Full text of DOI:10.1007/s004170100264

Graefe’s Arch Clin Exp Ophthalmol
(2001) 239:373–381
L A B O R AT O RY I N V E S T I G AT I O N
DOI 10.1007/s004170100264
Hans Hoerauf
Karin Kobuch
Joachim Dresp
Dirk-Henning Menz
Combined use of partially fluorinated alkanes,
perfluorocarbon liquids and silicone oil:
an experimental study
Abstract Background: Partially flu- pending on the type of FALK and ra-
Received: 8 March 2000
Revised: 15 January 2001
Accepted: 31 January 2001
Published online: 5 May 2001
© Springer-Verlag 2001
orinated alkanes (FALKs) are a new
class of substances which can be
tio used. (3) The lipophilic proper-
ties of FALKs and PFCLs could be
used in vitreoretinal surgery as an in- characterized by their dissolution in
traoperative tool and as a long-term
tamponade. The aim of this in vitro
native olive oil (F6H8=23.4 m%,
O44=16.7 m%, F6H6=12.3 m%,
study was (1) to investigate the solu- O62=5.3 m%, PFD=1.1 m%,
bility of FALKs in silicone oil during PFO=0.6 m%). (4) It was possible to
direct exchange, (2) to study their
combined use and solubility in
prepare “heavy silicone oil” e.g. by
adding 30 vol% F6H8, resulting in a
PFCLs, (3) to evaluate their lipophil- specific gravity of 1.08 g/ml, or by
ic properties and (4) to investigate
the possibility of preparing “heavy
silicone oil”. Methods: (1) Four dif-
ferent FALKs (F6H6, F6H8, O44 and
O62) were directly exchanged with
silicone oil (5,000 mPas). The dis-
solved amount of fluorocarbons in
the removed silicone oil was deter-
mined by gas chromatography and
by gravimetry. Furthermore, the dif-
fusion phenomena during the ex-
change process were studied. (2) The possible, if the right ratio is chosen.
behaviour of FALKs in PFCLs was (3) The solubility of FALKs in na-
investigated and the solubility of the tive olive oil may be an indicator for
resulting mixtures in silicone oil was their tissue penetration and may
measured. (3) The solubility of
FALKs and PFCLs in native olive oil term tamponade. (3) “Heavy silicone
adding 80 vol% O44, resulting in a
specific gravity of 1.25 g/ml, but de-
creasing the viscosity of the mixtures
dramatically. Conclusion: (1) If
FALKs are used as an intraoperative
tool, a direct exchange with silicone
oil should be avoided owing to their
capacity to dissolve in silicone oil,
resulting in a mixture with unpre-
dictable properties. (2) A combined
use with PFCLs and silicone oil is
This study is part of the work of the AGGE
J. Dresp and D.-H. Menz have a commer-
cial, proprietary or financial interest in re-
search or devices described in this study
H. Hoerauf (
Department of Ophthalmology,
University of Lübeck,
Ratzeburger Allee 160,
23538 Lübeck, Germany
)
✉
render feasible their use as a long-
was analysed. (4) Different FALKs
were added to silicone oil and mea-
surements of the resulting specific
gravity and the viscosity were per-
formed. Results: (1) FALKs dis-
solved in silicone oil up to the
oil” preparation using FALKs is pos-
sible, but the mixture needs further
evaluation in terms of emulsifica-
tion, mobilization of oligosiloxanes,
tissue penetration and long-term sta-
bility.
e-mail: hhoerauf@ophtha.mu-luebeck.de
Fax: +49-451-5003085
K. Kobuch
Department of Ophthalmology,
University of Regensburg, Germany
following values: F6H6=45 m%,
F6H8=54 m%, O44=100 m%,
O62=18 m%. (2) FALKs dissolved in
PFCL, thereby changing the physi-
cochemical properties of PFCL de-
J. Dresp
Bausch & Lomb Surgical, Munich,
Germany
D.-H. Menz
Pharm Pur GmbH, Augsburg, Germany

374
Introduction
Material and methods
Substances
Silicone oil and perfluorocarbon liquids (PFCLs) are
well-known substances, and many experimental and clin-
ical studies have classified their properties and their role
in the management of retinal detachments [1, 18,19]. In
the past decade PFCLs have become increasingly popular
as intraoperative tools in vitreoretinal surgery and have
facilitated the surgical management of complicated reti-
nal detachments [2, 29]. PFCLs have also been tested ex-
perimentally as a long-term vitreous substitute, showing
retinal damage in several studies [3, 4, 33]. Recently, a
new class of partially fluorinated alkanes (FALKs) [20,
32] was approved (CE-marked) in Europe for intraocular
The four different partially fluorinated alkanes used were F6H6,
F6H8, O44 and O62. F6H6 (perfluorohexylhexane) and F6H8 (perflu-
orohexyloctane) are already certified for medical use and commer-
cially available (Fluoron, Neu-Ulm, Germany; Geuder, Heidel-
berg, Germany). O44 (perfluorobutylbutane) and O62 (üperfluoro-
hexylethane) are still in clinical trials (Bausch&Lomb Surgical,
Munich, Germany; Pharm Pur, Augsburg, Germany). Chemical
and physical properties of the four different FALKs are listed in
Table 1. As perfluorocarbon liquids, ADATO OCTA (perfluorooc-
tane, PFO) and ADATO DECA (perfluorodecalin, PFD) were
used. They were obtained from Bausch &Lomb Surgical. Their
properties are listed in Table 1 as well. All studies were performed
using silicone oil with a viscosity of 5,000 mPas (ADATO
use [21]. Compared with PFCLs, FALKs have a different SIL-OL; Bausch &Lomb Surgical).
structure, a significantly lower specific gravity and a
higher potential to dissolve several lipophilic substances
and silicone oil [22]. They have been already investigated
Experiments
as a solvent for silicone oil adhesions on intraocular lens-
es [8, 9,15] and are recommended as a wash-out solution
Direct exchange of FALKs for silicone oil
The reduced specific gravity of FALKs compared with PFCLs has
after silicone oil removal. Recent studies have proven
their intraocular tolerance as a long-term endotamponade
experimentally [6, 14, 37] and clinically [12, 30].
PFCLs and silicone oil show no clinically significant
interactions during combined use. This clinical experi-
ence and the used intraoperative techniques cannot be
transferred to the new class of partially fluorinated al-
kanes. The exact knowledge of their interactions with
significant advantages during certain vitreoretinal manoeuvres
[10]. However, in selected clinical situations such as severe trac-
tion retinal detachment a direct exchange for silicone oil might be
preferred to avoid retinal slippage [36]. This experiment was per-
formed to investigate the solubility of the different FALKs in sili-
cone oil during direct exchange. In order to determine the mixing
ratio during direct exchange and to estimate the resulting risks, the
maximum solubility of FALKs in silicone oil and vice versa was
determined (a). The concentration reached during the exchange
silicone oil and PFCL is essential for their combined use was simulated under clinical conditions (b). Spontaneously occur-
in vitreoretinal surgery.
ring diffusion phenomena during direct contact were investigated
(c).
The presented in vitro study had the following aims:
(1) To characterize the solubility of four selected FALKs
(a) Measurement in cylindrical vials. In order to determine the re-
in silicone oil, which is an important property if a direct
exchange is intended; (2) to study their interaction with
PFCLs evaluating their capability to function as a tool
during macular translocation; (3) to measure the solubili-
ty of FALKs in native olive oil as a marker for lipophili-
city and tissue penetration capability; (4) to evaluate the
possibility of preparing “heavy silicone oil” by mixing
different FALKs with silicone oil and to assess the influ-
ence of the FALKs on the chemical properties of the re-
sulting substance.
spective saturation concentrations the substances in question were
mixed in 5-ml cylindrical glass vials. The solubility of two sub-
stances in each other depends on concentration and temperature.
The most common way to characterize the solubility is by a so-
called solubility diagram. The solubility of each FALK in silicone
oil, and vice versa, was measured using cylindrical vials. Silicone
oil was added to the selected FALK, and the selected FALK to sili-
cone oil, in small increments. The temperature at which the transi-
tion opacification occurred was determined for each ratio, creating
the solubility diagram. The temperature of the mixtures was
changed until they became opaque or cleared again. Solutions in a
molecular ratio which can be described as “FALK in silicone oil”
Table 1 Chemical and physical properties of semifluorinated alkanes (O62, O44, F6H6, F6H8) and PFCLs (PFO, PFD)
PFO
(perfluoro-
octane)
PFD
(perfluoro-
decalin)
O62
O44
F6H6
(perfluoro-
hexylhexane)
F6H8
(perfluoro-
hexyloctane)
(perfluoro-
(perfluoro-
hexylethane) butylbutane)
Purity
>99%
104°
1.78
1.27
1.4
14.0
438
>99%
142°
1.92
1.31
5.53
17.6
462
>99%
113.7°
1.62
>99%
102.8°
1.24
Boiling point (°C)
187
223
1.35
1.34
2.5
21.0
432
Density (g/cm3)
Refractive index (nD)
Viscosity (mPas)
Surface tension/air (mN/m)
Molecular weight (g/mol)
1.42
1.32
1.85
20.0
404
1.29
1.31
0.75
0.77
14.7
17.4
348
276

375
form so-called ideal solutions, whereas the ratio “silicone oil in stances, their penetration into the retina may be possible, particu-
FALK” is characterized by formation of aggregation systems. larly if they are used as long-term tamponades. Therefore we de-
Both forms of solution are transparent and cannot be differentiated termined the lipophilicity of F6H6, F6H8, O44 and O62.
macroscopically. The determination whether the mixture created
The current standard for determining the lipophilicity of flu-
was an ideal solution or just an aggregated system was performed orocarbons for biomedical applications is the measurement of the
by means of light scattering using a laser pointer (wavelength: CTSH value (critical temperature for solubility in n-hexane – the
630–660 nm, power:< 1 mW). At least two determinations were lower this temperature, the higher the lipophilicity). Because visu-
performed.
al observation of the opacification point served as an indicator for
lipophilicity of FALKs with CTSH values below 0°C, the preci-
(b) Measurements in a glass eye model. To simulate clinical condi- sion of this subjective method was not sufficient to determine the
tions optimally, a slow manual direct exchange in a glass eye mod- lipophilic properties of such compounds. In consequence, the sol-
el was performed, avoiding turbulences and maintaining a con- ubility of the selected FALKs and of PFCLs was measured in na-
trolled temperature of 37°C. The glass eye had two “sclerotomies” tive olive oil, another method for characterization of lipophilicity
to which syringes or infusion lines could be connected. The ex- behaviour.
change processes were identical to the clinical manoeuvres. After
Recent studies have shown that solubility in olive oil may
filling the glass eye with balanced salt solution (BSS) using the in- serve as a model for biological tissue penetration capability [16].
fusion line and subsequent exchange with one of the different The measurements were performed at room temperature by adding
FALKs by injection through the second opening, substitution of the characterized substance to native olive oil in cylindrical vials
FALK by silicone oil (5,000 mPas) was performed with a suction under constant mechanical stirring, until irreversible opacification
cannula. To minimize the influence of turbulence during the ex- occurred. Subsequently, an additional amount of fluorocarbon was
change process and to simulate intraoperative conditions, the sili- added to this mixture and it was evaluated at which temperature
cone oil was removed from the model eye after waiting 15 min. the native olive oil became clear again [16]. The solubility was de-
Removal was performed by aspiration using a syringe connected tected gravimetrically.
to the “sclerotomy site” and by injection of BSS via the infusion
line. The dissolved fluorocarbon molecules in the silicone oil were
detected using gas chromatography.
“Heavy silicone oil”
(c) Diffusion experiment in plastic syringes. This experimental set- Due to the solubility of FALKs in silicone oil, “heavy silicone oil”
up was chosen to create defined geometrical conditions and to in- can theoretically be created. Its higher specific gravity could have
vestigate different interactions between fluorocarbons and silicone marked advantages in the treatment of complicated retinal detach-
oil. Plastic syringes (5 ml) were placed in an upright position and ments, especially if inferior breaks or retinotomies are present.
filled with PFD, O44 and O62, then cooled down to temperatures The increased viscosity compared with FALKs could be advanta-
below 0°C. Subsequently silicone oil was added and the tempera- geous with respect to droplet formation and applicability.
ture was increased to 37°C. Fifteen minutes later the temperature
Therefore, changes in the specific gravity of silicone oil
was decreased again to “freeze” the diffusion profile. Samples at (5,000 mPas) were measured after adding different amounts of
different heights were taken using syringes and 20-gauge cannu- F6H6, F6H8 and O44. The viscosity was determined in a rheovis-
las. The amount of fluorocarbons in the different layers of the cometer according to Höppler. GC/MS was performed using a
mixture was measured by a head-space sampler (HS40 Perkin- Shimadzu GC 17AQP5000 coupled with a head space sampler
Elmer) coupled with gas chromatography (Shimadzu GC (HS40 Perkin-Elmer).
17AQP5000). This machine measures samples from a gas phase
which is in equilibrium with an underlying liquid phase. This tech-
nique allows detection of volatile compounds in silicone oil quali-
tatively and quantitatively.
Results
Direct exchange of FALKs with silicone oil
Combined use of FALKs, PFCL and silicone oil
(a) Measurement in cylindrical vials
FALKs were introduced into vitreoretinal surgery because of their
reduced specific gravity compared with PFCLs and have shown
advantages for the macular rotation manoeuvre. As FALKs and
silicone oil are miscible, a direct exchange of one of these two
substances for the other should be avoided. Therefore, before use
in surgery FALKs are mixed with PFCLs to avoid their dissolution
in silicone oil.
The solubility of mixtures of the four different FALKs in PFO
was analysed by adding the FALK/PFO mixtures to silicone oil at
a controlled temperature of 37°C until opacification, an indicator
of saturation, occurred. The substances were stirred until satura-
The solubility of O44, F6H8, F6H6 and O62 was measured
using cylindrical vials by adding silicone oil to the se-
lected FALK, and the selected FALK to silicone oil, in
small increments. The solubility diagram for F6H8 is
shown in Fig. 1. Only at a certain temperature are de-
fined ratios of FALK and silicone oil miscible [22, 23].
Outside these defined ratios the two substances are im-
tion was achieved. The amount of fluorocarbon mixtures dissolved miscible, creating a so-called mixture gap. At mixture
in silicone oil was measured gravimetrically. Furthermore, the sol-
gaps the solutions opacified, because they were not sta-
ubility of different ratios of FALK/PFO mixtures in silicone oil
ble, and the two substances separated to single phases
was analysed in order to determine a combination and a ratio
again. Another characteristic point is the change of the
solution structure at a molecular ratio of approximately
50%. Molecular ratios which can be described as “FALK
in silicone oil solutions” form so-called ideal solutions,
whereas “silicone oil in FALK solutions” are character-
which allows a direct exchange with silicone oil.
Lipophilicity of FALKs and PFCLs
The lipophilicity of a substance can be used as an indicator for its
tissue penetration properties. Since FALKs are lipophilic sub- ized by the formation of aggregation systems. That could

376
be demonstrated with the beam of a laser pointer: a solu-
tion of “silicone oil in FALK” (silicone oil:FALK ratio
<1) showed a Tyndall phenomenon and differed marked-
ly from a solution of “FALK in silicone oil” (silicone
oil:FALK ratio >1), where no Tyndall phenomenon
could be observed. If the temperature of the “silicone oil
in FALK” solution was increased the Tyndall phenome-
non disappeared, and if the temperature was diminished,
the phenomenon reoccurred, a typical indicator for self-
organization. Further cooling lead to the formation of
two separate phases. This behaviour illustrates the com-
plex interactions of the different components in such
mixtures, and it has to be considered that the individual
molecular structure of the FALK that is used has an im-
portant influence on the stability of the formed solutions.
The solubility of the different FALKs in silicone oil
(5,000 mPas) over the whole temperature range was: O44
> F6H8 > F6H6 > O62. If the temperature was reduced, the
solubility in silicone oil of all FALKs used was de-
creased. Assuming that the clinical exchange process is
performed at room temperature (25°C), the measured
values were: O44=100 m%, F6H8=54 m%, F6H6=45 m%
and O62=18 m%. There was a correlation between the
solubility in silicone oil, the lipophilicity and the molec-
ular structure of the FALKs. These values are the results
of double tests; therefore, no standard deviation is given.
The detected values are the upper limits of solubility of
FALKs in silicone oil during a direct exchange.
Fig. 1 Solubility diagram indicating the mixture gap in the system
ADATO SIL-OL 5000 and F6H8
(b) Measurements in the glass eye model
The quantity of FALKs measured dissolving in the sili-
cone oil samples was influenced markedly by the unsuit-
able surface properties of the glass eye, which showed a
considerable amount of visible FALK remnants after the
exchange process. Due to the adhesion forces of FALKs
Fig. 2 Mechanisms involved in direct exchange of FALKs and sil-
icone oil
and silicone oil to the glass surface, in contrast to clinical thin layer. However, if FALKs were added, a marked
conditions exact evaluation of the position of the inter- shift towards the silicone oil phase occurred and a larger
face shift during the exchange was not possible. Various mixing phase or interfacial layer developed (Fig. 3). The
modifications of the glass surface failed to improve its diffusion profile for PFD, O62 and O44 is shown in Fig. 4.
unsuitable properties for this experiment. The measured The steepness of the concentration gradient and the dif-
values were not reproducible and showed huge differ- fusion distance from the former phase boundary are
ences of more than 100% deviation. Visually the condi- strongly correlated with the solubility of the different
tions influencing the mixture could be described (Fig. 2), fluorocarbons in silicone oil.
but quantification was not possible.
Combined use of FALKs, PFCLs and silicone oil
(c) Diffusion experiment in plastic syringes
Initially, after adding the different FALKs to PFCL, inho-
In the diffusion experiment a shift in the position of the mogeneities could be detected with the naked eye which
interfaces could be observed as a clear indication of a disappeared after a few seconds without stirring. The sub-
dissolution process. The direction in which the dissolu- stances dissolved completely without later separation of
tion occurred could be described. After adding PFO to the phases. The resulting mixtures showed the same behav-
silicone oil only a minimal shift of the interface to the iour as PFCL in BSS, but revealed different chemical prop-
silicone oil was seen and the mixing phase appeared as a erties, including a higher solubility for silicone oil. There-

377
Table 2 Solubility of PFCLs (PFD, PFO) with 10 vol% FALK
(O62, O44, F6H8) in silicone oil (5,000 mPas) in contrast to the pure
FALK
PFD
PFO
Pure
(perfluoro- (perfluoro- FALKs
octane)
6.2 m%
decalin)
3.9 m%
O44
100 m%
18 m%
54 m%
(perfluorobutylbutane)
O62
5.8 m%
6.4 m%
3.7 m%
4.1 m%
(perfluorohexylethane)
F6H8
(perfluorohexyloctane)
fore the solubility of FALK/PFCL mixtures in silicone oil
was measured. Several ratios of FALK and PFCLs were
chosen. The resulting solubilities of the mixtures in a ratio
of “PFCL/FALK in silicone oil” ranged between the values
of the pure components. In contrast to the pure FALKs, sil-
icone oil was not miscible with mixtures of PFCL/FALKs.
The reason for this additional mixture gap was that the ad-
dition of the PFCL destroyed the solvent structure of sili-
cone oil in FALK. Therefore, an addition of PFCLs to sta-
ble “silicone oil in FALK” mixtures but also to
“PFCL/FALK in silicone oil” mixtures led to a spontane-
ous opacification or phase separation. Stable mixtures ex-
isted only in a limited range of “PFCL/FALK in silicone
oil” ratios, provided the sequence of addition of the single
components had been correct.
Fig. 3 Interaction of silicone oil and FALK or silicone oil and
PFCL
The results of solubility measurements for a ratio of
90 vol% PFO or PFD and 10 vol% FALK (O44, O62, F6H6
and F6H8) are presented (Table 2). For this ratio the solu-
bility for silicone oil was nearly the same as for pure
PFCLs (below 7 m%). In comparison, PFO could be dis-
solved up to 3.7–4.4 m% and PFD up to 5.8–6.4 m% in
silicone oils of different viscosities. Only PFCL/FALK
mixtures with nearly the same solubilities as the PFCLs
allowed a direct exchange with silicone oil, i.e. the FALK
concentration in the mixtures had to be 10% or less.
Lipophilicity of FALKs and PFCL
The solubility of PFO in native olive oil was 0.6 m% and
of PFD 1.1 m% respectively. The values measured for
FALKs were markedly higher: F6H8=23.4 m%, O44=
16.7 m%, F6H6=12.3 m% and O62=5.2 m%. The corre-
sponding molecular fraction correlated with their solu-
bility in silicone oil.
Mixtures of FALK in silicone oil – “heavy silicone oil”
The theoretically achievable changes in specific gravity
of silicone oil (5,000 mPas) by addition of different
Fig. 4 Diffusion profile of PFD, O62 and O44 in ADATO SIL-OL
5000

378
uble in one another. In former studies the different spe-
cific gravity of the two substances was used concomi-
tantly for combined long-term tamponade of the lower
and upper retina [28, 31]. In these studies no interaction
among PFCLs and silicone oil was noted. This situation
has changed markedly with the introduction of the so-
called partially fluorinated alkanes which are able to
form mixtures with silicone oil [22, 23]. The solubility
of the different FALKs in silicone oil varies significantly.
Therefore, exact knowledge of the different properties of
these new substances is extremely important if the vitre-
oretinal surgeon is to profit from their benefits and avoid
complications if a combined usage of FALKs and sili-
cone oil is intended.
Fig. 5 Theoretical changes of specific gravity by addition of dif-
ferent fluorocarbons to ADATO SIL-OL 5000: (....) indicates the
mixture gap of each fluorocarbon
Regarding solubility properties in the system
FALK/silicone oil, it has to be taken into account that so-
lutions of “FALKs in silicone oil” or “silicone oil in
FALKs (F6H8, O44 and O62) are shown in Fig. 5. The use FALK” show different effects. Usually the transition of a
of the FALK with the highest specific gravity (O62) “FALK in silicone oil” to a “silicone oil in FALK” solu-
yields mixtures with a specific gravity as high as possi- tion happens by a sudden change in solubility. This indi-
ble. However, the achievable increase in specific gravity cates that at this concentration structural changes in the
is limited by the solubility of FALKs in silicone oil. The- solution occur. In contrast to the purified PFCLs, the li-
oretically, an increase of specific gravity to 1.01 g/ml pophilic behaviour, expressed as the solubility of several
should be sufficient to obtain “heavy silicone oil”, but FALKs in silicone oil, cannot be predicted by the CTSH
the subsidence in water-containing media is also influ- value. The solubility in native olive oil is a better mea-
enced by surface tension and adhesion forces of sur- sure of the lipophilicity of FALKs [16], although struc-
rounding tissues. This could be observed in experiments tural influences in the solution have to be taken into ac-
with 20 vol% F6H8 in silicone oil, which was adequate to count. Especially in the case of small amounts of sili-
obtain a specific gravity of 1.02 g/ml, but showed an in- cone oil, the solubility can be dominated by structural in-
stability of the phase separation between “heavy silicone fluences. No simple correlation exists between the length
oil” and BSS, with floating and hovering of parts of the of the fluorinated or non-fluorinated part of the FALK
oily phase after moving the vial. To guarantee a stable and the solubility in silicone oil. Nevertheless, the ratio
tamponade which is heavier than water, a specific gravi- of length and molecular structure of both parts is of great
ty of at least 1.06 should be achieved in order to obtain a importance for the stability of the resulting solutions and
safe subsidence and phase separation inert tamponade. the dimension of the mixture gap in the system
Therefore, if F6H8 is used, a mixture near the saturation FALK/silicone oil. With decreasing solubility of FALK
limit is necessary for safe intraoperative use.
in silicone oil and of silicone oil in FALK, mixtures are
After addition of FALK the silicone oil changes its re- stable at higher temperatures only. During the exchange
fractive index and its original viscosity. An addition of process when the mixture ratios might be changed a seg-
10 vol% O62 to silicone oil (5,000 mPas) resulted in a regation may occur.
specific gravity of 1.04 g/ml and a decrease in viscosity
The following mechanisms may influence the amount
from 5,000 to 2,852 mPas. An addition of 30 vol% F6H8 of dissolved FALKs in silicone oil during their direct ex-
resulted in a specific gravity of 1.08 g/ml and a decrease change (see Fig. 2): (A) Diffusion through the contact in-
in viscosity to 1,136 mPas. Increasing the amount of terface of the two substances and (B) the dissolution by
FALK leads to further decrease in viscosity of the result- contact of silicone oil and FALK remnants on the retinal
ing mixture and thus may influence the risk of emulsifi- surface. This is enhanced by (C) turbulences during the
cation.
exchange process.
The silicone oil/FALK-mixtures described above
could be reduced if a larger volume of silicone oil would
pass the eye. However, this cannot be performed in clini-
cal routine without the risk of uncontrolled increase of
intraocular pressure.
Discussion
Direct exchange of FALKs with silicone oil
In the diffusion experiment the penetration depth of
Vitreoretinal surgeons are used to perfluorocarbon liq- the different FALKs correlated with their solubility. A
uids and silicone oil as intravitreal substances which do significant difference between PFCL and FALK could be
not show significant clinical interactions and are not sol- observed with a marked shift of the FALK/silicone oil in-

379
terface compared with the PFCL/silicone oil interface. the use of O62 alone. F6H6 has a lower specific gravity
However, this shift is also the prerequisite for a direct ex- than O62, but O62 shows the lowest solubility of the ex-
change, because opacifications are minimized this way.
amined FALKs in silicone oil. This is favourable for a
With increasing solubility of FALKs in silicone oil, direct exchange procedure and allows a clinically com-
the direct exchange becomes more difficult. If the ex- plete exchange.
change process is not conducted fast enough, complete
However, O62 has the highest specific gravity
exchange can become impossible. Only O62, the FALK (1.62 g/ml) among the presented FALKs. It seems that
with the lowest solubility in silicone oil, might be used if properties of FALKs other than specific gravity are of
a direct exchange cannot be avoided.
importance, especially interfacial tensions which facili-
Generally, an exchange from FALK to air should be tate the rotation manoeuvre. Further investigations will
preferred to avoid any dissolving of FALK in silicone oil have to clarify, whether the specific gravity is the critical
during the exchange process. However, it has to be kept property for the macular rotation procedure.
in mind that even if a clinically complete exchange of
Nevertheless, a lower specific gravity is favourable.
FALK with air is assumed, a small amount of FALK may The reduction of specific gravity of fluorocarbon can be
remain as a transparent layer on the retinal surface. This achieved by replacement of the fluorine atoms by hydro-
was already described for PFCL by Winter et al. [34]. gen atoms only. This, however, leads to a more hydrocar-
Since the enthalpy of the different FALKs is not known, bon type substance which shows a higher potential for
the exact vapour pressure at 37°C cannot be given. Us- solubility in silicone oil. The properties “low specific
ing fluorocarbons with a boiling point of 115°C at body gravity” and “high potential for solubility in silicone oil”
temperature results in a vapour pressure between 40 and coincide; thus, the substances F6H8 and O44 are not suit-
60 mbar. PFO, with a boiling point of 104°C and a va- able for combined use with silicone oil.
pour pressure of 43 mbar at 37°C, shows markedly better
Therefore, some general rules for the exchange proce-
evaporation than PFD, which has a boiling point of dure have to be followed:
142°C and a vapour pressure of 8.8 mbar. The FALKs
with a boiling point below 115°C have evaporation be- a) If PFCL is to be added to FALK intraoperatively, the
haviour similar to that of PFO and they are therefore
more suitable if an air exchange must be performed to
minimize remnants in the gas phase.
Therefore, the most suitable FALKs for an air ex-
change are O44 and O62. However, it has to be kept in
mind that an intraoperative fluid–air exchange is often
impossible in cases of severe tractional retinal detach-
ment. In these situations a PFCL should still be used.
resulting mixture should consist of at least 90% PFCL
before replacement with silicone oil is performed.
b) Even the least soluble FALK (O62) dissolves in sili-
cone oil more easily than PFCLs. Despite the fact that
the resulting silicone oil shows no visual differences
from “clean” silicone oil, viscosity and lipophilicity
are influenced in an unpredictable manner. Therefore,
direct exchange of FALK with silicone oil should be
considered with the least soluble FALK (O62) only.
c) If FALKs are already in contact with silicone oil, any
further contact between PFCLs and both other sub-
stances should be avoided.
Combined use of FALKs, PFCLs and silicone oil
FALKs and PFCLs can be mixed intraoperatively and
dissolve in each other. However, if a direct exchange of
the resulting mixture to silicone oil is planned, the se- Lipophilicity and tissue penetration capability
quence of substances should guarantee that the solubility of PFCLs and FALKs
of the PFCL for silicone oil has not increased. The use of
such mixtures could be considered during the macular The lipophilic properties of FALKs are important during
rotation procedure when a direct exchange is desired to their use as an intraoperative tool or vitreous substitute,
avoid retinal slippage [36]. To take advantage of the spe- because lipophilic compounds may penetrate ocular tissue
cific properties of a small bubble of FALK during the ro- like the retina or the optic nerve, as already shown for sili-
tation, its volume should not exceed 10 vol% of the sub- cone oil[5, 11, 13,26]. It has been shown that the determi-
sequently added PFCL. In this case the solubility of nation of solubility of FALKs in olive oil can be an ade-
PFCL is not markedly affected and complete direct ex- quate model for their tissue penetration capability [16].
change of the resulting mixture with silicone oil is guar-
All published results indicate that the examined
anteed. As compared with PFD analogues their solubility FALKs may have the capacity to penetrate tissue. As a
in silicone oil is low, PFO and PFO/FALK mixtures are result retinal damage might occur during long-term use.
favourable. If a larger amount (>10 vol%) of FALK is However, it remains speculative whether penetration
used, precipitations and opacifications during the direct through the inner limiting membrane (ILM) of the retina
exchange occur, which may cause severe complications takes place. The hydrated character of the ILM, which is
during the exchange process. An alternative method is classified as the basal membrane of the Müller cells,

380
should prevent a marked penetration of lipophilic mixture with a specific gravity of 1.08 g/ml. However,
FALKs.
this led also to a change in other properties, including a
The increased lipophilicity of FALKs compared with significant reduction in viscosity and a marked decrease
PFCLs also has advantages. Intraocularly retained in the refractive index. It is not yet known whether the
FALKs can dissolve in silicone oil better than the resulting mixture remains stable or whether it may lose
PFCLs, which have significantly lower lipophilicity. its higher specific gravity by emigration of the lipophilic
Small FALK bubbles should vanish in this way, resulting alkane. Due to lipophilic properties of the FALK/silicone
in a better optical performance of the silicone oil. It has oil solution, the capability of tissue penetration is un-
been reported that PFCL remnants can be mobilized by known and the anatomical and functional consequences
adding lipophilic FALKs [22]. Higher amounts of during a long-term tamponade can not be predicted. Fi-
FALKs (>10 vol%) markedly raise the solubility of nally, the mobilization of toxic oligosiloxanes in silicone
PFCLs, but unfortunately the risks caused by the insta- oil [25] by the addition of FALKs cannot be excluded.
bility of the mixtures so formed rise proportionally, as This hypothesis is supported by a study on the release of
mentioned above. Based on our observations concerning low-molecular-weight siloxanes and platinum from sili-
lipophilicity and segregation processes, preliminary tests cone breast implants [17]. These compounds are much
on diffusion mechanisms between different silicone more soluble in FALKs than silicone oil. Therefore, ex-
oil/FALK mixtures and aqueous solution like tissue cul- perimental in vivo studies are needed to evaluate this is-
ture media (DMEM) have been performed. First results sue.
indicate that lipophilic components leach into aqueous
media and form the basis for tissue penetration experi-
ments using retinal cultures.
Conclusion and clinical consequences
The new class of partially fluorinated alkanes offers
Mixtures of FALK and silicone oil – “heavy silicone oil” promising substances for vitreoretinal surgery, which
provide a variety of new properties. Basing on the results
A vitreous tamponade heavier than water would have presented in this study the following conclusions can be
marked advantages in the treatment of complicated de- drawn:
tachments of the lower retina, especially if large inferior
breaks or retinotomies are present. The use of PFCLs in a) This new class of vitreous substances cannot ade-
vitreoretinal surgery is limited mainly to a temporary
tamponade, and removal at the end of surgery is recom-
mended. Changes in the retinal architecture were ob-
served in animal studies using PFCLs as a long-term vit-
reous substitute [2, 3, 33]. Since these changes were
quately replace perfluorocarbon liquids in all their
current functions. Vitreoretinal surgeons should not
use them as if they were only another PFCL. General-
ly, direct replacement of FALKs by silicone oil should
be avoided.
found mainly in the lower retina, the high specific gravi- b) O44, the FALK with complete solubility in silicone oil
ty of the PFCLs was thought to be the reason for this
finding. Therefore FALKs, which have markedly lower
specific gravity, may offer advantages not only during
seems to be the best candidate as a wash-out solution
for silicone oil remnants which are not covered by hy-
drophobic substances.
certain vitreoretinal manoeuvres but also as a long term c) A combined use of PFCLs and FALKs, allowing di-
vitreous substitute [10]. Other disadvantages of PFCLs,
such as the combination of low viscosity and high spe-
cific gravity with the tendency toward emulsification and
dispersion, cannot be overcome by the use of FALKs. A
rect replacement of the resulting mixture by silicone
oil, is possible only if the right ratio of 9:1 is chosen.
This property is advantageous during macular rotation
if subsequent replacement by silicone oil is necessary.
new class of hydrofluorocarbons formed by oligo- d) FALKs are lipophilic compounds and their tissue pen-
merization of FALK may solve this problem [14].
Peyman et al. reported successful treatment of three
patients with fluorosilicone oil, which has a specific
etration capability is still unclear. Therefore, their use
as a heavy long-term vitreous tamponade should be
restricted to clinical studies at this time.
gravity of 1.31 g/ml [27]. However, toxic reactions after e) The preparation of “heavy silicone oil” by the addition
use of this oil were observed in experimental studies
[24]. An alternative preparation of “heavy silicone oil”,
which leads to a physical mixture and not to a chemical
compound, can be achieved by the addition of FALKs to
silicone oil [7,35]. Wolf et al. reported about the success-
ful use of such a mixture (Oxane HD) in a small series of
of FALKs to silicone oil is possible, but a marked re-
duction in viscosity and a shift in the refractive index
occur. The long-term stability, the tissue penetration
capability and the toxicity of the mixture have not
been sufficiently evaluated yet. The mixtures of FALKs
and silicone oil described need further research.
patients with inferior retinal pathologies. We found that f) Future clinical studies will have to define the exact
the use of approximately 30 vol% F6H8 resulted in a indications for the use of the FALKs presented here.

381
References
1. Chang S, Ozmert E, Zimmerman NJ
(1988) Intraoperative perfluorocarbon
liquids in the management of prolifera-
tive retinopathy. Am J Ophthalmol
106:668–674
2. Chang S, Sparrow JR, Iwamoto T,
Gershbein A, Ross R, Ortiz R (1991)
Experimental studies of tolerance to in-
travitreal perfluoro-n-octane liquid.
Retina 11:367–374
3. Eckardt C, Nicolai U (1993) Clinical
and histological findings after several
weeks of intraocular tamponade with
perfluorodecalin. Ophthalmologe
90:443–447
4. Eckardt C, Nicolai U, Winter M, Knop
E (1991) Experimental intraocular tol-
13. Knorr HL, Seltsam A, Holbach C,
Naumann GO. Intraocular silicone oil
tamponade. A clinico-pathologic study
of 36 enucleated eyes. Ophthalmologe
93:130–138
14. Kobuch KA, Hoerauf H, Menz D-H,
Eckert E, Gabel VP (2000) A new hy-
drofluorocarbon with increased viscos-
ity and reduced density for long-term
tamponade of the lower retina. Invest
Ophthalmol Vis Sci 41 [Suppl]:662
15. Langefeld S, Kirchhof B, Meinert H,
Roy T, Aretz A, Schrage NF (1999) A
new way of removing silicone oil from
the surface of silicone intraocular lens-
es. Graefe´s Arch Clin Exp Ophthalmol
237:201–206
27. Peyman GA, Smith RT, Charles H
(1987) Injection of fluorosilicone oil
and pars plana vitrectomy for complex
retinal detachment. Can J Ophthalmol
22:276–278
28. Peyman GA, Conway MD, Soike KF,
Clark LC (1991) Long-term vitreous
replacement in primates with intravit-
real vitreon or vitreon plus silicone oil.
Ophthalmic Surg 22:657–664
29. Peyman GA, Schulman JA, Sullivan B
(1995) Pefluorcarbon liquids in oph-
thalmology. Surv Ophthalmol
39:375–395
30. Schrage NF, Wong D, van Meurs J,
Macek M, Dahllke C, Kirchhof B
(2000) F6H8 als Endotamponade bei
PVR-Amotiones. Ophthalmologe
[Suppl 1]:176
erance to liquid perfluoroctane and per- 16. Lee TD, Arlauskas RA, Weers JG
fluoropolyether. Retina 11:375–384
5. Eckardt C, Nicolai U, Czank M,
Schmidt D (1992) Identification of sili-
cone oil in the retina after intravitreal
injection. Retina 12:17–22
6. Gabel VP, Kobuch K, Hoerauf H,
Menz DH, Dresp J, Laqua H (2000)
Specific gravity, the cause of retinal
damage induced by perfluorocarbon
liquids? Invest Ophthalmol Vis Sci 41
[Suppl]:662
(1996) Characterization of the lipo-
philicity of fluorocarbon derivatives
containing halogens or hydrocarbon
blocks. J Fluorine Chem 78:155–163
17. Lykissa ED, Kala SV, Hurley JB,
Ledovitz RM (1997) Release of low
molecular weight siloxanes and plati-
num from silicone breast implants.
Anal Chem 69:4912–4916
18. McCuen BW II, DeJuan E, Machemer
R (1985) Silicone oil in vitreoretinal
surgery. 1. Surgical techniques. Retina
5:189–197
31. Sparrow JR, Jayakumar A, Berrocal M,
Ozmert E, Chang S (1992) Experimen-
tal studies of the combined use of vitre-
ous substitutes of high and low specific
gravity. Retina 12:134–140
32. Turberg MP, Brady JE (1988) Semiflu-
orinated hydrocarbons: primitive sur-
factantmolecules. J Am Chem Soc
110:7797–7801
33. Velikay M, Wedrich A, Stolba U,
Datlinger P, Li Y, Binder S (1993) Ex-
perimental long-term vitreous replace-
ment with purified and non-purified
perfluorodecalin. Am J Ophthalmol
116:565–570
34. Winter M, Winter C, Wiechens B
(1999) Quantification of intraocular re-
tained perfluorodecaline after macro-
scopic removal. Graefe’s Arch Clin
Exp Ophthalmol 237:153–156
7. Genovesi-Ebert F, Rizzo S, Figus M,
Ferretti C, Nardi M (2000) Combina-
tion of silicone solvent (F6H8) and sili- 19. McCuen BW II, DeJuan E, Landers
cone oil (PDMS) in complex retinal
detachment. Invest Ophthalmol Vis Sci
41 [Suppl]:663
8. Hoerauf H, Menz DH, Dresp J, Laqua
H (1999) Use of O44 as a solvent for
silicone oil adhesions on intraocular
lenses. J Cataract Refract Surg
25:1392–1393
9. Hoerauf H, Menz DH, Dresp J, Laqua
H (1999) O44 - ein Lösungsmittel für
Silikonöladhäsionen auf Intraokularlin-
sen. Klin Monatsbl Augenheilkd
214:71–76
10. Kirchhof B (1999) Fluorkarbone in der
Netzhaut- Glaskörperchirurgie. Oph-
thalmochirurgie 11:153–158
MB, Machemer R (1985) Silicone oil
in vitreoretinal surgery. 2. Results and
complications. Retina 5:198–205
20. Meinert H, Knoblich A (1993) The use
of semifluorinated alkanes in blood
substitutes. Biomater Artif Cells Im-
mob Biotechnol 21:583–595
21. Meinert H: DE 19536504 A1;
PCT/EP96/03542
22. Meinert H, Roy T (2000) Semifluor-
inated alkanes – a new class of com-
pounds with outstanding properties for
use in ophthalmology. Eur J Ophthalm- 36. Wong D, Williams RL, German MJ
ol 10:189–197
23. Menz DH (1998) On the solubility
of silicone oil in fluorocarbons.
12th European Symposium on Fluorine
Chemistry, Berlin
35. Wolf S, Schön V, Meier P, Wiedemann
P (2000) Schweres Silikonöl als Endo-
tamponade bei komplizierter Net-
zhautablösung. Ophthalmologe [Suppl
1]:45
(1998) Exchange of perfluorodecalin
for gas or oil: a model for avoiding
slippage. Graefe’s Arch Clin Exp
Ophthalmol 236:234–237
11. Kirchhof B, Tavakolian U, Paulmann
H, Heimann K (1986) Histopathologic
findings in eyes after silicone oil injec-
tion. Graefe´s Arch Clin Exp
37. Zeana D, Becker J, Kuckelkorn R,
Kirchhof B (2000) Perfluorohexyloc-
tane as a long-term vitreous tamponade
in the experimental animal. Int
Ophthalmol 23:17–24
24. Miyamoto K, Refojo MF, Tolentino FI,
Fournier GA, Albert DM (1986) Fluor-
inated oils as experimental vitreous
substitutes. Arch Ophthalmol
Ophthalmol 224:34–37
12. Kirchhof B, Wong D, van Meurs J
(2000) Long term vitreous tamponade
by Perfluorohexyloctane (F6H8): a
clinical pilot study. Invest Ophthalmol
Vis Sci 41 [Suppl]:78
104:1053–1056
25. Nakamura K, Refojo MF, Crabtree DV,
Leang F-L (1990) Analysis and frac-
tionation of silicone and fluorosilicone
oils for intraocular use. Invest
Ophthalmol Vis Sci 31:2059–2069
26. Nishizono H, Ikurata Y, Tanuka M,
Soji T, Herbert DC (1993) Evidence
that Muller cells can phagocytize
egglecithin-coated silicone oil parti-
cles. Tissue Cell 25:305–310