Full text of DOI:10.1007/s004170100341

Graefe’s Arch Clin Exp Ophthalmol
(2001) 239:698–704
L A B O R AT O RY I N V E S T I G AT I O N
DOI 10.1007/s004170100341
Marc Veckeneer
Koen van Overdam
Jan Monzer
Ocular toxicity study of trypan blue injected
into the vitreous cavity of rabbit eyes
Karin Kobuch
Wilfried van Marle
Henk Spekreijse
Jan van Meurs
Abstract Background: To evaluate
the ocular toxicity of trypan blue
photoreceptors and marked disorga-
nization. Immunohistochemical
Received: 13 March 2001
Accepted: 27 June 2001
Published online: 22 August 2001
© Springer-Verlag 2001
(TB) injected into the vitreous cavity staining for rhodopsin was strongly
of rabbit eyes. TB is a dye that could reduced in those sections and stain-
be useful for staining epiretinal
membranes during vitrectomy sur-
gery. Methods: Ten New Zealand
White (NZW) rabbits underwent
ing for proliferation with Ki-67 was
positive. No histological abnormali-
ties were found in the upper retina of
the 0.2% TB-treated eyes or in any
gas-compression vitrectomy. Rabbits part of the retina of the 0.06% TB-
were divided into three groups to re- treated or control eyes. No histologi-
ceive injections of 0.1 ml basic salt
cal abnormalities were found in any
M. Veckeneer ( ) · K. van Overdam
✉
solution, 0.1 ml of a 0.06% TB solu- of the anterior chamber angle speci-
tion or 0.1 ml of a 0.2% TB solution. mens. Conclusions: Although no
Ocular toxicity was assessed by slit- signs of toxicity were found after the
lamp biomicroscopy, ophthalmosco- prolonged presence of TB at a con-
py, electroretinography and histolo-
gy. Results: Transient posterior cap-
sule opacification was noted in all
W. van Marle · J. van Meurs
Rotterdam Eye Hospital,
Schiedamse vest 180,
3011 BH Rotterdam, The Netherlands
e-mail: veckeneer@oogziekenhuis.nl
Fax: +31-10-4017690
centration of 0.06% in the vitreous
cavity of rabbit eyes, marked dam-
age occurred in the lower retina of
J. Monzer · K. Kobuch
Eye department,
University Hospital Regensburg,
Regensburg, Germany
animals. No significant reductions in 0.2% TB-treated eyes. The short-
a-wave or b-wave amplitudes were
found in any of the animals. Light
and electron microscopic examina-
tion of the inferior retina in the 0.2% higher concentration of TB could be
TB-treated eyes showed damaged unsafe.
term presence of TB at a concentra-
tion of 0.06% in the vitreous cavity
is harmless to the rabbit eye but a
H. Spekreijse
Department of Medical Physics,
Academisch Medisch Centrum,
Amsterdam, The Netherlands
tion may significantly simplify the procedure. Conse-
quently, less injury to the retina and fewer recurrences as
a result of more complete membrane removal could im-
prove clinical outcome.
Trypan blue (TB) is a well-known vital stain. Norn
[8] has reported the peroperative use of this dye (at a
concentration of 0.1%) in anterior segment surgery for
the vital staining of corneal endothelium. No ocular
complications were found during 8 years follow-up.
More recently, Melles et al. [6] have advocated the use
of TB (0.1 ml of a 0.06% solution) for the visualization
of the anterior capsule in cataract surgery. In a prelimi-
nary in vitro study (unpublished data) we evaluated the
Introduction
Peeling of epiretinal membranes (ERMs) is an essential
part of surgery for proliferative vitreoretinopathy (PVR)
and macular pucker [5, 7]. Precise identification and de-
lineation of the membrane is crucial for its successful re-
moval. Although some ERMs can be easily distinguished
from the underlying retina because of opacification due
to fibrosis or because of pigment deposition, others are
harder to identify, making their atraumatic surgical re-
moval more demanding. Notwithstanding important pro-
gress in surgical techniques and instrumentation, perop-
erative staining of membranes that improves visualiza-

699
potential of TB for the staining of ERMs. The dye was
applied immediately after the surgical removal of the
ERM or retinotomy specimen. TB stained the ERM in a
concentration-dependent manner (Fig. 1).
However, the toxic potential of TB is well known [1,
2]. Thorough evaluation of possible retinal toxicity was
therefore essential before further application in vitrecto-
my surgery would be acceptable. We studied the ocular
toxic potential of this dye in a rabbit animal model.
Materials and methods
Preparation of solutions
The TB solution used for this study was prepared by the hospital
pharmacist. It contained 0.6 mg/ml TB (for the 0.06% solution) or
2 mg/ml TB (for the 0.2% solution). Other constituents were (per
ml): 1.9 mg sodium monohydrogen orthophosphate, 0.3 mg sodi-
um dihydrogen orthophosphate, 6.2 mg sodium chloride and water
(resulting pH 7.4).
Injection
Ten New Zealand White (NZW) rabbits weighing 2–2.5 kg were
used in this study. The “Principles of laboratory animal care”
(NIH publication no. 85-23, revised 1985), the OPRR Public
Health Service Policy on the Human Care and use of Laboratory
Animals (revised 1986) and the US Animal Welfare Act,
as amended, were followed. The research was carried out in ac-
cordance with the Dutch National Laws on animal experiments
(Wet op dierproeven 1977, revised 1995; Dierproevenbesluit
1985)
The rabbits were divided into three groups to receive injec-
tions of 0.1 ml basic salt solution (BSS) (control; two animals),
0.1 ml 0.06% TB solution (four animals) or 0.1 ml 0.2% TB
solution (four animals). After baseline ocular examination and
electroretinography, the rabbits underwent gas-compression
vitrectomy. They were anaesthetized with a mixture of ketamine
HCl (50 mg/kg) and xylazine HCl (5 mg/kg). Proparacaine
0.5% was used for topical anaesthesia. The pupils were dilated
with 2.5% phenylephrine and 1% tropicamide. After ocular mas-
sage and paracentesis, 0.4 ml 100% C3F8 was injected into the
mid-vitreous. The injection was performed with a 30-gauge nee-
dle approximately 2–3 mm behind the limbus. At the time of
maximum gas expansion (3 days of gas compression) the eyes
were injected with 0.1 ml BSS, 0.1 ml 0.06% TB or 0.1 ml 0.2%
TB. No gas was removed artificially at any time during the
study.
Clinical examination
The cornea, lens, vitreous cavity and retina were examined with a
slit-lamp and indirect ophthalmoscope every other day for
4 weeks. The degree of lens opacification, resolution of the gas
bubble and fading of the blue color were recorded.
Fig. 1 Trypan blue induces a concentration-dependent staining of
the ERM in vitro

700
Electrophysiology
Electroretinograms (ERG) were obtained before and 4 weeks after
injection of BSS or TB. The xenon light stimulus was generated
by a PS-33 plus Grass photic stimulator (Grass Instruments,
Quincy, Mass.). The light passed through a diffuser at the corneal
surface in order to achieve full-field stimulation. Responses were
filtered to pass frequencies from 0.3 Hz to 1 kHz and were digi-
tized with a resolution of 1 mV (referred to the amplifier’s signal
input).
The rabbits were dark-adapted for 20 min. They were anaes-
thetized and the pupils were dilated as described above. Positive
electrodes (contact lens) were placed on the cornea with methyl-
cellulose, the negative-needle electrode was placed subcutaneous-
ly in the middle of the forehead and the ground electrode was
clipped to the earlobe with electric gel. The fellow eye was
patched to avoid any stimulation. Scotopic ERG responses were
recorded for six consecutive light levels ranging from 6.3×10–3 to
200 cd·s.m–2. Light attenuation was achieved with 0.9 log-unit
neutral-density filters (starting with five and decreasing to zero).
The average of three sweeps was taken for each step. A decrease
in amplitude of greater than 30% was considered significant.
Fig. 2 Photograph of the lens and ciliary body of a rabbit eye that
had been injected with 0.06% TB 4 weeks earlier. Note the rem-
nants of blue dye on the posterior capsule and ciliary body
Histology
The rabbits were killed 4 weeks after injection by an intracardiac
injection of pentobarbital while under anaesthesia. The eyes were
enucleated and the anterior segment was separated from the poste-
rior segment by an incision through the pars plana. Both segments
were immediately fixed in 2% paraformaldehyde and 3% glutaral-
dehyde in 0.1 M phosphate buffer. Anterior chamber angle, upper
retina and lower retina specimens were taken from each eye and
prepared for light microscopy (toluidine blue staining), transmis-
sion electron microscopy and immunohistochemical staining for
rhodopsin and Ki-67.
ERG recordings (recordings after 4 weeks compared
with preinjection recordings; Fig. 3).
Histology
Light and transmission electron microscopy
Compared to controls, light microscopic examination
demonstrated a normal, well-preserved retinal structure
in the 0.06% TB-treated group (Fig. 4A, B) as well as in
the superior retina of the 0.2% TB-treated group
(Fig. 4D). In contrast, striking alterations were found in
Results
Clinical examination
Some degree of posterior capsule opacification was seen certain areas of the lower retina of the eyes injected with
in all eyes but this cleared with time as the gas bubble 0.2% TB (Fig. 4E). The retinal architecture looked disin-
resolved. Blue discoloration of the anterior segment of tegrated, and the retinal layers were destroyed. Photore-
TB-treated eyes was striking during the first days after ceptors were reduced and outer segments had almost dis-
injection and was clearly more pronounced in the 0.2% appeared, as shown in detail in the electron micrograph
TB-treated eyes. Some blue staining, seen on slit-lamp shown in Fig. 4F. A normal morphology of the anterior
examination, seemed to occur at the iris stroma and the segment was preserved in all groups (Fig. 4I)
lens capsule. Although the color faded with time, some
degree of staining of the posterior lens capsule, the infe-
rior parts of the ciliary body, the pars plana region and Immunohistochemistry
the neighbouring retina of the eyes treated with TB was
still present at the time of enucleation 4 weeks after in- Immunohistochemical staining for rhodopsin confirmed a
jection (Fig. 2). Other than the capsule opacification and marked, and in some areas almost complete, loss of rho-
blue discoloration, the eyes remained quiet and the cor- dopsin in the lower retina of the 0.2% TB-treated group
neas clear throughout the experiment.
corresponding with the morphological loss of photorecep-
tor outer segments (Fig. 4G). This was in sharp contrast
with the images of the control and 0.06% TB-treated eyes
(Fig. 4C). Staining for proliferation with Ki-67 was nega-
tive in the retinas of the control and 0.06% TB-treated
Electrophysiology
No consistent and significant (>30%) reduction in the groups, whereas it was positive in the lower retinal speci-
amplitude of the a-wave or b-wave was found in any of mens of the 0.2% TB-treated group (Fig. 4H).

701
Fig. 3 ERG recordings from
each study group. No consis-
tently significant reduction in
amplitude was seen after
4 weeks as compared with pre-
injection recordings
Fig. 4 A Normal light microscopic image of the lower retina of a of a rabbit 4 weeks after intravitreal injection of 0.2% TB. Photo-
control animal. B Light microscopic image of the lower retina of a receptor reduction and outer segment loss can be seen in detail.
rabbit 4 weeks after 0.06% TB injection; no toxic changes are ap- G Immunohistochemical staining for rhodopsin is clearly reduced
parent. C Immunohistochemical staining for rhodopsin is normal in the lower retina of a rabbit 4 weeks after intravitreal injection of
in the lower retina of a rabbit 4 weeks after intravitreal injection of 0.2% TB. This corresponds with the morphological findings of
0.06% TB. D Light microscopic image of the upper retina of a photoreceptor outer segment loss. H Immunohistochemical stain-
rabbit 4 weeks after 0.2% TB injection; no toxic changes are ap- ing for Ki-67 indicates the presence of proliferative activity in the
parent. E Light microscopic image of the lower retina of a rabbit inferior retina of a rabbit 4 weeks after injection of 0.2% TB.
4 weeks after 0.2% TB injection; striking abnormalities can be I Light microscopic image of the anterior chamber angle of a rab-
seen. Note the reduction in photoreceptors and the disorganization bit 4 weeks after injection of 0.2% TB; the morphology is well
of the retinal layers. F Electron micrograph of the inferior retina preserved

702
Fig. 4 Legend see page 701

703
and pathological preparation 4 weeks after dye injection
(Fig. 2). This implies that certain intraocular tissues re-
mained in contact with the dye for the full length of the
experiment. The staining was faint and unequally distrib-
uted with more intense blue areas in the six o’clock re-
gion of the fundus periphery of the 0.2% TB-treated
eyes.
Discussion
A safe and effective method for enhancing the visualiza-
tion of ERMs would greatly facilitate their surgical re-
moval. This could lead to an improvement in the ana-
tomical and functional outcome of macular pucker and
PVR surgery. Several dyes have been advocated for use
in posterior segment surgery. Sodium fluorescein has
been used to stain the vitreous gel to facilitate complete
vitrectomy [9]. Kadonosono et al. [3] have recently de-
scribed the staining of the internal limiting membrane in
macular hole surgery with indocyanine green.
In a previous study we examined the potential of TB
for the staining of ERMs in vitro (unpublished data).
These ERMs were harvested at the time of vitrectomy
for PVR (grade C3 or higher) and macular pucker and
immediately stained with concentrations of TB ranging
form 0.06% to 1.0%. The staining appeared promptly
(within seconds) and consistently, and the degree of
staining was concentration-dependent (Fig. 1). TB has
been used as a surgical aid in the anterior segment with
good results and without any signs of toxicity [6, 8]. In
cataract surgery, the dye is injected under air for about
1 min to stain the anterior lens capsule. Considering the
in vitro staining of the ERMs, in vitrectomy surgery, TB
could also be injected under air and removed after only a
few minutes. This would prevent uncontrolled spreading
of the dye in the intraocular fluid.
The toxicity of TB is well documented. Several ani-
mal studies have indicated its teratogenic and carcino-
genic potential [1, 2]. In these experiments TB was ad-
ministered via intraperitoneal injection in dosages vary-
ing between 100 mg/kg body weight (teratogenic in rats)
and 300 mg/kg body weight (carcinogenic in rats). In
comparison, in cataract surgery in humans, TB is inject-
ed locally in the anterior chamber at a dose of about
0.005 mg/kg body weight for a short time after which it
is rinsed out. Nevertheless, a thorough investigation into
the posterior segment safety was indispensable before
further clinical use could be advocated. In cataract sur-
gery, it is known that despite active rinsing after applica-
tion, some dye will usually remain behind. When inject-
ed into the posterior segment, significant amounts of dye
can then remain in the eye after surgery. We therefore
chose to evaluate the ocular toxicity of TB injected into
the vitreous cavity using a spontaneous clearance experi-
mental model.
Clinical examination showed posterior capsule opaci-
fication in all animals, which cleared as the gas bubble
resolved. It could therefore be interpreted as a form of
gas-induced cataract. Blue discoloration in the anterior
segment of the TB-treated eyes was striking during the
first days after injection of the dye and faded slowly in
the following weeks. The spontaneous clearance of dye
from the vitreous cavity was remarkably slow and some
degree of staining persisted up to the time of enucleation
Histology revealed interesting differences between
the different dosage groups and the part of the retina ex-
amined. No abnormalities were found in any part of the
retina of the 0.06% TB-treated or control animals
(Fig. 4A, B). On the other hand, marked changes were
found on light microscopic examination of the inferior
retina of the 0.2% TB-treated animals (Fig. 4E). The ret-
inal organization into separate layers was lost. Most
prominent abnormalities occurred in the photoreceptors
where the loss of outer segments was striking. This was
confirmed by a severe reduction in the immunohisto-
chemical staining for rhodopsin in these retinal segments
(Fig. 4G). The positive staining for Ki-67 [4] in these
specimens indicates proliferative activity. This could in-
dicate a reactive proliferation in Muller cells as a re-
sponse to photoreceptor damage (Fig. 4H). Gravity and
the effect of gas compression causing a sequestration of
blue dye in the inferior part of the vitreous cavity seem
to explain why the inferior retina of the 0.2% TB-treated
group was selectively damaged.
Most of the data concerning potential TB toxicity in
the anterior segment concerns the corneal endothelium.
So far no toxicity towards the endothelium has been not-
ed [6, 8]. Very little is known about the interaction with
other anterior segment tissues. Our histological findings
indicate a preservation of normal morphology of the an-
terior chamber angle (corneal endothelium and trabecu-
lar meshwork in particular) in all animals despite the fact
that the dye was not actively removed in this study at
any time (Fig. 4I). The ERG responses did not deterio-
rate significantly in any of the animals (Fig. 3). Flash
ERG generates a mass response. Large areas of photore-
ceptor loss have to occur before the ERG amplitudes de-
crease significantly. Our histological findings demon-
strate that no generalized photoreceptor damage had oc-
curred. The results of this study nevertheless indicate
that the long-term intravitreal presence of 0.2% TB
could cause considerable damage to the retina.
The clinical use of TB as an aid in vitreoretinal sur-
gery could involve a short application time of about
1 min. The dye should be injected and removed under air
preventing generalized spread throughout the posterior
segment. In view of the fact that our study with a long
application time did not show signs of toxicity in the
0.06% TB-treated eyes, clinical studies to clarify the po-
tential of a low concentration TB solution for the stain-
ing of ERMs in posterior segment surgery seem justified.

704
References
1. Chung KT (1983) The significance of
azo-reduction in the mutagenesis and
carcinogenesis of azo dyes. Mutat Res
114:269–281
2. Ema M, Kanoh S (1982) Studies on the
pharmacological bases of fetal toxicity
of drugs (II). Effect of trypan blue on
the pregnant rats and their offspring.
Nippon Yakurigaku Zasshi 79:369–381
3. Kadonosono K, Itoh N, Uchio E,
Nakamura S, Ohno S (2000) Staining of
the internal limiting membrane in macu-
lar hole surgery. Arch Ophthalmol
118:1116–1118
4. Kreitz S, Fackelmayer FO, Gerdes J,
Knippers R (2000) The proliferation-
specific human Ki-67 protein is a con-
stituent of compact chromatin. Exp Cell
Res 261:284–292
5. Machemer R (1978) The surgical re-
moval of epiretinal macular membranes.
Klin Monatsbl Augenheilkd 173:36–42
6. Melles GR, de Waard PW, Pameyer JH,
Beekhuis HW (1999) Trypan blue cap-
sule staining to visualize the capsulo-
rhexis in cataract surgery. J Cataract Re-
fract Surg 25:7–9
7. Michels RG (1984) Surgery of retinal
detachment with proliferative vitreoreti-
nopathy. Retina 4:63–83
8. Norn MS (1980) Per operative trypan
blue vital staining of corneal endotheli-
um. Eight years’ follow up. Acta
Ophthalmol 58:550–555
9. Tamai M, Hara S, Mizuno K (1984) Pre-
injection with sodium fluorescein aids
discrimination in vitreous surgery. Oph-
thalmic Surg 15:761–764