corneal neovascularization argon · treatment of the vessel between these two points. intravenous...

Brit. J. Ophthal. (1976) 6o, 464

Corneal neovascularization treated with argon laser

PAUL M. H. CHERRY' AND ALEC GARNER2From the Pocklington Fye Transplantation Research Unit, Toronto', and the Pathology Department,Institute of Ophthalmology, London2

Corneal neovascularization can be troublesome,especially in an eye which is to have a corneal graft.A corneal transplant is more likely to be rejectedin a vascularized than a non-vascularized cornea.In this and many other clinical situations, there-fore, it is desirable that corneal new vessels beobliterated. Many treatments have been suggested,including corticosteroids, irradiation, thiotepa,cryotherapy, photocoagulation, cysteine, chon-droitin sulphuric acid, and dextran. One of us(PMHC) was, we believe, the first to report the useof the argon laser in the treatment of cornealneovascularization in man (Cherry, Faulkner,Shaver, Wise, and Witter, I1973). This was anopen trial in four patients, two with chemicalburns of the cornea and two with herpes simplexkeratitis. There were two failures, one success,and one partial success. It was impossible to drawany definite conclusions from so few cases, and itwas therefore decided to embark on a controlledtrial in rabbits.

Materials and methodsOne of the eyes of each of I I pigmented Dutch rabbitswere used as both case and control. Each animal weighedabout 2 kg. Two 7' o' or 8' o' black silk sutures wereplaced in the cornea about 3 mm in from the limbus,one at io o'clock and one at 2 o'clock. The animalswere anaesthetized with intravenous sodium pento-barbitone. Usually the cornea was perforated duringsuture placement with slight resultant loss of aqueousand shallowing of the anterior chamber. In every casethe chamber quickly re-formed and no anterior syne-chiae to the corneal perforation ensued. The sutureswere used as the stimulus for comeal new vessel forma-tion, which took about three weeks to reach the suturesfrom the limbus. When neovascularization was adequatethe sutures were removed. Comeal fluorescein angio-grams were then obtained under anaesthesia to docu-ment accurately the position of the vessels and deter-mine which were arteries and which were veins. Amodified Zeiss photoslit lamp camera was used for theangiography with a Balzar interference excitationfilter (FITC 4) and a Kodak Wratten 15 barrier filter.

Address for reprints: Paul M. H. Cherry, FRCS, Department ofOphthalmology, University of Toronto, 1 Spadina Crescent,Toronto, Canada

Because of the positioning of the sutures two sectorsof vessels were stimulated, one from the upper nasallimbus and the other from the upper temporal limbus.Two sectors of vessels in the same eye were neededso that one could be used as a control and the other asthe argon-treated area. The argon laser therapy wasperformed under sodium pentobarbitone anaesthesiaand repeated at intervals of several days until thevessels were completely occluded. (Only one animal,no. 2, was killed before the treated vessels were occluded.This was done deliberately to obtain some histologicalspecimens very soon after argon laser treatment.)The technique of laser therapy was as follows.

Both arteries and veins were treated as they crossedthe limbus and 2 to 3 mm to the corneal side of thelimbus. The first treatment of each vessel was 'down-stream' and the second 'upstream', thus trapping acolumn of blood which was able to absorb subsequenttreatment of the vessel between these two points.Intravenous fluorescein was usually given immediatelybefore the argon laser therapy so that more laser energywould be absorbed by the corneal vessels as they werebeing treated (Kohner, personal communication).Intravenous fluorescein was repeated during the treat-ment session to document the success or failure of eachstage of the treatment. When the treated sector ofcorneal vessels was occluded the fluorescein angiogramswere repeated at intervals of several days to make surethat the control vessels were still patent. The irisdamage during the experiment was also documentedwith fluorescein angiography until the changes werestabilized. The instrument used for the argon lasertherapy was manufactured by Coherent Radiation,Palo Alto, California, USA. The settings were usually50 F.m spot size, 0-2 s exposure time, and I 50 mW energy(see Table).When the animals were killed their treated eyes were

enucleated for histopathological study. Just beforesome of the animals were killed the pupil of thetreated eye was dilated and the fundus examined withthe indirect ophthalmoscope to determine if there wasany visible retinal damage. No damage was found.

Results

The results are summarized in the Table. Theargon laser was effective in obliterating cornealnew vessels, provided a continued stimulus forrevascularization of the cornea was not present.An active stromal infiltrate in the area of the suture

on June 15, 2020 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.60.6.464 on 1 June 1976. Dow

nloaded from

http://bjo.bmj.com/

Corneal neovascularization 465

Table Results of Argon laser treatment of corneal neovascularization in i i pigmented Dutch rabbits

Rabbitno.

2*

3

Suturesremovedon day:

0

0

0

Argonlasertreatmenton days:

9, 10, I3,

'4

No. oftreatments

4

I6 I

0, 3, 10, 6

13, 17,

42

4 0 0, 7, 10 3

5 0 II, 21 2

6 0 3, 7, 14 3

7 0 0, 7 2

No. ofshotsat eachtreatment

230

'34

I50

226

740

350

74547'

382

586291

394

2869

375

993I2

786

467280

747

368

45'

320

II39

88

I94

282

Powersetting(mW)

272 at ioo

92 at 200

I50 at 250

226 at 300

740

200

2283 at 150

586 at 250

2869

I50

150

I50

I50

Animalkilledon day:

I64

20

8o

Controlvesselspatent atdeath ofanimal

No (goneby day132)

Eye

Right

Yes Left

No (gone Leftby day69)

Nasal or

temporalvesselsectortreated

Temporal

Temporal

Nasal

77 Yes Right Nasal

9I Yes Left Nasal

25 Yes Right Temporal

I 8 No (goneby dav'4)

Left Temporal

8 0 0,4,28 3 2299

1456567

4322

100 32 Yes Right Temporal

9 0 7, 14, I 8 3

10 0 3, 7,I4 3

I I 0 3, 8 2

482

307239

1028

249415

316

980

833

288

1121

987 at I504I at 250

I028t

664 at I503I6 at 200

980

150

63 Yes Left Nasal

25 Yes Left Temporal

46 Yes Right Nasal

*Rabbit deliberately killed before occlusion to study histopathology very soon after argon laser treatmenttForty-one burns were at 100 ±Lm spot size not 50 xAm


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/

466 British Journal of Ophthalmology

scar was apparent during the sessions of argon lasertherapy in rabbits 3 and 8, and these were the twoanimals which required the greatest number oflaser shots (2869 and 4322 at six and three treat-ment sessions respectively). In both these animals,however, the laser-treated sector was successfullyoccluded after the suture scar inflammation hadresolved.When occluding the vessels an effort was made

to use the minimum amount of argon laser energypossible, thus limiting the iris damage. The mini-mum spot size (50 Im) was used (except for 41shots of ioo sum spot size in rabbit 9) together withan exposure time of 0o2 s: 150 mW of energy wasusually selected, but 100, 200, 250, and 300 mWwere occasionally used. Rabbit 8 had only ioo mWtreatment, and we suggest that this power metersetting is adequate so long as no stimulus forrevascularization of the cornea remains.We hoped that it would be enough to treat only

the arteries and that the veins, deprived of theirblood supply, would then spontaneously regress.But this was not found to be the case in rabbit i i(Figs I-5), in which all arteries in the treated sectorwere successfully obliterated by day 3 after sutureremoval; the veins were not treated. However, when FIG. 2 Rabbit iI. Pretreatment fluorescein angiogram,examined on day 8 the vein marked V2 was carrying right eye, showing 2 c'clock sector of vessels in arterialblood to the suture scar from the limbus and had phase

thus become the afferent part of a venous loop. V2was obliterated with the argon laser on day 8.The rabbit was next seen on day i i, when V3 was

... the afferent limb of a venous loop and V1 theefferent limb (Fig. 5). By this time the vesselswere of very fine calibre and V3 and V1 sponta-neously disappeared by day 14. Thus, two examplesof retrograde flow were demonstrated in thisrabbit, and it is therefore not sufficient to treatonly arteries when attempting to obliterate cornealnew vessels.

Before each session of argon laser therapyfluorescein was given intravenously to ensure thatmaximum laser energy would be taken up by thecorneal blood vessels (Kohner, personal communi-cation). The smaller and less obvious vesselscould also be visualized and therefore treated.Additionally, fluorescein given intravenously severaltimes during each treatment helped to show whenthe vessels had been occluded.

Apart from in rabbit 2, the laser therapy even-tually closed all the treated sectors of vessels. Thisrabbit was purposely killed four days after thefirst and only session of laser therapy in order thathistopathology of the acute lesion might be ob-served. In the remaining io rabbits the average

FIG. i Rabbit iI. Pretreatment photograph of 2 number of treatment sessions required to obliterateo'clock sector of vessels. Vi1, V2, and V3 are veins the vessels was 3. I. The animals were then ob.


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/


served for varying periods before being killed tomake sure that the control vessels remained patentand did not disappear spontaneously. The controlvessels in rabbits i, 3, and 7 did disappear spon-taneously at days I I8, 27, and 7 respectively aftersuccessful occlusion of the treated vessels. The

Lr~ ~ ~ V

FIG. 5 Rabbit i i. Fluorescein angiogram day iiafter argon laser treatment. "V." is afferent limb andV1 efferent limb of a venous loop. "V3" has filledfrom limbus and is beginning to empty into V,

control vessels of the other rabbits remainedpatent until they were killed.

Changes in the iris were observed at the time oflaser application in all animals. Energy not absorbedby the corneal blood vessels reached the iris andcaused a pigment disturbance on the surface. A

FIG. 3 As for Fig. 2 showing 2 o'clock sector of vessels black patch appeared on the surface of the irisin venous phase. V1, J72, and V3 are veins underneath the treated vessel and sometimes the

iris surface became slightly excavated, but byabout two months the excavation was almostimperceptible and the hyperpigmentation of theblack patch was less marked. In rabbit i theseiris changes were accompanied by very slightectropion uveae.

Fluorescein angiography was performed on thenon-vascularized eye of rabbit 9 two to three

V) hours after the application of 6oi (i o8 mW,VV3 5o rim, 0-2 s) argon laser burns to the iris (Figs

1 / 6-9). The black patches did not fluoresce, the/ ~~~~~~~pigment, acting as a curtain, preventing fluores-

cence from the underlying iris stromal vessels.Dots of fluorescein leakage, however, appeared

j~~~~~.~~~from many iris vessels, including those at theedge of the black patches, within 30 seconds ofintravenous fluorescein injection (Fig. 8), andthese dots slowly increased in size. There was alsoa massive outpouring of aqueous from the ciliarybody. Thirty to 6o seconds after intravenousfluorescein injection fluorescent aqueous started

FIG. 4 Rabbit i i. Photograph day i a a to enter the anterior chamber through the pupil,

- 4 after argonlaetramn.""iSafet rbad

laser treatment Veins V, and V3 are only vessels and after about five minutes the iris could hardlyremaining patent be seen in blue light because of the large amount of


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/

468 British Yournal of Ophthalmology

immediately beneath the epithelium, although inthree animals (I, 4, and 7) vascularization was moreobvious in the deep or midstroma. The vessels onthe treated side of the cornea were variably sur-rounded by scanty macrophages, lymphocytes, andproliferated fibrous scar tissue. In most instances

-rWE,W< >§74< _ the vessel would appear to have been completelydestroyed (Fig. i ia, b), while occasional lessseverely damaged vessels were occluded by throm-bus. Two corneae (rabbits 8 and io) showed smallsolitary aggregates of necrotic debris associatedwith trivial leucocytic infiltration between Desce-met's membrane and the lining endothelium(Fig. 12). These probably represented the site ofperforation into the anterior chamber of the suturesinserted into the cornea to induce vascular in-

FIG. 6 Rabbit 9. Photograph of right eye two to growth.three hours after application of 6oi (I00 mW, 50 ,um, Iris lesions were included in the sections of four0-2 s) argon laser burns to iris (corneal vessels not eyes (rabbits 2, 8, 9 (right eye), and I I). In eachinduced in this eye). Note black patches of iris damage case they presented as minute foci of stromal

pigment disturbance which had given rise to tinyaccumulations of melanin-laden macrophages nearto the anterior surface of the iris (Fig. I Ic). Therewas no sign of residual necrosis or inflammation,and the pigment epithelium on the posteriorsurface of the iris was intact in every instance.

Discussion

The power of the argon laser is confined to thegreen and blue regions of the spectrum with 8oper cent of the total confined to two wavelengths,5145 A (green) and 4880 A (blue) (Campbell,Rittler, Swope, and Wallace, I969). Energy atthese wavelengths is transmitted by the opticalsystem of the eye with relatively low absorptionand therefore high efficiency (Meyer-Schwickerath,

FIG. 7 As for Fig. 6. Black patches of iris damagemask fluorescence (I 5 s post-injection)

fluorescent aqueous in the anterior chamber. Therabbit's other eye was used for comparison andthere the entry of fluorescent aqueous into theanterior chamber via the pupil was much less.Three days after the application of the laser theamount of aqueous produced by the ciliary bodywas much less, and a week after the laser treatmentit had returned to normal (Fig. io). The focal irisleakage seen two to three hours after treatment(Figs 8, 9) had ceased. -

PATHOLOGYFIG. 8 As for Fig. 6. Dots offluorescein leakage

The corneal blood vessels seen in histological from many iris vessels including those at edge of blacksections were generally in the superficial stroma patches (30 s post-injection)


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/


I960). Pigment epithelium absorbs a high degreeof energy at these wavelengths, and a significantportion of the energy is also absorbed by blood(Geeraets, William, Chan, Ham, Guerry, andSchmidt, I962).The argon laser was effective in occluding

suture-induced corneal new vessels, the sutureshaving been removed before treatment with thelaser. In rabbits 3 and 8 an active infiltrate of thestromal suture scar remained after suture removal,and the argon laser did not occlude the vessels inthese animals until the inflammation subsided.We therefore suggest that the argon laser canocclude corneal blood vessels in rabbits so longas there is no, or very little, stimulus remainingfor revascularization. In these optimum circum-stances very little laser energy has to be used toocclude the vessels (50 Im, 100-I50 mW, 0o2 S).The most obvious complication of argon laser

therapy in these ii Dutch rabbits was damage tothe iris. This must particularly be borne in mindbecause of the iris route of corneal graft rejection(Jones, I973). If the argon laser is to be usedextensively to occlude corneal new vessels inman we think it will be to occlude corneal newvessels before keratoplasty. Therefore if argon lasertherapy to corneal vessels is complicated by anincreased chance of corneal graft rejection bythe iris route it must surely be discarded as ameans of treating corneal new vessels in thesecircumstances. Black iris patches, 'shadows' ofthe treated corneal vessels, appeared immediatelyon treating the vessels in our Dutch rabbits andwere caused by any laser energy not taken up bythe corneal vessels.

Unger, Perkins, and Bass (I974) showed thatlaser irradiation of the iris in pigmented Dutchrabbits is followed by a transient increase inintraocular pressure of up to 45 mmHg within iominutes, with a return to normal within 2 hours.They suggested that this increase is mediated byE-type prostaglandins. They also found a break-down of the blood-aqueous barrier with increasedpermeability to plasma proteins occurring in theciliary processes and not in the iris. Fluoresceinangiography performed in our experiment con-firmed the findings by Unger and others (I974).Angiograms two to three hours after the lasertreatment showed no fluorescence through thesurface black patches on the iris but there werefocal dots of fluorescent leakage from many irisvessels, including those at the edge of the blackpatches (Figs 8, 9). This leakage persisted for afew days but had always resolved within a week.There was also massive outpouring of fluorescentaqueous from the ciliary processes, through thepupil, in the fluorescein angiograms two to three

FIG. 9 As for Fig. 6. Fluorescent aqueous beginningto pour into anterior chamber through pupil. Dots offluorescein leakage from many iris vessels includingthose at edge of black patches. Black patches of irisdamage mask fluorescence (45 s post-injection)

FIG. I0 Rabbit 9. Fluorescein angiogram of righteye seven days after argon laser treatment to iris.Black patches of iris damage mask fluorescence. Thereis now no focal leakage offluorescein from the irisvessels (about I min post-injection)

hours after laser treatment. The intraocular pres-sure was not measured at this time or immediatelyafter laser treatment, but presumably it wastransiently raised. The breakdown of the blood-aqueous barrier resolved slowly and was also backto normal within a week.

Histopathological examination of the iris lesionsshowed minute foci of stromal pigment distur-bance which had given rise to tiny accumulationsof melanin-laden macrophages near the anterior


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/


surface of the iris (Fig. i ic). There was no sign ofresidual necrosis or inflammation and the pigmentepithelium on the posterior surface of the iris wasintact in every instance. This suggests that the irisdamage stabilizes about one week after argon lasertreatment of corneal blood vessels in pigmentedDutch rabbits and should not, therefore, neces-sarily contraindicate the use of the argon laser totreat corneal blood vessels in patients being con-sidered for keratoplasty. The corneal vessels couldbe occluded about two weeks before the proposedkeratoplasty, thus allowing the iris time to stabilizebefore the operation. It must also be mentionedthat further work by Unger and Brown (personalcommunication) has shown, so far, that the blood-aqueous barrier in humans is much more stableafter ruby laser therapy of the iris than that ofpigmented Dutch rabbits.The following technique of argon laser treatment

of corneal new vessels is recommended.i. Suppress the corneal lesion which is stimu-

lating the neovascularization with appro-priate treatment, thus reducing the stimulusfor revascularization to a minimum.

2. Perform intravenous fluorescein angiographyat the time of laser therapy in order todetermine which of the corneal vessels arearteries and which are veins and to visualizeeasily the smaller vessels. Furthermore, withfluorescein in the vessels at the time of lasertherapy the energy absorption by the vesselsis increased (Kohner, personal communica-tion).

3. Set the argon laser at I00 mW, 50 ,um, 02 sand treat the corneal arteries about 3 mm infrom the limbus, and then 'upstream' atthe limbus. Having isolated a column ofblood in this manner, treat between thesetwo points on all the arteries until thearteries are obviously occluded and the bloodflow in the veins therefore stagnant. Thelaser power can be increased to 150 mW, butit should not be necessary to use a greaterpower.

4. The retrograde flow in the veins (Figs 2-5)is at first very sluggish, taking about a day tobecome fully established. Even then, thecalibre of the venous loop vessels is alwaysless than before the arteries were occluded.However, this retrograde flow makes itnecessary to treat all vessels rather than justthe arteries. The veins are first treated at thelimbus, then 'upstream' 3 mm in from thelimbus, and then between the two pointsjust mentioned. They need far fewer lasershots than the arteries. They are treatedimmediately after the arteries, when the

4-,~~~

FIG. iiRabbit 2. (a) Section showing occluded bloodvessel in superficial stroma of cornea associated with alittle reactive scar tissue formation. Underlying irisshows small shallow excavation lined by pigment-filledmacrophages: iris otherwzise intact. Haematoxylin andeosin. X 92. (b) and (c) Higher magnifications ofoccluded corneal vessel and iris lesion seen in (a).Haematoxylin and eosin. x i 8o


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/

Coryzeal neovascularization 471

..... ..

..S

.... .... ..-li.;Ao

... .....M:N

Ai:

......d 4b

.....

....... .......

........

.... .....

-1h

..Ab4%::,::....

....

.. .......

..

Aft Aa-

b

(tIc)


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/


-4. .

t

FIG. I2 Rabbit io. Several treatedvessels in superficial corneal stromaassociated with mild scarring; smallfocus of debris between Descemet'smembrane and endothelium probablyrelated to earlier insertion of silksuture to induce vascularization.Haematoxylin and eosin. X I95

-_. w K .j s ;- :..

*b, vr _-*$t @* 9 .... 40

blood flow in the veins will be stagnant orvery sluggish.

5. If retreatment of the vessels is necessarysome days later it should be in a similarmanner to the first treatment session, butfewer laser shots should be needed to occludeall comeal vessels at the second and anysubsequent treatment sessions.

SummaryThe argon laser set at 50 Im, 100-150 mW, 0-2 Soccluded corneal blood vessels in pigmented Dutchrabbits provided the corneal lesion responsiblefor inducing vascularization was inactive. Afterarterial treatment with the argon laser retrogradeflow in untreated veins was demonstrated by

fluorescein angiography. Therefore all corneal newvessels should be treated, not just arteries. Minimaliris damage complicated the laser therapy, but thiswas not thought necessarily to contraindicate theuse of the argon laser to treat corneal blood vesselsin man. The iris damage was associated withoutpouring of aqueous from the ciliary processes,and it took up to a week for the blood-aqueousbarrier to return to its normal state.

We thank the Medical Research Council for donatingthe argon laser used in the experiment to Dr E. M.Kohner of the Hammersmith Hospital, London,England. We also thank Mr N. S. C. Rice and DrKohner for their advice, and the Department of Audio-Visual Communications, Institute of Ophthalmology,London, for preparing the figures.

References

CAMPBELL, C. J., RITTLER, M. C., SWOPE, C. H., and WALLACE, R. A. (I969) Amer. Jt. Ophthal., 67, 671CHERRY, P. M. H., FAULKNER, J. D., SHAVER, R. P., WISE, J. B., and WITTER, S. L. (,973) Ann. Ophthal., 5, 9IIGEERAETS, W. J., WILLIAM, R. C., CHAN, G., HAM, W. T. JR., GUERRY, D. III, and SCHMIDT, F. H. (I962)

Invest. Ophthal., I, 340JONES, B. R. (1973) 'Summing Up Present Knowledge and Problems of Corneal Graft Failure', Ciba

Foundation Symposium I5 (NS), p. 352. Elsevier, Excerpta Medica, North-Holland, AmsterdamMEYER-SCHWICKERATH, G. (i96o) 'Light Coagulation', p. i6. Mosby, St LouisUNGER, W. G., PERKINS, E. S., and BASS, M. S. (1974) Exp. Eye Res., 19, 367

._m 0 :.- -


http://bjo.bmj.com

/B

r J Ophthalm


nloaded from

http://bjo.bmj.com/

corneal neovascularization argon · treatment of the vessel between these two points. intravenous...

Documents