Clinical pharmacokinetics of the eye

Proctor Lecture
Saiichi Mishima

Pharmacokinetics of instilled drugs in the human eye is reviewed. The behavior of drugs in the
conjunctival cul-de-sac is discussed, and the loss rates with various vehicles are estimated.
Kinetics of intraocular drug penetration follows the same pattern in human and rabbit eyes.
From results of rabbit experiments, various pharmacokinetic coefficients are computed, includ-
ing the permeability of the corneal epithelium. Also, with use of published data in human
aqueous, the corneal permeability, apparent absorption, and elimination rate constants are
calculated for some drugs in the human eye. From the anesthetic response of the cornea the
apparent elimination rate constants of the surface anesthetics are obtained. The pupil response
is converted to a response parameter proportional to the biophase drug concentration. The time
course of its changes conforms with the kinetics of intraocular drug penetration, and the
apparent absorption and elimination rate constants are computed for various drugs. The latter
constant becomes smaller with increased ocular pigmentation. Use of the relative bioavailabil-
ity concept permits comparison of drug absorption efficiency among various vehicles. Phar-
macokinetic coefficients are also obtained for the cycloplegic responses. The intraocular pres-
sure response is analyzed, and it is suggested that the reduction ratio in the outflow pressure,
outflow resistance, and the aqueous formation rate be used for dose-response studies. The rate
of effect disappearance is defined and is given for three beta-adrenergic blocking agents.
INVEST OPHTHALMOL VIS SCI 21:504-541, 1981.)

Key words: pharmacokinetics, tears, corneal permeability, pupil,

accommodation, anesthetics, intraocular pressure, drugs, eye

T he purpose of medical treatment is to this principle, knowledge of the pharmacoki-

netics of instilled drugs in the human eye
achieve the desired effects of a given drug
and to avoid its untoward side effects. This would give a basis for logical use of ophthal-
may be realized by appropriate choice of the mic drugs.
therapeutic regimen based on the knowledge The availability of instilled drugs in the eye
of absorption, metabolism, and elimination of is * modified principally by three factors: (1)
the drug, i.e., its pharmacokinetics. 1 ' 2 On drug kinetics in the conjunctival cul-de-sac,
(2) the permeability of the cornea, and (3) the
From the Department of Ophthalmology, University of rate at which the drug is eliminated from the
Tokyo, School of Medicine.
eye. The drug kinetics in the conjunctival
This work has been supported by research grants
344602, 437037, and 587098 from the Ministry of Ed- cul-de-sac of the human subjects has been ap-
ucation and Culture of Japan. proached through studies using fluorescein3"6
Delivered April 30, 1981, at the Annual Meeting of The or sodium pertechnetate. 7 " 10 Intraocular
Association for Research in Vision and Ophthalmol- penetration of instilled drugs takes place
ogy, Sarasota, Fla.
mainly through the cornea, 11 " 13 and a volu-
Reprint requests: Saiichi Mishima, M.D., Department
of Ophthalmology, University of Tokyo School of Med- minous literature has accumulated on drug
icine, 7-3-1, Hongo, Bunkyo-ku, Tokyo, Japan, 113. penetration in rabbit eyes. 14 A series of works

504 0146-0404/81/100504+38$03.80/0 © 1981 Assoc. for Res. in Vis. and Ophthal., Inc.

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 505

of Robinson and co-workers15 22 has greatly physiologic turnover rate of 0.1 to 0.15 min ]
elucidated the pharmacokinetics of pilocar- the rate of tear flow is about 1 /x\ min" 1 . The
pine and fluorometholone in the rabbit eyes, maximum quantity of fluid that can be con-
but it is not readily apparent whether these tained in the cul-de-sac without overflow3 is
results can be directly extrapolated to the about 30 fx\. Since one drop of collyrium dis-
human eye. It is difficult to attempt to study pensed from a conventional ophthalmic bot-
this aspect in the human eye, and the results tle is 40 to 50 fi\, about 20 fi\ of the drug is
of a few studies 23 " 28 are not enough to spilled out from the lids at the time of instilla-
achieve a systematic knowledge of its phar- tion. The increased volume of fluid in the
macokinetics. For this reason the intraocular cul-de-sac is quickly delivered into the lacri-
penetration of instilled fluorescein was cho- mal drainage system by the pumping action of
sen as a model of drug transfer kinetics29 from the canaliculi associated with the blink move-
which the basic theory of ocular pharmacoki- ment. 35 The precorneal tear film is a stagnant
netics was developed. 30 fluid layer with a thickness 34 ' 36 of about 7 to 9
There are several ocular responses to drugs fim that is spread over the corneal epithelium
that can be quantified with reasonable accu- by a coacervate of mucin and is stabilized by
racy so as to make possible investigations on the superficial oily layer formed by meibo-
the dose-response relationship and the time mian gland secretion. 37 Therefore its mixing
course of the response changes. The mag- with the marginal tear fluid after drugs are
nitude of the response at a given time may be instilled takes place only by blink movements,
considered to be a function of the amount of which at the same time carry the instilled drug
the drug concurrently present in the bio- away from the cul-de-sac.
phase of the target tissue, lj 2 and therefore Because of the above mechanism, the tear
the time course of the response changes may film saturation with the instilled drug is in-
be analyzed according to the principles of complete. 6 After instillation of fluorescein so-
pharmacokinetics. This noninvasive method lution, the initial tear concentration, i.e., de-
was found to be fruitful in the pharmacoki- gree of saturation, increased slightly on in-
netic studies of instilled drugs in the human creasing the instilled volume from 5 to 20 /A,
eye, 6 ' 31> 32 and the results could be well ac- but further volume increase failed to raise the
counted for by the theory of transcorneal drug degree of saturation over a 46% level of the
transfer; they are also consistent with the instilled drug concentration. 6 Similarly, in
chemical analyses made in rabbit eyes. Sev- rabbit eyes an increase in the instillation vol-
eral pharmacokinetic coefficients computed ume was shown to result in a greater loss rate
from the studies of ocular responses may be from the tears but not to raise intraocular
useful for clinical purposes. 33 An attempt was drug penetration. 15 ' 16 On this basis, the in-
made in the present article to review these stillation volume of about 20 /xl has been
aspects of clinical pharmacokinetics in the thought to be adequate, 33 and this will be
human eye, with particular reference to topi- discussed later in relation to the intraocular
cally applied drugs. bioavailability of drugs.
The above phenomena imply that rapidly
Penetration kinetics of instilled drugs repeated instillations of a drug do not con-
Behavior of drugs in conjunctival cul-de- tribute to the increase of its bioavailability to
sac. An instilled drug penetrates the eye by the eye. 16 Furthermore, when two or three
absorption across the cornea from the pre- drugs are instilled at short intervals, they
corneal tear film.11"13 Thus the mixing and mutually dilute each other and reduce the
kinetic behavior of drugs in the tears have a availability of each of them. In addition, the
direct bearing on the efficiency of drug ab- increase in the drainage into the lacrimal sac
sorption by the eye. and the nose lead to an increase in the sys-
The conjunctival cul-de-sac normally con- temic absorption of the drug, hence the sys-
tains about 7 to 9 /x\ of tears3* 34; with a temic side effects. These systemic effects of

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
506 Mishima October 1981

the drug to mix with the precorneal tear film.

TEAR CORNEA The drug mixed in the tears will then be
diluted by tear secretion, and the concentra-
tion decrease is related to time by a single
exponential,3 i.e.

where C d and C d0 are the drug concentra-

tions in the tears at a given time and at time
zero, respectively, and a is the rate constant
of loss from the tears. The rate constant of
loss after instillation of fluorescein solution4' 6
ranges from 0.05 to 0.3 min"1, depending on
the degree of lacrimation. When the solution
causes stinging, the rate constant can be as
high 3 as 1 to 2 min"1.
There are numerous papers reporting that
the use of viscous vehicles improves the con-
tact of drugs with the cornea, thereby en-
hancing intraocular penetration and effects.
The increased contact of a drug with the cor-
nea may involve either or both of (1) higher
initial tear film saturation with the drug and
Fig. 1. Model of drug uptake by the cornea from (2) reduction of the rate of loss in the tears.
the tears. For explanation of symbols see text. Increase in the tear film saturation was noted
above, but the reduction of the loss rate ap-
topical ophthalmic drugs are of particular pears to be only slight; with sodium pertech-
concern in pediatric ophthalmology,38 and it netate, Tc99m, the loss rate was reported to
is important to take measures to reduce sys- be 0.06 to 0,1 min" 1 with viscous vehicles. 7 " 9
temic drug absorption. It is therefore neces- In rabbit eyes the loss rate decreased as the
sary to dispense appropriate drop size. Mau- viscosity of the solution increased, but it was
rice* suggested the development of a device stabilized at the level of about 0.1 min" 1 in
permitting instillation of precise volumes the viscosity range of 13 to 15 centistokes,
onto the cornea, which would be useful in and no further decrease was observed by in-
clinical practice. creasing the viscosity.15- 18> 20 This is in good
Addition of viscous vehicles (e.g., methyl- agreement with the observation in human
cellulose) could raise the initial tear film sat- eyes.
uration when the viscosity increases up to 15 Ointment is often used as the vehicle and
to 20 centistokes. 6 Further increase in the may be classified into two types: the simple
viscosity, however, failed to saturate the tear base (e.g., castor oil) and the compound
film any further, the limit being about 80% of base.39* 40 The compound bases are more
the instilled concentration. The higher tear often used, and the oil-in-water type emul-
film saturation that follows the addition of sion is particularly effective,39' 4i since the
viscous vehicles is probably due to slowing drug with low water solubility is contained at
the initial drainage of solution4' 6 and also to a high concentration in the oil phase, from
the prolonged retention of the thickened tear whence it is released into the water phase
film5; both mechanisms give more time for that becomes continuous with the tears. 21
After instillation these ointments break up
*Personal communication. into small droplets and remain in the cul-de-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 507

sac for a long time, acting as a depot of the

drug21' 39; some may be carried into the lac-
rimal drainage system and pass into the 10
nose. 42 The loss rate of tetracycline25 given in ^^ tV
castor oil is about 0.01 min" 1 and in ointment
is about 0.005 to 0.006 min"1. Z
Kinetics of intraocular penetration o
of drugs
Drug uptake by the cornea. No experimen-
<
CC 10
tal data are available for the corneal uptake of
LU
various drugs in the human eye, and we must
rely on the results obtained in rabbit eyes. To
O
z
i
compare the corneal uptake of various drugs o
o
it is necessary to compute tear-cornea drug
transfer either in terms of the transfer coeffi-
cient or of the permeability of the epithelial 10
barrier. Several papers in the literature have o
reported the corneal concentration changes z
shortly after drug instillation, and the per- LU
meability of the epithelial barrier may be cal- m
culated from their data. The permeability of
0 1 2 3 4 5 6
the rabbit and human corneas will then be
compared. TIME (HOURS)
During the very early period after instilla- Fig. 2. Time courses of the concentration changes
tion the drug is first mixed with the pre- of the drug in the cornea (upper curve) and in the
corneal tear film and is then taken up by the anterior chamber (lower curve) after instillation of
cornea. This process is depicted in Fig. 1, 50 /x\ of 1% 14C-befunolol in the albino rabbit eye.
and the transfer equation may be written Extrapolation of the corneal concentration curve
using equation 1 as gives [Cc]0. From the anterior chamber curve, the
apparent absorption constant, B, and the apparent
elimination rate constant, A, are obtained.
^jjf = kc.tlc C d0 e-« l - kc C c (2)
peak time occurs shortly after instillation.
where C c is the corneal concentration, C d0 is
The peak time, in rabbit eyes ranges from 5
the initial tear concentration, a is the rate
min22' 43 to about 30 min, 47 and this must be
constant of loss from the tears, k c d c is the
caused by variation of the loss rate in the
tear-cornea transfer coefficient in reference
tears. It is very difficult to accurately deter-
to the corneal volume, kc is the rate constant
mine the corneal concentrations in a very
of loss from the cornea, and t is the time.
short time after instillation, and the results of
Integration of the equation yields
only a few published papers could be ana-
lyzed with equation 3. In most published re-
(3)
a — kr sults, the concentration decreases in the cor-
nea after the peak time are shown (Fig. 2);
The peak time of the corneal concentration therefore, extrapolation of the curve to time
(Tc) is given by zero is possible to give the value of [C c ] 0 .
Remembering that a is far greater than kc,
(4)
one can write

Since the loss rate in the tears, a, is far

[CJ 0 = CJa
greater than that from the cornea, kc, the

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
508 Mishima October 1981

Table I. Permeability of the epithelial Mo, may then be given by integrating the
barrier (Kep) of the rabbit cornea above equation
to various drugs
Mo = Kep Q Cd0/a (7)
Source of
Drugs (x 10~4cmhr-1) data (ref.) Denoting the corneal volume as Vc, Mo
should be given by [CC]OVC, and equations 5
Pilocarpine 5.2-7.3 45
and 6 give
7-19 22
9-18A 46
Prednisolone 3.2-5.6 47 kc.dc Vc _ [ C c ] 0 Vc a (8)
acetate 5.8B 48 Q QC d 0
7.2 49
3.7-5.1 50 Thus the values of Kep were calculated from
Prednisolone Na 2.5-3.4 51 the data published in the literature, assuming
phosphate 9.6 52
5.8B 48 that Vc is 0.08 ml, Q is 2 cm2, a is 6 hr" 1 , and
Dexamethasone 33 53 C d0 is equal to the concentration of the in-
12 54 stilled drug. The results are given in Table I.
11 49
3.7B 48
There are many uncertainties in the experi-
Fluorometholone 3 55 mental conditions, which necessitate making
6 49 assumptions, as above, and the values are
5.8B 48
Timolol 14, 38 56, 57
therefore only approximate. Nevertheless,
Befunolol 13 Takase et al. E the values of Kep for the same drug obtained
Bupranolol 9 58 from different sources appear to be in fair
Epinephrine 1 Uda et al. E
agreement. In addition, it can be seen that
Dipyvalyl-epi- 53 Udaetal.E
nephrine the epithelial barrier of the cornea is more
Fluriprofen 26 Takase et al.K permeable to lipid-soluble than to lipid-
6-Aminohexanoic 12C 59 insoluble drugs.
acid
Hetrazan 9.6 60 Drug transfer into the anterior chamber.
Sodium 8-24D 61 Because of the rapid loss of drugs from the
Urea 7.5C 62
D-xylose 1.7C 62 conjunctival cul-de-sac, the uptake by the
Sucrose 1.4C 62 cornea is practically finished a few minutes
A
Corneal concentration not available; calculation was made from after instillation. The drug absorbed by the
peak aqueous concentration using equation 12.
B
cornea is then transferred into the anterior
Results of in vitro studies for the whole layers of the cornea.
c
Results of in vivo studies for the whole layers of the cornea. chamber, where the drug is carried away by
D
E
Calculated by the author from in vivo studies. aqueous flow and by diffusion into the blood
Unpublished results.
circulation in the anterior uvea. The transfer
kinetics may be represented by a two-com-
partment open model shown in Fig. 3. The
On the basis of this equation, [C c ] 0 can be
transfer equations applicable to this situation
used to calculate the tear-cornea transfer co-
were given by Jones and Maurice 30 and are
efficient, kc-dc.
Alternatively, the drug uptake by the cor- dCc
(9)
nea in the very early period after instillation d t
may be expressed as follows, since the loss
from the cornea during this period is neg- dC a — \c
*^a.ca C
'-'c — L-
"^a.ac C
*-la — \c
"M) C
'-'a (10)
d t
ligible:
(6)
The steady-state distribution ratio, rca, be-
-JJ-KepQ tween the cornea and the aqueous may be
where M is the amount of drug transferred, given by
Kep is the permeability of the epithelial bar-
rier, and Q is the area of the cornea. The total
amount of the drug absorbed by the cornea, where k cac and k cca are the aqueous-cornea

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 509

E
oi

c
o

c 10
0)
o
c \B
o
u \

10
2 3 4 5 6
Time (hour)
Fig. 4. Time course of fluorescein concentration
changes in the human eye after instillation of 10%
solution. The corneal concentration is in the upper
curve and the aqueous concentration in the lower
curve. The latter permits calculation of A and B,
the apparent elimination and absorption rate con-
stants. gca, Cornea-aqueous concentration ratio in
the parallel decrease phase.

Fig. 3. Model of drug transfer fron the cornea into

A + B = kc.ca + ka.ca + k0 (13)
the anterior chamber. For explanation of symbols AB = k0 kc.ca (14)
see text.
An example of the time course of concentra-
and cornea-aqueous transfer coefficients in tion changes in the cornea and the anterior
reference to the corneal volume, k a a c and chamber in the albino rabbit after instillation
k a c a are the transfer coefficients of the same of 1% befunolol solution is illustrated in Fig.
meaning but in reference to the anterior 2, and Fig. 4 depicts fluorescein concentra-
chamber volume, and k0 is the coefficient of tion changes in the cornea and the anterior
loss from the anterior chamber. chamber of the human eye after instillation of
A solution of equations 9, 10, and 11 gives 10% solution. In both cases the curves are
similar, and equation 12 describes the time
o kc.c (12)
Ca = — e" course of concentration changes in the an-
V a (B - A)
terior chamber. The anterior chamber con-
where Mo is the initial amount of drug ab- centration increases initially, reaches a peak,
sorbed by the cornea, Va is the distribution and then decreases in an exponential man-
volume in the anterior chamber, and t is the ner. The slope fitted to the decreasing phase
time. The constants A and B are given below: gives A (Figs. 2 and 4); A is therefore called

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sri.
510 Mishima October 1981

Table II. Pharmacokinetic coefficients in the albino rabbit eye, computed from
data in isotope experiments
V
A B
Drug (hr~>) (hr~<) (ml) (hr~>) (hr~l) Source of data (ref.)
Pilocarpine — 1.4 6.3 46
2.9(1.2)* 3(7)* — 1.1 8.1 22
Prednisolone acetate 0.6 2.2 0.21 0.6 2.6 49
0.7 1.4 0.3 0.6 3.3 50
Dexamethasone 0.5 2.6 0.19 0.44 3 49
0.45 2.2 0.18 0.4 2.5 53
0.7 2.5 0.23 0.6 2.8 54
Fluorometholone 0.8 3 0.19 0.7 3.5 49
0.4 3.5 0.4 3.8 55
Timolol 0.66 3.5 0.27 0.6 3.8 56
Beftinolol 0.54 2.3 0.32 0.5 2.4 Takase et al.t
Bupranolol 0.7 2.8 0.26 0.64 3 58

A = apparent elimination rate constant; B = apparent absorption rate constant.

* Numbers in parentheses were calculated for the corneal epithelium,
t Unpublished results.

the apparent elimination rate constant. Figs. the time course of drugs in the cornea and
2 and 4 also show graphic calculation of B; aqueous, as shown in Fig. 2. Consequently,
B is called the apparent absorption rate con- analysis by equations 12, 15, and 16 was con-
stant. Both A and B may also be calculated by ducted and the values for k c x a , k0, Va (an-
fitting equation 12 to the actual data by a terior chamber distribution volume), A (the
computer. 6 apparent elimination rate constant), and B
The values of A and B are a mixed function (the apparent absorption rate constant) were
of the cornea-aqueous transfer coefficient and calculated. Similar analysis was also per-
the loss coefficient in the anterior chamber. formed on our own data, and the results are
However, these coefficients may be com- given in Table II.
puted from the values of A and B. After the The values of k c x a are in a comparable
peak time has passed, the corneal and aque- range for steroids and beta-adrenergic block-
ous concentrations decrease at a similar rate ing agents, but the value for pilocarpine is
(Figs. 2 and 4), and the concentration ratio large. Pilocarpine was shown to be accumu-
C c /C a remains nearly constant during this lated at a high concentration in the corneal
period; this ratio is denoted as gca. Jones and epithelium, whereas the concentrations- in
Maurice30 showed that the following relation- the stroma and aqueous were similar22; it was
ships hold: thought that the corneal epithelium acts as a
depot of the drug and that the release rate
*c.ca
(15) from the epithelium is a limiting factor for
- rca/gca) aqueous drug kinetics. On the other hand,
k0 = B (1 - rca/gca) (16) the concentrations of timolol and befunolol in
the corneal epithelium were only about
The steady-state distribution ratio, r ca , is 20% higher than that in the stroma. 56 The
about 1.7 for fluorescein in the human eye, 63 stromal concentration was very much higher
but the values for drugs are completely un- than the aqueous concentration, indicating
known. Therefore they are assumed to be that the endothelial barrier is the limiting fac-
unity. tor for aqueous kinetics of these beta-adren-
The data of several previous papers using ergic blocking agents. The discrepancy in
radioactive compounds to study drug pene- values for kc ca may be attributed to the above
tration in rabbit eyes allow construction of difference in the mode of drug release from

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 511

Table III. Pharmacokinetic coefficients in the human eye for various drugs, calculated from
actual measurements of aqueous concentrations
A B Peak time
Drug (hr~l) (hr~l) (hr) (X 10-4cmhr-1) Source of data (ref.)
Fluorescein 0.20 1.04 2.2 0.15 4,30
Tetracycline* 0.7 0.8 1.3 0.1-0.17 25
Prednisolone 1.4 1.5 0.7 2.5-5 26
Triamcinolone
Tetrahydrotriamcinolone 1.4 1.5 0.7 23
Pilocarpinet 0.3-0.48 1.2-2.4 PI. 6 5-7 27,28
Chloramphenicol 0.5 1.2 1.2 15 24
A = apparent elimination rate constant; B = apparent absorption rate constant (equation 12 and text).
*For ointment.
t Values determined from the pupil responses; Kep (the permeability of the epithelial barrier of the cornea) was calculated from the peak
concentration in the anterior chamber.

Modified from Nagataki S and Mishima S: Int Ophthalmol Clin 20(3):33, 1980.

the cornea to the anterior chamber. The val- kinetics between the cornea and the aqueous
ues of k0 are similar for various drugs and humor.
they are in good agreement with the values Figs. 2 and 4 indicate that the transcorneal
reported in the literature, which were de- drug transfer into the anterior chamber fol-
termined with various substances.64 This is lows the basic kinetics of first order. Indeed,
understandable, since the major part of k0 is in both rabbit17 and human 28 eyes, the aque-
the rate of substance loss from the anterior ous concentration increases in proportion to
chamber by aqueous outflow. the increase in the concentrations of applied
It is of interest that the distribution vol- drugs. There are several papers reporting in-
ume, Va, for steroids is similar to the actual tracameral penetration of drugs in the human
volume of the anterior chamber, i.e., 227 to eyes, and these published data were sub-
250 fx\,64> 65 and this may be related to the fact jected to the analysis described above. The
that the albino iris does not accumulate ste- data scatter in the human experiments is
roids; that is, the concentrations in the aque- greater than that seen in experiments with
ous and iris are similar.53' 66 On the other rabbit eyes, but the values of the apparent
hand, the distribution volumes for the tested elimination rate constant A and the apparent
beta-adrenergic blockers are larger than this, absorption rate constant B were estimated.
indicating that these drugs are bound to the On the basis of equation 12, the value of Mo
intraocular tissues, and indeed, they are was computed from the peak aqueous concen-
found at a higher concentration in the iris and tration and the peak time, assuming that kc ca
the ciliary body than that found in the aque- is close to A and Va is close to the physical
ous humor. 56 ' 58 * Conrad and Robinson67 per- volume of the anterior chamber in man,68' 69
fused the anterior chamber of rabbits and de- i.e., 0.175 ml. Once Mo is known, the value
termined the distribution volume for pilo- of Kep, permeability of the epithelial barrier,
carpine to be 575 fx\ in albino eyes and 760 /nl may be calculated with equation 7. The re-
in pigmented eyes. Equation 12 and this sults of these calculations32 are listed in Table
large distribution volume of pilocarpine did III. It is of interest to find that the values of
not fit well to the data shown by Sieg and Kep for prednisolone and pilocarpine in the
Robinson22; the reason is not clear but it may human cornea are in the 50% to 70% range of
be related to the special mode of pilocarpine the value for the rabbit cornea. Because the
human cornea is thicker than the rabbit cor-
nea, a lower value of Kep in the human eye
*Takase M and Araie M: unpublished results. may be expected. In addition, we may say

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
512 Mishima October 1981

o 10
I
o
10'

10

0 1 2 3 4 5 %f/ 24
o 10 T 1 ME ( H O U R )
s:
Fig. 6. Befunolol concentration changes in the eye
0 1 2 3 4 5 6 of pigmented rabbit after instillation of 1% solu-
tion, A, Cornea; o, aqueous humor; m, iris; •,
T I ME ( H O U R ) ciliary body.
Fig. 5. Befunolol concentration changes in the al-
bino rabbit eye after instillation of 1% solution, A,
Cornea; o, aqueous humor; •, iris; •, ciliary body. traocular pharmacokinetics of this drug. 70
Several drugs are known to be bound to mel-
anin, and they are accumulated more readily
that the ocular penetration studies on the by pigmented than by albino tissues, 73 " 79
rabbit eye and their results are not far from which influences their pharmacokinetics in
what we can expect for the human eye and the eye. For example, Fig. 5 illustrates in
that these studies are useful, provided an ap- albino rabbits and Fig. 6 in pigmented rabbits
propriate conversion of the ocular dimension the time course of the concentration changes
is made. after the instillation of befunolol.* In the al-
Drug binding to tissues and pigments and bino eyes there is little drug accumulation in
its pharmacokinetic consequences. Follow-up the iris and ciliary body, and the steady state
of the aqueous pilocarpine concentration in of drug distribution is reached rapidly. As a
the albino rabbit eye for several hours after consequence, the concentrations in the aque-
instillation revealed that the kinetics of aque- ous, cornea, and anterior uvea decrease at a
ous pilocarpine could not be described by similar rate; equation 12 derived from the
equation 12.45> 70 Pilocarpine is known to be two-compartment model was applicable. On
bound to ocular tissues and serum protein, 71 the other hand, it is evident from Fig. 6 that
and the drug is probably accumulated, in the the anterior uvea of pigmented rabbits accu-
iris and the ciliary body tissues. The pharma- mulates the drug significantly. It is shown
cokinetics of pilocarpine was therefore inter- that the drugs accumulated in the pigmented
preted on the basis of four-compartment uveal tissues are released very slowly.76> 77> 80~82
model70 and the relevant transfer coefficients Thus the pigmented tissue acts as a reservoir
were calculated. Furthermore, pilocarpine of drugs, which accounts for long-lasting ef-
was shown to be metabolized in the ocular
tissues72 but too slowly to affect the in- *Takase M, Araie M, and Ishii Y: unpublished results.

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 513

fects in the pigmented eye. 73 ' 74' 76' 77» 8 °- 8 2

Indeed, the drug can be found in the pig- 1O
mented eye even 24 hr after instillation,
whereas in the albino eye the drug has disap- E
peared completely. The influence of ocular
pigmentation on the pharmacokinetics of the
human eye will be discussed in detail later in
this article.
I
Loss rate of drugs from tears and inraocu- LU
lar pharmacokinetics. We have considered O

the pharmacokinetics after instillation of con- I 10-

LU
ventional aqueous drug solution, which is lost z
rapidly from the conjunctival cul-de-sac so 0.
<
that corneal drug absorption is practically o
o
complete within a short time. However,
10"
when the drug is given in the form of oint-
ment, the loss rate in the tears is significantly
reduced, and this influences the pharmaco- 0 1 2 3 4 5 6 7 8
kinetics in the eye. Recently a drug delivery TIME (HOUR)
device was made available that controls dif-
Fig. 7. Computer-simulated time course of pilo-
fusional drug release from a reservoir by carpine concentration in the anterior chamber of
means of special membranes, thereby per- the human eye after instillation of 1% solution and
mitting almost constant drug release. 83 This its changes depending on the loss rate of the drug
device is made in the form of an ocular insert from the tears as indicated by percent per minute.
placed in the conjunctival cul-de-sac. Be- The following values were used for computation:
cause the drug is released at a constant rate tear volume 3 = 7 X 10" 3 cm3, Kep = 5 x 10"4 cm
(n per hour), the tear concentration will be hr" 1 , corneal area44 = 1.4 cm2, corneal thick-
maintained at a steady level; that is, [Cd] ness44 = 0.05 cm, k c x a = 0.38 hr" 1 , anterior
steady is equal to n/v, where v is the rate of chamber volume 68 - 69 = 0.175 cm3, k0 = 1.8 h r 1 .
tear flow per hour. With the loss rate from tears of 10% per minute,
the peak concentration is about 0.8 fxg ml" 1 , which
With the pharmacokinetic coefficients for agrees with the measurements in human eyes. 27 ' 28
pilocarpine, the intracameral pilocarpine con-
centrations have been computed for various
values of its rate of loss from the tears in the level, the aqueous concentration also reaches
human eye. The results are shown in Fig. 7. a steady state, with a time constant of k0, the
Reduction in the loss rate has two effects. coefficient of loss from the anterior chamber.
First, as indicated by equation 7, the amount The steady-state aqueous concentration and
of drug absorbed by the cornea is inversely tear concentration may then be
proportional to the loss rate, hence its reduc-
tion increases both the amount absorbed and (17)
the peak aqueous concentration (Fig. 8). Sec-
ond, the reduction in the loss rate in the tears where Kc is the overall permeability of the
retards the peak time (Fig. 9). When the loss corneal layers, Q is the corneal area, Va is the
rate becomes larger than 0.05 min" 1 , its distribution volume in the anterior chamber,
change has no significant influence on the k0 is the coefficient of loss from the anterior
peak time. Only when the loss rate is below chamber, and the suffix s means steady state.
0.02 min" 1 does the delay in the peak time Because the major barrier of the cornea is in
become evident. the epithelium, it may be assumed that Kc is
When the loss rate becomes zero, i.e., the almost equal to Kep. By inserting the values
tear concentration is maintained at a steady for pilocarpine in rabbit eyes, i.e., Kep =

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis, Sci.
514 Mishima October 1981

L 10
E
6)
z
g
< 1.0
Z
HI
o
z
o
o
LU
0.
0 10 20 30 40 50 60 70 80 90 100
LOSS RATE IN THE TEAR
( % PER MINUTE )
Fig. 8. Relationship between the peak aqueous concentration and the loss rate in the tears.
Results of computer simulation for pilocarpine as in Fig. 7.

cc 5

LU
&• 1

O 10 20 30 40 50 60 70 80 90 100
LOSS RATE IN THE TEAR
( % PER MINUTE )
Fig. 9. Relationship between the time of peak aqueous concentration and the loss rate in the
tears. Results of computer simulation for pilocarpine as in Fig. 7.

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 515

10 X 10"4 cm hr~\ k0 = 3 to 7 hr" 1 (Tables I

and II), and taking Q61 = 2 cm2 and Va64 =
0.25 cm 3 the relationship between the steady-
state pilocarpine concentrations in the 10
tears and the aqueous may be calculated:
[Ca]s = 2.3 - 1.1 x l O - ^ C J s . UJ
A A X •
Sendelbeck et al.84 placed the ocular pi-
locarpine insert, delivering 20 /xg hr" 1 , in the A it»
conjunctival cul-de-sac of the albino rabbit (/)
D • A DO
and found that the steady-state pilocarpine </)
UJ
concentration in the iris was about 0.6 fxg per
a «o
gm of tissue. If it is assumed that this is equal a o
to the aqueous concentration, the steady- 1.0
state tear concentration, [C d ] s , can be calcu-
lated to be 0.26 to 0.55 x 10" 3 gm ml" 1 . This 10 20 30 40
would be achieved if the tear flow rate is 0.6
to 1.3 fx\ min"1. The normal tear flow rate in T I M E ( m i n.)
the rabbit was reported 15 to be 0.55 /xl min"1.
Fig. 10. Changes in the corneal sensitivity thresh-
The agreement is very good if one considers
old after instillation of various concentrations (%)
that the insert may stimulate tear secretion, of lidocaine (•, 0.2; A, 0.4; A, 0.8; a, 1.2; •, 1.6; o,
and this indicates that the permeability of the 2.0). From Matsumoto T, Hayashi K, Tsuchisaka
epithelial barrier of the rabbit cornea to pilo- H, and Araie M: Jpn J Ophthalmol 25:335, 1981.
carpine given in Table I is reasonable.
(A of each solution, the authors determined
Pharmacokinetics of surface anesthetics the sensitivity threshold of the cornea in the
There is probably not a single day that an temporal intermediate zone with a Cochet-
opthahnologist can practice without use of Bonnet esthesiometer 88 until the corneal
surface anesthetics. Applanation tonometry, sensitivity returned to the predrug level. The
tonography, minor precedures, and even sensitivity threshold can be expressed in
major surgery are performed with the use of grams per square millimeter by using the
surface anesthetics. These agents can be toxic conversion table provided by the authors, 88
to the corneal epithelium, where changes of and it is plotted against the time after instil-
microvilli, cell desquamation, pit formation, lation in Fig. 10 for lidocaine and in Fig. 11
etc. are found to occur.85' 86 To minimize for benoxinate. The peak effect of lidocaine
these toxic effects and to promote appro- was usually reached 2 to 3 min after instilla-
priate use of the surface anesthetics in daily tion, and the peak effect of benoxinate oc-
practice, knowledge of the pharmacokinetics curred in about 1 to 2 min. Since the peak
of these drugs in the human eye should be time was very short, the duration of the effect
useful, and yet there appear to be no reports was calculated as the interval between the
on this aspect available in the literature. Ac- time of instillation and the time at which the
cordingly, Matsumoto et al.87 conducted sensitivity threshold returned to normal
studies of the surface anesthetics most often level. The logarithm of drug concentrations
utilized in clinical practice. and the duration of the effect are correlated
They prepared lidocaine solutions at six in Fig. 12 for lidocaine and in Fig. 13 for
concentrations in the range between 0.2% benoxinate. In both figures a linear correla-
and 2.0% as well as benoxinate solutions at tion is evident. The experiments repeated in
five concentrations from 0.08% to 0.20%, six young male volunteers with each drug
using commercial vehicles. The pH was 6.6 gave similar results in all cases.
in the former preparation and 5.5 in the lat- After instillation, the uncharged base of
ter. At 1 min intervals after instillation of 20 the anesthetics is absorbed by the lipid-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
516 Mishima October 1981

1.0

10
0.1
E
E
en
10
LU T I M E m i n.
Fig. 13. Relationship between the logarithm of
D benoxinate concentration and the duration of anes-
thetic effect. From Matsumoto T, Hayashi K,
if) Tsuchisaka H, and Araie M: Jpn J Ophthalmol
(/) 25:335, 1981.
LU
a: a •
Q. containing corneal epithelium. 89 It is then
D lost from this tissue by release into the cor-
1.0 neal stroma or back into the tears. If the epi-
thelial concentration of an anesthetic is de-
noted as C e and the rate constant of its elimi-
5 10 nation from the epithelium as ke, then the
I M E ( m i n. changes in epithelial concentration may be
given by substituting these symbols for C c
Fig. 11. Changes in the corneal sensitivity
and kc in equation 3. Since the loss rate in the
threshold after instillation of various concentra-
tears, a, is very large, the second term of the
tions (%) of benoxinate (A, 0.08; A, 0.10; a, 0.12; •,
0.15; •, 0.20). From Matsumoto T, Hayashi K,
equation becomes negligible. This makes the
Tsuchisaka H, and Araie M: Jpn J Ophthalmol single-compartment model applicable to the
25:335, 1981. present case.
Levy90' 91 demonstrated that the following
equation applies to the pharmacokinetics of a
single compartment:
1.0 Log C - Log C mln = T k e /2.3 (18)
that is, a linear relationship is seen between
the logarithm of the concentration and the
o duration of the effect (T). Thus the results in
z Fig. 12 and 13 may be analyzed on the basis
of this equation; the slope of the regression
I 0.1 line gives the rate constant of drug loss from
a the corneal epithelium, and extrapolation of
10 20 30 the line gives the C min , the least effective
T I M E I m i n.) concentration.
Fig. 12. Relationship between the logarithm of By means of the least-squares method, the
lidocaine concentration and the duration of anes- regression lines for the relationship were
thetic effect. From Matsumoto T, Hayashi K, computed for each individual. The elimina-
Tsuchisaka H, and Araie M: Jpn J Ophthalmol tion rate constant of the drug averaged
25:335, 1981. 0.081 ± 0.011 (S.E.) min" 1 for lidocaine and

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 517

_-©-
100
-o-

0
C
o

\0
(A

Ol_ —®=r

io8 16' To6 ro 5 ro4

mol/L
Fig. 14. Dose-response relationship in the isolated human sphincter muscle strip with car-
bachol (®) and pilocarpine (o). From Ohara K: Jpn J Ophthalmol 21:516, 1977.

0.10 ± 0.022 min 1 for benoxinate. The least and few data points did not allow exact calcu-
effective concentration averaged 0.10% ± lation of the constant for benoxinate. We
0.04% for lidocaine and 0.06% ± 0.007% for found three reports in the literature that gave
benoxinate. With these values for the elimi- the relationship between the concentrations
nation rate constant and the peak time of the of surface anesthetics and their effects on
effect, the loss rate of the surface anesthetics animal corneas. The elimination rate con-
in the tears may be calculated on the basis of stants from the corneal epithelium were
equation 4, with appropriate symbol changes. computed from their data: 0.08 min" 1 for
It was at least 1 min" 1 for lidocaine and 2.5 procaine, 0.10 min" 1 for butyn pH 5.5, and
min" 1 for benoxinate. Conrad et al.92 dem- 0.054 min" 1 for butyn pH 7.0 in rabbits94;
onstrated that drug dilution by tearing was and 0.048 min" 1 and 0.052 min" 1 for procaine
the least at physiologic pH and that it in- in guinea pigs. 95 ' 96 These values are in very
creased as the pH deviated toward both acid good agreement with the results in the
and alkaline sides. Thus the loss rate in the human cornea.
tears of the above order is not surprising with With a knowledge of the elimination rate
instillation of surface anesthetics that induce constant from the corneal epithelium, the
initial stinging reaction. Furthermore, it is least effective concentration, and the concen-
possible that the greater loss rate of benoxi- tration of the drug to be given, the duration
nate in tears than that of lidocaine is because of the effect (T) may be calculated by
of the more acidic pH of benoxinate.
2.3Log(C/Cmln)
Poise et al. 93 reported the average duration T = (19)
ke
of anesthetic effects in man for proparacaine
and benoxinate in three concentrations. Simi- Substituting values for benoxinate, one can
lar analysis of their data gave the elimination state that 0.4% benoxinate is effective for 19
rate constant from the corneal epithelium of min. Similarly, 4% lidocaine is effective for
0.08 min" 1 for proparacaine, but data scatter 45 min. The lasting effect of lidocaine is a

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
518 Mishima October 1981

muscle are largely alpha-adrenergic.97 Thus

2.0 - the pupil response is a simple system, from a
pharmacologic point of view, which is con-
venient for drug testing.
A mathematical expression of the time
cc
course of the pupil responses after single top-
I ical application of drugs has been attempted
x
(0 in human subjects, and a double-exponen-
J tial equation has been fitted to the time
S
OC course. 98 " 100 The usefulness of the pupil re-
sponse for pharmacokinetic study of a drug
was shown by Levy.90' 91 He indicated that
the decline of the mydriatic response in mice
after intravenous administration of an an-
-2.0 ticholinergic agent allows calculation of the
drug elimination rate constant, provided that
the appropriate conversion of the response
can be made on the basis of the dose-
response relationship. Subsequently, Smolen
10° 10' 10" 10
and associates101"104 carried out a series of
mol/ L
studies on the pupil response in the rabbit.
Fig. 15. Log-log plot of the response parameter On the basis of the dose-response relation-
and the drug concentration for the data shown in ship obtained by intravenous administration
Fig. 14. For explanation of the response parame- of the drug, they developed a function to
ter see text. From Ohara K: Jpn J Ophthalmol convert the pupil response into the parame-
21:516, 1977.
ters that indicate the drug concentration in
result of the high concentration of its avail- the biophase of the sphincter muscles. Fur-
able preparation and not of a difference in ther, they developed a computerized program
its elimination rate constant. Poise et al. 93 to estimate the bioavailability of the drug from
showed with applanation tonometry that cor- the pharmacologic data.104
neal sensation was absent when the corneal The pharmacokinetic study of the pupil re-
sensitivity threshold was about 6.6 gm mm" 2 sponse in the human subject was carried out
or higher, and below this level the patients by Yoshida and Mishima31 using topical
felt the touch of the tonometer. This sen- pilocarpine and tropicamide. An analysis of
sitivity threshold is reached at the peak time the dose-response relationship according to
after instillation of a 0.18% solution of benox- the principle of Wagner105 permitted calcula-
inate. The presently used 0.4% preparation tion of the response parameter (RP) that is
can maintain the absence of corneal sensation proportional to the drug concentration in the
for 8 min during applanation tonometry. biophase of the sphincter muscles. Study of
These calculations are in agreement with our the time course of the RP changes leads to
clinical experience. computation of the apparent elimination rate
constant and other relevant pharmacokinetic
Pharmacokinetics of pupil responses
coefficients for these drugs in the human eye.
The pupil response is elicited through the Subsequently, a similar analysis was con-
contraction and relaxation of two smooth ducted for various drugs, with use of our own
muscles, i.e., the sphincter and the dilator results and data reported in the literature.
muscles. The sphincter muscles are inner- This section describes conclusions of this type
vated largely by cholinergic nerves. The di- of analysis.
lator muscles are innervated mainly by ad- Dose-response relationship in isolated iris.
renergic nerves, and the receptors in the The iris is a porous structure that allows rapid

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 519

Z
P 80
U

(/)
Z 60
O
O
on
< 40

CL
D 12 16
0.
TIME (hours)
Fig. 16. Time course of miotic response after pilocarpine instillation. Dc, Control pupil diame-
ter; D, the experimental pupil diameter at a given time. •, 2.5 x 10~3; X, 5.0 X 10~3; •,
1.0 x 10~2; A, 2.0 X 10~2 g ml"'. From Yoshida S and Mishima S: Jpn J Ophthalmol 19:121,
1975.

access to the vicinity of the sphincter and di- From the maximum response (Rmax) and
lator muscles106' 107 of solutes even of mac- the response for a given molar concentration
romolecular size present in the anterior of the drug (R), the parameter R/(Rmax ~ R)
chamber. Thus a drug in the anterior cham- was calculated and is plotted in Fig. 15
ber can reach the biophase of these muscles in against the molar concentrations of the drug,
a very short time, and their response will cor- both on a logarithmic scale. A straight line
respond to that in the incubation bath of an in can be fitted to the relationship, and its slope
vitro system. Accordingly, the interaction of is close to unity. Thus one can describe the
drugs with isolated iris muscles will be con- relationship by
sidered first.
Fig. 14 shows the dose-response relation- R = qC (20)
ship in the sphincter muscle strip of the - R
human eye for carbachol and pilocarpine, de- where C is the drug concentration and q is
termined from the conventional cumulative the proportionality constant.
increase of the drug concentration in the in- Replotting on this principle was carried out
cubation medium. 108 The sphincter muscles for published response relationships: car-
have typical muscarinic receptors, 109 and bachol and the sphincter muscle strip of the
pilocarpine and carbachol are directly acting cat,111 relaxation effects of isoproterenol on
agonists.110 Since carbachol is considered a the bovine112 and rabbit113 sphincter strip, at-
full agonist, the intrinsic activity of pilocar- ropine and a whole-mount guinea pig iris,114
pine may be calculated by comparison of the adrenaline and acetylcholine and a whole
maximum responses. In the human iris this mount of cat iris,115 and Z-epinephrine and a
is 0.9, indicating that pilocarpine can be con- whole-mount guinea pig iris. 116 In all cases
sidered also to be a full agonist.108 the logarithmic plot of R/(Rmax ~~ R) against

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
520 Mishima October 1981

"x 100
o
a

Q
60
Z
g
40

20

2 4
TIME (hours)
Fig. 17. Time course of mydriatic response after tropicamide instillation. D c and D are as in
Fig. 16. •, 1.0 x 1(T4; X, 4.0 X 10"4; •, 1.0 x 10~3; A, 4.0 X 10"3 g ml"1. From Yoshida S
and Mishima S: Jpn J Ophthalmol 19:121, 1975.

logarithm of the drug concentration gave a a room with constant lighting. The pupil re-
linear relationship with the slope close to 1. sponse was then expressed in percent re-
Thus one can conclude that equation 20 de- sponse; if the diameter in the control eye is
scribes the dose-response relationship of the D c and the diameter in the experimental eye
sphincter and the dilator muscles. This equa- is D, then the percent response is given by
tion is similar in its form to the theoretical [(D c - D)/D c ] x 100 for pilocarpine and
equation derived by Ariens117 for drug in- [(D - D c )/D c ] X 100 for tropicamide. The
teraction with a one-receptor system. What- time course of the percent changes is shown
ever the theory may be concerning the man- in Fig. 16 for pilocarpine and in Fig. 17 for
ifestation of the drug effects, the applicability tropicamide. The peak responses were then
of the above equation to the in vitro iris prep- plotted against logarithm of the drug concen-
aration indicates that this is also the case for tration in Fig. 18 for pilocarpine, and a simi-
the relationship between the pupil response lar relationship can also be obtained for
and the drug concentration in the biophase of tropicamide. The relationships appear to be
the iris muscles of the living eye. linear.
Dose-response relationship of pupil re- The iris is an extremely mobile structure,
sponses. Yoshida and Mishima31 instilled 50 but unrestricted movement occurs only in
fx\ of various concentrations of pilocarpine so- the range of moderate pupil size; beyond the
lution in one eye of young subjects after at limit of 3 to 6 mm of the diameter, mechan-
least 15 min of dark adaptation, and the pu- ical restriction hinders iris movement.118* u 9
pil diameter of both eyes was determined Furthermore, it is impossible for the pupil
through measurements on infrared photo- diameter to become zero or to dilate beyond a
graphs taken in a dark room. Similarly, certain limit. Thus, with a high concentration
tropicamide experiments were carried out in of a drug, the response can no longer increase

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 521

#60

Q
o
Q

40
Z
o
Q.
C/)
LLJ
20

10" 10' 10" 10"

1
CONCENTRATION (gml" )
Fig. 18. Dose-response relationship obtained from the peak responses shown in Fig. 16. From
Yoshida S and Mishima S: Jpn J Ophthalmol 19:121, 1975.

with increase of the drug concentration; that D m a x , we can calculate the response parame-
is, a minimum or maximum pupil diameter ter, R/(Rmax ~ R)> for the pupil response.
must exist. Accordingly, the apparent linear For this purpose the pupil response, R, is
relationships shown in Fig. 18 covers only the redefined as follows. Let us denote the pupil
linear part of the sigmoid dose-response rela- diameter before the drug as D o and the di-
tionship. ameter at time of administration as D. We
The maximum or minimum possible size of may then write R = (D o — D) and Rmax =
the pupil may be determined from the results (D o — Dmln) for miotic response, and R =
of Loewenfeld and Newsome. 119 A light (D - Do) and Rmax = (D max - Do) for myd-
reaction was induced after pretreatment with riatic response. By expressing the pupil di-
0.5% physostigmine, leading to the smallest ameter in millimeters, the response parame-
attainable pupil diameter of 0.92 ± 0.13 mm ter (RP) may be written
(S.D.) in six young subjects. Thus 1 mm is a R
good approximation for the minimum pupil RP =
Rmax ~~
diameter, and this should also apply to
pilocarpine, since it is a full agonist for the _ Dp D for miotic response (21)
human iris sphincter muscles.108 The authors D - 1
also gave the largest pupil diameter in the
dark after topical administration of cyclopen- _ D Dp for mydriatic response (22)
tolate or cocaine, which was 7.94 ± 0.96 mm 8.5- D
with the former and 8.50 ± 0.71 mm with The response parameter, RP, was then calcu-
the latter drug. Yoshida and Mishima31 also lated from the pupil diameter at peak time
estimated the maximum pupil size to be 8.5 and was plotted against the drug concentra-
mm from their data, so that both estimates tion, both on a logarithmic scale. Fig. 19 is
are in agreement. the example for pilocarpine, and the similar
Using these values for minimum pupil di- relationship can also be obtained for tropic-
ameter, D mln , and the maximum diameter, amide. A linear correlation is seen and the

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
522 Mishima October 1981

10 Because intact sympathetic nerves in the iris

actively accumulate exogenously adminis-
0) tered norepinephrine,122 it is possible that
CD linear correlation does not hold between the
CO 10 1
administered amount and the final concen-
CO tration in the biophase of the dilator muscles.
a Thus the present analysis using RP is not
(0
applicable for the in vivo response to norepi-
nephrine and epinephrine.
I«
V)
Kinetics of pupil responses. The RP may be
calculated for the pupil diameters at various
a> time intervals after instillation, as shown in
DC
Fig. 16, and the results are plotted in Fig. 20.
The RP increases in proportion to the in-
Concentration (g-ml) crease of the drug concentrations, and the
shape of the curves for different drug concen-
Fig. 19. Log-log plot of the response parameter trations are congruent, indicating that the RP
and the pilocarpine concentrations from the data is indeed proportional to the biophase drug
shown in Fig. 16. For explanation of the response concentrations. Fig. 21 and 22 show the RP-
parameter see text. From Yoshida S and Mishima time curves for pilocarpine and tropicamide,
S: Jpn J Ophthalmol 19:121, 1975. respectively, and the curves are similar. The
following equation describes the curves:

slope of the regression line is very close to RP = m AB [e -A(t - t ) _ -B(t -

0 (24)

unity. Again one can assume the relationship (B-A)

to be expressed where m, A, and B are constants, and t and t0
are the time and the lag time required for the
RP = — Q Qns (23) response to begin after instillation. This
equation is basically the same as equation 12
As discussed previously, tear-aqueous drug for intraocular drug penetration; A may be
transfer follows first-order kinetics, and com- called the apparent elimination rate constant,
parison of equations 20 and 23 leads to the and B may be called the apparent absorption
conclusion that the RP calculated from the rate constant. Both A and B are related to the
pupil diameter is proportional to the drug rate of drug release from the cornea into the
concentration in the biophase of the iris anterior chamber and to the rate of loss from
muscles. the anterior chamber.
The mydriatic response of the human eye Similar analysis was made on the results
to norepinephrine120 was analyzed by the published in the literature. Some authors
log-log plot of the RP against dose. Similarly, gave the actual diameter of the pupil in their
the mydriatic response of the rabbit after figures, but others gave only the diameter
the intravitreal injection of norepinephrine121 change either as percentage of the predrug
was also analyzed. In both cases a linear cor- diameter or as the actual difference of the
relation was obtained, but the slope was 1.6 diameter. In the latter case the predrug di-
to 2. In the isolated iris the slope was 1 for ameter had to be assumed. One millimeter
epinephrine, and such deviation of the slope difference in the predrug diameter did not
from unity must be characteristic of the in give significant deviation in the pharmacoki-
vivo response for this drug. Significant de- netic coefficients, and therefore it was as-
viation of the slope from unity is attributed to sumed to be 7.6 mm in the dark (miotic ex-
nonlinear deviation of the biophase concen- periments) and 4 mm in the light (mydriatic
tration from the ambient concentration.117 experiments). The pharmacokinetic coeffi-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 523

101

0
0

2
cc
c.
X 10
0

6.1 o
a
0
c
o CC
a
</>
at
0£
10
2 4 6 8 10
0.01
Time (hours)
5 10
Hours Fig. 21. Time course of the response parameter
after instillation of 0.5% pilocarpine solution. Two
Fig. 20. Time course of the response parameter lines for calculation of the absorption and elimina-
changes after instillation of pilocarpine solution tion rate constants are shown. From Yoshida S and
with various concentrations, calculated from the Mishima S: Jpn J Ophthalmol 19:121, 1975.
data shown in Fig. 16. •. 2.5 x 10"3; X,
5.0 x 10"3; •, 1.0 x 10" 2.0 x 10"2 g ml. ues are in good agreement with those ob-
From Mishima S and Nagataki S: Contact In- tained by kinetic analysis (Table IV). The
traocular Lens Med J 4(3):22, 1978. pilocarpine miosis time study by Borgmann
and Wurster 125 was also analyzed by this
cients were thus calculated for various drugs method and the results are given in Table IV.
and are listed in Table IV. The least effective concentration may also be
Duration of effects and drug concentration. calculated by the dose-response method,
The RP of the pupil decreases in a single expo- i.e., by extrapolation of the relationship, as
nential manner after the peak time has passed. shown in Fig. 18. The least effective concen-
During this period of decay, the second term tration of pilocarpine in five young subjects
of equation 24 becomes negligible and the was 1.0 ± 0.6 X 10~3 gm ml" 1 by the dose-
theory of single-compartment would apply to response method and 1.3 ± 0.6 X 10~3 gm
this part of the response. Accordingly, the ml" 1 by the duration method. The least ef-
duration of the effect was defined as the time fective concentration of tropicamide in five
interval between the time of the peak re- young subjects was 2.6 ± 1.2 x 10~5 gm
sponse and the time of effect termination. The ml" 1 by the former and 1.7 ± 0.8 X 10"5 gm
logarithm of pilocarpine concentration and ml" 1 by the latter method. The results of
the duration of effect were correlated in Fig. both methods are in agreement.
23, and this was analyzed according to equa- Pharmacokinetic coefficients and ocular
tion 18, which allows calculations of the ap- pigmentation. We found two sets of data in
parent elimination rate constant and the least the literature that allow pilocarpine kinetics
effective concentration. This method of calcu- in subjects with light-colored and dark irides
lation may be called the duration method. to be compared. The results reported from
The apparent elimination rate constant was England124 and Germany 125 agree very well,
0.34 ± 0.10 hr" 1 for pilocarpine (Table IV) and the apparent elimination rate constant
and 0.82 ± 0.20 for tropicamide. These val- was 0.46 to 0.48 hr" 1 in subjects with a lightly

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
524 Mishima October 1981

io conclude that the apparent elimination rate

constant for these drugs decreases with in-
0) creasing ocular pigmentation.
0)
Chen and Poth128 pointed out that the
cO mydriatic action of cocaine, euphthalmine,
CO
a and ephedrine is less effective in black sub-
<D jects than in white subjects, and the effect in
(0 Chinese is intermediate. This is comparable
c 1OV
o to what we have seen in the apparent elimi-
a nation rate constant of pilocarpine. Similarly,
(0
0) numerous reports agree that the effects of var-
10
ious miotics and mydriatics are less intense
2 4 6 8 in more heavily pigmented subjects.129"131
Time (hours) Less marked effects with increasing pigmen-
tation are also reported for hypotensive effects
Fig. 22. Time course of the response parameter
of pilocarpine and epinephrine. 132 ' 133 Since
after instillation of 0.4% tropicamide solution. Two
these drugs are bound to the pigments of the
lines for calculation of the absorption and elimina-
tion rate constants are shown. From Yoshida S and anterior uvea,73"79 the decrease in the effects
Mishima S: Jpn J Ophthalmol 19:121, 1975. may be attributed to reduction in the biophase
drug concentration. In rabbits and guinea pigs
the mydriatic effects of various drugs were
pigmented iris and 0.36 to 0.37 hr 1 in sub- shown to decay at a slower rate in pigmented
jects with a dark iris (Table IV). The former strains than that in albino strains, and this was
value is close to the rate constant of 0.44 hr" 1 , attributed to the depot effect of the pigments,
computed from the results given in two re- which release bound drugs slowly.80"82 A dif-
ports from the United States. Although these ference in the pharmacokinetics of drugs be-
did not state the iris color of the subjects, it is tween the pigmented and albino rabbits is
quite possible that the majority had lightly shown in Figs. 5 and 6, and this difference
pigmented irides. The rate constant in Euro- supports the above explanation.
pean subjects with dark irides is similar to the Relative bioavailability and form of drug
value computed for Japanese subjects (Table delivery. Rapid loss of instilled drug solution
IV). Recently, Dr. J. W. Shell* at the School from the conjunctival cul-de-sac makes only
of Pharmacy of the University of California in a small fraction of a given drug available to
San Francisco gave us preliminary results of the eye. However, the use of various vehicles
his experiments, in which he found that the has been reported to increase the availabil-
half-life of pilocarpine effect decay averaged ity of the drug, thereby increasing its ef-
4.7 hr in four heavily pigmented subjects and fect.18' 20- "• 134"140 Efficiency of various
2 hr in three lightly pigmented subjects. methods of drug delivery in increasing drug
These values correspond to the apparent penetration may be studied by the use of the
elimination rate constants of 0.15 hr" 1 and concept of bioavailability. A comparison of
0.33 hr" 1 . the bioavailability of a drug between different
The apparent elimination rate constant of forms of delivery may be made by means of
homatropine was calculated from the average the area under the concentration-time curve.
time course reported by Gambill et al.99; it There is only one report in the literature 26
was 0.19 hr" 1 and 0.31 hr" 1 for subjects with that gives an estimate of the area under
dark and light eyes, respectively. We may concentration-time curve of a drug, predniso-
lone, in the anterior chamber of the human
eye after topical application, and on this basis
*Personal communication. the pharmacokinetic coefficients of predniso-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 525

3
O

D
Q

»-
U

10"

CONCENTRATION (g-mf1)
Fig. 23. Relationship between the duration of miotic effect and logarithm of pilocarpine
concentrations. For definition of the effect duration see text. From Yoshida S and Mishima S:
Jpn J Ophthalmol 19:121, 1975.

Table IV. Kinetic coefficients of the pupil responses

No. of B Lag time Peak time Source of data

Drug cases (hr~l) (t0, hr) (hr) (ref.)

Pilocarpine 5A 0.34 ± 0.10° — 31

5B 0.27 ± 0.05 1.84 ± 0.40° 31
D
5C 0.35 ± 0.12 1.20 ± 0.29 0.13 ± 0.075 1.6 6
6C 0.44 ± 0.14 2.44 ± 0.44 0.15 ± 0.08 1.0 3 cases (Fig. 1 of
123); 3 cases
(Figs. 1, 3, 5
of 98)
Blue iris 8 0.48 1 n Fig. 2 of 124
= Fig. 2 of 124
Brown iris 0.36 J °
Light irisA 0.46 125
Dark iris 0.37 125
Tropicamide 5B 0.84 ± 0.14D 2.22 ± 0.26D 31
3C 0.68 1.5 0.16 1.0 Unpublished data
1B 0.60 4.0 0.10 0.56 126
lc 0.64 ' 4.5 0.12 0.63 Fig. 5'
Hornatropine Allc 0.23 2.55 0.24 1.2 Fig. 4
Dark iris c 0.19 2.46 0.25 1.37 Fig. 6 • r>f QQ
Light iris c 0.31 Do = 2.07 0.25 1.33 Fig. 6 ot yy
Phenylephrine lc 0.63 2.05 0.34 1.17 Fig. 2
Hydroxyam- lc 0.47 2.53 0.34 1.16 Fig. 3
phetamine
Scopolamine 1B 0.025 Fig. 6 of 127
Atropine 1 0.017 Fig. 5 of 127

A = apparent elimination rate constant; B = apparent absorption rate constant; D o = predrug pupil diameter (assumed as indicated).
A
Calculated from the relationship between the duration and logarithm of concentration.
"Calculated by peeling method.
c
Calculated by a computer.
"Numbers are mean ± S.D.

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sri.
526 Mishima October 1981

20

10
=5
O
CD

0
10 20 30 40 50 60
Instilled volume (ul)
Fig. 24. Relationship between the relative bioavailability and the drop size of 0.5% pilocarpine
solution. Solid lines, Probable range of the mean. For definition of the relative bioavailability
see text. From Sugaya M and Nagataki S: Jpn J Ophthalmol 22:127, 1978.

lone in the human eye given in Table III were ability were studied by Sugaya and Nagataki.6
computed. 32 However, it is extremely difficult They prepared 0.5% pilocarpine solution at
to construct the complete time course of con- physiologic pH and instilled 5, 10, 20, and 50
centration changes by human experiments of fj\ to young normal subjects. The coefficient
this type. On the other hand, pupil changes of relative bioavailability was calculated from
can be followed easily for a long time without the RP-time curve by computer and is plot-
any strain to the subjects, and the RP that is ted in Fig. 24 against the volume of instilla-
proportional to the biophase concentration tion. The probable range of the mean was
can be obtained by simple calculation. Thus also computed and is given in the Fig. 24.
the study of the bioavailability in the human The coefficient of relative bioavailability in-
eye may be carried out with the area under the creases slightly with the drop-size increase
RP-time curve. Integration of equation 24 from 5 to 20 fi\, but no further augmentation
from time to infinity gives can be seen by increasing the drop size.
t = oo These results may be accounted for by the
RP dt = m (25) behavior of drugs in the conjunctival cul-de-
t= t sac, i.e., rapid drainage of instilled volume
If a drug with a given concentration is applied and poor tear film saturation with the drug.
in different forms, the bioavailabilities of Clinical implication of this phenomenon has
these preparations can be compared by means been discussed in the preceding section, and
of the values of m, hence it may be called the the volume of 20 /xl was thought to be practi-
coefficient of the relative bioavailability. cal for clinical use. 33
Drop size and bioavailability. The effects Viscous vehicles and bioavailability. The
of the volume of instillation on the bioavail- effects of viscous vehicles on the bioavailabil-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 527

ity were studied on the same principle with 10

the RP-time curve. 6 Pilocarpine solution of (Z
0.5% was prepared at physiologic pH and os- I
molarity, with addition of various concen- x
trations of hydroxypropylmethylcellulose o
E o
(HPMC). After instillation of 20 /*1, the RP
A*A A
was calculated and is plotted in Fig. 25,
where the results with 0.5% pilocarpine A
aqueous solution are also shown for compari-
O AO
son. The vehicle with 0.5% HPMC signifi-
cantly increases the RP, hence the intraocular +* 1
pilocarpine concentration. The coefficient of
relative bioavailability with the vehicle was
° O
greater than that with aqueous solution, and
Q. 0.5
the ratio averaged 3.1 ± 1.9 (S.D.) in five
normal subjects. 0)

The relative bioavailability and the viscos- c

o
ity of the solution are plotted in Fig. 26; the a
</>
probable range of the mean is also shown. a»
The bioavailability increases with increasing
viscosity of the vehicle, but no further rise
0.1 "
can be seen above the viscosity of 10 to 20 1 1

centistokes. Thus the viscosity of 10 to 20 10

centistokes is thought to be adequate for Hours
ophthalmic preparation. 33
From equation 12 it is evident that the bio- Fig. 25. Time course of the response parameter
availability is proportional to the amount of after instillation of 0.5% pilocarpine solution (•),
the drug initially absorbed by the cornea and 0.5% pilocarpine in 0.5% HPMC solution (o), and
is proportional to the degree of tear film sat- after application of 100 mg of 0.5% pilocarpine
ointment (A). From Sugaya M and Nagataki S: Jpn
uration. The degree of tear film saturation
J Ophthalmol 22:127, 1978.
was calculated for drop size and viscosity
studies on the basis of tear flow data with
fluorescein, and the relative bioavailability is 20) with aqueous pilocarpine solutions and
plotted in Fig. 27 against the degree of tear 6.8 ± 1.8 min (n = 20) with HPMC solu-
film saturation of pilocarpine. This shows that tion, but the apparent lag time with ointment
the above theory is indeed applicable to was 12.7 ± 2 . 5 min (n = 12). A significant
human eyes. delay occurs with ointment in the intraocular
Ointment and bioavailability. A typical penetration of the drug. The delay of the
RP-time curve after application of 100 mg of peak time is evident in Fig. 25; the peak time
0.5% pilocarpine ointment is shown in Fig. with aqueous solution is less than 1.5 hr but
25. As compared to aqueous solution, the with ointment it is about 2.5 hr. It has al-
bioavailability increased more than twice. In- ready been shown that the delay of the peak
spection of the RP-time curve with the oint- time is caused by a reduction in the loss rate
ment reveals two interesting aspects: (1) the from the tears to less than 0.01 min" 1 , and
RP increase occurs with a delay and (2) the this is the case with ointment. The delay in
time of peak RP takes place with a significant the intraocular drug penetration may be at-
delay, as compared with that for aqueous so- tributed to the mode of drug release from the
lution and HPMC solution. ointment, as discussed above.
The lag time, t0, given by equation 24 was Hydrophilic soft contact lens and bioavail-
computed 6 to be 7.7 ± 3.4 min (S.D.; n = ability. The use of soft contact lenses pre-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
528 Mishima October 1981

20

5
>
o
CD

10 20
Viscosity of solution (centistokes)
Fig. 26. Relationship between the relative bioavailability and the viscosity of the HPMC
solution containing 0.5% pilocarpine. Solid lines, Probable range of the mean. For definition of
the relative bioavailability see text. From Sugaya M and Nagataki S: Jpn J Ophthalmol 22:127,
1978.

soaked in a drug solution has been reported tact lens use is mainly caused by a burst of
to enhance the drug effects.141"145 Fig. 28 drug from the absorbed sites of the lens into
shows the RP-time curve after the 1 hr wear the tear film. The tear film concentration can
of a soft contact lens presoaked in 0.02% be very high, depending on the amount of
pilocarpine solution for 25 min. The RP-time drug absorbed by the soft contact lens, and
curves for 0.5% pilocarpine in aqueous so- this can be dangerous.
lution and in 0.5% HPMC solution are also Drugs may often be instilled with the soft
shown for comparison. The bioavailability is contact lens present in the eye. A set of the
greatly augmented by the soft contact lens, RP-time curves for pilocarpine instilled with
but the peak time shows no significant delay and without the soft contact lens in the eye
compared with that with aqueous solution. are shown in Fig. 29. In the case of Fig. 29,
The pupil response with pilocarpine in two of left panel, the RP is higher in the presence of
the published papers 141 ' 143 also indicates the lens than without it, but in Fig. 29, right
practically no delay in the time of peak re- panel, the result is reversed. In the latter
sponse. This shows that the loss rate of drugs case, it seems that the soft contact lens is
from the tears with this modality is not much acting as a barrier to the access of drugs to the
different from that with viscous vehicles. The cornea. However, when instillation is carried
soft contact lens absorbs various drugs at a out carefully to bring the drug under the soft
high concentration and releases them rapidly contact lens, one can increase the bioavail-
into the surrounding medium. The release ability of the drug. The fact that careless in-
rate was estimated to be about 0.03 to 0.04 stillation can greatly reduce the bioavailabil-
min" 1 from published data,142' 145 and this is ity in the presence of the soft contact lens is
in support of the above conclusion. Thus the of clinical importance, and this is worthwhile
increase in the bioavailability with soft con- warning.

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 529

lot

CD

10 0
E oo
co i
>» CO
a
> 0
o CO
ffi o
o
a
CO
0
50 100 cc
Initial concentration °/o
001
2 4 6
Fig. 27. Relative bioavailability and the degree of
tear film saturation. The initial tear film concen- Time (hours)
trations were estimated from drop size studies (o) Fig. 28. Time course of the response parameter
and viscosity studies (•) of the tear flow with after instillation of 0.5% pilocarpine solution (•),
fluorescein. The relative bioavailability is obtained instillation of 0.5% pilocarpine in 0.5% HPMC so-
through similar studies with 0.5% pilocarpine. lution (o), and after 1 hr wear of the soft contact
From Sugaya M and Nagataki S: Jpn J Ophthalmol lens (A) presoaked in 0.02% pilocarpine solution
22:127, 1978. for 25 min.

Drug delivery device. A drug delivery de- with the Pilo-20 points after the second day
vice in the form of an ocular insert was made leads to the conclusion that this method of
available,83' 84 and the inserts of pilocar- administration maintains a level equivalent to
pine 146 and corticosteroid147 were reported to the peak concentration after instillation of
be useful clinically. Fig. 30, top, shows the 0.4% to 0.5% pilocarpine solution.149
change in the release rate of pilocarpine in-
sert over 1 week. During the steady state, Pharmacokinetics of cycloplegic responses
this type of insert, Pilo-20, releases pilocar- Several reports are available that show the
pine at a rate of 20 /xg/hr, hence the tear film time course of the cycloplegic response to
concentration must also be at a steady state. various drugs. These data may be analyzed
Horie et al.148 tried this insert in 16 subjects, on the basis of the principles expounded here
and the average change in the pupil diameter so as to derive the pharmacokinetic co-
was converted to the RP. The RP-time curve efficients of these responses.
is shown in Fig. 30, bottom. The RP-time The dose-response relationship. Lom-
curve is seen to conform very well with the matzsch150 reported measurement of the re-
curve of the rate of pilocarpine release from sponse of isolated ciliary muscles to pilocar-
the insert. The horizontal line in Fig. 30, bot- pine, and his results can be interpreted ac-
tom, is the RP value at the peak time after cording to equation 20. In living monkey
single instillation of 1% pilocarpine solution eyes, Tornquist151 studied the dose-response
in the same group of subjects. Comparison relationship of pilocarpine-induced refractive

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Set.
530 Mishima October 1981

tr cr
i i

E E
a. a:
A
A A

A
A A A A
A
A

§ 0.1 0.1 A
Q.
A

5 10 5 1 0
Hours Hours
Fig. 29. Time course of the response parameter after instillation of 0.5% pilocarpine in the
presence (A) and absence (A) of soft contact lens in the eye. Left, Increased bioavailability;
right, decreased bioavailability by the presence of soft contact lens. From Mishima S and
Nagataki S: Contact Intraocular Lens Med J 4(3):22, 1978.

changes determining, in addition, the maxi- tion 23 is applicable to the cycloplegic re-
mum possible refractive change with high sponse. A similar confirmation was also ob-
doses of the drug. These results were ana- tained with the results of Smith.153
lyzed and it was found that equation 20 or Pharmacokinetic coefficients of cycloplegic
equation 23 is adaptable to pilocarpine- response. An example of the RP-time curve
induced refractive changes. In cycloplegic for the cycloplegic response to tropicamide is
responses the maximum response must be shown in Fig. 31. Again, we can see that
the state of zero amplitude of accommoda- equation 24 can describe the RP-time curve
tion, and the RP may be given by of the cycloplegic response. Accordingly, the
- A apparent elimination rate constant, the ap-
o
(26) parent absorption rate constant, the lag time,
RP =
and the peak time were calculated. A similar
where Ao is the amplitude of accommodation analysis was carried out on the data published
before the drug and A is the amplitude of in the literature, and these results are listed
accommodation after drug. From the peak in Table V.
response after instillation of various concen- The elimination rate constant of tropica-
trations of tropicamide, Yoshida152 calculated mide is larger than that found for the myd-
RP and plotted it against the instilled concen- riatic response. Likewise the elimination rate
tration of tropicamide, both on logarithmic constant of homatropine may be slightly larger
scales. She found a linear relationship, and than that seen for the mydriatic response, but
the slope of the regression line in five sub- for scopolamine and atropine the two types of
jects averaged 0.91 ± 0.21 (S.D.). Thus equa- responses have similar rate constants. It is

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 531

1100
"co 8 0
cc
% 60
CO 101
a> 4 0
f£ Pilo-20
1a 20 o
CO
o
o 2 3 4 5 6 7 I
Time (days)
o
10.

10
2 3
\
on TIME(hours)
Fig. 31. Time course of the response parameter in
the cycloplegic response to 0.5% tropicamide.
0.1 L k
Two lines for calculations of the apparent absorp-
0 1 2 3 4 5 6 7 tion and elimination rate constants are shown.
TIME (DAYS) From Yoshida S: Folia Ophthalmol Jpn 11:1009,
1976.
Fig. 30. Top, Time course of the release rate from
the pilocarpine ocular insert, Pilo-20 (by permis-
sion of Dr. J. W. Shell.); bottom, time course of as it passes from the iris root through the
the response parameter during 1 week with the ciliary muscles to the suprachoroid.158> 159
Pilo-20 in the eye. Dotted line, Peak response pa-
rameter after instillation of 1% pilocarpine solu- Pharmacokinetics of intraocular pressure
tion to the same group of subjects. The results are responses
the average in 16 eyes.
The intraocular pressure (IOP) can be de-
termined accurately, and its response to
interesting that the lag time and the peak antiglaucomatous drugs has been studied
time for cycloplegia are similar to the values by many investigators. In addition, various
obtained for the pupil responses. This prob- methods are available to study the aqueous
ably indicates that the time required for the humor dynamics in the human eye, and the
drug to reach the ciliary muscles is similar to mechanism by which the drug-induced hypo-
the time needed for access to the sphincter tensive reaction becomes manifest has been
muscles. The rapid access of cycloplegic the central subject of many investigations.
drugs to the ciliary muscles may be ac- The IOP does not respond directly to the
counted for by the outflow of the aqueous drugs as in the pupil reaction, but it is the
humor through the uveoscleral route, 156 ' 157 final result of complex mechanisms interact-
which would carry the drug with the aqueous ing with each other. Therefore the pharma-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Set.
532 Mishima October 1981

10

- I

oL
10 20 30
mmHg
Fig. 32. Relationship between the reduction in the IOP and the baseline IOP 2 hr after
instillation of 1% bupranolol solution. The correlation coefficient is 0.84. The regression equa-
tion is APj = 0.46 (Pj - 10.9). (Takase et al: unpublished results.)

Table V. Pharmacokinetic coefficients of cycloplegic responses

A B Lag time Peak time
Drug (hr-1) (hr~l) (hr) (hr) Source of data (ref.)
Tropicamide (n = 4) 1.54 ± 0.56 3.25 ± 1.52 0.1 ± 0.23 0.6 Computer fitting to data of 152
Cyclopentolate 0.44 2.3 0.15 1.05 Averaged curve from Fig. 1 of 154
Homatropine 0.35 1.3 0.2 1.6 Fig. 3 of 155
Scopolamine 0.022 2.9 0.1 1.8 Fig. 6 of 127
Atropine 0.012 1.44 0.1 3.4 Fig. 5 of 127
A = apparent elimination rate constant; B = apparent absorption rate constant (equation 24).

cokinetics of the IOP responses are compli- the pharmacokinetics of responses is the dose-
cated and can not be interpreted according to response relationship, but this aspect of the
a simple scheme that relates the response IOP response has been reported infrequently.
and the biophase drug concentrations. In- This may be because of various difficulties
deed, it is evident that the time course of the inherent to defining the drug-induced IOP
IOP response does not coincide with the time response, besides the necessity for careful
course of the intraocular drug concentration subject selection.160 Without medication the
and that a significant time lag exists between IOP is subject to diurnal variation,161 which
the two events. In the present article, there- presents a difficulty in determining the base
fore, the IOP responses to drugs are analyzed line IOP to which the drug-induced IOP
only from the phenomenologic point of view, changes are compared. Furthermore, the
and some comments are added concerning amplitude of the diurnal variation is not in-
the dose-response relationship and the dura- significant compared with the drug effects,
tion of the effects. particularly in glaucoma patients.
Dose-response relationship. The basis of In normal subjects, both eyes behave in a

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 533

Table VI. Correlation between IOP reduction in mm Hg and control level of IOP
Drug Regression equation (AP() No. of eyes Source of data (ref)

Pilocarpine
2% 0.358 (Pi-6.29 ) 19 0.82 168
2% 0.338 (Pi-7.92 ) 44 0.51 Kitazawa et al.t
4% 0.77 (Pi-15.65) 16 0.67 169
Z-Epinephrine
2% 0.768 (P, - 14.30) 22 0.85 170
Timolol
0.25% 0.33 (Pi - 0.90) 52 0.45 "
0.5% 0.37 (Pi - 4.23) 42 0.79
1.0% 0.44 (Pi - 2.34) 11 0.84
Atenolol
2% 0.697 (P, - 9.91) 15 0.86 Kitazawa et al. t
4% 0.60 (P, - 8.06) 14 0.90
Carteolol
0.5% 0.25 (P, - 1.24) 23 0.49
1% 0.54 (P, - 8.87) 47 0.54
Befunolol
0.125% 0.351 (P, - 15.1) 60 0.47 '
0.25% 0.395 (P, - 11.59) 60 0.56
0.5% 0.443 (P( - 8.04) 60 0.52
1% 0.697 (P, - 13.17) 60 0.72
Takase et al. t
Bupranolol
0.2% 0.15 (P, - 12.6) 30 0.82
0.5% 0.31 (P, - 11.3) 30 0.84
1% 0.46 (P, - 10.9) 30 0.84
* Correlation coefficient.
t Unpublished data.

Table VII. Reduction ratio in outflow pressure and outflow resistance

Drug P,-9 R, ££_ No. of eyes Source of data (ref.)

Pilocarpine (2%) 0.43 ± 0.20t 0.27 ± 0.13 19 168
Pilocarpine (4%) 0.48 0.49 ± 0.07 12 169
Physostigmine (1%) 0.47 0.47 ± 0.10 8 169
Z-Epinephrine (2%) 0.50 ± 0.10 0.52 ± 0.07 22 170
Acetazolamide (500 mg) 0.19 ± 0.12 0.07 ± 0.16 17 181
Timolol (0.5%) 0.42 ± 0.20 0.09 ± 0.38 16 183
P e (episcleral venous pressure) was assumed to be 9 mm Hg. R app is the reciprocal of the tonographic outflow facility (mm Hg min
l-)
tNumbers are mean ± S.D.

similar way, allowing drug testing in one eye error will be large, and this may be reduced
and use of the fellow eye as the control.162 by averaging repeated determinations. 170
This appears to be possible also in selected In correlating the IOP response with the
groups of glaucoma patients. 163 " 167 This is, drug concentration, the response has been
however, not necessarily the case, and the expressed in various ways: percentage reduc-
IOP values of the same eye before and after tion in reference to the control IOP, 163 ' 164> 171
administration of the drug have to be com- reduction of the IOP in mm Hg, 162 ' 167 and
pared. 168 " 170 Some drugs, when instilled in reduction ratio or the percentage reduction in
one eye, do not affect the contralateral eye, the outflow pressure. 165 ' 166 Such diversity
which allows both sides to be compared. 163~167 does not permit direct comparison of various
Others significantly affect the fellow eye and reports, and it is desirable to establish a
complicate this comparison in an individual method that will enable us to compare the
subject.171 Whatever control is adopted, the results of various studies on a common basis.

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
534 Mishima October 1981

The relationship between the drug-in- prises 4% to 12% of the total flow.157 Al-
duced 10P reduction and the control level of though it was shown to change in response to
the IOP will first be considered. An example drugs in monkey eyes, 179 we have currently
of this relationship is illustrated in Fig. 32, no clinical method of estimating it in the
where a linear correlation is significant: the human eye. It is therefore disregarded in the
higher the IOP, the larger the IOP reduction following discussion for simplicity of analysis.
with the same drug. This was first shown by When small changes occur in the outflow re-
Goodwin172 and was later recognized by Krill sistance and the rate of aqueous formation,
and Newell168 and Kronfeld.169- 170 Table VI and the IOP is reduced to a new steady-state
gives the regression equations for the rela- level, the following relationship may be de-
tionship from our own results and also from rived from the above equation to express
published data. Although the data scatter is these changes:
not small, the linear correlation is statistically
significant in all cases. The intercept of the A Pi = AZ (29)
regression line with the IOP axis averaged Pi - P e + F
9.0 ± 4 . 5 mm Hg (S.D.), which is in agree-
ment with the episcleral venous pressure re- The dose-response studies may be carried
ported by various investigators.68' 173~177 In out for the IOP changes, for the outflow resis-
addition, it is noted in Table VI that the slope tance, and also for the aqueous humor for-
of the relationship increases with increasing mation. This equation should relate the re-
concentration of the drug without signifi- sults of these studies.
cantly altering the intercept. This is in keep- The outflow resistance is the reciprocal of
ing with the report that the episcleral ve- the outflow facility of the outflow channel,
nous pressure does not change significantly which differs from the conventional tono-
with drugs such as pilocarpine,174 acetazol- graphic outflow facility, since the latter in-
amide, 174 epinephrine, 173) 177 and beta-ad- cludes the pseudofacility,178> 180 i.e., the sup-
renergic antagonists.175 On this basis, the re- pressibility of aqueous humor formation by
lationship between the IOP reduction elevating the IOP. The reciprocal of the to-
and the IOP (Ps) may be represented by nographic outflow facility is therefore de-
noted as Rapp,178 which is about 16% smaller
A Pi =_ X (27) than Rr, and the difference is probably 20% at
Pi - 9 the most.176' 178 Since the error of the tonog-
and the value X is varied by changes in the raphy can be larger than this, 182 one may
drug concentration. This can therefore be substitute R app for Rr as an approximation.
used to express the drug effects on the IOP, Consequently, the reduction ratios of the
i.e., the ratio of the outflow pressure re- outflow pressure and of the R app were calcu-
duction. lated from the published data, with the as-
Goldmann68 gave an equation relating the sumption that the episcleral venous pressure
rate of aqueous flow (F), the IOP (PJ, the is 9 mm Hg. The results are given in Table
episcleral venous pressure (Pe), and the re- VII. It is of interest to note that the reduction
sistance of outflow in the conventional out- ratio in the outflow pressure is close to the
flow route (R r ) 178 Subsequently, the uveo- reduction ratio of R app with pilocarpine and
scleral route of aqueous bulk flow became epinephrine, but they differ markedly with
known,156' 157 and this flow is insensitive to acetazolamide and timolol. This is in accor-
IOP changes.156 Incorporating these aspects dance with the understanding that the former
and denoting the rate of uveoscleral flow as U, group of drugs influences mainly the outflow
one may write resistance, and the latter group affects largely
the rate of aqueous humor formation.
(28) On the basis of the above theory, the use
F = -£- (P, - +U
of the reduction ratio of the outflow pres-
The uveoscleral flow in the human eye com- sure 165, 166 is thought to be more logical than

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 535

12

10

55
8
LU

0.08 0.125 0.25 0.5 1.0

(LED)
DRUG CONCENTRATION (%)
Fig. 33. Relationship between the average duration of the hypotensive effect and logarithm of
befunolol concentration. For definition of the effect duration see text. Vertical bars, S.D.. from
the results of eight normal subjects; LED, least effective concentration. (Takase et al: unpub-
lished results.)

other formulations of the IOP response, since the assumption of a fixed value of the
it allows direct correlation with the changes episcleral venous pressure would render a
in other parameters, i.e., the changes in the great error in the outflow pressure calcula-
outflow resistance and the rate of aqueous tions. Even if the episcleral venous pressure
humor formation. With the use of this ex- is measured individually, it is difficult to
pression of the IOP response, the results of apply the present method of calculation to
Harris and Galin163 and Drance and Nash165 the results with a few normal subjects. In
on pilocarpine effect are found to be almost fact, very low IOP is occasionally encoun-
identical. Similarly, the dose-response rela- tered in normal subjects after instillation of
tionship for the IOP effect of epinephrine 166 drugs (e.g., timolol), indicating a possibility
and its effect on the outflow resistance160 are that the episcleral venous pressure reduction
very similar. An extrapolation of the relation- actually occurs in these cases. Thus the pres-
ships gave the least effective concentration of ent method of analysis shown in Table VII is
about 0.1% to 0.15% for pilocarpine and applicable only in glaucoma and ocular hy-
0.1% to 0.2% for epinephrine. In addition, a pertension cases, where the outflow pressure
survey of various published results suggests is high enough to allow disregard of small
that the maximum drug effects on the outflow fluctuation of the episcleral venous pressure.
pressure reduction ratio would be around It must also be pointed out that the method
0.6 to 0.7. should be used on statistical basis for a large
When we study the drug effects on the number of cases and that the judgment
IOP of normal subjects, the initial outflow should be made with due consideration on
pressure is low and a slight fluctuation of the the error range of the techniques of exami-
episcleral venous pressure can occur177* 184; nation.

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
536 Mishima October 1981

Duration of hypotensive effects and drug was estimated with equation 18 to be about
concentration. In analogy to the pupil re- 15 hr. Taking the peak time of about 2 hr
sponse, the duration of the hypotensive effect into account, this drug is effective even 15 hr
of an antiglaucomatous drug may be defined after instillation, the result being consistent
as the interval between the time of peak re- with the report of Drance et al.186 Thus the
sponse and the time when the effect termi- knowledge of these two pharmacokinetic co-
nates. In the case of a long-acting antiglau- efficients is useful in determining the concen-
comatous drug, the termination time may be tration of the therapeutic preparation and the
obtained by extrapolation of the time course frequency of its application.
of the IOP changes. The duration of the effect
may then be correlated with the drug con- Concluding remarks
centration, but because of marked IOP fluc- The noninvasive method of analyzing the
tuation in an individual, such a correlation response in relation to the dose of the drug
could be seen only with use of the average enables us to estimate the rates at which
duration. An example of the correlation is some instilled drugs are absorbed and elimi-
shown in Fig. 33, where the average duration nated in the anterior segment of the human
of the hypotensive effects in eight normal eye. Knowing these pharmacokinetic coef-
subjects is plotted against the logarithm of ficients, it is possible to construct curves of
concentration of befunolol solution applied the intraocular drug concentration with vari-
topically.. The correlation is seen to be linear, ous modalities of drug application, e.g., sin-
and this may be analyzed by equation 18. gle or multiple drop instillation at certain in-
Since the relationship between the intraocu- tervals, application of ointment, or a constant-
lar drug concentration and the IOP response rate drug delivery device. This should be
is unknown, the slope of the regression line useful in planning the therapeutic regimen
was called the rate of effect disappearance according to the patient's need. However,
instead of the elimination rate constant of the drug therapy in practice is very complicated
drug. Extrapolation of the line gives the least and many factors must be taken into considera-
effective concentration, and this value agrees tion, such as slow development of effects187'188
well with the least effective concentration or individual differences in drug sensitivity
calculated by extrapolation of the dose- because of age, 189 heredity, progress of
response relationship.162 The least effective pathologic processes, or development of drug
concentration was 0.13% to 0.16% for bu- tolerance.160' 18T Thus the pharmacokinetic
pranolol, about 0.08% with befunolol, and coefficients obtained by the present analysis
about 0.045% for timolol. The rate of effect on the basis of a simple model may give only
disappearance was about 0.2 hr" 1 for bupran- a rough estimate of the dynamic state of drug
olol, 0.26 hr" 1 for befunolol, and 0.06 h r 1 for behavior in an actual clinical situation. It is
timolol. A similar linear correlation between hoped, however, that such pharmacokinetic
the duration of hypotensive effect and the results in the human eye will give us addi-
logarithm of the dose is also seen in pub- tional confidence in clinical practice.
lished data185 on the hypotensive effect of
colchicine in rabbit eyes. I thank Dr. D. M. Maurice for his kind advice during
the preparation of this manuscript. Acknowledgment is
The least effective concentration and the also due to Drs. J. W. Shell, M. Takase, Y. Kitazawa, S.
rate of effect disappearance permit estima- Nagataki, M. Araie, S. Shirato, and H. Yamauchi for
their cooperation and permission for the use of their
tion of the effect duration of a given concen-
unpublished works.
tration of a hypotensive drug. For example,
the least effective concentration of pilocar-
pine may be assumed to be about 0.1% and REFERENCES
the rate of effect disappearance about 0.35
1. Gibaldi M and Perrier D: Pharmacokinetics, Drugs
hr" 1 , a value corresponding to the apparent and Pharmaceutical Sciences, New York, 1975,
elimination rate constant of this drug in the Marcel Dekker, Inc., vol. 1.
iris. The effect duration of 8% pilocarpine 2. Wagner JG: Fundamentals of clinical pharmacoki-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 537

netics. Hamilton, 111., 1979, Drug Intelligence drug bioavailability I: Evaluation of fluorometh-
Publishing, Inc. olone. J Pharm Sci 64:931, 1975.
3. Mishima S, Gasset A, Klyce SD Jr, and Baum JL: 22. Sieg JW and Robinson JR: Mechanistic studies on
Determination of tear volume and tear flow. IN- transcorneal permeation of pilocarpine. J Pharm
VEST OPHTHALMOL 5:264, 1966. Sci 65:1816, 1976.
4. Adler CA, Maurice DM, and Paterson ME: The 23. Sugar J, Burde RM, Sugar A, Waltman ST, Kripa-
effect of viscosity of the vehicle on the penetration lani KJ, Weliky I, and Becker B: Tetrahydro-
of fluorescein into the human eye. Exp Eye Res triamcinolone and triamcinolone. I. Ocular pene-
11:34, 1971. tration. INVEST OPHTHALMOL 11:890, 1972.
5. Benedetto DA, Shah DO, and Kaufman HE: The 24. Beasley H, Boltralik JJ, and Baldwin HA: Chlor-
instilled fluid dynamics and surface chemistry of amphenicol in aqueous humor after topical appli-
polymers in the preocular tear film. INVEST cation. Arch Ophthalmol 93:184, 1975.
OPHTHALMOL 14:887, 1975. 25. Massey JY, Hanna C, Goodart R, and Wallace T:
6. Sugaya M and Nagataki S: Kinetics of topical Effect of drug vehicle on human ocular retention of
pilocarpine in the human eye. Jpn J Ophthalmol topically applied tetracycline. Am J Ophthalmol
22:127, 1978. 81:151, 1976.
7. Trueblood JH, Rossomondo RM, Carlton WH, and 26. Leibowitz HM, Berrospi AR, Kupferman A, Re-
Wilson LA: Corneal contact times of ophthalmic stropo GV, Galvis V, and Alvarez JA: Penetration
vehicles. Arch Ophthalmol 93:127, 1975. of topically administered prednisolone acetate into
8. Harberger R, Hanna C, and Boyd CM: Effects of the human aqueous humor. Am J Ophthalmol 83:
drug vehicles on ocular contact time. Arch Oph- 402, 1977.
thalmol 93:42, 1975. 27. Krohn DL and Breitfeller JM: Transcorneal flux of
9. Kitazawa Y, Takahashi O, Matsumoto T, Iinuma T, topical pilocarpine to the human aqueous. Am J
and Tateno Y: The effect of hydroxypropylmethyl- Ophthalmol 87:50, 1979.
cellulose on corneal contact times of ophthalmic 28. Krohn DL: Flux of topical pilocarpine to the
solution. Acta Soc Ophthalmol Jpn 81:1509, 1977. human aqueous. Trans Am Soc Ophthalmol 76:
10. Murai Y: Studies of tear flow using sodium per- 502, 1978.
technetate Tc"m. Jpn J Ophthalmol 20:283, 1976. 29. Maurice DM: Prolonged release systems and topi-
11. Maurice DM: Structures and fluids involved in the cally applied drugs. Sight Sav Rev 42:42, 1972.
penetration of topically applied drugs. Clinical 30. Jones RF and Maurice DM: New methods of mea-
pharmacology of the anterior segment. Int Oph- suring the rate of aqueous flow in man with fluores-
thalmol Clin 20(3):7, 1980. cein. Exp Eye Res 5:208, 1966.
12. Harris JE: Problems in drug penetration. In Sym- 31. Yoshida S and Mishima S: A pharmacokinetic anal-
posium on Ocular Therapy, Leopold IH, editor. ysis of the pupil response to topical pilocarpine and
St. Louis, 1968, The C. V. Mosby Co., vol. 3, pp. tropicamide. Jpn J Ophthalmol 19:121, 1975.
96-105. 32. Nagataki S and Mishima S: Pharmacokinetics of
13. Doane MG, Jensen AD, and Dohlman CH: Pene- instilled drugs in the human eye. Clinical pharma-
tration routes of topically applied eye medications. cology of the anterior segment. Holly, Int Oph-
Am J Ophthalmol 85:383, 1978. thalmol Clin 20(3):33, 1980.
14. Benson H: Permeability of the cornea to topically 33. Mishima S and Nagataki S: Pharmacology of
applied drugs. Arch Ophthalmol 91:313, 1974. ophthalmic solutions. Contact Intraocular Lens
15. Chrai SS, Patton TF, Mehta A, and Robinson JR: Med J 4(3):22, 1978.
Lacrimal and instilled fluid dynamics in rabbit 34. Ehlers N: The precorneal film. Biomicroscopical,
eyes. J Pharm Sci 62:1112, 1973. histological and chemical investigations. Acta Oph-
16. Chrai SS, Makoid MC, Eriksen ST, and Robinson thalmol Suppl. 81, 1965.
JR: Drop size and initial doseing frequency prob- 35. Maurice DM: The dynamics and drainage of tears.
lems of topically applied ophthalmic drugs. J Int Ophthalmol Clin 13:103, 1973.
Pharm Sci 63:333, 1974. 36. Mishima S: Some physiological aspects of the pre-
17. Chrai SS and Robinson JR: Corneal permeation of corneal tear film. Arch Ophthalmol 73:233, 1965.
topical pilocarpine nitrate in the rabbit. Am J 37. Holly FJ: Formation and stability of the tear film.
Ophthalmol 77:735, 1974. Int Ophthalmol Clin 13:73, 1973.
18. Chrai SS and Robinson JR: Ocular evaluation of 38. Patton TF and Robinson JR: Pediatric dosing con-
methylcellulose vehicle in albino rabbits. J Pharm siderations in ophthalmology. J Pediatr Ophthal-
Sci 63:1218, 1974. mol 13:171, 1976.
19. Mikkelson TJ, Chrai SS, and Robinson JR: Altered 39. Ehlers N: Pharmacology of the conjunctival sac. In
bioavailability of drugs in the eye due to drug- Drugs and Ocular Tissues, Dikstein S, editor.
protein interaction. J Pharm Sci 62:1648, 1973. Basel, 1977, S. Karger, pp. 23-56.
20. Patton TF and Robinson JR: Ocular evaluation of 40. Robin JS and Ellis PP: Ophthalmic ointments. Surv
polyvinyl alcohol vehicle in rabbits. J Pharm Sci Ophthalmol 22:335, 1978.
64:1312, 1975. 41. Klein M: Beitrag zur Resorption aus Augensalben.
21. Sieg JW and Robinson JR: Vehicle effects on ocular II. Mitteilung, Die Resorption des Atropins. Al-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
538 Mishima October 1981

brecht Von Graefes Arch Klin Exp Ophthalmol tration and hypotensive effect. Acta Soc Ophthal-
129:413, 1933. mol Jpn 83:537, 1979.
42. Scruggs J, Wallace T, and Hanna C: Route of ab- 59. Campbell RC, Neault RW, and Brubaker RF: The
sorption of drug and ointment after application to corneal penetration of 6-aminohexanoic acid. IN-
the eye. Ann Ophthalmol 10:267, 1978. VEST OPHTHALMOL 15:857, 1976.
43. Makoid MC, Sieg JW, and Robinson JR: Corneal 60. Lazar M, Lieberman TW, Furman M, and Leopold
drug absorption: an illustration of parallel first- IH: Ocular penetration of hetrazan in rabbits. Am J
order absorption and rapid loss of drug from ab- Ophthalmol 66:215, 1968.
sorption depot. J Pharm Sci 65:150, 1976. 61. Maurice DM: The permeability to sodium ions of
44. Maurice DM: The cornea and sclera. In The Eye, the living rabbit's cornea. J Physiol 112:367, 1951.
Davson H, editor. New York, 1969, Academic 62. Mishima S, Hattori E, and Yamauchi H: In vivo
Press, Inc., vol. 1, pp. 489-600. determination of the corneal permeability. Jpn J
45. Lazare R and Horlington M: Pilocarpine levels in Ophthalmol 15:183, 1971.
the eyes of rabbits following topical application. 63. Sawa M, Araie M, and Nagataki S: Permeability of
Exp Eye Res 21:281, 1975. the human corneal endothelium to fluorescein. Jpn
46. Patton TF and Francoeur M: Ocular bioavailability J Ophthalmol 25:60, 1981.
and systemic loss of topically applied ophthalmic 64. Kinsey VE and Reddy DVN: Chemistry and
drugs. Am J Ophthalmol 85:225, 1978. dynamics of aqueous humor. In The Rabbit in Eye
47. Kupferman A and Leibowitz HM: Topically ap- Research, Prince JH, editor. Springfield, 111.,
plied steroids in corneal disease. IV. The role of 1964, Charles C Thomas, Publisher, pp. 218-319.
drug concentration in stromal absorption of pred- 65. McMaster PRB and Macri FJ: Secondary aqueous
nisolone acetate. Arch Ophthalmol 91:377, 1974. humor outflow pathways in the rabbit, cat and
48. Hull DS, Hine JE, Delhauser F, and Hyndiuk RA: monkey. Arch Ophthalmol 79:297, 1968.
Permeability of the isolated rabbit cornea to corti- 66. Lippert U, Mosebach KO, and Dardenne MU: Un-
costeroids. Invest Ophthalmol 13:457, 1974. tersuchungen uber die Verteilung von Dexametha-
49. Yamauchi H, Kito H, and Uda K: Studies on in- son im Kaninchenauge nach lokaler Applikation.
traocular penetration and metabolism of fluoro- Klin Monatsbl Augenheilkd 164:225, 1974.
metholone in rabbits: a comparison between dexa- 67. Conrad JM and Robinson JR: Aqueous chamber
methasone and prednisolone acetate. Jpn J Oph- drug distribution volume measurement in rabbits.
thaimol 19:339, 1975. J Pharm Sci 66:219, 1977.
50. Schoenwald RD and Boltralik JJ: A bioavailability 68. Goldmann H: Abflussdruck, Minutenvolumen
comparison in rabbits of two steroids formulated as und Widerstand der Kammerwasserstromung des
high-viscosity gels and reference aqueous prepara- Menschen. Doc Ophthalmol 5:278, 1951.
tions. INVEST OPHTHALMOL VIS SCI 18:61, 1979. 69. Jones RF and Maurice DM: A simple photographic
51. Kupferman A and Leibowitz HM: Topically method of measuring the volume of the anterior
applied steroids in corneal disease. VI. Kinetics of chamber. Exp Eye Res 2:233, 1963.
prednisolone sodium phosphate. Arch Ophthalmol 70. Makoid MC and Robinson JR: Pharmacokinetics of
92:331, 1974. topically applied pilocarpine in the albino rabbit
52. Green K and Downs SJ: Prednisolone phosphate eye. J Pharm Sci 68:435, 1979.
penetration into and through the cornea. INVEST 71. Newsome DA and Stern R: Pilocarpine adsorption
OPHTHALMOL 13:316, 1974. by serum and ocular tissues. Am J Ophthalmol
53. Short C, Keats RH, Donovan EF, Wyman M, and 77:918, 1974.
Murdick PW: Ocular penetration studies. I. Topi- 72. Ellis PP, Littlejohn K, and Deitrich RA: Enzymatic
cal administration of dexamethasone. Arch Oph- hydrolysis of pilocarpine. INVEST OPHTHALMOL
thalmol 75:689, 1966. 11:747, 1972.
54. Kupferman A and Leibowitz HM: Topically 73. Seidenhamel RJ, Tye A, and Patil PN: An analysis
applied steroids in corneal disease. V. Dexa- of ephedrine mydriasis in relationship to iris pig-
methasone alcohol. Arch Ophthalmol 92:329, 1974. mentation in the guinea-pig eye in vitro. J Pharma-
55. Kupferman A and Leibowitz HM: Penetration of col Exp Ther 171:205, 1970.
fluorometholone. into the cornea and aqueous 74. Patil PN: Cocaine-binding by the pigmented and
humor. Arch Ophthalmol 93:425, 1975. the non-pigmented iris and its relevance to the
56. Araie M, Takase M, Yokoyama Y, and Kitagawa M: mydriatic effect. INVEST OPHTHALMOL 11:739,
Timolol maleate: pharmacokinetic analysis of ocu- 1972.
lar penetration in the rabbit eye and effects on 75. Lyons JS and Krohn DL: Pilocarpine uptake by
human aqueous humor dynamics. Acta Soc Oph- pigmented uveal tissue. Am J Ophthalmol 75:885,
thalmol Jpn 84:2139, 1980. 1973.
57. Schmitt CJ, Lotti VJ, and LeDourarec JC: Pene- 76. Patil PN and Jacobowitz D: Unequal accumulation
tration of timolol into the rabbit eye. Arch of adrenergic drugs by pigmented and nonpig-
Ophthalmol 98:547, 1980. mented iris. Am J Ophthalmol 78:470, 1974.
58. Komuro S, Nanba H, and Takase M: Studies of 77. Patil PN, Shimada K, Feller DR, and Malspeis L:
bupranolol therapy for glaucoma. I. Ocular pene- Accumulation of (—)14C-ephedrine by the pig-

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 539

merited and nonpigmented iris. J Pharmacol Exp activity of procaine. J Pharm Sci 51:1162, 1962.
Ther 188:342, 1974. 96. Smolen VF and Siegel FP: Procaine interaction
78. Barza M, Kane A, and Baum J: Marked differences with the corneal surface and its relation to anes-
between pigmented and albino rabbits in the con- thesia. J Pharm Sci 57:378, 1968.
centration of clindamycin in iris and choroid- 97. Ishikawa S, Oono S, and Hikita H: Drugs affecting
retina. J Infect Dis 139:203, 1979. the iris muscle. In Drugs and Ocular Tissues, Diks-
79. Barza M, Baum J, and Kane A': Inhibition of an- tein S, editor. Basel, 1977, S. Karger, pp. 288-382.
tibiotic activity in vitro by synthetic melanin. An- 98. Ogle KN, Whisnaut RA, and Hazelrig JB: Quanti-
timicrob Agents Chemother 10:569, 1976. tative study of pupil response to miotic drugs. IN-
80. Salazar M and Patil PN: An explanation of the long VEST OPHTHALMOL 5:176, 1966.
duration of mydriatic effect of atropine in eye. IN- 99. Gambill HD, Ogle KN, and Kearns TP: Mydriatic
VEST OPHTHALMOL 15:671, 1975. effect of four drugs determined with pupillograph.
81. Salazar M, Shimada K, and Patil PN: Iris pig- Arch Ophthalmol 77:740, 1967.
mentation and atropine mydriasis. J Pharmacol 100. Davanger M: The pupillary dilatation curve after
Exp Ther 197:79, 1976. mydriatics. Acta Ophthalmol 49:565, 1971.
82. Patil PN: Antimuscarinic effects of stereoisomers of 101. Smolen VF: Determination of time course of in
tropicamide on rabbit iris sphincter. INVEST vivo pharmacological effects from in vitro drug re-
OPHTHALMOL VIS SCI 17:65, 1978. lease testing. J Pharm Sci 60:878, 1971.
83. Shell JW and Baker RW: Diffusional systems for 102. Schoenwald RD and Smolen VF: Drug-absorption
controlled release of drugs to the eye. Ann Oph- analysis from pharmacological data. II. Trans-
thalmol 6:1037, 1974. corneal biophasic availability of tropicamide. J
84. Sendelbeck L, Moore D, and Urquart J: Compara- Pharm Sci 60:1039, 1971.
tive distribution of pilocarpine in ocular tissues of 103. Smolen VF: Applications of a pharmacological
the rabbit during administration by eye drop or by method of drug absorption analysis to the study of
membrane-controlled delivery systems. Am J Oph- the bioavailability characteristics of mydriatic
thalmol 80:274, 1975. drugs. Can J Pharm Sci 7:7, 1972.
85. Harnisch JP, Hoffmann F, and Dumitrescu L: 104. Smolen VF: Bioavailability and pharmacokinetic
Side-effects of local anesthetics on the corneal epi- analysis of drug responding systems. Ann Rev
thelium of the rabbit eye. Albrecht Von Graefes Pharmacol Toxicol 18:495, 1978.
Arch Klin Exp Ophthalmol 197:71, 1975. 105. Wagner JG: Kinetics of pharmacologic response. I.
86. Burstein NL and Klyce SD: Electrophysiologic and Proposed relationships between response and drug
morphologic effects of ophthalmic preparations on concentration in the intact animal and man. J
rabbit cornea epithelium. INVEST OPHTHALMOL VIS Theor Biol 20:173, 1968.
SCI 16:899, 1977. 106. Gregersen W: The tissue spaces in the human iris
87. Matsumoto T, Hayashi K, Tsuchisaka H, and Araie and their communication with the anterior cham-
M: Pharmacokinetics of surface anesthetics in the ber by way of the iridic crypts. Acta Ophthalmol
human cornea. Jpn J Ophthalmol 25:335, 1981. 36:819, 1958.
88. Cochet P and Bonnet R: L'esthesiometrie cor- 107. Smelser GK and Ishikawa T: Investigation on the
neenne. Realisation et interet pratique. Bull Soc porosity of the iris. In Proceedings of the Nine-
Ophtalmol Fr 61:541, 1961. teenth Congress on Ophthalmology. Bombay,
89. Swan KC and White NG: Corneal permeability. I. 1962, Times of India Press, vol. 1, pp. 612-623.
Factors affecting penetration of drugs into the cor- 108. Ohara K: Effects of cholinergic agonists on isolated
nea. Arch Ophthalmol 25:1043, 1942. iris sphincter muscles; a pharmacodynamic study.
90. Levy G: Relationship between elimination rate of Jpn J Ophthalmol 21:516, 1977.
drugs and rate of decline of their pharmacologic 109. Taylor SA: Quantitative studies on acetylcholine
effects. J Pharm Sci 53:342, 1964. receptors of the isolated iris sphincter muscle. J
91. Levy G: Kinetics of drug action in man. Acta Physiol 239:6, 1974.
Pharmacol 29:203, 1971. 110. Ohara K: Action sites of some cholinergic agonists
92. Conrad JM, Reay WA, Polcyn RE, and Robinson in the isolated iris sphincter muscles of rabbits. Jpn
JR: Influence of tonicity and pH on lacrimation and J Ophthalmol 22:280, 1978.
ocular drug bioavailability. J Parent Drug Assoc 111. Haris LS, Shimmyo M and Hughes J: Dose-
32:149, 1978. response of cholinergic agonists on cat irides. Arch
93. Poise KA, Keener RJ, and Jauregui MJ: Dose- Ophthalmol 91:299, 1974.
response effects of corneal anesthetics. Am J 112. Patil PN: Adrenergic receptors of the bovine iris
Optom Physiol Opt 55:8, 1978. sphincter. J Pharmacol Exp Ther 166:299, 1969.
94. Gerlough TD: The influence of pH on the activity 113. Mochizuki M and Mishima S: The adrenergic ef-
of certain local anesthetics as measured by the fects on cyclic AMP and tension of the sphincter
rabbit's cornea method. J Pharmacol Exp Ther pupillae of the rabbit. Doc Ophthalmol Proc Ser
41:307, 1931. 18:353, 1979.
95. Susina SV, Hitter FD, Siegel FP, and Blake MI: 114. Beaver WT and Riker EG: The quantitative
Effect of deuterium oxide on local anesthetic evaluation of autonomic drugs on the isolated

Downloaded from iovs.arvojournals.org on 05/24/2020

 Invest. Ophthalmol. Vis. Sci.
540 Mishima October 1981

eye. J Pharmacol Exp Ther 138:48, 1962. 134. Mueller WH and Deardorff DL: Ophthalmic vehi-
115. Quilliam JP: A quantitative method for the study of cles: the effect of methylcellulose on the penetra-
the reactions of the isolated cat's iris. J Physiol tion of homatropine hydrobromide through the
110:237, 1949. cornea. J Am Pharm Assoc 45:334, 1956.
116. Sears M: Adrenergic supersensitivity of the scor- 135. Haas JS and Merrill DL: The effect of methylcel-
butic iris. Trans Am Ophthalmol Soc 71:536, 1973. lulose on responses to solutions of pilocarpine. Am
117. Ariens EJ: Molecular Pharmacology. New York, J Ophthalmol 54:21, 1962.
1964, Academic Press, Inc., vol. 1. 136. Linn ML and Kones LT: Rate of lacrimal excretion
118. Newsome DA and Loewenfeld IE: Iris mechanics. of ophthalmic vehicles. Am J Ophthalmol 65:76,
II. Influence of pupil size on details of iris struc- 1968.
ture. Am J Ophthalmol 71:553, 1971. 137. Swanson AA, Jetter DJ, and Tucker P: Ophthalmic
119. Loewenfeld IE and Newsome DA: Iris mechanics. vehicles. II. Comparison of ointment and polyvinyl
I. Influence of pupil size on dynamics of pupillary alcohol 1.4%. Ophthalmologica 160:265, 1970.
movements. Am J Ophthalmol 71:347, 1971. 138. Waltman SR, Buerk K, and Foster CS: Effects of
120. Langham ME, Kitazawa Y, and Hart RW: Adren- ophthalmic ointments on intraocular penetration of
ergic responses in the human eye. J Pharmacol Exp topical fluorescein in rabbits and man. Am J
Ther 179:47, 1971. Ophthalmol 78:262, 1974.
121. Langham ME and Diggs EM: Quantitative studies 139. Hardberger RE, Hanna C, and Goodart R: Effects
of the ocular response to norepinephrine. Exp Eye of drug vehicles on ocular uptake of tetracycline.
Res 13:161, 1972. Am J Ophthalmol 80:133, 1975.
122. Bhattacherjee P: Uptake of (3H)noradrenaline by 140. Green K and Downs SJ: Ocular penetration of
the ocular tissues of rabbit. Exp Eye Res 9:73, pilocarpine in rabbits. Arch Ophthalmol 93:1165,
1970. 1975.
123. Lowenstein O and Loewenfeld IE: Effect of 141. Kaufman HE, Uotila MH, Gasse AR, Wood TO,
physostigmine and pilocarpine on iris sphincter of and Ellison ED: The medical uses of soft contact
normal man. Arch Ophthalmol 50:311, 1953. lenses. Trans Am Acad Ophthalmol Otolaryngol
124. Smith SA, Smith SE, and Lazare R: An increased 75:361, 1971.
effect of pilocarpine on the pupil by application of 142. Podos SM, Becker B, Asseff C, and Hartstein J:
the drug in oil. Br J Ophthalmol 62:314, 1978. Pilocarpine therapy with soft contact lenses. Am J
125. Borgmann H and Wurster W: Der Einfluss un- Ophthalmol 73:336, 1972.
terschiedlicher Konzentrationen und Vehikel auf 143. Maddox YT and Bernstein HN: An evaluation of
die Pilocarpin-Miosis. I. Unterschiedliche Konzen- the bionite hydrophilic contact lenses. Am J
trationen. Klin Monatsbl Augenheilkd 163:44, Ophthalmol 73:336, 1972.
1973. 144. Hull DS, Edelhauser HF, and Hyndiuk RA: Ocu-
126. Siebeck R and Diehl W: Verlau medikamentoser lar penetration of prednisolone and the hydrophilic
Mydriasis und Miosis unter Stabilisierung der contact lens. Arch Ophthalmol 92:413, 1974.
iibrigen pupillomotorischen Bedingungen. Al- 145. Ruben M and Watkins R: Pilocarpine dispensation
brecht Von Graefes Arch Klin Exp Ophthalmol for the soft hydrophilic contact lens. Br J Ophthal-
168:157, 1965. mol 59:455, 1975.
127. Marron J: Cycloplegia and mydriasis by use of at- 146. Armaly MF anf Rao KR: The effect of pilocarpine
ropine, scopolamine and homatropine-paredrine. ocusert with different release rates on ocular pres-
Arch Ophthalmol 23:340, 1940. sure. INVEST OPHTHALMOL 12:491, 1973.
128. Chen KK and Poth EJ: Racial differences as illus- 147. Allansmith MR, Lee JR, McClellan HB, and
trated by the mydriatic action of cocaine, euph- Dohlman CH: Evaluation of a sustained release
thalmine and ephedrine. J Pharmacol Exp Ther hydrocortisone ocular insert in humans. Trans Am
36:429, 1929. Acad Ophthalmol Otolaryngol 79:128, 1975.
129. Barbee RF and Smith WD: A comparative study of 148. Horie T, Takahashi O, Takase M, and Kitazawa Y:
mydriatic and cycloplegic agents. Am J Ophthal- Clinical evaluation of pilocarpine ocusert. I.
mol 44:617, 1957. Clinicopharmacological study. Ganka Rinsho Iho
130. Obianwu HO and Rand MJ: The relationship be- 71:888, 1977.
tween the mydriatic action of ephedrine and the 149. Tuchisaka H and Takase M: Effects of pilocarpine
colour of the iris. Br J Ophthalmol 49:264, 1965. ocusert upon IOP, pupil diameter and refraction in
131. Emiru VP: Response to mydriatics in the African. patients with ocular hypertension. Acta Soc
Br J Ophthalmol 55:538, 1971. Ophthalmol Jpn 82:219, 1978.
132. Melikian HE, Lieberman TW, and Leopold IH: 150. Lommatzsch P: Ueber Versuche mit Pilokarpin am
Ocular pigmentation and pressure and outflow re- isolierten Ciliarmuskel und Sphincter iridis. Al-
sponse to pilocarpine and epinephrine. Am J brecht Von Graefes Arch Klin Exp Ophthalmol
Ophthalmol 72:70, 1971. 165:487, 1963.
133. Harris LS and Galin MA: Effect of ocular pig- 151. Tornquist G: Comparative studies of the effect of
mentation on hypotensive response to pilocarpine. pilocarpine on the pupil and on the refraction in
Am J Ophthalmol 72:923, 1971. two species of monkey (Cercopithecus ethiops and

Downloaded from iovs.arvojournals.org on 05/24/2020

 Volume 21
Number 4 Clinical pharmacokinetics of the eye 541

Macacca irus). INVEST OPHTHALMOL 3:388, 1964. response and duration of action. Arch Ophthalmol
152. Yoshida S: Analysis of cycloplegic response to topi- 95:605, 1977.
cal tropicamide. Folia Ophthalmol Jpn 11:1009, 172. Goodwin RC: Changes in the intraocular pressure
1976. in normal and glaucomatous eyes following fur-
153. Smith SE: Dose-response relationships in tropi- methide. Am J Ophthalmol 34:1139, 1951.
camide-induced mydriasis cycloplegia. Br J Clin 173. Podos SM, Minas TF, and Macri FJ: A new in-
Pharmacol 1:37, 1974. strument to measure episcleral venous pressure.
154. Iwata O, Sakai A, Mori T, and Tsusaka H: Clinical Arch Ophthalmol 80:209, 1968.
use of cycloplegic-cyclogyl. Rinsho-Ganka (Jpn J 174. Linner E: Further studies of the episcleral venous
Clin Ophthalmol) 22:1542, 1968. pressure in glaucoma. Am J Ophthalmol 41:646,
155. Thorne FH and Murphey HS: Cycloplegics. Arch 1956.
Ophthalmol 22:274, 1939. 175. Wettrell K, Wilke K, and Pandolfi M: Effect of
156. Bill A: Aqueous humor dynamics in monkeys beta-adrenergic agonists and antagonists on re-
(Macacca irus and Cercopithecus ethiops). Exp Eye peated tonometry and episcleral venous pressure.
Res 11:195, 1971. Exp Eye Res 24:613, 1977.
157. Bill A and Phillips CI: Uveoscleral drainage of 176. Kupfer C and Sanderson P: Determination of
aqueous humour in human eyes. Exp Eye Res pseudofacility in the eye of man. Arch Ophthalmol
12:275, 1971. 80:194, 1968.
158. Inomata H, Bill A, and Smelser GK: Unconven- 177. Kupfer C, Gassterland D, and Ross K: Studies of
tional routes of aqueous humor outflow in cyno- aqueous humor dynamics in man. II. Measure-
molgus monkey (Macacca irus). Am J Ophthalmol ments in young normal subjects using acetazol-
73:893, 1972. amide and Z-epinephrine. INVEST OPHTHALMOL
159. Sherman SH, Green K, and Laties AM: The fate of 10:523, 1971.
anterior chamber fluorescein in the monkey eye. I. 178. Goldmann H: On pseudofacility. Bibl Ophthalmol
The anterior chamber outflow pathways. Exp Eye 76:1, 1968.
Res 27:159, 1978. 179. Bill A: Early effects of epinephrine on aqueous
160. Kronfeld PC: Dose-effect relationships as an aid in humor dynamics in vervet monkeys (Cercopithecus
the evaluation of ocular hypotensive drugs. INVEST ethiops). Exp Eye Res 8:35, 1969.
OPHTHALMOL 3:258, 1964. 180. Barany EH: A mathematical formulation of in-
161. Horie T and Kitazawa Y: The clinical significance of traocular pressure as dependent on secretion, ul-
diurnal pressure variation in primary open-angle trafiltration, bulk outflow, and osmotic reabsorp-
glaucoma. Jpn J Ophthalmol 23:310, 1979. tion of fluid. INVEST OPHTHALMOL 2:584, 1963.
162. Takase M, Komuro S, Nanba H, and Araie M: Ef- 181. Becker B and Constant MA: Experimental tonog-
fects of topical bupranolol hydrochloride on the in- raphy. The effect of carbonic anhydrase inhibitor
traocular pressure. Jpn J Ophthalmol 22:142, 1978. acetazolamide on aqueous flow. Arch Ophthalmol
163. Harris LS and Galin MA: Dose response analysis of 54:321, 1955.
pilocarpine-induced ocular hypotension. Arch 182. Becker B and Friedenwald JS: Clinical aqueous
Ophthalmol 84:605, 1970. outflow. Arch Ophthalmol 50:557, 1953.
164. Harris LS: Dose-response analysis of echothio- 183. Zimmerman TJ, Harbin R, Pett M, and Kaufman
phate iodide. Arch Ophthalmol 86:502, 1971. HE: Timolol and facility of outflow. INVEST
165. Drance SM and Nash PA: The dose response of OPHTHALMOL VIS SCI 16:623, 1977.
human intraocular pressure to pilocarpine. Can J 184. Wilke K: Early effects of epinephrine and pilocar-
Ophthalmol 6:9, 1971. pine on the intraocular pressure and the episcleral
166. Kitazawa Y: Dose response analysis of ocular venous pressure in the normal human eye. Acta
hypotensive effects of epinephrine and norepi- Ophthalmol 52:231, 1974.
nephrine. Jpn J Ophthalmol 16:30, 1972. 185. Bhattacherjee P and Eakins KE: The intraocular
167. Krieglstein GK and Leydhecker W: The dose- pressure lowering effect of colchicine. Exp Eye Res
response relationships of dipivalyl epinephrine in 27:649, 1978.
open-angle glaucoma. Albrecht Von Graefes Arch 186. Drance SM, Bensted M, and Schulzer M: Pilocar-
Klin Exp Ophthalmol 205:141, 1978. pine and intraocular pressure. Arch Ophthalmol
168. Krill AE and Newell FW: Effects of pilocarpine on 91:104, 1974.
ocular tension dynamics. Am J Ophthalmol 57:34, 187. Barany EH: A pharmacologist looks at medical
1964. treatment in glaucoma—in retrospect and in pros-
169. Kronfeld PC: The efficacy of combinations of ocu- pect. Am J Ophthalmol 86:80, 1979.
lar hypotensive drugs. Arch Ophthalmol 78:140, 188. Ballintine EJ and Garner LL: Improvement of co-
1967. efficient of outflow in glaucomatous eyes. Pro-
170. Kronfeld PC: Early effects of single and repeated longed local treatment with epinephrine. Arch
doses of Z-epinephrine in man. Am J Ophthalmol Ophthalmol 66:314, 1961.
72:1058, 1971. 189. Borthne A and Davanger M: Mydriatics and age.
171. Zimmerman TJ and Kaufman HE: Timolol, dose- Acta Ophthalmol 49:380, 1971.

Downloaded from iovs.arvojournals.org on 05/24/2020