Photokinetic Drug Delivery: Near infrared (NIR) Induced Permeation Enhancement of Bevacizumab, Ranibizumab and Aflibercept through Human Sclera
Photokinetic Drug Delivery: Near infrared (NIR) Induced Pe r m e a t i o n E n h a n c e m e n t o f B e v a c i z u m a b , Ra n i b i z u m a b and Aflibercept through Human Sclera
Steven A. Giannos 0
Edward R. Kraft 0
Zhen-Yang Zhao 0
Kevin H. Merkley 0
Jiyang Cai 0
0 Department of Ophthalmology and Visual Sciences, University of Texas Medical Branch , 301 University Blvd, Galveston, Texas 77555 , USA
1 Steven A. Giannos
Purpose Permeation studies, with near infrared (NIR) light and anti-aggregation antibody formulation, were used to investigate the in vitro permeation of bevacizumab, ranibizumab and aflibercept through human sclera. Methods A vertical, spherical Franz cell diffusion apparatus was used for this scleral tissue permeation model. A photokinetic ocular drug delivery (PODD) testing device accommodated the placement of NIR LEDs above the donor chambers. An adjustable LED driver/square wave generator provided electrical energy with a variable pulse rate and pulse width modulation (duty cycle). Results Exposure to non-thermal NIR light had no effect on mAbs with regard to monomer concentration or antibody binding potential, as determined by SE-HPLC and ELISA. The optimal LED wavelength was found to be 950 nm. Duty cycle power of 5% vs 20% showed no difference in permeation. When compared to controls, the combination of nonaggregating antibody formulation and NIR illumination provided an average transscleral drug flux enhancement factor of 3X. Conclusion Narrow wavelength incoherent (non-laser) light from an NIR LED source is not harmful to mAbs and can be used to enhance drug permeation through scleral tissue. The topical formulation, combined with pulsed NIR light irradiation, significantly improved scleral permeation of three anti-VEGF antibody drugs.
aflibercept; bevacizumab; permeation; photokinetic; ranibizumab; sclera
The number of people visually impaired in the world is
estimated to be 285 million, 39 million blind and 246 million
having low vision; 65% of people visually impaired and 82%
of all blind are 50 years and older. (
) Age-related macular
degeneration (AMD) is a progressive, degenerative disease of
the retina that occurs with increasing incidence with age and
ranks third among the global causes of visual impairment. (
Exudative AMD is caused by new, abnormal blood vessel
growth (neovascularization) in the subretinal layers, leading
to vascular leaks, bleeding, and progressive vision loss. (
Vascular endothelial growth factor (VEGF) is a signaling
protein produced by macrophages, retinal pigment epithelium
and Muller cells that stimulate vasculogenesis and
angiogenesis. Overexpression of VEGF has been implicated
in the development and progression of neovascular AMD. (
Drugs such as the monoclonal antibodies, bevacizumab,
ranibizumab and the fusion protein, aflibercept can inhibit
VEGF and control or slow those diseases. (
The commercial development of therapeutic monoclonal
antibodies commenced in the early 1980’s, and by 1986 the
first therapeutic monoclonal antibody (mAb), Orthoclone
OKT3, was approved for the prevention of kidney transplant
) As of 2015, the highly dynamic late-stage
commercial pipeline of recombinant therapeutics now includes
nearly 50 molecules. (
) The majority of approved antibody
drugs are used to treat cancer and inflammation. However,
two of these monoclonal antibodies, bevacizumab (Avastin®)
and FDA approved ranibizumab (Lucentis®), show
antiVEGF properties and may be used to treat age related
macular degeneration (AMD) and diabetic retinopathy. (
VEGF-Trap (Aflibercept, Intravitreal aflibercept injection
(IAI Eylea®)) is a soluble VEGF decoy receptor that consists of
the second immunoglobulin (Ig)-like domain of FLT1 and the
third Ig-like domain of KDR (kinase insert domain receptor)
linked to the IgG constant region (Fc). (
) Eylea®, the FDA
approved formulation of aflibercept for the treatment of wet
AMD, is administered as an intravitreal injection. (
Systemic drug administration for ophthalmic disease is
difficult because of poor drug permeability, due to a
bloodretinal barrier, encountered when targeting the posterior
) The retina has a unique position with regard to
pharmacokinetics in that the blood-retinal barrier (BRB)
separates the retina from the circulating blood. The BRB, which
forms complex tight junctions of retinal capillary endothelial
cells (inner BRB) and retinal pigment epithelial cells (outer
BRB), restricts nonspecific transport between the neural retina
and the circulating blood. (14)
Anti-VEGF drugs are delivered by repeated intravitreal
), which can bring about serious complications
including retinal detachment and infection, as well as being
painful and costly. (
) Alternative approaches such as
), sonophoresis (
) and photokinetic drug delivery
) have been used, non-invasively, to deliver drugs to the
back of the eye.
Photokinetic drug delivery, for transdermal and ocular
applications, is a new addition to noninvasive drug delivery
) Photokinetic facilitated drug delivery uses
selected pulsed illumination with a selected light wavelength,
directed onto drug molecules residing on a tissue surface.
The selected cyclic illumination causes unidirectional
translocation of the drug molecules from the tissue surface into the
tissue. Godley et al. recently described their early work with
photokinetics and the development of a device to deliver light
energy, especially in the application of transscleral drug
delivery, to the posterior segment of the eye. (
) This strategy
was applied and investigated as a simplified model of
scleralvitreous interface and unstirred gel mimicking the vitreous.
The focus of this research is to show the feasibility of
transscleral mAbs permeation, as well as enhanced permeation due
to NIR light irradiation. While developing and validating
methods for the photokinetic enhancement of transscleral
mAb permeation, using an in vitro Franz diffusion model, we
encountered the problem of mAb dilute solution instability in
phosphate buffered saline (PBS). Published tissue permeation
(Franz) cell studies typically use PBS as the recipient media at
37°C. Franz cell studies start with a recipient media
concentration of 0 of the test compound. We have identified a
phenomenon wherein low concentrations of antibodies
(<1150 ng/ml), in PBS or the manufacturer’s formula, lose
As the permeation experiment progresses; the
concentration of the test compound increases. However, at early points
in the tissue permeation experiment, the drug is still a very
dilute solution (i.e., at 2 h, it could be anywhere between 0 and
5000 ng/ml). For our experimental work on anti-VEGF
mAbs, we found that these conditions contributed to mAb
aggregation, causing reduced monomer concentrations and
decreased VEGF binding capacity.
Scleral drug permeation of large entities is generally
molecular weight and molecular size limited. (
aggregation, combining two or more individual antibodies,
would have a significant permeation rate limiting effect.
Doubling or tripling of a topically applied drug weight and
size by aggregation would render these large drug structures
impossible to permeate sclera. Antibody aggregation within
the topically applied drug composition had to be eliminated
before any assessment of passive and facilitated permeation
could be evaluated.
We have developed and used a novel, stabilized
formulation that prevents the antibodies from aggregating,
which is suitable for topical ocular delivery. (
) In vitro
studies demonstrate that this formulation, with concurrent
non-thermal NIR light irradiation, delivers clinically relevant
drug amounts, noninvasively, through scleral tissue using a
MATERIALS AND METHODS
L-Argininine, sodium phosphate dibasic anhydrous, sodium
phosphate monobasic monohydrate, sodium sulfate
anhydrous, phosphate buffered saline (10X), and HPLC water
were obtained from Fisher Scientific Fair Lawn, NJ.
α,α-trehalose dihydrate, polysorbate 80 (Tween® 80, low peroxide),
and sodium chloride, USP was purchased from
SigmaAldrich, St Louis Mo and normal saline 0.9% obtained from
Baxter Healthcare Corp., Deerfield, IL.
ELISA analytical kits for bevacizumab (kit
#AVA-EU51) and Ranibizumab (kit #LUC-E-U52) were obtained
from United Immunoassay Inc., San Bruno, CA USA.
Aflibercept was analyzed using an ELISA procedure as
described Celik et al. (
Antibodies bevacizumab (Avastin®) 25 mg/ml and
ranibizumab (Lucentis®) 10 mg/ml were obtained from
Genentech, South San Francisco CA. Intravitreal aflibercept
injection (IAI Eylea®) 48.2 mg/ml was obtained from
Regeneron, Tarrytown NY. (Note: bulk aflibercept as
provided by Regeneron is at 48.2 mg/ml while the common
pharmaceutical preparation Eylea® is formulated at 40.0 mg/ml).
All other reagents were of analytical or USP purity. Whole
globe human donor eye pairs, designated for research, were
obtained from Lone Star Lions Eye Bank, Manor, TX, from
Lions Eye Bank of Texas/ Baylor College of Medicine,
Houston Texas and from TBI Orlando/ Medical Eye Bank
Light Device and Experimental Setup
A vertical, spherical Franz cell diffusion apparatus was
adapted and used for the sclera tissue permeation model.
Spherical Franz cells (PermeGear, Inc., Hellertown, PA)
having a 9.1 mm diameter (0.65 cm2 diffusional area) were used.
A photokinetic ocular drug delivery (PODD)-modified Franz
cell testing device was configured so that it accommodated the
placement of NIR LEDs within the donor chamber. The
LEDs are placed 1 cm above the surface of the scleral
membrane. The cells were placed within an aluminum block on a
magnetic stir bar setup (manufactured by PermeGear, Inc.,
Hellertown, PA), with a custom built 37°C water bath to
provide even temperature control within the aluminum block, the
water bath and the permeation cells. The experimental
arrangement is shown in Fig. 1.
Spherical Franz cells were adapted for sclera photokinetic
permeation studies. A donor cell, containing bevacizumab,
ranibizumab or aflibercept in Formula 14 (F14, described
Fig. 1 Schematic of Franz cell diffusion apparatus.
below) was on top of the scleral membrane and receiver cell.
The receiver cell was filled with F14 as well. Samples for
chemical analysis were taken from the side arm of the Franz
apparatus with volume replacement. Control cells were set up
the same, but without the NIR LED. NIR LEDs were driven
by a custom built square wave pulse generator.
An adjustable square wave pulse generator provided
pulsed electrical energy from 400 to 1000 cycles per second
(CPS) with a variable pulse width modulation duty cycle
circuit. This circuit is adjustable from 0.1% to 99% percent ON
pulse duration (e.g., ON 50%, OFF 50% of the time, 50%
duty cycle; or ON 10%, OFF 90%, 10% duty cycle; or ON
20%, OFF 80%, 20% duty cycle and so on). The electrical
current for the driver of the LEDs was adjusted to about 5 to 8
times above the continuous wave current level recommended
for the LED drive current in order to provide the desired
equivalent rated optical power output with short duration
pulsed drive current. In general, the short duration electrical
pulse (i.e. 10–20% duty cycle) with the over-rated drive
current, provided a minor (≤ 1°C) donor compartment
temperature increase above ambient conditions. Combination
adjustments in pulse widths in conjunction with electrical drive
currents were used to produce the required optical power
without excess or extraneous LED device heat or heat from
absorbed light energy. The temperature of the drug donor
compartment was ≤1°C above ambient Franz cell apparatus
Drug Permeation Experiments
Bevacizumab, ranibizumab or aflibercept was placed in
Formulation 14 (F14), which contains 100 mM sodium
phosphate buffer, 0.3% NaCl, 7.5% trehalose, 10 mM arginine
and 0.04% Tween 80 at a pH of 6.78. (
) The composition
was used as a diluent for the drug donor solution and also used
as the drug recipient media in the permeation cell studies.
Aflibercept (as provided), ranibizumab and bevacizumab
had different packaged concentrations (43.8, 10 and 25 mg/
ml respectively), and were diluted to 2.5 mg/ml. A total
standardized dosage of 1 mg was provided by placing 0.4 ml of the
2.5 mg/ml solution in the Franz cell. The antibody solution
was left in place for one hour then removed from the donor
Whole globe human donor eye pairs, designated for research,
were obtained from Lone Star Lions Eye Bank, Manor, TX,
from Lions Eye Bank of Texas/ Baylor College of Medicine,
Houston Texas and from TBI Orlando/ Medical Eye Bank of
Florida. Eye pairs were positioned in moist chambers with
transport cages and shipped on wet ice or otherwise stored at
4°C. The average time of death to time of permeation study
start was 84.48 h, ranging from 39 to 171 h. The eyes were
dissected into 4 sections from anterior to posterior along the
axis of the muscle insertion points to assure that blood vessel
penetrations would not be within the permeation area and be
positioned under the Franz cell flange. Anterior tissue sections
from limbus to equator were mounted onto 0.65cm2
hemispheric Franz cells and clamped into place. Cells were checked
to make sure that there was no leakage through the tissue
between the donor and receiver chambers and also no leakage
from the donor or receiver chambers to the exterior of the
Franz cell. Each group of Franz cells, the 4 sections from each
eye, were randomly designated to either a control or a
photokinetic group; i.e. right eye verses left eye. Results were
multiplied by 1.54 to provide amount of drug permeated per
1cm2 for convenience and comparison, as is generally
After temperature equilibration, the donor chamber was
emptied and dried. The test drug in formula 14, 400ul of 2.5 mg/
ml (1 mg total), was added into the donor chambers. The
control group was held in the dark. The photokinetic group
had a selected NIR LED positioned 1 cm above the tissue
surface. Experimental time started when the drug solution
was placed in the donor chamber and NIR LED turned ON
together. The drug solution, with the concurrent NIR
irradiation, was allowed to sit for 60 min. After 60 min, the drug
solution was removed from the donor chamber and the NIR
LED was turned OFF. In our studies, we did not vary the
length of time for the application of the antibody solution or
the length of time of NIR irradiation. These parameters were
held constant at 60 min.
Serial 400ul samples were taken from the center of the
recipient chamber with a 3″ x 20G needle and syringe at the
selected time points. Sample volume was replaced with
degassed F14. Samples were centrifuged at 1090 relative
centrifugal force (RCF) for 5 min, at 23°C, to precipitate any
Enzyme-Linked Immunosorbent Assay (ELISA)
Bevacizumab and ranibizumab ELISA assays were performed
as per manufacturer’s instructions except for a substitution of
the base analytical standard material which was taken from
the pharmaceutical preparations obtained and diluted as
described below. Aflibercept ELISA was performed as described
by Celik et al. (
), except for a substitution of the base
analytical standard material which was taken f rom the
pharmaceutical preparations obtained and diluted as
described below. The ELISA method dynamic ranges are as
follows: bevacizumab 1–281.25 ng/ml, ranibizumab
1125 ng/ml and aflibercept 1-125 ng/ml.
Size-Exclusion High Performance Liquid
Analytical size-exclusion chromatography was performed
using an Agilent HPLC system HP1100 from Agilent
Technologies (Santa Clara, CA) with a UV detector. Studies
were performed using a TSKgel UltraSW Aggregate
7.8 mm × 30 cm, 3 μm SEC column with TSKgel UltraSW
guard column (Tosoh Bioscience LLC, King of Prussia, PA).
Mobile phase comprising 85% 100 mM sodium sulfate in
100 mM phosphate buffer in HPLC water (adjusted to
pH 6.68) with 15% acetonitrile/ 0.1% trifluoroacetic acid
was used at a flow rate of 0.6 ml/min. Sample injections were
100 μl in volume. The eluted protein was monitored by UV
Absorbance at 212 nm. Silanized HPLC sample vials and
silanized vial inserts from Agilent Technologies (Santa Clara,
CA), were used throughout.
Data is presented as mean ± Standard Deviation (SD). Group
to group analysis was conducted with student’s two-tailed
ttest using Microsoft Excel. In all cases, a p value ≤0.05 is
considered to be significant.
In Vitro Studies
Bevacizumab/NIR Light Stability Study
It is well known that ultraviolet light may degrade an
antibody. However, we were interested in finding if non-thermal,
non-ionizing infrared light had any detrimental effects on
antibodies. In this case a volume of 25 mg/ml bevacizumab was
diluted with F14 to a concentration of 500 ng/ml. This
material was used as a starting composition and separated into two
groups; control and light exposure, n = 4/group. Plastic
chambers (3 ml) were fitted with tight lids, housing 5 mm
950 nm NIR LEDs with light directed toward the bottom of
the chambers. Each chamber received 600 μl of the 500 ng/
ml bevacizumab in F14. LEDs were positioned 3 cm above
the surface of the antibody solution. The chambers were
placed in a 37°C water bath. In the light control group, the
LED was turned ON providing 24.5 ± 2.5 mW power,
fluency of 3.15 ± 0.15 W/cm2 pulsed at 1000 cycles per second
(Hz) with a duty cycle of 20% (20% ON time) measured at
the LED surface. At 3 cm distance from the LED to the
surface of the solution, the power was 12 mW ± 1.0 mW and a
fluency of 1.25 W ± 0.25. The non-light control group was
shielded from light exposure. Samples were taken at 1 h and
5 h and examined by ELISA and SE-HPLC.
A standard dilution range of 1125 to 4.3945 ng/ml starting
with 25 mg/ml bevacizumab in F14 was made and divided for
both analytical methods. ELISA standard dilutions and test
samples were run on the same plate at the same time.
SEHPLC standard dilution and test samples were run
consecutively on the same method setup. Standard dilution curves
were derived and used to determine the concentration of the
subject test samples for each method.
Photokinetic/Bevacizumab Antibody Permeation
Different IR Wavelengths
In this study, 5 different NIR wavelengths were tested, with
bevacizumab in F14, to identify the best condition for
photokinetic drug delivery. Human sclera was used as the test
membrane. The drug donor was 400ul of 2.5 mg/ml bevacizumab
in each Franz cell, which was removed at one hour and the
LED was turned off. Samples were taken at 5 and 8 h. The
conditions tested were: 830 nm/1000cps/5% duty cycle (ON
time 5% vs OFF time 95%), 950 nm/1000cps/5% duty cycle,
1300 nm/1000cps/5% duty cycle, 1450 nm/1000cps/5%
duty cycle, 3400 nm/400cps/2.5% duty cycle. There was no
heat generated from the LEDs at these duty cycles. All samples
were analyzed by ELISA.
ELISA vs. SE-HPLC
An in vitro study, using ranibizumab, was run to test the
correlation of ELISA results to SE-HPLC analysis results using F14
and NIR light. In this study, the NIR irradiation was applied
for 1 h; dark control vs 950 nm/1000cps/20% duty cycle
(27mw/cm2). The controls were shielded from light. For this
experiment, a single eye donor, with 4 sclera sections per eye,
was used. Right eye vs. left eye was tested. The drug donor was
400ul of 2.5 mg/ml bevacizumab in each Franz cell and was
removed at one hour and LED was turned off. Samples of
400ul were taken at 2, 3, 5, 8, 10, 12 and 24 h.
SE-HPLC was used to accurately quantitate the
monomeric antibody concentration over a wide dynamic range.
SEHPLC, however, is not an indicator of biological activity.
ELISA, which is an additional method to determine
concentration, has a generally limited dynamic range. ELISA is not a
robust quantitative method, but it clearly demonstrates the
biological activity of the antibody. Together, quantification
was determined and correlated using physico-chemical and
biological activity methods.
The samples were split in two for SE-HPLC and ELISA
analysis. SE-HPLC analysis was started immediately. Derived
SE-HPLC values determined the dilution strategy of samples
for ELISA to bring into the usable range of the method. The
ELISA range was 1-200 ng/ml. The SE-HPLC range was
4.4–18,000 ng/ml. ELISA samples were diluted 2X -25X as
determined by SE-HPLC results. ELISA was run for the 3, 5,
8, 12 & 24 h time points (samples from 2 h and 10 h time
points were omitted due to limited space on the ELISA test
Formula 14/Bevacizumab Sclera Permeation Feasibility
An initial feasibility in vitro human sclera permeation study was
performed, with bevacizumab, to replicate F14 exposure to
NIR light (n= 18). In this study, NIR irradiation of 950 nm/
1000cps/5% duty cycle (8mw/cm2) was used. The controls
were shielded from light and the NIR irradiation was applied
for 1 h. After this time point, the NIR light was turned off and
the drug solution was removed. Samples were taken at 5, and
Same Wavelength (950 nm), Different Duty Cycle (5% vs. 20%)
In this study, the same wavelength LED (950 nm) was tested,
with bevacizumab at two different duty cycles, in order to
study the impact of power level on photokinetic drug delivery.
Human sclera was used as the test membrane. The drug
donor was 400ul of 2.5 mg/ml bevacizumab in each Franz cell
and was removed at one hour and LED was turned off.
Samples (400 μl) were taken at 2, 3, 5, 8, 10 and 12 h. The
conditions tested were: dark control vs 950 nm/1000cps/5%
duty cycle (8mw/cm2) and dark control vs 950 nm/1000cps/
20% duty cycle (27mw/cm2). Samples were analyzed by
Ranibizumab with NIR 950 nm
Ranibizumab was diluted to 2.5 mg/ml and then 400ul (1 mg)
was placed in each Franz cell (n = 16). The photokinetic cells
were irradiated with 950 nm light, 1000cps, 20% duty cycle
(27 mW/ cm2) for the exposure hour. The donor drug
solution removed at one hour. Control cells were held in the dark
for the exposure hour. Samples (400 μl) were taken at 2, 3, 5,
8, 10, 12 and 24 h.
Aflibercept with NIR 950 nm
Aflibercept was diluted to 2.5 mg/ml and then 400ul (1 mg)
was placed in each Franz cell. The photokinetic cells were
irradiated with 950 nm light, 1000cps, 20% duty cycle
(27 mW/ cm2) for the exposure hour. The donor drug
solution was removed at one hour. Control cells were held in the
dark for the exposure hour. Samples (400 μl) were taken at 2,
3, 5, 8, 10, 12 and 24 h.
In Vitro Studies
Bevacizumab/NIR Light Stability Study
Figure 2 shows the result of NIR light exposure using both
E L I S A a n d S E - H P L C m e t h o d s f o r a n a l y s i s . T h e
bevacizumab concentrations obtained for the control (no
NIR light) and light exposed (NIR light) groups, as analyzed
by SE-HPLC, were both about 550 ng/ml at 1 and 5 h,
respectively. The bevacizumab concentrations obtained for the
one hour control and light exposed group were also about
550 ng/ml, using ELISA as the analytical method. At 5 h,
both the control and the light exposed groups showed a slight
increase in concentration, when analyzed by ELISA. This
slight increase in concentration may be attributed to the
inherent error in ELISA methodology and/or standard curve
dilution or interpolation errors. In any event, it appears that
exposure to relatively high amounts of non-thermal NIR light
does not have a detrimental effect on bevacizumab diluted
with formula 14.
Different NIR Wavelengths
Figure 4 shows the results from the feasibility experiments
(n= 18/group). The control bevacizumab permeation (no
NIR light) was 2671 ± 1967 ng/cm2 at 5 h (Fig. 4a) and
5094 ± 2523 ng/cm2 at 8 h (Fig. 4b). The NIR light irradiated
samples (950 nm/1000cps/5% duty cycle (8 mW/cm2)) were
11,390 ± 7872 ng/cm2 at 5 h (Fig. 4a) and 19,859 ±
11,226 ng/cm2 at 8 h (Fig. 4b). This test successfully shows that
mAbs may be evaluated using this equipment, formula and
method. The results also show that using NIR light significantly
(p < 0.05) increases transscleral drug delivery of bevacizumab,
as well as providing a reliable 4X enhancement of permeation.
Figure 5 shows the results of a study comparing the power
output from the LEDs (5% vs. 20% duty cycle) at a wavelength
of 950 nm. The two mAb permeation profiles (LED
irradiated) are also compared against the permeation profile without
LED irradiation (dark control). When compared, both of the
LED enhanced mAb permeation profiles are about 3 times
that of the dark controls. However, when comparing mAb
enhanced permeation from 5% duty cycle power against
20% duty cycle power, both conditions led to nearly identical
results in the irradiated sclera. Judging from these results,
there is no significant difference in mAb transscleral
permeation when using NIR light at 5% or 20% duty cycle power.
ELISA vs. SE-HPLC
Ranibizumab by ELISA and SE-HPLC correlate closely. The
minor differences, seen in Fig. 6, are probably due to dilution
factors and inherent error associated with ELISA. There are
values of about ≤9% difference; therefore there is no
significant difference in ELISA analysis vs SE-HPLC analysis.
Ranibizumab with NIR 950 nm
With the one hour application of ranibizumab and concurrent
950 nm NIR irradiation, shown in Fig. 7, the flux values are
twice as much as the control flux values (Fig. 7a). This overall
2X increase of ranibizumab flux, when compared to control
values is consistent over the 24 h time period. The cumulative
amount of drug delivered, when compared to control values,
is about 2.3X after 24 h (Fig. 7b).
Aflibercept with NIR 950 nm
With the one hour application of aflibercept and concurrent
950 nm NIR irradiation, the flux values are twice as much as
8 Hrs. ng/ml (SD)
the control flux values, as shown in Fig. 8a. This overall 2-3X
increase of aflibercept flux, when compared to control values
is consistent over the 24 h time period. The cumulative
amount of drug delivered, shown in Fig. 8b, is about 3X,
when compared to controls after 24 h.
Comparison of Passive Transscleral Permeation
Earlier experiments reveal a passive, transscleral permeation
due only to the mAbs being formulated in F14. Data from the
ranibizumab, aflibercept and bevacizumab permeation
controls (shown in Fig. 9) was compared in order to illustrate the
enhancement without any light irradiation. The rate of
permeation is directly proportional to the molecular weight of the
We became interested in the stability of dilute mAb solutions
when we were developing and validating methods for mAb
photokinetic transscleral permeation. mAbs are usually
formulated in concentrations of 1 mg/ml or higher for
intraocular injections. In general, most pharmaceutical antibody
preparations contain 10-150 mg antibody/ml providing long term
storage stability. Due to our planned work with Franz cells
and tissue permeation, we needed to develop methods for
working with dilute solutions of mAbs.
To briefly explain, the Franz Cell chamber (Fig. 1) is an
in vitro tissue permeation assay frequently used in
transmembrane feasibility and formulation development. The Franz
Cell apparatus consists of two primary chambers separated by
a membrane. The test product is applied to the membrane via
the top chamber (donor). The bottom chamber (receiver)
contains fluid from which samples are taken at regular intervals for
analysis. This testing determines the amount of active
compound that has permeated the membrane at each point in time.
Published tissue permeation (Franz cell) studies typically
use PBS as the recipient media at 37°C. Franz cell studies start
with a recipient media at 0 concentration. As the permeation
experiment progresses, the concentration of the drug in the
receiver chamber increases. However, at early points in the
experiment, the drug is still a very dilute solution.
Body temperature and low drug concentrations in PBS (or
other common media) contribute greatly to mAb aggregation,
causing reduced measured concentrations and decreased
VEGF binding activity. Therefore, a need arose to quickly
characterize mAb dilute solutions and develop a formulation
that would allow for the stable and accurate handling of these
solutions in our permeation experiments.
Based upon the manufacturer’s formulation, excipient
substitutions were screened with dilutions of standard
concentrations of ranibizumab. An excipient formulation, that
we here designate as Formula 14 (F14), developed in our lab,
was chosen, which preserves and stabilizes antibody products
such as bevacizumab, ranibizumab and aflibercept. Standard
dilutions of bevacizumab, ranibizumab and aflibercept were
prepared in PBS, manufacture’s formulation, and the new
formulation. These were analyzed by HPLC and ELISA.
) The novelty of both using and correlating the two
analytical methods together ensured that the permeated drug
through the scleral tissue was unchanged mAb and that the
mAb was still biologically active.
Non-invasive transscleral drug delivery has been
accomplished through iontophoresis, (
) sonophoresis (
and photokinetics. (
) Eljarrat-Binstock et al. described
the iontophoretic transscleral delivery of methotrexate in
rabbit eyes. (33) Pescina, et al., used iontophoresis to deliver
transscleral delivery of bevacizumab and dexamethasone in
preclinical models. (
) Pulsed high-intensity focused
ultrasound has been also studied recently in order to facilitate drug
delivery across the sclera noninvasively. Cheung et al. used
ultrasound (sonophoresis) to enhance the intrascleral
penetration of protein, increasing the diffusivity by 1.6-folds while
causing no damage to the retinal tissues. (
) Suen et al. also
found that low-frequency ultrasound significantly enhanced
the penetration of macromolecules via transscleral route. (
Photokinetic ocular drug delivery (PODD) describes drug
permeation enhancement through the use of pulsed infrared
light for ophthalmic applications. Briefly, it is hypothesized that
if a drug molecule in a pharmacologically acceptable
formulation is placed on the surface of the sclera/cornea and cyclically
illuminated with a selected wavelength of NIR light at a
selected pulse rate, that drug molecule and/or tissue would absorb
the light, which would result in molecular bond vibrations. This
cyclic molecular bond stretching and relaxing would in turn
cause a molecular kinetic motion. The resulting cyclic physical
shape change of the molecule and/or tissue may cause gross
movement and result in the facilitated diffusion of the molecule
into and through the sclera/cornea membrane. (
Fig. 6 ELISA vs. HPLC values ≤9% difference, no significant difference
(control n = 8 and 950nm n = 8). Same time point samples handled the same
and split in two for analysis. HPLC samples did not require dilution for analysis.
ELISA samples required several step dilutions of 100 to 300 times (F14) to
bring samples into ELISA kit range of 1-100 ng/ml. All ELISA samples plated in
duplicate on same plate.
Bevacizumab was tested to verify that NIR light was not
damaging to the protein. Both ELISA and SE-HPLC
methods (shown in Fig. 2) were used for analysis. The
bevacizumab concentrations obtained for the control (no
NIR light) and light exposed (NIR light) groups, as analyzed
by SE-HPLC, were both about the same quantity at 1 and 5 h,
respectively. The bevacizumab concentrations obtained for
the one hour control and light exposed group were also about
the same quantity, using ELISA as the analytical method. At
5 h, both the control and the light exposed groups showed a
slight increase in concentration, when analyzed by ELISA.
This slight increase in concentration may be attributed to
the inherent error in ELISA methodology and/or standard
curve dilution or interpolation errors. For example, Barregard
et al., studying urinary 8-oxo-7,8-dihydro-2′-deoxyguanosine
(8-oxodG); a widely used biomarker of oxidative stress, found
that chromatographic assays showed high agreement across
urines from different subjects, whereas ELISAs showed far
more inter-laboratory variation and generally overestimated
levels, compared to the chromatographic assays. (
Additionally, errors in plate manufacturing, handling,
loading, or mistakes in the usage of reagents are all common causes
of well variance in ELISA. (
In our in vitro studies, we did not vary the length of time for
the application of the antibody solution or the length of time of
NIR irradiation. These parameters were held constant at
60 min. This 60 min treatment protocol is central to our
envisioned mode of therapy. We found that applying pulsed
NIR light at 950 nm increased the amount of drug to
permeate through scleral tissue (Figs. 3, 4, 5, 6, 7, and 8). Figure 9
shows the amount of passive transscleral permeation of each
mAb as derived from the control values of the previous figures.
We investigated several different LEDs, with variable
attainable light powers for each, as previously described. The
distance of the LED to the tissue surface was constant,
however the irradiated power from each test condition varied.
During the NIR irradiation of the mAb solution study, we
used LEDs positioned 3 cm above the surface of the drug
solution (950 nm). The LED driver currents were adjusted
to provide an irradiation power of 12 mW ± 1 mW at 3 cm
away from the LED. These solutions were continuously
irradiated for 5 h, with samples taken at 1 and 5 h; the irradiated
samples and the dark control samples were treated the same.
The total light dosage (power x time) at 5 h exceeded all the
Franz test conditions of one hour light exposure used in any
tissue permeation experiment. In the Franz cell permeation
experiments, the LED was positioned at 1 cm above the tissue
surface and approximately 0.5 cm above the drug donor. This
1 cm distance was selected for the reason that the average
vertex distance of spectacle lenses (the distance between the
front of the cornea and the back surface of a corrective lens),
when properly fitted, is 8–12 mm. Many refractive surgeons
assume an average vertex distance of 12 to 14 mm for all
patients without measuring vertex distance. The average true
vertex distance has been reported to be 20.4 mm with a range
of 10 to 34 mm. (
In our studies, pulsed NIR light was remotely directed onto
aqueous mAb solutions, as well as mAb solutions applied onto
the surface of scleral tissue. The NIR LEDs were electronically
driven with short ON period pulses. Duty cycle or pulse width
modulation specifications of 5–20% (5–20% of the total cycle
time being ON) eliminates heat buildup within the LED itself.
) Absorbed light energy transformed into heat is the
primary degrading pathway for exposed tissues, cells and proteins.
Direct heating, for example, from contact with the LED itself
could be another source of damage. (
In our configuration, the LED is not in contact with the
drug, drug solution or tissue surface. Therefore, the possibility
of direct heat damage from the LED is not an issue. For light
energy to be transformed into heat requires that the light be
absorbed. Water is the primary light absorbing material at
950 nm. Our direct temperature measurements of the Franz
drug donor solutions indicated minimal temperature rises of
≤1°C above the 37°C water bath. This value, being well
below a destructive thermal injury temperature, indicates that
the possibility of heat damage to tissues or proteins is
Non-thermal red to NIR light is non-ionizing and is
thought to be of no consequence and may actually be
beneficial to ocular tissues. (
) We are not aware of any reports
wherein pulsed non-thermal NIR light in the 950 nm range, at
the conditions we describe, have caused damage to tissues or
These results demonstrate that narrow wavelength
incoherent (non-laser) light from an NIR light emitting diode
(LED) source can be used for drug delivery and that
nonionizing pulsed NIR light with these characteristics is not
harmful to the drug molecule.
In addition to drug permeation enhancement, we also saw
a drug depot effect. A drug depot effect can be defined as a
body area in which a substance, e.g., a drug, can be
accumulated, deposited, or stored and from which it can be
distributed. We observed that although the drug solution was in
contact with the sclera for a relatively short time - one hour with
concurrent IR light, and then both the drug solution and light
removed - there was continuous and enhanced drug diffusion
of the drug over the next 23 h when compared to Bno light^
controls. The application of NIR light generates a
photokinetic action, thereby enhancing drug absorption into the outer
layers of scleral/corneal tissue, which leads to an overall 2-3X
increase in drug permeation over a prolonged amount of time.
In this sense, there is improved and enhanced tissue
penetration, depot effect and sustained release drug action.
Intravitreal injection of anti-VEGF therapies are
commonly performed every 4–8 weeks to treat retinal disease
(Table II). This treatment regimen is based mainly on the
patient tolerability of repeated injections, rather than
maintaining optimum binding concentrations of anti-VEGF drugs
to their target. Aflibercept’s competitive advantage over
ranibizumab is attributed to an every-two-month intravitreal
injection treatment compared to a once-a-month injection
required with ranibizumab. This indicates that patients prefer
Comparing once-a-month intravitreal injection to a more
frequent (1 or 2 times a week) noninvasive topical treatment
has to take into account the therapeutic concentration
requirement, not the dose administered. Injected anti-VEGF
mAb concentrations, in the vitreous, have large initial peaks
with long low concentration troughs. (
) The biological
halflife of bevacizumab and ranibizumab is found by about day 8
after injection. By day 14, the vitreal concentration of injected
mAbs is generally insignificant. More frequent dosing, every
Table II Current clinical
intravitreal injection protocols
*Concentration in vitreous after injection. Assume vitreous body is about 4 ml
2 weeks, has been shown to have greater efficacy. (
more effective treatment would be for more frequent dosing;
however patents will generally not tolerate an every-2-week
treatment regime. Non-invasive frequent topical dosing may
not reach the high peak levels of an injection, but may provide
a constant therapeutic level within the eye. This concept is the
foundation of non-invasive ocular delivery, (44) as well as for
controlled release from reservoir implants for posterior
segment disease. (
Any topical mAb ocular therapy will require modification
of the antibody carrier composition to be physiologically
acceptable (pH, osmolarity and viscosity) as well as
pharmacologically acceptable (excipients used to prevent antibody
aggregation) for application onto the eye. Ocular compositions
need to be non-irritating and provide enough resident time for
ocular tissue absorption. We have previously determined that
anti-VEGF antibodies undergo significant aggregation unless
we modify the carrier composition. (
) Our composition
provides monomeric antibody forms that can be used for tissue
permeation in a clinical setting.
Our Franz cell drug donor; a topical formulation with a
concentration of 2.5 mg/ml, was based on recent anti-VEGF
topical therapies for corneal neovascularization (
well as prior permeation cell work (27) (see Table III). In these
prior studies, existing compositions were diluted with saline
and applied to the eye or eye tissue. The concentration of
2.5 mg/ml was used as a starting point for the studies to relate
to prior work by others as well as provide a method to suspend
antibodies in our anti-aggregation composition.
The in vitro flux data provides information for the in vivo
situation. However, this information cannot fully simulate
and account for the actual limited drug residence time on
the tissue surface, due to tear washout time, or the restrictions
from multiple tissue layers under the sclera. On the other
hand, the in vitro conditions may not include increases in drug
flux due to increased distribution from drug donor/tissue
contact such as eye blinking or the increased available permeation
area found when the entire anterior eye surface is considered.
There also remains an added possible drug reservoir effect
and prolonged drug elution from the conjunctiva into the eye.
As shown in our permeation model results, we have
attained clinically relevant drug fluxes with a non-stirred,
one-hour application of 0.4 ml of a 2.5 mg/ml solution
(1 mg dose) related to one square cm of sclera permeation
area. We used sclera sections from human eyes immediately
adjacent to the cornea representing a clinically relevant tissue
treatment location and tissue thickness. In the case of
ranibizumab, which is clinically dosed at 0.5 mg /injection,
the monthly dose is 500,000 ng. In our experiment, the dose
given was 1,000,000 ng. Our transscleral results were
Table III Topical anti-VEGF agent
concentrations of various studies
FITC labeled Bevacizumab
60,000 ng/24 h passively and 150,000 ng/24 h using 60 min
of NIR. If a one hour NIR treatment was given once a week
for 4 weeks, this would equal 600,000 ng which is comparable
to the injection dose.
Drug permeation into and through tissues is time,
concentration and carrier dependent. The Franz cell
model predicts in vivo permeation and drug fluxes under
specified conditions. The translation from predicted
fluxes with the permeation cell model into actual
achievable in vivo drug fluxes can only be done by direct in vivo
models. We are currently planning pre-clinical studies to
evaluate the mAb formulation and photokinetic drug
We first developed a novel pharmacologically acceptable
formulation that prevented antibody aggregation and then
combined it, in vitro, with our enhanced, near infrared (NIR)
lightbased transscleral drug delivery system.
In vitro human sclera permeation methods were adapted to
include concurrent irradiation with non-thermal, pulsed
950 nm NIR light from an LED source. We observed that
although the drug solution was in contact with the sclera for a
relatively short time - one hour with concurrent NIR light,
and then both the drug solution and light removed - there
was continuous and enhanced drug diffusion as well as drug
elution from the sclera over the next 23 h.
Pulsed, non-thermal NIR irradiation is non-visible light,
which is more preferred; being less annoying to a patient,
harmless to eye tissues. Exposure to pulsed, non-thermal
NIR light has no degrading effect on mAbs, when
comprised with our formula. The optimal LED wavelength
for enhanced permeation was 950 nm. The duty cycle
(ON time: total cycle time) of 5% vs 20% showed no
difference in enhanced permeation. The topical formulation,
combined with pulsed NIR light irradiation, significantly
improved scleral permeation of three anti-VEGF agents
compared to control conditions.
The combination of a non-aggregating antibody
formulation and pulsed NIR irradiation, provided an average
enhancement factor of 3.0 (range is 2.5–3.5X), when compared
to passive permeation. This combination of stabilized
antibody with light irradiation provides clinically relevant drug
amounts to be delivered into the eye using a one-hour
treatment avoiding needle injection. A one-hour non-invasive,
drug/NIR treatment, given once a week for 4 weeks, would
equal the same dose given by a monthly injection. Our model
serves as a good case for the importance of topical
compositions preventing antibody aggregation as well as an indication
that topical antibody delivery for posterior segment treatment
is possible and likely.
ACKNOWLEDGMENTS AND DISCLOSURES. The au
thors wish to acknowledge and thank Genentech, South San
Francisco CA, for the gift of Lucentis and Regeneron for the
gift of IAI Eylea. The authors Steven A. Giannos and Edward
R. Kraft own significant interest in the IP of the technology
described. The authors Zhen-Yang Zhao, Kevin H. Merkley
and Jiyang Cai have no competing interests.
Open Access This article is distributed under the terms of the
Creative Commons Attribution 4.0 International License
permits unrestricted use, distribution, and reproduction in any
medium, provided you give appropriate credit to the original
author(s) and the source, provide a link to the Creative
Commons license, and indicate if changes were made.
1. Pascolini D , Mariotti SP . Global estimates of visual impairment: 2010. Br J Ophthalmol . 2012 ; 96 ( 5 ): 614 - 8 .
2. Congdon N , O'Colmain B , Klaver CCW , Klein R , Muñoz B , Friedman DS , et al. Causes and prevalence of visual impairment among adults in the United States . Arch Ophthalmol . 2004 ; 122 : 477 - 85 .
3. Martin DF , Maguire MG , Ying GS , Grunwald JE , Fine SL , Jaffe GJ , et al. Ranibizumab and bevacizumab for Neovascular agerelated macular degeneration the CATT research group . N Engl J Med . 2011 ; 364 ( 20 ): 1897 - 908 .
4. Stewart MW . The expanding role of vascular endothelial growth factor inhibitors in ophthalmology . Mayo Clin Proc . 2012 ; 87 ( 1 ): 77 - 88 .
5. Keane PA , Sadda SR . Development of anti-VEGF therapies for intraocular use: a guide for clinicians . J Ophthalmol . 2012 ; 2012 : 1 - 13 .
6. Weber M , Sennlaub F , Souied E , Cohen SY , Behar-Cohen F , Milano G , et al. Review and Expert opinion in age related macular degeneration. Focus on the pathophysiology, angiogenesis and pharmacological and clinical data . Journal Francais D Ophtalmologie . 2014 ; 37 ( 7 ): 566 - 79 .
7. Ecker DM , Jones SD , Levine HL . The therapeutic monoclonal antibody market . Mabs-Austin . 2015 ; 7 ( 1 ): 9 - 14 .
8. Reichert JM . Antibodies to watch in 2015 . Mabs-Austin. 2015 ; 7 ( 1 ): 1 - 8 .
9. Osaadon P , Fagan XJ , Lifshitz T , Levy J. A review of anti-VEGF agents for proliferative diabetic retinopathy . Eye (Lond) . 2014 ; 28 ( 5 ): 510 - 20 .
10. Yu DC , Lee JS , Yoo JY , Shin H , Deng HX , Wei YQ , et al. Soluble vascular endothelial growth factor decoy receptor FP3 exerts potent antiangiogenic effects . Mol Ther . 2012 ; 20 ( 5 ): 938 - 47 .
11. Stewart MW . Clinical and differential utility of VEGF inhibitors in wet age-related macular degeneration: focus on aflibercept . Clin Ophthalmol . 2012 ; 6 : 1175 - 86 .
12. Campbell M , Humphries MM , Humphries P . Barrier modulation in drug delivery to the retina . Methods Mol Biol . 2013 ; 935 : 371 - 80 .
13. Cunha-Vaz J . Blood-Retinal Barrier A2 - Dartt. In: Darlene A, editor. Encyclopedia of the eye . Oxford: Academic Press; 2010 . p. 209 - 15 .
14. Hosoya KI , Tachikawa M. Inner blood-retinal barrier transporters: role of retinal drug delivery . Pharm Res . 2009 ; 26 ( 9 ): 2055 - 65 .
15. Tufail A , Patel PJ , Egan C , Hykin P , da Cruz L , Gregor Z. Bevacizumab for neovascular age related macular degeneration (ABC trial): multicentre randomised double masked study . BMJ 2010 ; 340 ( c2459 ): 1 - 10 .
16. Falavarjani KG , Nguyen QD . Adverse events and complications associated with intravitreal injection of anti-VEGF agents: a review of literature . Eye (Lond) . 2013 ; 27 ( 7 ): 787 - 94 .
17. Li SK , Hao JS . Transscleral passive and iontophoretic transport: theory and analysis . Expert Opin Drug Deliv . 2018 ; 15 ( 3 ): 283 - 99 .
18. Lafond M , Aptel F , Mestas JL , Lafon C . Ultrasound-mediated ocular delivery of therapeutic agents: a review . Expert Opin Drug Deliv . 2017 ; 14 ( 4 ): 539 - 50 .
19. Godley BF , Rowe-Rendleman CL , Kraft E , Kulp G , editors. Transsceral drug delivery to the posterior segment of the eye . Boca Raton: CRC Press; 2013 .
20. Kraft ER , Kulp GA , Godley BF , Koutrouvelis AP . Photokinetic ocular drug delivery methods and apparatus . In: USPTO, editor. United States Patent & Tradmark Office. United States: The Board of Regents , The University of Texas System; 2011 . p. 30 .
21. Godley BF , Kraft ER , Giannos SA , Zhao ZY , Haag AM , Wen JW . Photokinetic drug delivery: light-enhanced permeation in an in vitro eye model . J Ocul Pharmacol Th . 2015 ; 31 ( 10 ): 650 - 7 .
22. Ambati J , Canakis CS , Miller JW , Gragoudas ES , Edwards a , Weissgold DJ , et al. Diffusion of high molecular weight compounds through sclera . Invest Ophthalmol Vis Sci . 2000 ; 41 : 1181 - 5 .
23. Giannos SA , Kraft ER , Zhao ZY , Merkley KH , Cai J . Formulation Stabilization and Disaggregation of Bevacizumab, Ranibizumab and Aflibercept in Dilute Solutions . Pharm Res . 2018 ; 35 ( 4 ): 78 .
24. Kraft ER , Giannos SA , Godley BF . Antibody and Protein Therapeutic Formulations and Uses Thereof . The Board of Regents , The University of Texas System; 2017 . US Patent Application, Serial Number PCT/US/53185.
25. Celik N , Scheuerle A , Auffarth GU , Kopitz J , Dithmar S. Intraocular pharmacokinetics of Aflibercept and vascular endothelial growth factor-a . Invest Ophthalmol Vis Sci . 2015 ; 56 ( 9 ): 5574 - 8 .
26. Eljarrat-Binstock E , Pe'er J , Domb AJ . New techniques for drug delivery to the posterior eye segment . Pharm Res . 2010 ; 27 ( 4 ): 530 - 43 .
27. Pescina S , Ferrari G , Govoni P , Macaluso C , Padula C , Santi P , et al. In-vitro permeation of bevacizumab through human sclera: effect of iontophoresis application . J Pharm Pharmacol . 2010 ; 62 ( 9 ): 1189 - 94 .
28. Chopra P , Hao JS , Li SK . Iontophoretic transport of charged macromolecules across human sclera . Int J Pharm . 2010 ; 388 ( 1-2 ): 107 - 13 .
29. Suen WL , Wong HS , Yu Y , Lau LC , Lo AC , Chau Y. Ultrasoundmediated transscleral delivery of macromolecules to the posterior segment of rabbit eye in vivo . Invest Ophthalmol Vis Sci . 2013 ; 54 ( 6 ): 4358 - 65 .
30. Razavi Mashoof A. High intensity focused ultrasound in ophthalmology : part one, transscleral drug delivery : part two, infrared thermography for scalable acoustic characterization, an application in the manufacture of a glaucoma treatment device . In.: Université Claude Bernard - Lyon I; 2014 .
31. Shah R , Zderic V . Ultrasound-enhanced drug delivery through sclera . J Acoust Soc Am . 2009 ; 125 ( 4 ): 2680 - 0 .
32. Cheung ACY , Yu Y , Tay D , Wong HS , Ellis-Behnke R , Chau Y . Ultrasound-enhanced intrascleral delivery of protein . Int J Pharm . 2010 ; 401 ( 1-2 ): 16 - 24 .
33. Eljarrat-Binstock E , Domb AJ , Orucov F , Frucht-Pery J , Pe'er J. Methotrexate delivery to the eye using transscleral hydrogel iontophoresis . Curr Eye Res . 2007 ; 32 ( 7-8 ): 639 - 46 .
34. Barregard L , Moller P , Henriksen T , Mistry V , Koppen G , Rossner P , et al. Human and methodological sources of variability in the measurement of urinary 8- Oxo-7 , 8 - dihydro-2 '- deoxyguanosine . Antioxid Redox Sign . 2013 ; 18 ( 18 ): 2377 - 91 .
35. Lilyanna S , Ming Wei Ng E , Moriguchi S , Pang Chan S , Kokawa R , Hung Huynh S , Jenny Chong PC , Xia Ng Y , Mark Richards A , Ng TW , Wah Liew O. Variability in Microplate Surface Properties and Its Impact on ELISA . The Journal of Applied Laboratory Medicine: An AACC Publication . 2017 .
36. Weiss RA , Berke W , Gottlieb L , Horvath P. Clinical importance of accurate refractor vertex distance measurements prior to refractive surgery . J Refract Surg . 2002 ; 18 ( 4 ): 444 - 8 .
37. Bozkurt A , Onaral B . Safety assessment of near infrared light emitting diodes for diffuse optical measurements . Biomed Eng Online . 2004 ; 3 ( 1 ): 9 .
38. Eells JT , Gopalakrishnan S , Valter K. Near-infrared Photobiomodulation in retinal injury and disease . Retinal Degenerative Diseases: Mechanisms and Experimental Therapy . 2016 ; 854 : 437 - 41 .
39. Geneva II . Photobiomodulation for the treatment of retinal diseases: a review . Int J Ophthalmol . 2016 ; 9 ( 1 ): 145 - 52 .
40. Whelan HT , Wong-Riley MTT , Eells JT , VerHoeve JN , Das R , Jett M. DARPA soldier self care: rapid healing of laser eye injuries with light emitting diode technology . In. RTO HFM Symposium on Combat Casualty Care in Ground Based Tactical Situations: Trauma Technology and Emergency Medical Procedures. St . Petersburg Beach, Florida; 2004 .
41. Stewart MW . Predicted biologic activity of intravitreal bevacizumab . Retina . 2007 ; 27 ( 9 ): 1196 - 200 .
42. Stewart MW , Rosenfeld PJ , Penha FM , Wang F , Yehoshua Z , Bueno-Lopez E , et al. Pharmacokinetic rationale for dosing every 2 weeks versus 4 weeks with intravitreal ranibizumab, bevacizumab, and aflibercept (vascular endothelial growth factor trap-eye) . Retina . 2012 ; 32 ( 3 ): 434 - 57 .
43. Foss AJE , Childs M , Reeves BC , Empeslidis T , Tesha P , DharMunshi S , et al. Comparing different dosing regimens of bevacizumab in the treatment of neovascular macular degeneration: study protocol for a randomised controlled trial . Trials . 2015 ; 16 : 85 .
44. Gaudana R , Ananthula HK , Parenky A , Mitra AK . Ocular drug delivery . AAPS J . 2010 ; 12 : 348 - 60 .
45. Thrimawithana TR , Young S , Bunt CR , Green C , Alany RG . Drug delivery to the posterior segment of the eye . Drug Discov Today . 2011 ; 16 ( 5-6 ): 270 - 7 .
46. Park Y-R , Chung SK . Inhibitory effect of topical Aflibercept on corneal neovascularization in rabbits . Cornea . 2015 ; 34 ( 10 ): 1303 - 7 .
47. Krizova D , Vokrojova M , Liehneova K , Studeny P. Treatment of corneal neovascularization using anti-VEGF bevacizumab . J Ophthalmol . 2014 ; 2014 : 1 - 7 .
48. Ozdemir O , Altintas O , Altintas L , Ozkan B , Akdag C , Yuksel N . Comparison of the effects of subconjunctival and topical antiVEGF therapy (bevacizumab) on experimental corneal neovascularization . Arq Bras Oftalmol . 2014 ; 77 ( 4 ): 209 - 13 .
49. Kadar T , Amir A , Cohen L , Cohen M , Sahar R , Gutman H , et al. Anti-VEGF therapy (bevacizumab) for sulfur mustard-induced corneal neovascularization associated with delayed Limbal stem cell deficiency in rabbits . Curr Eye Res . 2014 ; 39 ( 5 ): 439 - 50 .
50. Ferrari G , Dastjerdi MH , Okanobo A , Cheng SF , Amparo F , Nallasamy N , et al. Topical Ranibizumab as a treatment of corneal neovascularization . Cornea . 2013 ; 32 ( 7 ): 992 - 7 .
51. Ahmed A , Berati H , Nalan A , Aylin S . Effect of bevacizumab on corneal neovascularization in experimental rabbit model . Clin Exp Ophthalmol . 2009 ; 37 : 730 - 6 .
52. Dastjerdi MH , Al-Arfaj KM , Nallasamy N , Hamrah P , Jurkunas UV , Pineda R , et al. Topical bevacizumab in the treatment of corneal neovascularization . Arch Ophthalmol . 2009 ; 127 ( 4 ): 381 - 9 .
53. Koenig Y , Bock F , Horn F , Kruse F , Straub K , Cursiefen C . Shortand long-term safety profile and efficacy of topical bevacizumab (Avastin(a (R))) eye drops against corneal neovascularization . Graef Arch Clin Exp . 2009 ; 247 ( 10 ): 1375 - 82 .
54. Bock F , Konig Y , Kruse F , Baier M , Cursiefen C. Bevacizumab (Avastin) eye drops inhibit corneal neovascularization . Graef Arch Clin Exp . 2008 ; 246 ( 2 ): 281 - 4 .
55. Kim SW , Ha BJ , Kim EK , Tchah H , Kim T -i. The effect of topical bevacizumab on corneal neovascularization . Ophthalmology . 2008 ; 115 ( 6 ): e33 - 8 .
56. DeStafeno JJ , Kim T. Topical bevacizumab therapy for corneal neovascularization . Arch Ophthalmol . 2007 ; 125 ( 6 ): 834 - 6 .