- Email: [email protected]
Safety and pharmacodynamics of suprachoroidal injection of triamcinolone acetonide as a controlled ocular drug release model
Journal of Controlled Release 203 (2015) 109–117
Contents lists available at ScienceDirect
Journal of Controlled Release journal homepage: www.elsevier.com/locate/jconrel
Safety and pharmacodynamics of suprachoroidal injection of triamcinolone acetonide as a controlled ocular drug release model Mei Chen a, Xiaoli Li a, Jinkun Liu a, Yin Han a, Lingyun Cheng a,b,⁎ a b
Institute of Ocular Pharmacology, School of Ophthalmology and Optometry, Wenzhou Medical University, 270 Xueyuan Road, Wenzhou, Zhejiang 325027, China Jacob's Retina Center at Shiley Eye Center, Department of Ophthalmology, University of California San Diego, 9415 Campus Point Drive, La Jolla, CA 92037-0946, United States
a r t i c l e
i n f o
Article history: Received 1 December 2014 Received in revised form 7 February 2015 Accepted 15 February 2015 Available online 17 February 2015 Chemical compounds studied in the article: Triamcinolone acetonide (PubChem CID: 6436) Indocyanine green (PubChem CID: 11967809) Saline (PubChem CID: 5234) Lipopolysaccharide (PubChem CID: 11970143) Haemotoxylin (PubChem CID: 10603) Eosin (PubChem CID: 173745) Formalin (PubChem CID: 712) Paraffin (PubChem CID: 109453) Pentobarbital sodium (PubChem CID: 16219847) Xylazine hydrochloride (PubChem CID: 68554) Keywords: Suprachoroidal injection Injection volume Drug formulation Triamcinolone acetonide (TA) Indocyanine green (ICG) Pharmacodynamics Experimental uveitis Rabbit eyes
a b s t r a c t Suprachoroidal injection is an emerging technique for drug delivery to the posterior segment, which is hard to reach by non-invasive approaches. However, the injection technique varies and the associated ocular safety is not well understood. In addition, it is not clear if drug formulation is a major factor in optimizing pharmacodynamics using this technique. The current study was designed to compare the suprachoroidal injection of different drug formulations and to characterize the safety and pharmacodynamics of triamcinolone acetonide (TA) delivered by this technique. Both indocyanine green (ICG) solution and TA suspension, at 50 μL, 100 μL, and 150 μL, were suprachoroidally injected and intraocular pressure (IOP) tonometry, fundus photography, and electroretinography were performed over multiple time points up to eight weeks. After 50 μL TA (Kenalog-40) suprachoroidal injection, 4–5 animals at 7 time points were sacrificed for aqueous, vitreous, retina, and plasma collections. TA was quantitated using ultra-performance liquid chromatography tandem mass spectrometry. For comparative efficacy study, 50 μL (2 mg) suprachoroidal TA versus 20 mg subtenon TA were performed 4 weeks before induction of experimental uveitis with 10 ng of intravitreal lipopolysaccharide. After suprachoroidal injection, IOP had an acute elevation, higher volume caused higher IOP (p b 0.0001). Equivalent volume of ICG solution led to a significantly smaller IOP elevation than after TA suprachoroidal injection. This finding suggests better distribution of ICG solution than TA suspension in the suprachoroidal space. Following a 50 μL suprachoroidal injection, peak TA concentration in the aqueous was below 1 ng/mL. In contrast, the posterior vitreous and retina had 1912 ng/mL and 400,369 ng/mL TA, respectively. Maximum TA in plasma was 11.6 ng/mL. Drug exposure to the posterior retina was 523,910 times more than that to the aqueous and 29,516 times more than systemic TA exposure. In the treatment of lipopolysaccharide-induced uveitis, compared with 20 mg subtenon injection, suprachoroidal 2 mg TA demonstrated much better efficacy with significantly less aqueous humor cells and lower vitreous opacity scores (p b 0.05). Histology showed much less vitreous inflammation in the suprachoroidal injection group (p b 0.0001). It seems that a 50 μL suprachoroidal injection of TA was well tolerated in rabbit eyes and demonstrated excellent penetration into the posterior retina, providing better therapeutic effect than subtenon 20 mg TA. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Retinal diseases such as diabetic retinopathy or macular degeneration are chronic and refractory. Treatment of these diseases is challenging due to their chronic nature, which mandates frequent intravitreal injections of therapeutics or even a surgical procedure to implant a sustained drug delivery device at the vitreous base [1,2]. Intravitreal drug delivery is invasive and exposes patients to risks of retinal drug toxicity [3], lens injury and vitreous hemorrhage [4], and even intraocular infections [4]. Subconjunctival or subtenon injection of therapeutic ⁎ Corresponding author at: Institute of Ocular Pharmacology, School of Ophthalmology and Optometry, Wenzhou Medical College, 270 Xueyuan Road, Wenzhou, Zhejiang 325027, China. E-mail address: [email protected] (L. Cheng).
http://dx.doi.org/10.1016/j.jconrel.2015.02.021 0168-3659/© 2015 Elsevier B.V. All rights reserved.
agents is considered to be safer than intravitreal injection and these transscleral drug delivery routes have been shown to be more effective than systemic drug administration [5]. Though transscleral drug delivery is a better option than systemic drug administration, it is less effective than intravitreal injection in the treatment of various macular edemas [6]. Drugs placed on the sclera's surface encounter vigorous clearance from conjunctival and episcleral blood and lymphatic circulation [7]. In addition, the sclera itself is a major barrier for drugs to penetrate and reach the retina. To overcome the blockade of the sclera, several groups have investigated suprachoroidal drug delivery [8]. To date, suprachoroidal drug delivery seems to be feasible; however, the safety of suprachoroidal drug delivery remains controversial depending on the therapeutic agent and technique. Eimnabl et al. [9] reported that suprachoroidal injection of poly-(ortho ester) (POE) or 1% sodium hyaluronate led to fundus pigment disturbance and retinal atrophy.
110
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
three escalating volumes, 50 μL, 100 μL, and 150 μL because human vitreous volume (4.5 mL) is 3 times more than rabbit vitreous (1.5 mL). We hypothesize that a drug solution would behave differently than a drug suspension (Kenalog-40) during and after injection into the suprachoroidal space. In the current study, indocyanine green (ICG) was used to model a drug solution for its easy visual identification and Kenalog-40 as a suspension.
Suprachoroidal cannulation and drug delivery were also reported to be damaging to the outer retina, including photoreceptors, in the primate eye model [10]. The most recent studies from Tyagi et al. and Patel et al. were encouraging and a good safety profile was depicted using a 34 gauge needle or a microneedle for the suprachoroidal injection [11, 12]. From the current literature, efficiency of drug diffusion to the retina after suprachoroidal injection varies from drug to drug. For example, the amount of Avastin reaching the retina from suprachoroidal delivery was limited and cleared more quickly compared with intravitreal delivery [13]. In contrast, sodium fluorescein seems to behave differently [11, 12]. Though suprachoroidal delivery may be different from subtenon and intravitreal routes, a sustained drug delivery formulation is equally important for all these ocular delivery routes. Patel et al. demonstrated that suprachoroidally delivered dextran or Bevacizumab was cleared within a day while 1 μm sized polymeric particles remained in the eyes for over 2 months [11]. To date, triamcinolone acetonide seems to be the most promising drug formulation for suprachoroidal delivery to treat retinal diseases due to its low solubility and sustained release property [10,14]. However, no long-term safety and pharmacodynamics study of non-cannulated suprachoroidal TA is available yet. The current study was designed to evaluate the feasibility and safety of various volumes of TA suprachoroidal injection and to characterize the pharmacokinetics of the vitreous and retina as well as therapeutic efficacy in an experimental uveitis model using subtenon TA injection as a comparative modality.
All rabbits were anesthetized by an intramuscular injection of pentobarbital sodium (20 mg/kg) and xylazine (4 mg/kg). In addition, 0.5% proparacaine ophthalmic drops (Alcon) were used topically. Under the direct view of a surgical microscope (F18; Leica, Wetzlar, Germany), superonasal conjunctival peritomy was performed with a radial cut along the superior rectus muscle to expose the sclera. A 30 gauge needle was fit into an extension tubing (extension set 15.2 cm ID 8 mm OD 1.6 mm, EAGLE LABS, Rancho Cucamonga, CA) which was then connected to a 250 μL Hamilton syringe (HAMILTON). The Hamilton syringe was loaded into a repeating dispenser with which each actuation delivers 1/50 of the total capacity of that syringe [23]. The injection was made 9 mm behind the limbus at the globe meridian of 2 o'clock for the right eye and 11 o'clock for the left eye. After the procedure, the conjunctiva and Tenon were restored by an 8-0 suture and ofloxacin eye ointment was applied after the last examination.
2. Materials and methods
2.4. Safety study
2.1. Study design
For ICG suprachoroidal injection study, our purpose is to compare how a solution differs from a suspension like TA in terms of easiness of injection. In addition, we wanted to estimate the distributing area of the injected within the suprachoroidal space after a solution injection versus a suspension injection. We are not interested in knowing if ICG itself is toxic. Therefore, two eyes of each animal were used to reduce animal usage. For suprachoroidal TA injection, only one eye of each animal was used and the fellow eye was left untouched for control purpose. Following suprachoroidal injection, the eyes were monitored by ocular tomometry, indirect ophthalmoscopy, fundus photography, electroretinography (ERG), and histology after sacrifice as previously reported [24].
Intravitreal injection is invasive and the associated risks can be escalating with repeated injections. Vitreous itself has no blood vessels and is an immunologically privileged space. Therefore, it is prone to opportunistic infections. Development of a safer ocular drug delivery route has become an unmet need. Our group has been focusing on periscleral drug delivery which includes transscleral [15–19] and suprachoroidal routes. The current study was designed to compare traditional subtenon drug delivery with the emerging technique of suprachoroidal drug delivery in the context of medical care of posterior segment eye diseases. Thus far, when comparing suprachoroidal drug delivery with intravitreal or transscleral drug delivery, drug level was computed from the mixture of choroid and retina for suprachoroidally injected eyes [20,11,12]. This can lead to an inflated drug level because the drug delivered to suprachoroidal space was sampled into the complex of choroid and retina. In the current study, only the vitreous and retina were individually sampled for the drug quantitation. The retina is distinctively separated from the choroid by the durable Bruch's membrane that provides a natural barrier to avoid cross-contamination between the retina and the choroid during the dissection [21,15,22]. For the in vivo studies, Japanese big-eared white rabbits were used and animal handling was in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Visual Research. The body weights of the rabbits were between 2.2 kg and 3.5 kg with a mean of 2.7 kg ± 0.35 kg. Both genders were randomly used. 2.2. Determination of injection volumes The suprachoroidal cavity is a virtual space that is not present in the normal eye. It is important to know how much fluid can be safely injected into this virtual space without it flushing back and without causing sustained high intraocular pressure. The current study uses the rabbit eye as a model for human eye application. Using intravitreal injection volume as a reference in which 50 μL can be safely administered into rabbit vitreous without paracentesis, we elected to test
2.3. Suprachoroidal injection procedure
2.5. Pharmacokinetic study Twenty-eight Japanese big-eared white rabbits were used for this study and only one eye of each rabbit was injected. After suprachoroidal injection of 50 μL of triamcinolone acetonide suspension (2 mg), 4 to 5 rabbits were sacrificed for sampling of the eye tissue at each time point (1-day, 3-day, 1-week, 2-week, 4-week, 6-week, and 8-week). Following globe nucleation, the globe was snap frozen as we described previously [15,21] and the globe was cut into three sections. The first cut was through a circumferential line at pars plana 2 mm from the limbus, and the second cut was through the midline of globe equator. The aqueous, anterior vitreous, anterior retina, posterior vitreous, and posterior retina were dissected into individual pre-weighed vials and kept at − 80 °C until drug quantitation. Each cut was made with a new 23 blade and the cutting was done from the opposite of the suprachoroidal injection site towards the injection site to avoid cross-contamination. Blood was also sampled from the ear vein at each time point for plasma TA quantitation. Fellow eyes were not analyzed because the drug source for these eyes would be from blood circulation and TA in the blood sample was very low. Pharmacokinetics of subtenon TA injection was not concurrently performed because such a study was already conducted and the results were available in the literature [15].
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
2.6. Comparative efficacy study Twelve Japanese big-eared white rabbits were used for this study. The study was designed to test and compare the therapeutic sustainability of TA in the retina and vitreous following suprachoroidal space injection versus traditional subtenon injection. The dose used for suprachoroidal injection was 2 mg in 50 μL (determined from the preceding study) and was 20 mg for subtenon injection, which is the commonly used dose in clinical practice [25,26]. Four weeks after the TA injection, all rabbits' right eyes were intravitreally injected with 10 ng of lipopolysaccharide (InvivoGen, San Diego, CA) in 50 μL of saline through the pars plana (1.5 to 2 mm from limbus). The left eyes were untouched. Vitritis was induced at 4 weeks because the preceding pharmacokinetics showed TA concentration well above therapeutic level at this time point after suprachoroidal injection, and lower drug level after subtenon injection [15]. At 24 h, 48 h, and 72 h aqueous cells were graded under slit-lamp biomicroscope using standardization of uveitis nomenclature for reporting clinical data [27]. At the same time points, vitreous haze was graded from fundus photographs using a 0-8 grading scale [28]. After the last examination at 72 h and immediately before sacrifice, 50 μL of aqueous was sampled using a 30 gauge needle attached to a 1 mL syringe under direct view of a surgical microscope. Twenty microliters of the aqueous was used to create smear for inflammatory cell count. After sacrifice the eye globes were fixed in 10% formalin for paraffin histology processing. Five micrometer paraffin sections were prepared for HE (Hematoxylin & Eosin) staining and light microscopy.
2.7. Statistical analysis The categorical data such as incidence of the complications were tabulated and analyzed with multivariate logistic regression to model the important predictors for the likelihood of complications. In addition, a trend test was performed by coding the three doses into ordinal scales. For continuous data such as intraocular pressure (IOP), multivariate linear regression was performed to model the predictors, such as formulation and volume, and their interaction. For ERG data, since only one eye of each rabbit was injected with TA, paired t test was used for each type of ERG to compare the treated eye with its fellow eye. For aqueous humor inflammatory cell count and vitreous haze grading, the count or grading was conducted three times, once a day, generalized estimating equation (GEE) was used within the SAS (version 9.4) environment to compare inflammation between the two groups while adjusting for the evaluating time. Similarly, inflammatory cell counts from aqueous humor smears and vitreous on HE stained sections at multiple locations
111
were analyzed using GEE. All analyses were performed using either JMP 11 or SAS 9.4 statistical software (SAS Institute Inc, Cary, NC). The pvalues smaller than 0.05 were considered to be significant. 3. Results 3.1. Assessing optimal injection volumes Our hypothesis was that a solution may differ from a suspension in aspects of injection and associated complications. Both ICG and TA were tested at volumes of 150 μL, going down to 100 μL, then 50 μL. From a pilot study using 150 μL we found that IOP was too high to be measured by a Tonopen; therefore a Schiotz tonometer was used throughout the study. The complications observed include backflow from injection entry, needle injury causing choroidal hemorrhage, choroidal capillary bleeding (Choroid hemorrhage away from needle entry), and serous sensory retina elevation from penetrating through Bruch's membrane (Fig. 1). The incidence of these complications is summarized in Table 1. Logistic regression using volume and agent as predictors found that volume was a significant contributor to these complications but agent was not (Table 1). Further statistical analysis showed that all three complications (backflow, p = 0.0003; localized serous retinal elevation, p = 0.0002; choroid hemorrhage away from needle entry, p b 0.0001) had a higher incidence with higher injection volumes (Cochran Armitage Trend Test). After these complications were pooled, the logistic regression for independent variables of agent and volume both became significant predictors. The drug suspension (TA) tended to have a higher rate of complications (p = 0.0192, nominal logistic fit) after the volumes were adjusted. 3.2. Intraocular pressure (IOP) assessment After suprachoroidal injection the eyes were monitored for their IOP change over time (Fig. 2). For both solution (ICG) and suspension (TA), the magnitude of IOP increase was significantly associated with the injection volume; higher volume caused higher IOP (p b 0.0001, multivariate linear regression). In addition, IOP rise after suprachoroidal injection of a drug suspension (TA) was significantly higher than after injection of a drug solution (ICG) (p b 0.0001, multivariate linear regression). 3.3. Area of ICG diffusion within suprachoroidal space To determine the area of ICG diffusion within the suprachoroidal space, an indirect ophthalmoscope and a 28-diopter Volk lens were
Fig. 1. The fundus images are showing the complications from suprachoroidal injections. The left frame is demonstrating localized serous retinal detachment (thin arrow) from injecting fluid due to the needle tip penetrating Bruch's membrane at some point. The middle frame is demonstrating superfacial choroidal hemorrhage, next to the optic nerve and below the medullary ray (thin arrow), purportedly caused from acute IOP change. The right frame is demonstrating deeper choroidal hemorrhage (thin arrow) at needle entry near the vortex vein (thicker arrow). The vortex veins can be seen at superior peripheral retina and indicated with the thicker arrows. The images are not crystal clear due to residual cornea edema from high IOP.
112
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Table 1 Injection volumes and complications. Injection agents
Nb
Volume
Backflow from needle entry
Localized serous retinal elevation
Choroid hemorrhage away from needle entry
Choroid bleeding at needle entry
ICG ICG ICG TA TA TA p Valuea
12 12 12 6 28 37
150 μL 100 μL 50 μL 150 μL 100 μL 50 μL
2/12 0/12 0/12 2/6 2/28 0/37 0.0026 (volumes); 0.99 (agents)
4/12 1/12 0/12 2/6 1/28 1/37 0.0036 (volumes); 0.90 (agents)
1/12 1/12 0/12 6/6 7/28 0/37 0.0003 (volumes); 0.99 (agents)
0/12 0/12 0/12 0/6 1/28 2/37 0.71 (volumes); 0.17 (agents)
ICG = indocyanine green. TA = triamcinolone acetonide. a Effect likelihood ratio tests. b N = numbers of eyes used.
used. The green area was measured by the number of disc diameters on each o'clock meridian (12 meridians) from the optic nerve towards the peripheral. Then the number of optic diameters at each meridian was summed for each eye. For this purpose, 3 rabbits and six eyes were used to measure the suprachoroidal ICG area by indirect ophthalmoscopy before sacrifice and then after sacrifice by a caliper under a dissecting microscope using the vertical diameter of the optic nerve head. Two observers conducted the measurements independently. The numbers from indirect ophthalmoscopy and caliper measurement were in good agreement (Overall Concordance Correlation Coefficient = 0.8959 with confidence intervals from 0.8425 to 0.9338, n = 72) [29]. On average, at the 12 o'clock meridian there are 5 disc diameters between the optic nerve and ora serrata; 6.5 disc diameters at nasal (at 9 for OD and at 3 for OS); 6.5 disc diameters at temporal (at 3 for OD and at 9 for OS); and 7 disc diameters at inferior meridian (6 O'clock). The average summed disc diameters of the 12 meridians was 75.6. This number was used as the total fundus area and was divided into the measurements of green area for each study eye to estimate the percentage of retina area stained by ICG as shown in Fig. 3. Roughly, 15%, 45%, and 65% of suprachoroidal space was filled with ICG following 50 μL, 100 μL, and 150 μL injections. The three sample images are demonstrated in Fig. 4.
3.4. Electroretinography (ERG) and histology ERG examination was performed on TA injected animals. The right eye of each rabbit received TA while the left eye was untouched to serve as control. A paired t-test was used to compare the amplitudes of right eyes versus left eyes by ERG type. No statistically significant difference was found (Fig. 5). In the histology analysis, ICG was identifiable in suprachoroidal space as dark green color and no abnormality was noted in TA injected eyes (Fig. 6). 3.5. Pharmacokinetic study Based on the preceding safety studies, 50 μL of TA suprachoroidal injection was used for the ocular pharmacokinetic study. The key pharmacokinetic parameters are summarized in Table 2 and the drug concentration–time curves are presented in Fig. 7. The triamcinolone in aqueous was below 1 ng/mL for the whole study period. In contrast, posterior vitreous and retina had peak concentrations of 1912 ng/mL and 400,369 ng/mL, respectively. The maximum TA in plasma was 11.6 ng/mL. Drug exposure to the posterior retina was 527,869 times more than that to the aqueous and 29,516 times more
Fig. 2. Intraocular pressure change over time following suprachoroidal injection of various volumes of ICG or TA. The elevated IOPs returned back to baseline or near baseline 20 min after the injection. Each error bar is constructed using 1 standard error from the mean.
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 3. Box plot of percentage of fundus involved with the suprachoroidal ICG at different injection volumes, depicting the ICG distribution within suprachoroidal space.
than systemic TA exposure (Table 2). The TA detected at the last time point of 8 weeks in the posterior retina was above 10 ng/g. 3.6. Comparative efficacy of pretreatment Four weeks after subtenon (20 mg) or suprachoroidal (2 mg) TA injection, 10 ng of intravitreal lipopolysaccharide was administered to cause clinically notorious uveitis in both groups. However, both anterior and posterior uveitides were more severe in the subtenon TA group. Aqueous humor cell grading was significantly higher in subtenon eyes (least square mean = 2.87) than that in suprachoroidal injected eyes (least square mean = 1.42, p = 0.0012, Fig. 8) and aqueous humor cells decreased significantly over 72 h (β = −0.04, p b 0.0001). Similarly, vitreous haze grading was significantly higher in subtenon group eyes (least square mean = 6.13) than in suprachoroidal group eyes (least square mean = 4.53, p = 0.04, Fig. 8); vitreous haze did not show significant change during the 72 hour study period (β = −0.007, p = 0.36) (Fig. 9). In addition, the infected eyes had significantly lower intraocular pressure than their fellow eyes (12.58 mm Hg, p b 0.0001) though there was no significant difference between the subtenon TA injected eyes (least square mean = 7.04 mm Hg) and suprachoroidal TA injected eyes (least square mean = 7.73 mm Hg, p = 0.46). IOP showed a statistically significant decrease over the 72 hour (β = −0.06, p b 0.0001). The inflammatory cells in aqueous humor were also counted from a smear. Seven spots were sampled for counting under 20 × objective
113
Fig. 5. The b-wave amplitude was compared between the right eyes (OD, TA injected eyes) and the left eyes (OS, control eyes) within each type of ERG (paired t-test).
lens and 10 × eyepieces. Each spot is a 20 × field. Four evenly spaced spots along the smear edge and three spots on the middle longitudinal axis were counted for cells. The same paradigm was used for each smear and counting was performed by a masked observer. The least square mean cell number for subtenon TA eyes was 23/field but only 9/field for suprachoroidal TA eyes (p = 0.0007). Histology and microscopy revealed that the main change was vitritis. Inflammatory cells were variably distributed in the vitreous cavity and posterior chamber. Vitreous inflammatory cells were counted at four locations: in front of the optic nerve, behind the lens, and superior peripheral vitreous and inferior peripheral vitreous. At each location, four random fields under a 40 × objective lens were photographed (64 ×) and the cells were counted. Statistics showed that significantly more inflammatory cells per field were present in the subtenon group than in the suprachoroidal group (132 versus 52, p b 0.0001, Fig. 10). 4. Discussion Suprachoroidal injection for drug delivery to the posterior segment is an emerging method. Understanding its delivery mechanism and possible lateral side effects is the first step to deploy this technology. So far, a few reports are available using either cannulation or proprietary hollow microneedle [11,13,30]. The current study demonstrates that suprachoroidal injection is feasible with a 30 gauge sharp needle attached to a Hamilton syringe loaded on a repeating dispenser. 50 μL of Kenalog-40 can be consistently delivered with minimal complication.
Fig. 4. Sample fundus images with different volumes of suprachoroidal ICG injection, taken shortly after the injection. The images are not crystal clear due to residual corneal edema from high IOP.
114
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 6. Microscopic images showing the retina (R), retinal pigment epithelium (RPE, shorter arrows), choroid (Ch), sclera (S), and the junction line between the choroid and sclera (longer arrows). The image in the left panel came from an eye injected with suprachoroidal ICG at 2-weeks after the injection; the image in the right panel came from an eye injected with suprachoroidal TA 2 weeks after the injection. The dark green ICG can be seen at the junction line between the choroid and sclera in the left panel (thin arrows). The image on the left panel was lightly stained with hematoxylin on purpose to reveal the ICG color.
The complications become more of a risk with the increase of injection volume. In addition, TA suspension seems to have more complications than ICG injection. However, it should be noted that there were no complications with ICG injections and only one serous retinal elevation out of 37 eyes injected with 50 μL of TA suspension. Thus far, the largest suprachoroidal volume tested in vivo was 100 μL injected into the rabbit eye [9]. In that study, local retinal atrophy after a 100 μL suprachoroidal sodium hyaluronate injection was noted on fundus photograph [9]. Olsen et al. cannulated pig suprachoroidal space and injected 70 μL of bevacizumab mixed with hyaluronic acid (50 μL bevacizumab + 20 μL of hyaluronic acid). Ocular pharmacokinetics was reported without mentioning any complication [13]. In the case of the pig, the vitreous volume is more than double the volume of rabbit vitreous [31]. From the current study, 50 μL TA did not show obvious adverse ocular effects except for one injection procedure related complication (localized serous retinal elevation at the needle tip). When the volume escalated to 100 μL, 7 out of 28 eyes showed choroidal hemorrhage away from the needle entry site. We postulate that this may be caused by acute IOP change because this type of hemorrhage was observed more frequently in 150 μL injections. In addition, this type of hemorrhage occurred less in ICG injected eyes and ICG induced IOP elevation was also significantly less than that from injection of TA. The difference of IOP elevation between these two agents may be due to poorer expansion and distribution of TA suspension in suprachoroidal space. The choroidal hemorrhage at the needle entry site is, we think, from choroidal
injury by the needle tip; the hemorrhage disappeared within 1 to 2 weeks without significant findings in histology. Compared with the current injection technique, choroidal cannulation may require much more training and cannulation itself is more traumatizing [10]. To improve the safety of suprachoroidal injection, a proprietary microhollow needle was fabricated and in an ex vivo study it could consistently deliver up to 35 μL into enucleated eye globes of rabbit, pig, and human. In a subsequent in vivo study using rabbit eyes, 50 μL of suspension could be delivered within 15 s and waiting an additional 30 s before removing the microneedle from the eye can prevent excessive reflux [11]. The microneedle's short length allows more reflux than a longer needle; however, with a short needle there is less risk of choroidal damage. The current study is the first comprehensive in vivo study to investigate the relationship between injection volume and IOP change, and the relationship between drug formulations (ICG vs. TA suspension). As expected, larger injection volumes induced higher IOP elevation. The elevated IOP returned to pre-injection level within 20 min except for the eyes injected with 150 μL. An interesting finding from the current study is that IOP elevation was significantly higher after the injection of TA suspension than after the injection of ICG even when the injection volumes were the same. This phenomenon may be explained by the postulation that ICG expands further within suprachoroidal space than TA suspension that expands more towards the vitreous and causes IOP elevation.
Table 2 Pharmacokinetic parameters following suprachoroidal TA injection. Ocular tissues
Rsq adjusted
Half-life (days)
Tmax (days)
Cmax (ng/mL or g)
SE_Cmax
Tlast (days)
Clast (ng/mL or g)
AUClast
MRTlast (days)
Aqueous Posterior vitreous Posterior retina Plasma
0.85 0.99 0.97 0.74
9.85 3.78 3.41 16.7
1.00 3.00 3.00 1.00
0.69 1911.62 400,368.55 11.6
0.19 605.39 117,120.49 7.01
28.00 56.00 56.00 56
0.10 0.18 11.18 0.87
7.94 22,016.71 4,191,281.90 141.65
8.93 8.20 7.66 17.02
Rsq adjusted: R square adjusted. Tmax: The time at which mean drug concentration was the maximum. Cmax: Mean maximum drug concentration. SE = standard error. Tlast = the last observed time point. Clast = concentration at the last observed time point. AUClast: Area under the curve calculated to the last observed time. MRTlast: Mean residence time calculated to the last observed time.
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 7. TA (triamcinolone acetonide) concentration–time curves of aqueous, posterior vitreous, posterior retina, and plasma.
The current pharmacokinetic study following 50 μL of suprachoroidal TA suspension demonstrated high TA concentration in both the retina and vitreous. In contrast, TA in aqueous humor was very low, below 1 ng/mL. During the whole study period (2 months), the TA exposure to aqueous humor (AUC = 7.94 ng·day/mL) was 2773 times lower than the AUC of posterior vitreous and roughly 527,869 times lower than the AUC of posterior retina. This can be a major advantage for suprachoroidal injection over subtenon injection because high TA in aqueous humor can cause steroid-related glaucoma and cataract. This point is supported by the fact that intravitreal TA causes significantly more IOP elevation than subtenon TA and the aqueous humor TA concentration in the former was much higher [32] than the latter [16]. After subtenon injection of 40 mg of TA, TA concentration in the aqueous and the vitreous was similar, with peak TA concentration in aqueous humor being 1244 ng/mL in the rabbit eye and 197 ng/mL in the human eye [15,16]. The TA systemic exposures following 2 mg suprachoroidal injection and following 40 mg subtenon injection in rabbit were similar [15]. These facts indicate that suprachoroidal TA injection is a better option that uses less drug, provides higher TA concentrations in the retina and vitreous, leads to minimal aqueous humor drug exposure, and has similar systemic TA exposure to subtenon injection. In addition, suprachoroidal injection should be less vulnerable to procedure related opportunistic infections than intravitreal injection because the vitreous is an immunologically privileged culture medium.
115
Suprachoroidal injection is an emerging method and its ocular safety and efficacy needs thorough inspection in preclinical studies. In addition, it is also expected that not every compound or formulation will fit this delivery route. Based on the literature available, a readily sustainable formulation seems to be the right fit for this technique [9,11,13]. Triamcinolone acetonide is a hydrophobic small molecule drug. Its suspension is commercially available and has been used as a sustained release formulation through intravitreal or subtenon injection to treat various chorioretinal disorders [33–35]. TA has been evaluated with suprachoroidal injection and Olsen et al. reported that TA remained in the local ocular tissue for at least 120 days following a 12 μL TA suspension suprachoroidal cannulation [10]. In that study, the pig eye was used and a more concentrated TA suspension (1.5 or 3 mg in 12 μL) was cannulated. At postcannulation day 60, much more TA (0.1 mg/g) was found in the retina compared with the current study (11 ng/g). The difference could stem from the use of different animal species as well as different TA concentration and injection methods. Another in vivo study by Gilger et al. evaluated the effect of TA administered into the suprachoroidal space (SCS) using a microneedle and compared its efficacy with intravitreal TA injections in a porcine model of acute posterior segment inflammation [14]. The acute ocular inflammation was induced in the pig 24 h after suprachoroidal TA injection. The current study was designed differently in that we aimed to test the efficacy of sustained drug release following suprachoroidal TA injection and used concurrent subtenon TA injection as a control. The study results were very encouraging. The results demonstrated the superiority of suprachoroidal 2 mg TA to subtenon 20 mg TA in reducing acute inflammation in a rabbit model of uveitis. This was the first pharmacodynamics study to evaluate suprachoroidal TA efficacy in a long-term setting. Key uveitis parameters, such as vitreous haze and inflammatory cell counts, were significantly reduced in the eyes pretreated with suprachoroidal TA. In summary, it seems to be feasible to consistently deliver 50 μL of TA suspension into the rabbit suprachoroidal space using a 30 gauge hypodermic needle with the help of Hamilton syringe and repeating dispenser. There was no backflow with this volume, but backflow was observed when delivering 100 μL and 150 μL. ICG seems to be easier to deliver than TA suspension, with less procedure-related complications and less IOP elevation. No functional abnormality was observed in any dose group by ERG examination. Ocular pharmacokinetics of TA delivered by suprachoroidal injection demonstrated much lower TA concentration in aqueous humor but much higher in posterior vitreous and retina than that from a traditional 40 mg TA subtenon injection. Suprachoroidal injection of 2 mg of TA provided stronger therapeutic effect in reducing acute experimental vitritis, even at 4 weeks following the initial single injection, than the traditional subtenon injection of 20 mg of TA did. Suprachoroidal injection may become a promising drug delivery method to the back of the eye for slow release drug formulations.
Fig. 8. Slit-lamp based microphotographs of the anterior segment. The left panel was from an eye in the suprachoroidal TA injection group and the right panel from an eye in the subtenon TA injection group. The aqueous humor was more turbid and exudates were present in the pupil area of the subtenon TA injected eye. In addition, due to inflammation the pupil was less dilated in the right image compared with the pupil in the left image.
116
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 9. This composite image is depicting various vitreous hazes from the vitritis induced by intravitreal lipopolysaccharide. The upper panel images are from a single eye with a suprachoroidal space (SCS) injection of 2 mg of TA and the lower panel images are from a single eye with a posterior subtenon (PST) 20 mg of TA injection. The vitreous clarity in the first image of the upper panel at 24 h (24 h, grade 3) is clearer than at 48 h (48 h, grade 5) and 72 h (72 h, grade 4). In general, vitreous clarity is significantly better in the upper panel images than in the lower panel images (24 h, grade 6; 48 h, grade 7; 72 h, grade 8).
Fig. 10. Microphotographs from HE stained histology sections, demonstrating inflammatory cell infiltration (asterisk) at the inferior vitreous cavity that is indicated by the ora sarrata (thin arrow) and part of the pars plana (thick arrow). Inflammatory cells were more significant in the eye with subtenon TA injection (right panel) compared with that in the eye with suprachoroidal TA injection (the left panel). The retinas of both eyes were relatively normal looking in anatomical structures.
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Acknowledgments Funding/support: National Natural Science Foundation of China, Grant No. 31271022. Statement of conflicts of interest The authors have no financial/conflicting interests in any aspect of this study. References [1] I. Krebs, L. Schmetterer, A. Boltz, R. Told, V. Vecsei-Marlovits, S. Egger, U. Schonherr, A. Haas, S. Ansari-Shahrezaei, S. Binder, M.R. Grp, A randomised double-masked trial comparing the visual outcome after treatment with ranibizumab or bevacizumab in patients with neovascular age-related macular degeneration, Br. J. Ophthalmol. 97 (2013) 266–271. [2] G.J. Jaffe, Reimplantation of a fluocinolone acetonide sustained drug delivery implant for chronic uveitis, Am J. Ophthalmol. 145 (2008) 667–675. [3] A.A. Seawright, R.D. Bourke, R.J. Cooling, Macula toxicity after intravitreal amikacin, Aust. N. Z. J. Ophthalmol. 24 (1996) 143–146. [4] A.E. Fung, P.J. Rosenfeld, E. Reichel, The International Intravitreal Bevacizumab Safety Survey: using the internet to assess drug safety worldwide, Br. J. Ophthalmol. 90 (2006) 1344–1349. [5] O. Weijtens, R.C. Schoemaker, A.F. Cohen, F.P.H.T.M. Romijn, E.G.W.M. Lentjes, J. van Rooij, J.C. van Meurs, Dexamethasone concentration in vitreous and serum after oral administration, Am. J. Ophthalmol. 125 (1998) 673–679. [6] M.A. Bonini-Filho, R. Jorge, J.C. Barbosa, D. Calucci, J.A. Cardillo, R.A. Costa, Intravitreal injection versus sub-Tenon's infusion of triamcinolone acetonide for refractory diabetic macular edema: a randomized clinical trial, Invest. Ophthalmol. Vis. Sci. 46 (2005) 3845–3849. [7] S.H. Kim, C.J. Galban, R.J. Lutz, R.L. Dedrick, K.G. Csaky, M.J. Lizak, N.S. Wang, G. Tansey, M.R. Robinson, Assessment of subconjunctival and intrascleral drug delivery to the posterior segment using dynamic contrast-enhanced magnetic resonance imaging, Invest. Ophthalmol. Vis. Sci. 48 (2007) 808–814. [8] M.R. Robinson, S.S. Lee, H. Kim, S. Kim, R.J. Lutz, C. Galban, P.M. Bungay, P. Yuan, N.S. Wang, J. Kim, K.G. Csaky, A rabbit model for assessing the ocular barriers to the transscleral delivery of triamcinolone acetonide, Exp. Eye Res. 82 (2006) 479–487. [9] S. Einmahl, M. Savoldelli, F. D'Hermies, C. Tabatabay, R. Gurny, F. Behar-Cohen, Evaluation of a novel biomaterial in the suprachoroidal space of the rabbit eye, Invest. Ophthalmol. Vis. Sci. 43 (2002) 1533–1539. [10] T.W. Olsen, X. Feng, K. Warner, S.R. Conston, D.H. Sierra, D.V. Folden, M.E. Smith, J.D. Cameron, Cannulation of the suprachoroidal space: a novel drug delivery methodology to the posterior segment, Am J. Ophthalmol. 142 (2006) 777–787. [11] S.R. Patel, D.E. Berezovsky, B.E. McCarey, V. Zarnitsyn, H.F. Edelhauser, M.R. Prausnitz, Targeted administration into the suprachoroidal space using a microneedle for drug delivery to the posterior segment of the eye, Invest. Ophthalmol. Vis. Sci. 53 (2012) 4433–4441. [12] P. Tyagi, R.S. Kadam, U.B. Kompella, Comparison of suprachoroidal drug delivery with subconjunctival and intravitreal routes using noninvasive fluorophotometry, Plos One 7 (2012). [13] T.W. Olsen, X. Feng, K. Wabner, K. Csaky, S. Pambuccian, J.D. Cameron, Pharmacokinetics of pars plana intravitreal injections versus microcannula suprachoroidal injections of bevacizumab in a porcine model, Invest. Ophthalmol. Vis. Sci. 52 (2011) 4749–4756. [14] B.C. Gilger, E.M. Abarca, J.H. Salmon, S. Patel, Treatment of acute posterior uveitis in a porcine model by injection of triamcinolone acetonide into the suprachoroidal space using microneedles, Invest. Ophthalmol. Vis. Sci. 54 (2013) 2483–2492. [15] K. Nan, S. Sun, Y. Li, J. Qu, G. Li, L. Luo, H. Chen, L. Cheng, Characterisation of systemic and ocular drug level of triamcinolone acetonide following a single sub-Tenon injection, Br. J. Ophthalmol. 94 (2010) 654–658. [16] L. Shen, Y. You, S. Sun, Y. Chen, J. Qu, L. Cheng, Intraocular and systemic pharmacokinetics of triamcinolone acetonide after a single 40-mg posterior subtenon application, Ophthalmology 117 (2010) 2365–2371.
117
[17] W. Du, S. Sun, Y. Xu, J. Li, C. Zhao, B. Lan, H. Chen, L. Cheng, The effect of ocular pigmentation on transscleral delivery of triamcinolone acetonide, J. Ocul. Pharmacol. Ther. 29 (2013) 633–638. [18] X. Liu, Y. Li, Y. Zhang, W. Du, S. Sun, B. Lin, H. Chen, L. Cheng, Comparison of intraocular pressure elevation after anterior versus posterior subtenon triamcinolone acetonide acetate injection: a retrospective study, Retina 32 (2012) 1838–1843. [19] Y. Meng, S. Sun, J. Li, K. Nan, B. Lan, Y. Jin, H. Chen, L. Cheng, Sustained release of triamcinolone acetonide from an episcleral plaque of multilayered poly-epsiloncaprolactone matrix, Acta Biomater. 10 (2014) 126–133. [20] M.Z. Wang, W. Liu, Q.J. Lu, H.Y. Zeng, S.M. Liu, Y.K. Yue, H.T. Cheng, Y. Liu, M. Xue, Pharmacokinetic comparison of ketorolac after intracameral, intravitreal, and suprachoroidal administration in rabbits, Retina J. Ret. Vit. Dis. 32 (2012) 2158–2164. [21] L.Y. Cheng, K.Y. Hostetler, J. Lee, H.J. Koh, J.R. Beadle, K. Bessho, M. Toyoguchi, K. Aldern, J.M. Bovet, W.R. Freeman, Characterization of a novel intraocular drugdelivery system using crystalline lipid antiviral prodrugs of ganciclovir and cyclic cidofovir, Invest. Ophthalmol. Vis. Sci. 45 (2004) 4138–4144. [22] H.Y. Wang, J. Chhablani, W.R. Freeman, J.R. Beadle, K.Y. Hostetler, K. Hartmann, L. Conner, K.A. Aldern, L. Pearson, L.Y. Cheng, Intraocular safety and pharmacokinetics of hexadecyloxypropyl-cidofovir (HDP-CDV) as a long-lasting intravitreal antiviral drug, Invest. Ophthalmol. Vis. Sci. 52 (2011) 9391–9396. [23] L. Cheng, M. Toyoguchi, D.J. Looney, J. Lee, M.C. Davidson, W.R. Freeman, Efficient gene transfer to retinal pigment epithelium cells with long-term expression, Retina 25 (2005) 193–201. [24] Y. Li, H. Chen, J. Hou, Y. Zhang, G. Li, X. Gu, L. Luo, H. Shen, J. Qu, L. Cheng, Further characterization of ocular safety profile of commercially available preserved and preservative-free triamcinolone acetonide, Retina 32 (2012) 364–374. http://dx. doi.org/10.1097/IAE.0b013e31821e1f7c. [25] J. Toda, H. Fukushima, S. Kato, Injection of triamcinolone acetonide into the posterior sub-tenon capsule for treatment of diabetic macular edema, Retina 27 (2007) 764–769. http://dx.doi.org/10.1097/IAE.1090b1013e318030bfcd. [26] A.A. Okada, T. Wakabayashi, Y. Morimura, S. Kawahara, E. Kojima, Y. Asano, T. Hida, Trans-Tenon's retrobulbar triamcinolone infusion for the treatment of uveitis, Br. J. Ophthalmol. 87 (2003) 968–971. [27] D.A. Jabs, R.B. Nussenblatt, J.T. Rosenbaum, Standardization of Uveitis Nomenclature Working Group. Standardization of uveitis nomenclature for reporting clinical data. Results of the First International Workshop, Am J. Ophthalmol. 140 (2005) 509–516. [28] J.L. Davis, B. Madow, J. Cornett, R. Stratton, D. Hess, V. Porciatti, W.J. Feuer, Scale for photographic grading of vitreous haze in uveitis, Am J. Ophthalmol. 150 (2010) 637–641 (e631). [29] S.B. Crawford, A.S. Kosinski, H.M. Lin, J.M. Williamson, H.X. Barnhart, Computer programs for the concordance correlation coefficient, Comput. Methods Prog. Biomed. 88 (2007) 62–74. [30] S.R. Patel, A.S.P. Lin, H.F. Edelhauser, M.R. Prausnitz, Suprachoroidal drug delivery to the back of the eye using hollow microneedles, Pharm. Res. Dordr. 28 (2011) 166–176. [31] A. Shafiee, G.L. McIntire, L.C. Sidebotham, K.W. Ward, Experimental determination and allometric prediction of vitreous volume, and retina and lens weights in Gottingen minipigs, Vet. Ophthalmol. 11 (2008) 193–196. [32] P.M. Beer, S.J. Bakri, R.J. Singh, W. Liu, G.B. Peters III, M. Miller, Intraocular concentration and pharmacokinetics of triamcinolone acetonide after a single intravitreal injection, Ophthalmology 110 (2003) 681–686. [33] R.A. Alshareef, S.J. Garg, J. Hsu, J. Vander, C. Park, M.J. Spirn, Intravitreal triamcinolone acetonide injection for macular edema due to central retinal vein occlusion persisting despite multiple intravitreal bevacizumab injections, J. Ocul. Pharmacol. Ther. 30 (2014) 512–516. [34] Q. Liu, Y. Hu, H. Yu, L. Yuan, J. Hu, A. Atik, M. Guan, D. Li, X. Li, S. Tang, Comparison of intravitreal triamcinolone acetonide versus intravitreal bevacizumab as the primary treatment of clinically significant macular edema, Retina 35 (2014) 272–279. [35] S. Sonoda, T. Sakamoto, T. Yamashita, H. Otsuka, M. Shirasawa, N. Kakiuchi, E. Uchino, H. Terasaki, H. Kawano, Effect of intravitreal triamcinolone acetonide or bevacizumab on choroidal thickness in eyes with diabetic macular edema, Invest. Ophthalmol. Vis. Sci. 55 (2014) 3979–3985.
Contents lists available at ScienceDirect
Journal of Controlled Release journal homepage: www.elsevier.com/locate/jconrel
Safety and pharmacodynamics of suprachoroidal injection of triamcinolone acetonide as a controlled ocular drug release model Mei Chen a, Xiaoli Li a, Jinkun Liu a, Yin Han a, Lingyun Cheng a,b,⁎ a b
Institute of Ocular Pharmacology, School of Ophthalmology and Optometry, Wenzhou Medical University, 270 Xueyuan Road, Wenzhou, Zhejiang 325027, China Jacob's Retina Center at Shiley Eye Center, Department of Ophthalmology, University of California San Diego, 9415 Campus Point Drive, La Jolla, CA 92037-0946, United States
a r t i c l e
i n f o
Article history: Received 1 December 2014 Received in revised form 7 February 2015 Accepted 15 February 2015 Available online 17 February 2015 Chemical compounds studied in the article: Triamcinolone acetonide (PubChem CID: 6436) Indocyanine green (PubChem CID: 11967809) Saline (PubChem CID: 5234) Lipopolysaccharide (PubChem CID: 11970143) Haemotoxylin (PubChem CID: 10603) Eosin (PubChem CID: 173745) Formalin (PubChem CID: 712) Paraffin (PubChem CID: 109453) Pentobarbital sodium (PubChem CID: 16219847) Xylazine hydrochloride (PubChem CID: 68554) Keywords: Suprachoroidal injection Injection volume Drug formulation Triamcinolone acetonide (TA) Indocyanine green (ICG) Pharmacodynamics Experimental uveitis Rabbit eyes
a b s t r a c t Suprachoroidal injection is an emerging technique for drug delivery to the posterior segment, which is hard to reach by non-invasive approaches. However, the injection technique varies and the associated ocular safety is not well understood. In addition, it is not clear if drug formulation is a major factor in optimizing pharmacodynamics using this technique. The current study was designed to compare the suprachoroidal injection of different drug formulations and to characterize the safety and pharmacodynamics of triamcinolone acetonide (TA) delivered by this technique. Both indocyanine green (ICG) solution and TA suspension, at 50 μL, 100 μL, and 150 μL, were suprachoroidally injected and intraocular pressure (IOP) tonometry, fundus photography, and electroretinography were performed over multiple time points up to eight weeks. After 50 μL TA (Kenalog-40) suprachoroidal injection, 4–5 animals at 7 time points were sacrificed for aqueous, vitreous, retina, and plasma collections. TA was quantitated using ultra-performance liquid chromatography tandem mass spectrometry. For comparative efficacy study, 50 μL (2 mg) suprachoroidal TA versus 20 mg subtenon TA were performed 4 weeks before induction of experimental uveitis with 10 ng of intravitreal lipopolysaccharide. After suprachoroidal injection, IOP had an acute elevation, higher volume caused higher IOP (p b 0.0001). Equivalent volume of ICG solution led to a significantly smaller IOP elevation than after TA suprachoroidal injection. This finding suggests better distribution of ICG solution than TA suspension in the suprachoroidal space. Following a 50 μL suprachoroidal injection, peak TA concentration in the aqueous was below 1 ng/mL. In contrast, the posterior vitreous and retina had 1912 ng/mL and 400,369 ng/mL TA, respectively. Maximum TA in plasma was 11.6 ng/mL. Drug exposure to the posterior retina was 523,910 times more than that to the aqueous and 29,516 times more than systemic TA exposure. In the treatment of lipopolysaccharide-induced uveitis, compared with 20 mg subtenon injection, suprachoroidal 2 mg TA demonstrated much better efficacy with significantly less aqueous humor cells and lower vitreous opacity scores (p b 0.05). Histology showed much less vitreous inflammation in the suprachoroidal injection group (p b 0.0001). It seems that a 50 μL suprachoroidal injection of TA was well tolerated in rabbit eyes and demonstrated excellent penetration into the posterior retina, providing better therapeutic effect than subtenon 20 mg TA. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Retinal diseases such as diabetic retinopathy or macular degeneration are chronic and refractory. Treatment of these diseases is challenging due to their chronic nature, which mandates frequent intravitreal injections of therapeutics or even a surgical procedure to implant a sustained drug delivery device at the vitreous base [1,2]. Intravitreal drug delivery is invasive and exposes patients to risks of retinal drug toxicity [3], lens injury and vitreous hemorrhage [4], and even intraocular infections [4]. Subconjunctival or subtenon injection of therapeutic ⁎ Corresponding author at: Institute of Ocular Pharmacology, School of Ophthalmology and Optometry, Wenzhou Medical College, 270 Xueyuan Road, Wenzhou, Zhejiang 325027, China. E-mail address: [email protected] (L. Cheng).
http://dx.doi.org/10.1016/j.jconrel.2015.02.021 0168-3659/© 2015 Elsevier B.V. All rights reserved.
agents is considered to be safer than intravitreal injection and these transscleral drug delivery routes have been shown to be more effective than systemic drug administration [5]. Though transscleral drug delivery is a better option than systemic drug administration, it is less effective than intravitreal injection in the treatment of various macular edemas [6]. Drugs placed on the sclera's surface encounter vigorous clearance from conjunctival and episcleral blood and lymphatic circulation [7]. In addition, the sclera itself is a major barrier for drugs to penetrate and reach the retina. To overcome the blockade of the sclera, several groups have investigated suprachoroidal drug delivery [8]. To date, suprachoroidal drug delivery seems to be feasible; however, the safety of suprachoroidal drug delivery remains controversial depending on the therapeutic agent and technique. Eimnabl et al. [9] reported that suprachoroidal injection of poly-(ortho ester) (POE) or 1% sodium hyaluronate led to fundus pigment disturbance and retinal atrophy.
110
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
three escalating volumes, 50 μL, 100 μL, and 150 μL because human vitreous volume (4.5 mL) is 3 times more than rabbit vitreous (1.5 mL). We hypothesize that a drug solution would behave differently than a drug suspension (Kenalog-40) during and after injection into the suprachoroidal space. In the current study, indocyanine green (ICG) was used to model a drug solution for its easy visual identification and Kenalog-40 as a suspension.
Suprachoroidal cannulation and drug delivery were also reported to be damaging to the outer retina, including photoreceptors, in the primate eye model [10]. The most recent studies from Tyagi et al. and Patel et al. were encouraging and a good safety profile was depicted using a 34 gauge needle or a microneedle for the suprachoroidal injection [11, 12]. From the current literature, efficiency of drug diffusion to the retina after suprachoroidal injection varies from drug to drug. For example, the amount of Avastin reaching the retina from suprachoroidal delivery was limited and cleared more quickly compared with intravitreal delivery [13]. In contrast, sodium fluorescein seems to behave differently [11, 12]. Though suprachoroidal delivery may be different from subtenon and intravitreal routes, a sustained drug delivery formulation is equally important for all these ocular delivery routes. Patel et al. demonstrated that suprachoroidally delivered dextran or Bevacizumab was cleared within a day while 1 μm sized polymeric particles remained in the eyes for over 2 months [11]. To date, triamcinolone acetonide seems to be the most promising drug formulation for suprachoroidal delivery to treat retinal diseases due to its low solubility and sustained release property [10,14]. However, no long-term safety and pharmacodynamics study of non-cannulated suprachoroidal TA is available yet. The current study was designed to evaluate the feasibility and safety of various volumes of TA suprachoroidal injection and to characterize the pharmacokinetics of the vitreous and retina as well as therapeutic efficacy in an experimental uveitis model using subtenon TA injection as a comparative modality.
All rabbits were anesthetized by an intramuscular injection of pentobarbital sodium (20 mg/kg) and xylazine (4 mg/kg). In addition, 0.5% proparacaine ophthalmic drops (Alcon) were used topically. Under the direct view of a surgical microscope (F18; Leica, Wetzlar, Germany), superonasal conjunctival peritomy was performed with a radial cut along the superior rectus muscle to expose the sclera. A 30 gauge needle was fit into an extension tubing (extension set 15.2 cm ID 8 mm OD 1.6 mm, EAGLE LABS, Rancho Cucamonga, CA) which was then connected to a 250 μL Hamilton syringe (HAMILTON). The Hamilton syringe was loaded into a repeating dispenser with which each actuation delivers 1/50 of the total capacity of that syringe [23]. The injection was made 9 mm behind the limbus at the globe meridian of 2 o'clock for the right eye and 11 o'clock for the left eye. After the procedure, the conjunctiva and Tenon were restored by an 8-0 suture and ofloxacin eye ointment was applied after the last examination.
2. Materials and methods
2.4. Safety study
2.1. Study design
For ICG suprachoroidal injection study, our purpose is to compare how a solution differs from a suspension like TA in terms of easiness of injection. In addition, we wanted to estimate the distributing area of the injected within the suprachoroidal space after a solution injection versus a suspension injection. We are not interested in knowing if ICG itself is toxic. Therefore, two eyes of each animal were used to reduce animal usage. For suprachoroidal TA injection, only one eye of each animal was used and the fellow eye was left untouched for control purpose. Following suprachoroidal injection, the eyes were monitored by ocular tomometry, indirect ophthalmoscopy, fundus photography, electroretinography (ERG), and histology after sacrifice as previously reported [24].
Intravitreal injection is invasive and the associated risks can be escalating with repeated injections. Vitreous itself has no blood vessels and is an immunologically privileged space. Therefore, it is prone to opportunistic infections. Development of a safer ocular drug delivery route has become an unmet need. Our group has been focusing on periscleral drug delivery which includes transscleral [15–19] and suprachoroidal routes. The current study was designed to compare traditional subtenon drug delivery with the emerging technique of suprachoroidal drug delivery in the context of medical care of posterior segment eye diseases. Thus far, when comparing suprachoroidal drug delivery with intravitreal or transscleral drug delivery, drug level was computed from the mixture of choroid and retina for suprachoroidally injected eyes [20,11,12]. This can lead to an inflated drug level because the drug delivered to suprachoroidal space was sampled into the complex of choroid and retina. In the current study, only the vitreous and retina were individually sampled for the drug quantitation. The retina is distinctively separated from the choroid by the durable Bruch's membrane that provides a natural barrier to avoid cross-contamination between the retina and the choroid during the dissection [21,15,22]. For the in vivo studies, Japanese big-eared white rabbits were used and animal handling was in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Visual Research. The body weights of the rabbits were between 2.2 kg and 3.5 kg with a mean of 2.7 kg ± 0.35 kg. Both genders were randomly used. 2.2. Determination of injection volumes The suprachoroidal cavity is a virtual space that is not present in the normal eye. It is important to know how much fluid can be safely injected into this virtual space without it flushing back and without causing sustained high intraocular pressure. The current study uses the rabbit eye as a model for human eye application. Using intravitreal injection volume as a reference in which 50 μL can be safely administered into rabbit vitreous without paracentesis, we elected to test
2.3. Suprachoroidal injection procedure
2.5. Pharmacokinetic study Twenty-eight Japanese big-eared white rabbits were used for this study and only one eye of each rabbit was injected. After suprachoroidal injection of 50 μL of triamcinolone acetonide suspension (2 mg), 4 to 5 rabbits were sacrificed for sampling of the eye tissue at each time point (1-day, 3-day, 1-week, 2-week, 4-week, 6-week, and 8-week). Following globe nucleation, the globe was snap frozen as we described previously [15,21] and the globe was cut into three sections. The first cut was through a circumferential line at pars plana 2 mm from the limbus, and the second cut was through the midline of globe equator. The aqueous, anterior vitreous, anterior retina, posterior vitreous, and posterior retina were dissected into individual pre-weighed vials and kept at − 80 °C until drug quantitation. Each cut was made with a new 23 blade and the cutting was done from the opposite of the suprachoroidal injection site towards the injection site to avoid cross-contamination. Blood was also sampled from the ear vein at each time point for plasma TA quantitation. Fellow eyes were not analyzed because the drug source for these eyes would be from blood circulation and TA in the blood sample was very low. Pharmacokinetics of subtenon TA injection was not concurrently performed because such a study was already conducted and the results were available in the literature [15].
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
2.6. Comparative efficacy study Twelve Japanese big-eared white rabbits were used for this study. The study was designed to test and compare the therapeutic sustainability of TA in the retina and vitreous following suprachoroidal space injection versus traditional subtenon injection. The dose used for suprachoroidal injection was 2 mg in 50 μL (determined from the preceding study) and was 20 mg for subtenon injection, which is the commonly used dose in clinical practice [25,26]. Four weeks after the TA injection, all rabbits' right eyes were intravitreally injected with 10 ng of lipopolysaccharide (InvivoGen, San Diego, CA) in 50 μL of saline through the pars plana (1.5 to 2 mm from limbus). The left eyes were untouched. Vitritis was induced at 4 weeks because the preceding pharmacokinetics showed TA concentration well above therapeutic level at this time point after suprachoroidal injection, and lower drug level after subtenon injection [15]. At 24 h, 48 h, and 72 h aqueous cells were graded under slit-lamp biomicroscope using standardization of uveitis nomenclature for reporting clinical data [27]. At the same time points, vitreous haze was graded from fundus photographs using a 0-8 grading scale [28]. After the last examination at 72 h and immediately before sacrifice, 50 μL of aqueous was sampled using a 30 gauge needle attached to a 1 mL syringe under direct view of a surgical microscope. Twenty microliters of the aqueous was used to create smear for inflammatory cell count. After sacrifice the eye globes were fixed in 10% formalin for paraffin histology processing. Five micrometer paraffin sections were prepared for HE (Hematoxylin & Eosin) staining and light microscopy.
2.7. Statistical analysis The categorical data such as incidence of the complications were tabulated and analyzed with multivariate logistic regression to model the important predictors for the likelihood of complications. In addition, a trend test was performed by coding the three doses into ordinal scales. For continuous data such as intraocular pressure (IOP), multivariate linear regression was performed to model the predictors, such as formulation and volume, and their interaction. For ERG data, since only one eye of each rabbit was injected with TA, paired t test was used for each type of ERG to compare the treated eye with its fellow eye. For aqueous humor inflammatory cell count and vitreous haze grading, the count or grading was conducted three times, once a day, generalized estimating equation (GEE) was used within the SAS (version 9.4) environment to compare inflammation between the two groups while adjusting for the evaluating time. Similarly, inflammatory cell counts from aqueous humor smears and vitreous on HE stained sections at multiple locations
111
were analyzed using GEE. All analyses were performed using either JMP 11 or SAS 9.4 statistical software (SAS Institute Inc, Cary, NC). The pvalues smaller than 0.05 were considered to be significant. 3. Results 3.1. Assessing optimal injection volumes Our hypothesis was that a solution may differ from a suspension in aspects of injection and associated complications. Both ICG and TA were tested at volumes of 150 μL, going down to 100 μL, then 50 μL. From a pilot study using 150 μL we found that IOP was too high to be measured by a Tonopen; therefore a Schiotz tonometer was used throughout the study. The complications observed include backflow from injection entry, needle injury causing choroidal hemorrhage, choroidal capillary bleeding (Choroid hemorrhage away from needle entry), and serous sensory retina elevation from penetrating through Bruch's membrane (Fig. 1). The incidence of these complications is summarized in Table 1. Logistic regression using volume and agent as predictors found that volume was a significant contributor to these complications but agent was not (Table 1). Further statistical analysis showed that all three complications (backflow, p = 0.0003; localized serous retinal elevation, p = 0.0002; choroid hemorrhage away from needle entry, p b 0.0001) had a higher incidence with higher injection volumes (Cochran Armitage Trend Test). After these complications were pooled, the logistic regression for independent variables of agent and volume both became significant predictors. The drug suspension (TA) tended to have a higher rate of complications (p = 0.0192, nominal logistic fit) after the volumes were adjusted. 3.2. Intraocular pressure (IOP) assessment After suprachoroidal injection the eyes were monitored for their IOP change over time (Fig. 2). For both solution (ICG) and suspension (TA), the magnitude of IOP increase was significantly associated with the injection volume; higher volume caused higher IOP (p b 0.0001, multivariate linear regression). In addition, IOP rise after suprachoroidal injection of a drug suspension (TA) was significantly higher than after injection of a drug solution (ICG) (p b 0.0001, multivariate linear regression). 3.3. Area of ICG diffusion within suprachoroidal space To determine the area of ICG diffusion within the suprachoroidal space, an indirect ophthalmoscope and a 28-diopter Volk lens were
Fig. 1. The fundus images are showing the complications from suprachoroidal injections. The left frame is demonstrating localized serous retinal detachment (thin arrow) from injecting fluid due to the needle tip penetrating Bruch's membrane at some point. The middle frame is demonstrating superfacial choroidal hemorrhage, next to the optic nerve and below the medullary ray (thin arrow), purportedly caused from acute IOP change. The right frame is demonstrating deeper choroidal hemorrhage (thin arrow) at needle entry near the vortex vein (thicker arrow). The vortex veins can be seen at superior peripheral retina and indicated with the thicker arrows. The images are not crystal clear due to residual cornea edema from high IOP.
112
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Table 1 Injection volumes and complications. Injection agents
Nb
Volume
Backflow from needle entry
Localized serous retinal elevation
Choroid hemorrhage away from needle entry
Choroid bleeding at needle entry
ICG ICG ICG TA TA TA p Valuea
12 12 12 6 28 37
150 μL 100 μL 50 μL 150 μL 100 μL 50 μL
2/12 0/12 0/12 2/6 2/28 0/37 0.0026 (volumes); 0.99 (agents)
4/12 1/12 0/12 2/6 1/28 1/37 0.0036 (volumes); 0.90 (agents)
1/12 1/12 0/12 6/6 7/28 0/37 0.0003 (volumes); 0.99 (agents)
0/12 0/12 0/12 0/6 1/28 2/37 0.71 (volumes); 0.17 (agents)
ICG = indocyanine green. TA = triamcinolone acetonide. a Effect likelihood ratio tests. b N = numbers of eyes used.
used. The green area was measured by the number of disc diameters on each o'clock meridian (12 meridians) from the optic nerve towards the peripheral. Then the number of optic diameters at each meridian was summed for each eye. For this purpose, 3 rabbits and six eyes were used to measure the suprachoroidal ICG area by indirect ophthalmoscopy before sacrifice and then after sacrifice by a caliper under a dissecting microscope using the vertical diameter of the optic nerve head. Two observers conducted the measurements independently. The numbers from indirect ophthalmoscopy and caliper measurement were in good agreement (Overall Concordance Correlation Coefficient = 0.8959 with confidence intervals from 0.8425 to 0.9338, n = 72) [29]. On average, at the 12 o'clock meridian there are 5 disc diameters between the optic nerve and ora serrata; 6.5 disc diameters at nasal (at 9 for OD and at 3 for OS); 6.5 disc diameters at temporal (at 3 for OD and at 9 for OS); and 7 disc diameters at inferior meridian (6 O'clock). The average summed disc diameters of the 12 meridians was 75.6. This number was used as the total fundus area and was divided into the measurements of green area for each study eye to estimate the percentage of retina area stained by ICG as shown in Fig. 3. Roughly, 15%, 45%, and 65% of suprachoroidal space was filled with ICG following 50 μL, 100 μL, and 150 μL injections. The three sample images are demonstrated in Fig. 4.
3.4. Electroretinography (ERG) and histology ERG examination was performed on TA injected animals. The right eye of each rabbit received TA while the left eye was untouched to serve as control. A paired t-test was used to compare the amplitudes of right eyes versus left eyes by ERG type. No statistically significant difference was found (Fig. 5). In the histology analysis, ICG was identifiable in suprachoroidal space as dark green color and no abnormality was noted in TA injected eyes (Fig. 6). 3.5. Pharmacokinetic study Based on the preceding safety studies, 50 μL of TA suprachoroidal injection was used for the ocular pharmacokinetic study. The key pharmacokinetic parameters are summarized in Table 2 and the drug concentration–time curves are presented in Fig. 7. The triamcinolone in aqueous was below 1 ng/mL for the whole study period. In contrast, posterior vitreous and retina had peak concentrations of 1912 ng/mL and 400,369 ng/mL, respectively. The maximum TA in plasma was 11.6 ng/mL. Drug exposure to the posterior retina was 527,869 times more than that to the aqueous and 29,516 times more
Fig. 2. Intraocular pressure change over time following suprachoroidal injection of various volumes of ICG or TA. The elevated IOPs returned back to baseline or near baseline 20 min after the injection. Each error bar is constructed using 1 standard error from the mean.
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 3. Box plot of percentage of fundus involved with the suprachoroidal ICG at different injection volumes, depicting the ICG distribution within suprachoroidal space.
than systemic TA exposure (Table 2). The TA detected at the last time point of 8 weeks in the posterior retina was above 10 ng/g. 3.6. Comparative efficacy of pretreatment Four weeks after subtenon (20 mg) or suprachoroidal (2 mg) TA injection, 10 ng of intravitreal lipopolysaccharide was administered to cause clinically notorious uveitis in both groups. However, both anterior and posterior uveitides were more severe in the subtenon TA group. Aqueous humor cell grading was significantly higher in subtenon eyes (least square mean = 2.87) than that in suprachoroidal injected eyes (least square mean = 1.42, p = 0.0012, Fig. 8) and aqueous humor cells decreased significantly over 72 h (β = −0.04, p b 0.0001). Similarly, vitreous haze grading was significantly higher in subtenon group eyes (least square mean = 6.13) than in suprachoroidal group eyes (least square mean = 4.53, p = 0.04, Fig. 8); vitreous haze did not show significant change during the 72 hour study period (β = −0.007, p = 0.36) (Fig. 9). In addition, the infected eyes had significantly lower intraocular pressure than their fellow eyes (12.58 mm Hg, p b 0.0001) though there was no significant difference between the subtenon TA injected eyes (least square mean = 7.04 mm Hg) and suprachoroidal TA injected eyes (least square mean = 7.73 mm Hg, p = 0.46). IOP showed a statistically significant decrease over the 72 hour (β = −0.06, p b 0.0001). The inflammatory cells in aqueous humor were also counted from a smear. Seven spots were sampled for counting under 20 × objective
113
Fig. 5. The b-wave amplitude was compared between the right eyes (OD, TA injected eyes) and the left eyes (OS, control eyes) within each type of ERG (paired t-test).
lens and 10 × eyepieces. Each spot is a 20 × field. Four evenly spaced spots along the smear edge and three spots on the middle longitudinal axis were counted for cells. The same paradigm was used for each smear and counting was performed by a masked observer. The least square mean cell number for subtenon TA eyes was 23/field but only 9/field for suprachoroidal TA eyes (p = 0.0007). Histology and microscopy revealed that the main change was vitritis. Inflammatory cells were variably distributed in the vitreous cavity and posterior chamber. Vitreous inflammatory cells were counted at four locations: in front of the optic nerve, behind the lens, and superior peripheral vitreous and inferior peripheral vitreous. At each location, four random fields under a 40 × objective lens were photographed (64 ×) and the cells were counted. Statistics showed that significantly more inflammatory cells per field were present in the subtenon group than in the suprachoroidal group (132 versus 52, p b 0.0001, Fig. 10). 4. Discussion Suprachoroidal injection for drug delivery to the posterior segment is an emerging method. Understanding its delivery mechanism and possible lateral side effects is the first step to deploy this technology. So far, a few reports are available using either cannulation or proprietary hollow microneedle [11,13,30]. The current study demonstrates that suprachoroidal injection is feasible with a 30 gauge sharp needle attached to a Hamilton syringe loaded on a repeating dispenser. 50 μL of Kenalog-40 can be consistently delivered with minimal complication.
Fig. 4. Sample fundus images with different volumes of suprachoroidal ICG injection, taken shortly after the injection. The images are not crystal clear due to residual corneal edema from high IOP.
114
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 6. Microscopic images showing the retina (R), retinal pigment epithelium (RPE, shorter arrows), choroid (Ch), sclera (S), and the junction line between the choroid and sclera (longer arrows). The image in the left panel came from an eye injected with suprachoroidal ICG at 2-weeks after the injection; the image in the right panel came from an eye injected with suprachoroidal TA 2 weeks after the injection. The dark green ICG can be seen at the junction line between the choroid and sclera in the left panel (thin arrows). The image on the left panel was lightly stained with hematoxylin on purpose to reveal the ICG color.
The complications become more of a risk with the increase of injection volume. In addition, TA suspension seems to have more complications than ICG injection. However, it should be noted that there were no complications with ICG injections and only one serous retinal elevation out of 37 eyes injected with 50 μL of TA suspension. Thus far, the largest suprachoroidal volume tested in vivo was 100 μL injected into the rabbit eye [9]. In that study, local retinal atrophy after a 100 μL suprachoroidal sodium hyaluronate injection was noted on fundus photograph [9]. Olsen et al. cannulated pig suprachoroidal space and injected 70 μL of bevacizumab mixed with hyaluronic acid (50 μL bevacizumab + 20 μL of hyaluronic acid). Ocular pharmacokinetics was reported without mentioning any complication [13]. In the case of the pig, the vitreous volume is more than double the volume of rabbit vitreous [31]. From the current study, 50 μL TA did not show obvious adverse ocular effects except for one injection procedure related complication (localized serous retinal elevation at the needle tip). When the volume escalated to 100 μL, 7 out of 28 eyes showed choroidal hemorrhage away from the needle entry site. We postulate that this may be caused by acute IOP change because this type of hemorrhage was observed more frequently in 150 μL injections. In addition, this type of hemorrhage occurred less in ICG injected eyes and ICG induced IOP elevation was also significantly less than that from injection of TA. The difference of IOP elevation between these two agents may be due to poorer expansion and distribution of TA suspension in suprachoroidal space. The choroidal hemorrhage at the needle entry site is, we think, from choroidal
injury by the needle tip; the hemorrhage disappeared within 1 to 2 weeks without significant findings in histology. Compared with the current injection technique, choroidal cannulation may require much more training and cannulation itself is more traumatizing [10]. To improve the safety of suprachoroidal injection, a proprietary microhollow needle was fabricated and in an ex vivo study it could consistently deliver up to 35 μL into enucleated eye globes of rabbit, pig, and human. In a subsequent in vivo study using rabbit eyes, 50 μL of suspension could be delivered within 15 s and waiting an additional 30 s before removing the microneedle from the eye can prevent excessive reflux [11]. The microneedle's short length allows more reflux than a longer needle; however, with a short needle there is less risk of choroidal damage. The current study is the first comprehensive in vivo study to investigate the relationship between injection volume and IOP change, and the relationship between drug formulations (ICG vs. TA suspension). As expected, larger injection volumes induced higher IOP elevation. The elevated IOP returned to pre-injection level within 20 min except for the eyes injected with 150 μL. An interesting finding from the current study is that IOP elevation was significantly higher after the injection of TA suspension than after the injection of ICG even when the injection volumes were the same. This phenomenon may be explained by the postulation that ICG expands further within suprachoroidal space than TA suspension that expands more towards the vitreous and causes IOP elevation.
Table 2 Pharmacokinetic parameters following suprachoroidal TA injection. Ocular tissues
Rsq adjusted
Half-life (days)
Tmax (days)
Cmax (ng/mL or g)
SE_Cmax
Tlast (days)
Clast (ng/mL or g)
AUClast
MRTlast (days)
Aqueous Posterior vitreous Posterior retina Plasma
0.85 0.99 0.97 0.74
9.85 3.78 3.41 16.7
1.00 3.00 3.00 1.00
0.69 1911.62 400,368.55 11.6
0.19 605.39 117,120.49 7.01
28.00 56.00 56.00 56
0.10 0.18 11.18 0.87
7.94 22,016.71 4,191,281.90 141.65
8.93 8.20 7.66 17.02
Rsq adjusted: R square adjusted. Tmax: The time at which mean drug concentration was the maximum. Cmax: Mean maximum drug concentration. SE = standard error. Tlast = the last observed time point. Clast = concentration at the last observed time point. AUClast: Area under the curve calculated to the last observed time. MRTlast: Mean residence time calculated to the last observed time.
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 7. TA (triamcinolone acetonide) concentration–time curves of aqueous, posterior vitreous, posterior retina, and plasma.
The current pharmacokinetic study following 50 μL of suprachoroidal TA suspension demonstrated high TA concentration in both the retina and vitreous. In contrast, TA in aqueous humor was very low, below 1 ng/mL. During the whole study period (2 months), the TA exposure to aqueous humor (AUC = 7.94 ng·day/mL) was 2773 times lower than the AUC of posterior vitreous and roughly 527,869 times lower than the AUC of posterior retina. This can be a major advantage for suprachoroidal injection over subtenon injection because high TA in aqueous humor can cause steroid-related glaucoma and cataract. This point is supported by the fact that intravitreal TA causes significantly more IOP elevation than subtenon TA and the aqueous humor TA concentration in the former was much higher [32] than the latter [16]. After subtenon injection of 40 mg of TA, TA concentration in the aqueous and the vitreous was similar, with peak TA concentration in aqueous humor being 1244 ng/mL in the rabbit eye and 197 ng/mL in the human eye [15,16]. The TA systemic exposures following 2 mg suprachoroidal injection and following 40 mg subtenon injection in rabbit were similar [15]. These facts indicate that suprachoroidal TA injection is a better option that uses less drug, provides higher TA concentrations in the retina and vitreous, leads to minimal aqueous humor drug exposure, and has similar systemic TA exposure to subtenon injection. In addition, suprachoroidal injection should be less vulnerable to procedure related opportunistic infections than intravitreal injection because the vitreous is an immunologically privileged culture medium.
115
Suprachoroidal injection is an emerging method and its ocular safety and efficacy needs thorough inspection in preclinical studies. In addition, it is also expected that not every compound or formulation will fit this delivery route. Based on the literature available, a readily sustainable formulation seems to be the right fit for this technique [9,11,13]. Triamcinolone acetonide is a hydrophobic small molecule drug. Its suspension is commercially available and has been used as a sustained release formulation through intravitreal or subtenon injection to treat various chorioretinal disorders [33–35]. TA has been evaluated with suprachoroidal injection and Olsen et al. reported that TA remained in the local ocular tissue for at least 120 days following a 12 μL TA suspension suprachoroidal cannulation [10]. In that study, the pig eye was used and a more concentrated TA suspension (1.5 or 3 mg in 12 μL) was cannulated. At postcannulation day 60, much more TA (0.1 mg/g) was found in the retina compared with the current study (11 ng/g). The difference could stem from the use of different animal species as well as different TA concentration and injection methods. Another in vivo study by Gilger et al. evaluated the effect of TA administered into the suprachoroidal space (SCS) using a microneedle and compared its efficacy with intravitreal TA injections in a porcine model of acute posterior segment inflammation [14]. The acute ocular inflammation was induced in the pig 24 h after suprachoroidal TA injection. The current study was designed differently in that we aimed to test the efficacy of sustained drug release following suprachoroidal TA injection and used concurrent subtenon TA injection as a control. The study results were very encouraging. The results demonstrated the superiority of suprachoroidal 2 mg TA to subtenon 20 mg TA in reducing acute inflammation in a rabbit model of uveitis. This was the first pharmacodynamics study to evaluate suprachoroidal TA efficacy in a long-term setting. Key uveitis parameters, such as vitreous haze and inflammatory cell counts, were significantly reduced in the eyes pretreated with suprachoroidal TA. In summary, it seems to be feasible to consistently deliver 50 μL of TA suspension into the rabbit suprachoroidal space using a 30 gauge hypodermic needle with the help of Hamilton syringe and repeating dispenser. There was no backflow with this volume, but backflow was observed when delivering 100 μL and 150 μL. ICG seems to be easier to deliver than TA suspension, with less procedure-related complications and less IOP elevation. No functional abnormality was observed in any dose group by ERG examination. Ocular pharmacokinetics of TA delivered by suprachoroidal injection demonstrated much lower TA concentration in aqueous humor but much higher in posterior vitreous and retina than that from a traditional 40 mg TA subtenon injection. Suprachoroidal injection of 2 mg of TA provided stronger therapeutic effect in reducing acute experimental vitritis, even at 4 weeks following the initial single injection, than the traditional subtenon injection of 20 mg of TA did. Suprachoroidal injection may become a promising drug delivery method to the back of the eye for slow release drug formulations.
Fig. 8. Slit-lamp based microphotographs of the anterior segment. The left panel was from an eye in the suprachoroidal TA injection group and the right panel from an eye in the subtenon TA injection group. The aqueous humor was more turbid and exudates were present in the pupil area of the subtenon TA injected eye. In addition, due to inflammation the pupil was less dilated in the right image compared with the pupil in the left image.
116
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Fig. 9. This composite image is depicting various vitreous hazes from the vitritis induced by intravitreal lipopolysaccharide. The upper panel images are from a single eye with a suprachoroidal space (SCS) injection of 2 mg of TA and the lower panel images are from a single eye with a posterior subtenon (PST) 20 mg of TA injection. The vitreous clarity in the first image of the upper panel at 24 h (24 h, grade 3) is clearer than at 48 h (48 h, grade 5) and 72 h (72 h, grade 4). In general, vitreous clarity is significantly better in the upper panel images than in the lower panel images (24 h, grade 6; 48 h, grade 7; 72 h, grade 8).
Fig. 10. Microphotographs from HE stained histology sections, demonstrating inflammatory cell infiltration (asterisk) at the inferior vitreous cavity that is indicated by the ora sarrata (thin arrow) and part of the pars plana (thick arrow). Inflammatory cells were more significant in the eye with subtenon TA injection (right panel) compared with that in the eye with suprachoroidal TA injection (the left panel). The retinas of both eyes were relatively normal looking in anatomical structures.
M. Chen et al. / Journal of Controlled Release 203 (2015) 109–117
Acknowledgments Funding/support: National Natural Science Foundation of China, Grant No. 31271022. Statement of conflicts of interest The authors have no financial/conflicting interests in any aspect of this study. References [1] I. Krebs, L. Schmetterer, A. Boltz, R. Told, V. Vecsei-Marlovits, S. Egger, U. Schonherr, A. Haas, S. Ansari-Shahrezaei, S. Binder, M.R. Grp, A randomised double-masked trial comparing the visual outcome after treatment with ranibizumab or bevacizumab in patients with neovascular age-related macular degeneration, Br. J. Ophthalmol. 97 (2013) 266–271. [2] G.J. Jaffe, Reimplantation of a fluocinolone acetonide sustained drug delivery implant for chronic uveitis, Am J. Ophthalmol. 145 (2008) 667–675. [3] A.A. Seawright, R.D. Bourke, R.J. Cooling, Macula toxicity after intravitreal amikacin, Aust. N. Z. J. Ophthalmol. 24 (1996) 143–146. [4] A.E. Fung, P.J. Rosenfeld, E. Reichel, The International Intravitreal Bevacizumab Safety Survey: using the internet to assess drug safety worldwide, Br. J. Ophthalmol. 90 (2006) 1344–1349. [5] O. Weijtens, R.C. Schoemaker, A.F. Cohen, F.P.H.T.M. Romijn, E.G.W.M. Lentjes, J. van Rooij, J.C. van Meurs, Dexamethasone concentration in vitreous and serum after oral administration, Am. J. Ophthalmol. 125 (1998) 673–679. [6] M.A. Bonini-Filho, R. Jorge, J.C. Barbosa, D. Calucci, J.A. Cardillo, R.A. Costa, Intravitreal injection versus sub-Tenon's infusion of triamcinolone acetonide for refractory diabetic macular edema: a randomized clinical trial, Invest. Ophthalmol. Vis. Sci. 46 (2005) 3845–3849. [7] S.H. Kim, C.J. Galban, R.J. Lutz, R.L. Dedrick, K.G. Csaky, M.J. Lizak, N.S. Wang, G. Tansey, M.R. Robinson, Assessment of subconjunctival and intrascleral drug delivery to the posterior segment using dynamic contrast-enhanced magnetic resonance imaging, Invest. Ophthalmol. Vis. Sci. 48 (2007) 808–814. [8] M.R. Robinson, S.S. Lee, H. Kim, S. Kim, R.J. Lutz, C. Galban, P.M. Bungay, P. Yuan, N.S. Wang, J. Kim, K.G. Csaky, A rabbit model for assessing the ocular barriers to the transscleral delivery of triamcinolone acetonide, Exp. Eye Res. 82 (2006) 479–487. [9] S. Einmahl, M. Savoldelli, F. D'Hermies, C. Tabatabay, R. Gurny, F. Behar-Cohen, Evaluation of a novel biomaterial in the suprachoroidal space of the rabbit eye, Invest. Ophthalmol. Vis. Sci. 43 (2002) 1533–1539. [10] T.W. Olsen, X. Feng, K. Warner, S.R. Conston, D.H. Sierra, D.V. Folden, M.E. Smith, J.D. Cameron, Cannulation of the suprachoroidal space: a novel drug delivery methodology to the posterior segment, Am J. Ophthalmol. 142 (2006) 777–787. [11] S.R. Patel, D.E. Berezovsky, B.E. McCarey, V. Zarnitsyn, H.F. Edelhauser, M.R. Prausnitz, Targeted administration into the suprachoroidal space using a microneedle for drug delivery to the posterior segment of the eye, Invest. Ophthalmol. Vis. Sci. 53 (2012) 4433–4441. [12] P. Tyagi, R.S. Kadam, U.B. Kompella, Comparison of suprachoroidal drug delivery with subconjunctival and intravitreal routes using noninvasive fluorophotometry, Plos One 7 (2012). [13] T.W. Olsen, X. Feng, K. Wabner, K. Csaky, S. Pambuccian, J.D. Cameron, Pharmacokinetics of pars plana intravitreal injections versus microcannula suprachoroidal injections of bevacizumab in a porcine model, Invest. Ophthalmol. Vis. Sci. 52 (2011) 4749–4756. [14] B.C. Gilger, E.M. Abarca, J.H. Salmon, S. Patel, Treatment of acute posterior uveitis in a porcine model by injection of triamcinolone acetonide into the suprachoroidal space using microneedles, Invest. Ophthalmol. Vis. Sci. 54 (2013) 2483–2492. [15] K. Nan, S. Sun, Y. Li, J. Qu, G. Li, L. Luo, H. Chen, L. Cheng, Characterisation of systemic and ocular drug level of triamcinolone acetonide following a single sub-Tenon injection, Br. J. Ophthalmol. 94 (2010) 654–658. [16] L. Shen, Y. You, S. Sun, Y. Chen, J. Qu, L. Cheng, Intraocular and systemic pharmacokinetics of triamcinolone acetonide after a single 40-mg posterior subtenon application, Ophthalmology 117 (2010) 2365–2371.
117
[17] W. Du, S. Sun, Y. Xu, J. Li, C. Zhao, B. Lan, H. Chen, L. Cheng, The effect of ocular pigmentation on transscleral delivery of triamcinolone acetonide, J. Ocul. Pharmacol. Ther. 29 (2013) 633–638. [18] X. Liu, Y. Li, Y. Zhang, W. Du, S. Sun, B. Lin, H. Chen, L. Cheng, Comparison of intraocular pressure elevation after anterior versus posterior subtenon triamcinolone acetonide acetate injection: a retrospective study, Retina 32 (2012) 1838–1843. [19] Y. Meng, S. Sun, J. Li, K. Nan, B. Lan, Y. Jin, H. Chen, L. Cheng, Sustained release of triamcinolone acetonide from an episcleral plaque of multilayered poly-epsiloncaprolactone matrix, Acta Biomater. 10 (2014) 126–133. [20] M.Z. Wang, W. Liu, Q.J. Lu, H.Y. Zeng, S.M. Liu, Y.K. Yue, H.T. Cheng, Y. Liu, M. Xue, Pharmacokinetic comparison of ketorolac after intracameral, intravitreal, and suprachoroidal administration in rabbits, Retina J. Ret. Vit. Dis. 32 (2012) 2158–2164. [21] L.Y. Cheng, K.Y. Hostetler, J. Lee, H.J. Koh, J.R. Beadle, K. Bessho, M. Toyoguchi, K. Aldern, J.M. Bovet, W.R. Freeman, Characterization of a novel intraocular drugdelivery system using crystalline lipid antiviral prodrugs of ganciclovir and cyclic cidofovir, Invest. Ophthalmol. Vis. Sci. 45 (2004) 4138–4144. [22] H.Y. Wang, J. Chhablani, W.R. Freeman, J.R. Beadle, K.Y. Hostetler, K. Hartmann, L. Conner, K.A. Aldern, L. Pearson, L.Y. Cheng, Intraocular safety and pharmacokinetics of hexadecyloxypropyl-cidofovir (HDP-CDV) as a long-lasting intravitreal antiviral drug, Invest. Ophthalmol. Vis. Sci. 52 (2011) 9391–9396. [23] L. Cheng, M. Toyoguchi, D.J. Looney, J. Lee, M.C. Davidson, W.R. Freeman, Efficient gene transfer to retinal pigment epithelium cells with long-term expression, Retina 25 (2005) 193–201. [24] Y. Li, H. Chen, J. Hou, Y. Zhang, G. Li, X. Gu, L. Luo, H. Shen, J. Qu, L. Cheng, Further characterization of ocular safety profile of commercially available preserved and preservative-free triamcinolone acetonide, Retina 32 (2012) 364–374. http://dx. doi.org/10.1097/IAE.0b013e31821e1f7c. [25] J. Toda, H. Fukushima, S. Kato, Injection of triamcinolone acetonide into the posterior sub-tenon capsule for treatment of diabetic macular edema, Retina 27 (2007) 764–769. http://dx.doi.org/10.1097/IAE.1090b1013e318030bfcd. [26] A.A. Okada, T. Wakabayashi, Y. Morimura, S. Kawahara, E. Kojima, Y. Asano, T. Hida, Trans-Tenon's retrobulbar triamcinolone infusion for the treatment of uveitis, Br. J. Ophthalmol. 87 (2003) 968–971. [27] D.A. Jabs, R.B. Nussenblatt, J.T. Rosenbaum, Standardization of Uveitis Nomenclature Working Group. Standardization of uveitis nomenclature for reporting clinical data. Results of the First International Workshop, Am J. Ophthalmol. 140 (2005) 509–516. [28] J.L. Davis, B. Madow, J. Cornett, R. Stratton, D. Hess, V. Porciatti, W.J. Feuer, Scale for photographic grading of vitreous haze in uveitis, Am J. Ophthalmol. 150 (2010) 637–641 (e631). [29] S.B. Crawford, A.S. Kosinski, H.M. Lin, J.M. Williamson, H.X. Barnhart, Computer programs for the concordance correlation coefficient, Comput. Methods Prog. Biomed. 88 (2007) 62–74. [30] S.R. Patel, A.S.P. Lin, H.F. Edelhauser, M.R. Prausnitz, Suprachoroidal drug delivery to the back of the eye using hollow microneedles, Pharm. Res. Dordr. 28 (2011) 166–176. [31] A. Shafiee, G.L. McIntire, L.C. Sidebotham, K.W. Ward, Experimental determination and allometric prediction of vitreous volume, and retina and lens weights in Gottingen minipigs, Vet. Ophthalmol. 11 (2008) 193–196. [32] P.M. Beer, S.J. Bakri, R.J. Singh, W. Liu, G.B. Peters III, M. Miller, Intraocular concentration and pharmacokinetics of triamcinolone acetonide after a single intravitreal injection, Ophthalmology 110 (2003) 681–686. [33] R.A. Alshareef, S.J. Garg, J. Hsu, J. Vander, C. Park, M.J. Spirn, Intravitreal triamcinolone acetonide injection for macular edema due to central retinal vein occlusion persisting despite multiple intravitreal bevacizumab injections, J. Ocul. Pharmacol. Ther. 30 (2014) 512–516. [34] Q. Liu, Y. Hu, H. Yu, L. Yuan, J. Hu, A. Atik, M. Guan, D. Li, X. Li, S. Tang, Comparison of intravitreal triamcinolone acetonide versus intravitreal bevacizumab as the primary treatment of clinically significant macular edema, Retina 35 (2014) 272–279. [35] S. Sonoda, T. Sakamoto, T. Yamashita, H. Otsuka, M. Shirasawa, N. Kakiuchi, E. Uchino, H. Terasaki, H. Kawano, Effect of intravitreal triamcinolone acetonide or bevacizumab on choroidal thickness in eyes with diabetic macular edema, Invest. Ophthalmol. Vis. Sci. 55 (2014) 3979–3985.