Investigative Ophthalmology & Visual Science, Vol. 29, No. 10, October 1988
Copyright © Association for Research in Vision and Ophthalmology
Pathogenesis of Blepharoconjunctivitis Complicating
13-cis-Retinoic Acid (Isotretinoin) Therapy in a
Laboratory Model
Robert W. Lambert ond Ronald E. Smith
Systemic treatment of adult male New Zealand albino rabbits with 13-cis-retinoic acid (isotretinoin)
resulted in a reduction in the size of the meibomian gland. Clinical signs of toxicity included weight
loss, alopecia, dry skin and mild conjunctiva! erythema with crusting on the eyelid margin. Histopathologic findings included thickening of duct and ductule epithelium, decrease in acinar tissue, accentuation of basaloid cells and evidence of periacinar fibrosis. The model presents the first experimental
data to indicate that systemic 13-cis-retinoic acid effects meibomian gland structure in a laboratory
model. Future functional studies of this model may yield important insights into the relationships
between meibomian gland morphology, function, the ocular surface and the pathogenesis of blepharoconjunctivitis. Invest Ophthalmol Vis Sci 29:1559-1564,1988
Materials and Methods
Isotretinoin (13-cis-retinoic acid) is used for the
treatment of acne vulgaris and keratinizing dermatoses.1'2 Therapeutic effects of the drug as used for
dermatologic conditions include decreased sebum
production, changes in the surface lipid composition
and inhibition of keratinization.23 One of the more
common side effects of this therapy is blepharoconjunctivitis, which has been reported in 20-50% of
such patients. 12 ' 4 Although Fraunfelder and colleagues2 suggested that meibomian gland dysfunction
may be implicated, to date there has been no evidence to support this hypothesis. Since studies in patients undergoing isotretinoin therapy for acne have
revealed a reduction in the size of skin sebaceous
glands,5'6 and since there are similarities between skin
sebaceous glands and the meibomian glands,7 we
studied animals fed 13-cis-retinoic acid to determine
if pathologic changes occur in the meibomian gland.
Herein we describe the histopathology of the rabbit
meibomian gland after systemic treatment with Accutane® (Hoffman-La Roche, Inc., Nutley, NJ).
Eighteen male New Zealand albino rabbits were
treated with isotretinoin at the following doses: 1-2
mg/kg/day (Group I, n = 6), 10 mg/kg/day (Group II,
n = 6), and 20 mg/kg day (Group III, n = 6) for a
period of 12 weeks. The low dose was chosen to approximate the clinical dose, while the medium and
higher doses were chosen based on the data of
Kamm.8 The drug was administered by oral intubation, using peanut oil as a vehicle. A control group of
six additional animals received the vehicle only. Prior
to initiation of treatment, the lids of each animal were
evaluated by biomicroscopy.7 All meibomian glands
appeared normal.
13-cis-retinoic acid (RO 4-3780) was a gift from
Hoffmann-La Roche, Inc. It was supplied as a pure
dry powder that was stored at —20°C. Prior to its use,
the 13-cis-retinoic acid was brought to room temperature and dissolved in a known quantity of the vehicle.
At the end of the 12-week treatment period, the
animals were killed and the eyelids were removed and
placed in fixative for 24 hr. The tissue was then dissected into rectangular blocks containing two or three
glands per block and processed for histologic examination. Only glands taken from the central portion of
the lipid were sectioned for morphometric analysis.
This was done to minimize the error introduced by
regional differences in the size of the rabbit meibomian glands from temporal to nasal.
At least three glands per animal were serially sectioned (10 fim section thickness, 150-200 sections/
From the Department of Ophthalmology, University of Southern California School of Medicine, and Estelle Doheny Eye Institute, Los Angeles, California.
Supported in part by National Institutes of Health Core Center
grant EY-03040 (Bethesda, Maryland) and by a grant from Research to Prevent Blindness, Inc.
Submitted for publication: January 13, 1988; accepted April 7,
1988.
Reprint requests: Ronald E. Smith, MD, 1355 San Pablo Street,
Los Angeles, CA 90033.
1559
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933367/ on 06/18/2017
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / October 1988
1560
Vol. 29
B
Fig. 1. (A) Normal rabbit eyelid. The central duct (*) extends the length of the gland and connects to acini by means of short ductules
(arrowhead). C - conjunctiva. Original magnification X50, hematoxylin and eosin. (B) Tracing of an outline of the meibomian gland
illustrated above.
gland); every tenth section was stained with hematoxylin and eosin, assessed by light microscopy and
subjected to morphometric analysis.
This study conformed to the ARVO Resolution on
the Use of Animals in Research.
Morphometric Analysis
The Micro-Plan II Digitizer (Donsanto Corp.) was
used to determine cross-sectional area. The section
was projected onto a digitizing tablet and an outline
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933367/ on 06/18/2017
ISOTRETINOIN COMPLICATIONS / Lamberr and Smith
No. 10
1561
Fig. 2. Acini from normal
rabbit meibomian gland
containing cells at various
stages of holocrine differentiation. A single layer of
nondifferentiated basaloid
cells is present at the periphery of acini (arrowhead). Original magnification X250, hematoxylin and
eosin.
of the entire gland was traced with a cursor (Fig. 1).
The following parameters were determined:
1. Cross-sectional area (Ax) in mm2 for every tenth
section of a serially sectioned gland.
2. Mean cross-sectional area (A) as the mean of all
cross sectional areas measured for a particular
group.
3. Meibomian gland volume (Vx) in mm3 for each
serially sectioned gland, according to the Delesse principle9:
Vx = SAXT
2AX = sum of all cross-sectional areas for gland
T = thickness of tissue section represented by
a single slide, ie, T = 0.1 mm as every
tenth section was sectioned.
4. Mean meibomian gland volume (V) as the
mean of all gland volumes calculated for a particular group.
Results
At the time of sacrifice, all animals treated with
13-cis-retinoic acid showed signs of systemic toxicity.
These included weight loss, alopecia of the chest, abdomen and distal extremities, dry skin and mild erythema with crusting on the eyelid margins. Animals
in the control group showed no signs of toxicity. His-
topathologic examination of the control tissue revealed no structural abnormalities. The "normal"
rabbit meibomian gland is a simple branched acinar
gland that extends the approximate length of the tarsal plate (Fig. 1). A central duct courses throughout
the length of the gland and connects with individual
acini by short ductules. Acinar cells can be identified
at various stages of holocrine differentiation (Fig. 2).
Animals treated with 13-cis-retinoic acid showed a
marked reduction in the size of the meibomian
glands (Fig. 3); this was confirmed by morphometric
analysis. The mean cross-sectional surface area of the
meibomian gland was reduced by 15% in Group I, by
21% in Group II and by 26% in Group III. This was
reflected in a similar reduction of the meibomian
gland volume; ie, mean volumes were reduced by
18% in Group I, by 25% in Group II and by 28% in
Group III. Results are presented in Table 1.
Meibomian acinar tissue appeared to be diminished and histopathologic changes were observed in
all experimental animals treated with isotretinoin.
These changes included thickening of the epithelium
lining the ducts and ductules, decrease in the number
and size of acini and a reduction in the frequency of
lipid-laden acinar cells (Fig. 3). Meibomian glands of
animals treated with a high dose of isotretinoin
(Group III) showed evidence of a periacinar fibroblastic activity replacing acinar tissue (Fig. 4). Casts of
acinar cells were frequently observed in the acinar
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933367/ on 06/18/2017
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Ocrober 1988
1562
Vol. 2 9
Fig. 3. Treated meibomian gland (Group II, 10 mg isotretinoin daily X12 weeks). Epithelium lining the duct and acini is thickened
(arrowhead) and acini are reduced in size. Original magnification X65, hematoxylin and eosin.
lumen and in the central duct of meibomian glands
from this group of animals (Fig. 5).
Discussion
Treatment of adult New Zealand albino rabbits
with 13-cis-retinoic acid (1-20 mg/kg daily for 12
weeks) resulted in clinical signs of toxicity in all animals. These were present in animals receiving the
Table 1. Rabbit meibomian gland cross-section
areas (A) in mm2 and volume (V) in mm3 following
oral administration of 13-cis-retinoic acid
A (mm2)
Control
Group I
13-cis-retinoic acid
1-2 mg/kg/bw
Group II
13-cis-retinoic acid
10 mg/kg/bw
Group III
13-cis-retinoic acid
20 mg/kg/bw
Data are mean ± SD, n = 6,
* P < 0.005 (paired t-test).
t P < 0.01 (paired t-test).
V (mm3)
1.85 ±0.07
3.93 ± 0.25
1.55 ±0.04t
3.23 ± 0.4f
1.50 ±0.07*
2.97 + 0.16*
1.42 ±0.16*
2.87 ±0.15*
clinical dose (Group I) and progressed in severity to
Group III, where all animals showed clinical signs
similar to a generalized hypervitaminosis A. The histopathology and statistical analysis of gland volume
showed clearly that systemic 13-cis-retinoic acid affects the morphology and reduces the size of the meibomian gland at each dose studied. Isotretinoin appears to inhibit the ability of the meibomian acinar
cell to differentiate, while stimulating the epithelium
lining the ducts and acini to proliferate.
Histopathologic changes observed in our animal
model are compatible with those of Landthaler and
associates,l0 who demonstrated an inhibitory effect of
13-cis-retinoic acid on human skin sebaceous glands.
In their model, 13-cis-retinoic acid appeared to exert
its action by inhibiting the differentiation of sebaceous gland acinar cells. This results in a diminution
of skin sebaceous gland size and, presumably, decreases the production of sebum. In the sebaceous
gland, a reduction in sebum production is believed to
be the therapeutic basis for the treatment of acne
vulgaris.1' In our animal model, the consequence of a
reduction in gland tissue may be mirrored in the adverse ocular side effects associated with isotretinoin
therapy, the most common being blepharoconjunctivitis, which has been reported in 20 to 50% of pa-
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933367/ on 06/18/2017
No. 10
ISOmETINOIN COMPLICATIONS / Lamberr ond Smirh
1563
Fig. 4. Treated meibomian acini {20 mg isotretinoin daily X12 weeks). A
periacinar fibrosis appears
to be replacing actnar tissue
(arrowhead). Original magnification X35O, toluidine
blue.
tients.12 Several hypotheses have been suggested to
explain retinoic-induced blepharoconjunctivitis. According to Fraunfelder, La Braico and Meyer,2 this
type of blepharitis arises from abnormal meibomian
gland function. They suggest that treatment with 13cis-retinoic acid results in a decrease in the meibo-
Fig. 5. Treated meibomian gland (20 mg isotretinoin daily X12 weeks).
Casts of intact differentiated
holocrine cells are seen in
the lumen of the gland
(arrow). Original magnification X425, toluidine blue.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933367/ on 06/18/2017
1564
INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / October 1988
mian lipid component of the tear film which decreases tear break-up time and consequently destabilizes the tear film, leading to a "dry eye" syndrome.
Our present study does not address the meibomian
lipid component of the tear film; however, it seems
likely that the histopathologic changes observed in
retinoid treated rabbit meibomian glands might influence this factor. Additional studies would be required to determine if our model provides a basis for
the hypothesis that changes in meibomian gland morphology result in decreased meibum excretion and a
subsequent reduction in the lipid component of the
tear film. Alternatively, Rismondo and Ubels3 have
detected the presence of 13-cis-retinoic acid in the
tear fluid of patients undergoing retinoid therapy for
acne. They suggested that a lipolytic action of the
retinoid may further destabilize the tear film, compounding the effects described by Fraunfelder and
associates.2
The experimental evidence we have presented indicates that 13-cis-retinoic acid affects meibomian
gland structure in a laboratory model. This is not
only of potential importance in understanding the
pathogenesis of retinoid-induced blepharoconjunctivitis, but also in contributing to our understanding of
the relationship between the meibomian gland and
the complex question of lid margin disease. Future
functional studies in our animal model may yield
important insights into the relationship between meibomian gland structure, function, and the ocular surface. Similar studies have not yet been performed to
confirm our findings in man.
Key words: blepharoconjunctivitis, meibomian gland dysfunction, morphometry, toxicity, hypervitaminosis A
Vol. 29
References
1. Milson J, Jones DH, King K, and Cunliffe WJ: Ophthalmological effects of 13-cis-retinoic acid therapy for acne vulgaris. Br
JDermatol 107:491, 1982.
2. Fraunfelder FT, LaBraico JM, and Meyer SM: Adverse ocular
reactions possibly associated with isotretinoin. Am J Ophthalmol 100:534, 1985.
3. Rismondo V and Ubels JL: Isotretinoin in lacrimal gland fluid
and tears. Arch Ophthalmol 105:416, 1987.
4. Blackman HJ, Peck GL, Olsen TG, and Bergsma DR: Blepharoconjunctivitis: A side effect of 13-cis-retinoic acid therapy for
dermatologic diseases. Ophthalmology 86:753, 1979.
5. Plewig G, Wagner A, Nikolowski J, and Landthaler M: Effects
of two retinoids in animal experiments and after clinical application in acne patients: 13-cis-retinoic Ro 4-3780 and aromatic
retinoid Ro 10-9359. In Retinoids: Advances in Basic Research and Therapy, Orfanos CE, Braun-Falco O, Farber EM,
Grupper C, Polano MK, Schuppli R, editors. New York,
Springer-Verlag, 1981, pp. 219-235.
6. Landthaler M, Kummermehr J, Wagner A, and Plewig G:
Inhibitory effects of 13-cis-retinoic acid on human sebaceous
glands. Arch Dermatol Res 269:297, 1980.
7. Jester JV, Rife L, Nii D, Luttrull JK, Wilson L, and Smith RE:
In vivo biomicroscopy and photography of meibomian glands
in a rabbit model of meibomian gland dysfunction. Invest
Ophthalmol Vis Sci 22:660, 1982.
8. Kamm J: Toxicology, carcinogenicity, and teratogenicity of
some orally administered retinoids. J Am Acad Dermatol
6:652, 1982.
9. Weissmann A, Bowden J, Frank B, Horwitz SN, and Frost P:
Morphometric studies of the hamster flank organ: An improved model to evaluate pharmacologic effects on sebaceous
glands. J Invest Dermatol 82:522, 1984.
10. Landthaler M, Kummermehr J, Wagner A, and Plewig G:
13-cis-retinoic acid inhibits sebaceous glands in animals and
humans: Planimetry and in-vitro-autoradiography. J Invest
Dermatol 74:453, 1980.
11. Strauss JS, Stranieri AM, Farrell LN, and Downing DT: The
effect of marked inhibition of sebum production with 13-cisretinoic acid on skin surface lipid composition. J Invest Dermatol 74:66-67, 1980.
Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933367/ on 06/18/2017