficacies were insufficient in clinical. In this study, we evaluated ocular pharmacokinetics and pharmacological activities of regorafenib and pazopanib after ocular instillation in multiple animal species. In rats, both regorafenib and pazopanib showed high enough concentrations in the posterior eye tissues to inhibit VEGFR. In laser‐induced rat CNV model, regorafenib showed clear reduction in CNV area. On the other hand, the concentrations of regorafenib and pazopanib in the posterior eye tissues were much lower after ocular instillation in rabbits and monkeys compared to those in rats. Pazopanib did not show any improvement in monkey model. Regorafenib was nano‐crystalized to improve its drug delivery to the posterior eye tissues. The nano‐crystalized formulation of regorafenib showed higher concentrations in the posterior segments in rabbits compared to its microcrystal suspension. From these studies, large interspecies differences were found in ocular delivery to the posterior segments after ocular instillation. Such large interspecies difference could be the reason for the insufficient efficacies of regorafenib and pazopanib in clinical studies. Nano‐crystallization was suggested to be one of the effective ways to overcome this issue.


| INTRODUC TI ON
Age-related macular degeneration (AMD) is a chronic disease in the posterior eye segment and the leading cause of severe vision impairment and legal blindness in patients over 65 years old in Western populations. [1][2][3][4] It has been estimated that the number of patients with AMD is going to increase in the next few decades because 25% of Asian people are going to be over 60 years old by 2050. 5 AMD is classified into two subgroups, the non-neovascular (nonexudative or dry) form and the neovascular (exudative or wet) form of the disease. 6 Choroidal neovascularization (CNV) is the hallmark of neovascular AMD and causes leaking fluid, lipids and blood, and those leads to fibrous scarring. 3 Vascular endothelial growth factor (VEGF) is a regulator of neovascularization and plays a causal role in the formation of CNV. 7,8 Anti-VEGF therapy has remarkably improved visual outcomes in neovascular AMD patients, and ranibizumab and aflibercept are mainly used as anti-VEGF agents. Intravitreal injections of ranibizumab and aflibercept not only prevent vision loss but also lead to significant visual gain. [9][10][11][12][13][14] However, the current anti-VEGF therapy has multiple issues: invasive and frequent injections, high financial costs, and risk of ocular and systemic adverse events. [14][15][16] To overcome these problems, small molecule inhibitors of VEGF receptors (VEGFR) have been developed as eye-drop formulation. 6 Regorafenib and pazopanib are multiple tyrosine kinase inhibitors which mainly targets VEGFRs. Regorafenib has been approved for the treatment of metastatic colorectal cancer, metastatic gastrointestinal stromal tumor and hepatocellular carcinoma, and pazopanib for advanced renal cell carcinoma and advanced soft tissue sarcoma. 17,18 Eye-drop formulations of regorafenib (ophthalmic suspensions) and pazopanib (ophthalmic solutions) had been clinically developed for the treatment of neovascular AMD. However, the clinical development was terminated because of the lack of efficacy. [19][20][21] The eye-drop formulations of regorafenib and pazopanib showed significant improvements in nonclinical laser-induced CNV models in our study or another study, 22 however, reasons for their poor efficacies in humans are still unclear.
Nano-crystalization of drugs has been studied for the purpose of improving ocular drug delivery. Compared to conventional micro-sized crystals, nano-crystalization increase the surface area of particles which lead to increased dissolution rate and apparent solubility of drugs. In addition, nano-crystals are beneficial for long time retainment on eye surface. Nano-crystals can be retained in the cul-de-sac and its large surface area allows long-term adhesion to the eye surface. [23][24][25][26] In this study, in order to consider the discrepancy observed in the efficacies between animal models and clinical studies, interspecies difference in ocular pharmacokinetics was investigated for regorafenib and pazopanib after ocular instillation. From the results, large interspecies difference was found in the ocular delivery to the posterior segments. The concentrations of regorafenib and pazopanib in the posterior segments were high in rats but low in rabbits and monkeys. The impact of nano-crystalization was also investigated for regorafenib. The concentrations of regorafenib in the posterior eye segments were increased by nano-crystalization. Our results indicated the importance to consider interspecies difference in ocular delivery to the posterior segments. Our results also suggested that nano-crystalization is an effective way to increase ocular delivery in animal species where poor ocular delivery was observed with conventional microcrystal suspension.

| Animals
All the animal tests in this study were conducted in accordance with the Guide for the Care and Use of Laboratory Animals. All the experimental protocols used in this study were approved by the

| Pharmacokinetic studies in rabbits and monkeys
In rabbits, regorafenib eye-drops at 24.1 mg/mL (20 μL/eye) or pazopanib eye-drops at 5.0 mg/mL (20 μL/eye) was administered as a single topical dose, and blood samples and eye tissue samples (the choroid/retina) were collected at 1.5 hours after ocular instillation of regorafenib eye-drops (n = 3 for each sample) or pazopanib eye-drops (n = 2 for each sample). In addition, nano-crystalized regorafenib ophthalmic formulations with nano-particles at concentrations of 1.7 mg/mL (average particle size: 97.42 nm) and 2.1 mg/mL (average particle size: 233.8 nm) were administered to rabbits and studied pharmacokinetics in the same manner described above.
In monkeys, regorafenib eye-drops at 17.1 mg/mL (50 μL/eye) or pazopanib eye-drops at 5.0 mg/mL (30 μL/eye) was administered as a single topical dose, and eye tissue samples (the cornea, iris/ciliary body, retina, and choroid, n = 2 for each sample) were collected at 4 hours after ocular instillation of regorafenib eye-drops or pazopanib eye-drops.
In the rabbit and monkey studies, sample preparation and determination of drug concentrations were performed in the same manner for the rat pharmacokinetic study.

| Pharmacological study of regorafenib eyedrops in laser-induced rat CNV model
Pharmacological study in laser-induced rat CNV model was conducted by referring to and optimizing the previous research. 27 Animals with no ocular abnormalities were used for the study. The animals were arranged in ascending order of body weights and animals close to the median of the body weights were serially allocated into groups (13 animals per group).

| Induction of CNV
After the induction of mydriasis by instillation of Mydrin-P ophthalmic Solution has been confirmed, the animals were anesthetized by intramuscular injection with a mixture solution of ketamine hydrochloride/xylazine hydrochloride (7/1, v/v). The ocular fundus of the eye of each animal was observed using a slit lamp (SL-130; Carl Zeiss Meditec AG). Laser was irradiated on eight sites of the retina using a Multicolor Laser Photocoagulator (MC-300; NIDEK Co., Ltd.) to induce CNV. The laser irradiation conditions were as follows; wavelength: 532 nm, spot size: 80 μm, irradiation time: 0.05 seconds, and laser output: 120 mW.

| Administration of test articles
Regorafenib eye-drops and aflibercept were used as test articles.
Just after the laser irradiation (the day of the laser irradiation was the start date of repeated ocular instillation), vehicle eye-drops and regorafenib eye-drops at 21.2 mg/mL were topically administered twice daily to the eyes induced CNV for 14 days. Just after the laser irradiation, aflibercept (200 μg/eye) was administered by intravitreal injection once. Photographs of CNV sites were taken using a confocal microscope (ECLIPSE TE2000-U; Nikon Instech Co., Ltd.). The areas of CNV sites were calculated using ImageJ Software (National Institutes of Health).

| CNV evaluation
The CNV areas of the laser-irradiated eight sites/eye, excluding data unsuitable for evaluation (eg, irradiation sites could not be confirmed; ill-defined border; damaged sample), were used for analysis.
The mean value of CNV areas was calculated for each animal with at least three irradiation sites available for analysis. The mean value was regarded as individual representative value.

| Statistical analysis
Dunnett test was performed between in the vehicle eye-drops dosed group and the each test article dosed group. Stat Light #03 and Stat Light #04 (Yukms Co. Ltd., Kanagawa, Japan) were used these statistical analyses at a two-sided significance level of 5%.

| Pharmacological study of pazopanib eye-drops in laser-induced monkey CNV model
Pharmacological study in laser-induced monkey CNV model was conducted by referring to and optimizing the previous researches. [28][29][30] Animals were allocated to groups to minimize bias in the incidence and severity of CNV lesions and body weight between the groups.

| Induction of CNV
After the induction of mydriasis by instillation of Mydrin-P ophthalmic solution has been confirmed, animals were anesthetized by intramuscular injection with a mixture solution of ketamine hydro-

| Administration of test articles
Pazopanib eye-drops and aflibercept were used as test articles.

| Fluorescein fundus angiography and CNV grade evaluation
Fluorescein fundus angiography and CNV grade evaluation were

| Statistical analysis
Fisher's exact test was performed for the incidence of Grade 4 le-

| Ocular pharmacokinetics of regorafenib and pazopanib in rats
Regorafenib and pazopanib were administered to one designated eye of rats, and the concentrations in the choroid/sclera, retina and plasma were measured (Figure 1). In addition, the ocular and plasma pharmacokinetic parameters were calculated ( Table 1)

| Species difference of ocular pharmacokinetics of regorafenib and pazopanib in rats, rabbits, and monkeys
Regorafenib and pazopanib concentrations in the posterior eye tissues were evaluated in rats, rabbits, and monkeys after a single ocular instillation (Table 2). Surprisingly, the regorafenib and pazopanib concentrations in the posterior eye tissues of the dosed eye were much lower in rabbits and monkeys compared to rats.
Large species difference was observed in ocular pharmacokinetics of regorafenib and pazopanib after ocular instillation between animals.

| Ocular distribution of regorafenib and pazopanib in monkeys
More detailed ocular distribution including the anterior eye tissues were evaluated for regorafenib and pazopanib after a single ocular instillation to monkeys (Table 3)

| Laser-induced rat CNV model
The inhibitory effects of regorafenib on laser-induced CNV were evaluated in rats after repeated ocular instillations (21.2 mg/mL, twice daily, 14 days). Aflibercept was also tested as a positive control by a single intravitreal injection (200 μg/eye). In the aflibercept dosed group, CNV was significantly suppressed (P < .05). In the regorafenib eye-drops dosed group, significant reduction in CNV area was also observed (P < .05), and the efficacy was comparable to that of aflibercept ( Figure 2).

| Laser-induced monkey CNV model
The inhibitory effects of pazopanib (5.0 mg/mL, 4 times daily, 35 days) on laser-induced CNV were evaluated in monkeys after repeated ocular instillations ( Figure 3 and Table 4). Aflibercept (500 μg/eye) was also tested as a positive control by a single intra-

| Nano-crystal to increase drug delivery to the posterior eye tissues
In order to evaluate the impact of nano-crystalization on the ocular delivery to the posterior eye segments, regorafenib formulations were prepared with multiple mean particle sizes: 97.42, 233.8, and 6400 nm and used in the pharmacokinetic study in rabbits (Table 5)  formulations.

| D ISCUSS I ON
In this study, we evaluated interspecies differences in ocular pharmacokinetics after ocular instillation of regorafenib and pazopanib in rats, rabbits, and monkeys. In rats, regorafenib and pazopanib    each animal. In this study, we found out that there is a significant interspecies differences in total drug concentrations in the posterior eye tissues and pharmacological activities after ocular instillation of regorafenib and pazopanib among rat, rabbit, and monkey, though, further studies are needed in order to consider ocular pharmacokinetics and pharmaceutical activities in the multiple animals and also in human.
Nano-crystals, also regarded as 'nano-suspensions', are drug crystals with nanometer-size particles and can be applied to insoluble drugs. 25 Conventional ophthalmic formulations result in less than 5% bioavailability of dosed drugs to the eye tissues because of the ocular structural barriers and rapid draining by lacrimal fluid. 33 Nano-crystal technology can increase the adhesiveness to tissue surface, the dissolution rate, and apparent solubility of poorly soluble drugs. [23][24][25][26] Therefore, it is expected that nano-crystalization can improve drug delivery to the posterior segment of eye after ocular instillation of insoluble drugs, such as regorafenib. On the other hand, pazopanib is a soluble drug, so that is the reason why we judged that it was difficult to apply nano-crystal technology to pazopanib. We developed nano-crystallized regorafenib ophthalmic formulations with particle sizes ranging from around 100 nm to 230 nm, and studied to see if nano-crystallization is useful to increase the drug delivery of regorafenib to the posterior eye tissues. Although no exposure of regorafenib in the choroid/retina was observed with micro-crystal suspension (average particle size: 6.4 μm) at a higher concentration of 24.13 mg/mL in rabbits, de- In conclusion, significant species differences were observed in drug delivery to the posterior eye tissues after ocular instillation of regorafenib and pazopanib among rat, rabbit, and monkey, and it is possible that the insufficient exposure of these drugs in the posterior eye tissues was one of causes of the lack of efficacies in rabbits, monkeys, and also patients with neovascular AMD. Nano-crystal technology seems to be useful to increase drug delivery of regorafenib to the posterior segment of eye. Further studies are needed in order to consider causes of the lack of efficacy of these drugs in clinical and develop nano-crystalized regorafenib ophthalmic formulation which expresses enough efficacies in patients with neovascular AMD.

ACK N OWLED G EM ENTS
I am grateful to all the related staffs for technical assistance with the animal experiments. I would like to thank Kyowa Kirin Co., Ltd.
and University of Shizuoka for giving me to conduct the research.

D I SCLOS U R E
We state that there is no conflict of interest in our manuscript.