Pharmacokinetics of Ranibizumab after Intravitreal Administration in Patients with Retinal Vein Occlusion or Diabetic Macular Edema Yi Zhang, PhD,1 Zhenling Yao, PhD,1 Nitin Kaila, PhD,2 Peter Kuebler, PharmD,1 Jennifer Visich, PhD,1 Mauricio Maia, PhD,1 Lisa Tuomi, PharmD,1 Jason S. Ehrlich, MD, PhD,1 Roman G. Rubio, MD,1 Peter A. Campochiaro, MD3 Objective: To describe the systemic pharmacokinetics of ranibizumab after intravitreal administration in patients with retinal vein occlusion (RVO) or diabetic macular edema (DME). Design: A population approach of nonlinear mixed-effect pharmacokinetics modeling based on serum concentrations of ranibizumab measured at various times after intravitreal administration. Participants: Patients with RVO (n ¼ 441) and DME (n ¼ 435) from 4 large, randomized, phase 3 clinical trials of monthly ranibizumab intravitreal administration. Methods: A 1-compartment pharmacokinetics model with first-order absorption and elimination rate constants previously developed in patients with age-related macular degeneration (AMD) was fitted separately to RVO and DME data. Population pharmacokinetic parameters and interindividual variability were estimated for each model. Baseline covariates were evaluated for potential effects on systemic pharmacokinetics. Model performance was validated using general diagnostic plots and a visual predictive check. Main Outcome Measures: Ranibizumab disposition was determined in RVO and DME patients and compared with that previously seen in AMD patients. Results: The AMD pharmacokinetics model correctly predicted the measured serum ranibizumab concentration data for RVO and DME patients. Most observed data points were within the simulated 90% confidence interval, indicating that systemic ranibizumab concentrations were comparable among AMD, RVO, and DME patients. No disease-related covariates were identified by the population pharmacokinetics analysis. Conclusions: The systemic pharmacokinetics of ranibizumab were similar among patients with AMD, RVO, or DME. Disease-related differences and patient demographics, measured in this study, did not lead to variability in ocular elimination or in systemic exposure of ranibizumab after intravitreal administration. In all disease processes tested, ranibizumab exits the eye slowly and then is eliminated rapidly from the circulation, thus minimizing systemic exposure. Ophthalmology 2014;121:2237-2246 ª 2014 by the American Academy of Ophthalmology.
Vascular endothelial growth factor (VEGF)-A, a potent endothelial-specific mitogen usually found as a homodimer, plays an important role in the pathogenesis of several vascular diseases of the eye.1e4 Pathologic changes (including neovascularization, hemorrhages, microaneurysms, and retinal vascular leakage) seen in neovascular age-related macular degeneration (AMD), retinal vein occlusions (RVOs), nonproliferative and proliferative diabetic retinopathy, and diabetic macula edema (DME) have been linked to elevated intraocular levels of VEGF. In animal models, increased levels of VEGF induce retinopathy with features similar to those observed in diabetic retinopathy,2 cause macular edema,3 and stimulate neovascularization from the deep capillary bed of the retina.4 Similarly, in animal models of intraocular neovascularization, neutralization of VEGF suppresses neovascularization.5,6 Ranibizumab (Lucentis; Genentech, Inc, South San Francisco, CA) is an affinity-matured, humanized, monoclonal 2014 by the American Academy of Ophthalmology Published by Elsevier Inc.
antibody antigen binding fragment (Fab) engineered specifically for the eye to inhibit potently all known biologically active isoforms of VEGF-A.7 Intravitreal administration of ranibizumab in patients with neovascular AMD, RVO, or DME causes rapid and sustained improvements in visual acuity, reduction in rates of additional vision loss, and improvement in macular edema.8e13 Treatment with intravitreal ranibizumab also reduced the rate of diabetic retinopathy worsening and substantially improved diabetic retinopathy severity in many patients with DME.14 The United States Food and Drug Administration and similar regulatory agencies in many other countries have approved ranibizumab for the treatment of neovascular AMD, macular edema occurring after RVO, and DME. The pharmacokinetics of ranibizumab after intravitreal administration in patients with AMD is best described by a 1-compartment model with first-order absorption into, and first-order elimination from, the systemic circulation.15 http://dx.doi.org/10.1016/j.ophtha.2014.05.012 ISSN 0161-6420/14
2237
Ophthalmology Volume 121, Number 11, November 2014 Table 1. Retinal Vein Occlusion and Diabetic Macular Edema Studies Included in the Analysis Retinal Vein Occlusion Study Study design Eligibility criteria Ranibizumab dose (mg/eye) Regimen Controlled treatment period (mos) Sampling scheme
No. of patients treated with ranibizumab No. of patients with PK samples collected No. of patients with 1 measurable concentration No. of total PK samples collected No. of measurable PK concentrations* Mean no. of doses per patient Mean no. of evaluable samples per patient
BRAVO (n ¼ 397)
CRUISE (n ¼ 392)
Multicenter, prospective, randomized, doublemasked, sham injection-controlled, registration, phase 3 trial Adult patients with macular edema secondary to BRVO or CRVO 0.3 or 0.5 Monthly 6 Day 7 after first dose; days 3, 7, and 14 after third dose; and months 3 and 6 before injection 264 261 225
261 259 216
871 424 5.73 1.88
870 419 5.7 1.94
Diabetic Macular Edema RIDE (n ¼ 382)
RISE (n ¼ 377)
Multicenter, prospective, randomized, doublemasked, sham injection-controlled, registration, phase 3 trial Adult patients with diabetic macular edema 0.3 or 0.5 Monthly 24 All patients: screening month 2 and day 7 after dose, month 12 and month 24 before dose; PK subset: day 7 after dose, month 2 and days 3 and 14 after dose, and month 3 before dose 252 250 244 243 216 218 1030 445 21.55 2.08
1137 549 21.53 2.52
BRAVO ¼ Ranibizumab for the Treatment of Macular Edema following Branch Retinal Vein Occlusion: Evaluation of Efficacy and Safety study; BRVO ¼ branch retinal vein occlusion; CRUISE ¼ Ranibizumab for the Treatment of Macular Edema following Central Retinal Vein Occlusion Study: Evaluation of Efficacy and Safety; CRVO ¼ central retinal vein occlusion; DME ¼ diabetic macular edema; PK ¼ pharmacokinetic; RIDE and RISE ¼ A Study of Ranibizumab Injection in Subjects with Clinically Significant Macular Edema with Center Involvement Secondary to Diabetes Mellitus. *Unmeasurable samples had concentrations less than the limit of detection.
Vitreous elimination half-life was calculated to be 9 days, and after this slow egress from the eye, the intrinsic systemic elimination half-life was calculated to be approximately 2 hours. The experimentally determined half-life after a single injection of 0.5 mg ranibizumab was 7.19 days.16 Regardless, the systemic-to-vitreous exposure ratio is extremely low, approximately 1:90 000.15 The systemic pharmacokinetics of ranibizumab have not yet been described in the RVO and DME patient populations. The objective of this analysis was to evaluate whether systemic ranibizumab exposure is similar in patients with RVO or DME compared with that seen in patients with AMD.
Methods Clinical Trials Included in the Analysis Two phase 3 clinical studies testing the effect of ranibizumab in patients with RVO were included in the population pharmacokinetics analysis: Ranibizumab for the Treatment of Macular Edema following Branch Retinal Vein Occlusion: Evaluation of Efficacy and Safety study (ClinicalTrials.gov identifier, NCT00486018)10,17 and Ranibizumab for the Treatment of Macular Edema following Central Retinal Vein Occlusion Study: Evaluation of Efficacy and Safety study (ClinicalTrials.gov identifier, NCT00485836).9,18 Two phase 3 clinical trials testing the effect of ranibizumab in patients with DME also were included: A Study of Ranibizumab Injection in Subjects with Clinically Significant Macular Edema with Center Involvement Secondary to Diabetes Mellitus (Clinical Trials.gov identifier, [RIDE] NCT00473382) and A Study of Ranibizumab Injection in Subjects with Clinically Significant
2238
Macular Edema with Center Involvement Secondary to Diabetes Mellitus (ClinicalTrials.gov identifier, [RISE], NCT00473330).12 The trial designs, treatment frequency and duration, and pharmacokinetics sampling schemes are summarized in Table 1. The clinical trials included in this analysis were conducted in accordance with the Declaration of Helsinki, United States Food and Drug Administration regulations, and the Health Insurance Portability and Accountability Act. The respective institutional review boards approved the study protocols and all enrolled participants provided informed consent in writing.
Ranibizumab Pharmacokinetic Assay After intravitreal administration of ranibizumab, the concentration of drug in serum was measured using a homogenous bridging enzyme-linked immunosorbent assay developed at Genentech, Inc. Briefly, serum samples were incubated overnight with biotinylated recombinant human VEGF and rabbit antiranibizumab polyclonal antibody. The complex formed by ranibizumab and these 2 molecules was captured onto a streptavidin-coated plate and quantified after sequential addition of horseradish peroxidase conjugate and a peroxidase substrate solution. The resulting colorimetric reaction was read on a plate reader. The lower limit of quantitation of this assay was 0.075 ng/ml, and concentrations lower than the lower limit of quantitation were considered less than reportable (LTR) for human serum samples. The assay used in the AMD population had a lower limit of quantitation of 0.3 ng/ml.15,19 The assay used in the present study measures both VEGF-bound and free ranibizumab. It was fully validated and shown to measure ranibizumab in human serum with acceptable accuracy, as well as interassay and intraassay precision: all measured values in accuracy experiments were within 10% of the spiked ranibizumab concentration and precision experiments resulted in coefficient of variation values of less than 10% at all ranibizumab concentration levels tested. Stability of
Zhang et al
Pharmacokinetics of Ranibizumab
Table 2. Descriptive Summary of Baseline Covariates Parameter
Retinal Vein Occlusion: Pooled BRAVO and CRUISE (n [ 441)
Median age (range), yrs Median CrCL (ml/min) Median BLVA (range), no. of letters Median BLA (range) Median BCFT (range), mm Median AONP (mm2) TTDX (day) BIOP (mmHg) IOP (mmHg) DIOP (mmHg)
67.0 (38e91) 81.1 (10.9e217.2) 53.0 (9e79) NA 587 (126e1651) 0 (0e13.22) 60 (0e1058) 15 (0e28) 17 (4e50) 3 (8 to 31)
Diabetic Macular Edema: Pooled RIDE and RISE (n [ 434) 62 91.5 58 9.2 449
(21e88) (10.6e375.4) (15e85) (0.0e16.0) (82e1241) NA NA NA NA NA
ANOP ¼ area of nonperfusion; BCFT ¼ baseline central foveal thickness; BIOP ¼ intraocular pressure before injection; BLA ¼ baseline leakage area; BLVA ¼ baseline visual acuity; BRAVO ¼ Ranibizumab for the Treatment of Macular Edema following Branch Retinal Vein Occlusion: Evaluation of Efficacy and Safety study; CrCL ¼ creatinine clearance; CRUISE ¼ Ranibizumab for the Treatment of Macular Edema following Central Retinal Vein Occlusion Study: Evaluation of Efficacy and Safety; DIOP ¼ difference between BIOP and IOP; IOP ¼ intraocular pressure (after injection); NA ¼ not applicable; RIDE and RISE ¼ A Study of Ranibizumab Injection in Subjects with Clinically Significant Macular Edema with Center Involvement Secondary to Diabetes Mellitus; RVO ¼ retinal vein occlusion; TTDX ¼ time to diagnosis. The total number of patients was 785. The type of RVO was the only categorical covariate tested, and there were no values missing.
ranibizumab in human serum also was evaluated, and it was shown to be stable for up to 4 freezeethaw cycles.
Pharmacokinetic Analysis
Ranibizumab Serum Concentration Data
The pharmacokinetic analysis comprised 3 stages: (1) assessing similarity in systemic exposure between RVO, DME, and AMD patients with a previously developed population-based pharmacokinetics model in AMD (the AMD population pharmacokinetics base model is shown in Fig 1)15; (2) development of pharmacokinetics
The pharmacokinetics sampling schedule can be found in Table 1. Serum concentration data were excluded from the analysis if they met 1 or more of the following criteria: measurable concentrations before the first dose, outliers (defined as concentration data with weighted residuals of more than 5 after fitting data to the base model using nonlinear mixed-effects modeling [NONMEM program versions 6 and 7; ICON Development Solutions, Ellicott City, Maryland]), data points that were found by inspection of visual predictive check plots to exhibit pharmacologically implausible spikes or drops in concentration (e.g., peak concentration at trough sampling time), and patients who did not have any measurable serum ranibizumab concentration after administration. Less-than-reportable samples after dose administration were included in the analysis for RVO. We used the maximum likelihood method20 for parameter estimation. A sensitivity analysis was performed to examine the impact of discarding LTR data20 on the base model estimation. No difference was found in parameter estimation between the maximum likelihood method and the sensitivity analysis performed to examine the impact of discarding LTR data. For the DME analysis, we used the sensitivity analysis performed to examine the impact of discarding LTR data to handle LTR samples, in which LTR observations were excluded from the analysis.
Covariates Included for the Analysis Patient characteristics measured at baseline were evaluated for covariate effects on the disposition of ranibizumab (Table 2). If covariates were missing in less than 20% of patients, continuous covariates were imputed for those patients at the median value. Categorical covariates were imputed for the most frequent category. If covariates were missing in more than 20% of patients, the missing covariates were not imputed and the covariate was excluded in covariate analysis.
Ranibizumab vitreous dA vit dt
Ka x Avit
Avit (0) = Dose Ka Systemic circulation dA sys dt
Ka x Avit
Asys Vc
CL Figure 1. Schematic representation of the 1-compartment model of the disposition of ranibizumab after intravitreal administration with first-order absorption into, and first-order elimination from, the systemic circulation. Asys ¼ amount of ranibizumab in the systemic compartment; Avit ¼ amount of ranibizumab in the vitreous compartment; CL ¼ clearance; Ka ¼ rate of systemic absorption/rate of vitreous elimination; Vc ¼ apparent volume of the central compartment.
2239
Ophthalmology Volume 121, Number 11, November 2014
Serum ranibizumab concentration (ng/ml)
A
0.5 mg/eye 5.00
0.50
LLOQ
0.05
0.01 0
10
20
30
40
Time (day)
B
0.5 mg/eye
Serum ranibizumab concentration (ng/ml)
5.00
1.00 0.50
0.10
LLOQ 0.05 0
10
20
30
Time after dose (day) Figure 2. Graphs showing simulated and observed ranibizumab serum concentration in (A) retinal vein occlusion patients and (B) diabetic macular edema patients using population pharmacokinetics model in age-related macular degeneration patients. LLOQ ¼ lower limit of quantitation.
Table 3. Population Parameter Estimates for the Retinal Vein Occlusion and Diabetic Macular Edema Final Model
Parameter
Final Age-Related Macular Degeneration Model19 (Percentage Relative Standard Error of the Estimates)
Retinal Vein Occlusion (Percentage Coefficient of Variation of the Estimate)
Diabetic Macular Edema (Percentage Coefficient of Variation of the Estimate)
CL/F* (l/day) Ka (1/day) Vc/F (l)
24.1 (4.52) 0.0806 (7.33) 3.01 (13.3)
28.5 (2.11) 0.106 (4.53) 2.54 (Fixed)
24.8* (31.3) 0.0969 (26.8) 2.77 (270)
CL ¼ clearance; CL/F ¼ apparent CL, CL divided by bioavailability (F); Ka ¼ rate of systemic absorption (rate of vitreous elimination); Vc/F ¼ apparent volume of the central compartment, or in short for a 1-compartmental model, volume of distribution. Table shows typical values of parameters and interpatient percentage coefficient of variation of the estimate. *The median creatinine clearance (CrCL) in the diabetic macular edema population was 91.5 ml/min; in this table, CL/F for the diabetic macular edema population was scaled to CrCL ¼ 65.2 ml/min, that is, the median CrCL in the age-related macular degeneration population.
2240
Zhang et al
Pharmacokinetics of Ranibizumab
Table 4. Individual Post Hoc Apparent Clearance (Clearance Divided by Bioavailability [F]) per Renal Function Groups22
Renal Function RVO study Normal Mild impairment Moderate impairment Severe impairment DME study Normal Mild impairment Moderate impairment Severe impairment
Estimated Creatinine Clearance (ml/min)
No. of Patients (% of Total)
Ranibizumab Apparent Clearance (Clearance Divided by Bioavailability (F)), Mean ± Standard Deviation (l/day)
>80 50e80 30e50 30
272 191 50 12
(51.8) (36.4) (9.5) (2.3)
31.610.8 20.55.20 15.93.45 9.362.69
>80 50e80 30e50 30
267 120 40 7
(61.5) (27.6) (9.2) (1.6)
36.810.2 25.26.4 19.34.6 12.42.8
DME ¼ diabetic macular edema; RVO ¼ retinal vein occlusion.
models for RVO and DME patients; and (3) validation of final models (with covariates added). We evaluated the adequacy of the AMD population pharmacokinetics model in describing data from RVO and DME patients by simulating the serum concentrations in RVO and DME patients with the AMD population pharmacokinetics model. For comparison, simulated serum concentrations were plotted against time from last dose and were overlaid with observed concentrations in RVO and DME, respectively. To develop the models for RVO and DME, the population pharmacokinetics analysis was performed using the nonlinear mixed-effects modeling program. Typical population values of pharmacokinetics parameters, as well as interpatient variability on those parameters, were estimated. Covariate model building was conducted using the step-wise reduction method based on the likelihood ratio test. The difference (d) in objective functions (asymptotically chi-square distributed) was used to compare alternative models. All model comparisons were performed at the P ¼ 0.01 level (d ¼ 7.88 for 1 degree of freedom). We evaluated the performance of the model in describing the observed data with standard diagnostics tools such as goodnessof-fit plots, where observed concentrations were compared with predicted concentrations. Model stability was assessed with the bootstrap approach, where parameter estimation was performed using randomly drawn data from the original data set. The estimation confidence intervals (CIs) were compared with the original point estimates. A visual predictive check was undertaken to evaluate the predictive performance of the final population pharmacokinetics model (with all covariates) for RVO and DME patients. We performed the simulations using final parameter estimates. Simulation and observations were plotted together against time from last dose for RVO and DME, respectively.
Results There were a total of 843 measurable ranibizumab serum concentrationetime records from 441 patients with RVO and 994 measurable concentrationetime records from 434 patients with DME (Table 1). Ranibizumab serum concentrations between days 3 and 30 after monthly 0.3-mg or 0.5-mg intravitreal dosing in patients with AMD, RVO, and DME were low relative to the concentration (IC50) necessary to inhibit the biological activity of VEGF-A by 50% IC50 (3 ng/ml).21 The number of patients with 1 or more sample with ranibizumab serum concentrations above the
IC50 of 3 ng/ml was 34 of 238 for AMD, 13 of 441 for RVO, and 6 of 434 for DME. All samples with ranibizumab concentrations of more than 3 ng/ml were collected at day 7 or earlier after ranibizumab administration. The AMD pharmacokinetics model accurately described the observed ranibizumab serum concentration data for both RVO and DME patients. The plot overlaying simulation and observation showed most observed RVO and DME data falling within the simulated 90% CI, indicating that the systemic ranibizumab concentrations were comparable among AMD, RVO, and DME patients (Fig 2). Final model parameters are shown in Table 3. Apparent serum clearance and rate of systemic absorption of ranibizumab in RVO and DME patients were consistent with those observed in AMD patients.15 No disease-related (RVO or DME) covariates, such as baseline visual acuity, baseline central foveal thickness, or area of nonperfusion, were identified. Some of the covariates were not included in the DME analysis because they had been found to be insignificant in the RVO analysis. Similar to AMD, the effect of creatinine clearance on apparent serum clearance was statistically significant for both RVO and DME. Individual serum ranibizumab clearance estimates by renal function groups22 are presented in Table 4 for both RVO and DME. Differences in creatinine clearance were estimated to account for 46% of the variability in apparent serum clearance of ranibizumab in RVO patients. A DME patient with creatinine clearance of 50 ml/minute had 30% lower total apparent serum clearance of ranibizumab compared with a DME patient with creatinine clearance of 90 ml/minute. Only approximately 10% of DME patients in this analysis had a creatinine clearance of 50 ml/minute or less. The fit of the model predictions to observed data for a few representative patients is illustrated in Figure 3. Figure 3A shows predicted and observed ranibizumab serum concentrations in RVO patients plotted against time, starting with the first dose (from day 0 to day 180). Figure 3B shows ranibizumab serum concentrations in DME patients plotted against time since most recent dose (from day 0 to day 30). Both plots illustrate that, on the individual level, predicted concentrations fit the observations well. During visual predictive check, most of the observed concentrations were within the simulated 90% CI using the RVO or DME final model, suggesting that the final covariate model accurately predicts ranibizumab serum concentration data (Fig 4). The final population pharmacokinetics model was fit repeatedly to 1000 datasets obtained by bootstrapping the original database. Median values of the bootstrapped parameters were consistent with the point estimates obtained from the final model (Table 5), and the 5% to 95% CIs were narrow.
2241
Ophthalmology Volume 121, Number 11, November 2014
5.00
DV PRED IPRED
Serum ranibizumab concentration (ng/ml)
Serum ranibizumab concentration (ng/ml)
A
0.50
DV PRED IPRED
0.50
0.01
0.01 50
0
100
150
50
0
100
Time (day)
5.00
DV PRED IPRED
Serum ranibizumab concentration (ng/ml)
Serum ranibizumab concentration (ng/ml)
5.00
0.50
0.01
DV PRED IPRED
0.50
0.01 0
50
100
0
150
50
100
Time (day)
B
DV PRED IPRED
5.00
Serum ranibizumab concentration (ng/ml)
Serum ranibizumab concentration (ng/ml)
150
Time (day)
DV PRED IPRED
5.00
0.50
0.05
0.50
0.05 0
10
20
30
0
40
10
Time after dose (days)
20
30
40
Time after dose (days)
DV PRED IPRED
DV PRED IPRED
5.00
Serum ranibizumab concentration (ng/ml)
5.00
Serum ranibizumab concentration (ng/ml)
150
Time (day)
0.50
0.05
0.50
0.05 0
10
20
30
Time after dose (days)
40
0
10
20
30
40
Time after dose (days)
Figure 3. Goodness-of-fit plots for the final population pharmacokinetic model showing individual predicted (IPRED) results (ng/ml [black dashed line]), population predicted (PRED) results (ng/ml [solid line]) overlaid on observed concentrations (DV) as a function of time after dose (days) (horizontal gray dashed line is the less than reportable value of 0.075 ng/ml) for (A) retinal vein occlusion and (B) diabetic macular edema.
2242
A
Serum ranibizumab concentration (ng/ml)
Zhang et al
Pharmacokinetics of Ranibizumab
5.00
0.50
LLOQ
0.05 0.01 0
10
20
30
40
Time after dose (day)
Serum ranibizumab concentration (ng/ml)
B
5.00
1.00 0.50
0.10
LLOQ 0.05 0
10
20
30
Time after dose (day) Figure 4. Visual predictive check and goodness-of-fit plots of (A) retinal vein occlusion and (B) diabetic macular edema final covariate models. LLOQ ¼ lower limit of quantitation.
Discussion Results from the current population pharmacokinetics analysis demonstrate that the systemic exposure and pharmacokinetics of ranibizumab are comparable in RVO, DME, and AMD patients. The pharmacokinetics parameter estimates were comparable among the 3 indications with moderate numerical difference. Estimation of apparent serum clearance of ranibizumab was slightly lower in the AMD population compared with the RVO or DME populations. This was because the AMD population was older and had reduced renal function on average compared with the RVO and DME populations. The difference in estimations of systemic absorption of ranibizumab between AMD and the other 2 indications was likely the result of differences in sampling time. The AMD pharmacokinetics data
set included samples obtained as early as 1 hour and 1 day after dose administration, whereas the RVO and DME data sets did not include these early time points, which are most informative for estimating systemic absorption. Therefore, the numerical difference in estimates of systemic absorption of ranibizumab is not likely to reflect any meaningful differences in the rate of exit of ranibizumab from eyes of patients with AMD, RVO, or DME. The very similar pharmacokinetics profiles seen in patients with AMD, DME, or RVO suggest that there are no disease-related differences in the disposition of ranibizumab after intravitreal injection. Similar to the situation in AMD patients, systemic exposure of ranibizumab was low in patients with RVO or DME. This explains the generally favorable systemic safety profile seen in patients receiving monthly intravitreal injections of ranibizumab in several phase 3 clinical trials. It
2243
Ophthalmology Volume 121, Number 11, November 2014 Table 5. Bootstrap of Parameter Estimation for Retinal Vein Occlusion and Diabetic Macular Edema Bootstrap Final Model, Median (5%e95% Confidence Interval) Parameter Typical CL/F (l/day) Typical Ka (/day) CrCL on CL/F uCL/F (%) uKa (%) spop (%)
Retinal Vein Occlusion 28.4 (24.8e29.5) 0.107 0.624 29.4 39.5 24.6
(0.0989e0.117) (0.495e0.747) (25.5e40.2) (29.1e58.1) (17.6e28.3)
Diabetic Macular Edema 30.4 (29.3e31.6) 0.097 0.597 30.8 26.5 28.2
(0.093e0.101) (0.516e0.664) (27.3e34.0) (21.9e29.9) (26.0e30.7)
CL/F ¼ apparent clearance, clearance divided by bioavailability (F); CrCL ¼ creatinine clearance; Ka ¼ rate of systemic absorption (rate of vitreous elimination); uCL/F ¼ intersubject variability as percent coefficient of variation of the estimate of apparent clearance, clearance divided by bioavailability (F); uKa ¼ intersubject variability as percent coefficient of variation of the estimate of rate of systemic absorption (rate of vitreous elimination); spop ¼ standard deviation describing the proportional component of residual variability.
should be noted that the molecular characteristics of ranibizumab (high affinity for VEGF obtained by affinity maturation, extremely slow dissociation rate when in complex with VEGF, small size, and absence of an Fc region) were designed specifically to combine potent inhibition of VEGF in the intraocular compartment with minimal VEGF inhibition in extraocular blood vessels and tissues because of its rapid elimination from the systemic circulation.7,23,24 This is an important pharmacologic property of ranibizumab because inhibition of systemic VEGF may impair normal angiogenesis and interfere with the integrity of the endothelial and epithelial surfaces, vascular tone, and the filtration function of the kidneys. This can cause systemic VEGF inhibition class side effects such as hypertension, proteinuria, arterial thromboembolic events, gastrointestinal perforation, and wound-healing complications.25 Because full-length antibodies have an Fc fragment that increases residence time in circulation, it is expected that bevacizumab and the Fc-containing aflibercept have greater systemic exposure compared with ranibizumab, and this was demonstrated to be the case (Avery R, Castellarin AA, Steinle NC, et al. Comparison of systemic pharmacokinetics following anti-VEGF intravitreal injections of ranibizumab, bevacizumab, and aflibercept. Paper presented at: AAO Annual Meeting; November 2013; Las Vegas, Nevada). Consistent with this observation, the alternative treatments to that used in the Inhibit VEGF in Age-Related Choroidal Neovascularization clinical trial showed that patients with neovascular AMD treated with intraocular injections of bevacizumab had a significant reduction in serum VEGF levels compared with patients treated with ranibizumab.26 These data were replicated in another clinical trial.27 Thus, on theoretical grounds, it is expected that Fc-containing VEGF antagonists may carry a higher risk of systemic adverse events than ranibizumab, but the rates of such side effects are very low, and efficacy studies are not powered to identify differences in relatively rare events. However, such
2244
differences might exist because in the Comparison of AgeRelated Macular Degeneration Treatments Trials, patients with neovascular AMD treated with intraocular injections of bevacizumab had a significantly greater number of systemic serious adverse events at the 1- and 2-year time points than seen in patients treated with ranibizumab.13,28 The authors of the study discussed whether this difference truly indicates increased bevacizumab risk or was simply the result of chance and suggested that because they did not understand the mechanism of many of the serious adverse events, the possibility that the difference is the result of chance may be favored. However, it may not be prudent to attribute the difference to chance simply because the serious adverse events do not conform to expectations regarding mechanisms of blocking VEGF. Similar to the situation in AMD patients, creatinine clearance was identified to be a statistically significant covariate of apparent serum clearance of ranibizumab in RVO and DME patients. It is unlikely that reduced renal function would cause a clinically meaningful decrease in apparent serum clearance of ranibizumab and a consequent increase in systemic exposure that would affect the efficacy or safety of ranibizumab. Efficacy should not be influenced by systemic exposure because ocular compartments are the sites of action for ranibizumab and renal function does not affect the elimination of ranibizumab from the vitreous. Safety is unlikely to be influenced because systemic exposure in RVO and DME patients was low and not expected to be pharmacologically active even in patients with mild to moderate renal impairment. The severe renal impairment in approximately 2% of the RVO and DME patients was not found to be associated with significantly increased rates of VEGF-related serious adverse events. Moreover, creatinine clearance had no statistically or clinically meaningful effect on vitreous elimination of ranibizumab quantified by the absorption rate constant (Ka). Hence, renal impairment is not expected to affect vitreous exposure to ranibizumab. No other disease-related factors or baseline patient characteristics were found to have significant effects on ranibizumab serum pharmacokinetics. This suggests that differences in disease states among patients with AMD, RVO, or DME do not contribute to the variability in pharmacokinetics. In conclusion, results from this population pharmacokinetics analysis demonstrate that serum levels of ranibizumab after monthly intravitreal injections are similar in RVO or DME patients compared with those seen in patients with AMD. This suggests that regardless of whether a patient is being treated for AMD, RVO or DME, intravitreal injection of the same amount of ranibizumab is likely to lead to similar intraocular concentrations, similar egress from the eye into systemic circulation, and ultimately similar serum concentrations. In patients with AMD, RVO, or DME, systemic exposure is comparable. The low systemic exposure to ranibizumab after intraocular injections in all patient populations studied is reassuring and likely to account for the low rates of systemic complications seen in phase 3 ranibizumab treatment trials regardless of treatment indication.
Zhang et al
Pharmacokinetics of Ranibizumab
Acknowledgments. Support for third-party writing and formatting assistance for this manuscript by Rebecca Jarvis, PhD, CMPP, was provided by Genentech, Inc.
References 1. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med 2003;9:669–76. 2. Tolentino MJ, Miller JW, Gragoudas ES, et al. Intravitreous injections of vascular endothelial growth factor produce retinal ischemia and microangiopathy in an adult primate. Ophthalmology 1996;103:1820–8. 3. Ozaki H, Hayashi H, Vinores SA, et al. Intravitreal sustained release of VEGF causes retinal neovascularization in rabbits and breakdown of the blood-retinal barrier in rabbits and primates. Exp Eye Res 1997;64:505–17. 4. Okamoto N, Tobe T, Hackett SF, et al. Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization. Am J Pathol 1997;151:281–91. 5. Aiello LP, Pierce EA, Foley ED, et al. Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins. Proc Natl Acad Sci U S A 1995;92:10457–61. 6. Adamis AP, Shima DT, Tolentino MJ, et al. Inhibition of vascular endothelial growth factor prevents retinal ischemiaassociated iris neovascularization in a nonhuman primate. Arch Ophthalmol 1996;114:66–71. 7. Ferrara N, Damico L, Shams N, et al. Development of ranibizumab, an anti-vascular endothelial growth factor antigen binding fragment, as therapy for neovascular age-related macular degeneration. Retina 2006;26:859–70. 8. Brown DM, Kaiser PK, Michels M, et al; ANCHOR Study Group. Ranibizumab versus verteporfin for neovascular agerelated macular degeneration. N Engl J Med 2006;355:1432–44. 9. Brown DM, Campochiaro PA, Singh RP, et al; CRUISE Investigators. Ranibizumab for macular edema following central retinal vein occlusion: six-month primary end point results of a phase III study. Ophthalmology 2010;117:1124–33. 10. Campochiaro PA, Heier JS, Feiner L, et al; BRAVO Investigators. Ranibizumab for macular edema following branch retinal vein occlusion: six-month primary endpoint results of a phase III study. Ophthalmology 2010;117:1102–12. 11. Diabetic Retinopathy Clinical Research Network, Elman MJ, Aiello LP, Beck RW, et al. Randomized trial evaluating ranibzumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology 2010;117:1064–77. 12. Nguyen QD, Brown DM, Marcus DM, et al; RISE and RIDE Research Group. Ranibizumab for diabetic macular edema: results from 2 phase III randomized trials: RISE and RIDE. Ophthalmology 2012;119:789–801. 13. CATT Research Group. Ranibizumab and bevacizumab for neovascular age-related macular degeneration. N Engl J Med 2011;364:1897–908. 14. Ip MS, Domalpally A, Hopkins JJ, et al. Long-term effects of ranibizumab on diabetic retinopathy severity and progression. Arch Ophthalmol 2012;130:1145–52.
15. Xu L, Lu T, Tuomi L, et al. Pharmacokinetics of ranibizumab in patients with neovascular age-related macular degeneration: a population approach. Invest Ophthalmol Vis Sci 2013;54: 1616–24. 16. Krohne TU, Liu Z, Holz FG, Meyer CH. Intraocular pharmacokinetics of ranibizumab following a single intravitreal injection in humans. Am J Ophthalmol 2012;154: 682–6. 17. Brown DM, Campochiaro PA, Bhisitkul RB, et al. Sustained benefits from ranibizumab for macular edema following branch retinal vein occlusion: 12-month outcomes of a phase III study. Ophthalmology 2011;118:1594–602. 18. Campochiaro PA, Brown DM, Awh CC, et al. Sustained benefits from ranibizumab for macular edema following central retinal vein occlusion: twelve-month outcomes of a phase III study. Ophthalmology 2011;118:2041–9. 19. Lowe J, Maia M, Wakshull E, et al. Development of a novel homogenous electrochemiluminescence assay for quantitation of ranibizumab in human serum. J Pharm Biomed Anal 2010;52:680–6. 20. Beal SL. Ways to fit a PK model with some data below the quantification limit. J Pharmacokinet Pharmacodyn 2001;28: 481–504. 21. Yu L, Liang XH, Ferrara N. Comparing protein VEGF inhibitors: in vitro biological studies. Biochem Biophys Res Comm 2011;408:276–81. 22. Center for Drug Evaluation and Research. Guidance for Industry. Guidance for industry. Pharmacokinetics in patients with impaired renal functiondstudy design, data analysis, and impact on dosing and labeling. Rockville, MD: Center for Biologics Evaluation and Research; May 1998. Available at: http://www.fda.gov/downloads/Drugs/GuidanceComplianceR egulatory%20Information/Guidances/UCM072127.pdf. Accessed on April 2, 2014. 23. Chen Y, Wiesmann C, Fuh G, et al. Selection and analysis of an optimized anti-VEGF antibody: crystal structure of an affinity-matured Fab in complex with antigen. J Mol Biol 1999;293:865–81. 24. Lowe J, Araujo J, Yang J, et al. Ranibizumab inhibits multiple forms of biologically active vascular endothelial growth factor in vitro and in vivo. Exp Eye Res 2007;85:425–30. 25. Chen HX, Cleck JN. Adverse effects of anticancer agents that target the VEGF pathway. Nat Rev Clin Oncol 2009;6: 465–77. 26. IVAN Study Investigators, Chakravarthy U, Harding SP, Rogers CA, et al. Ranibizumab versus bevacizumab to treat neovascular age-related macular degeneration: one-year findings from the IVAN randomized trial. Ophthalmology 2012;119:1399–411. 27. Zehetner C, Kirchmair R, Huber S, et al. Plasma levels of vascular endothelial growth factor before and after intravitreal injection of bevacizumab, ranibizumab and pegaptanib in patients with age-related macular degeneration, and in patients with diabetic macular oedema. Br J Ophthalmol 2013;97: 454–9. 28. Comparison of Age-Related Macular Degeneration Treatments Trials (CATT) Research Group, Martin DF, Maguire MG, Fine SL, et al. Ranibizumab and bevacizumab for treatment of neovascular age-related macular degeneration: two-year results. Ophthalmology 2012;119:1388–98.
2245
Ophthalmology Volume 121, Number 11, November 2014
Footnotes and Financial Disclosures Originally received: October 16, 2013. Final revision: February 3, 2014. Accepted: May 13, 2014. Available online: July 4, 2014.
Manuscript no. 2013-1746.
1
Genentech, Inc., South San Francisco, California.
2
Quantitative Solutions, Menlo Park, California.
3
The Wilmer Eye Institute, Johns Hopkins School of Medicine, Baltimore, Maryland. Presented at: Association for Research in Vision and Ophthalmology Annual Meeting, May 2011 and May 2012, Fort Lauderdale, Florida. Financial Disclosure(s): The author(s) have made the following disclosure(s): Yi Zhang: Employee e Genentech, Inc. (South San Francisco, California). Zhenling Yao: Employee e Genentech, Inc. (South San Francisco, California). Nitin Kaila: Consultant e Genentech, Inc. (South San Francisco, California); Employee e Quantitative Solutions. Peter Kuebler: Employee e Genentech, Inc. (South San Francisco, California). Jennifer Visich: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche. Mauricio Maia: Employee e Genentech, Inc. (South San Francisco, California). Lisa Tuomi: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche.
2246
Jason S. Ehrlich: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche. Roman G. Rubio: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche. Peter A. Campochiaro: Consultant e Genentech, Inc. (South San Francisco, California), GlaxoSmithKline, Regeneron, Aerpio Therapeutics (employer, Johns Hopkins University, receives compensation), and Elan, Gene Signal, Norvorx (receives personal compensation); Data and Safety Monitoring Committee e Advanced Cell Technology, Regeneron; Financial support e Genentech, GlaxoSmithKline, Genzyme, Oxford BioMedica; Equity owner e Graybug, Inc. Genentech, Inc., South San Francisco, California, provided support for the study; participated in study design and conduct of the study; provided data collection, management, and interpretation; as well as reviewed and approved the manuscript. Abbreviations and Acronyms: AMD ¼ age-related macular degeneration; CI ¼ confidence interval; DME ¼ diabetic macular edema; LTR ¼ less than reportable; RVO ¼ retinal vein occlusion; VEGF ¼ vascular endothelial growth factor. Correspondence: Peter A. Campochiaro, MD, The Wilmer Eye Institute, Johns Hopkins Hospital School of Medicine, 719 Maumenee, 600 North Wolfe St, Baltimore, MD 21287-9277. E-mail: [email protected].
Vascular endothelial growth factor (VEGF)-A, a potent endothelial-specific mitogen usually found as a homodimer, plays an important role in the pathogenesis of several vascular diseases of the eye.1e4 Pathologic changes (including neovascularization, hemorrhages, microaneurysms, and retinal vascular leakage) seen in neovascular age-related macular degeneration (AMD), retinal vein occlusions (RVOs), nonproliferative and proliferative diabetic retinopathy, and diabetic macula edema (DME) have been linked to elevated intraocular levels of VEGF. In animal models, increased levels of VEGF induce retinopathy with features similar to those observed in diabetic retinopathy,2 cause macular edema,3 and stimulate neovascularization from the deep capillary bed of the retina.4 Similarly, in animal models of intraocular neovascularization, neutralization of VEGF suppresses neovascularization.5,6 Ranibizumab (Lucentis; Genentech, Inc, South San Francisco, CA) is an affinity-matured, humanized, monoclonal 2014 by the American Academy of Ophthalmology Published by Elsevier Inc.
antibody antigen binding fragment (Fab) engineered specifically for the eye to inhibit potently all known biologically active isoforms of VEGF-A.7 Intravitreal administration of ranibizumab in patients with neovascular AMD, RVO, or DME causes rapid and sustained improvements in visual acuity, reduction in rates of additional vision loss, and improvement in macular edema.8e13 Treatment with intravitreal ranibizumab also reduced the rate of diabetic retinopathy worsening and substantially improved diabetic retinopathy severity in many patients with DME.14 The United States Food and Drug Administration and similar regulatory agencies in many other countries have approved ranibizumab for the treatment of neovascular AMD, macular edema occurring after RVO, and DME. The pharmacokinetics of ranibizumab after intravitreal administration in patients with AMD is best described by a 1-compartment model with first-order absorption into, and first-order elimination from, the systemic circulation.15 http://dx.doi.org/10.1016/j.ophtha.2014.05.012 ISSN 0161-6420/14
2237
Ophthalmology Volume 121, Number 11, November 2014 Table 1. Retinal Vein Occlusion and Diabetic Macular Edema Studies Included in the Analysis Retinal Vein Occlusion Study Study design Eligibility criteria Ranibizumab dose (mg/eye) Regimen Controlled treatment period (mos) Sampling scheme
No. of patients treated with ranibizumab No. of patients with PK samples collected No. of patients with 1 measurable concentration No. of total PK samples collected No. of measurable PK concentrations* Mean no. of doses per patient Mean no. of evaluable samples per patient
BRAVO (n ¼ 397)
CRUISE (n ¼ 392)
Multicenter, prospective, randomized, doublemasked, sham injection-controlled, registration, phase 3 trial Adult patients with macular edema secondary to BRVO or CRVO 0.3 or 0.5 Monthly 6 Day 7 after first dose; days 3, 7, and 14 after third dose; and months 3 and 6 before injection 264 261 225
261 259 216
871 424 5.73 1.88
870 419 5.7 1.94
Diabetic Macular Edema RIDE (n ¼ 382)
RISE (n ¼ 377)
Multicenter, prospective, randomized, doublemasked, sham injection-controlled, registration, phase 3 trial Adult patients with diabetic macular edema 0.3 or 0.5 Monthly 24 All patients: screening month 2 and day 7 after dose, month 12 and month 24 before dose; PK subset: day 7 after dose, month 2 and days 3 and 14 after dose, and month 3 before dose 252 250 244 243 216 218 1030 445 21.55 2.08
1137 549 21.53 2.52
BRAVO ¼ Ranibizumab for the Treatment of Macular Edema following Branch Retinal Vein Occlusion: Evaluation of Efficacy and Safety study; BRVO ¼ branch retinal vein occlusion; CRUISE ¼ Ranibizumab for the Treatment of Macular Edema following Central Retinal Vein Occlusion Study: Evaluation of Efficacy and Safety; CRVO ¼ central retinal vein occlusion; DME ¼ diabetic macular edema; PK ¼ pharmacokinetic; RIDE and RISE ¼ A Study of Ranibizumab Injection in Subjects with Clinically Significant Macular Edema with Center Involvement Secondary to Diabetes Mellitus. *Unmeasurable samples had concentrations less than the limit of detection.
Vitreous elimination half-life was calculated to be 9 days, and after this slow egress from the eye, the intrinsic systemic elimination half-life was calculated to be approximately 2 hours. The experimentally determined half-life after a single injection of 0.5 mg ranibizumab was 7.19 days.16 Regardless, the systemic-to-vitreous exposure ratio is extremely low, approximately 1:90 000.15 The systemic pharmacokinetics of ranibizumab have not yet been described in the RVO and DME patient populations. The objective of this analysis was to evaluate whether systemic ranibizumab exposure is similar in patients with RVO or DME compared with that seen in patients with AMD.
Methods Clinical Trials Included in the Analysis Two phase 3 clinical studies testing the effect of ranibizumab in patients with RVO were included in the population pharmacokinetics analysis: Ranibizumab for the Treatment of Macular Edema following Branch Retinal Vein Occlusion: Evaluation of Efficacy and Safety study (ClinicalTrials.gov identifier, NCT00486018)10,17 and Ranibizumab for the Treatment of Macular Edema following Central Retinal Vein Occlusion Study: Evaluation of Efficacy and Safety study (ClinicalTrials.gov identifier, NCT00485836).9,18 Two phase 3 clinical trials testing the effect of ranibizumab in patients with DME also were included: A Study of Ranibizumab Injection in Subjects with Clinically Significant Macular Edema with Center Involvement Secondary to Diabetes Mellitus (Clinical Trials.gov identifier, [RIDE] NCT00473382) and A Study of Ranibizumab Injection in Subjects with Clinically Significant
2238
Macular Edema with Center Involvement Secondary to Diabetes Mellitus (ClinicalTrials.gov identifier, [RISE], NCT00473330).12 The trial designs, treatment frequency and duration, and pharmacokinetics sampling schemes are summarized in Table 1. The clinical trials included in this analysis were conducted in accordance with the Declaration of Helsinki, United States Food and Drug Administration regulations, and the Health Insurance Portability and Accountability Act. The respective institutional review boards approved the study protocols and all enrolled participants provided informed consent in writing.
Ranibizumab Pharmacokinetic Assay After intravitreal administration of ranibizumab, the concentration of drug in serum was measured using a homogenous bridging enzyme-linked immunosorbent assay developed at Genentech, Inc. Briefly, serum samples were incubated overnight with biotinylated recombinant human VEGF and rabbit antiranibizumab polyclonal antibody. The complex formed by ranibizumab and these 2 molecules was captured onto a streptavidin-coated plate and quantified after sequential addition of horseradish peroxidase conjugate and a peroxidase substrate solution. The resulting colorimetric reaction was read on a plate reader. The lower limit of quantitation of this assay was 0.075 ng/ml, and concentrations lower than the lower limit of quantitation were considered less than reportable (LTR) for human serum samples. The assay used in the AMD population had a lower limit of quantitation of 0.3 ng/ml.15,19 The assay used in the present study measures both VEGF-bound and free ranibizumab. It was fully validated and shown to measure ranibizumab in human serum with acceptable accuracy, as well as interassay and intraassay precision: all measured values in accuracy experiments were within 10% of the spiked ranibizumab concentration and precision experiments resulted in coefficient of variation values of less than 10% at all ranibizumab concentration levels tested. Stability of
Zhang et al
Pharmacokinetics of Ranibizumab
Table 2. Descriptive Summary of Baseline Covariates Parameter
Retinal Vein Occlusion: Pooled BRAVO and CRUISE (n [ 441)
Median age (range), yrs Median CrCL (ml/min) Median BLVA (range), no. of letters Median BLA (range) Median BCFT (range), mm Median AONP (mm2) TTDX (day) BIOP (mmHg) IOP (mmHg) DIOP (mmHg)
67.0 (38e91) 81.1 (10.9e217.2) 53.0 (9e79) NA 587 (126e1651) 0 (0e13.22) 60 (0e1058) 15 (0e28) 17 (4e50) 3 (8 to 31)
Diabetic Macular Edema: Pooled RIDE and RISE (n [ 434) 62 91.5 58 9.2 449
(21e88) (10.6e375.4) (15e85) (0.0e16.0) (82e1241) NA NA NA NA NA
ANOP ¼ area of nonperfusion; BCFT ¼ baseline central foveal thickness; BIOP ¼ intraocular pressure before injection; BLA ¼ baseline leakage area; BLVA ¼ baseline visual acuity; BRAVO ¼ Ranibizumab for the Treatment of Macular Edema following Branch Retinal Vein Occlusion: Evaluation of Efficacy and Safety study; CrCL ¼ creatinine clearance; CRUISE ¼ Ranibizumab for the Treatment of Macular Edema following Central Retinal Vein Occlusion Study: Evaluation of Efficacy and Safety; DIOP ¼ difference between BIOP and IOP; IOP ¼ intraocular pressure (after injection); NA ¼ not applicable; RIDE and RISE ¼ A Study of Ranibizumab Injection in Subjects with Clinically Significant Macular Edema with Center Involvement Secondary to Diabetes Mellitus; RVO ¼ retinal vein occlusion; TTDX ¼ time to diagnosis. The total number of patients was 785. The type of RVO was the only categorical covariate tested, and there were no values missing.
ranibizumab in human serum also was evaluated, and it was shown to be stable for up to 4 freezeethaw cycles.
Pharmacokinetic Analysis
Ranibizumab Serum Concentration Data
The pharmacokinetic analysis comprised 3 stages: (1) assessing similarity in systemic exposure between RVO, DME, and AMD patients with a previously developed population-based pharmacokinetics model in AMD (the AMD population pharmacokinetics base model is shown in Fig 1)15; (2) development of pharmacokinetics
The pharmacokinetics sampling schedule can be found in Table 1. Serum concentration data were excluded from the analysis if they met 1 or more of the following criteria: measurable concentrations before the first dose, outliers (defined as concentration data with weighted residuals of more than 5 after fitting data to the base model using nonlinear mixed-effects modeling [NONMEM program versions 6 and 7; ICON Development Solutions, Ellicott City, Maryland]), data points that were found by inspection of visual predictive check plots to exhibit pharmacologically implausible spikes or drops in concentration (e.g., peak concentration at trough sampling time), and patients who did not have any measurable serum ranibizumab concentration after administration. Less-than-reportable samples after dose administration were included in the analysis for RVO. We used the maximum likelihood method20 for parameter estimation. A sensitivity analysis was performed to examine the impact of discarding LTR data20 on the base model estimation. No difference was found in parameter estimation between the maximum likelihood method and the sensitivity analysis performed to examine the impact of discarding LTR data. For the DME analysis, we used the sensitivity analysis performed to examine the impact of discarding LTR data to handle LTR samples, in which LTR observations were excluded from the analysis.
Covariates Included for the Analysis Patient characteristics measured at baseline were evaluated for covariate effects on the disposition of ranibizumab (Table 2). If covariates were missing in less than 20% of patients, continuous covariates were imputed for those patients at the median value. Categorical covariates were imputed for the most frequent category. If covariates were missing in more than 20% of patients, the missing covariates were not imputed and the covariate was excluded in covariate analysis.
Ranibizumab vitreous dA vit dt
Ka x Avit
Avit (0) = Dose Ka Systemic circulation dA sys dt
Ka x Avit
Asys Vc
CL Figure 1. Schematic representation of the 1-compartment model of the disposition of ranibizumab after intravitreal administration with first-order absorption into, and first-order elimination from, the systemic circulation. Asys ¼ amount of ranibizumab in the systemic compartment; Avit ¼ amount of ranibizumab in the vitreous compartment; CL ¼ clearance; Ka ¼ rate of systemic absorption/rate of vitreous elimination; Vc ¼ apparent volume of the central compartment.
2239
Ophthalmology Volume 121, Number 11, November 2014
Serum ranibizumab concentration (ng/ml)
A
0.5 mg/eye 5.00
0.50
LLOQ
0.05
0.01 0
10
20
30
40
Time (day)
B
0.5 mg/eye
Serum ranibizumab concentration (ng/ml)
5.00
1.00 0.50
0.10
LLOQ 0.05 0
10
20
30
Time after dose (day) Figure 2. Graphs showing simulated and observed ranibizumab serum concentration in (A) retinal vein occlusion patients and (B) diabetic macular edema patients using population pharmacokinetics model in age-related macular degeneration patients. LLOQ ¼ lower limit of quantitation.
Table 3. Population Parameter Estimates for the Retinal Vein Occlusion and Diabetic Macular Edema Final Model
Parameter
Final Age-Related Macular Degeneration Model19 (Percentage Relative Standard Error of the Estimates)
Retinal Vein Occlusion (Percentage Coefficient of Variation of the Estimate)
Diabetic Macular Edema (Percentage Coefficient of Variation of the Estimate)
CL/F* (l/day) Ka (1/day) Vc/F (l)
24.1 (4.52) 0.0806 (7.33) 3.01 (13.3)
28.5 (2.11) 0.106 (4.53) 2.54 (Fixed)
24.8* (31.3) 0.0969 (26.8) 2.77 (270)
CL ¼ clearance; CL/F ¼ apparent CL, CL divided by bioavailability (F); Ka ¼ rate of systemic absorption (rate of vitreous elimination); Vc/F ¼ apparent volume of the central compartment, or in short for a 1-compartmental model, volume of distribution. Table shows typical values of parameters and interpatient percentage coefficient of variation of the estimate. *The median creatinine clearance (CrCL) in the diabetic macular edema population was 91.5 ml/min; in this table, CL/F for the diabetic macular edema population was scaled to CrCL ¼ 65.2 ml/min, that is, the median CrCL in the age-related macular degeneration population.
2240
Zhang et al
Pharmacokinetics of Ranibizumab
Table 4. Individual Post Hoc Apparent Clearance (Clearance Divided by Bioavailability [F]) per Renal Function Groups22
Renal Function RVO study Normal Mild impairment Moderate impairment Severe impairment DME study Normal Mild impairment Moderate impairment Severe impairment
Estimated Creatinine Clearance (ml/min)
No. of Patients (% of Total)
Ranibizumab Apparent Clearance (Clearance Divided by Bioavailability (F)), Mean ± Standard Deviation (l/day)
>80 50e80 30e50 30
272 191 50 12
(51.8) (36.4) (9.5) (2.3)
31.610.8 20.55.20 15.93.45 9.362.69
>80 50e80 30e50 30
267 120 40 7
(61.5) (27.6) (9.2) (1.6)
36.810.2 25.26.4 19.34.6 12.42.8
DME ¼ diabetic macular edema; RVO ¼ retinal vein occlusion.
models for RVO and DME patients; and (3) validation of final models (with covariates added). We evaluated the adequacy of the AMD population pharmacokinetics model in describing data from RVO and DME patients by simulating the serum concentrations in RVO and DME patients with the AMD population pharmacokinetics model. For comparison, simulated serum concentrations were plotted against time from last dose and were overlaid with observed concentrations in RVO and DME, respectively. To develop the models for RVO and DME, the population pharmacokinetics analysis was performed using the nonlinear mixed-effects modeling program. Typical population values of pharmacokinetics parameters, as well as interpatient variability on those parameters, were estimated. Covariate model building was conducted using the step-wise reduction method based on the likelihood ratio test. The difference (d) in objective functions (asymptotically chi-square distributed) was used to compare alternative models. All model comparisons were performed at the P ¼ 0.01 level (d ¼ 7.88 for 1 degree of freedom). We evaluated the performance of the model in describing the observed data with standard diagnostics tools such as goodnessof-fit plots, where observed concentrations were compared with predicted concentrations. Model stability was assessed with the bootstrap approach, where parameter estimation was performed using randomly drawn data from the original data set. The estimation confidence intervals (CIs) were compared with the original point estimates. A visual predictive check was undertaken to evaluate the predictive performance of the final population pharmacokinetics model (with all covariates) for RVO and DME patients. We performed the simulations using final parameter estimates. Simulation and observations were plotted together against time from last dose for RVO and DME, respectively.
Results There were a total of 843 measurable ranibizumab serum concentrationetime records from 441 patients with RVO and 994 measurable concentrationetime records from 434 patients with DME (Table 1). Ranibizumab serum concentrations between days 3 and 30 after monthly 0.3-mg or 0.5-mg intravitreal dosing in patients with AMD, RVO, and DME were low relative to the concentration (IC50) necessary to inhibit the biological activity of VEGF-A by 50% IC50 (3 ng/ml).21 The number of patients with 1 or more sample with ranibizumab serum concentrations above the
IC50 of 3 ng/ml was 34 of 238 for AMD, 13 of 441 for RVO, and 6 of 434 for DME. All samples with ranibizumab concentrations of more than 3 ng/ml were collected at day 7 or earlier after ranibizumab administration. The AMD pharmacokinetics model accurately described the observed ranibizumab serum concentration data for both RVO and DME patients. The plot overlaying simulation and observation showed most observed RVO and DME data falling within the simulated 90% CI, indicating that the systemic ranibizumab concentrations were comparable among AMD, RVO, and DME patients (Fig 2). Final model parameters are shown in Table 3. Apparent serum clearance and rate of systemic absorption of ranibizumab in RVO and DME patients were consistent with those observed in AMD patients.15 No disease-related (RVO or DME) covariates, such as baseline visual acuity, baseline central foveal thickness, or area of nonperfusion, were identified. Some of the covariates were not included in the DME analysis because they had been found to be insignificant in the RVO analysis. Similar to AMD, the effect of creatinine clearance on apparent serum clearance was statistically significant for both RVO and DME. Individual serum ranibizumab clearance estimates by renal function groups22 are presented in Table 4 for both RVO and DME. Differences in creatinine clearance were estimated to account for 46% of the variability in apparent serum clearance of ranibizumab in RVO patients. A DME patient with creatinine clearance of 50 ml/minute had 30% lower total apparent serum clearance of ranibizumab compared with a DME patient with creatinine clearance of 90 ml/minute. Only approximately 10% of DME patients in this analysis had a creatinine clearance of 50 ml/minute or less. The fit of the model predictions to observed data for a few representative patients is illustrated in Figure 3. Figure 3A shows predicted and observed ranibizumab serum concentrations in RVO patients plotted against time, starting with the first dose (from day 0 to day 180). Figure 3B shows ranibizumab serum concentrations in DME patients plotted against time since most recent dose (from day 0 to day 30). Both plots illustrate that, on the individual level, predicted concentrations fit the observations well. During visual predictive check, most of the observed concentrations were within the simulated 90% CI using the RVO or DME final model, suggesting that the final covariate model accurately predicts ranibizumab serum concentration data (Fig 4). The final population pharmacokinetics model was fit repeatedly to 1000 datasets obtained by bootstrapping the original database. Median values of the bootstrapped parameters were consistent with the point estimates obtained from the final model (Table 5), and the 5% to 95% CIs were narrow.
2241
Ophthalmology Volume 121, Number 11, November 2014
5.00
DV PRED IPRED
Serum ranibizumab concentration (ng/ml)
Serum ranibizumab concentration (ng/ml)
A
0.50
DV PRED IPRED
0.50
0.01
0.01 50
0
100
150
50
0
100
Time (day)
5.00
DV PRED IPRED
Serum ranibizumab concentration (ng/ml)
Serum ranibizumab concentration (ng/ml)
5.00
0.50
0.01
DV PRED IPRED
0.50
0.01 0
50
100
0
150
50
100
Time (day)
B
DV PRED IPRED
5.00
Serum ranibizumab concentration (ng/ml)
Serum ranibizumab concentration (ng/ml)
150
Time (day)
DV PRED IPRED
5.00
0.50
0.05
0.50
0.05 0
10
20
30
0
40
10
Time after dose (days)
20
30
40
Time after dose (days)
DV PRED IPRED
DV PRED IPRED
5.00
Serum ranibizumab concentration (ng/ml)
5.00
Serum ranibizumab concentration (ng/ml)
150
Time (day)
0.50
0.05
0.50
0.05 0
10
20
30
Time after dose (days)
40
0
10
20
30
40
Time after dose (days)
Figure 3. Goodness-of-fit plots for the final population pharmacokinetic model showing individual predicted (IPRED) results (ng/ml [black dashed line]), population predicted (PRED) results (ng/ml [solid line]) overlaid on observed concentrations (DV) as a function of time after dose (days) (horizontal gray dashed line is the less than reportable value of 0.075 ng/ml) for (A) retinal vein occlusion and (B) diabetic macular edema.
2242
A
Serum ranibizumab concentration (ng/ml)
Zhang et al
Pharmacokinetics of Ranibizumab
5.00
0.50
LLOQ
0.05 0.01 0
10
20
30
40
Time after dose (day)
Serum ranibizumab concentration (ng/ml)
B
5.00
1.00 0.50
0.10
LLOQ 0.05 0
10
20
30
Time after dose (day) Figure 4. Visual predictive check and goodness-of-fit plots of (A) retinal vein occlusion and (B) diabetic macular edema final covariate models. LLOQ ¼ lower limit of quantitation.
Discussion Results from the current population pharmacokinetics analysis demonstrate that the systemic exposure and pharmacokinetics of ranibizumab are comparable in RVO, DME, and AMD patients. The pharmacokinetics parameter estimates were comparable among the 3 indications with moderate numerical difference. Estimation of apparent serum clearance of ranibizumab was slightly lower in the AMD population compared with the RVO or DME populations. This was because the AMD population was older and had reduced renal function on average compared with the RVO and DME populations. The difference in estimations of systemic absorption of ranibizumab between AMD and the other 2 indications was likely the result of differences in sampling time. The AMD pharmacokinetics data
set included samples obtained as early as 1 hour and 1 day after dose administration, whereas the RVO and DME data sets did not include these early time points, which are most informative for estimating systemic absorption. Therefore, the numerical difference in estimates of systemic absorption of ranibizumab is not likely to reflect any meaningful differences in the rate of exit of ranibizumab from eyes of patients with AMD, RVO, or DME. The very similar pharmacokinetics profiles seen in patients with AMD, DME, or RVO suggest that there are no disease-related differences in the disposition of ranibizumab after intravitreal injection. Similar to the situation in AMD patients, systemic exposure of ranibizumab was low in patients with RVO or DME. This explains the generally favorable systemic safety profile seen in patients receiving monthly intravitreal injections of ranibizumab in several phase 3 clinical trials. It
2243
Ophthalmology Volume 121, Number 11, November 2014 Table 5. Bootstrap of Parameter Estimation for Retinal Vein Occlusion and Diabetic Macular Edema Bootstrap Final Model, Median (5%e95% Confidence Interval) Parameter Typical CL/F (l/day) Typical Ka (/day) CrCL on CL/F uCL/F (%) uKa (%) spop (%)
Retinal Vein Occlusion 28.4 (24.8e29.5) 0.107 0.624 29.4 39.5 24.6
(0.0989e0.117) (0.495e0.747) (25.5e40.2) (29.1e58.1) (17.6e28.3)
Diabetic Macular Edema 30.4 (29.3e31.6) 0.097 0.597 30.8 26.5 28.2
(0.093e0.101) (0.516e0.664) (27.3e34.0) (21.9e29.9) (26.0e30.7)
CL/F ¼ apparent clearance, clearance divided by bioavailability (F); CrCL ¼ creatinine clearance; Ka ¼ rate of systemic absorption (rate of vitreous elimination); uCL/F ¼ intersubject variability as percent coefficient of variation of the estimate of apparent clearance, clearance divided by bioavailability (F); uKa ¼ intersubject variability as percent coefficient of variation of the estimate of rate of systemic absorption (rate of vitreous elimination); spop ¼ standard deviation describing the proportional component of residual variability.
should be noted that the molecular characteristics of ranibizumab (high affinity for VEGF obtained by affinity maturation, extremely slow dissociation rate when in complex with VEGF, small size, and absence of an Fc region) were designed specifically to combine potent inhibition of VEGF in the intraocular compartment with minimal VEGF inhibition in extraocular blood vessels and tissues because of its rapid elimination from the systemic circulation.7,23,24 This is an important pharmacologic property of ranibizumab because inhibition of systemic VEGF may impair normal angiogenesis and interfere with the integrity of the endothelial and epithelial surfaces, vascular tone, and the filtration function of the kidneys. This can cause systemic VEGF inhibition class side effects such as hypertension, proteinuria, arterial thromboembolic events, gastrointestinal perforation, and wound-healing complications.25 Because full-length antibodies have an Fc fragment that increases residence time in circulation, it is expected that bevacizumab and the Fc-containing aflibercept have greater systemic exposure compared with ranibizumab, and this was demonstrated to be the case (Avery R, Castellarin AA, Steinle NC, et al. Comparison of systemic pharmacokinetics following anti-VEGF intravitreal injections of ranibizumab, bevacizumab, and aflibercept. Paper presented at: AAO Annual Meeting; November 2013; Las Vegas, Nevada). Consistent with this observation, the alternative treatments to that used in the Inhibit VEGF in Age-Related Choroidal Neovascularization clinical trial showed that patients with neovascular AMD treated with intraocular injections of bevacizumab had a significant reduction in serum VEGF levels compared with patients treated with ranibizumab.26 These data were replicated in another clinical trial.27 Thus, on theoretical grounds, it is expected that Fc-containing VEGF antagonists may carry a higher risk of systemic adverse events than ranibizumab, but the rates of such side effects are very low, and efficacy studies are not powered to identify differences in relatively rare events. However, such
2244
differences might exist because in the Comparison of AgeRelated Macular Degeneration Treatments Trials, patients with neovascular AMD treated with intraocular injections of bevacizumab had a significantly greater number of systemic serious adverse events at the 1- and 2-year time points than seen in patients treated with ranibizumab.13,28 The authors of the study discussed whether this difference truly indicates increased bevacizumab risk or was simply the result of chance and suggested that because they did not understand the mechanism of many of the serious adverse events, the possibility that the difference is the result of chance may be favored. However, it may not be prudent to attribute the difference to chance simply because the serious adverse events do not conform to expectations regarding mechanisms of blocking VEGF. Similar to the situation in AMD patients, creatinine clearance was identified to be a statistically significant covariate of apparent serum clearance of ranibizumab in RVO and DME patients. It is unlikely that reduced renal function would cause a clinically meaningful decrease in apparent serum clearance of ranibizumab and a consequent increase in systemic exposure that would affect the efficacy or safety of ranibizumab. Efficacy should not be influenced by systemic exposure because ocular compartments are the sites of action for ranibizumab and renal function does not affect the elimination of ranibizumab from the vitreous. Safety is unlikely to be influenced because systemic exposure in RVO and DME patients was low and not expected to be pharmacologically active even in patients with mild to moderate renal impairment. The severe renal impairment in approximately 2% of the RVO and DME patients was not found to be associated with significantly increased rates of VEGF-related serious adverse events. Moreover, creatinine clearance had no statistically or clinically meaningful effect on vitreous elimination of ranibizumab quantified by the absorption rate constant (Ka). Hence, renal impairment is not expected to affect vitreous exposure to ranibizumab. No other disease-related factors or baseline patient characteristics were found to have significant effects on ranibizumab serum pharmacokinetics. This suggests that differences in disease states among patients with AMD, RVO, or DME do not contribute to the variability in pharmacokinetics. In conclusion, results from this population pharmacokinetics analysis demonstrate that serum levels of ranibizumab after monthly intravitreal injections are similar in RVO or DME patients compared with those seen in patients with AMD. This suggests that regardless of whether a patient is being treated for AMD, RVO or DME, intravitreal injection of the same amount of ranibizumab is likely to lead to similar intraocular concentrations, similar egress from the eye into systemic circulation, and ultimately similar serum concentrations. In patients with AMD, RVO, or DME, systemic exposure is comparable. The low systemic exposure to ranibizumab after intraocular injections in all patient populations studied is reassuring and likely to account for the low rates of systemic complications seen in phase 3 ranibizumab treatment trials regardless of treatment indication.
Zhang et al
Pharmacokinetics of Ranibizumab
Acknowledgments. Support for third-party writing and formatting assistance for this manuscript by Rebecca Jarvis, PhD, CMPP, was provided by Genentech, Inc.
References 1. Ferrara N, Gerber HP, LeCouter J. The biology of VEGF and its receptors. Nat Med 2003;9:669–76. 2. Tolentino MJ, Miller JW, Gragoudas ES, et al. Intravitreous injections of vascular endothelial growth factor produce retinal ischemia and microangiopathy in an adult primate. Ophthalmology 1996;103:1820–8. 3. Ozaki H, Hayashi H, Vinores SA, et al. Intravitreal sustained release of VEGF causes retinal neovascularization in rabbits and breakdown of the blood-retinal barrier in rabbits and primates. Exp Eye Res 1997;64:505–17. 4. Okamoto N, Tobe T, Hackett SF, et al. Transgenic mice with increased expression of vascular endothelial growth factor in the retina: a new model of intraretinal and subretinal neovascularization. Am J Pathol 1997;151:281–91. 5. Aiello LP, Pierce EA, Foley ED, et al. Suppression of retinal neovascularization in vivo by inhibition of vascular endothelial growth factor (VEGF) using soluble VEGF-receptor chimeric proteins. Proc Natl Acad Sci U S A 1995;92:10457–61. 6. Adamis AP, Shima DT, Tolentino MJ, et al. Inhibition of vascular endothelial growth factor prevents retinal ischemiaassociated iris neovascularization in a nonhuman primate. Arch Ophthalmol 1996;114:66–71. 7. Ferrara N, Damico L, Shams N, et al. Development of ranibizumab, an anti-vascular endothelial growth factor antigen binding fragment, as therapy for neovascular age-related macular degeneration. Retina 2006;26:859–70. 8. Brown DM, Kaiser PK, Michels M, et al; ANCHOR Study Group. Ranibizumab versus verteporfin for neovascular agerelated macular degeneration. N Engl J Med 2006;355:1432–44. 9. Brown DM, Campochiaro PA, Singh RP, et al; CRUISE Investigators. Ranibizumab for macular edema following central retinal vein occlusion: six-month primary end point results of a phase III study. Ophthalmology 2010;117:1124–33. 10. Campochiaro PA, Heier JS, Feiner L, et al; BRAVO Investigators. Ranibizumab for macular edema following branch retinal vein occlusion: six-month primary endpoint results of a phase III study. Ophthalmology 2010;117:1102–12. 11. Diabetic Retinopathy Clinical Research Network, Elman MJ, Aiello LP, Beck RW, et al. Randomized trial evaluating ranibzumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology 2010;117:1064–77. 12. Nguyen QD, Brown DM, Marcus DM, et al; RISE and RIDE Research Group. Ranibizumab for diabetic macular edema: results from 2 phase III randomized trials: RISE and RIDE. Ophthalmology 2012;119:789–801. 13. CATT Research Group. Ranibizumab and bevacizumab for neovascular age-related macular degeneration. N Engl J Med 2011;364:1897–908. 14. Ip MS, Domalpally A, Hopkins JJ, et al. Long-term effects of ranibizumab on diabetic retinopathy severity and progression. Arch Ophthalmol 2012;130:1145–52.
15. Xu L, Lu T, Tuomi L, et al. Pharmacokinetics of ranibizumab in patients with neovascular age-related macular degeneration: a population approach. Invest Ophthalmol Vis Sci 2013;54: 1616–24. 16. Krohne TU, Liu Z, Holz FG, Meyer CH. Intraocular pharmacokinetics of ranibizumab following a single intravitreal injection in humans. Am J Ophthalmol 2012;154: 682–6. 17. Brown DM, Campochiaro PA, Bhisitkul RB, et al. Sustained benefits from ranibizumab for macular edema following branch retinal vein occlusion: 12-month outcomes of a phase III study. Ophthalmology 2011;118:1594–602. 18. Campochiaro PA, Brown DM, Awh CC, et al. Sustained benefits from ranibizumab for macular edema following central retinal vein occlusion: twelve-month outcomes of a phase III study. Ophthalmology 2011;118:2041–9. 19. Lowe J, Maia M, Wakshull E, et al. Development of a novel homogenous electrochemiluminescence assay for quantitation of ranibizumab in human serum. J Pharm Biomed Anal 2010;52:680–6. 20. Beal SL. Ways to fit a PK model with some data below the quantification limit. J Pharmacokinet Pharmacodyn 2001;28: 481–504. 21. Yu L, Liang XH, Ferrara N. Comparing protein VEGF inhibitors: in vitro biological studies. Biochem Biophys Res Comm 2011;408:276–81. 22. Center for Drug Evaluation and Research. Guidance for Industry. Guidance for industry. Pharmacokinetics in patients with impaired renal functiondstudy design, data analysis, and impact on dosing and labeling. Rockville, MD: Center for Biologics Evaluation and Research; May 1998. Available at: http://www.fda.gov/downloads/Drugs/GuidanceComplianceR egulatory%20Information/Guidances/UCM072127.pdf. Accessed on April 2, 2014. 23. Chen Y, Wiesmann C, Fuh G, et al. Selection and analysis of an optimized anti-VEGF antibody: crystal structure of an affinity-matured Fab in complex with antigen. J Mol Biol 1999;293:865–81. 24. Lowe J, Araujo J, Yang J, et al. Ranibizumab inhibits multiple forms of biologically active vascular endothelial growth factor in vitro and in vivo. Exp Eye Res 2007;85:425–30. 25. Chen HX, Cleck JN. Adverse effects of anticancer agents that target the VEGF pathway. Nat Rev Clin Oncol 2009;6: 465–77. 26. IVAN Study Investigators, Chakravarthy U, Harding SP, Rogers CA, et al. Ranibizumab versus bevacizumab to treat neovascular age-related macular degeneration: one-year findings from the IVAN randomized trial. Ophthalmology 2012;119:1399–411. 27. Zehetner C, Kirchmair R, Huber S, et al. Plasma levels of vascular endothelial growth factor before and after intravitreal injection of bevacizumab, ranibizumab and pegaptanib in patients with age-related macular degeneration, and in patients with diabetic macular oedema. Br J Ophthalmol 2013;97: 454–9. 28. Comparison of Age-Related Macular Degeneration Treatments Trials (CATT) Research Group, Martin DF, Maguire MG, Fine SL, et al. Ranibizumab and bevacizumab for treatment of neovascular age-related macular degeneration: two-year results. Ophthalmology 2012;119:1388–98.
2245
Ophthalmology Volume 121, Number 11, November 2014
Footnotes and Financial Disclosures Originally received: October 16, 2013. Final revision: February 3, 2014. Accepted: May 13, 2014. Available online: July 4, 2014.
Manuscript no. 2013-1746.
1
Genentech, Inc., South San Francisco, California.
2
Quantitative Solutions, Menlo Park, California.
3
The Wilmer Eye Institute, Johns Hopkins School of Medicine, Baltimore, Maryland. Presented at: Association for Research in Vision and Ophthalmology Annual Meeting, May 2011 and May 2012, Fort Lauderdale, Florida. Financial Disclosure(s): The author(s) have made the following disclosure(s): Yi Zhang: Employee e Genentech, Inc. (South San Francisco, California). Zhenling Yao: Employee e Genentech, Inc. (South San Francisco, California). Nitin Kaila: Consultant e Genentech, Inc. (South San Francisco, California); Employee e Quantitative Solutions. Peter Kuebler: Employee e Genentech, Inc. (South San Francisco, California). Jennifer Visich: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche. Mauricio Maia: Employee e Genentech, Inc. (South San Francisco, California). Lisa Tuomi: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche.
2246
Jason S. Ehrlich: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche. Roman G. Rubio: Employee e Genentech, Inc. (South San Francisco, California); Equity owner e Roche. Peter A. Campochiaro: Consultant e Genentech, Inc. (South San Francisco, California), GlaxoSmithKline, Regeneron, Aerpio Therapeutics (employer, Johns Hopkins University, receives compensation), and Elan, Gene Signal, Norvorx (receives personal compensation); Data and Safety Monitoring Committee e Advanced Cell Technology, Regeneron; Financial support e Genentech, GlaxoSmithKline, Genzyme, Oxford BioMedica; Equity owner e Graybug, Inc. Genentech, Inc., South San Francisco, California, provided support for the study; participated in study design and conduct of the study; provided data collection, management, and interpretation; as well as reviewed and approved the manuscript. Abbreviations and Acronyms: AMD ¼ age-related macular degeneration; CI ¼ confidence interval; DME ¼ diabetic macular edema; LTR ¼ less than reportable; RVO ¼ retinal vein occlusion; VEGF ¼ vascular endothelial growth factor. Correspondence: Peter A. Campochiaro, MD, The Wilmer Eye Institute, Johns Hopkins Hospital School of Medicine, 719 Maumenee, 600 North Wolfe St, Baltimore, MD 21287-9277. E-mail: [email protected].
