PEER REVIEWED

Newer Molecular Targets For the Management of DME

A number of candidates could lead to new treatments.

ASHISH SHARMA, MD · BARUCH D. KUPPERMANN, MD, PhD

Multiple treatment approaches have evolved for the management of diabetic macular edema. While anti-VEGF therapy seems to be promising,¹ according to leading researchers, a “VEGF +” approach might be more effective, particularly in the case of DME, compared to the VEGF-only approach, which is more relevant for the management of AMD.

In keeping with this concept, researchers have been searching for new molecular targets other than VEGF to make DME therapy more effective. This article provides detailed information regarding newer molecular targets that have appeared promising in early clinical trials.

RAF KINASE (ANTISENSE TECHNOLOGY)

RAF proto-oncogene serine/threonine-protein kinase is also known as proto-oncogene c-RAF or c-Raf. We know that c-Raf is a principal component of the first described mitogen-activated protein kinase (MAPK) pathway: ERK1/2 signaling.²

c-Raf acts as a MAP3 kinase, initiating the entire kinase cascade. The hypothesis is that numerous growth factors (VEGF, Bfgf, IGF-1, EPO, HGF and integrin) signal through c-Raf.

As single-stranded, synthetically prepared strands of deoxynucleotide sequences, usually between 18 and 21 nucleotides in length, antisense oligonucleotides (As-ODNs) complement the mRNA sequence of the target gene.

As-ODNs bind cognate mRNA sequences selectively by sequence-specific hybridization, leading to cleavage or disabling of the mRNA and, thus, inhibition of expression of the target gene (Figure 1).

Figure 1. Mechanism of action of intercellular adhesion molecule 1 (ICAM-1) antisense oligonucleotides. Before entering cells, the ICAM-1 antisense oligonucleotides may be damaged by nucleases. Different chemical modifications (eg, the attachment of phosphorothiate groups) prevent damage by nucleases. The negative charges on the antisense oligonucleotides facilitate their binding to the positively charged cell membranes. In the cytoplasm, ICAM-1 antisense oligonucleotides bind to the specific sequence on ICAM-1 messenger RNA (mRNA) molecules. The formation of oligonucleotide–mRNA duplexes activates the ribonuclease H (RNase H) enzyme, which cuts ICAM-1 mRNA, preventing the synthesis of ICAM-1 protein.

CREDIT: CAMBRIDGE UNIVERSITY PRESS

iCo-007 (iCo Therapeutics Inc., Vancouver, Canada) is a second-generation antisense drug that targets c-Raf kinase to treat diabetic eye disease. It is more potent, more stable, and less inflammatory than the first-generation molecule fomivirsen (Vitravene, Isis Pharmaceuticals, Carlsbad, CA).

iCo-007 inhibits c-Raf expression and blocks MAP kinase signaling. It has a favorable ocular pharmacokinetic profile, and a demonstrated half-life of six to eight weeks in rabbits and monkeys after intravitreal injection.

Evidence From Clinical Trials of iCo-007

iCo has completed a phase 1, open label dose-escalation study in 15 patients with diffuse DME and a six-month follow-up after a single intravitreal injection of iCo-007 (doses ranging between 110 µg and 1,000 µg).

The study included patients with diffuse DME within 300 µm of the foveal center, OCT at baseline >250 µm, and BCVA at baseline of 60 ±15 ETDRS letters (20/63 to 20/500 Snellen). The investigators then divided the patients into four cohorts (a total of 15 patients, six in the last cohort).

The study results suggested no drug-related adverse effects, such as ocular inflammation, increased IOP, and systemic exposure, even at the highest doses. Pharmacokinetic test results indicated that iCo-007 was below the detectable level of 2.00 ng/mL in the blood plasma.

As a secondary endpoint, mean change in reduction of excess retinal thickness compared to baseline was 40%, with a 69% improvement in visual acuity compared to baseline at the end of week 24.³

In 2011, iCo announced a physician-sponsored randomized, multicenter, phase 2 study of iCo-007 as monotherapy or in combination with ranibizumab (Lucentis, Genentech, South San Francisco, CA) or laser for DME involving the foveal center (the iDEAL Study).

The study is under way across 26 sites throughout the United States. The iDEAL study will follow patients for 12 months. The investigators are randomizing the patients into one of the following therapies⁴:

•   iCo-007 monotherapy (350 µg);
•   iCo-007 monotherapy (700 µg);
•   iCo-007 (350 µg) and laser photocoagulation; or
•   iCo-007 (350 µg) and ranibizumab (0.5 mg).

The primary endpoint of the iDEAL trial is a change in VA from baseline to month 8. Secondary endpoints include VA at month 12, retinal thickness as measured by optical coherence tomography at months 8 and 12, duration of the effect of iCo-007 at month 12, and safety.

The phase 1 study revealed no drug-related side effects and possible signs of anatomical changes (reduction of macular edema) in some patients at the end of the follow-up period. The drug persisted for long period of time in some patients, with an increase in OCT thickness noted soon after the injection in most patients.

In January 2013, iCo announced midway through the iDEAL study that it had found no serious drug-related adverse events in subjects receiving repeated doses of iCo-007.

The company also announced that with experience in 149 patients so far, if the study continues on track, the company will be able to report on primary endpoint data for all iDEAL study patients by the fourth quarter of 2013.

THE RTP801 GENE

Ischemia, hypoxia, and oxidative stress can induce the expression of the RTP801 gene. Quark (Fremont, CA) discovered the RTP801 gene using the BiFAR target discovery platform, which identifies the critical genes and proteins that, when inhibited, will reverse disease phenotypes. Quark has patented the RTP801 gene and protein sequences, specific antibodies, and gene inhibition for several diseases.⁵

PF-655 (Quark) is a 19-nucleotide methylated double-stranded siRNA that targets the RTP801 gene. Pfizer (New York, NY) has an exclusive worldwide license on PF-655.

PF-655’s mechanism of action differs from VEGF inhibitors. It reduces retinal blood vessel leakage in diabetic mice. The retinas of the RTP801-knockout mice in a retinopathy of prematurity model showed significant decreases in retinal neovascularization and in numbers of apoptotic cells in the inner nuclear layer.⁶

Investigators also observed reduced pathology when evaluating RTP801-knockout mice in several nonocular disease models, including cigarette smoke–induced lung injury and emphysema.⁷ These findings suggest a broader role for RTP801 in the stress-related diseases (Figure 2).

Figure 2. Role of RTP801 as a regulator of multiple inflammatory targets. Example of emphysema protection in RTP801 knockout mice RTP801 inhibits mTOR complex-1 (mTORC1) activity, resulting in downregulation of hypoxia-inducible factor-1 (HIF-1) and vascular endothelial growth factor (VEGF), which are known to inhibit the production of reactive oxygen species (ROS) that can cause apoptosis of epithelial cells of the alveolar wall. RTP801 may also directly activate the production of ROS and alveolar cell apoptosis in response to stress signals, inducing obliteration of the pulmonary alveoli. A detrimental effect in the lungs is also mediated through the induction of NF-κB by RTP801—possibly via inhibition of mTORC1—that leads to inflammation and alveolar cell proteolysis. Emphysema-promoting factors are shown in red, whereas emphysema-protecting factors are represented in green.

CREDIT: NATURE PUBLISHING COMPANY

Quark completed a phase 1 dose-escalation trial of PF-655 in 2010 (in doses up to 3 mg) in AMD patients. The company assessed the drug’s safety and tolerability for patients enrolled over a 24-month period. No drug-related adverse events occurred.

The investigators reported anatomical changes in the retina and choroid 14 days after injection of PF-655, as well as changes in VA. A great majority of the patients had improved or stable vision 14 days after a single injection.

Moreover, some of the patients (18 of the 27 patients enrolled) showed improvement on a VA test (mean improvement of four letters) after a single injection of the siRNA.⁸ While Quark designed the study to test for the safety and not to assess the efficacy of PF-655, these changes were clinically meaningful. They created a platform for a phase 2 study in DME and wet AMD.

Clinical Trials of PF-655

To evaluate the safety and efficacy of three doses (0.4-mg, 1-mg, and 3-mg) of PF-655 in comparison to laser to treat DME, Quark enrolled 184 patients with BCVA between 20/40 and 20/320 in a multicenter, prospective, masked, randomized, active-controlled, phase 2, interventional clinical trial.

The investigators randomly assigned the patients to the three dosage groups or to laser. The main outcome measurement was change in BCVA from baseline. All dose levels showed improvement at 12 months, but only the 3-mg group showed even a trend toward greater improvement in BCVA from baseline than laser (respectively 5.77 vs 2.39 letters; P=0.08; two-sided alpha = 0.10). The investigators terminated the study before the end of month 12 based on their predetermined futility criteria.

Nevertheless, they found PF-655 safe and well-tolerated, with few treatment-related side effects. At 12 months, discontinuation rates in the PF-655 groups had exceeded those in the laser group, and they were inversely correlated with dose levels.⁹

The MATISSE dose-escalation study of PF-0655 is now under way at 51 locations. It is a two-part study. The first part (Stratum I) is an open-label, dose-escalation, safety, tolerability, and pharmacokinetic study, in which investigators will give the drug to all subjects. Stratum I will determine the maximum tolerated dose and any dose-limiting toxicities.

The second part (Stratum II) is a prospective, randomized, multicenter, double-masked, dose-ranging study evaluating the efficacy and safety of PF-655 alone and in combination with ranibizumab versus ranibizumab alone in patients with DME.

Stratum I will enroll a maximum of 24 subjects with low vision, inclusive of possible intermediate doses, in up to four cohorts of three to six subjects. Stratum II will enroll approximately 240 subjects with DME in up to four cohorts of 60 measurable subjects at a ratio of 1:1:1:1.¹⁰

The primary outcome measurements will be to determine the safety and dose-limiting toxicities and pharmacokinetics of a single intravitreal injection of PF-655 in subjects with low vision (Stratum I) at six months postinjection, along with the safety, tolerability, and ability of PF-655 alone and in combination with ranibizumab to improve visual acuity compared to ranibizumab alone in subjects with DME 30 days after their last injections.

The secondary outcome measurements include anatomical changes in retina and retinal nerve fiber layer morphology on fundus photography and SD-OCT.

The phase 1 study in AMD patients revealed no drugrelated side effects, with possible signs of anatomical changes and VA improvement leading to the phase 2 trial for DME. Investigators expect to complete the phase 2 MATISSE study in July 2014.

THE TIE2 RECEPTOR

Good health relies on a healthy endothelial lining of the blood vessels. Damage to this lining can cause leakage of blood proteins and other components, which is known as vascular leakage. This leakage can be local or systemic.

Tie2 is a receptor for the angiopoietin family of growth factors. Angiopoietin-1 (Angl) and angiopoietin-2 (Ang2) are the natural Tie2 agonist and antagonist, respectively. Circulating levels of Ang2 are involved in vascular leakage and pathologic angiogenesis.

Ang2 inhibits Tie2 signaling by binding to the Tie2 receptor, which compromises vascular integrity and promotes vascular leakage and pathologic angiogenesis¹¹ (Figure 3). AKB-9778 (Aerpio Therapeutics Inc., Cincinnati, OH) is a small-molecule, Tie2 activating agent that blocks these effects.

Figure 3. Role of Tie 2 in regulation of angiogenesis. In response to stimuli such as hypoxia, VEGF induces vasculogenesis and endothelial cell proliferation. Ang1–Tie2 interactions mediate vessel maturation and maintain vessel integrity through the recruitment of peri-endothelial cells. Ang2 blocks Ang1–Tie2 signaling, loosening vascular structure and exposing the endothelium to inducers of angiogenesis such as VEGF. In the presence of VEGF, endothelial cells migrate and proliferate to form new capillary sprouts and blood vessels. Ang2 expression in the absence of VEGF stimulation leads to vessel regression and apoptosis

CREDIT: AMERICAN HEART ASSOCIATION

AKB-9778 inhibits the human protein tyrosine phosphatase ß (HPTPß) enzyme, which acts as a negative regulator of the Tie2 receptor. Inhibiting this negative regulator restores Tie2 signaling, reversing the effects of Ang2-induced vascular destabilization.¹¹

AKB-9778 has demonstrated its efficacy in a wide range of preclinical testing, including retinopathy and macular edema. As a result, Aerpio plans to develop AKB-9778 for DME initially, with further indications hoped for based on demonstrated efficacy.¹¹

AKB-9778 in Clinical Trials

Aerpio enrolled 48 healthy volunteers in a phase 1 safety and efficacy study of single ascending doses of AKB-9778. The subjects tolerated the drug well, leading to a phase 1b/2a study of AKB-9778 in patients with DME.

That trial is currently under way.¹² It will enroll up to 24 patients at six sites throughout the United States.¹³ Investigators will evaluate up to four dose levels of subcutaneous AKB-9778. They will administer doses daily for 28 days.

The primary outcome measurement is ocular and systemic safety. The secondary outcome measurements include pharmacokinetics of the drug and changes in OCT-measured retinal thickness and BCVA.

INTEGRINS

Integrin peptide therapy is a novel modality for the treatment for retinal vascular diseases, including DME. Integrins are cell surface receptors that perform a number of functions, including cell signal transduction, mediation of attachments between cells, and regulation of the cell cycle.

Integrins play an important role by interacting extra-cellularly with important proteins, such as collagen and fibronectin, and by intracellularly regulating cell survival, proliferation, and trafficking. By interacting with specific ligands, the peptide interferes with the angiogenic cascade at multiple points and inhibits cell adhesion (Figure 4).

Figure 4. Model of integrin-stimulated tyrosine phosphorylation and signalling pathways. Integrin receptor engagement by ligands such as fibronectin or vitronectin stimulates focal adhesion kinase (FAK) autophosphorylation at Tyr397, creating a binding site for the SH2 domain of c-Src. Recruitment and activation of Src-family protein tyrosine kinases (PTKs) can lead to enhanced phosphorylation of FAK at other sites, such as Tyr925 (which creates a binding site for the Grb2 adaptor protein). Integrin activation of both FAK and Src-family PTKs can promote Shc tyrosine phosphorylation and Grb2 binding to Shc at Tyr317. Grb2 binding to these signalling complexes can potentiate the translocation of the GDP–GTP exchange protein Sos to the plasma membrane, leading to enhanced GTP exchange on Ras. Activation of the ERK mitogen-activated protein kinase (MAPK) cascade is one target for the actions of GTP-bound Ras. Ras can also activate phosphoinositide 3-kinase (PI 3-kinase), which might provide cell survival signals through the activation of targets such as the Akt protein serine/threonine kinase. PI 3-kinase might also facilitate coupling of Ras to the Raf-1 kinase, leading to enhanced activation of the ERK MAP kinase pathway. Src-family association with FAK also potentiates the association and tyrosine phosphorylation of p130Cas. Crk and Nck adaptor protein binding to p130Cas might lead to enhanced cell migration through the activation of pathways involving the Rac GTPase or the JNK MAP kinase cascade. Additional integrin-stimulated signalling events involve the c-Abl PTK and the ILK protein serine/threonine kinase, which might facilitate gene expression and cell-cycle progression events, respectively.

CREDIT: ELSEVIER

Integrin peptide therapy works in many different ways compared to anti-VEGF. Rather than just targeting VEGF, it inhibits integrin. It is a small oligopeptide, compared to large monoclonal antibody, and it is almost one-thousandth the size of a monoclonal antibody.

ALG-1001 (Allegro Ophthalmics, LLC, San Juan Capistrano, CA) is the first entity for integrin peptide therapy. The peculiar shape, size, and configuration of ALG-1001 allow it to inhibit multiple integrin receptor sites successfully.

Scientists attempted integrin inhibition in the past, but were not successful because molecules could not block all of the sites, and they targeted a single receptor. Apart from multitarget deficiency, frequent administration was required due to its short half-life. ALG-1001 binds to all integrin receptors involved with retinal angiogenesis and has a long-lasting effect.

Clinical Trials of ALG-1001

A phase 1 study of ALG-1001 in humans included 15 subjects with advanced DME to assess safety and efficacy. These patients had BCVA of 20/100 or worse, some had early proliferative DR, and many were refractory to standard of care. After a washout period of 90 days with no anti-VEGF, steroid, or laser treatment, the patients received three intravitreal 2.5-mg injections at monthly intervals as standalone therapy.

Follow-up continued for three months after the last treatment. In these 15 patients, the mean age at enrollment was 62.5 years old, baseline BCVA was 1.0 logMAR (20/200 Snellen equivalent), and baseline central macular thickness on OCT was 519 μm . Four patients had received previous anti-VEGF therapy, and six had received previous laser.

No subjects in the study showed loss of BCVA or an increase in CMT on OCT. No serious or significant adverse events were seen during follow-up. The events that the investigators observed were mostly related to the injection and were minor and transient.

Among serious ocular events, one subject had transient IOP elevation that resolved spontaneously, and two had transient mild intraocular inflammation after injection that resolved quickly with topical steroid treatment. Three subjects were lost to follow-up for reasons not related to the study.

Preliminary Indications of Efficacy

In addition to the safety results, preliminary indications of efficacy were seen in these 15 subjects. Mean BCVA in all subjects improved from 1.0 (20/200) at baseline to 0.81 (20/125) at 60 days (last treatment). At 150 days, mean BCVA remained at 0.75 (20/125). Final follow-up in this group occurred at three months after treatment had ceased.

This modest mean improvement of an average of approximately 2 lines of vision (persisting for 90 days after treatment) does not tell the whole story, however. The investigators considered eight patients who demonstrated at least 3 lines of improvement at day 90 to be responders, and they analyzed these patients separately from a group of seven nonresponders.

The responders improved from a mean of 1.08 (20/200) at baseline to 0.7 (20/100) at 60 days and 0.7 (20/100) at 150 days. Nonresponders started at 0.91 (20/160) and remained at 0.94 (20/160) at 60 days, with slight improvement to 0.76 (20/125) at 150 days.

Regarding anatomic outcomes, the mean central macular thickness of 519 μm at baseline in all subjects decreased to a mean of 387 μm at 150 days. Among responders, the mean baseline of 563 μm decreased to a mean of 307 μm at 150 days, while nonresponders started at 468 μm and ended at 481 μm at 150 days. Notably, for patients considered nonresponders, no significant loss of BCVA or macular thickness occurred.¹⁴

To summarize, clinical evaluation in patients with endstage DME has shown safety and signs of efficacy with several types of evaluations, including VA and central macular thickness on OCT. The effects of the drug appear to last for three months after treatment.

KALLIKREIN

Recent studies have shown that several components of the kallikrein kinin system (KKS), including plasma kallikrein, factor XII, and kininogen, appear in the vitreous of patients with advanced diabetic retinopathy.

Preclinical studies in rodents showed that activation of plasma kallikrein in the vitreous increased retinal vascular permeability, while inhibition reduced diabetes- and hypertension-induced retinal leakage.

These findings indicate that activation of the intraocular plasma kallikrein pathway can cause excessive retinal vascular permeability, which can result in DME. KKS contains two separate and independently regulated serine proteases (plasma kallikrein and tissue kallikrein), which generate bradykinin peptides.

Tissue kallikrein circulates in the retina and ciliary body. Recent studies have indicated that plasma kallikrein inhibitors could reduce retinal vascular permeability (Figure 5, page 23). Intravitreal recombinant plasma kallikrein produces retinal vascular leakage and hemorrhage, while both kinin B1 and B2 receptor agonists induce retinal edema Kallikrein inhibitors and peptide-based B1 receptor antagonists reduce or block retinal vascular permeability in rats.

Figure 5. Schematic diagram of the renin–angiotensin system and kallikrein–kinin system. Angiotensin-converting enzyme is strategically poised to regulate the balance between Ang II and bradykinin. ACE, angiotensin-converting enzyme; Ang, angiotensin.

CREDIT: OXFORD UNIVERSITY PRESS

FOV-2304 (Fovea Pharmaceuticals SA, Paris, France) is a nonpeptide selective B1 receptor antagonist. It consistently blocks retinal vascular permeability, inhibits leukocyte adhesion, and reduces several inflammatory mediators Blockade of the KKS appears to have potential as a target for new DME therapies.¹⁶

Potential for New DME Therapies

The findings of a study by Pouliot et al suggested that B1R plays a role in retinal vascular damage in the early stages of diabetes. Hyperglycemia results in expression of B1R in the retina and subsequent production of reactive oxygen species, which enhance the expression of several proinflammatory mediators (COX-2, IL-1β, ICAM-1, VEGF, and HIF-1α).

With B1R, these mediators could enhance vascular permeability in the diabetic retina, resulting in diabetic retinopathy. LF22-0542, a highly potent antagonist of human B1R, could offer a therapeutic approach in diabetic retinopathy.¹⁷

KalVista Pharmaceuticals (Boston, MA) is developing novel plasma kallikrein inhibitors, and it is hoping to market products for both intravitreal and oral administration.

Intravitreal plasma kallikrein inhibitors could very well be efficacious in improving symptoms of the disease, as well as preserving VA and slowing disease progression, while oral formulations have the potential for use in therapeutic areas beyond the eye.¹⁸

CONCLUSION

Unearthing anti-VEGF for the management of retinal diseases has revolutionized the field of retina. It has led vision researchers to explore the area of new molecular targets to overcome the limitations of anti-VEGF.

Understanding the molecular biology of DME has resulted in the identification of new potential therapeutic targets, which could result in novel therapeutic options. Early clinical trials results are promising, and we are looking forward to the end results with the hope of having effective, long-lasting, affordable treatment options for the management of DME. RP

REFERENCES

1. Diabetic Retinopathy Clinical Research Network. Intravitreal ranibizumab for diabetic macular edema with prompt versus deferred laser treatment: three-year randomized trial results. Ophthalmology. 2012;119:2312-2318.

2. Kyriakis JM, App H, Zhang XF, et al. Raf-1 activates MAP kinase-kinase Nature. 1992;358:417-421.

3. Boyer D. iCo-007 for treatment of diffuse diabetic macular edema: phase 1, dose escalation, open label clinical trial. iCo Therapeutics Web site. Available at: http://www.icotherapeutics.com/_resources/presentations/presentation_david_boyer.pdf. Accessed May 21, 2013.

4. A randomized, multi-center, phase II study of the safety, tolerability and bioactivity of repeated intravitreal injections of iCo-007 as monotherapy or in combination with ranibizumab or laser photocoagulation in the treatment of diabetic macular edema (the iDEAL Study). Clinical Trials.gov Web site. Available at: http://www.clinicaltrials.gov/ct2/show/NCT01565148. Accessed May 21, 2013.

5. Shoshani T, Faerman A, Mett I, et al. Identification of a novel hypoxia-inducible factor 1-responsive gene, RTP801, involved in apoptosis. Mol Cell Biol. 2002;22:2283-2293.

6. Brafman A, Mett I, Shafir M, et al. Inhibition of oxygen-induced retinopathy in RTP801-deficient mice. Invest Ophthalmol Vis Sci. 2004;45:3796-3805.

7. Yoshida T, Mett I, Bhunia AK, et al. Rtp801, a suppressor of mTOR signaling, is an essential mediator of cigarette smoke-induced pulmonary injury and emphysema. Nat Med. 2010;16:767-773.

8. Pipeline. Silence Therapeutics Web site. Available at: http://silence-therapeutics.com/pipeline/. Accessed May 21, 2013.

9. Nguyen QD, Schachar RA, Nduaka CI, et al. Dose-ranging evaluation of intravitreal siRNA PF-04523655 for diabetic macular edema (the DEGAS study). Invest Ophthalmol Vis Sci. 2012;53:7666-7674.

10. PF-04523655 dose escalation study, and evaluation of PF-04523655 with/without ranibizumab in diabetic macular edema (DME) (MATISSE).

ClinicalTrials.gov Web site. Available at: http://clinicaltrial.gov/ct2/show/NCT01445899. Accessed May 21, 2013.

11. Tie-2 activators. Aerpio Web site. Available at: http://www.aerpio.com/pipeline/aerpio-9778.html. Accessed May 21, 2013.

12. Aerpio announces positive phase 1 data on first-in-class Tie2 activator, AKB-9778, in development for diabetic macular edema [press release]. Business Wire Web site. Available at: http://www.businesswire.com/news/home/20120320005096/en/Aerpio-Announces-Positive-Phase-1-Data-First-in-class. Accessed May 21, 2013.

13. Safety and pilot efficacy of AKB-9778 in subjects with diabetic macular edema. ClinicalTrials.gov Web site. Available at: http://clinicaltrial.gov/ct2/show/NCT01702441. Accessed May 21, 2013.

14. Kuppermann BD. Integrin peptide therapy for the treatment of vascular eye diseases. Retin Today. 2013(2):60-62.

15. Feener EP. Plasma kallikrein and diabetic macular edema. Curr Diab Rep. 2010;10:270-275.

16. Pruneau D, Bélichard P, Sahel JA, et al. Targeting the kallikrein-kinin system as a new therapeutic approach to diabetic retinopathy. Curr Opin Investig Drugs. 2010;11:507-514.

17. Pouliot M, Talbot S, Sénécal J, et al. Ocular application of the kinin B1 receptor antagonist LF22-0542 inhibits retinal inflammation and oxidative stressin streptozotocin-diabetic rats. PLoS One. 2012;7:e33864.

18. KalVista Pharmaceuticals wins £2.4m biomedical catalyst grant to further develop oral plasma kallikrein inhibitors as a treatment for diabetic macular edema [press release]. KalVista Pharmaceuticals Web site. Available at: http://www.kalvista.com/news/37/131/KalVista-Pharmaceuticals-Wins-2-4m-Bio-medical-Catalyst-Grant-to-Further-Develop-Oral-Plasma-Kallikrein-Inhibitors-as-a-Treatment-for-Diabetic-Macular-Edema.html. Accessed May 21, 2013.