Ординатура / Офтальмология / Английские материалы / Diabetes and Ocular Disease Past, Present, and Future Therapies 2nd edition_Scott, Flynn, Smiddy_2009

234 Diabetes and Ocular Disease

133.Blankenship GW, Machemer R. Long-term diabetic vitrectomy results: report of a 10-year follow-up. Ophthalmology. 1985;92:503–506.

134.Gollamudi SR, Smiddy WE, Schachat AP, et al. Long-term survival rate after vitreous surgery for complications of diabetic retinopathy. Ophthalmology. 1991;98:18–22.

236 Diabetes and Ocular Disease

Several corticosteroids have been evaluated; however, triamcinolone acetonide was chosen because of its long half-life of 18.6 days in the nonvitrectomized eye [5], lack of toxicity, and the fact that it was generally well tolerated [3,6]. In addition, there is a considerable body of literature describing the efficacy of triamcinolone administration in various ocular diseases including uveitis [7,8], macular edema secondary to ocular trauma or retinal vascular disease [9], proliferative diabetic retinopathy (PDR) [10], intraocular proliferation such as proliferative vitreoretinopathy [11], and choroidal neovascularization from age-related macular degeneration [12,13].

The Rationale for Using Intravitreal Triamcinolone Acetonide in Diabetic Retinopathy. Aiello et al. [14] reported increased levels of VEGF in eyes with PDR, and Funatsu et al. [15] have reported increased levels of VEGF in eyes with diabetic macular edema (DME). The cause of DME is believed to be multifactorial, because of breakdown of the blood–retinal barrier secondary to alterations of various antipermeability factors, such as interleukin-6 and VEGF.

Corticosteroids have antiangiogenic, antifibrotic, and antipermeability properties. The principal effects of corticosteroids are stabilization of the blood–retinal barrier, resorption of exudation, and down-regulation of inflammatory stimuli [16–20].

Experimentally, corticosteroids have been shown to reduce inflammatory mediators including interleukin-5, interleukin-6, interleukin-8, prostaglandins, interferon-gamma, and tumor necrosis factor [16–18]; decrease levels of VEGF, a potent permeability factor [19,20]; and improve blood–retinal barrier function [21]. Several known mechanisms of action of corticosteroids could explain the stabilization of the blood–retinal barrier including stabilization of cell and lysosomal membranes [22], reduction of the release [22] or synthesis [23] of prostaglandins, inhibition of cellular proliferation [24], blockage of macrophage recruitment in response to macrophage inhibitory factor, inhibition of phagocytosis by mature macrophages, and decreased polymorphonuclear infiltration into injured tissues [25]. Triamcinolone acetonide in particular has been shown to have an antiangiogenic effect. It has been shown to inhibit basic fibroblast growth factor–induced migration and tube formation in choroidal microvascular endothelial cells and down-regulate metalloproteinase-2 (MMP-2) [26], decrease permeability and down-regulate intercellular adhesion molecule-1 (ICAM-1) expression in vitro [27], and decrease major histocompatibility complex-II (MHC-II) antigen expression [28].

Penfold and associates found that triamcinolone acetonide significantly decreased MHC-II expression consistent with immunocytochemical observations, which revealed condensed microglial morphology [28]. The modulation of subretinal edema and microglial morphology correlated with in vitro observations suggesting that down-regulation of inflammatory markers and endothelial cell permeability are significant features of the mode of action of triamcinolone acetonide. In another study, Penfold et al. investigated the capacity of triamcinolone to modulate the expression of adhesion molecules and permeability using a human

epithelial cell line (ECV304) as a model of the outer blood–retinal barrier [27]. They found that triamcinolone modulated transepithelial resistance of TER and ICAM-1 expression in vitro, suggesting that reestablishment of the blood–retinal barrier and down-regulation of inflammatory markers are the principal effects of intravitreal triamcinolone in vivo. These studies indicate that triamcinolone has the potential to influence cellular permeability, including the barrier function of the retinal pigment epithelium.

Why Deliver Triamcinolone by Intravitreal Injection? Topical corticosteroids have been shown to penetrate the anterior segment [29], but not the posterior segment. Topical corticosteroids sometimes reduce cystoid macular edema occurring after cataract extraction, by reducing the anterior segment inflammation causing the cystoid macular edema. Posterior subtenon corticosteroids may be useful in the treatment of DME [30], but they take longer to diffuse into the posterior segment, and placement over the macula may be variable. In addition, there is considerable systemic absorption with a subtenon injection, which may adversely influence blood sugar levels in diabetic patients. The best way to circumvent the blood–ocular barrier is by direct intravitreal injection. Intravitreal triamcinolone has been shown to deliver high initial concentrations to the target tissue and provide effective levels for at least 3 months [5].

Preclinical Evaluation of the Safety of Intravitreal Triamcinolone. A single, pure, intravitreal triamcinolone acetonide injection has been shown to be well tolerated in rabbit eyes [3]. Electroretinographic data showed no significant differences between treated and control eyes and both light and electron microscopy were normal in both groups. Hida and associates [6] investigated the vehicles of six commercially available depot corticosteroids in rabbit eyes and found no toxic effect on the retina and lens with the vehicle in Kenalog (commercially available triamcinolone acetonide) at levels two times higher than in the marketed drug. However, preservatives present in the vehicle for Kenalog including benzyl alcohol were shown in the same report to have toxic effects on the retina in other steroid preparations. Triescence (Alcon, Fort Worth, TX) is a preservative-free preparation of triamcinolone acetonide that is approved by the Food and Drug Administration (FDA) for intraocular use. It comes as a 1-mL vial containing a suspension of triamcinolone acetonide with concentration of 40 mg/mL. Recommended dosing is 1 to 4 mg (25–100 mL) administered intravitreally. Trivaris (Allergan, Irvine, CA) was also recently approved as a preservative-free gel for intravitreal injection. It was used in the National Eye Institute-sponsored clinical trials evaluating intraocular corticosteroids for macular edema (e.g., SCORE, DRCR.net trials). It comes in a blister pack as a 0.1-mL vial with 8 mg of triamcinolone acetonide; therefore, the concentration is 8 mg/0.1 mL. As an off-label use, Kenalog (Bristol-Myers Squibb, Peapack, NJ) can be administered for the same indication. Kenalog contains benzyl alcohol as a preservative; many retinal surgeons prefer to use nonpreserved Kenalog, after a compounding pharmacy removes the preservative [31]. The standard concentration of this preparation is 40 mg/mL.

238 Diabetes and Ocular Disease

Effi cacy of Intravitreal Triamcinolone for Diabetic Macular Edema. In the first report of intravitreal corticosteroids for DME, Martidis and colleagues [32] injected 4 mg of intravitreal triamcinolone into 16 eyes with clinically significant macular edema (CSME) that failed to respond to at least two previous sessions of laser photocoagulation. With all patients in this retrospective study completing 1- and 3-month follow-up, and 50% completing 6 or more months, the mean improvement in visual acuity was 2.4, 2.4, and 1.3 Snellen lines at the 1-, 3-, and 6-month follow-up intervals, respectively. The central macular thickness measured by optical coherence tomography (OCT) decreased by 55%, 57.5%, and 38%, respectively, from a baseline mean thickness of 540 microns. Reinjection was performed in three of eight eyes after 6 months because of recurrence of macular edema.

Another interventional case series included 26 eyes of 20 patients who received 25 mg of intravitreal triamcinolone acetonide by repeatedly filtering and concentrating the standard triacminolone acetonide preparation, for treatment of diffuse DME [33]. Visual acuity improved from 20/166 at baseline to 20/105 over a mean of 6.64 months follow-up (P < 0.001), compared with a control group of 16 patients who underwent macular grid laser photocoagulation. Seventeen (81%) of 21 injected eyes with a follow-up period of more than 1 month had improved visual acuity; in contrast, the visual acuity of the control group did not change significantly.

In the Intravitreal Steroid Injection Study (ISIS) [34], a prospective, pilot study, 30 eyes of 30 patients were randomized to receive either 2 or 4 mg intravitreal triamcinolone for CSME present at least 3 months and refractory to focal laser treatment. Mean change in visual acuity at 3 months compared to baseline was 7.1 letters (P = 0.01) in the 2-mg group and 12.5 letters in the 4-mg group (P < 0.0001). There was no significant difference in visual gain between the 2- and 4-mg dose groups (P = 0.11). Vision improved >15 letters at 3 months in 23% (3/13) of the 2-mg group and in 33% (5/15) of the 4-mg group (P = 0.69), and 0% (0/11) and 21% (3/14) at 6 months, respectively (P = 0.23). Visual improvement was more likely in cystoid-type DME than diffuse DME.

In a small, randomized study, Massin and associates [35] evaluated 12 patients with bilateral DME unresponsive to laser photocoagulation. One eye was randomized to receive 4 mg intravitreal triamcinolone, and the other eye was observed. All patients were followed for at least 3 months; seven had a follow-up of 6 months. The baseline central macular thickness measured by OCT was 510 microns in injected eyes and 474 microns in control eyes. Retinal thickness improved to 207 microns (P < 0.001) in injected eyes and was not improved in control eyes (506 microns) at 4 weeks after injection. Retinal thickness remained improved in the injected eyes after 12 weeks (207 microns) and was significantly (P = 0.005) different from the control eyes (469 microns). The difference between the central macular thickness of injected and control eyes was no longer significant at 24 weeks because of the recurrence of macular edema in 5 of 12 (42%) injected eyes. Despite the improved anatomic results demonstrated by OCT, at no time was the difference between the visual acuity scores measured on the Early Treatment Diabetic Retinopathy Study (ETDRS) chart for injected and control eyes significant.

The Diabetic Retinopathy Clinical Research network (DRCR.net) conducted a prospective randomized trial comparing laser treatment with intravitreal injection of either 1- or 4-mg doses of Trivaris (Allergan, Irvine, CA), a new formulation of preservative-free triamcinolone acetonide for DME. Eight hundred forty study eyes of 693 subjects with DME involving the fovea and with visual acuity of 20/40 to 20/320 were randomized to focal/grid photocoagulation (n = 330), 1 mg intravitreal triamcinolone (n = 256), or 4 mg intravitreal triamcinolone (n = 254). Retreatment was given for persistent or new edema at 4-month intervals. At 4 months, mean visual acuity was better in the 4-mg triamcinolone group than in either the laser group (P < 0.001) or the 1-mg triamcinolone group (P = 0.001). However, by 1 year, there were no significant differences among groups in mean visual acuity. At the 16-month visit and extending through the primary outcome at 2 years, the mean change in visual acuity was better in the laser group than in the other two groups (at 2 years, P = 0.02 comparing the laser and 1-mg groups, P = 0.002 comparing the laser and 4-mg groups, and P = 0.49 comparing the 1- and 4-mg groups). Treatment group differences in the visual acuity outcome could not be attributed solely to cataract formation. Optical coherence tomography results generally paralleled the visual acuity results.

The DRCR.net study concluded that over a 2-year period, focal/grid photocoagulation is more effective and has fewer side effects than 1- or 4-mg doses of preservative-free intravitreal triamcinolone for most patients with DME who have characteristics similar to the cohort in this clinical trial. However, 40% of eyes in the study had received no previous focal laser treatment and 79% of eyes were phakic. In addition, the preparation of triamcinolone used was a gel and may have different diffusion characteristics compared to the triamcinolone suspensions used in the other studies. In the ISIS study, patients with a cystoid component of macular edema did better with triamcinolone than eyes with diffuse DME. This finding was not replicated in the DRCR.net. We also do not know what the visual results of the DRCR.net would be if those phakic eyes had cataract extraction—in the study, cataract grading was subjective, and cataract may have accounted for more of the vision loss than was believed to be the case. In addition, the treatments were given every 4 months, and perhaps visual results would have been better had they been given at 3-month intervals or even more frequently. Nevertheless, the DRCR.net study does suggest that focal laser is the first line treatment for DME, with consideration of the addition of triamcinolone if a large cystoid component is present, or if two or three focal laser sessions fail to produce the desired visual or OCT outcome.

Dosage of Intravitreal Triamcinolone. Reported doses of triamcinolone used for the treatment of macular edema include 2, 4, and 20 to 25 mg. The optimal dose of intravitreal triamcinolone is not known. In the United States, 4 mg (in 0.1 cc) is the most commonly used dosage. In Germany, the literature reports using 20 to 25 mg, which is prepared by repeatedly filtering and concentrating the triamcinolone preparation. Studies comparing the efficacy, complications, and duration of different doses of triamcinolone in different diseases are necessary to determine the optimum dose. In the ISIS trial, the 4-mg dose was found to be more effective than the 2-mg dose.

240 Diabetes and Ocular Disease

Safety of Intravitreal Triamcinolone. Potential complications of intravitreal steroid injections include endophthalmitis, retinal detachment, retinal tears, vitreous hemorrhage, increased intraocular pressure, and cataract formation [36].

Intraocular Pressure (IOP) Rise. The most common adverse effect of intravitreal triamcinolone is increased IOP [37–39]. Bakri et al. reported a pressure rise of 5 mmHg or greater in 49% of 43 eyes, and a pressure rise of 10 mmHg or greater in 28%, within 12 weeks after a 4-mg intravitreal triamcinolone injection (Table 12.1) [38]. The mean time for an IOP rise of 5 mmHg or greater to occur was 4.1 weeks, and the mean time to reach maximum IOP was 6.6 weeks, although the follow-up interval was variable. The difference between the mean preinjection IOP (15.12 mmHg, n = 43) and the maximum postinjection IOP (20.74 mmHg, n = 43) was statistically significant (P < 0.0001). All eyes in this study responded adequately to topical ocular hypotensive medications within 6 months [38]. Jonas [37] reported that after intravitreal injection of a higher dose of 25 mg of triamcinolone acetonide, an IOP elevation developed in about 50% of eyes, starting 1 to 2 months after the injection. In the vast majority, IOP was normalized by topical medications, and returned to normal values without further medication about 6 months after the injection. Wingate [36] reviewed 113 patients at a single time point (3 months) after a 4-mg intravitreal triamcinolone injection and found that 32% had a rise of 5 mmHg or greater, and 11% had a pressure rise of 10 mmHg or greater. In most of these series, IOP was controlled with topical medications; however, there have been reports [40] of eyes with uncontrolled IOP elevations undergoing trabeculectomy, or even vitrectomy to remove the triamcinolone. In the DRCR.net study, IOP increased from baseline by 10 mmHg or more at any visit in 4% in the focal laser group, 16% in the 1-mg triamcinolone group, and 33% of eyes in the 4-mg triamcinolone group.

Infectious Endophthalmitis. Acute-onset bacterial endophthalmitis was reported in 0.87% of 922 intravitreal triamcinolone acetonide injections (95% confidence interval of 0.38–1.70%) in a retrospective, multicenter review [41]. Potential predisposing risk factors in these eight patients included non-insulin dependent diabetes mellitus (n = 5), injection from a multi-use Kenalog bottle (n = 2), filtering blebs (n = 1), and blepharitis (n = 1). The median time to presentation was 7.5 days (range, 1–15 days). Mycobacterium chelonae endophthalmitis has also been reported

Table 12.1. Intraocular pressure elevation after 4 mg intravitreal triamcinolone

after intravitreal steroid injections [42]; the eye eventually underwent enucleation. In the DRCR.net study, there were no cases of endophthalmitis or inflammatory pseudoendophthalmitis after 1649 intravitreal injections performed in the trial.

Noninfectious Endophthalmitis with Pseudohypopyon. Acute, noninfectious endophthalmitis has been reported following intravitreal triamcinolone injection in 0.87% to 5% of cases [43–46]. Features differentiating inflammatory vitritis from true endophthalmitis include earlier onset, better visual acuity at presentation, lack of growth on culture or organisms on gram stain, and better final visual acuity, all associated with inflammatory vitritis. Median time to presentation in noninfectious endophthalmitis was 1.5 days [44] versus 7.5 days in infectious endophthalmitis [40]. A pseudohypopyon consisting of triamcinolone acetonide particles may occur after intravitreal injection [47]. One technique to help differentiate a pseudohypopyon from true endophthalmitis is to place the patient on their side. If the hypopyon settles in the dependent region, then it is more likely pseudo and not real.

Other Complications. Intravitreal triamcinolone accelerates cataractogenesis and intravitreal injection may cause retinal detachment and vitreous hemorrhage. The sudden onset of cataract after needle entry into the lens is uncommon but it did occur in reported series [48]. Retinal detachment can occur if the needle penetrates through the retina and causes a retinal break, or if an induced posterior vitreous detachment precipitates a retinal tear. Injecting the triamcinolone acetonide slowly may decrease this risk.

SUSTAINED-RELEASE INTRAVITREAL CORTICOSTEROIDS

Several extended-release steroid preparations are currently undergoing clinical trials. The Retisert implant (developed by Bausch and Lomb and Control Delivery Systems), which is implanted surgically in a similar fashion to the ganciclovir sustained-release implant, releases fluocinolone acetonide in a linear release pharmacokinetics over 3 years. In the phase 3 trial for DME, 80 patients were randomized to receive standard of care (macular grid laser or observation) (n = 28) or either a 0.5-mg (n = 11) or a 2-mg (n = 41) Retisert implant. Enrollment for the 2-mg dose was discontinued early because of an increased incidence of steroid related complications (e.g., IOP, cataract). The 12-month results (Pearson et al. Association for Research and Vision in Ophthalmology, Fort Lauderdale, FL, 2004) reported that more patients randomized to the 0.5-mg group than the standard of care group had resolution of central macular edema (48.8% vs. 25.0%; P = 0.047) and improvement in central retinal thickness (P = 0.003). More patients treated with 0.5 mg Retisert than standard of care gained 15 letters from baseline (19.5% vs. 7.1%; not statistically significant). More patients receiving standard of care lost 15 letters from baseline (14.3% vs. 4.7%). Overall, 70% of patients in the Retisert group, compared with 50% in the standard of care group, had stable visual acuity (P = 0.08).

242 Diabetes and Ocular Disease

The overall incidence of serious ocular adverse events (e.g., increased IOP, cataracts, retinal detachment, vitreous hemorrhage) in the study eye over 12 months was 58.5% in patients receiving the 0.5 mg implant and 10.7% in the standard of care group. The proportion of patients with an increase in IOP ≥30 mmHg in the study eye was higher in the 0.5 mg-group (19.5%) than in the standard of care group (0.0%) with three of eight patients in the 0.5-mg group undergoing a trabeculectomy. In addition, cataract progression at 12 months was 0.0% in the standard of care group versus 54.8% of the 31 patients in the 0.5-mg implant group who had not undergone cataract surgery prior to enrollment in the study. No patients required implant removal or withdrew from the study due to an adverse event.

The Posurdex implant (Allergan, Irvine, CA) is a biodegradable copolymer consisting of 70% dexamethasone (350 or 700 µg ) and 30% polylactic–glycolic acid. The copolymer hydrolyzes to lactic and glycolic acids. Lactic acid is metabolized to H2O and CO2; glycolic acid is either excreted or enzymatically converted to other metabolized species. The implant is delivered via an applicator as an in-office procedure. A phase 2 trial randomized patients with macular edema due to a variety of causes to receiving placebo, or the 350or 700-µg Posurdex implant. Three-month results in the subgroup of patients with DME (n = 172) showed a two or more line visual improvement in 34% of eyes receiving the 700 µg implant, compared with 24.5% receiving the 350 µg implant and 12.7% undergoing observation (Haller JA, American Academy of Ophthalmology, New Orleans, LA, October 2004). The retinal thickness was comparable in all three groups at baseline. At day 90, the observation group had an average increase in thickness of 0.31 microns, while the 350-µg group had an average decrease in thickness of 72 microns and thickness in the 700-µg group decreased on average 157 microns.

ANTIANGIOGENIC AGENTS

The antiangiogenic and antipermeability properties of anti-VEGF agents could potentially be useful in the treatment of diabetic retinopathy. Since DME is a VEGF-mediated disease, intravitreal anti-VEGF may be useful in the treatment of DME. The advantage of intravitreal anti-VEGF agents over intravitreal triamcinolone is that they do not cause cataract or glaucoma, whereas triamcinolone does. However, the half-life of anti-VEGF agents is considerably less than that of triamcinolone; thus patients may need more frequent injections. Ranibizumab (Lucentis, Genentech, San Francisco, CA) is currently in phase 3 clinical testing for DME. Another anti-VEGF agent, Macugen (pegaptanib sodium, formerly NX1838; Eyetech Pharmaceuticals), has completed phase 2 testing for diabetic macular edema.

Ranibizumab is a humanized, antigen-binding fragment (Fab) of a secondgeneration, recombinant mouse monoclonal antibody directed toward VEGF. It consists of two parts: a nonbinding human sequence (humanized), making it less antigenic in humans, and a high-affinity binding epitope (Fab fragment) derived from the mouse, which serves to bind the antigen [49]. It is produced via a plasmid, containing the appropriate gene sequence, inserted into an Escherichia coli

expression vector that undergoes large-scale fermentation. This is drained, and the supernatant collected and purified to produce the active drug. Ranibizumab is FDA approved for choroidal neovascularization due to age-related macular degeneration and there are ongoing clinical trials using Ranibizumab for diabetic macular edema.

In a pilot study [50], 10 eyes of 10 patients with DME involving the center of the macula and best-corrected visual acuity (BCVA) in the study eye between 20/63 and 20/400 were treated with ranibizumab. Three intravitreal injections of ranibizumab (0.3 or 0.5 mg each injection) were administered on day 0, month 1, and month 2, and observation until month 24. Of the 10 patients enrolled, 5 received 0.3 mg and 5 received 0.5 mg ranibizumab. At month 3, 4 of 10 patients gained 15 letters or more, 5 of 10 gained 10 letters or more, and 8 of 10 gained 1 or more letters. At month 3, the mean decrease in retinal thickness of the center point of the central subfield was 45.3 ± 196.3 microns for the 0.3-mg group and 197.8 ± 85.9 microns for the 0.5-mg group.

The anti-VEGF pegylated aptamer Macugen is a polyethylene-glycol (PEG) conjugated oligonucleotide with high specificity and affinity for the major soluble human VEGF isoform, VEGF165. Pegylation decreases the clearance of the drug from the vitreous following intravitreal injection. Aptamers are chemically synthesized short strands of RNA or DNA (oligonucleotides) designed to bind to specific molecular targets based on their three-dimensional structure, and are made using SELEX technology (Systematic Evolution of Ligands by EXponential enrichment). Macugen is an aptamer composed of 28 nucleotide bases that avidly binds and inactivates VEGF165 [51]. Macugen is FDA approved for treatment of choroidal neovascularization in age-related macular degeneration and has completed phase 2 testing for diabetic macular edema [52].

In the Macugen for diabetic macular edema phase 2 trial, 172 patients were randomized to receive intravitreal pegaptanib (0.3, 1, and 3 mg) or sham injections at study entry, week 6, and week 12. Additional injections and/or focal photocoagulation were performed as needed for another 18 weeks. Median visual acuity was better at week 36 with 0.3 mg pegaptanib (20/50), as compared with sham (20/63) (P = 0.04). More patients receiving 0.3 mg pegaptanib gained 10 or more letters of vision (34% vs. 10%, P = 0.003) and 15 or more letters (18% vs. 7%, P = 0.12). Mean central retinal thickness decreased by 68 microns with 0.3 mg pegaptanib, versus an increase of 4 microns with sham (P = 0.02). Focal laser treatment was necessary in fewer subjects in each pegaptanib arm (0.3 mg vs. sham, 25% vs. 48%; P = 0.04).

A retrospective analysis of this trial identified patients with retinal neovascularization [53]. Changes in retinal neovascularization were assessed on fundus photographs and fluorescein angiograms graded at a reading center. Scatter panretinal photocoagulation (PRP) before study enrollment was permitted, but not within 6 months of randomization and study entry. Of the 172 participants, 19 had retinal neovascularization in the study eye at baseline. Excluding 1 who had PRP 13 days before randomization and 2 with no follow-up photographs, 1 of the remaining 16 subjects had PRP during study follow-up. Of these 16 subjects, 8 of 13 (62%) in a pegaptanib treatment group (including the one who received PRP), 0 of 3 in the