Full text of DOI:10.1007/s00417-002-0561-0

Graefe’s Arch Clin Exp Ophthalmol
(2002) 240:989–995
L A B O R AT O RY I N V E S T I G AT I O N
DOI 10.1007/s00417-002-0561-0
Sylvia Mertens
Jürgen Bednarz
Katrin Engelmann
Evidence of toxic side effects
of perfluorohexyloctane after vitreoretinal
surgery as well as in previously established
in vitro models with ocular cell types
Abstract Background: Cases of
ocular irritation have been observed
after early clinical trials using per-
fluorohexyloctane (F6H8) as endo-
tamponade. In our clinic two of three ly lower extinctions for vital cells as
eyes developed severe inflammatory- well as a non-significant decrease in
like reactions after intermediate-term proliferation compared with con-
endothelial cell morphology was
documented. Results: After 5 days
incubation with F6H8, cultures of
RPE and HCEC showed significant-
Received: 18 February 2002
Revised: 19 July 2002
Accepted: 20 August 2002
Published online: 21 September 2002
© Springer-Verlag 2002
tamponade. These cases will be
depicted, serving as background for
the experimental study. To elucidate
possible toxic effects of F6H8 on
trols. Analysis by means of fluores-
cence microscopy after treatment
with F6H8 or perfluorodecaline re-
vealed decreased cell densities
different ocular cell types and corne- (F6H8 > perfluorodecaline) within
al tissue we applied our previously
established in vitro models to inves-
tigate effects of F6H8 on cultured
ocular cells in comparison with
contact areas. The endothelium of do-
nor corneas incubated in presence of
F6H8 developed circumscribed ne-
crotic areas. Conclusions: Decreased
perfluorodecaline. Methods: Vitality amounts of vital cells cannot be ex-
and proliferation of cultured
human corneal endothelial cells
(HCEC) and human retinal pigment
plained solely by mechanical effects
or nutritional deficit due to direct
contact, since F6H8 has a lower spe-
Presented at the 99th meeting of the
Deutsche Ophthalmologische Gesellschaft,
Berlin, Germany, 29 September 2001
epithelial cells (RPE) were measured cific weight than perfluorodecaline.
after incubation with F6H8 or per-
fluorodecaline for up to 5 days.
Vitality was evaluated using the
Live/Dead assay, and proliferation
was determined according to BrdU
incorporation. Additionally the en-
dothelium of donor corneas was
incubated with F6H8 for 5 days and
The ability of the remaining cells to
proliferate revealed that they were not
irreversibly damaged. Due to the high
lipophilicity of F6H8 interactions
with cellular lipoprotein membranes
as well as other toxic effects have to
be considered and should further be
investigated prior to clinical use.
S. Mertens ( ) · J. Bednarz · K. Engelmann
✉
Klinik und Poliklinik für Augenheilkunde
des Universitätsklinikums
Hamburg-Eppendorf, Martinistrasse 52,
20246 Hamburg, Germany
e-mail: smertens@uke.uni-hamburg.de
Tel.: +49-40-428032334
Fax: +49-40-428034481
uted to their relatively high specific gravity [18, 22, 23,
24]. Partially fluorinated alkanes present physical prop-
Introduction
Liquid perfluorocarbons are established intraoperative erties comparable to liquid perfluorocarbons [11, 16, 17].
tools for manipulation in difficult vitreoretinal surgery Perfluorohexyloctane (C14F13H17; F6H8) is a semifluor-
[5, 20]. Due to their high specific gravity (1.7–2.1 g/cm3) inated alkane with lipophilic and hydrophilic properties
hydrokinetic stabilisation of the inferior retina can be but a density of only 1.35 g/cm3. Furthermore, it is a bio-
achieved. Retinal damage by mechanical effects after compatible solvent for silicon oil [13]. Because of its
prolonged contact with liquid perfluorocarbons is attrib- lower density and its biocompatibility, F6H8 has recently

990
Live/Dead assay
been used as an intraoperative tool and intermediate-
term vitreous tamponade in difficult surgical procedures
[25]. Zeana et al. found good intraocular intermediate-
term tolerance in a rabbit model for up to 9 weeks. How-
ever, in some animals use of F6H8 led to ocular irritation
with whitish precipitates and widespread proliferative
vitreoretinopathy (PVR) [25]. Hiscott et al. [8] excised
membranes after long-term use of F6H8 and detected
histological and immunohistochemical features typical
for PVR membranes with macrophages and giant cells.
Kobuch et al. [12] observed vascular constriction, slight
disarrangement of the plexiform layers, hypertrophy of
Müller cells and vacuolisation of the inner retinal surface
in rabbit eyes after 6 weeks’ tamponade.
Vitality of cells was determined using the Live/Dead Viability/
Cytotoxicity kit (Molecular Probes) containing the two compo-
nents calcein and ethidium homodimer. Calcein penetrates the cell
membrane, and esterases in the cytoplasm of vital cells render
it fluorescent. Ethidium homodimer penetrates not intact but
damaged cell membranes and then adheres to nucleic acids of
damaged or dead cells, leading to red fluorescent nuclei.
Cell cultures of human RPE and HCEC were seeded on micro-
titer plates and cultivated until cell layers were semiconfluent.
Next, cultures were incubated with an additional 50 µl perfluoro-
decaline or F6H8 per hole for 1, 3 and 5 days. Controls did not
receive these additives. Dying procedure, analysis and documenta-
tion were as described before [18].
Proliferation assay
In our clinic, F6H8 was used as an intermediate-term
vitreous tamponade in three patients who suffered from Cultures of RPE and HCEC were treated with F6H8 or perfluoro-
decaline as described for the Live/Dead assay. After the incubation
complicated retinal detachment. Two of these patients
developed inflammatory-like complications 8–14 weeks
after surgery.
period, F6H8 or perfluorodecaline was removed by washing the
cultures twice with PBS. Subsequently, the cells were cultured for
a further 20–24 h in the presence of 10 mM bromodeoxyuridine
(BrdU). Incorporation of BrdU was determined by means of
monoclonal antibodies as described by Mertens et al. [18].
Based on this early clinical experience, in vitro tests
were performed to determine possible toxic side effects
of this new drug applying the Live/Dead assay and a pro-
liferation assay on human retinal pigment epithelium
(RPE) and human corneal endothelial cells (HCEC). To
assess the effect also in an ocular tissue model in vitro,
two pairs of human donor corneas were incubated with
F6H8. These tests had previously been established in our
lab to elucidate possible toxic effects of perfluorodeca-
line [18]. Moreover, ocular cell cultures have been
shown to be useful in quality testing of media and sera
for organ culture [7].
Organ culture
Two pairs of human corneas not suitable for transplantation due to
low endothelial cell density and/ or hepatitis C infection were
stored in 50-ml tissue culture flasks. Donor ages were 73 and
80 years, post mortem times 58 and 6 h. Organ culture medium
was minimal essential medium supplemented with 2% FCS [2]. In
order to determine the effect of F6H8 on corneal endothelium,
corneas were preserved in 24-well plates, endothelial side up, in
minimal essential medium supplemented with 2% FCS and 6%
dextran T500. In each pair, 100 µl F6H8 was applied to the endo-
thelium of one cornea. The fellow corneas were not treated and
served as controls. After 1, 3 and 5 days endothelial cell densities
were determined using a phase-contrast microscope (Labovert FS,
Leitz) and documented photographically [2].
Materials and methods
Statistics
Vitreoretinal surgery using F6H8
Live/Dead assay and proliferation assay were repeated four times
using RPE cells from four different donors and four different sub-
cultures of the HCEC cell line. Controls were always performed
using cells of the same preparation. The extinction values were
normalised to the values obtained with control cultures. The sig-
nificance of the results was determined using Student’s t-test.
A standard three-port vitrectomy, retinal reattachment and endolaser
coagulation was performed in all three patients (patient 1, female,
aged 45 years; patient 2, female, aged 75 years; patient 3, male,
aged 78 years). F6H8 was used as endotamponade and filled the
whole vitreous cavity.
Removal of the tamponade was performed after 6 weeks
(patient 3), 8 weeks (patient 1) and 12 weeks (patient 2).
Results
Cell cultures
Two out of three patients developed symptoms of intra-
ocular irritation with white precipitates in the vitreous
after intermediate-term endotamponade with F6H8.
RPE cells of four different human donors were isolated and culti-
vated using the method described by Sobottka Ventura et al. [21].
RPE cells were not transfected.
Immortalised HCEC [2] transfected with pRNS [14], a plasmid
containing the genes for the SV40 small and large T-antigen, were
cultivated in medium F99. a 1:1 mixture of Ham’s F12 (Invitro-
gen) and M199 (Invitrogen), supplemented with 5% fetal calf
serum (FCS; Seromed). For subcultivation cells were trypsinised
(0.05%/0.02% trypsin/EDTA in phosphate-buffered saline), seeded
onto new culture flasks or microtiter plates, and cultivated in an
atmosphere containing 5% CO2 at 37°C.
Brief case reports
Case 1
Diagnosis. Central retinal detachment due to multiple
retinal holes, myopia magna and posterior staphyloma.

991
Therapy. Standard three-port vitrectomy, F6H8 endo-
tamponade, endolaser treatment.
Follow-up. No intraocular irritation 6 weeks postopera-
tively; F6H8 removal after 8 weeks.
Complications. Development of retrolental and preretinal
grey–white membranes or fibrin-like structures. After
F6H8 removal retinal redetachment in four quadrants.
Following vitrectomy and silicone oil endotamponade
resulted in hypotonia.
Case 2
Diagnosis. Vitreal and subretinal hemorrhage with reti-
nal detachment due to CNV in age related macular de-
generation.
Fig. 1 Anterior segment of the eye of 75-year-old patient E.S.
after F6H8 vitreous tamponade for 14 weeks. F6H8 dispersed and
penetrated into the anterior chamber (arrows)
Therapy. Standard three-port vitrectomy, lavage of
subretinal blood, F6H8 endotamponade, endolaser and
endocryocoagulation.
Follow-up. Follow-up examinations not performed due
to non-compliance of the patient until 13 weeks after
surgery.
Complications. Dispersion and penetration of F6H8 into
the anterior chamber (Fig. 1). Development of mem-
brane-like structures similar to case 1, decompensation
of intraocular pressure (45 mmHg). Retinal detachment
after F6H8 removal.
Case 3
Fig. 2 Effect of perfluorodecaline (PFD) and F6H8 on vitality of
Diagnosis. Amotio non-sanata (inferior quadrants) with human RPE. Prolonged contact with PFD and F6H8 caused a pro-
gressive decrease of the fluorescent signal emitted by vital or dead
risk of macular detachment state post conventional sili-
cone plombage.
cells. Five-day incubation with PFD led to mean extinctions of
56% for vital cells and 59% for dead cells compared with untreat-
ed control cultures. F6H8 incubation led to mean extinctions of
Therapy. Standard three-port vitrectomy, F6H8 endotam-
ponade, endolaser coagulation. Because of previous
complications F6H8 removal was performed after
6 weeks without inflammatory complications.
38% for vital cells and 41% for dead cells after 5 days. Signifi-
cance (Student’s t-test): * P<0.05, ** P<0.01
Human retinal pigment epithelial cells
Follow-up. The retina remained attached.
Unexpected complications occurred during the post- Contact of cultured RPE cells with F6H8 and perfluoro-
operative course in two of three patients. Morphological decaline led to a reduction in the fluorescent signals
changes in both patients occurred in the anterior vitreous emitted by vital and damaged cells respectively. The
segment related to the posterior surface of the lens. effect was more evident with F6H8 than after contact
There are grounds for suspicion that interaction of the with perfluorodecaline. In detail, the values relative to
vitreous tamponade with the retinal surface, persistent controls for the signals emitted by living cells measured
anterior vitreous cortex or lens material led to toxic side after contact with F6H8 were 53.2%, 38.2% and 38.4%
effects and inflammatory processes. Based on this hy- after 1, 3 and 5 days respectively (Fig. 2). The corre-
pothesis and our clinical observations, in vitro tests were sponding values after contact with perfluorodecaline
performed using non-transformed cells of ocular origin.
were 84.4%, 62.1% and 55.8% (Fig. 2). Thus, the reduc-

992
Fig. 4 Effect of perfluorodecaline (PFD) and F6H8 on cultured
transfected HCEC by means of the Live/Dead assay compared
with controls (100%). One-day contact induced a significant
decrease in extinctions for vital and dead cells. Significantly
reduced values (3–14%) observed after incubation with F6H8
revealed areas of nearly complete cell necrosis. Significance
(Student’s t-test): * P<0.05, ** P<0.01
Fig. 3 Mean proliferative activity after incubation of RPE cultures
with perfluorodecaline (PFD) or F6H8 for up to 5 days. Cell pro-
liferation after contact with the substances was not significantly
different from that in controls
tion in fluorescence was higher after contact with F6H8
than after contact with perfluorodecaline at all time
points. An analogous result was found for fluorescent
signals emitted by damaged cells. After contact with per- with perfluorodecaline, as already seen in cultured RPE
fluorodecaline the emissions measured at 1, 3, and cells. The emission from vital cells was 58.1%, 34.6%
5 days reached 84.7%, 62.2%, and 58.8%, respectively, and 29.1% of the control values at 1 day, 3 days and
of the values in the control cultures (Fig. 2). Cultures in- 5 days, respectively, and that from damaged cells was
cubated with F6H8 exhibited only 55.5%, 40.3% and 58.1%, 35.2% and 33.6%.
41.6% of the fluorescent signal emitted by untreated cell
cultures (Fig. 2).
The decrease in emission characteristic for vital cells
was not correlated with an increase in the fluorescent
Although contact of cell cultures with F6H8 or perflu- signal emitted by damaged cells. Analysis of the stained
orodecaline caused a decrease in fluorescent signals cell cultures by means of fluorescence microscopy re-
characteristic for living as well as for damaged cells, the vealed that after treatment with F6H8 or perfluorodeca-
relation between these two signals remained rather con- line a decrease in cell density occurred within the re-
stant during the observation period.
gions of F6H8 or perfluorodecaline application (Fig. 5).
After removal of the substances from cultures, the re- Therefore the overall decrease in fluorescence corre-
maining proliferative capacity was determined and com- sponded to an overall decrease in cell number.
pared with controls. No significant changes in prolifera-
The cells remaining after incubation in the presence
tive activity after incubation with either substance were of F6H8 and perfluorodecaline were analysed for their
observed. Extinctions of 90.6% (t-test: P=0.019) after proliferation activity as indicated by incorporation of
5-day contact with perfluorodecaline and 91.9% (t-test: BrdU. Contact with perfluorodecaline did not affect
P=0.104) with F6H8 were measured (Fig. 3).
BrdU incorporation into HCEC significantly. After con-
tact with F6H8 a decrease in proliferation activity com-
pared with untreated cells was observed, but the differ-
ence was not statistically significant (P=0.072).
Human corneal endothelial cells
Cultured HCEC responded to contact with F6H8 or per-
fluorodecaline in the same manner as RPE cells but even
more sensitively. As shown in Fig. 4, after only 1 day’s
Organ-cultured corneas
contact with F6H8 the signal emitted by living cells had The endothelium of cultured human corneas did not
decreased to only 6.9% of that in untreated cells. Further show a significant decrease in cell density after incuba-
contact caused an ongoing decrease to 4.3% after 3 days tion with F6H8 compared with controls. Circumscribed
and 3.5% after 5 days. At the same points in time, the areas of cell necrosis developed at the interface between
signal emitted by damaged cells reached only 14.2%, medium and F6H8 after 3 days. Since only two pairs of
9.8% and 7.4%, respectively, of that emitted by the con- human corneas were incubated with F6H8 no statistical
trols (Fig. 4). The effect was less evident after contact evaluation of mean cell density was attempted.

993
Fig. 5a, b Cultured HCEC
after incubation with a F6H8
and b perfluorodecaline for
3 days. Cultures were analysed
using the Live/Dead assay. In
areas which had been in contact
with F6H8 or perfluorodecaline
cell density was dramatically
decreased
In our study we used cultures of human ocular cells
which may come into contact with F6H8 after its use as
Discussion
Retinal complications after F6H8 endotamponade in two endotamponade. Both cell types, primary cultured RPE
of three eyes were unexpected and led to severe decrease [21] and transfected HCEC [1, 3], retained morphological
in visual acuity. Removal of the newly formed membrane- and functional characteristics during culture and have
like structures which had developed between the inner ret- been used successfully in previous test systems to evaluate
inal layer and F6H8 tamponade proved difficult. The ma- toxic effects [2, 7, 18]. Furthermore, we brought these cul-
terial was extremely fragile and could not be peeled off tured cells in direct contact to F6H8. Therefore we believe
like a membrane. Structures were especially situated in that this model reflects the clinical situation after endo-
the anterior part at the vitreous basis and retrolental. De- tamponade more accurately than other in vitro test sys-
spite the absence of signs of inflammation in the anterior tems used before. In cultures of both cell types, contact
segment, findings were comparable to endophthalmitis- with F6H8 led to a decrease in the number of vital cells.
like reactions. Interaction between F6H8 and remnants of Only cells which had been in contact with F6H8 seemed
the vitreous cortex in phacic eyes has to be considered. to be damaged, however. After removal of this substance,
The origin of precipitates and membranes especially at in- the remaining cells of the cultures exhibited normal prolif-
terfaces, e.g. lens capsule, anterior vitreous and between eration activity. Possible mechanical and barrier effects of
retina and endotamponade, remain to be elucidated. Ob- the substances during incubation have to be considered
servation of these clinical cases led us to perform an in since cells were seeded onto the floor of culture plates
vitro trial to analyse possible toxic effects of F6H8.
[18]. The reasons for the significant effect of F6H8 remain
F6H8 is licensed for use in complicated retinal de- unclear, but our results yield some evidence of interaction
tachments especially in the inferior part of the eye. In between F6H8 and cellular structures.
Germany animal studies [25] and standard toxicity tests
Ten years ago low- molecular-weight components
were performed prior to clinical use. Toxicity tests were were shown to be responsible for toxic effects seen after
performed using HeLa, MonoMac 6, Molt 4 and Raji silicone oil application [19]. In the case of F6H8 the ob-
cell cultures [17]. These cells have been under culture served effects cannot be explained by such contamina-
conditions for many decades and have lost some meta- tions since these preparations are stated to be 100% pure.
bolic features of the original cell types. They are mainly
used for toxicity testing to evaluate effects on the house- after F6H8 application than after use of perfluorodeca-
keeping metabolism of cells [15].
line. The hydrogenation of F6H8 causes a polarisation
The effect of cell loss was much more pronounced

994
and renders this substance more lipophilic [10]. Thus [9] described emulsification after combined use of sili-
F6H8 may interfere with the cell membrane, membrane cone oil and perfluorohexylhexane, a semifluorinated
proteins or other hydrophobic proteins such as albumin. alkane, for macular rotation. Besides emulsification and
Interaction with cell membranes may change the adher- opacification, proteins as well as lipids, blood cells,
ence of the cells to their matrix. This would explain the fibrin, albumin and fibrinogen have been found after
occurrence of cell-free areas in cultures treated with endotamponade in a human eye. A breakdown of the
F6H8. In contrast, the endothelium of donor corneas is blood–retina barrier has to be expected in this case.
additionally stabilised by tight junctions between the Kobuch et al. [12] described alteration of vascular struc-
endothelial cells. These tight junctions are not yet pres- tures 6 weeks after injection of a semifluorinated alkane
ent in non-confluent cell cultures. Therefore, cultured (O62) into the rabbit eye. Vasoconstriction may cause
cells may be more susceptible to F6H8-associated dam- disturbance of the blood–retina barrier, resulting in
age. Nevertheless, necrosis developed also in the endo- leakage of protein. This would explain the presence of
thelium of cultured organ corneas at the interphase be- albumin demonstrated in the whitish precipitates seen
tween F6H8 and medium. Thus, F6H8 seemed to exert after the use of hydrofluorocarbon liquids (Spindler, un-
its effects not only by influencing the adherence of the published communication, DOG 2001) as well as the
cells to their matrix. Compared with an in vivo model, occurrence of PVR-like membranes with macrophages
cell cultures do not present vascular structures. There- and giant cells [8]. Why these reactions have been
fore, observation of protein leakage and macrophages observed after different times of incubation remains
described histologically as whitish membranes and PVR unclear.
formation cannot be observed using this in vitro model.
Additional in vitro tissue tests and animal studies
Whitish precipitates observed in patient eyes may be should be considered prior to further clinical application
blood derived or of cellular origin. Hoerauf and Laqua of F6H8 in patients with retinal detachment.
References
1. Aboalchamat B, Engelmann K,
Böhnke M, Bednarz J (1999) Morpho-
logical and functional analysis of SV40
transfected human corneal endothelial
cells after transplantation. Exp Eye Res
69:547–552
8. Hiscott P, Colthurst, Wong D, Lois N
(2001) Epiretinal membranes and long-
term perfluorohexyloctane tamponade.
ARVO Abstract 2305
9. Hoerauf H, Laqua H (2002) Severe
emulsification after combined use of
a partially fluorinated alkane and sili-
cone oil. Graefes Arch Clin Exp
Ophthalmol 240:131–136
15. Marquardt H, Schäfer G (1994)
Lehrbuch der Toxikologie, vol I.
Wissenschaftsverlag, Mannheim, p 138
16. Meinert H, Geister U (1992) Semiflu-
orinated compounds in ophthalmology.
J Vitreoretina 1:31–35
2. Bednarz J, Aboalchamat B,
17. Meinert H, Roy T (2000) Semifluor-
inated alkanes – a new class of com-
pounds with outstanding properties
for use in ophthalmology. Eur
Engelmann K (2000) Characterization
of immortalized human corneal endo-
thelial cells [abstract]. Invest
10. Hoerauf H, Kobuch K, Dresp J,
Menz D (2001) Combined use of
Ophthalmol Vis Sci 41:448
J Ophthalmol 10:189–197
3. Bednarz J, Teifel M, Friedel P,
Engelmann K (2000) Immortalization
of human corneal endothelial cells
using electroporation protocol optimi-
sed for human corneal endothelial and
human retinal pigment epithelial cells.
Acta Ophthalmol Scan 78:130–136
partially fluorinated alkanes, perfluoro- 18. Mertens S, Bednarz J, Richard G,
carbon liquids and silicone oil: an ex-
perimental study. Graefes Arch Clin
Exp Ophthalmol 239:373–381
Engelmann K (2000) Effect of perflu-
orodecaline on human retinal pigment
epithelium and human corneal endo-
thelium in vitro. Graefes Arch Clin
Exp Ophthalmol 238:181–185
19. Nakamura K, Refojo MF, Crabtree DV,
Pastor J, Leong FL (1991) Ocular
toxicity of low-molecular-weight
components of silicone and fluorosili-
cone oils. Invest Ophthalmol Vis Sci
32:3007–3020
20. Peyman G, Schulman J, Sullivan B
(1995) Perfluorocarbon liquids in
ophthalmology. Surv Ophthalmol
39:375–395
21. Sobottka Ventura A, Böhnke M,
Löliger C, Kühnl P, Winter R,
Engelmann K (1996) HLA-Typisierung
von Spenderhornhäuten mit hohen
post-mortem-Zeiten. Ophthalmologe
93:262–267
11. Kirchhof B (1999) Fluorkarbone in
der Netzhaut-Glaskörperchirurgie.
Ophthalmochirurgie 11:153–158
4. Böhnke M (1991) Corneal preservation 12. Kobuch K, Menz DH, Hoerauf H,
in organ culture. Curr Opin Ophthalm-
ol 2:432–442
Dresp JH, Gabel VP (2001) New sub-
stances for intraocular tamponades:
perfluorocarbon liquids, hydrofluoro-
carbon liquids and hydrofluorocarbon-
oligomers in vitreoretinal surgery.
Graefes Arch Clin Exp Ophthalmol
239:635–642
5. Chang S, Zimmerman HJ, Iwamoto T
(1997) Experimental vitreous replace-
ment with perfluortributylamine. Am
J Ophthalmol 103:29–37
6. Engelmann K, Friedl P (1989) Optimi-
zation of culture conditions for human
corneal endothelial cells. In Vitro Cell
Dev Biol 25:1065–1072
7. Engelmann K, Sobottka-Ventura A,
Drexler D, Staude HJ (1998) A sensi-
tive method for testing the quality of
organ cultured media of individual
medium components in a cornea bank.
Graefes Arch Clin Exp Ophthalmol
236:312–319
13. Langefeld S, Schrage N, Meinert H,
Roy T, Aretz A, Kirchhof B (1999)
A new way of removing silicone oil
from the surface of silicone intraocular
lenses. Graefes Arch Clin Exp
Ophthalmol 237:201–206
14. Litzkas P, Jha KK, Ozer HL (1984)
Efficient transfer of cloned DNA into
human diploid cells: protoplast fusion
in suspension. Mol Cell Biol
4:2549–2552

995
22. Stolba U (1996) Perfluorocarbone als
Glaskörperersatz – klinische und
experimentelle Ergebnisse. In:
Faulborn J, Freyler H, Göttinger W
(eds) Spectrum Augenheilkunde 10
Suppl 19. Springer, Berlin Heidelberg
New York, pp 3–28
23. Wilbanks G, Apel A, Jolly S,
Devenyi R, Rootman D (1996) Per-
fluorodecaline corneal toxicity: five
case reports. Cornea 15:329–334
24. Winter M, Beehrende S, Menz DH,
Pfister G (1999) Retinale Druck-
belastung durch Perfluordecalin.
Ophthalmologe 1 [Suppl]:103
25. Zeana D, Becker J, Kuckelhorn,
Kirchhof B (1999) Perfluorhexyloctane
as a long-term vitreous tamponade
in the experimental animal. Int
Ophthalmol 23:17–24