1461-5347/98/$ see front matter 1998 Elsevier Science. All rights reserved.

PII: S1461-5347(98)00087-X
M Most conventional ophthalmic dosage forms
are simplistic. It is usual that water-soluble drugs
are delivered through topical administration in
an aqueous solution
1
, and water-insoluble drugs
are administered topically as an ointment or
aqueous suspension. The major deficiencies of
these conventional dosage forms include poor
ocular drug bioavailability, pulse-drug entry after
topical administration, systemic exposure be-
cause of nasolacrimal duct drainage, and a lack
of effective systems for drug delivery to the pos-
terior segment of ocular tissue.
Poor ocular drug bioavailability is the result of
ocular anatomical and physiological constraints,
which include the relative impermeability of the
corneal epithelial membrane, tear dynamics,
nasolacrimal drainage
2
, and the high efficiency
of the bloodocular barrier
3
. It is standard for
only 1% or less of a topically applied dose to be
absorbed across the cornea and thus reach the
anterior segment of the eye
4,5
.
Pulse entry is a common, and yet highly un-
desirable, pharmacokinetic characteristic associ-
ated with eye drops
6
. The initial high drug con-
centration found in tears, followed by a rapid
decline, poses a potential risk of toxicity, and
suggests a requirement for frequent dosing.
Attempts to overcome the toxicity associated
with the high initial concentration without a
requirement for frequent dosing form a challeng-
ing task, particularly in the case of potent drugs.
Nasolacrimal drainage is the major factor for
precorneal drug loss that leads to poor ocular
bioavailability. It is also the major route of entry
into the circulatory system for drugs that are
applied through topical administration
7,8
. For
potent drugs, the systemic exposure through na-
solacrimal drainage after topical administration
can be sufficiently high to cause systemic toxicity.
A recognized example is timolol; systemic tox-
icity has been reported for the ophthalmic solu-
tion of timolol following topical administration
9
.
The delivery of drugs to the posterior segment
of ocular tissue is prevented by the same factors
that are responsible for the poor ocular bioavail-
ability. In addition, the bloodretinal barrier lim-
its the effectiveness of the intravenous route in
posterior drug delivery. To date, the most accept-
able method for posterior drug delivery is intra-
vitreal injection, and yet although effective, the
intravitreal injection procedure is associated with
a high risk of complications
3
.
Early attempts
A considerable amount of effort has been made
in ophthalmic drug delivery since the 1970s. The
various approaches attempted in the early stages
can be divided into two main categories:
bioavailability improvement and controlled re-
lease drug delivery. The former was attempted by
the methods listed in Table 1, and the latter was
attempted by various types of inserts (Table 2)
and nanoparticles. After initial investigations,
some approaches were dropped quickly, whereas
others were highly successful and led to mar-
keted products.
Bioavailability improvement
Viscosity enhancers. Attempts to increase the viscosity
of the formulation, in order to prolong precorneal
residence time and improve bioavailability,
Recent developments in ophthalmic
drug delivery Shulin Ding
Shulin Ding
Allergan
2525 Dupont Drive
Irvine, CA 92623
USA
tel: 1 714 246 6218
fax: 1 714 246 6981
e-mail:
Ding_Shulin@Allergan.com
reviews research focus
328
PSTT Vol. 1, No. 8 November 1998
Recent research efforts in ophthalmic drug delivery have mainly
focused on systems in which drugs may be administered in the form
of eye-drops. As a result of these efforts, significant advancements
have been made in the following areas: in situ-forming gels, oil-
in-water emulsions, colloidal drug delivery systems (liposomes and
nanoparticles) and microparticulates. Protein and peptide delivery,
posterior drug delivery and non-aqueous vehicles are three new
areas of interest in ophthalmic drug delivery, and this review will
discuss recent progress and specific development issues relating to
these drug delivery systems.
formed a popular strategy in the early research stages of oph-
thalmic drug delivery. The viscosity enhancers were often hy-
drophilic polymers, such as various celluloses, polyvinyl
alcohol and polyacrylic acid. The conceptual simplicity, ease of
manufacture, and initially promising data in animal models
meant that, despite the lack of human data, in the 1980s many
of the viscosity enhancers could be incorporated into eye-drop
formulations. Subsequently, the effect of these viscosity en-
hancers on drug bioavailability was demonstrated to be mini-
mal in humans
10
, and their clinical significance was modest at
best
11
. In rabbits, it was found that the effect was more pro-
nounced, and reached a maximum level at a viscosity of ap-
proximately 15150 cps
12
. A review of the research efforts in
this area was presented by Lee and Robinson in 1986
4
. Today,
hydrophilic polymers continue to be used in formulations of
ophthalmic products, but their function is more for patient
comfort and/or reasons of bioadhesion rather than viscosity
enhancement.
Gels. Gel formation is an extreme case of viscosity enhance-
ment through the use of viscosity enhancers. Many of the vis-
cosity enhancers form a viscous gel at high concentrations in
water. Despite the extremely high viscosity, gels achieve only a
limited improvement in bioavailability, and the dosing fre-
quency can be decreased to once a day at most
13
. The high
viscosity, however, results in blurred vision and matted eyelids,
which substantially reduce patient acceptability. To date, only
two ophthalmic gels are listed in the Physicians Desk Reference for
Ophthalmology: Pilopine HS

gel, commercialized in 1986 by

Alcon, and more recently Mercks Timoptic-XE

.
Penetration enhancers. In the early stages of ophthalmic drug de-
livery research, chelating agents, preservatives, surfactants and
bile salts were studied as possible penetration enhancers. The
effort was soon diminished, however, due to the local toxicity
associated with the enhancers. Current research efforts are fo-
cused first on developing an understanding of the mechanisms
and regulations of drug transport, before proposing specific
methods for penetration enhancement
14
.
Prodrugs. The principle of prodrugs is to enhance corneal
drug permeability through modification of the hydrophilicity
(or lipophilicity) of the drug. Prodrugs have proven to be of
value, but also difficult to design and develop, as evidenced by
the fact that only one ophthalmic prodrug, Allergans Propine

,
has enjoyed successful development and commercialization.
No new prodrug has been introduced to the ophthalmic mar-
ket since the commercialization of Propine

in 1980.
Liposomes. The use of liposomes as a topically administered
ocular drug delivery system began in the early stages of research
into ophthalmic drug delivery. The results of early studies were
329
PSTT Vol. 1, No. 8 November 1998 reviews research focus
Table 1. Early attempts to increase bioavailability
Approach Advantage Disadvantage Marketed product
Viscosity enhancers
4,1012
(such Droppable formulation Minimal effects in humans in Present in many current
as hydroxylpropyl methyl-cellulose, Ease in manufacture terms of improvement in products
carboxylmethyl cellulose, polyvinyl Slight prolonged precorneal bioavailability
alcohol, carbomers) residence time
Gels
13
Slightly prolonged precorneal Limited value in terms of Pilopine HS gel
residence time improvement in bioavailability
Reduced systemic exposure Blurred vision
Matted eyelids
Penetration enhancer
4,14
Enhanced corneal drug penetration Ocular irritation and toxicity None
Prodrugs
4
Enhanced corneal drug penetration Extensive pharmacokinetic and Propine
pharmacologic information is
required for proper design
Prodrug is considered as a new
drug entity
Liposomes
1520
Droppable Low drug loading None
Biocompatible Inadequate aqueous stability
Biodegradable Manufacturing difficulties
Potential for bioavailability for sterile preparations
improvement, toxicity reduction,
sustained release, and site-
specific delivery
favourable for lipophilic drugs
1517
, but disappointing for
hydrophilic drugs
1719
. The physicochemical properties of the
encapsulated drug apparently had a significant impact on the
subsequent penetration of the drug into tissues. When the en-
capsulated drug was hydrophilic, the effectiveness of liposomes
as drug carriers was compromised by the rapid escape of the
drug from the interior of the liposomes into the surrounding
water. Further studies revealed that liposomes must be adsorbed
onto corneal and conjunctival surfaces for drugs to penetrate
into ocular tissues, and the number of contact sites on these sur-
faces was limited
17,20
. It was concluded that liposomes may be
suitable for ocular drug delivery, provided they had an affinity
for, and were able to bind to, ocular surfaces, and released
their contents at optimal rates
20
. Since the discovery of contact
sites on ocular surface tissues, the effect of liposome surface
properties on drug delivery has provided a focus for research,
and new findings will be described in the section detailing
recent developments.
Controlled drug delivery
Nanoparticles. The potential use of polymeric colloidal particles as
ophthalmic drug delivery systems can be traced back to the
late 1970s. The first two systems studied in this area were the
pilocarpinecellulose acetate hydrogen phthalate (CAP) latex
system
4,21
and Piloplex
4,22
. The pilocarpineCAP latex system
consisted of pilocarpine adsorbed onto CAP particles that were
suspended in an aqueous solution (pH 4.5). The average size of
the particles was approximately 300 nm. The CAP particles co-
agulated in the ocular cul-de-sac because of a pH shift from 4.5
to 7.4 (the pH of tears), and pilocarpine was slowly released
from the coagulated CAP mass. The other early system, Piloplex,
was an aqueous polymeric dispersion of pilocarpine salt and the
co-polymer of laurylmethacrylated-acrylic acid. Piloplex was
tested in humans and demonstrated limited success in sustained
drug release
4
. Neither CAP latex nor Piloplex entered further
commercial development because of various development
issues, such as non-biodegradability, local toxicity and difficulty
in the large-scale manufacture of sterile preparations. Subsequent
research on nanoparticles has focused on biodegradable and/or
bioadhesive nanospheres and nanocapsules, which will be
covered in the section detailing recent developments.
Inserts. Solid inserts were introduced to the ophthalmic mar-
ket 50 years ago. The earliest official record of a solid insert was
described in the 1948 British Pharmacopoeia. It was an atropine-
containing gelatin wafer, Lamellae. The potential of solid inserts
for sustained or controlled drug release was subsequently rec-
ognized, and, in the 1970s and 1980s, numerous systems were
developed using various polymers, and different drug release
principles were applied. The various types of inserts are catego-
rized according to physicochemical properties in Table 2 (see
references 6 and 2325 for review). Although the advantages of
inserts in sustained or controlled drug release have been clearly
demonstrated, there is not a high level of acceptance amongst
patients, particularly the elderly. The low level of patient
330
PSTT Vol. 1, No. 8 November 1998 reviews research focus
Table 2. Various types of ophthalmic inserts
Type Description Marketed product
Erodible inserts
6,25
The fabrication polymer is hydrophobic but biodegradable. Drug is None
released through the erosion of the surface of the insert
Soluble inserts
6,23,25
The fabrication polymer is hydrophilic and water soluble Lacrisert
Drug release characteristics: Soluble ophthalmic drug insert
Diffusion control for soluble drugs (SODI)
Dissolution control for less soluble drugs
Hydrophilic but water-insoluble The fabrication polymer is hydrophilic but water-insoluble None
inserts
6,25
Drug release characteristics:
Diffusion control for soluble drugs
Dissolution control for less soluble drugs
Inserts using osmotic A polymeric matrix in which the drug is dispersed as discrete None
system
6,24
small domains. Upon placement in the cul-de-sac, tears are
imbibed into the matrix because of an osmotic pressure gradient
created by the drug, whereupon the drug is dissolved and released
Membrane-controlled The drug core is surrounded by a hydrophobic polymer membrane; Ocusert
diffusional inserts
6,23
this controls the diffusion of the drug from the core to the outside
acceptance is due primarily to the following issues: difficulty
with self insertion, foreign body sensation and inadvertent loss
from the eye. To date, only two insert products are listed in the
Physicians Desk Reference for Ophthalmology (Ocusert and Lacrisert),
and pharmaceutical manufacturers are not actively developing
inserts for commercialization.
Recent developments
An important lesson learned from earlier efforts in ophthalmic
drug delivery is the necessity of balancing the technologies of
sustained drug release or bioavailability improvement with pa-
tient comfort and ease of use. It is generally agreed that the
preferred topical ophthalmic drug delivery system would be
administered in eye-drop form, without causing blurred
vision and irritation. The preferred system would also
provide improved bioavailability, site-specific delivery, and/or
continuous drug release
2,4,22
.
With this in mind, recent research efforts have focused
largely on systems and technologies in which drugs can be ad-
ministered as an eye drop. Significant advancements have been
made in the following areas:

in situ-forming gels;

oil-in-water emulsions;

colloidal drug delivery systems (liposomes and nano-

particles);

microparticulates.
One in situ-forming gel, Timoptic-XE

, has been commer-

cially available since 1994, and a cyclosporine-containing oil-
in-water ophthalmic emulsion is in late Phase III development.
The development of liposomes, nanoparticles and micropar-
ticulates for topical ophthalmic applications has proven to be
extremely challenging. Despite recent progress, major obsta-
cles remain to be overcome before commercialization can be
realized, and their potential as injectable products for posterior
drug delivery is under current investigation. Finally, three
new areas of interest in ophthalmic drug delivery include
protein/peptide delivery, intraocular drug delivery and non-
aqueous vehicle.
In situ-forming gels
Major progress has been made recently in ophthalmic gel tech-
nology. This progress has been in the form of improved levels of
patient acceptance, and, more specifically, the development of
droppable gels (in situ-forming gels). The droppable gels are
liquid upon instillation, and they undergo a phase transition in
the ocular cul-de-sac to form a viscoelastic gel, and this provides
a response to environmental changes. Parameters that can change
and trigger the phase transition of droppable gels include
pH, temperature and ionic strength. Examples of potential
ophthalmic droppable gels reported in the literature include:

gelling triggered by a change in pH CAP latex

2,22,26
, cross-
linked polyacrylic acid and derivatives such as carbomers
and polycarbophil;

gelling triggered by temperature change poloxamers

2,22,26
,
methyl cellulose and Smart Hydrogel;

gelling triggered by ionic strength change Gelrite

2,2628
and alginate29.
In addition, a combination system has been reported
30,31
, in
which a thermally induced gelling material (methyl cellulose)
is physically combined with a pH-induced gelling material
(carbomer). The resultant system was able to achieve the in situ-
gelling property with less total polymer content.
Among all droppable gel systems, only Gelrite has been used
in the formulation of a prescription ophthalmic product that
has been approved by the US Food and Drug Administration
(FDA). Timoptic-XE

, timolol in Gelrite, was developed as an

extension of the Timoptic ophthalmic solution product line.
Through formulation into a droppable gel, the dosing fre-
quency of Timoptic was reduced from twice-a-day to once-
a-day and without the loss of ease of instillation.
Other droppable gel systems under assessment for clinical
applications are DuraSite

and Smart Hydrogel. DuraSite

uses polycarbophil, a cross-linked polyacrylic acid, to achieve

the desired in situ-gelling property
32
. Marketed drugs, such
as levobunolol, pilocarpine, and diclofenac, have been refor-
mulated in DuraSite

, and these new formulations are being

evaluated in clinical studies. In addition, AquaSite, an over-
the-counter tear product formulated in DuraSite

, has been
made commercially available.
Smart Hydrogel is a graft co-polymer of polyacrylic acid
and a poloxamer
33
. As with poloxamers, Smart Hydrogel has
the property of reverse thermal gelation, but it requires only a
low polymer concentration (1 to 3%) to gel at body tempera-
ture. In addition, Smart Hydrogel is bioadhesive because of
its polyacrylic acid component.
As described, gel systems are limited in their ability to both
improve bioavailability and facilitate sustained drug release.
Once- or twice-a-day dosing is the typical expectation from
these gel systems.
Oil-in-water emulsions
Emulsion technology has been applied to oral, parenteral and
dermatological dosage forms for many years. However, it
wasnt until the 1980s that it was used in the development
of ophthalmic formulations. This delay can be attributed to
the following:
331
PSTT Vol. 1, No. 8 November 1998 reviews research focus

the difficulty in manufacturing sterile emulsions with

acceptable sterility assurance and globule size control;

the difficulty in achieving acceptable long-term stability;

the difficulty in achieving a high comfort level.

Most of the early investigational ophthalmic emulsions mimic
the formulations of parenteral emulsions, and use phospholipids
and pluronics as emulsifiers. Because phospholipids are sensitive
to oxygen, antioxidants are incorporated into these emulsion
formulations to improve their shelf-life. Even with antioxidants,
however, the phospholipid-containing emulsions are still limited
in room temperature stability.
The in vivo data obtained from studies of these early formu-
lations confirm that emulsions can be effective topical ophthalmic
drug delivery systems
34,35
, with a potential for sustained drug re-
lease
36,37
. Naveh and co-workers noted that the intraocular-
pressure-reducing effect of a single, topically administered dose
of pilocarpine emulsion lasted for more than 29 h in albino rab-
bits, whereas that of the generic pilocarpine solution lasted only
five hours
36
.
Oil-in-water emulsions are particularly useful in the delivery
of water-insoluble drugs. Previously, ointments and suspensions
were the only two available options, but the former suffered from
poor patient acceptance because of blurred vision and matted
eyelids, and the latter appeared to have problems with particle
irritation, poor bioavailability, and changes in polymorphism and
particle size upon storage. The newly developed oil-in-water
emulsion offers a third option that has the advantages of an oint-
ment without its drawbacks.
In the oil-in-water emulsion, the water-insoluble drug is solu-
bilized in the internal oil phase, thereby remaining in the pre-
ferred solution state. By keeping the drug in solution, the issue of
potential absorption because of slow dissolution of solid drug
particles is avoided. In addition, the blurred vision caused by oils
is minimized by the water in the external phase. Furthermore, the
concentration of the drug in the oil phase can be adjusted to max-
imize thermodynamic activity, thus enhancing drug penetration.
As technologies have advanced, the obstacles preventing the
development of ophthalmic emulsions have gradually been
removed. The improvements in both machinery and aseptic
processing allow for the reproducible manufacture of a sterile
product with greater assurance than was previously possible.
Moreover, new types of emulsifiers that are safe, non-irritating
and chemically unique have become available. Some of these
emulsifiers have demonstrated a remarkable ability to stabilize
emulsions. Using novel polymeric emulsifiers, a newly formu-
lated cyclosporine ophthalmic emulsion demonstrates excellent
room temperature stability and extremely low ocular
irritation
38,39
. This emulsion is in Phase III clinical studies for
the treatment of dry eye disease.
Liposomes
During the early stages of research into ophthalmic drug de-
livery, it was widely observed that positively-charged lipo-
somes have a greater affinity for ocular tissues than neutral or
negatively charged liposomes
19,40
. Recent studies have con-
firmed that positively charged liposomes are superior in their
ability to increase both precorneal drug retention and drug
bioavailability. For example, through the use of -scintigra-
phy, the precorneal drainage rate of positively-charged lipo-
somes was demonstrated to be slower than that of neutral or
negatively charged liposomes
41
. Similarly, the addition of
stearylamine to a liposomal preparation enhanced the corneal
absorption of dexamethasone valerate
42
, and the enhancement
was believed to be because of the positive charges conferred
by stearylamine onto the liposomes. Finally, the mydriatic re-
sponse in rabbits dosed with atropine containing positively-
charged multilamellar vesicles (MLV) liposomes was pro-
longed when compared to the response in animals dosed with
atropine containing neutral or negatively charged MLV
43
.
The greater affinity of positively charged liposomes for
ocular tissues has been attributed to the negatively charged
corneal surface: the corneal epithelium is thinly coated with
negatively charged mucin to which the positive surface-
charge of the liposomes may adsorb more strongly
44
.
The use of bioadhesive polymers to coat liposomes is a
strategy designed to prolong the precorneal retention of lipo-
somes
45,46
. Using Carbopol 1342 as the bioadhesive polymer
at a pH of 5, coated pilocarpine-containing liposomes were
shown to produce both a larger area under the meiotic inten-
sity curve and a longer duration of action, when compared
with the non-coated liposomes
45
. In a similar study conducted
by Davies et al.
46
, the Carbopol-coated liposomes were retained
in the precorneal area longer than the non-coated liposomes,
but the mydriatic response of the entrapped drug, tropi-
camide, was not enhanced or prolonged.
The potential of liposomes as a topical ophthalmic drug de-
livery system is limited because of their limited drug loading ca-
pacity and stability. In addition, large-scale manufacture of sterile
liposomes is expensive and technically challenging. Recently,
researchers have begun to explore the potential of liposome
technology in posterior drug delivery. By injecting liposome
preparations subconjunctivally or intravitreally, a sustained
drug release effect has been demonstrated in some studies
22,47
.
Nanoparticles
When appropriately formulated for ophthalmic drug delivery,
nanoparticles may provide sustained drug release and pro-
longed therapeutic activity. To achieve this, the particles must
be retained in the ocular cul-de-sac after topical adminis-
tration, and the entrapped drug must be released from the
332
PSTT Vol. 1, No. 8 November 1998 reviews research focus
particles at an appropriate rate. If the drug leaks out of the par-
ticles too fast, then there will be little sustained drug release.
On the other hand, if the release is too slow, then the concen-
tration of the drug in the tears may be too low to allow ade-
quate drug penetration into ocular tissues.
To enhance particle retention in the ocular cul-de-sac, it is
highly desirable to fabricate the particles with bioadhesive ma-
terials. Without bioadhesion, nanoparticles are eliminated
from the precorneal site almost as quickly as aqueous solu-
tions. For example, in an ocular disposition study conducted
by Wood et al.
48
, nanoparticles fabricated using poly-hexyl-2-
cyanoacrylate were rapidly cleared from the precorneal site
after topical administration, and only approximately 1% of the
instilled dose adhered to corneal and conjunctival surfaces.
Biodegradation is also a highly desirable property for the
fabrication materials of nanoparticles. This is primarily due to
safety concerns.
In recent studies of nanoparticles, the most commonly used
polymers are various poly(alkylcyanoacrylates), poly--capro-
lactone and polylactic-co-glycolic acid, all of which are
biodegradable polymers that undergo hydrolysis in tears. In
one study conducted by Marchal-Heussler et al.
49
, the efficiency
of betaxolol delivery was compared among nanoparticles
made of these three types of polymers. The results showed that
poly--caprolactone nanoparticles yielded the highest pharma-
cological activity, and the agglomeration of these nanoparticles
in the conjunctival sac was thought to be responsible for the
good level of performance. Using poly--caprolactone as the
fabrication polymer and carteolol as the entrapped drug,
Marchal-Heussler and coworkers further demonstrated that
nanocapsules outperformed nanospheres in the reduction of
intraocular pressure
50
. This improvement was attributed to the
fact that carteolol remained in solution in the oil-core of the
nanocapsules, and therefore was more readily available for
penetration into ocular tissues.
Physicochemical properties of nanoparticles also have an
effect on their efficiency in drug delivery. Among the param-
eters studied, both the surface charge and the binding of the
drug to the particles were found to be more important than
the drug-loading percentage
51
.
Coating nanoparticles with positively charged bioadhesive
polymers is a recent strategy designed to further enhance the
interaction between nanoparticles and the negatively charged
corneal surface. The positively charged coating, however, may
not necessarily improve the performance of the nanoparticles.
A recent publication reported that chitosan-coated nanocap-
sules doubled the ocular bioavailability of indomethacin when
compared with the non-coated preparation, whereas poly-
L-lysine-coating failed to improve bioavailability
52
. The authors
suggested that the positive surface charge was not critical
because both chitosan and poly-L-lysine conferred positive
charges onto the particles. It was suggested instead that the
specific nature of chitosan was responsible for the bioavailabil-
ity improvement.
The efficiency of drug delivery upon co-administration of
nanoparticles with viscous or bioadhesive polymers was also
explored through the use of pilocarpine-loaded albumin
nanoparticles
53
. The results showed that most of the polymers
studied had a minimal impact on delivery efficiency. The co-
administration of mucin, however, resulted in a significant
improvement in drug bioavailability.
The potential use of nanoparticles as an ophthalmic drug
delivery system has been demonstrated in numerous studies
for either hydrophobic or hydrophilic drugs (see Refs 21 and
22 for review). Despite the promising in vivo results, many is-
sues must be resolved before an ophthalmic product can
be developed using this technology. The major development
issues for nanoparticles include formulation stability, control
of particle size, control of the rate of drug release, and large-
scale manufacture of sterile preparations. Similar to the research
direction taken for liposomes, research has begun to focus
on the use of this technology in posterior drug delivery.
Microparticulates
Microparticulates are drug-containing, micron-sized poly-
meric particles suspended in a liquid medium. Drugs can be
physically dispersed in the polymer matrix or covalently bound
to the polymer backbone
54
. Upon topical instillation, the parti-
cles reside in the ocular cul-de-sac, and the drug is released
from the particles through diffusion, chemical reaction,
and/or polymer degradation. Microparticulates are larger than
nanoparticles, which may make them better suited for sus-
tained or controlled drug release, but the larger size may make
them less tolerable.
Biodegradation, bioadhesion and biocompatibility are the
desired properties for the fabrication polymers of ophthalmic
microparticulates. The following are examples of published
biodegradable microparticulates, in which the in vivo efficacy
performance is reportedly superior to that of the correspond-
ing conventional dosage forms:

microspheres of methylprednisolone chemically linked to

hyaluronate esters
55
;

pilocarpine-loaded albumin or gelatin microspheres

56
;

acyclovir-loaded chitosan microspheres

57
.
An interesting system that uses collagen pieces or particles has
recently been developed. The Collasomes system was developed
to replace the collagen shield, a type of soluble ophthalmic
insert that offered sustained drug release, but caused blurred
333
PSTT Vol. 1, No. 8 November 1998 reviews research focus
vision and was difficult to insert
22
. Collasomes efficiently
deliver hydrophilic molecules, and are well-tolerated by
patients
58
. Technically they are not microparticulates because
their size is in the millimeter range, but they can reportedly be
instilled as an eye drop.
Of all the available ocular drug delivery approaches,
microparticulates and solid inserts are the only two technolo-
gies that have the potential to achieve truly controlled and sus-
tained drug release. Microparticulate technology has the ad-
vantage of superior patient acceptability because the particles
are suspended in a medium that can be topically administered
as an eye drop, or intraocularly administered as an injection.
However, the manufacture and control of large-scale manufac-
turing of sterile microparticulates is very challenging and ex-
pensive. Therefore the economic value of microparticulates for
commercialization is significantly reduced.
To date, only one microparticulate ophthalmic prescription
product, Betoptic

S, is on the US market. By binding betaxolol

to ion-exchange resin particles, Betoptic

S retards drug release

in the tear and enhances drug bioavailability. Betoptic

S 0.25%
is bioequivalent to Betoptic Solution 0.5% in lowering intra-
ocular pressure. By reducing the drug strength by half, and
slowing down the drug-release rate in tears, Betoptic

S signifi-
cantly improves the ocular comfort of Betoptic Solution
59
.
Posterior drug delivery
Intravitreal drug injection is the current therapy for posterior
segment disorders. The procedure is associated with a high risk
of complications, particularly when frequent, repeated injec-
tions are required. Thus, sustained-release technologies are
being proposed
3,60
, and the possible benefits of using lipo-
somes, nanoparticles or microparticulates in intravitreal injec-
tions are under current investigation for posterior drug deliv-
ery. Initial results indicate that these technologies do retard
drug release after injection, but they introduce other problems,
such as intravitreal toxicity of the drug carriers, interference
with vision and difficulties in the large-scale manufacture of
sterile preparations
3,60
.
Alternative routes of administration include subconjunctival
injection of sustained-release drug carriers, and the use of im-
plants; that is, osmotic minipumps
3,61
, biodegradable scleral
plugs
62
, and biodegradable poly-(caprolactone) matrices
63
. In
1996, Vitrasert
61
, an intravitreal implant of ganciclovir, re-
ceived FDA approval for the treatment of cytomegalovirus
retinitis. See Ref. 3 for a comprehensive review of intraocular
delivery systems.
Delivery of proteins and peptides
Although small peptides (molecular weight less than 5000)
can be absorbed via the topical, ocular route into ocular tissues
or the circulatory system, larger peptides and proteins cannot
be absorbed without the aid of penetration enhancers
64,65
,
which have not been well-accepted because of local toxicity.
Recently, less irritating ocular penetration enhancers have been
found for insulin. They can reportedly improve the absorption
of insulin via the topical, ocular route and achieve a biologi-
cally effective systemic level of insulin
63
. Cytochalasin B, a
cytoskeletal modulator, is another potential enhancer of
ocular penetration with low toxicity
64
.
Non-aqueous vehicle
A non-aqueous, comfortable vehicle is desired for topical
ophthalmic drug delivery because of drug degradation trig-
gered by water or the premature leakage of the drug from the
delivery system in the presence of water. Typically, lipophilic
vehicles (that is, mineral oils and vegetable oils) have been
used, but they are poorly accepted by patients due to blurred
vision and matted eyelids. Reconstitution with water prior to
use is unacceptable because of cost and patient compliance
issues.
Perfluorocarbons, or fluorinated silicone liquids, have re-
cently been suggested as good non-aqueous vehicles for topi-
cal ophthalmic drug delivery
66,67
. They are chemically and bio-
logically inert, and have a low surface tension, excellent
spreading characteristics and close-to-water refractive indices.
Perfluorocarbons have been studied for years as a blood substi-
tute
68
; minimal systemic toxicity is expected via the topical
route.
Future directions
A few new products have been commercialized as a result of
the research into ophthalmic drug delivery. The performance
of these new products, however, is still far from being perfect.
An ideal system should be able to achieve an effective drug
concentration at the target tissue for an extended period of
time, while minimizing systemic exposure. In addition, the
system should be both comfortable and easy to use. Patient
acceptance will continue to be emphasized in the design of
future ophthalmic drug delivery systems.
Major improvements are required in each of the technol-
ogies discussed in this review. Some approaches are relatively
easy to manufacture, but are limited in their ability to provide
sustained drug release. Other approaches are promising with
regard to sustained drug release, but are difficult to manufac-
ture. Stability is a major issue with particulates and liposomes.
A reasonable strategy to circumvent the drawbacks of indi-
vidual technologies is to combine technologies. Reported ex-
amples include liposomes and nanoparticles in droppable gels
and liposomes and nanoparticles coated with bioadhesive
polymers.
334
PSTT Vol. 1, No. 8 November 1998 reviews research focus
Much remains to be learned regarding the delivery charac-
teristics of emulsions and perfluorocarbons. Combining these
two new technologies with earlier delivery approaches may be
another interesting area for exploration.
Acknowledgement
The author wishes to thank Orest Olejnik for his support and
critical review of this manuscript.
References
01 Lang, J.C. (1995) Adv. Drug Deliv. Rev. 16, 3943
02 Desai, S.D. and Blanchard, J. (1995) in Encyclopedia of Pharmaceutical
Technology (Swarbrick, J. and Boylan, J.C., eds), pp. 4375, Marcel
Dekker, New York, NY, USA
03 Peyman, G.A. and Ganiban, G.J. (1995) Adv. Drug Deliv. Rev. 16, 107123
04 Lee, V.H.L. and Robinson, J.R. (1986) J. Ocul. Pharmacol. Ther. 2, 67108
05 Mezei, M. and Meisner, D. (1993) in Biopharmaceutics of Ocular Drug Delivery
(Edman, P., ed.), pp. 91104, CRC Press, Boca Raton, FL, USA
06 Shell, J.W. (1985) Drug Dev. Res. 6, 245261
07 Urtti, A. and Salminen, L. (1993) Surv. Ophthalmol. 37, 435456
08 Chang, S-C. and Lee, V.H.L. (1987) J. Ocul. Pharmacol. Ther. 3, 159169
09 McMahon, C.D. et al. (1979) Am. J. Ophthalmol. 88, 736738
10 Chrai, S.S. et al. (1973) J. Pharm. Sci. 62, 11121121
11 Adler, C.A., Maurice, D.M. and Paterson, M.E. (1971) Exp. Eye Res. 11,
3442
12 Patton, T.F. and Robinson, J.R. (1975) J. Pharm. Sci. 64, 13121315
13 Zignani, M., Tabatabay, C. and Gurny, R. (1995) Adv. Drug Deliv. Rev. 16,
5160
14 Liaw, J. and Robinson, J.R. (1993) in Ophthalmic Drug Delivery Systems
(Mitra, A.K., ed.), pp. 369381, Marcel Dekker, New York, NY, USA
15 Smolin, G. et al. (1981) Am. J. Ophthalmol. 91, 220225
16 Singh, K. and Mezei, M. (1983) Int. J. Pharm. 16, 339344
17 Stratford, R.E. et al. (1983) Int. J. Pharm. 13, 263
18 Benita, S. et al. (1984) J. Microencapsulation 1, 203206
19 Singh, K. and Mezei, M. (1984) Int. J. Pharm. 19, 263269
20 Lee, V.H.L., Takemoto, K.A. and Iimoto, D.S. (1984) Curr. Eye Res. 3,
585591
21 Zimmer, A. and Kreuter, J. (1995) Adv. Drug Deliv. Rev. 16, 6173
22 Le Bourlais, C. et al. (1998) Prog. Retinal Eye Res. 17, 3358
23 Saettone, M.F. and Salminen, L. (1995) Adv. Drug Deliv. Rev. 16, 95106
24 Saettone, M.F. (1993) in Biopharmaceutics of Ocular Drug Delivery (Edman, P.,
ed.), pp. 6179, CRC Press, Boca Raton, FL, USA
25 Bawa, R. (1993) in Ophthalmic Drug Delivery Systems (Mitra, A.K., ed.),
pp. 223259, Marcel Dekker, New York, NY, USA
26 Gurny, R. et al. (1987) in Ophthalmic Drug Delivery. Biopharmaceutical,
Technological and Clinical Aspects (Vol. 11) (Saettone, M.S., Bucci, G. and
Speiser, P., eds), pp. 2736, Fidia Research Series, Liviana Press, Padova,
Italy
27 Rozier, A. et al. (1989) Int. J. Pharm. 57, 163168
28 Sanzgiri,Y.D. et al. (1993) J. Control. Release 26, 195201
29 Cohen, S. et al. (1997) J. Control. Release 44, 201208
30 Joshi, A., Ding, S. and Himmelstein, K. (1993) US Patent 5,252,318,
12 October
31 Kumar, S., Haglund, B.O. and Himmelstein, K. (1994) J. Ocul. Pharmacol.
Ther. 10, 4756
32 Patel, R. et al. (1994) US Patent 5,340,572, 23 August
33 Ron, E.S. et al. (1996) Proc. Int. Symp. Control. Release Bioact. Mat. 23
34 Garty, N. et al. (1994) Invest. Ophthalmol.Vis. Sci. 35, 2175
35 Muchtar, S. et al. (1992) Ophthalmic Res. 24, 142149
36 Naveh, N., Muchtar, S. and Benita, S. (1994) J. Ocul. Pharmacol. Ther. 10,
509520
37 Bar-Ilan, A. et al. (1994) Reg. Immunol. (USA) 6, 166168
38 Ding, S., Tien, W. and Olejnik, O. (1995) US Patent 5,474,979,
12 December
39 Ding, S. and Olejnik, O. (1997) Pharm. Res. 14, S41, Abstract 1127
40 Schaeffer, H.E. and Krohn, D.K. (1982) Invest. Ophthalmol.Vis. Sci. 22,
220227
41 Fitzgerald, P. et al. (1987) Int. J. Pharm. 40, 8184
42 Taniguchi, K. et al. (1988) J. Pharmacobiodyn. 11, 607611
43 Meisner, D., Pringle, J. and Mezei, M. (1989) Int. J. Pharm. 55,
105113
44 Shek, P.N. and Barber, R.F. (1987) Biochim. Biophys. Acta 902, 229236
45 Durrani, A.M. et al. (1992) Int. J. Pharm. 88, 409415
46 Davies, N.M. et al. (1992) Pharm. Res. 9, 11371144
47 Meisner, D. and Mezei, M. (1995) Adv. Drug Deliv. Rev. 16, 7593
48 Wood, R.W. et al. (1985) Int. J. Pharm. 23, 175183
49 Marchal-Heussler, L. et al. (1992) STP Pharm. Sci. 2, 98104
50 Marchal-Heussler, L. et al. (1993) Pharm. Res. 10, 386390
51 Marchal-Heussler, L. et al. (1990) Int. J. Pharm. 58, 115122
52 Calvo, P., Vila-Jato, L. and Alonso, M.J. (1997) Int. J. Pharm. 153,
4150
53 Zimmer, A.K. et al. (1995) J. Control. Release 33, 3146
54 Joshi, A. (1994) J. Ocul. Pharmacol.Ther. 10, 2945
55 Kyyronen, K. et al. (1992) Int. J. Pharm. 80, 161169
56 Leucuta, S.E. (1989) Int. J. Pharm. 54, 7178
57 Genta, I. et al. (1997) J. Pharm. Pharmacol. 49, 737742
58 Kaufman, H.E. et al. (1994) J. Ocul. Pharmacol.Ther. 10, 1727
59 Jani, R. et al. (1994) J. Ocul. Pharmacol.Ther. 10, 5767
60 Metrikin, D.C. and Anand, R. (1994) Curr. Opin. Ophthalmol. 5, 2129
61 Ashton, P. et al. (1994) J. Ocul. Pharmacol.Ther. 10, 691701
62 Hashizoe, M. et al. (1994) Arch. Ophthalmol. 112, 13801384
63 Borhani, H., Rahimy, M.H. and Peyman, G.A. (1993) Invest. Ophthalmol.
Vis. Sci. 34(Suppl.), 1488
64 Harris, D., Liaw, J-H. and Robinson, J.R. (1992) Adv. Drug Deliv. Rev. 8,
331339
65 Chiou, G.C.Y. (1994) J. Ocul. Pharmacol.Ther. 10, 9399
66 Meadows, D.L. (1992) US Patent 5,173,298, 22 December
67 Meadows, D.L. (1996) US Patent 5,518,731, 21 May
68 Riess, J.G. and Krafft, M.P. (1997) Artif. Cells Blood Substitutes and
Immobilization Biotechnol. 25, 4352
335
PSTT Vol. 1, No. 8 November 1998 reviews research focus