Darapladib

Darapladib, a Lipoprotein-Associated Phospholipase A2 Inhibitor, in Diabetic Macular Edema:A 3-Month Placebo-Controlled Study

Giovanni Staurenghi, MD,1 Li Ye, MS,2 Mindy H. Magee, PhD,2 Ronald P. Danis, MD,3
John Wurzelmann, MD, MPH,4 Peter Adamson, PhD,5 Megan M. McLaughlin, MS,2 for the Darapladib DME Study Group*

Purpose: To investigate the potential of lipoprotein-associated phospholipase A2 inhibition as a novel mechanism to reduce edema and improve vision in center-involved diabetic macular edema (DME).
Design: Prospective, multicenter, randomized, double-masked, placebo-controlled phase IIa study.
Participants: Fifty-four center-involved DME patients randomized 2:1 to receive darapladib (n 36) or placebo (n 18).
Methods: Darapladib 160 mg or placebo monotherapy was administered orally once daily for 3 months, and patients were followed up monthly for 4 months.
Main Outcome Measures: Mean change from baseline in best-corrected visual acuity (BCVA) and the center subfield and center point of the study eye at month 3 as determined by spectral-domain optical coherence tomography.
Results: Five patients in the study received intravitreal antievascular endothelial growth factor rescue therapy before the day 90 assessment, 2 of 36 (6%) in the darapladib arm and 3 of 18 (17%) in the placebo arm. Administration of 160 mg darapladib for 3 months resulted in statistically significant mean improvements, from baseline to month 3, in BCVA of 4.1 Early Treatment Diabetic Retinopathy Study (ETDRS) letters (95% confidence interval [CI], 2.3e5.8) and of 57 mm in central subfield thickness (95% CI, 84 to 30) in the study eyes. An increase in BCVA of 1.7 ETDRS letters (95% CI, 1.0 to 4.4) and a decrease in center subfield thickness of 34 mm (95% CI, 75 to 6.8) for the placebo group were not significant. No ocular severe adverse events (SAEs) or SAEs considered related to darapladib were reported. One SAE of myocardial infarction, not considered related to darapladib, was reported, and 1 SAE of severe diarrhea was reported in a placebo patient, subsequently with- drawn from the study. Study eye ocular adverse events (AEs) and nonocular AEs were similar between treatment groups.
Conclusions: Once-daily oral darapladib administered for 3 months demonstrated modest improvements in vision and macular edema that warrant additional investigation of this novel lipoprotein-associated phospholipase A2 inhibitory mechanism for the treatment of DME. Ophthalmology 2015;■:1e7 ª 2015 by the American Academy of Ophthalmology.

Macular edema is a major cause of vision loss in diabetic patients and is a direct consequence of inner blooderetinal barrier breakdown.1,2 This breakdown is likely a result of ischemia and hypoxia-driven secretion of cytokines and growth factors, the best known being vascular endothelial growth factor (VEGF). Sight-threatening, center-involved macular edema may be ameliorated by intravitreal anti-VEGF therapy.3 However, factors in addition to VEGF play a role in the disease process, and significant room for improvement in drug delivery and response to treatment exists.

Oxidative stress may be an important contributor to the vasopermeability and inner blooderetinal barrier dysfunc- tion associated with diabetic retinopathy and diabetic mac- ular edema (DME). It is hypothesized that oxidation of lipids creates free radicals that damage local tissue and initiate an inflammatory cascade. Lipoprotein-associated phospholipase A2 (Lp-PLA2) is a member of the super family of phospholipase A2 enzymes that is involved in hydrolysis of oxidized phospholipids.4 Lipoprotein- associated phospholipase A2 is secreted by cell types that

© 2015 by the American Academy of Ophthalmology Published by Elsevier Inc.

http://dx.doi.org/10.1016/j.ophtha.2014.12.014 1
ISSN 0161-6420/15

Ophthalmology Volume ■, Number ■, Month 2015

play a major role in the systemic inflammatory response to injury, including macrophages, monocytes, lymphocytes, T lymphocytes, and mast cells. Lipoprotein-associated phos- pholipase A2 is associated predominately with oxidized low-density lipoprotein in human circulation.5
Lipoprotein-associated phospholipase A2 uniquely cleaves oxidized, but not unmodified, phospholipids and has been shown to generate significant quantities of nonesteri- fied fatty acids and lysophosphatidylcholine (lyso-PC), both of which are proinflammatory. Lysophosphatidylcholine has been implicated in leukocyte activation, induction of apoptosis, and mediation of endothelial dysfunction.5,6 It is also a known endothelial permeability factor in a number of

inhibitors to explore the role of this enzyme in DME path- ologic features. Importantly, such a therapeutic intervention may yield a method of treating DME. The Lp-PLA2 inhibitor darapladib has been investigated in a phase III study of nearly 16 000 patients with coronary heart disease, approximately one third of whom had diabetes mellitus and received either once-daily darapladib (at a dose of 160 mg) or placebo for a median of 3.7 years.25 This is the first clinical study to investigate the potential of darapladib to reduce edema and improve vision in DME patients.

Methods

vascular beds.7e9 One study has suggested that lyso-PC may

signal indirectly through the VEGF receptor 2 pathway, resulting in VEGF receptor 2edependent, VEGF- independent signaling.9
It is notable that lyso-PC levels are higher in the low- density lipoprotein of diabetic patients than in healthy controls and are correlated with Lp-PLA2 levels.10 Additionally, lyso-PC was higher in patients with pre- proliferative and proliferative diabetic retinopathy, as well as those with nephropathy. Treatment with simvastatin (an 3-hydroxy-3-methylglutaryl-coenzyme A [HMGCoA]- reductase inhibitor), which lowers low-density lipoprotein in humans and thus lowers substrate levels, reduced both circulating Lp-PLA2 and lyso-PC. Statin treatment also is associated with an improvement in DME, hard exudates, and fluorescein leakage.11,12 There are significant associa- tions of serum lipids with the development and progression of diabetic eye disease.13e16
Inhibition of Lp-PLA2 and the production of lyso-PC may be protective against hyperglycemia-related compro- mise of the blooderetinal barrier and bloodebrain barrier. Darapladib is a potent and reversible orally active inhibitor of Lp-PLA2 and reduces the presence of extravascular immunoglobulin G in brain sections of hyperglycemic, hy- percholesterolemic pigs.17,18 Interestingly, darapladib also suppressed monocyte chemotactic protein 1 (MCP-1; 45%) and its receptor (86%) in this pig model.19 Monocyte chemotactic protein 1 is strongly upregulated in diabetic eye disease and is increased in response to lyso-PC.20,21 Additionally, MCP-1 gene-deficient mice are resistant to hyperglycemia-induced retinal vascular permeability and show reduced retinal leukocyte infiltration compared with wild-type mice.21
In addition to MCP-1, other inflammatory mediators such as interleukin 6 have been implicated in DME.20,22 Both interleukin 6 and MCP-1 are impacted significantly with effective DME therapeutics such as triamcinolone.23 Interestingly, Lp-PLA2 inhibitors also have been shown to impact interleukin 6 levels in a number of preclinical models, as has darapladib in the clinic.21,24 Studies using an alternative Lp-PLA2 inhibitor, SB435495, dosed sys- temically in Brown Norway rats demonstrated suppression of retinal vascular albumin leakage secondary to sustained hyperglycemia (Canning P, unpublished data, 2015).
The action of Lp-PLA2 in generating products from oxidized phospholipids, which are potent mediators of retinal and cerebral vascular permeability, allows us to use selective

Study Design and Objectives
This study was a prospective, multicenter, randomized, double- masked, placebo-controlled phase IIa trial assessing the efficacy, safety, tolerability, pharmacokinetics, and pharmaco- dynamics (Lp-PLA2 inhibition) of darapladib monotherapy administered for 3 months to center-involved DME patients. The study was conducted according to the ethical principles of the Declaration of Helsinki and was approved by the appro- priate ethics committees. All study participants provided written informed consent. The clinical trial was registered at Clinical- Trials.gov as NCT01506895.
The study had a primary objective to determine the effect of darapladib administered as once-daily oral doses for 3 months on the coprimary end points of best-corrected visual acuity (BCVA) and spectral-domain (SD) optical coherence tomography (OCT)e determined center subfield retinal thickness in adult patients with center-involved DME. Secondary objectives were to determine the effect of darapladib administered as oral once-daily doses for 3 months on (1) retina anatomic features, (2) safety and tolerability, and (3) pharmacokinetics and pharmacodynamics in adult patients with center-involved DME.

Patients
Patients were treatment naïve or previously treated with a 90-day exclusion period for ranibizumab (Lucentis; Genentech, South San Francisco, CA) or laser, and 180 days for bevacizumab or steroids, were 18 years of age or older, and had DME in the study eye based on investigator-determined fluorescein angiography. Patients were enrolled if they had center subfield thickness of more than 330 mm on Spectralis SD OCT (Heidelberg Engineering, Carlsbad, CA) or more than 310 mm on Cirrus SD OCT; (Zeiss, Dublin, CA) and BCVA between 24 and 78 letters (Snellen equivalent of approximately 20/320e20/32). Patients were excluded if they had evidence of vitreomacular traction, active proliferative diabetic retinopathy, ischemic maculopathy, or choroidal neovascularization; were using systemic anti- angiogenics; had additional eye disease that could compromise BCVA; had uncontrolled diabetes as indicated by hemoglobin A1c of more than 10% at screening; or were women of childbearing potential. At any time during the study, including the follow-up period, standard of care rescue was given based on the clinical judgment of the ophthalmologist. Rescue treatment was to be considered strongly for patients whose center subfield thickness had increased by 60 mm or 15% from baseline or whose BCVA decreased by more than 5 letters compared with baseline and was maintained at 2 consecutive visits. Patients requiring rescue treat- ment continued receiving study medication and were followed up for the remainder of the study.

Randomization and Masking
Eligible patients were stratified at randomization based on baseline BCVA (>50 letters and 50 letters in study eye) and were ran- domized 2:1 to receive darapladib 160 mg or placebo. All study assessments were made by the principal investigator, sub- investigative staff, and clinical study staff, who were masked to
treatment.
Treatments
The treatment schedule is summarized in Figure 1. One hundred sixty milligrams of darapladib, or placebo to match, was administered orally once daily.
Assessments
Efficacy. Spectral-domain optical coherence tomography and BCVA using electronic Early Treatment Diabetic Retinopathy Study (ETDRS) visual acuity were performed in patients at screening, baseline, and monthly during the 3-month treatment period and at follow-up. Fluorescein angiography and color fundus photography were assessed at the screening and month 3 visits. Final SD OCT interpretations, as well as interpretation of fluo- rescein angiography and color fundus photography, were per- formed by a central reading center (University of Wisconsin Madison Fundus Photograph Reading Center, Madison, WI). Best- corrected visual acuity certification was performed by the EMMES Corporation (Rockville, MD).
Safety. Adverse events (AEs), serious AEs (SAEs), concomi- tant medications, clinical chemistry findings, urinalysis results, vital signs, and ophthalmic assessments were studied. Adverse events were coded by preferred term and primary system organ class according to the Medical Dictionary for Regulatory Activities. Pharmacokinetics. Pharmacokinetics were assessed by serial blood samples collected over an 8-hour period at the month 1 or month 2 visits, when darapladib was expected to be at steady state. Human potassium (K2) ethylene diamine tetraacetic acid (EDTA) plasma samples were analyzed for darapladib using a validated analytical method based on protein precipitation, followed by high- pressure liquid chromatography tandem mass spectrometry anal- ysis. The assay was validated over the darapladib concentration range of 0.10 to 50 ng/ml, and the lower limit of quantification was
0.1 ng/ml using a 50-ml aliquot of human plasma. Samples that exceeded the upper end of the validated concentration range were diluted and reassayed.
Quality-control samples were analyzed with each batch of samples against separately prepared calibration standards. Quality- control samples and calibration standards were prepared using

Figure 1. Schematic showing the study design. BL ¼ baseline; FU ¼ follow-up; M ¼ month; S ¼ screening.

independently prepared stock solutions of darapladib reference materials. For the analysis to be acceptable, no more than one third of the quality-control results were to deviate from the nominal concentration by more than 15%, and at least 50% of the results from each quality-control concentration were to be within 15% of nominal. Pharmacokinetic analysis of plasma darapladib concentrationetime data was conducted using noncompartmental Model 200 (for extravascular administration) of WinNonlin Pro- fessional Edition version 5.2 (Pharsight Corporation, Mountain View, CA). Actual elapsed time from dosing was used to estimate all individual plasma pharmacokinetic parameters for evaluable subjects. Given that samples were collected over a duration of 8 hours after dose administration, to estimate area under the plasma concentrationetime curve over the dosing interval of 24 hours, the predose concentration of each subject was assumed to be the concentration 24 hours after dosing. This assumption is reasonable given the samples were collected during a steady state.
Pharmacodynamics (Lipoprotein-Associated Phospholipase A2 Inhibition). Lipoprotein-associated phospholipase A2 activity was measured in plasma using a colorimetric assay, as described previously.26
Pharmacokinetic and Pharmacodynamic Relationship. The relationship between darapladib plasma concentration and plasma Lp-PLA2 activity data was analyzed using nonlinear mixed effect modeling as implemented in the computer program NONMEM version 7 (ICON). The first-order conditional estimation with interaction method was used in the model development process.
Statistical Analyses
For the primary analysis, the change from baseline in BCVA and SD OCT center point at month 3 for the darapladib arm was analyzed by a Bayesian probability model. Normal distributions and mildly informative prior distributions were assumed. A 95% probability that the mean BCVA change at month 3 would be in the range of
0.16 to 15.84 ETDRS letters (0.20 standard deviation) and a 95% probability that the SD OCT change at month 3 would be in the range of 26 to 150 was assumed for the prior distribution for the darapladib arm. As designed with a sample size of 30 completers in the darapladib arm, there would be a more than 80% probability to detect a mean BCVA increase of more than 4 letters at month 3 if the true population mean was more than 5.2 letters and the standard deviation was 7.7 letters and to detect a more than 65-mm change if the true population mean SD OCT decrease at month 3 was more than 92 mm and the standard deviation was 113 mm. The posterior probabilities of mean BCVA and SD OCT changes of more than 4 letters and more than 65 mm at month 3 were computed.
For secondary analysis, the change from baseline in BCVA and
SD OCT center subfield were analyzed separately using a mixed- model repeated measures method, fitting visit, treatment, and visit-by-treatment interaction as fixed effects and baseline data as the covariate. Within each patient, the observations across visits were treated as repeated measures. The variance and covariance for observations across visits within a patient was assumed to be the same for all patients and to have an unspecified structure. The analyses were performed using observed cases; missing assess- ments or assessments after rescue were not included in the analysis data set.

Results

Study Population
A total of 54 patients were enrolled in the darapladib phase IIa study: 36 received 160 mg darapladib daily for 90 days and
18 received placebo. Patient demographics are described in

Ophthalmology Volume ■, Number ■, Month 2015

Table 1. Baseline Characteristics

Characteristics

Placebo (n [ 18)

Darapladib 160 mg Once Daily (n [ 36)

Mean age (yrs) 64.8 63.6
Male gender 12 (67) 25 (69)

White/European heritage

15 (83) 33 (92)

BMI (kg/m2) 27.4 (21e36) 30.2 (19e42)
Diabetes duration (yrs) 23.1 (8e56) 15.3 (1e39) HbA1c (%) 7.63 (5.9e9.3) 7.16 (5.1e9.8) BCVA (ETDRS letters) 66 (46e78) 65 (26e78)

Center subfield thickness (mm)

460 (312e737) 495 (326e805)

Figure 3. Bar graph showing the least squares mean change from baseline

Previous PDR 3 (17) 6 (17)

(BL) in spectral-domain optical coherence tomographyedetermined center

Previous medication for DME

5 (28) 13 (36)

subfield thickness, with standard error for all patients (range, 20/32e20/320 at BL) and the subset with 20/40 to 20/320 vision at BL.

Previous laser 15 (83) 24 (67)

BCVA best-corrected visual acuity; BMI body mass index; DME diabetic macular edema; ETDRS Early Treatment Diabetic Retinopathy Study; HbA1c hemoglobin A1c; PDR proliferative diabetic retinopathy.
Data are no. (%) or mean (range) unless otherwise indicated.

Table 1. There were no deviations from the inclusion and exclusion criteria or major deviations during study conduct that would be anticipated to affect the results of the study. Five patients in the study received intravitreal anti-VEGF rescue therapy before the day 90 assessment: 2 of 36 (6%) in the darapladib arm (1 ranibi-
zumab and 1 bevacizumab) and 3 of 18 (17%) in the placebo arm
(2 ranibizumab and 1 bevacizumab).
Efficacy Outcomes
Bayesian analysis showed a 60% probability of mean BCVA in- crease from baseline of 4 letters or more and an 83% probability of mean SD OCT center point decrease from baseline of 65 mm or more at the day 90 visit for the darapladib arm, given data from the study eye observed cases. Mean changes from baseline at day 90 in the coprimary end points (BCVA and SD OCT central subfield thickness) were an increase of 4.06 ETDRS letters (95% confi- dence interval [CI], 2.31e5.80; n 34) and a decrease of 57 mm (95% CI, 84.1 to 29.9; n 33) for darapladib and an increase
of 1.67 letters (95% CI, 0.99 to 4.34) and a decrease of 34.1 mm
(95% CI, 75 to 6.8) for placebo (n 14; Figs 2 and 3).
A post hoc analysis of the mean change from baseline in the subset of patients with 20/40 to 20/320 vision at baseline resulted

Figure 2. Bar graph showing the least squares mean change from baseline (BL) in best-corrected visual acuity (BCVA), with standard error for all pa-

in an increase of 5.17 ETDRS letters (95% CI, 3.00e7.34; n 23) and a decrease of 53 mm (95% CI, 86 to 20; n 22) for
darapladib and an increase of 0.61 letters (95% CI, 2.64 to 3.68)
and a decrease of 16 mm (95% CI, 64 to 32) for placebo (n 10; Figs 2 and 3). Five subjects in the darapladib arm (15%) and 0 subjects in the placebo arm gained at least 2 lines ( 10 ETDRS letters) of vision (Table 2). Post hoc analysis showed an 18% mean increase from baseline (from 56% to 74%) in darapladib-treated study eyes achieving 20/40 or better vision versus a 7% decline (from 57% to 50%) in the placebo-treated study eyes. Change from baseline at day 90 for the coprimary end points for individual patients receiving darapladib or placebo is shown in Figure 4A, B.

Safety
No deaths were reported. No ocular SAEs or SAEs considered related to darapladib were reported. One SAE of myocardial infarction, not considered related to darapladib, was reported and 1 SAE of severe diarrhea was reported in a patient who received placebo and subsequently was withdrawn from the study. Study eye ocular AEs (Table 3) and nonocular AEs (Table 4) were similar between treatment groups.

Pharmacokinetics
All 36 subjects contributed pharmacokinetic samples for quantifi- cation of darapladib concentrations in plasma. After repeat dose administration, darapladib was quantifiable in all samples at all time points from all subjects. The time course of darapladib plasma concentration in the 36 subjects seemed to have 2 types of profiles: one with a distinct rise in darapladib concentration, signifying an absorption phase followed by distribution and elimination of

Table 2. Patients with Best-Corrected Visual Acuity Change from Baseline by Category

Category (Early Treatment
Diabetic Retinopathy Study Placebo Darapladib 160 mg
Letters) (n [ 14) Once Daily (n [ 34)
Change ≤15 0 (0) 1 (3)
14 ≤ change ≤ 10 0 (0) 4 (12)
9 ≤ change ≤ 5 3 (21) 11 (32)
4 ≤ change ≤ -4 10 (71) 16 (47)
—5 ≤ change 1 (7) 2 (6)

Data are no. (%).

tients (range, 20/32e20/320) and the subset with 20/40 to 20/320 vision at BL.

Figure 4. Scatterplots showing the change from baseline for coprimary end points in patients receiving (A) darapladib or (B) placebo. Note that an investigator-determined value for central subfield thickness is presented for the patient denoted by a square in (A) because the day (D) 90 spectral-domain optical coherence tomography (SD OCT) image was corrupted and could not be transferred to the central reading center. BCVA ¼ best-corrected visual acuity.

darapladib from the plasma (n 29/36), whereas the other type of profile seemed to be relatively flat, with no apparent absorption phase occurring (n ¼ 7/36). Pharmacokinetic parameters in DME patients (n ¼ 36) showed that the geometric mean of the area under

the receiver operating characteristic curve from 0 to 24 hours and maximum concentration observed were 555.6 ng $ h/ml (co- efficient of variation [CV]% 56.9%) and 39.2 ng/ml (CV% 61.9%), respectively.

Ophthalmology Volume ■, Number ■, Month 2015

be a predictor of response to therapy makes a comparison to a trial with a similar average baseline vision important.3 The approximately 4-letter (ETDRS) and 60-mm central subfield thickness improvements observed with 160 mg darapladib at month 3 are intermediate to those observed with laser and ranibizumab, which provide approximately 2.5- and 7-letter (ETDRS) increases and 40- and 125-mm decreases, respec- tively, in a study with similar baseline visual acuity over the same period.3

assessment

AE

mab also was analyzed and did indeed show an increased improvement in mean change from baseline in BCVA (þ5

¼ adverse event.

Pharmacodynamics (Lipoprotein-Associated Phospholipase A2 Inhibition)
Trough inhibition of Lp-PLA2 activity relative to baseline as measured by colorimetric assay was, on average, 63% (CV% ¼ 15%).
Pharmacokinetic and Pharmacodynamic Relationship
The relationship between plasma darapladib concentrations and plasma Lp-PLA2 activity was described best by a direct-effect inhib- itory maximum response relationship. The population mean plasma concentration inhibiting Lp-PLA2 activity by 50% was estimated at
6.76 ng/ml, with interindividual variability of 29.6%. The population mean baseline Lp-PLA2 activity was estimated at 92.2 nmol/minute per milliliter with interindividual variability of 23.8% in DME patients. The Hill coefficient describing the slope of the concentrationeresponse curve was estimated at 0.72, indicating that the concentrationeLp-PLA2 activity response was not very steep.

Discussion

In this clinical report investigating Lp-PLA2 inhibition for diabetic eye disease, darapladib reduced macular edema and improved vision in center-involved DME patients. To interpret these results in the context of other clinical trials, a consideration that baseline visual acuity has been shown to

Table 4. Nonocular Adverse Events

ETDRS letters) versus the total darapladib-treated group ( 4 ETDRS letters).27
Mean improvements were supported by a responder analysis. Beyond measurement variability, a 5-letter (1-line) change in BCVA would be significant for an individual.28 With this in mind, 47% of the darapladib-treated patients had a mean change from baseline of at least 1 line of vision in their study eye versus 21% of placebo-treated subjects. Additionally, there was an increase from baseline in darapladib-treated study eyes achieving driving vision (20/ 40 or better) versus a decline in the placebo study eyes.
Administration of darapladib 160 mg for 3 months was safe and well tolerated. This study demonstrated that oral admin- istration of darapladib modestly reduced edema and improved vision in center-involved DME patients. Lipoprotein- associated phospholipase A2 inhibition is a novel mecha- nism for potential treatment of diabetic eye disease, and it seems to be distinct from that of the anti-VEGF standard of care. This mechanism warrants additional investigation to determine whether it can improve outcomes for DME patients.

References

1. Antonetti DA, Barber AJ, Bronson SK, et al. Diabetic reti- nopathy: seeing beyond glucose-induced microvascular disease. Diabetes 2006;55:2401–11.
2. Curtis TM, Gardiner TA, Stitt AW. Microvascular lesions of diabetic retinopathy: clues towards understanding pathogen- esis? Eye 2009;23:1496–508.
3. Elman MJ, Bressler NM, Qin H, et al; Diabetic Retinopathy Clinical Research (DRCR) Network. Randomized trial evalu- ating ranibizumab plus prompt or deferred laser or triamcin-

AEs

Placebo (n [ 18)

Darapladib 160 mg Once Daily (n [ 36)

olone plus prompt laser for diabetic macular edema. Ophthalmology 2010;117:1064–77.
4. MacPhee CH, Moores KE, Boyd HF, et al. Lipoprotein-

No. of patients with nonocular AE
Nonocular AEs in more than 1 patient

9 (50%) 14 (39%)

associated phospholipase A2, platelet-activating factor ace- tylhydrolase, generates two bioactive products during the oxidation of low-density lipoprotein: use of a novel inhibitor. Biochem J 1999;338:479–87.
5. Wilensky RL, Macphee CH. Lipoprotein-associated phos- pholipase A (2) and atherosclerosis. Curr Opin Lipidol 2009;20:415–20.
6. Lavi S, McConnell JP, Rihal CS, et al. Local production of lipoprotein-associated phospholipase A2 and lysophosphati- dylcholine in the coronary circulation: association with early

AE ¼ adverse event.

coronary atherosclerosis and endothelial dysfunction in humans. Circulation 2007;115:2715–21.

7. Huang F, Subbaiah PV, Holian O, et al. Lysophosphati- dylcholine increases endothelial permeability: role of PKCal- pha and RhoA cross talk. Surgery 2005;138:464–73.
8. Yan S, Chai H, Wang H, et al. Effects of lysophosphati- dylcholine on monolayer cell permeability of human coronary artery endothelial cells. Surgery 2005;289:L176–85.
9. Fujita Y, Yoshizumi M, Izawa Y, et al. Transactivation of fetal liver kinase-1/kinase-insert domain-containing receptor by lysophosphatidylcholine induces vascular endothelial cell proliferation. Endocrinology 2006;147:1377–85.
10. Iwase M, Sonoki K, Sasaki N, et al. Lysophosphatidylcholine contents in plasma LDL in patients with type 2 diabetes mellitus: relation with lipoprotein-associated phospholipase A2 and effects of simvastatin treatment. Atherosclerosis 2008;196:931–6.
11. Gupta A, Gupta V, Thapar S, Bhansali A. Lipid-lowering drug atorvastatin as an adjunct in the management of diabetic macular edema. Am J Ophthalmol 2004;137:675–82.
12. Panagiotoglou TD, Ganotakis ES, Kymionis GD, et al. Ator- vastatin for diabetic macular edema in patients with diabetes mellitus and elevated serum cholesterol. Ophthalmic Surg Lasers Imaging 2010;41:316–22.
13. Chew EY, Klein ML, Ferris FL III, et al. Association of elevated serum lipid levels with retinal hard exudate in diabetic retinopathy. Early Treatment Diabetic Retinopathy Study (ETDRS) report 22. Arch Ophthalmol 1996;114:1079–84.
14. Raman R, Rani PK, Kulothungan V, et al. Influence of serum lipids on clinically significant versus nonclinically significant macular edema: SN-DREAMS report number 13. Ophthal- mology 2010;117:766–72.
15. Benarous R, Sasongko MB, Qureshi S, et al. Differential as- sociation of serum lipids with diabetic retinopathy and diabetic macular edema. Invest Ophthalmol Vis Sci 2011;52:7464–9.
16. Yau JW, Rogers SL, Kawasaki R, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care 2012;35:556–64.
17. Blackie JA, Bloomer JC, Brown MJ, et al. The identification of clinical candidate SB-480848: a potent inhibitor of lipoprotein- associated phospholipase A2. Bioorg Med Chem Lett 2003;13: 1067–70.

Footnotes and Financial Disclosures
Originally received: July 28, 2014.
Final revision: November 20, 2014.
Accepted: December 12, 2014.
Available online: ■■■. Manuscript no. 2014-1179.
1 Department of Biomedical and Clinical Science “Luigi Sacco,” Sacco Hospital, University of Milan, Milan, Italy.
2 GlaxoSmithKline, King of Prussia, Pennsylvania.
3 Department of Ophthalmology and Visual Sciences, University of Wis- consin, Madison, Wisconsin.
4 GlaxoSmithKline, Research Triangle Park, North Carolina.
5 GlaxoSmithKline, Stevenage, United Kingdom.
Presented as a poster at: American Academy of Ophthalmology Annual Meeting, New Orleans, Louisiana, November 2013.
*Darapladib DME Study Group is available online at www.aaojournal.org. Financial Disclosure(s):
The author(s) have made the following disclosure(s): G.S.: Consultant –
GlaxoSmithKline
L.Y.: Employee – GlaxoSmithKline M.H.M.: Employee – GlaxoSmithKline
R.P.D.: Consultant – GlaxoSmithKline, King of Prussia, Pennsylvania
18.
Acharya NK, Levin EC, Clifford PM, et al. Diabetes and hypercholesterolemia increase blood-brain barrier permeability and brain amyloid deposition: beneficial effects of the LpPLA2 inhibitor darapladib. J Alzheimers Dis 2013;35:179–98.
19. Wilensky RL, Shi Y, Mohler ER 3rd, et al. Inhibition of lipoprotein-associated phospholipase A2 reduces complex coronary atherosclerotic plaque development. Nat Med 2008;14:1059–66.
20. Funatsu H, Noma H, Mimura T, et al. Association of vitreous inflammatory factors with diabetic macular edema. Ophthal- mology 2009;116:73–9.
21. Gonçalves I, Edsfeldt A, Ko NY, et al. Evidence supporting a key role of Lp-PLA2-generated lysophosphatidylcholine in human atherosclerotic plaque inflammation. Arterioscler Thromb Vasc Biol 2012;32:1505–12.
22. Yoshimura T, Sonoda KH, Sugahara M, et al. Comprehensive analysis of inflammatory immune mediators in vitreoretinal diseases. PLoS One 2009:e8158.
23. Sohn HJ, Han DH, Kim IT, et al. Changes in aqueous con- centrations of various cytokines after intravitreal triamcinolone versus bevacizumab for diabetic macular edema. Am J Oph- thalmol 2011;152:686–94.
24. Mohler ER, Ballantyne CM, Davidson MH, et al. The effect of darapladib on plasma lipoprotein-associated phospholipase A2 activity and cardiovascular biomarkers in patients with stable coronary heart disease or coronary heart disease risk equiva- lent. JACC J Am Coll Cardiol 2008;51:1632–41.
25. STABILITY Investigators. Darapladib for preventing ischemic events in stable coronary heart disease. N Engl J Med 2014;370:1702–11.
26. Gora S, Lambeau G, Bollinger JG, et al. The proinflammatory mediator platelet activating factor is an effective markers of inflammation, renal function, and hemodynamic stress. Arte- rioscler Thromb Vasc Biol 2006;26:1586–93.
27. Nguyen QD, Brown DM, Marcus DM, et al. Ranibizumab for diabetic macular edema. Ophthalmology 2012;119:789–801.
28. Beck RW, Moke PS, Turpin AH, et al. A computerized method of visual acuity testing: adaptation of the early treat- ment of diabetic retinopathy study testing protocol. Am J Ophthalmol 2003;135:194–205.

J.W.: Employee – GlaxoSmithKline Research Triancle Park, North Carolina P.A.: Employee – GlaxoSmithKline Stevenage, United Kingdom
M.M.M.: Employee – GlaxoSmithKline King of Prussia, Pennsylvania
Supported by GlaxoSmithKline, King of Prussia, Pennsylvania. The sponsor participated in the design of the study; conduct of the study; data collection, management, analysis, and interpretation; and preparation, re- view, and approval of the manuscript.
Abbreviations and Acronyms:
AE ¼ adverse event; BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; DME ¼ diabetic macular edema; ETDRS ¼ Early Treatment Diabetic Retinopathy Study; Lp- PLA2 ¼ lipoprotein-associated phospholipase A2; lyso- PC ¼ lysophosphatidylcholine; MCP-1 ¼ monocyte chemotactic protein; OCT ¼ optical coherence tomography; SAE ¼ severe adverse event; SD ¼ spectral domain; VEGF ¼ vascular endothelial growth factor.
Correspondence:
Giovanni Staurenghi, MD, Department of Biomedical and Clinical Science “Luigi Sacco,” Sacco Hospital, University of Milan, Milan, Italy. E-mail: [email protected].