The Role of Inflammation in Diabetic Eye Disease

MEDICAL ED NG UC UI AT A CONTINUING TIN IO CON N MEDICAL EDUCATION PUBLICATION CME

ISSUE 23 The Role of In ammation in Diabetic Eye Disease

DANIEL F. KIERNAN, MD Some degree of in ammation is involved in nearly all forms of diabetic eye disease. Prompt recognition that antiin ammatory therapy may be warranted, particularly in cases of diabetic macular edema, may lead to improved outcomes. OD OS FIGURE 1 Color montage stereo photographs of a patient’s right and Vascular endothelial growth factor (VEGF) inhibitors are left fundi demonstrating moderate non-proliferative DR and hard yellow a signifi cant therapeutic advancement in the treatment of dia- exudates characteristic of clinically signifi cant diabetic maculopathy. betic retinopathy (DR) and diabetic macular edema (DME).1 (Images courtesy of Dr. Kiernan.) But it is clear from multiple lines of evidence that DR and DME are complex, multifactorial conditions for which anti-VEGF trols.6 Further, the levels of these cytokines, chemokines, and injections are not a panacea.2 Indeed, anti-VEGF therapy is adhesion molecules are found to be increased in proportion ineff ective or insuffi ciently eff ective for a substantial portion of with disease severity.6 patients.3-5 Given the pathophysiological role of infl ammation While VEGF may be a dominant contributor to pathologic in DR and (especially chronic) DME, antiinfl ammatory drugs neovascularization in DR and DME, other factors, including can be an essential component of treatment. infl ammatory mediators, are likely involved. Treatment with anti-VEGF agents, while eff ective, may not entirely address the INFLAMMATORY MEDIATORS underlying infl ammatory processes involved in these diseases. In patients with diabetes, elevated blood glucose and advanced glycosylation end products cause dysregulation of EVIDENCE FROM PIVOTAL STUDIES the retinal vasculature, including pericyte and endothelial In the phase 3 RISE and RIDE trials, patients with DME cell loss, capillary dropout, and consequent ischemia. Th is who were randomized to receive monthly injections of active dysregulation, in turn, leads to the release of a host of proin- drug (0.3 or 0.5 mg ranibizumab) showed strong visual acuity fl ammatory growth factors, including VEGF, and cytokines.2 Aqueous and vitreous samples taken from patients with DR and DME consistently show elevated levels of infl amma- See INSIDE for: tory molecules such as interleukin (IL)-1α, IL-1β, IL-6, IL-8, Antiinfl ammatory Eff ects of Amniotic Membrane interferon-γ, intercellular adhesion molecule (ICAM)-1, and by Scheff er CG Tseng, MD, PhD monocyte chemotactic protein (MCP-1) versus normal con-

To obtain CME credit for this activity, go to http://cme.ufl .edu/ed/self-study/toai/ Supported byTopics an unrestricted in OCULAR educational ANTIINFLAMMATORIES grant from Shire. 1 gains and reductions in central foveal TOPICS IN OCULAR ANTIINFLAMMATORIES, ISSUE 23 thickness.5 Th ose who had initially been STATEMENT OF NEED Explorer® 6.0 or newer, or Mozilla® Firefox® 2.0 or newer randomized to receive sham injections The control of ocular infl ammation is a critical aspect of (JavaScript™ and Java™ enabled). For Mac® users: Mac OS® medical and surgical ophthalmic practice. Despite their X 10.4 (Tiger®) or newer; Safari™ 3.0 or newer, Mozilla® were able to crossover and receive side eff ects, antiinfl ammatory drugs are used to treat a Firefox® 2.0 or newer; (JavaScript™ and Java™ enabled). ranibizumab 0.5 mg aft er 24 months. very wide range of conditions throughout the eye, from Internet connection required: Cable modem, DSL, or ocular surface disease and allergic conjunctivitis to poste- better. Interestingly, at 36 months, anatomical rior segment conditions. Use of antiinfl ammatory agents is also critical in ocular surgery, contributing greatly to DATE OF ORIGINAL RELEASE September 2018. Ap- improvements similar to the original patient comfort and positive outcomes. proved for a period of 12 months. treatment groups were seen in the group The ocular antiinfl ammatory landscape is changing as ACCREDITATION STATEMENT research reveals more about the role of infl ammation in originally randomized to sham (aver- This activity has been planned and implemented in ac- a range of ocular conditions and as new antiinfl ammatory cordance with the accreditation requirements and poli- age OCT thickness of 194.1 μm in the agents enter the market.1,2 Twenty years ago, for example, cies of the Accreditation Council for Continuing Medical the idea of using a topical corticosteroid to treat dry eye sham/0.5 mg group, versus 223.4 μm Education (ACCME) through the joint providership of the and/or allergic conjunctivitis was viewed with alarm; University of Florida College of Medicine and Candeo in the 0.3 mg group and 201.9 μm in today, it is accepted practice. Clinical/Science Communications, LLC. The University of the 0.5 mg group). But the visual acuity Although corticosteroids and nonsteroidal antiinfl am- Florida College of Medicine is accredited by the ACCME matory drugs (NSAIDs) have been the mainstays of the to provide continuing medical education for physicians. gains in the group originally random- ocular anti-infl ammatory armamentarium, a number of ized to sham did not catch up to those new agents with novel mechanisms of action (and new CREDIT DESIGNATION STATEMENT 3 ocular drug delivery systems) have come to market or are The University of Florida College of Medicine designates 3,4 in the two active treatment groups. Th e being made ready for market. this enduring material for a maximum of 1 AMA PRA proportions of patients who had gained As indications expand and change, and as new drugs, Category 1 Credit™. Physicians should claim only the formulations, and delivery systems become available, credit commensurate with the extent of their participa- at least 15 Early Treatment Diabetic clinicians require up-to-date protocols for drug selec- tion in the activity. Retinopathy Study (ETDRS) letters from tion and use. Such protocols are also needed for routine (but nevertheless off -label) uses of corticosteroids and EDITORIAL BOARD/FACULTY ADVISORS baseline to 36 months were 19.2% for NSAIDs because important diff erences in effi cacy, safety, Marguerite B. McDonald, MD, FACS, practices at sham/0.5 mg, 36.8% for 0.3 mg, and and tolerability exist between these classes and among Ophthalmic Consultants of Long Island, and is a clinical formulations within each of these classes.5,6 professor of ophthalmology at the New York University 40.2% for 0.5 mg in RIDE; and 22.0%, By putting the latest published evidence into the context School of Medicine. She is also an adjunct clinical profes- 51.2%, and 41.6%, respectively, in RISE.3 of current clinical practice, Topics in Ocular Antiinfl amma- sor of ophthalmology at Tulane University Health Sciences tories equips ophthalmologists to maintain competen- Center. Dr. McDonald is a consultant for Allergan, Alcon, Looking at the eff ect of a single year cies and narrow gaps between their actual and optimal Bausch + Lomb, BlephEx, FOCUS Laboratories, Shire, of monthly ranibizumab treatment, the infl ammation management practices, across the range and J&J Vision. of clinical situations in which current and novel ocular Victor L. Perez, MD, is a professor of ophthalmology at mean number of letters gained was 2.8 antiinfl ammatories may be used. the Duke University School of Medicine. He is also the in the sham/0.5 mg group, versus 10.6 director of Duke Eye Center’s Ocular Immunology Center REFERENCES and Ocular Surface Program. Dr. Perez is a consultant for letters in the 0.3 mg group and 11.1 1. Song JS, Hyon JY, Lee D, et al. Current practice pattern Allergan, Shire, EyeGate, and TopiVert. He is also a stock letters in the 0.5 mg group (though the for dry eye patients in South Korea: a multicenter study. shareholder for EyeGate. Korean Journal of Ophthalmology. 2014;28(2):115-21. Matthew J. Gray, MD, is an assistant professor in the groups at these timepoints were not 2. Ciulla TA, Harris A, McIntyre N, Jonescu-Cuypers C. Treat- department of ophthalmology at the University of fully comparable, given the delay in ment of diabetic macular edema with sustained-release Florida College of Medicine. He states that in the past 12 3 glucocorticoids: intravitreal triamcinolone acetonide, months, he has not had a fi nancial relationship with any treatment for the original sham group). dexamethasone implant, and fl uocinolone acetonide commercial organization that produces, markets, resells, The results of the delayed treatment implant. Expert Opin Pharmacother. 2014;15(7):953-9. or distributes healthcare goods or services consumed by group portray an overall picture of an 3. Maya JR, Sadiq MA, Zapata LJ, et al. Emerging therapies or used on patients relevant to this manuscript. for noninfectious uveitis: what may be coming to the Daniel F. Kiernan, MD, practices at Ophthalmic Consul- ongoing infl ammatory condition that clinics. J Ophthalmol. 2014;2014:310329. tants of Long Island in Lynbrook, Mineola, and Rockville 4. Sheppard JD, Torkildsen GL, Lonsdale JD, et al, and Centre, NY. He has received grant/research support from results in permanent architectural and/ the OPUS-1 Study Group. Lifi tegrast ophthalmic solu- Allergan, Alimera Sciences, and Graybug. Dr. Kiernan is or neurodegenerative changes, including tion 5.0% for treatment of dry eye disease: results also a consultant for Allergan, Alimera Sciences, Genen- of the OPUS-1 phase 3 study. Ophthalmology. 2014 tech, Regeneron, and Mallinckrodt Pharmaceuticals, and neural cell loss and damage, fi brosis, and Feb;121(2):475-83. is on the speakers’ bureau for Allergan, Alimera Sciences, pigmentary changes. Th is points to the 5. Fong R, Leitritz M, Siou-Mermet R, Erb T. Loteprednol Genentech, and Regeneron. benefi t of earlier treatment and perhaps etabonate gel 0.5% for postoperative pain and infl am- Scheff er CG Tseng, MD, PhD, is the medical director of mation after cataract surgery: results of a multicenter the Ocular Surface Center in Miami, FL and serves as the to more comprehensive antiinfl amma- trial. Clin Ophthalmol. 2012;6:1113-24. chief technology offi cer of TissueTech, Inc. in Miami, FL. He 6. Singer M, Cid MD, Luth J, et al. Incidence of corneal melt states that he has received grant/research support from tory management. in clinical practice: our experience vs a meta-analysis National Institutes of Health and the National Eye Institute. Th e Diabetic Retinopathy Clinical of the literature. Clin Exp Ophthalmol. 2012;S1:003. Dr. Tseng is also a stock shareholder for TissueTech, Inc. Research Network (DRCR.net)’s Proto- OFF-LABEL USE STATEMENT DISCLAIMER col I explored the eff ects of ranibizumab This work may discuss off -label uses of medications. Participants have an implied responsibility to use the newly acquired information to enhance patient outcomes and GENERAL INFORMATION 0.5 mg plus prompt (within 1 week) or professional development. The information presented in deferred (within ≥ 24 weeks) focal/grid This CME activity is sponsored by the University of Florida this activity is not meant to serve as a guideline for patient College of Medicine and is supported by an unrestricted care. Procedures, medications, and other courses of diag- laser photocoagulation; sham injection educational grant from Shire. nosis and treatment discussed or suggested in this activity plus prompt laser; and intravitreal tri- The University of Florida College of Medicine designates should not be used by clinicians without evaluation of this activity for a maximum of 1 AMA PRA Category 1 their patients’ conditions and possible contraindications amcinolone injection 4 mg plus prompt Credit™. There is no fee to participate in this activity. In or dangers in use, applicable manufacturer’s product laser. The overall outcomes favored order to receive CME credit, participants should read information, and comparison with recommendations of the report, and then take the posttest. A score of 80% other authorities. ranibizumab with either prompt or is required to qualify for CME credit. Estimated time to deferred laser, though interestingly, the complete the activity is 60 minutes. On completion, take COMMERCIAL SUPPORTERS the test online at http://cme.ufl .edu/ed/self-study/toai/ This activity is supported by an unrestricted educational outcomes for triamcinolone plus laser System requirements for this activity are: For PC us- grant from Shire. were comparable to the ranibizumab ers: Windows® 2000, XP, 2003 Server, or Vista; Internet groups for pseudophakic patients.7,8

2 Topics in OCULAR ANTIINFLAMMATORIES To obtain CME credit for this activity, go to http://cme.ufl .edu/ed/self-study/toai/ A post-hoc analysis of pooled data from the two ranibizumab treatment arms found a signifi cant association between CORE CONCEPTS the improvement in best-corrected visual acuity (BCVA) at 12 ✦ Infl ammation is among the pathophysiological 4 weeks and at 1 and 3 years. Researchers stratifi ed patients by mechanisms at work in diabetic eye disease. level of treatment response (change in BCVA from baseline ✦ Elevated levels of infl ammatory mediators are found to week 12): < 5 letter improvement, 5 to 9 letters of improve- in aqueous and vitreous samples taken from patients ment, or ≥ 10 letters of improvement. Th ey found that only a with DR and DME and increase in proportion to disease minority of those with the lowest level of improvement at 12 severity. weeks would go on to experience clinically signifi cant gains ✦ Evidence from clinical studies of anti-VEGF treatments in BCVA with continued treatment over the ensuing 1 to 3 4 suggest that damage due to infl ammatory changes can years. Th us, where initial response to anti-VEGF therapy result from delayed treatment. is minimal, looking to alternative or adjunctive treatments,

including corticosteroids, makes sense. ✦ In general, initial response to anti-VEGF therapy in DME is predictive of long-term response. IMAGING AND MONITORING ✦ A switch to or addition of corticosteroid may be At present, spectral-domain optical coherence tomogra- warranted in cases of poor initial response to anti-VEGF phy (SD-OCT) is the gold-standard for assessing treatment therapy. response in DME; fl uorescein angiography is also commonly ✦ There is limited evidence that adding steroid to anti-VEGF used to thoroughly characterize the extent of nonperfusion in persistent DME is of benefi t to visual acuity. and vascular leakage. Ultra-widefi eld photography and an- ✦ SD-OCT and ultra-widefi eld angiography are allowing giography are increasingly favored to help identify areas of for more detailed, comprehensive examination and peripheral ischemia, which may contribute to more center- treatment monitoring. involved disease by generating cytokines and other infl amma- ✦ Intravitreal corticosteroid implants oﬀ er long-term tory mediators (Figures 1 and 2).9 Some researchers have found antiinfl ammatory coverage but are accompanied by risks that DME with subfoveal neuroretinal detachments detectable of cataract progression and IOP elevation. on SD-OCT may represent a specifi c disease presentation as- ✦ Novel anti-VEGF agents and combinations, plus sociated with a higher concentration of infl ammatory media- alternative routes of drug administration, may be on the tors, such as IL-6, in the vitreous.6 Hyperrefl ective spots or foci horizon for DR and DME treatment. may be another SD-OCT indicator of DR and DME severity.6 A number of studies have looked at a retinal “ischemic index” (essentially, a ratio of the nonperfused retinal area to ANTIINFLAMMATORY TREATMENT OPTIONS the total retinal area), demonstrating that eyes with larger Currently, in the absence of DME, the only approved areas of retinal nonperfusion and neovascularization tended treatment for diabetic retinopathy is ranibizumab 0.3 mg. to have more severe, recalcitrant DME.9 Ultra-widefi eld imag- However, it is clear from clinical studies of the other available ing may be helpful in guiding targeted sectoral laser therapy anti-VEGF agents and the corticosteroids approved for DME to address areas of peripheral ischemia which may indirectly that these treatments provide secondary improvements in contribute to more central disease and might otherwise be diabetic retinopathy severity scores.5,11,12 missed. DRCR.net researchers are conducting a prospective Most of my patients with center-involving, clinically study to explore the impact of ultra-widefi eld imaging on the signifi cant DME receive anti-VEGF injections as fi rst-line ability to predict disease progression over time.10 therapy, and I turn to corticosteroids when there is an inadequate anatomical and/or visual response aft er the fi rst two to three monthly injections. Options include preservative-free triamcinolone, used off -label; the biodegradable dexamethasone intravitreal implant (Ozurdex 0.7 mg, Allergan,), which lasts 3 to 6 months; and the fl uocinolone acetonide (FAc) intravitreal implant (Iluvien 0.19 mg, Alimera Sciences), which lasts up to 36 months. Because of the risk of elevated intraocular pressure (IOP) resulting in incisional surgery during the FAME clinical trial, the FDA indication for the FAc implant specifi es a prior trial of corticosteroids to test for a clinically signifi cant rise in IOP.13-15 OD OS In my practice, if IOP is not signifi cantly elevated aft er 3 to FIGURE 2 Widefi eld montage fl uorescein angiographic images of a 4 weeks with topical corticosteroids, intravitreal triamcino- patient’s right and left fundi demonstrating ischemic peripheral areas due to lone, or the dexamethasone implant, or is controllable with proliferative DR treated with pan-retinal photocoagulation. Shadowing due to IOP-lowering drops, I may proceed to the longer-acting FAc an inferior vitreous hemorrhage is present OD. (Images courtesy of Dr. Kiernan.) implant for patients who have persistent fl uid as seen on optical

To obtain CME credit for this activity, go to http://cme.ufl .edu/ed/self-study/toai/ Topics in OCULAR ANTIINFLAMMATORIES 3 subcutaneous adrenocorticotropic hormone (ACTH) gel, has a broad indication for ophthalmic infl ammatory conditions, and I have used it successfully in some patients with recalcitrant posterior uveitis or multifocal choroiditis who happen to be diabetic. ACTH binds to melanocortin receptors and is thought to exert its antiinfl ammatory eff ects primarily by stimulating glucocorticoids and potentially through other mechanisms.17,18 Its use is uncommon in ophthalmology largely because there is a paucity of relevant, controlled clinical data 17,18 OD OS about it. In general, adverse events are likely to be similar in nature to those of systemic corticosteroids.17 FIGURE 3 Spectral domain optical coherence tomography images of a On the horizon a micro-injection preparation of triam- patient’s right and left maculae demonstrating persistent DME before (top images) and after intravitreal anti-VEGF injections. The decision was made cinolone acetonide, intended for administration to the supra- to initiate intravitreal corticosteroid therapy to manage this persistent DME. choroidal space, is being investigated in phase 2 trials as an (Images courtesy of Dr. Kiernan.) adjunct to afl ibercept for DME.19 Further, a novel anti-VEGF agent, brolucizumab, which has shown robust phase 3 data for neovascular age-related macular degeneration, will also be coherence tomography and therefore may need longer-term investigated in the treatment of DME.20,21 Other agents that antiinfl ammatory treatment (Figure 3). combine VEGF and angiopoietin-2 inhibition are also being investigated clinically for DME.22,23 GENERAL CONSIDERATIONS Similar to the results from clinical trials, I fi nd that about REMAINING QUESTIONS 30% of the patients I treat with the dexamethasone intravit- Ultimately, the data to guide our adjunctive use of an- real implant develop elevated IOP, which is manageable with tiinfl ammatory agents to treat DR and DME is somewhat topical therapy in the majority of cases. In some situations (eg, limited. Th e recently published DRCR.net Protocol U ex- when IOP is above 30 mm Hg while on ocular hypotensive plored the addition of a dexamethasone implant to ongoing drops) I may refer patients to a glaucoma specialist for con- ranibizumab injections in patients with persistent DME aft er sideration of laser trabeculoplasty or other options. at least three anti-VEGF injections.24,25 Patients were random- In addition to monitoring IOP, cataract progression is ized to receive ranibizumab plus dexamethasone implant or a concern for patients receiving intravitreal corticosteroid. ranibizumab plus sham injection. Th is investigation found However, clinical trials of these implants generally show good that at 24 weeks, although the addition of dexamethasone visual results for patients who start out phakic and undergo had a more pronounced eff ect on retinal thickness, it did not cataract surgery.14,15 improve visual acuity over and above continued ranibizumab When cataracts are visually signifi cant and surgery is therapy.24,25 Likewise, a recent systematic review found low- indicated, I prefer to have a longer-acting antiinfl ammatory quality evidence (mostly from trials involving triamcinolone product on board, such as the dexamethasone or the long- and bevacizumab) to suggest that, in general, the combination acting FAc implant. Th is helps to ensure that the retina is of anti-VEGF and corticosteroid injections did not provide as fl at as possible and may also help limit the infl ammatory signifi cant visual benefi t over monotherapy.26 impact of the surgery. Another consideration for cataract Yet infl ammation management in DR and DME, particu- surgery, which typically prompts a referral, is preoperative larly recalcitrant cases, remains important. While adjunctive dry eye disease.16 Dry eye symptoms are common in patients therapy may or may not be warranted in many cases, a switch with diabetes, and there is evidence of altered tear parameters to corticosteroid treatment because of inadequate response to and reduced corneal nerve density in these patients, both of anti-VEGF therapy or, occasionally, systemic safety concerns, which can impact surgical outcomes.16 is oft en the reasonable choice.

LOOKING AHEAD Daniel F. Kiernan, MD, practices at Ophthalmic Consultants of Long Island Opportunities remain to address the underlying infl am- in Lynbrook, Mineola, and Rockville Centre, NY. He has received grant/ matory component of DR and DME with suﬃ cient power research support from Allergan, Alimera Sciences, and Graybug. Dr. Kiernan as well as appropriate safety. Th e manifestations of diabetes is also a consultant for Allergan, Alimera Sciences, Genentech, Regeneron, and and diabetic eye disease vary from patient to patient, and we Mallinckrodt Pharmaceuticals, and is on the speakers’ bureau for Allergan, do not currently have a way to readily measure, for example, Alimera Sciences, Genentech, and Regeneron. Medical writer Jennifer Zweibel the level of infl ammatory cytokines in aff ected eyes. So, with of Candeo Clinical/Science Communications, LLC, assisted in the preparation current technology, it is diﬃ cult to customize a proactive an- of this manuscript. tiinfl ammatory treatment plan for DR or DME; our approach is necessarily to let response to initial treatment guide us. One systemic therapy that has been available for decades, KIERNAN REFERENCES continue on page 9

4 Topics in OCULAR ANTIINFLAMMATORIES To obtain CME credit for this activity, go to http://cme.ufl .edu/ed/self-study/toai/ Antiin ammatory E ects of Amniotic Membrane CORE CONCEPTS ✦ Amniotic membrane therapy is widely used in SCHEFFER CG TSENG, MD, PHD Over the the management of ocular surface disorders with past two decades, amniotic membrane has proved to infl ammation. It has two distinct types of ocular surface be an e ective treatment modality for an increasing usage: as a permanent surgical graft or as a biological bandage or patch. number of ocular surface diseases with in ammation. Its ophthalmic applications should continue to grow, thanks ✦ In ocular surface repair or reconstruction, the most important properties of human amniotic membrane to the new discovery about a crucial active component include the antiinfl ammatory eff ect through suppression and, accordingly, a greater understanding of the tissue’s of proinfl ammatory cells and the ability to promote antiin ammatory and other therapeutic actions. epithelialization. Amniotic membrane also possesses potent antiscarring and antiangiogenic actions that are benefi cial for surface restoration in a variety of ocular Amniotic membrane, the innermost layer of the fetal mem- surface pathologies. brane complex, consists of a thick basement membrane, a thin ✦ The mechanism of action of amniotic membrane’s epithelium, and an avascular stroma. It is a versatile biological therapeutic eff ects involves certain signaling molecules and tissue with clinical uses in multiple medical fi elds including regulatory factors present in the tissue. A recently identifi ed ophthalmology. Th e ophthalmic indications for amniotic matrix component named HC-HA/PTX3 is likely one of the membrane therapy encompass a wide range of conjunctival active components responsible for amniotic membrane’s and corneal conditions where infl ammation control and tissue antiinfl ammatory and other therapeutic actions. repair are desirable (Table I). A few examples are persistent ✦ The HC-HA/PTX3 complex exerts broad antiinfl ammatory epithelial defects, corneal ulcers, pterygium, chemical burns, actions and regenerative healing by orchestrating recurrent epithelial erosion, Stevens-Johnson syndrome, and multiple cellular actions on the ocular surface. limbal stem cell defi ciency. Th is article focuses on amniotic ✦ Cryopreservation devitalizes living cells, retains the native membrane’s antiinfl ammatory eff ect and its therapeutic im- architecture of amniotic membrane and maintains the plications in the management of ocular surface disorders. quantity and activity of key biological signals present in the tissue such as HC-HA/PTX3A. Dehydration, by HISTORICAL PERSPECTIVES comparison, is a harsher process with detrimental eff ects on the tissue’s properties. Amniotic membrane use is not a recent development in ophthalmology. Its history goes back to 1940, when De Rötth ✦ Amniotic membrane therapy has shown promise for a role in the management of DED. Its clinical benefi ts for moderate- to-severe DED include not only reduction of ocular surface TABLE I infl ammation but also regeneration of corneal nerves. Ophthalmic Indications for Amniotic Membrane Therapy

AS A SURGICAL GRAFT fi rst documented the successful use of a fetal membrane graft • Pterygium and pinguecula for repair of conjunctival defects.1 Several years later, Sorsby • Bulbar conjunctival scarring and colleagues reported using amniotic membrane as a tem- • Conjunctivochalasis porary patch to treat ocular burns.2,3 Th ese initial studies, • Infectious keratitis and scleritis however, did not lead to further investigation into amniotic • Symptomatic bullous keratopathy membrane’s ocular use. For fi ve decades, the method was • Partial limbal stem cell defi ciency largely neglected and hardly ever mentioned in the literature. • Total limbal stem cell defi ciency with limbal transplantation In 1995, my colleague Kim and I reported using amniotic • Glaucoma surgery (fi ltering bleb leakage, tube exposure, etc.) membrane transplantation for ocular surface reconstruction in a rabbit limbal stem cell defi ciency model.4 We found that AS A BIOLOGICAL BANDAGE/PATCH glycerin-preserved human amniotic membrane promotes • Refractory ulcerative keratitis corneal recovery in rabbits aft er total corneal epithelial re- • Acute chemical/thermal burns moval and a limbal lamellar keratectomy. It appeared that • Stevens-Johnson syndrome the amniotic membrane graft acted as a substrate to support • Persistent epithelial defect regeneration of limbal stem cells, just as good topsoil in the • Recurrent corneal erosion garden supports the growth of newly planted seeds. Th is dis- • Dry eye disease covery helped explain why limbal stem cell transplantation • Haze after refractive surgery performed back then would sometimes fail with no obvious

To obtain CME credit for this activity, go to http://cme.ufl .edu/ed/self-study/toai/ Topics in OCULAR ANTIINFLAMMATORIES 5 reason for failure, a problem that we had struggled for years to address. Th e rabbit study suggested for the fi rst time that the answer lies in the lack of a supporting stromal environment, and that amniotic membrane transplantation as an adjunctive measure can improve the outcome of stem cell transplantation. Th is is now a well-established concept. In retrospect, our work marked the revival of amniotic membrane transplantation in the fi eld of ophthalmology. From then on, the technique truly took hold and, with the advent of the cryopreservation method, in 1997, grew rapidly in popu- FIGURE 1 Amniotic membrane’s modes of action. A) As a permanent graft, to fi ll in stromal defect and allow the epithelium to grow on larity. Numerous clinical studies took place, establishing the top of the graft. B) As a biological bandage, to control ocular surface evidence basis for the therapy’s eﬃ cacy. infl ammation and promote healing.(Modifi ed from Hosam Sheha, MD, PhD & Scheﬀ er CG Tseng, MD, PhD, Clinical Atlas Clinical of Procedures in Ophthalmic and THE MOLECULAR MECHANISM Oculofacial Surgery 2012.) Th e therapeutic eff ect of amniotic membrane is believed to originate from its innate wound healing, antiscarring, antian- as either a permanent surgical graft or a temporary biological giogenic, and, most importantly, its antiinfl ammatory proper- bandage or patch (Figure 1). When used as a permanent graft , ties. Indeed, when cryopreserved amniotic membrane received amniotic membrane is surgically transplanted—by either su- approval through Request for Designation from the Food and tures or fi brin glue—to the cornea, conjunctiva, or other tissue Drug Administration (FDA) for ocular surface reconstruc- planes to fi ll in a void for repair or reconstruction (Figure 2A tion, in 2001, it was approved to promote healing through and B). Th e membrane will be remodeled during wound heal- antiinfl ammatory, antiscarring, and antiangiogenic eff ects. ing to become part of the host tissue, a process that may take Amniotic membrane has long been thought to produce a whole month. When applied as a temporary ocular surface these biological actions through the release of certain regula- dressing, the membrane exerts its biological actions without tory factors within the tissue.5 However, it is only recently that full integration into the host tissue (Figure 2C). It needs to be the molecular identity of such crucial components has begun removed upon healing, though in cases where infl ammation to come to light. One may naturally assume that amniotic is intense, the tissue tends to undergo dissolution. membrane’s complex actions is likely to be based on a sym- Amniotic membrane’s structural integrity and biological phony of biological molecules, but our cumulative research properties are aff ected diff erently by the processing methods. over more than a decade suggests otherwise. In 2006, we fi rst In the US, preservation of amniotic membrane relies on two reported that amniotic membrane stromal matrix promotes main types of techniques: cryopreservation, which involves apoptosis of activated macrophages to suppress proinfl amma- freezing of the tissue at very low temperatures; and dehydration tory responses.6,7 We then demonstrated in an experimental by heat drying following high salt extraction. Th is dehydrated study in 2008 that such antiinfl ammatory activity is retained amniotic membrane has not been cleared by FDA to possess in soluble amniotic membrane extract.8 In 2009, we success- the aforementioned anti-infl ammatory and antiscarring ac- fully identifi ed and purifi ed heavy chain-hyaluronic acid/ tion, and is cleared only as a wound covering. In a recent study pentraxin 3 (HC-HA/PTX3) for the fi rst time from amniotic of a head-to-head analysis of cryopreservation and dehydra- membrane extract. Th is HC-HA/PTX3 complex is a unique tion using biochemical and functional assays, we found that matrix component that retains fresh amniotic membrane’s cryopreserved amniotic membrane maintains the quantity multifactorial antiinfl ammatory actions.9 and activity of HC-HA/PTX3, while dehydrated tissues are Further studies have shown that HC-HA/PTX3 is a key signaling molecule in amniotic membrane that orchestrates multiple cellular actions to suppress infl ammation and promote regenerative healing.7 Th e complex promotes apoptosis of macrophages and suppresses lipopolysaccharide-induced macrophage infi ltration of the cornea. It facilitates apoptosis of activated neutrophils and lymphocytes and inhibits activation of helper T cells. Such antiinfl ammatory eff ect in turn leads to enhanced wound repair and decreased scar formation, and the HC-HA/ PTX3 complex confers a direct antiscarring eff ect as well—by inhibiting activities of fi broblasts on the ocular surface.9-12 Th e molecule also has profound infl uences on the behavior of vascular endothelial cells and, as a result, an antiangiogenic eff ect. FIGURE 2 Sutureless application of amniotic membrane. A and B) Amniotic membrane graft secured with fi brin glue after pterygium excision. C) Self-retained biological bandage applied to the ocular surface MODES OF USAGE for neurotrophic keratitis. (Images Courtesy of Hosam Sheha, MD, PhD.) On the ocular surface, amniotic membrane can be used

6 Topics in OCULAR ANTIINFLAMMATORIES To obtain CME credit for this activity, go to http://cme.ufl .edu/ed/self-study/toai/ structurally compromised and almost completely defi cient in gravate tear fi lm insuffi ciency and the self-perpetuating cycle the complex’s integral components.13 of ocular surface deterioration. Rich in neurotrophic factors, particularly nerve growth factor (NGF), amniotic membrane A UNIQUE ANTIINFLAMMATORY THERAPY is thought to facilitate regeneration of corneal nerves.19 Th i s Th e HC-HA/PTX3 complex stands out as a unique class is supported by the effi cacy amniotic membrane therapy has of antiinfl ammatory therapy. Conventional antiinfl ammatory demonstrated in not only restoring corneal sensitivity in eyes agents, such as corticosteroids, nonsteroidal antiinfl ammatory with DED but also pain control in patients with neuropathic drugs (NSAIDs), and cyclosporine, tend to target a specifi c ac- corneal pain.15,20 tion of one particular cell type. Th e HC-HA/PTX3 complex in Amniotic membrane’s regenerative eff ect distinguishes it the amniotic membrane, by comparison, acts on neutrophils, from conventional antiinfl ammatory therapies such as cyclo- macrophages, and lymphocytes that mediate both innate and sporine, corticosteroids, and NSAIDs, which are potentially adaptive immune responses. By downregulating multiple types deleterious to corneal nerves.21-23 Another advantage of amni- of infl ammatory cells and addressing various cellular actions, otic membrane is its lasting eff ectiveness. While conventional HC-HA/PTX3 exerts a broader and likely more potent anti- therapies require a daily maintenance dose to be eff ective, infl ammatory eff ect. Meanwhile, HC-HA/PTX3 has not been amniotic membrane is able to produce more than 3 months associated with any adverse eff ects, perhaps because it aff ects of symptom improvement aft er a single placement of several activated cells but spares normal, resting cells.14 days.15-17 Such sustained therapeutic eff ect is likely resulted Th e broad antiinfl ammatory action of HC-HA/PTX3 from regeneration of corneal nerves and raises the possibility means that amniotic membrane therapy could be an eff ective of reducing the use of concomitant topical medications to treatment for diff erent types of ocular surface infl ammation. minimize the risk of side eff ects. Of course, infl ammation arises from the immune system’s responses to an injury or a noxious insult, and that underlying A PROMISING FUTURE injury or insult will need to be addressed separately. When Today, cryopreserved amniotic membrane comes in many treating an infection, for example, one must consider using diff erent forms to reduce infl ammation and improve ocular antimicrobials besides quelling the infl ammatory response set surface health, including a sutureless, in-offi ce therapeutic off by the infection. Th is is precisely where amniotic membrane device that is available for use in primary eye care. Th e fi eld therapy can be particularly useful in treating ocular surface is continuing to evolve with biologic graft therapies and even disorders—as a complementary adjunctive therapy for infl am- pharmaceutical products. One may envision that the platform mation control and enhanced healing. technology based on HC-HA/PTX3 may help develop new therapeutics in reducing infl ammation to orchestrate regen- A THERAPEUTIC ROLE FOR DRY EYE erative healing. In recent years, researchers have begun to explore the use of amniotic membrane therapy for the treatment of dry eye Sche er CG Tseng, MD, PhD, is the medical director of the Ocular Surface disease (DED), in which infl ammation plays a central patho- Center in Miami, FL and serves as the chief technology o cer of TissueTech, genic role. Th e common, multifactorial disease has many dif- Inc. in Miami, FL. He states that he has received grant/research support from ferent anatomical forms, but all enter the fi nal common vicious National Institutes of Health and the National Eye Institute. Dr. Tseng is also circle of infl ammation leading to ocular surface disruption a stock shareholder for TissueTech, Inc. Medical writer Ying Guo, MBBS, PhD, and clinical signs and symptoms. Various antiinfl ammatory assisted in the preparation of this manuscript. therapies are used to treat DED, including corticosteroids, cyclosporine, and, most recently, lifi tegrast. Still, there is a REFERENCES 1. De Rötth A. Plastic repair of conjunctival defects with fetal membranes. Arch refractory population that does not respond well to conven- Ophthalmol. 1940. 23:522-55. tional forms of treatment and needs new therapeutic regimes. 2. Sorsby A, Symons HM. Amniotic membrane grafts in caustic burns of the Recent studies provide increasing evidence supporting the eye. Br J Ophthalmol. 1946(30):337-45. 3. Sorsby A, Haythorne J, Reed H. Further experience with amniotic membrane use of amniotic membrane as a therapy for moderate-to-severe grafts in caustic burns of the eye. Br J Ophthalmol. 1947(31):409-18. DED.15-17 In a retrospective chart review at multiple clinical 4. Kim JC, Tseng SCG. Transplantation of preserved human amniotic mem- sites, amniotic membrane treatment rapidly reduced signs and brane for surface reconstruction in severely damaged rabbit corneas. Cornea. 1995;14:473-84. symptoms of DED and accelerated corneal healing in patients 5. Tseng SC, Espana EM, Kawakita T, et al. How does amniotic membrane with refractory DED.17 Th e favorable outcomes are attributed work? Ocul Surf. 2004;2(3):177-87. to amniotic membrane’s potent infl ammatory eff ect, but the 6. Li W, He H, Kawakita T, Espana EM, Tseng SC. Amniotic membrane induces apoptosis of interferon-gamma activated macrophages in vitro. Exp treatment’s benefi ts for DED patients likely go beyond that. Eye Res. 2006;82(2):282-92. In the course of chronic DED, infl ammation of the ocular 7. Tseng SC. HC-HA/PTX3 purifi ed from amniotic membrane as novel surface can cause corneal nerve endings to gradually degener- regenerative matrix: insight into relationship between infl ammation and regeneration. . 2016;57(5):ORSFh1-8. ate and, eventually, disappear.18 Corneal nerves, by mediating Invest Ophthalmol Vis Sci blink and tearing refl exes, play an important role in maintain- ing the healthy state of the corneal epithelium and a stable tear fi lm. Injury or loss of these sensory nerves will further ag- TSENG REFERENCES continue on page 9

To obtain CME credit for this activity, go to http://cme.ufl .edu/ed/self-study/toai/ Topics in OCULAR ANTIINFLAMMATORIES 7 EXAMINATION QUESTIONS TOPICS IN OCULAR ANTIINFLAMMATORIES | ISSUE 23 This CME activity is sponsored by the University of Florida College of Medicine and is supported by an unrestricted educational grant from Shire. Participants must score at least 80% on this exam in order to receive credit. The University of Florida College of Medicine designates this enduring material for a maximum of 1 AMA PRA Category 1 Credit™. To take this exam and obtain credit, please take the test online at http://cme.ufl.edu/ed/self-study/toai/. Expires: August 31, 2019.

1. Which of the following 4. Which of the following 8. Which of the following is an inflammatory mediators have been statements is correct about HC-HA/ advantage of amniotic membrane found at elevated levels in aqueous PTX3? compared with conventional topical or vitreous samples from diabetic A. It retains the antiinflammatory antiinflammatory therapies? eyes? actions of amniotic membrane A. Broad antiinflammatory actions A. IL-6 B. It induces apoptosis of B. Long duration of action B. IFN-γ both activated and resting C. Few detrimental effects on the C. ICAM-1 inflammatory cells ocular surface D. All of the above C. It regulates innate but not D. All of the above adaptive immune responses 2. The fluocinolone acetonide D. Its synthetic form is under 9. Post-hoc analysis of the intravitreal implant: investigation in clinical trials DRCR.net Protocol I data found A. Delivers a bolus of drug that which of the following? tapers over 12 months 5. In addition to inflammation A. Response to the first B. Delivers a continuous dose of reduction, what therapeutic effect ranibizumab injection predicted FAc over 36 months does amniotic membrane exert in outcomes at 5 years the treatment of DED? C. Is a common first-line B. Response after the first three treatment for center-involving A. Increased tear production ranibizumab injections predicted DME B. Improved corneal sensitivity outcomes at 3 years D. Is a common first-line C. Accelerated epithelial healing C. There was no clear association treatment for proliferative DR D. Both B and C between early and long-term response to ranibizumab 3. What is the retinal ischemia 6. Amniotic membrane has D. Ranibizumab was superior to index? been found to suppress the sham injection at 3 months A. The number of clock-hours proinflammatory responses of: affected by nonperfusion A. Macrophages 10. Which of the following is B. The ratio of perfused to B. Fibroblasts NOT an initial FDA-approved claim for cryopreserved amniotic nonperfused retina C. Vascular endothelial cells membrane in ocular surface C. The area of nonperfused retina D. Epithelial cells reconstruction? over the total retinal area A. Antiinflammatory effect D. The percentage of the retinal 7. Which of the following SD- B. Antiscarring effect periphery that is nonperfused OCT features may be associated C. Antiangiogenic effect with DME severity? D. Analgesic effect A. Subfoveal neuroretinal detachments B. Epiretinal membrane C. Hyperreflective spots D. Both A and C

8 Topics in OCULAR ANTIINFLAMMATORIES To obtain CME credit for this activity, go to http://cme.ufl.edu/ed/self-study/toai/ KIERNAN REFERENCES from page 4 1. Virgili G, Parravano M, Evans JR, Gordon I, Lucenteforte E. Anti-vascular 15. Campochiaro PA, Brown DM, Pearson A, et al. Sustained delivery fluocino- endothelial growth factor for diabetic macular oedema: a network meta- lone acetonide vitreous inserts provide benefit for at least 3 years in patients analysis. Cochrane Database Syst Rev. 2017 Jun 22;6:CD007419. with diabetic macular edema. Ophthalmology. 2012;119(10):2125-32. 2. Semeraro F, Cancarini A, dell’Omo R, Rezzola S, Romano MR, Costagliola 16. Markoulli M, Flanagan J, Tummanapalli SS, Wu J, Wilcox M. The impact C. Diabetic retinopathy: vascular and inflammatory disease.J Diabetes Res. of diabetes on corneal nerve morphology and ocular surface integrity. Ocul 2015;2015:582060. doi: 10.1155/2015/582060. Epub 2015 Jun 7. Surf. 2018;16:45-57. 3. Brown DM, Nguyen QD, Marcus DM, et al. Long-term outcomes of ranibi- 17. Agarwal A, Hassan M, Sepah YJ, Do DV, Nguyen QD. Subcutaneous zumab therapy for diabetic macular edema: the 36-month results from two repository corticotropin gel for non-infectious panuveitis: Reappraisal of an phase III trials: RISE and RIDE. Ophthalmology. 2013;120(10):2013-22. old pharmacologic agent. Am J Ophthalmol Case Reports. 2016;4:78-82. 4. Gonzalez VH, Campbell J, Holekamp NM, et al. Early and long-term 18. Wright AD, Sharma P, Zweifel B, Oh L. Suppression of acute uveitis follow- responses to anti–vascular endothelial growth factor therapy in diabetic ing treatment with repository corticotropin injection. J Immunol. 2017;198(1) macular edema: analysis of Protocol I data. Am J Ophthalmol. 2016;172:72-9. (Suppl):127.9. 5. Brown DM, Schmidt-Erfurth U, Do DV, et al. Intravitreal aflibercept for 19. Clearside Biomedical. Suprachoroidal CLS-TA with intravitreal aflibercept diabetic macular edema 100-week results from the VISTA and VIVID versus aflibercept alone in subject with diabetic macular edema (TYBEE). studies. Ophthalmology. 2015;122(10)2044-52. ClinicalTrials.gov Identifier: NCT03126786. https://clinicaltrials.gov/ct2/ 6. Vujosevic S, Simo R. Local and systemic inflammatory biomarkers of show/NCT03126786. Accessed May 30, 2018. diabetic retinopathy: an integrative approach. Invest Ophthalmol Vis Sci. 20. Dugel P, et al. A Comparison of the anatomical efficacy of brolucizumab and 2017;58:BIO68-75. aflibercept in neovascular age-related macular degeneration (nAMD): an 7. Elman MJ, Aiello LP, Beck RW, et al. Randomized trial evaluating ranibi- analysis over 16 weeks of matched treatment in the HAWK and HARRIER zumab plus prompt or deferred laser or triamcinolone plus prompt laser for studies. Presented at the Association for Research in Vision and Ophthalmol- diabetic macular edema. Ophthalmology. 2010;117(6):1064-77. ogy Annual Meeting. Honolulu, HI; April 29-May 3, 2018. 8. Bressler SB, Glassman A, Almukhtar T, et al. Five-year outcomes of 21. Novartis Pharmaceuticals. Study of Efficacy and Safety of Brolucizumab vs. ranibizumab with prompt or deferred laser versus laser or triamcinolone Aflibercept in Patients with Visual Impairment Due to Diabetic Macular plus deferred ranibizumab for diabetic macular edema. Am J Ophthalmol. Edema (KESTREL). ClinicalTrials.gov Identifier: NCT03481634. https:// 2016;164:57-68. clinicaltrials.gov/ct2/show/NCT03481634. Accessed May 30, 2018. 9. Patel RD, Messner LV, Teitelbaum B, Michel KA, Hariprasad SM. Char- 22. Regeneron Pharmaceuticals. Anti-vasculaR Endothelial Growth Factor acterization if ischemic index using ultra-widefield fluorescein angiography plUs Anti-angiopoietin 2 in Fixed comBination therapY: Evaluation for the in patients with focal and diffuse recalcitrant diabetic macular edema.Am J Treatment of Diabetic Macular Edema (RUBY). ClinicalTrials.gov Identi- Ophthalmol. 2013;155(6):1038-44. fier: NCT02712008. https://clinicaltrials.gov/ct2/show/NCT02712008. 10. DRCR.net. Peripheral diabetic retinopathy (DR) lesions on ultra-widefield Accessed May 30, 2018. fundus images and risk of DR worsening over time. http://drcrnet.jaeb.org/ 23. Hoffman-La Roche. A Study of RO6867461 in Participants with Center-in- Studies.aspx?RecID=239. Accessed May 31, 2018. volving Diabetic Macular Edema (BOULEVARD). ClinicalTrials.gov Iden- 11. Wykoff CC, Chakravarthy U, Campochiaro PA, Bailey C, Green K, Cunha- tifier: NCT02699450. https://clinicaltrials.gov/ct2/show/NCT02699450. Vaz J. Long-term effects of intravitreal 0.19 mg fluocinolone acetonide -im Accessed May 30, 2018. plant on progression and regression of diabetic retinopathy. Ophthalmology. 24. Maturi RK, Glassman AR, Liu D, et al. Effect of adding dexamethasone 2017;124;4:440-9. to continued ranibizumab treatment in patients with persistent diabetic 12. Iglicki M, Zur D, Busch C, Okada M, Loewenstein A. Progression of diabetic macular edema: A DRCR Network phase 2 randomized clinical trial. JAMA retinopathy severity after treatment with dexamethasone implant: a 24-month Ophthalmol. 2018;136(1):29-38. cohort study. Acta Diabetologica. 2018;55:626-35. 25. Krick TW, Bressler NM. Recent clinically relevant highlights from the 13. Regillo CD, Callanan DG, Do DV, et al. Use of corticosteroids in the treat- Diabetic Retinopathy Clinical Research Network. Curr Opin Ophthalmol. ment of patients with diabetic macular edema who have a suboptimal response 2018 May;29(3):199-205. to anti-VEGF: recommendations of an expert panel. Ophthalmic Surg Lasers 26. Mehta H, Hennings C, Gillies MC, Nguyen V, Campain A, Fraser-Bell Imaging Retina. 2017;48:291-301. S. Anti-vascular endothelial growth factor combined with intravitreal 14. Boyer DS, Yoon YH, Belfort R Jr., et al. Three-year, randomized, sham- steroids for diabetic macular oedema. Cochrane Database Syst Rev. 2018 Apr controlled trial of dexamethasone intravitreal implant in patients with diabetic 18;4:CD011599. doi:10.1002/14651858.CD011599.pub2. macular edema. Ophthalmology. 2014;121(10):1904-14.

TSENG REFERENCES continued from page 7 8. He H, Li W, Chen SY, Zhang S, et al. Suppres- matrix. J Cell Physiol. 1999;179(3):325-35. nerve structure and function in patients with sion of activation and induction of apoptosis in 3. Cooke M, Tan EK, Mandrycky C, He H, non-sjogren dry eye: clinical correlations. Invest RAW264.7 cells by amniotic membrane extract. O’Connell J, Tseng SC. Comparison of cryopre- Ophthalmol Vis Sci. 2013;54(8):5144-50. Invest Ophthalmol Vis Sci. 2008;49(10):4468-75. served amniotic membrane and umbilical cord tis- 19. Touhami A, Grueterich M, Tseng SC. The role 9. He H, Li W, Tseng DY, et al. Biochemical sue with dehydrated amniotic membrane/chori- of NGF signaling in human limbal epithelium characterization and function of complexes on tissue. J Wound Care. 2014;23(10):465-74,476. expanded by amniotic membrane culture. Invest formed by hyaluronan and the heavy chains of 14. He H, Zhang S, Tighe S, Son J, Tseng SC. Ophthalmol Vis Sci. 2002;43(4):987-94. inter-alpha-inhibitor (HC*HA) purified from Immobilized Heavy Chain-Hyaluronic Acid 20. Morkin MI, Hamrah P. Efficacy of self-retained extracts of human amniotic membrane. J Biol Polarizes Lipopolysaccharide-activated Mac- cryopreserved amniotic membrane for treat- Chem. 2009;284(30):20136-46. rophages toward M2 Phenotype. J Biol Chem. ment of neuropathic corneal pain. Ocul Surf. 10. Li W, He H, Chen YT, Hayashida Y, Tseng 2013;288(36):25792-803. 2018;16(1):132-8. SC. Reversal of myofibroblasts by amni- 15. John T, Tighe S, Sheha H, et al. Corneal nerve 21. Namavari A, Chaudhary S, Chang JH, et al. Cy- otic membrane stromal extract. J Cell Physiol. regeneration after self-retained cryopreserved closporine immunomodulation retards regener- 2008;215(3):657-64. amniotic membrane in dry eye disease. J Oph- ation of surgically transected corneal nerves. In- 11. Lee SB, Li DQ, Tan DT, Meller DC, Tseng thalmol. 2017;2017:6404918. vest Ophthalmol Vis Sci. 2012;53:732-40. SC. Suppression of TGF-beta signaling in both 16. Cheng AM, Zhao D, Chen R, et al. Accelerated 22. Lee HK, Ryu IH, Seo KY, Hong S, Kim HC, normal conjunctival fibroblasts and pterygial restoration of ocular surface health in dry eye Kim EK. Topical 0.1% prednisolone lowers body fibroblasts by amniotic membrane.Curr disease by self-retained cryopreserved amniotic nerve growth factor expression in kerato- Eye Res. 2000;20(4):325-34. membrane. Ocul Surf. 2016;14(1):56-63. conjunctivitis sicca patients. Ophthalmology. 12. Tseng, SC, Li DQ, Ma X. Suppression of 17. McDonald MB, Sheha H, Tighe S, et al. 2006;113(2):198-205. transforming growth factor-beta isoforms, Treatment outcomes in the DRy Eye Amniotic 23. Gaynes BI, Onyekwuluje A. Topical ophthalmic TGF-beta receptor type II, and myofibroblast Membrane (DREAM) study. Clin Ophthalmol. NSAIDs: a discussion with focus on nepaf- differentiation in cultured human corneal 2018;12:677-81. enac ophthalmic suspension. Clin Ophthalmol. and limbal fibroblasts by amniotic membrane 18. Labbé A, Liang Q, Wang Z, et al. Corneal 2008;2(2):355-68.

To obtain CME credit for this activity, go to http://cme.ufl.edu/ed/self-study/toai/ Topics in OCULAR ANTIINFLAMMATORIES 9