RISKS OF INTRAVITREOUS INJECTION:
`A COMPREHENSIVE REVIEW
`
`RAMA D. JAGER, MD, MBA,* LLOYD PAUL AIELLO, MD, PHD,* SAMIR C. PATEL, MD,†
`EMMETT T. CUNNINGHAM, JR., MD, PHD, MPH‡
`
`Purpose: To evaluate the prevalence of the most common serious adverse events
`associated with intravitreous (IVT) injection.
`Methods: A systematic search of the literature via PubMed from 1966 to March 1, 2004,
`was conducted to identify studies evaluating the safety of IVT injection. Data submitted in New
`Drug Applications to the U.S. Food and Drug Administration for drugs administered into the
`vitreous were included where available. Serious adverse events reported in each study were
`recorded, and risk per eye and risk per injection were calculated for the following serious
`adverse events: endophthalmitis, retinal detachment, iritis/uveitis, intraocular hemorrhage,
`ocular hypertension, cataract, and hypotony. Rare complications also were noted.
`Results: Data from 14,866 IVT injections in 4,382 eyes were analyzed. There were 38
`cases of endophthalmitis (including those reported as pseudoendophthalmitis) for a prev-
`alence of 0.3% per injection and 0.9% per eye. Excluding cases reported specifically as
`pseudoendophthalmitis, the prevalence of endophthalmitis was 0.2% per injection and
`0.5% per eye. Retinal detachment, iritis/uveitis, ocular hypertension, cataract, intraocular
`hemorrhage, and hypotony were generally associated with IVT injection of specific com-
`pounds and were infrequently attributed by the investigators to the injection procedure
`itself. Retinal vascular occlusions were described rarely in patients after IVT injection, and
`it was unclear in most cases whether these represented true injection-related complica-
`tions or chance associations.
`Conclusion: The risk of serious adverse events reported after IVT injection is low.
`Nevertheless, careful attention to injection technique and appropriate postinjection mon-
`itoring are essential because uncommon injection-related complications may be associ-
`ated with permanent vision loss.
`RETINA 24:676 –698, 2004
`
`Over the last 2 decades, the use of intravitreous
`
`(IVT) injection has gained increasing acceptance
`in the therapeutic management of many intraocular
`
`From *the Beetham Eye Institute, Joslin Diabetes Center and
`Department of Ophthalmology, Harvard Medical School, Boston,
`Massachusetts; the †Department of Ophthalmology and Visual
`Science, University of Chicago, Chicago, Illinois; and the ‡Depart-
`ment of Ophthalmology, New York University School of Medi-
`cine, New York, New York.
`Samir C. Patel and Emmett T. Cunningham, Jr., are employees
`of Eyetech Pharmaceuticals, Inc., New York, NY.
`Although the term “intravitreal” is used colloquially quite often,
`we have used “intravitreous” as the grammatically correct and
`preferred term in this review.
`Reprints: Dr. Emmett T. Cunningham, Jr., Eyetech Pharmaceu-
`ticals, Inc., 3 Times Square, 12th Floor, New York, NY 10036;
`e-mail: emmett.cunningham@eyetech.com
`
`diseases, particularly disorders affecting the posterior
`segment. A highly effective and frequently used
`means of administering antiviral agents in the treat-
`ment of cytomegalovirus (CMV) retinitis, direct in-
`jection of antiviral agents into the vitreous of patients
`with acquired immunodeficiency syndrome maxi-
`mizes intraocular drug levels while minimizing the
`risk of toxicity associated with systemic administra-
`tion of these agents.1–11 In addition, IVT injection of
`various gases has been used as a less-invasive alter-
`native to scleral buckling for the management of ret-
`inal detachment
`in the setting of pneumatic reti-
`nopexy12–14 and for
`the administration of
`tissue
`plasminogen activator ([TPA] Retavase; Centocor,
`Malvern, PA)15,16 in the treatment of submacular hem-
`
`676
`
`Celltrion Exhibit 1054
`Page 1
`
`

`

`RISKS OF IVT INJECTION • JAGER ET AL
`
`677
`
`orrhage and acute central retinal vein occlusion. Nee-
`dle aspiration of the vitreous—procedurally similar to
`IVT injection—is used routinely to biopsy the vitre-
`ous. In addition to the long-standing practices of vit-
`reous Gram staining, microbial culture, and sensitivity
`analysis in the setting of suspected endophthalmitis,17
`vitreous sampling may also be used to obtain DNA for
`polymerase chain reaction– based analyses for patients
`suspected of having necrotizing herpetic retinitis18 –22
`or toxoplasmic retinochoroiditis.23
`Recent investigations into the treatment of retinal
`neovascularization, retinal edema, and posterior seg-
`ment inflammation have led to the development of
`new biologic and pharmacologic agents that are opti-
`mally administered directly into the vitreous. IVT
`injection of these compounds is being investigated,
`both as a method of achieving vitreous concentrations
`beyond those obtainable with systemic administration
`and as a means of avoiding potential systemic adverse
`effects. Several of these investigational agents, such as
`the therapeutic aptamer oligonucleotide pegaptanib
`sodium (Macugen; Eyetech Pharmaceuticals, New
`York, NY)24 –26 and the monoclonal antibody frag-
`ment ranibizumab (Lucentis; Genentech, San Fran-
`cisco, CA),27,28 are currently undergoing clinical eval-
`uation for the treatment of neovascular age-related
`macular degeneration (AMD) and, in the case of pe-
`gaptanib sodium, for diabetic macular edema and ret-
`inal vein occlusion. In addition, the off-label use of
`IVT triamcinolone acetonide (Kenalog; Bristol-Myers
`Squibb, New York, NY) injection is under investiga-
`tion for a number of disorders, including macular
`edema29,30 and retinal neovascularization.31–34
`Because the potential advantages of IVT injection
`have become more widely appreciated and the number
`of possible applications has grown, questions have
`arisen regarding risks associated with this route of
`administration. Several potential complications of
`IVT injection, such as endophthalmitis, retinal detach-
`ment, traumatic cataract, and intraocular hemorrhage,
`can be vision threatening. A sufficient body of litera-
`ture now exists to support a thorough review of the
`risks associated with IVT injection in managing ocular
`diseases. To that end, as background, we present a
`brief historical overview of the use of IVT injection in
`humans over the last century and a synopsis of re-
`cently published studies on the pharmacokinetic prop-
`erties of agents administered directly into the vitreous.
`We then present the results of a comprehensive, sys-
`tematic review of the literature from which we calcu-
`lated prevalence estimates for the most common com-
`plications associated with IVT injection.
`The analyzed reports of IVT injection varied con-
`siderably in size, design, and indication, and in most
`
`instances, the reported rates of the various complica-
`tions were quite low. Although efforts were made to
`distinguish the risks associated with the specific
`agents being administered from the risks related to the
`IVT injection procedure per se, in some instances,
`such distinctions were difficult to make. Together,
`these factors limit to some extent the generalizability
`of cross-study comparisons and analyses. Despite
`these inherent shortcomings, to our knowledge, this re-
`view represents the most extensive assessment of the
`risks associated with IVT injection compiled to date.
`
`A Historical Perspective on the Use of IVT
`Injection
`
`IVT injection has been used in the treatment of
`human ocular disease for nearly a century. Figure
`124,25,27,28,34 – 48 presents a timeline of important ad-
`vances in the use of this technique from its earliest
`therapeutic application through the present. Although
`this timeline is intended to highlight some of the major
`achievements in the development of IVT therapeutics,
`it is not meant to be an exhaustive compilation or to
`acknowledge the many excellent investigative studies
`that served as a foundation for these advances. Ini-
`tially reported in 1911 by Ohm35 as a means to intro-
`duce air for retinal tamponade and repair of detach-
`ment,
`the IVT administration of pharmaceutical
`agents was pioneered in the mid-1940s with the use of
`penicillin to treat endophthalmitis.36,37 Unfortunately,
`at that time drug administration often was delayed for
`days or even weeks after the infection became estab-
`lished, making most of these early attempts unsuccessful.
`The technique was used infrequently, therefore.
`During the 1950s and 1960s, the use of IVT injec-
`tion still was limited to the administration of air38 or
`silicone oil39 in the treatment of retinal detachment.
`By the 1970s,
`the advent of newer antimicrobial
`agents, combined with the continued poor success of
`alternative treatment options, led to renewed interest
`in IVT therapy for endophthalmitis. Animal studies
`demonstrating the safety of this route of administra-
`tion49,50 were followed by the publication of two case
`series describing successful
`treatment of endoph-
`thalmitis using IVT injection in patients.40,41 Although
`still considered experimental at that time,51 wider use
`of IVT injection to treat endophthalmitis was being
`advocated due to the poor treatment outcomes re-
`ported with systemic administration of antibiotics,
`which generally produced suboptimal drug levels in
`the vitreous.52
`The development of IVT injection for the treatment
`of ophthalmic conditions other than endophthalmitis
`and retinal detachment lagged even further behind,
`
`Celltrion Exhibit 1054
`Page 2
`
`

`

`678
`
`RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES ● 2004 ● VOLUME 24 ● NUMBER 5
`
`Fig. 1. A timeline of important advances in the use of intravitreous (IVT) injections to treat human ocular diseases.
`
`perhaps because of perceived risks related to the pro-
`cedure and because endophthalmitis and retinal de-
`tachment generally have the greatest likelihood for
`acute and irreversible vision loss. Although IVT in-
`jection of corticosteroids was evaluated in an animal
`model of ocular inflammation in the early 1980s,53,54
`there were no publications describing the use of IVT
`corticosteroids in humans until the 1990s. The first
`new application for IVT injection was not tried until
`
`1982, when a pilot study assessing the efficacy and
`safety of 5-fluorouracil delivered as an IVT injection
`for the prevention of postvitrectomy fibroblast prolif-
`eration in patients with proliferative retinopathy was
`initiated.42 This was followed in 1987 by the use of
`IVT ganciclovir sodium (Cytovine; Roche Pharma-
`ceuticals, Nutley, NJ) in the treatment of CMV reti-
`nitis in a patient with acquired immunodeficiency
`syndrome.43
`
`Celltrion Exhibit 1054
`Page 3
`
`

`

`RISKS OF IVT INJECTION • JAGER ET AL
`
`679
`
`The gradually increasing confidence in the safety of
`IVT injection that emerged as a result of these studies
`led to the evaluation of IVT administration of a num-
`ber of other agents throughout the 1990s. Blanken-
`ship44 demonstrated in 1991 that dexamethasone was
`well tolerated but of no therapeutic value for postvit-
`rectomy treatment of diabetic retinopathy. Penfold et
`al34 in a pilot study demonstrated that triamcinolone
`acetonide, a longer-acting corticosteroid, was well tol-
`erated in patients with exudative AMD. Other novel
`applications of IVT injection in the 1990s included the
`use of methotrexate for treatment of ocular lympho-
`ma45 and the injection of TPA for the management of
`submacular hemorrhage.47 In 1998, fomivirsen so-
`dium (Vitravene; Isis Pharmaceuticals, Carlsbad, CA),
`an antisense compound used to treat active CMV
`retinitis, was the first compound to be approved spe-
`cifically for IVT injection by the U.S. Food and Drug
`Administration, representing a milestone achievement
`in the use of IVT therapy (Fig. 1).46
`The pace of development of new applications for
`IVT injection continues to accelerate, with three new
`investigational drug products in clinical trials in the
`early 2000s. These include the vascular endothelial
`growth factor inhibitors ranibizumab27and pegaptanib
`sodium24 –26 for the treatment of neovascular AMD
`and ovine hyaluronidase (Vitrase; ISTA Pharmaceuti-
`cals, Irvine, CA),48 which was recently approved by
`the U.S. Food and Drug Administration as a disper-
`sion agent for other injected drugs and is still seeking
`approval for the treatment of vitreous hemorrhage.55
`The use of IVT injection as a method for localized
`adenovirus vector–mediated gene transfer to treat ret-
`initis pigmentosa and AMD is also being ex-
`plored.56,57 With numerous novel ophthalmic thera-
`pies currently poised to enter into clinical trials, it is
`likely that the number of drugs under development for
`IVT injection will continue to increase rapidly.
`
`Pharmacokinetic Characteristics of Compounds
`Injected Into the Vitreous
`
`Drug concentrations in the vitreous are determined
`not only by the amount of drug given but also by the
`distribution and clearance of such compounds.58,59 As
`with other routes of drug administration, pharmacoki-
`netic characteristics are dependent on both the ana-
`tomical and physiologic features at the site of admin-
`istration58 – 60 and the physicochemical properties of
`the agent administered.61,62 Although several investi-
`gators have explored the pharmacokinetic properties
`of the IVT injection of selected compounds in animal
`in hu-
`models58 – 60,62,63 and to a limited extent
`
`mans64 – 66 over the past 2 decades, this topic remains
`complex and incompletely understood.
`The composition of the vitreous is unique in that it
`is a highly hydrated, avascular, gelatinous body con-
`taining 98% water. Vitreous structural elements com-
`posed of type II collagen and hyaluronic acid occupy
`⬍1% of the total volume.61 The human eye contains 3
`mL to 4 mL of vitreous humor.67 The vitreous can
`move during eye motion, particularly in the elderly
`who normally have a more liquefied vitreous than
`younger individuals.61 By 80 years of age, approxi-
`mately one half of the vitreous in most people is
`estimated to exist in a liquid state.68 Although little is
`known about the effect of age on the disposition of
`compounds administered directly into the vitreous,
`such age-related changes in vitreous characteristics
`merit consideration because many indications for IVT
`injection affect elderly patients disproportionately.
`Although there is a relative barrier between the
`anterior and the posterior segments of the eye, injected
`substances move throughout the vitreous fairly readily
`by either diffusion or bulk flow. Whereas bulk flow
`tends to be the primary means of movement when the
`vitreous is formed, when the vitreous body is partly
`removed, degenerated, or collapsed,
`the exchange
`movement tends to be due largely to diffusion. Gra-
`dients exist in both directions between the vitreous
`and plasma. These gradients are a result of several
`mechanisms, including diffusion and bulk flow pro-
`cesses within the vitreous body as mentioned above,
`the presence of physiologic blood– ocular barriers vis-
`à-vis active and passive transport, and metabolism
`within the eye itself.69
`Alterations of the normal anatomy and physiology
`of the eye and the size of the administered compound
`have been demonstrated to be factors that impact the
`distribution and elimination of drugs from the vitreous
`in animal studies. A history of vitrectomy, aphakia, or
`pseudophakia has been shown to increase the rate of
`distribution and clearance of amphotericin B, for ex-
`ample. Radiolabeled amphotericin B elimination from
`aphakic, vitrectomized rabbit eyes occurred in 1.8
`days compared with 15.1 days for normal rabbit
`eyes.59 A similarly rapid clearance of triamcinolone
`acetonide in aphakic, vitrectomized rabbit eyes (6.5
`days) compared with clearance from aphakic, nonvit-
`rectomized eyes (16.8 days) and normal, phakic, non-
`vitrectomized eyes (41 days) was reported by Schin-
`dler et al.60 In this latter study, high-performance
`liquid chromatography was unable to detect the drug
`in 5 of 6 normal rabbit eyes at 21 days after IVT
`injection. Uptake of triamcinolone acetonide by ocular
`tissues is apparently quite slow, because ⬍1% of the
`compound was found in the iris ciliary body, lens,
`
`Celltrion Exhibit 1054
`Page 4
`
`

`

`680
`
`RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES ● 2004 ● VOLUME 24 ● NUMBER 5
`
`retina, pigment epithelium, and sclera at 72 hours after
`injection.63
`The pharmacokinetic properties of compounds in
`the vitreous have been observed to change in the
`setting of inflammation or infection, such as endoph-
`thalmitis. Using a rabbit model, Coco et al58 estimated
`that the half-life of vancomycin in normal, uninflamed
`eyes was ⬇62 hours, compared with 14 hours in eyes
`in which endophthalmitis had been experimentally
`induced by injecting Staphylococcus aureus into the
`vitreous, a ⬎4-fold increase. The reverse was true for
`plasma concentrations. These data led the investiga-
`tors to suggest that the elimination of vancomycin was
`enhanced in the setting of inflammation or infection,
`most probably due to breakdown of the blood–retinal
`barrier, which when intact acts to limit drug clearance.58
`The size of the administered compound has been
`shown to affect the half-life in the vitreous. Full-
`length radiolabeled humanized monoclonal antibodies
`(molecular weight, ⬇150 kd) injected into the vitreous
`of rhesus macaques were found not to penetrate the
`inner limiting membrane of the retina, while Fab an-
`tibody fragments (molecular weight, ⬇50 kd) diffused
`through the neural retina to the retinal pigment epi-
`thelial layer within 1 hour.62 Consequently, the half-life
`in the vitreous was 5.6 days for the full-length antibody
`and 3.2 days for the Fab antibody fragment.62
`Specific pharmacokinetic profiling of compounds
`administered into the vitreous in humans has been
`limited. Two studies have evaluated the duration of
`detectable concentrations of triamcinolone acetonide
`in aqueous humor samples.64,65 Neither study at-
`tempted to identify the route of elimination of triam-
`cinolone acetonide from the eye or the levels of com-
`pound that could be achieved in various ocular
`compartments after IVT injection. Pharmacokinetic
`studies of pegaptanib sodium injected into the vitreous
`humor of rhesus monkeys have shown that vitreous
`humor and plasma concentrations were linearly re-
`lated to the dose administered with a half-life of ⬇4
`days. In addition, pegaptanib sodium sampled from
`the vitreous at 28 days was fully active.67
`Recent computer simulations have suggested that
`both anterior and posterior routes of elimination may
`be important in the clearance of drugs from the vitre-
`ous.61 Although the crystalline lens is highly imper-
`meable to water and to many drugs, the anterior hya-
`loid membrane, which separates the vitreous from the
`aqueous, is thin and porous, offering little resistance to
`diffusion. Because aqueous turnover is very rapid,
`with a residence time in the eye of ⬇2.5 hours,61 in
`aphakic and pseudophakic patients aqueous clearance
`can represent a major route of elimination of com-
`pounds injected into the vitreous. In contrast, elimi-
`
`nation by the posterior route is facilitated by active
`unidirectional drug transport at the posterior vitreous
`surface, as demonstrated by studies in which fluores-
`cein movement through the vitreous was monitored.66
`The computer model predicts that a relatively small
`molecule will be quickly cleared from the vitreous by
`absorption through the retina or by release into the
`aqueous.61 In contrast, a larger molecule would dif-
`fuse more slowly, relying on bulk flow to transverse
`the vitreous, and clearance will occur primarily
`through the retina and not the aqueous.61 Although
`these findings are largely hypothetical in the absence
`of empirical data, they do suggest that the pharmaco-
`kinetic properties of compounds administered into the
`vitreous can be complex and influenced by a number
`of factors.
`
`Assessment of Complications Associated With the
`Use of IVT Injection
`
`Methods
`
`A systematic review of the literature via PubMed
`from 1966 to March 1, 2004, using the search terms
`“intravitreous,” “intravitreal,” and “endophthalmitis”
`was conducted to identify studies and case series
`reporting the safety of IVT injection. The search was
`limited to primary reports published in English. Pub-
`lications also were retrieved using the “related arti-
`cles” function of PubMed, and bibliographies from all
`articles selected for analysis were reviewed to identify
`additional citations. More than 220 references were
`reviewed.
`The following types of reports were included: (a) all
`randomized, controlled human clinical
`trials using
`IVT injection; (b) prospective or retrospective case
`series that included ⱖ20 eyes; and (c) safety data from
`U.S. Food and Drug Administration New Drug Ap-
`plications for compounds not yet approved for use in
`the United States where adverse events related to IVT
`injection were reported. The following types were
`excluded: (a) reports in which compounds were ad-
`ministered into the vitreous through sclerotomy inci-
`sions during vitrectomy; (b) reports in which com-
`pounds were administered by IVT injection in
`conjunction with other ocular procedures, such as
`scleral buckling or vitrectomy; (c) reports in which
`IVT injections were administered in eyes as a treat-
`ment for endophthalmitis; and (d) reports that, in the
`opinion of the authors, lacked relevant data, provided
`data of unacceptable quality, or duplicated data sets
`found in more-comprehensive publications. There
`were no exclusions based on the length of the fol-
`low-up period.
`
`Celltrion Exhibit 1054
`Page 5
`
`

`

`RISKS OF IVT INJECTION • JAGER ET AL
`
`681
`
`The following information was extracted from each
`report: (a) information on study design; (b) indication
`for IVT injection; (c) description, dose, and volume of
`substance injected; (d) injection procedure and any
`preinjection preparation, such as the use of topical
`antibiotics; (e) number of eyes in which IVT injec-
`tions were administered; (f) total number of injections
`administered; (g) any complications or serious adverse
`events noted; and (h) duration of follow-up. The rate,
`or prevalence, of seven serious adverse events— en-
`dophthalmitis, retinal detachment,
`iritis/uveitis,
`in-
`traocular hemorrhage, ocular hypertension, cataract,
`and hypotony—was recorded. Both prevalence per
`eye and prevalence per injection were then calculated.
`Other rare, but potentially serious, adverse events
`deemed by the reviewers to be possibly related to IVT
`injection also were noted.
`
`Results
`
`Publications identified via the literature search on
`IVT injection that met the study criteria outlined under
`Methods are summarized in Table 1,1– 8,10,11,70 Table
`2,29 –34,71–73and Table 3,12–16,25,48,74 –76 and results of
`combined analyses are presented in Table 4. CMV
`retinitis in patients with acquired immunodeficiency
`syndrome was the indication that yielded the greatest
`number of publications analyzed, with a total of 10
`(Table 1). The antiviral compounds cidofovir (Vistide;
`Gilead Sciences, Foster City, CA),4 –7 ganciclovir
`(with or without intravenous foscarnet sodium [Fos-
`cavir; AstraZeneca Pharmaceuticals, Wilmington,
`DE]),1–3,10,11 and fomivirsen8,70 were used in 909
`eyes. There were no reports describing IVT adminis-
`tration of foscarnet sodium alone with sufficient num-
`bers of eyes to meet the inclusion criteria for this
`review. Treatment of CMV retinitis entailed adminis-
`tering multiple IVT injections, usually once every
`week or every other week for extended periods, re-
`sulting in both the greatest number of total injections
`(n ⫽ 10,839) and the greatest number of injections per
`eye. It should be noted, however, that patients with
`CMV retinitis are at greatly increased risk for retinal
`detachment due to the presence of large areas of
`retinal necrosis. Whether acquired immunodeficiency
`syndrome–associated immunosuppression per se pre-
`disposes human immunodeficiency virus–positive pa-
`tients to procedure-related bacterial
`infections is
`largely unknown.
`Another common use for IVT therapy was triam-
`cinolone acetonide in the off-label treatment of retinal
`neovascularization29,31–34 and macular edema,29,71–73
`with nine publications reporting treatment complica-
`tions. These included three prospective studies,31,33,34
`
`two case series,30,32 and four retrospective chart re-
`views (Table 2).29,71–73 In these reports, a total of
`1,703 eyes received 1,739 injections of triamcinolone
`acetonide. The three largest retrospective chart re-
`views were conducted specifically to investigate the
`prevalence of endophthalmitis and provided informa-
`tion on 1,442 eyes and 1,466 injections.71–73
`The remaining published prospective studies and
`case series covered the use of IVT injection for a
`spectrum of indications (Table 3), including IVT in-
`jection of gas during pneumatic retinopexy to treat
`retinal detachment,12–14,74,75 IVT injection of hyal-
`uronidase48 and TPA15 for vitreous hemorrhage and
`submacular hemorrhage, respectively, IVT injection
`of TPA for retinal vein occlusion,16 IVT injection of a
`vascular endothelial growth factor– blocking agent for
`choroidal neovascularization in patients with neovas-
`cular AMD,25 and IVT injection of methotrexate to
`treat ocular lymphoma.76
`Sufficient data were available from the literature
`search to estimate individually the prevalence of com-
`plications associated with IVT injection of antiviral
`compounds and triamcinolone acetonide, while data
`for all other
`indications for
`IVT injection were
`grouped together. Complications considered the most
`serious and thereby warranting specific scrutiny in-
`cluded endophthalmitis,
`retinal detachment,
`iritis/
`uveitis, ocular hypertension, intraocular hemorrhage,
`cataract, and hypotony. Whenever possible, complica-
`tions caused by the IVT injection procedure were
`distinguished from those attributable to the condition
`being treated or those related to the administration of
`specific therapeutic compounds. The following sec-
`tion provides an estimation of the prevalence of each
`complication and includes a brief discussion of each
`complication within the context of overall risk of IVT
`injection.
`
`Endophthalmitis
`Reports of endophthalmitis associated with admin-
`istration of antiviral agents. The prevalence of en-
`dophthalmitis in eyes with CMV retinitis treated with
`IVT therapy was estimated from the studies listed in
`Table 1 to be 1.3% (12/909) per eye and 0.1% (12/
`10,839) per injection. In the largest clinical trial re-
`viewed, two cases of endophthalmitis were reported in
`association with 1,791 IVT injections of fomivirsen in
`330 eyes.8 In the second largest trial reviewed, a case
`series involving 156 eyes and 2,890 IVT injections of
`ganciclovir, endophthalmitis was diagnosed in four
`eyes, two of which were lost to total retinal necrosis.1
`Microbial culture results were not presented in either
`report. Two other series involving patients receiving
`IVT ganciclovir also identified cases of endophthalmi-
`
`Celltrion Exhibit 1054
`Page 6
`
`

`

`682
`
`RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES ● 2004 ● VOLUME 24 ● NUMBER 5
`
`injected
`increasedvolume
`secondaryto
`
`SubfovealHE
`
`orbitalcellulitis
`opticatrophy;1
`degeneration;1
`RVO;4macular
`retinaltears;1
`4CME;3RAO;5
`
`synechiae
`posterior
`irreversibleHP;18
`
`13transientand3
`
`infection
`2extraocularCMV
`macularscarring;
`irreversibleHP;1
`
`2transientand2
`
`detachment
`
`1choroidal
`
`retinalnecrosis
`
`RDatareasof
`
`infection
`7nonocularCMV
`5opticdiskatrophy,
`
`0.3
`
`0.4
`
`0.5
`
`0.2
`
`1.7
`
`0.5
`
`0.1
`
`3.1
`28
`
`4.8
`44
`
`1
`
`5.9
`54
`
`2.6
`24
`
`5.919.7
`54
`
`179
`
`1.3
`12
`
`1
`
`3
`
`2
`
`3
`
`6
`
`33
`
`54
`
`21¶
`
`126
`
`33
`
`2
`
`16
`
`10
`
`18
`
`1
`
`4㛳
`
`2
`
`4
`
`2
`
`29
`
`5
`
`1
`
`5
`
`5
`
`2
`
`1
`
`4
`
`1
`
`SAEs/Comments
`
`Other
`
`HP
`
`CT
`
`OH‡
`
`HE
`
`I/U
`
`RD
`
`EO
`
`RS,retrospectivestudy.
`HP,hypotony;CS,caseseries;CMV,cytomegalovirus;POS,prospectiveopenstudy;RCT,randomized,controlledtrial;CME,cystoidmacularedema;RAO,retinalarteryocclusion;RVO,retinalveinocclusion;
`SAEs,seriousadverseevents;IVT,intravitreous;EO,endophthalmitis;RD,retinaldetachment;I/U,iritis/uveitis/vitritis/anteriorchamberinflammation;HE,hemorrhage;OH,ocularhypertension;CT,cataract;
`¶Includes20retinalHEsand1hyphema.
`
`injectionbutwereincluded.
`
`㛳SixRDswerereported:twooccurredsubsequenttovitrectomyandwerenotcounted,andfourdidnotoccurimmediatelyaftertheprocedureandwereconsideredbytheauthorstobeunrelatedtothe
`§Atotalof115eyeswereadministeredinjections;safetydatawerepresentedonlyfor93eyesthathadatleast1monthoffollow-up.
`‡Includesintraocularpressureelevationsthatwerereportedonapatient-specificbasis.
`†Oneinjectionpereyewascountedwhenthenumberofadministeredinjectionswasnotprovided.
`*Blankspacesindicatenoreportedevents.
`
`10,839
`1,583
`
`909
`74
`
`CMVretinitisGanciclovir/2mg/8–132wk(range)
`
`CS
`
`injection(%)
`Prevalenceper
`
`Prevalencepereye
`
`(%)
`
`Total
`
`Young,199810
`
`341
`
`31
`
`CMVretinitisGanciclovir/2mg/20.4wk(mean),2–59wk(range)
`
`RS
`
`Young,199211
`
`1,791
`
`330
`
`Fomivirsen/165or330␮g/1–1,024d(range)
`
`CMVretinitis
`
`RCT
`
`Vitravene,20028,70
`
`CMVretinitisGanciclovir/400␮g/9wk(mean),1–39wk(range)
`
`POS
`
`Cochereau-Massin,19912
`
`CMVretinitisGanciclovir/350␮g/19.8⫾2.9wk(mean),1wk–32mo
`
`(range)
`
`Compound/Dose/Follow-up
`
`Indication
`
`CS
`
`Type
`Study
`
`Baudouin,19961
`
`Study
`
`Table1.SAEsReportedfromStudiesInvolvingIVTAdministrationofAntiviralCompounds*
`
`168
`
`37
`
`3,027
`
`710
`
`2,890
`
`Injections†
`
`No.
`
`53
`
`24
`
`57
`
`64
`
`156
`
`Eyes
`No.
`
`246
`46
`
`93§
`27
`
`Cidofovir/20␮g/177d(mean),31–674d(range)
`Cidofovir/10␮g/65.3d(mean),17–140d(range)
`
`CMVretinitis
`CMVretinitis
`
`Cidofovir/20␮g/15wk(mean),0–58wk(range)
`
`CMVretinitis
`
`Cidofovir/20␮g/87.8d(mean),38–186d(range)
`
`CMVretinitis
`
`CMVretinitisGanciclovir/400␮g/53.1wk(mean),6–156wk(range)
`
`POS
`
`CS
`CS
`
`CS
`
`CS
`
`Taskintuna,19977
`Taskintuna,19976
`
`Rahhal,19965
`
`Kirsch,19954
`
`Hodge,19963
`
`Celltrion Exhibit 1054
`Page 7
`
`

`

`RISKS OF IVT INJECTION • JAGER ET AL
`
`683
`
`14P-EO;10culture
`
`positive
`
`14P-EO;10culture
`
`positive
`
`13
`
`9.9**
`9.9**
`100
`
`36.6
`38.3
`
`—7P-EO;0culture
`
`positive
`
`—7P-EO;2culture
`
`positive
`
`—0P-EO;8culture
`
`positive
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`—
`
`0.6
`
`0.6
`1.4
`1.4
`24
`
`8
`
`9
`
`7
`
`1HRVO
`
`13
`
`100
`
`5
`
`NR§
`
`9
`
`NR§
`
`4
`
`4
`
`35
`
`31
`
`4
`
`21
`
`1,739
`
`104
`
`440
`
`922
`
`273
`
`34
`
`27
`
`78
`
`75
`
`16
`43
`
`OtherSAEs/Comments
`
`HP
`
`CT
`
`OH‡
`
`HE
`
`I/U
`
`RD
`
`EO
`
`Injections†
`
`No.
`
`pseudoendophthalmitis;CRVO,centralretinalveinocclusion;ME,macularedema.
`degeneration;HRVO,hemiretinalveinocclusion;CS,caseseries;Inj,injection;NR,notreported;POS,prospective,openstudy;CME,cystoidmacularedema;P-EO,noninfectiousendophthalmitis/
`ocularhypertension;CT,cataract;HP,hypotony;RS,retrospectivestudy;DME,diabeticmacularedema;CNV,choroidalneovascularization;RCT,randomized,controlledtrial;AMD,age-relatedmacular
`SAEs,seriousadverseevents;IVT,intravitreous;TCA,triamcinoloneacetonide;EO,endophthalmitis;RD,retinaldetachment;I/U,iritis/uveitis/vitritis/anteriorchamberinflammation;HE,hemorrhage;OH,
`**IncludedwerestudiesinwhichthenumberofCTs(n⫽8)andthenumberofphakiceyes(n⫽81)onlywerereported.
`¶Culture-positiveinfectiousEOwasdistinguishedfromnoninfectionsEO/pseudo-EOinthesestudies.TheprevalenceofEOwascalculatedwithandwithoutcasesofnoninfectiousED/pseudo-EO.
`
`prevalenceoftheotheradverseevents.
`
`㛳RetrospectivechartreviewsdesignedtodeterminethefrequencyofEOafterTCAinjection.Datawerenotobtainedforotheradverseevents.Theeyesandinjectionswerenotincludedincalculatingthe
`§CTreportedasameanindex(ofopacification)thatincreasedsignificantlyinpatientsreceivingTCAinjection.
`‡Includesintraocularpressureelevationsthatwerereportedonapatient-specificbasis.
`†Oneinjectionpereyewascountedwhenthenumberofadministeredinjectionswasnotprovided.
`*Blankspacesindicatenoeventsreported.Dashesindicatedataexcludedfromcalculations.
`
`Table2.SAEsReportedfromStudiesInvolvingIVTAdministrationofTCA*
`
`(PhakicEyes)
`TotalNo.Eyes
`
`Compound/Dose/
`
`Follow-up
`
`TCA/4mg/3–18mo
`
`(range)
`
`5.7mo(mean)
`mo(range);3rdInj:
`mo(mean),3.9–8.9
`(range);2ndInj:5.7
`3.1–19.57mo
`7.46mo(mean),
`TCA25mg/1stInj:
`
`TCA/25mg/6.64⫾
`
`6.1mo(mean)
`
`TCA/4mg/12mo
`
`(min)
`
`TCA/4mg/6–12mo
`TCA/4mg/12wk(min)
`
`(min)
`
`Subtotal
`
`AMD
`
`DME
`
`AMD
`
`AMD
`
`AMD
`
`DME,CNV,other
`
`Indication
`
`POS
`
`CS
`
`CS
`
`RCT
`
`RCT
`RS
`
`Type
`Study
`
`Penfold,199534
`
`Jonas,200330
`
`Jonas,200332
`
`Gillies,200333
`
`Danis,200031
`Bakri,200329
`
`Study
`
`1,703
`
`104
`
`416
`
`922
`
`261(147)
`
`30(NR)
`
`26(18)
`
`71(48)
`
`75(56)
`
`16(7)
`43(18)
`
`TCA/1or4mg/NR
`
`TCA/4mg/2–9mo
`
`(range)
`
`DME,CME
`
`DME,CME,
`
`CRVO
`
`RS
`
`RS
`
`Prevalenceperinjection
`
`excludingP-EO(%)
`
`excludingP-EO(%)
`
`Prevalencepereye
`
`Prevalenceperinjection(%)
`Prevalencepereye(%)
`
`Total
`
`Roth,200373
`
`Nelson,200372
`
`TCA/4mg/1–9mo
`
`(range)
`
`CME,DME,CNV
`
`RS
`
`Moshfeghi,200371
`ThefollowingstudiesareusedincalculatingtheprevalenceofEOonly.储¶
`
`Celltrion Exhibit 1054
`Page 8
`
`

`

`684
`
`RETINA, THE JOURNAL OF RETINAL AND VITREOUS DISEASES ● 2004 ● VOLUME 24 ● NUMBER 5
`
`vitreous,17hyphema,3retinal
`3retinascars;1RVO;HEsincluded123
`blindness;3toxoplasmaleyeinfections;
`strabismus,5blindness;1transient
`ptosis;2maculardegeneration;5
`posteriorcapsule