Reports 627
Volume 13
Number 8
Perfluoropentane
in experimental ocular
surgery. IAN J. CONSTABLE.
Air has long been advocated as a mechanical
aid to both anterior segment'• 2 and retinal detachment surgery.v However, the fact that air
is absorbed within three to four days has lead
to the investigation of gases which might last
longer in tissues.'''7
One factor which helps determine the absorption rate of gases is aqueous solubility.1- n Thus,
argon lasts longer than the more soluble carbon
dioxide/ but neither outlasts air. A second factor
is molecular weight. Hence, two insoluble inorganic gases of large molecular weight, sulfur
hexafluoride and octafluorocyclobutane, are both
retained in the aqueous and vitreous cavity longer
than air.'1-"
Many of the family of perfluoronated carbon
compounds are highly insoluble. With increasing
molecular weight, however, the boiling point
drops, so that the six-carbon compounds are
liquid at body temperature. However, the fivecarbon compounds such as perfluoropentane
(CiFu) have a boiling point of about 30° C ,
so that they are liquid at room temperature, but
gaseous at body temperature.
Liquid volumes ranging from approximately
0.2 p\ to 2 iA of perfluoropentane were drawn
up in a chilled, sterile microsyringe. These volumes
were then immediately injected into the anterior
chamber (14 eyes), vitreous cavity (14 eyes),
or subconjunctival tissues (6 eyes) of rabbits.
The rabbits were anesthetized with pentobarbital
sodium and postoperative mydriasis was maintained with atropine. In each situation this minute
amount of liquid was immediately heated by the
tissues beyond its boiling point and converted
into bubbles of gas, which quickly coalesced
into one or two large bubbles (Fig. 1). The
material was injected from a dependent position,
so that the gas bubbles rose away from the injection site.
In the anterior chamber 0.5 to 2 ix\ of perfluoropentane expanded immediately to occupy
one-tenth to one-half the normal aqueous volume.
No suture was used, and an immediate rise in
intraocular pressure was avoided by displacement
of aqueous out the injection site. In each case
the gas expanded a further 300 to 600 per cent
in volume over the following seven days (Fig.
2), presumably due to the trapping of dissolved
tissue gases."
In four eyes injected with 2 jtil of liquid perfluoropentane, secondary expansion of the gas
completely replaced the aqueous humor. Each
of these eyes developed acute glaucoma, resulting
in a cloudy cornea, deepened anterior chamber,
posterior synechia, and generalized lens opacities.
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Fig. 1. Bubbles of vaporized perfluoropentane
immediately after injection of 0.2 fA of liquid
into the anterior chamber of a rabbit eye.
Fig. 2. Maximum secondary expansion of a perfluoropentane gas bubble seven days after injection
of 0.2 fi\ of liquid into the anterior chamber.
Note fibrin strands (—>) and localized corneal
haze above.
In another ten eyes injected with less than 0.5
jul of perfluoropentane, secondary expansion of the
gas did not completely fill the anterior chamber,
and intraocular pressure did not rise (Fig. 2).
Slit-lamp examination revealed fibrin deposits and
a flare in the residual aqueous in the first few
days in all ten eyes (100 per cent), such as is
seen in rabbits after paracentesis alone. After
one week, localized corneal haze developed in
six eyes (60 per cent) and anterior subcapsular
lens opacities in two eyes (20 per cent). In
each instance the signs of injury were confined
to the area directly in contact with the gas bubble.
Corneal haze was not seen in four eyes (40 per
cent), and in each of these the bubbles were
628 Reports
Fig. 3. Multiple perfluoropentane gas bubbles
in the vitreous cavity of a rabbit eye twelve
weeks after injection of 0.5 /*! of liquid.
Fig. 4. Subconjunctival perfluoropentane gas four
months after injection. There are no signs of irritation, but the gas bubble appears encapsulated.
multiple or remained small enough to move around
in the anterior chamber. The corneal haze cleared
completely in all six affected eyes when the
bubble became smaller after three weeks. The
lens opacity (two eyes) did not progress after
the second week, but persisted as a localized
anterior subcapsular plaque. The volume of gas
began to slowly decrease two weeks after injection, finally disappearing from the anterior chamber
by ten to twelve weeks. No corneal opacities
nor synechia were seen in any of these ten eyes
two months after complete resorption of the gas.
Injection of 1 to 2 /A of perfluoropentane into
the vitreous cavity of six rabbit eyes resulted in
a shallow anterior chamber and acute glaucoma
in each case (100 per cent), These eyes developed a cloudy cornea and generalized cataract.
When 0.5 n\ of perfluoropentane was injected
into the vitreous (eight eyes), both the imme-
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
Investigative Ophthalmology
August 1974
diate and the delayed expansion of the gas was
accommodated in all eyes without a rise in
intraocular pressure. The gas expanded to a
maximum volume in four days of about 20 per
cent of the vitreous cavity as judged ophthalmoscopically. The gas volume began to decrease
three weeks after injection, to about 10 per cent
of the vitreous volume at six weeks, and to less
than 3 per cent at twelve weeks (Fig. 3). Several
minute gas bubbles persisted in four eyes (50
per cent) after six months. Clinical examination
revealed no inflammatory cells, opacities, or scar
tissue formation in the vitreous during the period
of observation. No opacities or vacuoles developed in the lens of any of these eyes. Histologic examination of two eyes at four and six
months revealed no abnormalities in the lens
or retina, and no fibroblastic proliferation in the
vitreous.
Subconjunctival injection of perfluoropentane
(six eyes) resulted in extensive ballooning of
these tissues. Gas migrated right around the limbus and posteriorly into the orbit. Again, the
gas caused no clinical signs of irritation, and
was slowly resorbed in each case over six months.
The residual subconjunctival gas at four months
was localized in a bleb (Fig. 4) in three eyes
(50 per cent), but was no longer present in any
animals at six months.
These preliminary studies show that perfluoropentane will persist in biological tissues and
cavities for much longer than other gases reported. It causes very little inflammation, but
is definitely irritative to the corneal endothelium
and lens if in immobile direct contact. A similar
irritative effect has been noted with air and other
gases.H> a"° However, when the gas bubbles are
able to move around in the aqueous or vitreous
cavities, evidence of toxicity is not seen clinically.
This previously observed fact is also true for perfluoropentane, even though it persists for months.
Expansion of sulfur hexafluoride in the vitreous
can be controlled by admixture with air.G To
test this hypothesis with perfluoropentane, the
liquid was added to air in a syringe in a cold
room (4° C.) in concentrations of 1 and 5 /xl
per cubic centimeter of air. After warming to
37° C. in an incubator, a small bubble of each
concentration was injected into the anterior chamber (four eyes) and vitreous cavity (four eyes)
of rabbits. The bubbles were examined and photographed daily.
Perfluoropentane in a concentration of 1 fi\ per
cubic centimeter of air showed no clinical evidence of expansion in any of the eyes. The gas
bubble only began to decrease in size after
three weeks. In concentrations of 5 fi\ per cubic
centimeter, the bubble expanded 200 to 300 per
cent in volume over three days in all eyes, in
both the aqueous and vitreous.
Reports 629
Volume 13
Number 8
Attempts to add perfluoropentane directly to
an air bubble already injected into the aqueous
or vitreous proved unreliable. It was not possible
to measure such small amounts of the perfluoropentane in these situations.
Long-term evaluation is necessary to determine
if perfluoropentane is sufficiently nontoxic for
clinical use. However, these experiments suggest
important principles in the search for an inert
expander of tissue spaces or cavities. Any insoluble chemical of relatively high molecular weight
which will vaporize below 37° C. might be
expected to persist in a gaseous phase in the
body for a prolonged period of time.
From the Department of Retina Research,
Retina Foundation, 20 Staniford St., Boston, Mass.
02114. Supported by funds from the Retina
Foundation and by Center Grant EY-00227 of the
National Eye Institute, National Institutes of
Health. Submitted for publication Feb. 8, 1974.
Reprint requests: Editorial Services, Retina Foundation, 20 Staniford St., Boston, Mass. 02114.
Key words: perfluoropentane, anterior chamber,
vitreous, subconjunctival tissues, vaporization,
toxicity.
REFERENCES
1. Selinger, E.: Injection of air into the anterior
chamber after cataract extraction, Am. J.
Ophthalmol. 20: 827, 1937.
2. MacMillan, J. A.: Injection of air as factor
in maintaining filtration after corneo-scleral
trephining in glaucoma, Arch. Ophthalmol.
22: 968, 1939.
3. Rosengren, B.: t)ber die Behandlung der
Netzhautablbsung mittelst Diathermie und
Luftinjektion in den Glaskorper, Acta. Ophthalmol. 16: 3, 1938.
4. Schenk, H.: Experimentelle und klinische
Untersuchungen zur Frage der Resorption
gasformiger Stoffe aus dem glasskorper, Albrecht von Graefes Arch. Klin. Ophthalmol.
161: 252, 1959.
5. Widder, W.: Tierversuche iiber die verweildauer verschiedener Glaskorperimplantate, Albrecht von Graefes Arch. Klin. Ophthalmol.
164: 550, 1962.
6. Norton, E. W. D.: Intraocular gas in the
management of selected retinal detachments,
Trans. Am. Ophthalmol. Otolaryngol. 77: 85,
1973.
7. Vygantas, C. M., Peyman, G. A., Daily, M.
J., et al.: Octafluorocyclobutane and other
gases for vitreous replacement, Arch. Ophthalmol. 90: 235, 1973.
8. Van Horn, D. L., Edelhauser, H. F., Aaberg,
T. M., et al.: In vivo effects of air and sulfur
hexafluoride gas on rabbit corneal endothelium,
INVEST. OPHTHALMOL. 11: 1028,
1972.
9. Constable, I. J., and Swann, D. A.: Vitreous
substitution with gases. Arch. Ophthalmol.
(in press).
Downloaded From: http://iovs.arvojournals.org/ on 06/17/2017
A model of anisotropic factors contributing
to retinal venous nicking. B.
SMILEY,
JONATHAN
D.
CARROLL
WIRTSCHAFTER,
AND JAMES F. LAFFERTY.
A laboratory model consisting of fluid flow
through Penrose tubing arrangements in a pressure chamber was used to explain the "nicking"
phenomenon. The effects of various forces possibly at play in the regions of arteriovenous
crossings were considered. The triangular or
"shark's tooth" shape of the underlying vein observed by investigators can be produced by the
forces generated by the movement of a hypertensive artery. The narrowed lumen of the vein
and, therefore, the narrowed blood column at the
crossings can explain the ophthahnoscopic appearance of "nicking."
This investigation determined, in a laboratory
model, the effects of various forces which may act
to deform the veins in the region of retinal arteriovenous crossings. Histopathologic reconstructions
from eyes with hypertensive retinopathy have
revealed progressive changes in the shape of the
cross-section of nicked retinal veins.1' - At such
crossings, the veins have triangular or shark toothshaped cross-sections instead of . the indented
cylinders normally seen. These deformations extend for essentially equal distances both proximal
and distal to the crossings. The shape of the crosssection undergoes progressive translation from
triangular through elliptical to circular. The axis
of the ellipse is at right angles to the internal
limiting membrane of the retina. In this study the
retinal vein was treated as a collapsible tube and
the question asked was what forces were necessary
to produce the observed deformations.
Previous laboratory studies of collapsible tubes
have dealt with water flowing through circular rubber tubes rigidly supported at either end of a pressure chamber.3-4 In such models the wall properties
of the tubing were essentially uniform around the
circumference while the wall properties of the
retinal veins are not uniform because 20 per cent
of the circumference of the venous perimeter is
attached to the overlying, stiffer artery at the
crossing. Such inequality of directional properties
is known as anisotropy. In addition, the retinal
vein is not suspended in a homogeneous media
but is embedded in and attached to the retinal
structures which are also likely to subject the
vein to anisotropic forces. The deformed shape
of the venous cross-section may result from the
interaction of anisotropic forces exerted by the
hypertensive artery on the anisotropic structure
of the vein and the surrounding retina.
A collapsible piece of rubber tubing was connected to fittings at either end of clear Lucite
cylinder 20 cm. long and 8 cm. in diameter (Fig.