Uveitis in Therapy Immunomodulatory• 36 chapter

were observed during the chlorambucil therapy. In summary, shortterm chlorambucil therapy for refractory uveitis in Behçet’s disease was found effective in controlling the disease in two-thirds of patients.

Miserocchi et al.28 reported their experience with chlorambucil in 56 eyes of 28 patients with various uveitides (Behçet’s disease, juvenile idiopathic arthritis, pars planitis, sympathetic ophthalmia, idiopathic uveitis, Crohn’s disease, and human leukocyte antigen (HLA)-B27- associated uveitis) unresponsive to oral corticosteroids and other immunosuppressive therapy. The median duration of treatment was 12 months (range, 4–50 months); the median daily dosage was 8 mg (range, 4–22 mg). Median follow-up period was 46 months (range, 4–166 months). Vision improved or stabilized in 82% of patients, prednisone was successfully discontinued in 68% of patients, and 50% of patients had inactive disease without systemic therapy at the end of follow-up. However, 25% of patients had to discontinue therapy because of side-effects (temporary amenorrhea, unacceptable gastro­ intestinal intolerance, and progressive leucopenia).

Goldstein et al.29 treated 53 patients with intractable sight-threatening uveitis with short-term (average 16 weeks) high-dose chlorambucil (average 20 mg/day). Seventy-seven percent of patients treated were in remission with an average follow-up of 4 years. Forty-seven percent had at least 2 lines of improvement in Snellen visual acuity after treatment, with an average gain of 3.5 lines. By the time the chlorambucil regimen was stopped, 78% of patients had discontinued oral corticosteroids. Adverse effects include secondary amenorrhea (26% of females), nonophthalmic herpes zoster (12%), testicular atrophy, and erectile dysfunction. None of the patients had developed any malignancy as of their last follow-up.

Efficacy and comparison with other agents

O’Duffy et al.30 assessed the results of various treatments among 21 patients with Behçet’s disease. Treatment with chlorambucil, 0.1 mg/ kg daily, compared favorably with corticosteroids. Uveitis and visual acuities improved in 5 of 7 eyes when the patients were treated with chlorambucil, whereas improvement was observed in only 4 of 13 eyes when treatment consisted of corticosteroids. Toxicity from chlorambucil included leukopenia (2 patients), thrombocytopenia (1), bronchopneumonia (1), and amenorrhea (1). The duration of chlorambucil therapy averaged 1.8 years.

Drug mechanism, systemic and ocular complications and toxicity, drug interactions, and contraindications are summarized in Table 36.3.2,6,8

Summary and key points

Alkylating agents, in view of their potential toxicity, can be a reasonable choice for patients with ocular inflammatory disease resistant to other forms of second-line immunosuppressive agents. The common properties of alkylating agents are summarized in Table 36.3. Chlorambucil therapy can be associated with unpredictable and sudden pancytopenia, and for that reason many clinicians prefer to use cyclophosphamide. However, cyclophosphamide can be associated with hemorrhagic cystitis and an increased risk of bladder cancer but is most effective in Wegener granulomatosis and mucous membrane pemphigoid.

SUMMARY

Comparison outcome studies have shown unequivocally that immunomodulatory therapy, such as the calcineurin inhibitors, the antimetabolites, and the alkylating agents, should have a much more important role to play today in the care of patients with ocular inflammation. Clinician scientists caring for patients with uveitis and ocular inflammatory diseases should aim toward the goal of no tolerance to any degree of inflammation. When employed properly and monitored appropriately, the potential adverse events of immunomodulatory therapy are manageable and may be acceptable when benefits to patients can be demonstrated.

REFERENCES

1.Kapturczak MH, Meier-Kriesche HU, Kaplan B. Pharmacology of calcineurin antagonists. Transplantation Proc 2004;36(2 Suppl):25S-32S.

2.Kurup SK, Chan CC. Immunotherapeutic approaches in ocular inflammatory diseases. Arch Immunol Ther Exp 2005;53:484–496.

3.Ozdal PC, Ortac S, Taskintuna I, et al. Long-term therapy with low dose cyclosporin A in ocular Behçet’s disease. Doc Ophthalmol 2002;105:301– 312.

4.Kilmartin DJ, Forrester JV, Dick AD. Cyclosporin A therapy in refractory non-infectious childhood uveitis. Br J Ophthalmol 1998;82:737–742.

5.Murphy CC, Greiner K, Plskova J, et al. Cyclosporine vs tacrolimus therapy for posterior and intermediate uveitis. Arch Ophthalmol 2005;123:634– 641.

6.Jabs DA, Rosenbaum JT, Foster CS, et al. Guidelines for the use of immunosuppressive drugs in patients with ocular inflammatory disorders: recommendations of an expert panel. Am J Ophthalmol 2000;130:492–513.

7.Gerber DA, Bonham CA, Thomson AW. Immunosuppressive agents: recent developments in molecular action and clinical application. Transplant Proc 1998;30:1573–1579.

8.Becker MD, Smith JR, Max R, et al. Management of sight-threatening uveitis: new therapeutic options. Drugs 2005;65:497–519.

9.Sloper CM, Powell RJ, Dua HS. Tacrolimus (FK506) in the treatment of posterior uveitis refractory to cyclosporine. Ophthalmology 1999;106:723–

728.

10.Hogan AC, McAvoy CE, Dick AD, et al. Long-term efficacy and tolerance of tacrolimus for the treatment of uveitis. Ophthalmology 2007;114:1000– 1006.

11.Thorne JE, Jabs DA, Qazi FA, et al. Mycophenolate mofetil therapy for inflammatory eye disease. Ophthalmology 2005;112:1472–1477.

12.Siepmann K, Huber M, Stübiger N, et al. Mycophenolate mofetil is a highly effective and safe immunosuppressive agent for the treatment of uveitis: a retrospective analysis of 106 patients. Graefes Arch Clin Exp Ophthalmol 2006;244:788–794.

13.Sobrin L, Christen W, Foster CS. Mycophenolate mofetil after methotrexate failure or intolerance in the treatment of scleritis and uveitis. Ophthalmology 2008;115:1416–1421.

14.Doycheva D, Deuter C, Stuebiger N, et al. Mycophenolate mofetil in the treatment of uveitis in children. Br J Ophthalmol 2007;91:180–184.

15.Samson CM, Waheed N, Baltatzis S, et al. Methotrexate therapy for chronic noninfectious uveitis: analysis of a case series of 160 patients. Ophthalmology 2001;108:1134–1139.

16.Heiligenhaus A, Mingels A, Heinz C, et al. Methotrexate for uveitis associated with juvenile idiopathic arthritis: value and requirement for additional anti-inflammatory medication. Eur J Ophthalmol 2007;17:743– 748.

17.Hardwig PW, Pulido JS, Erie JC, et al. Intraocular methotrexate in ocular diseases other than primary central nervous system lymphoma. Am J Ophthalmol 2006;142:883–885.

18.Taylor SR, Habot-Wilner Z, Pacheco P, et al. Intraocular methotrexate in the treatment of uveitis and uveitic cystoid macular edema. Ophthalmology 2009;116:797–801.

19.Yazici H, Pazarli H, Barnes CG, et al. A controlled trial of azathioprine in Behçet’s syndrome. N Engl J Med 1990;322:281–285.

20.Vianna RN, Ozdal PC, Deschênes J, et al. Combination of azathioprine and corticosteroids in the treatment of serpiginous choroiditis. Can J Ophthalmol 2006;41:183–189.

21.Kim SJ, Yu HG. The use of low-dose azathioprine in patients with Vogt–Koyanagi–Harada disease. Ocul Immunol Inflamm 2007;15:381–387.

22.Schatz CS, Uzel JL, Leininger L, et al. Immunosuppressants used in a steroid-sparing strategy for childhood uveitis. J Pediatr Ophthalmol Strabismus 2007;44:28–34.

23.Akpek EK, Jabs DA, Tessler HH, et al. Successful treatment of serpiginous choroiditis with alkylating agents. Ophthalmology 2002;109:1506–1513.

24.Durrani K, Papaliodis GN, Foster CS. Pulse IV cyclophosphamide in ocular inflammatory disease: efficacy and short-term safety. Ophthalmology 2004;111:960–965.

25.Ozyazgan Y, Yurdakul S, Yazici H, et al. Low dose cyclosporin A versus pulsed cyclophosphamide in Behçet’s syndrome: a single masked trial. Br J Ophthalmol 1992;76:241–243.

26.Tessler HH, Jennings T. High-dose short-term chlorambucil for intractable sympathetic ophthalmia and Behçet’s disease. Br J Ophthalmol 1990;74:353–357.

27.Mudun BA, Ergen A, Ipcioglu SU, et al. Short-term chlorambucil for refractory uveitis in Behçet’s disease. Ocul Immunol Inflamm 2001;9: 219–229.

28.Miserocchi E, Baltatzis S, Ekong A, et al. Efficacy and safety of chlorambucil in intractable noninfectious uveitis: the Massachusetts Eye and Ear Infirmary experience. Ophthalmology 2002;109:137–142.

29.Goldstein DA, Fontanilla FA, Kaul S, et al. Long-term follow-up of patients treated with short-term high-dose chlorambucil for sight-threatening ocular inflammation. Ophthalmology 2002;109:370–377.

30.O’Duffy JD, Robertson DM, Goldstein NP. Chlorambucil in the treatment of uveitis and meningoencephalitis of Behçet’s disease. Am J Med 1984;76: 75–84.

KEY FEATURES

Vascular endothelial growth factor (VEGF) blockers currently used to treat eye diseases have included monoclonal antibodies, antibody fragments, and an aptamer. A different method of achieving VEGF blockade in retinal diseases includes the concept of a cytokine trap. Cytokine traps are being evaluated for the treatment of various diseases that are driven by excessive cytokine levels. Traps consist of two extracellular cytokine receptor domains fused together to form a human immunoglobulin G (IgG).1 VEGF Trap-Eye is a soluble fusion protein which combines ligand-binding elements taken from the extracellular components of VEGF receptors 1 and 2 fused to the Fc portion of IgG1.2 This protein contains all human amino acid sequences, which minimizes the potential for immunogenicity in human patients (Figure 37.1).2

VEGF has two main receptors in normal biological systems: VEGFR1 and VEGFR2. VEGFR2 is responsible for most of the endothelial cellproliferating activity of VEGF,3,4 whereas VEGFR1 mediates other activities of VEGF, such as its chemoattractant properties.4 Both receptors are important for the angiogenic-promoting properties of VEGF.3,4 Due to the more recent discovery of other VEGF family members, the original VEGF molecule is now often referred to as VEGF-A. Although several VEGF family members have now been identified, VEGF-A and placental growth factor (PlGF) are the forms believed to be most involved in angiogenesis.2,5 Other family members that exist in mammals include VEGF-B, VEGF-C, and VEGF-D.3,4

VEGF-A is important for vascular permeability, which is important for normal physiological processes such as wound healing but can potentially increase leakage in pathological disease states, such as neovascular age-related macular degeneration (AMD).3,4 VEGF-A also appears to promote the survival of hematopoietic stem cells; the migration of these cells appears to be promoted via interactions with VEGFR1.4 It has been suggested VEGF-A is a critical factor for the recruitment of monocytes and macrophages in inflammatory neovascularization.6

VEGF-B may be particularly important for cardiovascular development and cardiovascular angiogenesis. Although the role of VEGF-B in angiogenesis is not completely understood, there is evidence that it may be involved in inflammatory angiogenesis in various disease states, and may also modulate endothelial proliferation and vessel growth.7 VEGF-C and -D appear to be involved in the development of the lymphatic system.4

A large body of recent research indicates that PlGF may be another factor important for promotion of pathological neoangiogenesis.8,9 There is specific evidence for the involvement of PlGF in neoangiogenesis in eye pathologies. For example, absence of PlGF prevents pathologic angiogenesis and vascular leakage in animal models of cancer, retinal ischemia, and wound healing.10 In the rabbit cornea and the chick chorioallantoic membrane, PlGF was shown to be as effective as VEGF in inducing angiogenesis.11 In mice with experimental choroidal neovascularization (CNV), PlGF (as well as VEGF) was upregulated during initial CNV development. In PlGF null mice or in wild-type mice treated with an anti-VEGFR-1 antibody, a significant reduction in the incidence and severity of laser-induced CNV was observed.12 PlGF was present in the vitreous and in the neovascular membranes in humans with proliferative diabetic retinopathy (PDR).13,14 It

therefore seems that PlGF is a critical factor involved in pathological neoangiogenesis.

Unlike the currently available monoclonal VEGF antibody and the VEGF antibody fragment, VEGF Trap-Eye binds VEGF-A, VEGF-B, and PlGF.2,5,6,15,16 Figure 37.1 illustrates the characteristics of VEGF TrapEye. VEGFR1 and VEGFR2 normally bind VEGF-A with binding affinity of 10–30 and 100–300 pM kD, respectively. VEGF Trap-Eye has a much higher binding affinity of approximately 0.5 pM kD. The currently available VEGF antibody fragment binds VEGF-A with an affinity of ≤99–179 pM kD,17 which has approximately 10-fold greater affinity than the currently available VEGF monoclonal antibody.15 The higher binding affinity of VEGF Trap-Eye to VEGF-A, as well as binding to VEGF-B and PlGF, may confer a differential antiangiogenic activity relative to currently available anti-VEGF agents.2,5,6,15,16 The clinical development of VEGF Trap-Eye is intended to explore the potential relevance of these pharmacologic differences.

INTRODUCTION AND HISTORY

Early studies demonstrated that treatment with a truncated, soluble VEGFR1 molecule almost completely suppressed VEGF bioactivity and thereby virtually halted corpus luteum angiogenesis in a rat ovulation model.18 A soluble VEGFR chimeric protein exhibited greater VEGF inhibition than inducible VEGF gene ablation in newborn mice. The VEGF receptor chimera almost completely halted VEGF-mediated growth and was lethal in these animals.19 Several other studies have similarly demonstrated that decoy receptors provide an effective means of blocking VEGF and its angiogenic effects.20–22

The current VEGF Trap-Eye evolved from a parental VEGF-Trap.2,5,23,24 In this initial trap, the first three Ig domains of VEGFR1 were fused to the Fc region of human IgG1. Three additional VEGF traps were then engineered based on that initial molecule; these were called VEGF Trap B1 (in which a highly basic 10-amino-acid sequence was removed from the third Ig domain of the parental trap), VEGF Trap B2 (in which the entire first Ig domain from VEGF Trap B1 was removed), and VEGF-Trap R1R2. VEGF-Trap R1R2 was created via the fusion of the second Ig domain of VEGFR1 with the third domain of VEGFR2.2 These modifications enhanced VEGF-A affinity for the R1R2 trap relative to the other traps constructed and improved the pharmacokinetic properties of the VEGF-Trap R1R2. The initial parental VEGF-Trap had very high affinity for VEGF-A and binding affinity for PlGF, but was a strongly positively charged molecule that bound the extracellular matrix in addition to VEGF-A and PlGF. Modifications resulted in a less positively charged molecule that retained high affinity for VEGF-A and VEGF-B as well as PlGF, but did not nonspecifically bind the extracellular matrix.2,16

Different formulations of VEGF-Trap are currently being developed for oncology and ophthalmologic indications. VEGF-Trap for oncology (aflibercept) has been designed to be administered via intravenous infusion. VEGF Trap-Eye is intended to be delivered via intravitreal injection. Therefore, the molecule itself undergoes additional purification steps to avoid irritation of the eye. Therefore, VEGF Trap-Eye is formulated as an iso-osmotic formulation consisting of buffers used at

• 37Trapchapter-VEGF Proteins: Fusion

66

77

Kinase Kinase

kDa 10–30 pM kDa 100–300 pM

Figure 37.1  Vascular endothelial growth factor (VEGF) Trap-Eye (R1R2) combines the second domain of VEGF receptors 1 (seen in blue) and the third domain of VEGF receptor 2 (seen in red) fused to the Fc portion of IgG1. VEGFR1 and VEGFR2 bind VEGF with affinity of 10–30 and 100–300 kD, respectively. Both are normally linked to a kinase, which mediates intracellular signal transduction. VEGF Trap-Eye has a binding affinity of approximately 0.5 pM.

different concentrations as well as a different concentration of buffers than in the oncology formulation.

PHARMACOLOGY

The pharmacokinetics and ocular distribution of 500 g intravitreal VEGF Trap-Eye were studied in rabbits. Maximum vitreous concentrations of free VEGF Trap-Eye were approximately 500 g/ml 0.25–6 hours following the 500-g injection, and the VEGF Trap-Eye vitreous elimination half-life was about 4.5 days. VEGF Trap-Eye was detected in both retina and choroid, from which it was eliminated with a similar half-life. Ten days after the injection, maximal plasma total VEGF TrapEye levels were 1.6 g/ml. At week 4, vitreous free VEGF Trap-Eye was 10-fold greater than levels of excess bound VEGF Trap-Eye. Therefore, ocular VEGF production would likely be completely suppressed for greater than 6 weeks following an intravitreal injection based on these results.25

In early experiments, modifications led to improved binding affinity for VEGF-Trap R1R2 relative to the parental VEGF-Trap, since parental VEGF-Trap had a binding affinity (kD) of about 5 pM, whereas VEGF Trap-R1R2 had a kD of approximately 1 pM.2 Furthermore, its VEGF receptor-binding domains have a high affinity for PlGF – about 45 pM. A large body of recent research indicates that PlGF may be another factor that is important for promoting pathological neoangiogenesis.8,9 Based on initial studies demonstrating high-affinity and improved pharmacokinetics, the development of this molecule has advanced to human clinical trials, including for the treatment of wet AMD and diabetic macular edema (DME). Experimental studies with intravitreal injections indicate that VEGF Trap-Eye should penetrate all layers of the retina (molecular weight 110 000) with minimal systemic exposure.26 In an animal model (diabetic rats), an intravitreal injection of VEGF Trap-Eye was distributed to all retinal layers, with little, if any, systemic exposure.27 The agent also displays an extended half-life, allowing for long-term blockade.5

A recent modeling study predicted that the VEGF-binding activity of VEGF Trap-Eye would far exceed that of the anti-VEGF monoclonal antibody ranibizumab, based on the higher binding affinity and longer

Figure 37.2  Vascular endothelial growth factor (VEGF)-Trap binds VEGF dimers (seen as two blue ovals) as a decoy receptor, in a different manner than the way in which VEGF antibodies bind VEGF dimers. VEGF-Trap binds the VEGF dimer on both sides of the dimer “like two hands on a football.” The trap prevents VEGF from activating its cell surface receptor. In contrast, anti-VEGF antibodies bind in a manner that allows the VEGF dimer to interact further with other molecules.

intravitreal half-life of VEGF Trap-Eye. Approximately 1.15 mg VEGF Trap-Eye is an equimolar concentration to 0.5 mg ranibizumab. This concentration of VEGF Trap-Eye (1.15 mg) would have binding activity at 79 days, comparable to 0.5 mg ranibizumab binding activity at 30 days. According to this study 0.5, 2, and 4 mg VEGF Trap-Eye injected intravitreally would have binding activity at 73, 83, and 87 days respectively, comparable to ranibizumab 0.5 binding activity at 30 days.28 A similar study predicted that binding activity for the monoclonal antibody ranibizumab would be roughly comparable to that of the anti-VEGF monoclonal antibody bevacizumab, implying that VEGF Trap-Eye should have more extended binding activity than either anti-VEGF monoclonal antibody.29 This study hypothesizes that VEGF Trap-Eye should have more durable effects than currently available treatments, which would likely result in longer intervals between doses.

VEGF Trap-Eye theoretically binds VEGF more tightly than native receptors or monoclonal antibodies. This suggests that a much lower dose of VEGF Trap-Eye may be used versus anti-VEGF monoclonal antibodies.2 VEGF TRAP-Eye (R1R2) blocks all VEGF-A isoforms and PlGF.2,5,6 It also blocks all isoforms of VEGF-B.16 Consistent with this comprehensive receptor-blocking effect, VEGF Trap-Eye was shown in vitro to block several biological effects of VEGF, including potent blockade of the activation of VEGFR by VEGF, complete blockade of VEGFR2-induced phosphorylation in cultured human umbilical vein endothelial cells, and blockade of VEGF-induced proliferation of NIH/3T3 cells transfected with a VEGFR2/TrkB chimeric receptor.2

DRUG MECHANISM

The VEGF-Trap forms a stable and inert complex with VEGF.30 Figure 37.2 illustrates differences in how VEGF-Trap binds VEGF dimers as a decoy receptor, in contrast to the manner in which VEGF antibodies bind VEGF dimers. VEGF normally binds its receptors as homoor heterodimers.3 VEGF-Trap binds the VEGF dimer on both sides of the dimer “like two hands on a football.” The trap binds VEGF very tightly and prevents it from activating its cell surface receptor. In contrast, anti-VEGF antibodies bind in a manner that allows the VEGF dimer to interact further with other molecules. Since, in contrast to antibody complexes, the VEGF-Trap forms inert complexes with VEGF,30 this minimizes the potential for VEGF-Trap to interact with more than one VEGF-Trap molecule, for example, thus multimeric complexes will not