Intraocular Toxicity and Pharmacokinetics of Candesartan in a Rabbit Model

Abstract

Purpose.: To investigate the intravitreal toxicity and pharmacokinetics of candesartan, a selective type 1 angiotensin II receptor blocker, in rabbit eyes.

Methods.: For the toxicity study, 15 white rabbits were divided into three groups (five rabbits each). Different candesartan doses, namely 0.5, 1, and 2 mg in 0.1 mL, were injected into the vitreous of the right eye in each of the five rabbits. The vehicle solution was injected into the left eye as a control. ERG was recorded at 1, 3, and 7 days after injection. Retinal histology was examined by light microscope and transmission electron microscope. For pharmacokinetics analysis, one eye of each of the 30 rabbits received an intravitreal injection of candesartan (1 mg/0.1 mL). The concentration of candesartan in the vitreous was measured by a liquid chromatograph-triple quadrupole mass spectrometer at 12, 24, 36, 48, 60, and 72 hours after injection.

Results.: No significant difference in ERG was found between the study and the control eyes of the 0.5-mg group. The dark-adapted b-wave amplitudes decreased significantly at −10-dB intensities of stimulation in the 1-mg group. The b-wave amplitudes were significant at all intensities in the 2-mg group. Histologic studies revealed normal retinal morphology and structures in all eyes. The half-life of candesartan was 6.8 hours in the rabbit eyes.

Conclusions.: Intravitreal injection of 0.5 mg candesartan would be safe in the rabbit eyes. The half-life of candesartan was short in the vitreous, and modification of the delivery method would be required to extend the action duration for clinical applications.

Angiotensin (AT)-II induces vascular contraction and has a role in blood pressure control. AT-II controls the growth of vascular smooth muscle cells and has a proliferative effect on them. ¹ AT-II also stimulates the induction of various growth factors, including vascular endothelial growth factor (VEGF), which is associated with the induction of ocular neovascularization. ^{2 –4}  

Several studies have suggested that the upregulated renin-angiotensin system (RAS) may associated with the progression of diabetic retinopathy. ^{5 –8} Vitreous levels of AT-II and VEGF are higher in proliferative diabetic retinopathy (PDR) than nondiabetic retinopathy. ^5,6 Inhibition of the angiotensin converting enzyme (ACE) has been reported to be associated with a reduction of PDR, ^{9 –11} suggesting that inhibition of the RAS may prevent and treat diabetic retinopathy. 

Various ACE inhibitors and AT-II receptor blockers are being used for antihypertension medication. Although several studies were conducted to evaluate the impact of these medications for diabetic retinopathy, no significant efficacy has been shown ^{9 –11} except in the Diabetic Retinopathy Candesartan Trials (DIRECT) program, which studied the effect of oral candesartan, a selective type 1 AT-II receptor blocker, for diabetic retinopathy. ^{12 –14} Candesartan is considered the only drug to have a proven effect on diabetic retinopathy. However, efficacy was not found in the group with relatively severe diabetic retinopathy in spite of systemic medication of the large dose. Because an increase in the drug amount might bring about systemic side effects, ¹⁵ intravitreal injection is used to avoid these side effects and to increase intraocular concentration of the drug more effectively than oral medication. 

The toxicity and pharmacokinetics of candesartan in the eye after intravitreal injection have not been studied. In the present study, the retinal toxicity and pharmacokinetic profile after intravitreal injection of candesartan were studied in rabbits to provide safety information for future studies. 

Materials and Methods

Animal Preparations and Drug Administration

Forty-five New Zealand White rabbits (Covance, Princeton, NJ), each weighing approximately 2 kg, were used. All animals used in this study were treated in accordance with the institutional guidelines of Pusan National University and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

All injection procedures were performed as follows: rabbits were anesthetized with intramuscularly administered tiletamine plus zolazepam (Zoletil, Virbac, France) 25 mg/1 mL, and ventilated room air. The pupils were fully dilated with 2.5% phenylephrine and cyclopentolate hydrochloride (1%). After instillation of 0.5% proparacaine for topical anesthesia, intravitreal injections were performed using a 26-gauge needle attached tuberculin syringe. The needle was inserted 2 mm posterior to the limbus and was advanced toward the center of the vitreous until the tip was viewed with an operating microscope. A volume of 0.1 mL was slowly injected. After the injection, the fundus was observed for arterial pulsation. 

For the toxicity study, 15 rabbits were divided into three groups (five rabbits each). Candesartan (Tecoland Co., Edison, NJ) was dissolved in 20% ethyl alcohol and 80% normal saline. Different candesartan single doses, namely 0.5 mg, 1 mg, and 2 mg in 0.1 mL, were injected into the vitreous of the right eye in each of five rabbits. As a control, the vehicle solution was injected into the left eye of each animal. 

For pharmacokinetics analysis, one eye of each of the 30 rabbits received an intravitreal single injection of candesartan (1 mg/0.1 mL), and the other eye received no injection. 

Electroretinography

Electroretinograms (ERGs) were recorded in each of the five rabbits at 1 day, 3 days, and 7 days after intravitreal injection using a commercial ERG system (RETIport32; Roland Instrument, Brandenburg, Germany). The rabbits were placed in a dark room for adaptation for 1 hour. Pupils were dilated by instillation of 2.5% phenylephrine and 1% cyclopentolate hydrochloride. Reference and ground electrodes were placed on the lateral canthus and the ear, respectively, after shaving. Active electrodes (ERG Jet; Fabrinal SA, La Chaux-de-Fonds, Switzerland) were put on the cornea, and the head was positioned in the Ganzfeld dome. Dark-adapted ERGs were recorded with stimulation by −25- to 0-dB white flashes (increasing by 5 dB; 0.0095–3.004 cd · s/m²). Signals were amplified with a bandpass of 1 to 300 Hz. Amplitudes and implicit times of the a-wave and b-wave were measured and compared between the tested eyes using paired t-test. 

Histologic Examination

For histologic studies, the rabbits were euthanatized with an overdose of intravenous thiopental sodium after ERG recording at 7 days, and the eyes were enucleated. The anterior segment was removed, and a full-thickness specimen including the retina and sclera was obtained at the same distance inferiorly from the optic disc. The tissues were prefixed with 2.5% glutaraldehyde (4°C; phosphate buffer, pH 7.2) and were postfixed with 1% osmium tetroxide in the same buffer. The materials were dehydrated with a series of the graded ethyl alcohol and were embedded in epoxy resin (Epon 812 mixture). Thick sections (1 μm) were stained with 1% toluidine blue for light microscopy. Thin sections (50∼60 nm) were prepared by using an ultramicrotome (Reichert SuperNova; Leica, Wetzlar, Germany) and were double stained with uranyl acetate and lead citrate. Thin sections were examined with a transmission electron microscope (JEM 1200EX-II; JEOL). 

Pharmacokinetic Analysis

The injected eyes of five rabbits in each group were enucleated at 12, 24, 36, 48, 60, and 72 hours each for sampling of the drug. Eyes were frozen at −80°C immediately after enucleation. Whole vitreous was isolated from each eye by dissection of ocular tissue. The vitreous was mixed with 3 mL methanol, the mixture was centrifuged with 12,000 rpm for 10 minutes at 4°C, and the resultant supernatant was filtered with 0.22-μm micropore. Samples of 25 μL were analyzed using a liquid chromatograph-triple quadrupole mass spectrometer (6410 Triple Quadrupole LC/MS; Agilent Technologies, Palo Alto, CA). C-18 columns (Capcell Pak, 4.6 × 250 mm; Shisheido, Tokyo, Japan) were used. The flow rate used was 1 mL/min, and the mobile phase consisted of acetonitrile and water (0.1% formic acid). 

Pharmacokinetic parameters such as half-life, area under curve (AUC), mean residence time (MRT), and clearance were determined (WinNonlin Pro version 5.2; Pharsight, Mountain View, CA). 

Statistical Analysis

Data are expressed as mean ± SD. Statistical analyses were carried out using a paired t-test. P ≤ 0.05 was considered statistically significant. 

Results

After the injections, indirect ophthalmoscopy was used to view the injected candesartan, and white crystalline materials were observed dispersing in the vitreous. The crystalline material decreased at 1 day and was not observed at 3 or 7 days after 0.5-mg injection (Fig. 1). In the 2-mg group, some crystalline materials were observed in the inferior vitreous at 7 days after injection. 

Figure 1.

 

The crystalline of candesartan (arrow) is present in the inferior vitreous 1 day after a 0.5-mg intravitreal injection.

Electroretinography

No significant difference was found on ERGs between the experimental and the control eyes of the 0.5-mg group (Fig. 2). The dark-adapted, b-wave amplitudes decreased significantly at −10-dB intensities of stimulation in the 1-mg group at 1 and 3 days after injection (Fig. 3). ERG changes were significant at all intensities in the 2-mg group at days 1, 3, and 7 (Figs. 2, 3). 

Figure 2.

 

Dark-adapted ERG, elicited by light stimuli of different intensities from −25 to 0 dB, were recorded. Experimental (A) and control (B) eyes at 7 days after 0.5-mg intravitreal injection of candesartan. Experimental (C) and control (D) eyes 7 days after 2-mg injection. B-wave amplitudes (C) were decreased more than in the control eyes.

Figure 3.

 

Response-intensity data for the dark-adapted ERG b-wave. The amplitudes of the b-waves (μV) were compared by paired t-test. ●, experimental eyes; ■, control eyes. *P < 0.05.

There were no statistically significant differences between the implicit times of the a-waves and b-waves and between the amplitudes of the a-waves in the all groups (Fig. 4). 

Figure 4.

 

Response-intensity data for the dark-adapted ERG a-wave. The amplitudes of the a-waves (μV) were compared by paired t-test. There were no statistically significant differences at amplitude of a-wave in the all groups. ●, experimental eyes; ■, control eyes.

Histologic Examination

Retinal histology was examined by light microscope and transmission electron microscope. Histologic examination of all experimental eyes and control eyes showed normal retinal structures (Figs. 5, 6). 

Figure 5.

 

Photomicrographs of toluidine blue–stained sections of experimental retinas at 7 days after intravitreal injection of 0.5 mg (A), 1 mg (B), and 2 mg (C) candesartan. (D) Control retina (original magnification, ×400). The retinal ganglion cell layer is facing upward in all photographs. Micrograph of an experimental eye shows the retina and choroids are normal, with no inflammatory response or toxicity to the photoreceptor.

Figure 6.

 

Electron micrographs of experimental retinas at 7 days after intravitreal injection of candesartan. Photomicrographs show ganglion cell and axons with normal structures (A, 2-mg group; B, control group; original magnification, ×5000; scale bar, 1 μm) and normal inner nuclear and inner plexiform layers (C, 2-mg group; D, control group; original magnification, ×2000; scale bar, 2 μm). (E) 2-mg group. Normal photoreceptor outer segments and retinal pigment epithelial layer compared with control group (F; original magnification, ×6000; scale bar, 1 μm).

Pharmacokinetic Analysis

The mean concentration of candesartan in the vitreous was 10.29 ± 1.79 μg/mL at 12 hours, 2.86 ± 0.67 at 24 hours, 1.22 ± 0.29 at 36 hours, 0.21 ± 0.05 at 48 hours, 0.10 ± 0.01 at 60 hours, and 0.02 ± 0.01 at 72 hours after injection. Changes in concentration in the vitreous over time are illustrated in Figure 7. Candesartan was not detected in the vitreous of the noninjected eye. 

Figure 7.

 

Candesartan concentration in vitreous, assessed by liquid chromatograph-triple quadrupole mass spectrometer over time after a 1-mg intravitreal injection (A). Logarithmic plotting of the average concentration of candesartan with time (B). Data were expressed as mean ± SD.

The calculated half-life, AUC_0-∞, MRT, and clearance of candesartan were approximately 6.83 hours, 398.0 h · μg/mL, 7.66 hours, and 2.51 mL/h, respectively (Table 1). 

Table 1.

 

Pharmacokinetic Parameters of Candesartan 1 mg in the Vitreous of Rabbit Eyes

Half-Life (h)	AUC0−∞ (h · μg/mL)	MRT (h)	Clearance (mL/h)
6.83	398.0	7.66	2.51

 AUC, area under curve; MRT, mean residence time.

Discussion

The RAS is known as a hormone system that regulates blood pressure, vascular contraction, and water balance. When blood volume is low, the kidneys secrete renin. Renin stimulates the production of AT-I from angiotensinogen. ACE converts AT-I to AT-II. AT-II causes blood vessels to constrict, resulting in increased blood pressure. AT-II also stimulates the secretion of the hormone aldosterone from the adrenal cortex. 

Recently, the RAS has been reported to be present in local tissues playing roles other than those mentioned. The eye also contains all the major components of the RAS. Angiotensinogen is present in the nonpigmented ciliary epithelium. ¹⁶ Renin, ACE, chymase, and AT-II type 1 receptors are expressed in the retina. ^{17 –19}  

VEGF and AT-II are increased locally in the diabetic eye. ^{5 –8,20} These growth factors induce angiogenesis ^20,21 and can increase the risk for PDR. They also increase the retinal vascular permeability, ²² with the associated risk of macular edema. In particular, AT-II appears to potentiate VEGF-induced angiogenesis. ⁸ Funatsu et al. ⁵ found a highly significant correlation between the levels of AT-II and VEGF in the vitreous fluid in patients with PDR undergoing vitrectomy. Inhibition of AT-II formation with ACE inhibition results in the suppression of VEGF expression experimentally ²³ and clinically. ²⁴  

The importance of RAS inhibition to reduce the risk for diabetic retinopathy in humans is suggested in several studies. ^{9 –11} However, the clinical efficacy of RAS inhibition in diabetic retinopathy was not satisfactory compared with that in the other vascular events, such as occur in the heart, brain, and kidney of diabetic patients. Although lisinopril, an ACE inhibitor, demonstrated some efficacy in diabetic retinopathy, it was negated after correction of HbA_1C difference at baseline. ¹¹  

Recently, published reports ^13,14 of the Diabetic Retinopathy Candesartan Trials (DIRECT) program showed promising results. The DIRECT studies randomly assigned 5231 patients with type 1 or type 2 diabetes into groups receiving placebo or 32 mg candesartan daily. The result of the DIRECT-Prevent 1 study was that candesartan significantly reduced the incidence of diabetic retinopathy in type 1 diabetes, even after adjustment of changes in blood pressure during the trial. ¹³ The DIRECT Protect-2 study also showed significant regression of retinopathy in patients with type 2 diabetes and mild diabetic retinopathy. ¹⁴ Based on the evidence revealed in these studies, it is suggested that candesartan should be considered in hypertensive patients with diabetes. ²⁵  

However, DIRECT studies failed to show a significant difference in the progression of diabetic retinopathy, and no regression of retinopathy was noted in patients with moderate to moderately severe diabetic retinopathy. Furthermore, candesartan did not reduce the progression to proliferative retinopathy or the incidence of macular edema. The inferior results in retinopathy compared with other macrovascular events might be caused by the lower penetration of systemic medication into the local RAS on the eye because of the blood-retinal barrier. The 32-mg dose of candesartan cilexetil is already a large dose for systemic medication, and further increases may raise risk side effects such as vasovagal episode, peripheral edema, and atrial arrhythmia. ¹⁵ Hence, intravitreal injection of candesartan can be an option to increase the efficacy of managing diabetic retinopathy. 

A study of intravitreal injection of RAS inhibitor has not been reported. In our study, we evaluated retinal toxicity and the pharmacokinetics profile after intravitreal injection of candesartan. Intravitreal injection, which has a relatively low risk of complication, is commonly practiced to treat choroidal neovascularization, diabetic macular edema, and macular edema caused by retinal vein occlusion. 

Candesartan is a selective type 1 AT-II receptor blocker with no agonist activity. Candesartan cilexetil is used orally for hypertension treatment. It is a well-tolerated and effective antihypertensive drug. In addition, it dissociates slowly from the AT-II type 1 receptor and causes a long-lasting inhibition of the AT-II–mediated vascular contraction. ²⁶ To enhance bioavailability, a prodrug, candesartan cilexetil, is administered and is rapidly and completely converted to the active compound during absorption from the gastrointestinal tract. ¹² We used candesartan instead of candesartan cilexetil in our study because absorption from the vitreous after intravitreal injection is not needed to convert the active compound. 

In our study, intravitreal injection of 0.5 mg candesartan showed no ERG changes in the rabbit eyes. Compared with 16 mg candesartan cilexetil of the common daily oral dose for humans, 0.5 mg candesartan is a large dose. On the assumption of equal distribution, the dose of 0.5 mg in the eye is approximately 200 times higher than that after taking 16 mg orally. These results imply that intraocular toxicity of candesartan is very low. In the other aspect, a much higher dose of candesartan may be delivered to the target tissues, including the retina and the retinal vessels, by intravitreal injection rather than by oral medication. However, the half-life of candesartan of 6.8 hours in our study seems short compared with that of triamcinolone, ²⁷ bevacizumab, ²⁸ and ranibizumab. ²⁹ Further studies about the delivery system for a long duration of action of candesartan would be required. 

In summary, intravitreal injection of 0.5 mg candesartan was safe in rabbit eyes. However, an adequate therapeutic level for the suppression of the intraocular RAS system and the efficacy of treating diabetic retinopathy should be evaluated. The half-life of candesartan was short in the vitreous. Other delivery systems for a long duration of action would be required for clinical applications. 

Footnotes

 Disclosure: J.E. Lee, None; D.W. Lim, None; H.J. Park, None; J.H. Shin, None; S.M. Lee, None; B.S. Oum, None