British Journal of Ophthalmology 1995; 79: 1083-1087

Diode laser contact transscleral retinalphotocoagulation: a clinical study

Dominic A McHugh, Steven Schwartz, Jonathan G Dowler, Michael Ulbig, RolfK Blach,Peter A Hamilton

Moorfields EyeHospital, LondonD A McHughS SchwartzJ G DowlerM UlbigR K BlachP A Hamilton

Correspondence to:Mr Dominic McHugh,King's College Hospital,Denmark Hill, LondonSE5 9RS.Accepted for publication8 August 1995

AbstractAim-To examine the clinical efficacy ofcontact transscleral retinal photocoagula-tion with a diode laser.Methods -Transscleral retinal photo-coagulation was performed on 36 eyes. Theconditions treated included peripheralretinal breaks associated with retinaldetachments (30 eyes) and giant retinaltears (six eyes). Of the 30 eyes withretinal detachments, 28 underwent trans-scleral photocoagulation to the site ofdrainage of subretinal fluid in an attemptto reduce the risk ofhaemorrhage.Results-Threshold lesions were obtainedwith irradiances of between 95 4 W/cm2and 191 W/cm2. Satisfactory chorioretinaladhesion was achieved in all eyes withretinal breaks and giant retinal tears. Theonly significant complications of treat-ment encountered were punctatechoroidal haemorrhages (three eyes).Drainage related choroidal haemor-rhage following earlier photocoagulationoccurred in two of 28 eyes.Conclusions-This study confirms theclinical potential of transscleral diodelaser photocoagulation in the therapy ofsurgical retinal conditions.(BrJ7 Ophthalmol 1995; 79: 1083-1087)

Semiconductor diode lasers (wavelength ofemission, 790-840 nm, infrared) have beenemployed in ophthalmic therapy since 1988.Clinical trials have demonstrated their com-parable efficacy with argon lasers in the

Figure 1 Transscleral delivery of diode laser radiation using a specialised probe. Thetarget area and the transilluminated aiming beam are visualised with a 20 D lens.

treatment of retinal vascular conditions andglaucoma.1-3 The portability and reliability ofdiode lasers confer ergonomic advantages overion lasers; and the infrared emission wave-length has the biophysical advantages of goodtransmission through haemorrhagic mediaopacities and nuclear sclerotic cataract,together with negligible absorption withinmacular xanthophyll.45The initial delivery systems employed

transpupillary irradiation of the target zone(in conjunction with, for example, a slit-lampmicroscope). More recently, however, the highscleral transmission of near infrared radiationhas prompted interest in the employment ofdiode lasers for transscleral therapy. Pilotstudies have reported the effectiveness of diodelaser cyclophotocoagulation in the treatmentof refractory glaucoma and histological workand a pilot clinical study have indicatedthe potential of transscleral retinal therapy('retinopexy').68 Stated advantages include theability to bypass media opacities, which wouldotherwise attenuate the treatment beam, andalso less marked inflammatory side effects incomparison with transscleral cryotherapy.The aims of the current study were to

examine the potential range of applications oftransscleral diode laser retinopexy in the treat-ment of surgical retinal conditions, to optimisetreatment variables, and to assess whether thisnovel modality had significant advantages overconventional techniques.

Materials and methods

LASERAll treatments were performed using a semi-conductor diode laser, with an emission wave-length of 810 nm (Iris Medical InstrumentsInc, Mountain View, CA, USA). The maxi-mum available power was 2 W and the expo-sure duration could be incrementally increasedup to a maximum of 9 seconds. Laser deliverywas accomplished via a transscleral probe,which was applied to the external scleralsurface (Fig 1).The probe tip incorporated a 2 X 2 mm

glass prism, which had a 1P5 mm diameteroutput face, to deliver laser energy perpendic-ular to the device handle and shaft. The faceof this prism protruded 0-8 mm above theshaft shank, permitting indentation at thetreatment site. The target area was visualisedwith an indirect ophthalmoscope while atransilluminating aiming beam from a redemitting diode laser facilitated precise place-ment of the probe over the target area.

1083

on June 10, 2021 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.79.12.1083 on 1 Decem

ber 1995. Dow

nloaded from

http://bjo.bmj.com/

McHugh, Schwartz, Dowler, Ulbig, Blach, Hamilton

Figure 2 Acute burns produced by transscleral diode lasertreatment of a peripheral retinal break. Note the grey-whiteappearance of the lesions, which is typical of irradiationwith a wavelength of 810 nm.

Optics inside the probe produced an aerialimage of the laser fibre optic, 0 9-1 *0 mm indiameter at the 'ideal' treatment plane located0 5 mm from the output face of the prism.Residual beam divergence near the prism facewas only approximately 80 mrad full angle.This means that within 1 mm from the outputface of the prism, the laser illuminated spotdid not change by more than 0 1 mm.A foot switch activated the laser and in all

cases the desired visible end point was the grey-ish-white appearance typical of a thresholddiode laser lesion. The strong forward scatteringcharacteristics of sclera imply that the diameterof the treatment beam was not significantlyincreased during its passage through the sclera.

PATIENTSA total of 36 eyes in 36 patients underwentdiode laser retinopexy. The mean age was 47(range 18-45) years and the mean period ofreview was 8 (range 4-12) months. There werea number of indications for transscleral diodelaser retinopexy:

Figure 3 (A) Six monthsafter treatment ofa retinalbreak during retinaldetachment surgery, thediode laser lesions havebecome pigmented scars(arrowed). (B) In the sameeye, prophylacticcryotherapy had beenapplied 2 years previously.Note the more extensivepigment epithelial atrophycompared with the laserscars.

(a) Retinal tears of up to 2 clock hours inextent, which were associated with retinaldetachment (30 eyes). After drainage of sub-retinal fluid (where appropriate), each breakwas surrounded by two rows of laser bums.This was then followed by application of a scle-ral explant and, if necessary, air or gas injec-tion.

(b) Irradiation of the site of drainage of sub-retinal fluid. In those cases of retinal detach-ment which required drainage (28 eyes), thetechnique of external suture needle drainagewas applied.9 With the aim of inducingchoroidal vascular closure and thus reducingthe risk of drainage related haemorrhage, apattern of 16 exposures was applied to thelocation selected for drainage.

(c) Giant retinal tears (six eyes). Tears of 3clock hours or more in extent were treated withvitrectomy and silicone oil exchange. A doublerow of diode laser burns was administered tothe posterior edge of the break and the 'horns'of the tear.During treatments for these conditions, the

initial power setting was 500 mW and thetarget area irradiated until a reaction wasobserved. The power was increased in 250mW increments if a lesion was not observedafter an exposure of up to 2-3 seconds. Thepower was decreased following an intenseblanching reaction, an audible 'popcorn'effect, or a haemorrhagic lesion. All procedureswere performed subconjunctivally underperibulbar or general anaesthesia. Careful notewas made of any peroperative or subsequentside effects of treatment. Patient review was at2 weeks, 6 weeks, and 12 weeks after therapy,then at 3 monthly intervals.

ResultsThe irradiances required for threshold lesionsranged between 95 4 W/cm2 and 191 W/cm2(mean 133 (SD 27.20) W/cm2), with treat-ment variables 0-75-1.5 W power and 1-4second pulse duration. Fundus pigmentationinfluenced the treatment variables selected,the two non-white patients (who wereAsian-Indian) requiring lower irradiances than

Fig 3B

1084

iF

Diode laser contact transscleral retinal photocoagulation: a clinical study

the white patients (95 4 W/cm2; power 0 75Wand pulse duration 1 second). Higher powerswere required in the presence of subretinalfluid or haemorrhage. Other relevant factorsincluded the need to place the probe preciselyperpendicular to the scleral surface (which wasindicated by how sharply defined the aimingbeam appeared), and whether treatment wasbeing applied through an extraocular muscle.Retinal blanching could usually be seen even incases where the fundal view was obscured bysubretinal fluid, cataract, or blood. It wasnoticeable that even extensive treatments didnot induce chemosis, or lid oedema, as is com-monly observed after cryotherapy.

In those eyes with retinal tears and retinaldetachments, effective chorioretinal lesionswere produced with transscleral therapy (Fig2). Retinal reattachment was initially achievedin all the eyes, although the subsequent forma-tion of new breaks required a second (non-diode) vitrectomy procedure in two eyes. Inneither case was this thought to be a diodelaser related complication. Punctate haemor-rhages were observed at the centre of severallesions in three eyes, but reduction in powereliminated this effect in subsequent exposures.

Retinal burns 12 weeks after treatmentappeared as pigmented scars. One eye hadundergone prophylactic cryotherapy severalyears previously. In contrast with diode laserscars there was a greater amount of chorio-retinal atrophy associated with the cryopexyscar (Fig 3). Retinopexy before transscleralsuture needle drainage of subretinal fluid wasfollowed by drainage induced haemorrhage of2-3 disc diameters in area in two of 28 eyes.No drainage related complications wereobserved in the remaining eyes (Fig 4).

Retinal reattachment was achieved in all ofthe eyes treated for giant retinal tears. Onestriking feature was the mild degree ofpostoperative inflammation. The relativelyextensive photocoagulation required inducedrelatively little intravitreal pigment dispersionor uveitis, in comparison with that commonlyseen with extensive retinal cryotherapy. In oneeye an excessively high power induced severalpunctate choroidal haemorrhages, but thesehad resolved by 2 weeks postoperatively.

DiscussionThe results of this study suggest the potentialof transscleral diode laser therapy in the treat-ment of surgical conditions of the retina andthat there may be significant advantages overtechniques currently employed.The principal aims of retinal reattachment

surgery are to appose the neurosensory retinato the underlying retinal pigment epitheliumand create a chorioretinal adhesion around allretinal breaks. The three principal methodsavailable to produce a chorioretinal scar arediathermy, cryotherapy, and photocoagula-tion. The ideal requirement of a transscleraldelivery system is the ability to induce afocal tissue effect that is limited to the outerretina and inner choroid, with minimal damageto the inner retina, outer choroid, or sclera.

Fig 4C

Figure 4 (A) Transscleral diode laser retinopexy to thearea selectedfor drainage ofsubretinalfluid. The aim wasto decrease the risk ofdrainage related choroidalhaemorrhage. (B) Drainage ofsubretinalfluid using asuture needle at the site of earlier photocoagulation.(C) Diode laser scars in the site ofdrainage ofsubretinalfluid (arrowed).

;, w:. R

i...,,. 's w

#* i,.::*:S't \

k.:Mb-.

w ,hElS

Fig 4A

Fig 4B

1085

on June 10, 2021 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.79.12.1083 on 1 Decem

ber 1995. Dow

nloaded from

http://bjo.bmj.com/

McHugh, Schwartz, Dowler, Ulbig, Blach, Hamilton

Table 1 Transmission and absorption characteristics of visible and infrared wavelengthswithin ocular tissues

Argon green Krypton red Diode Nd:YAG(514-5 nm) (647 nm) (810 nm) (1064 nm)

Scleral transmission (%)1314Non-contact 12 18 35 53Indentation 33 43 71 77

Absorption (O/o)Melanin* 72 60 35 12

*Choroid and retinal pigment epithelium.15

Before the introduction of cryotherapy,diathermy was the preferred method of induc-ing a chorioretinal adhesion in the treatment ofretinal breaks.'0 The physical principles ofdiathermy involve the generation of radio-frequency energy that is converted to heat asit propagates through biological tissue. It is,therefore, relatively non-selective in its site ofaction and, in addition to producing a chorio-retinal lesion, also causes scleral destructionand weakening and extensive choroidaldestruction, with damage to vortex veins or theposterior ciliary arteries or nerves.

Cryotherapy, which produces an inflamma-tory reaction and thus chorioretinal adhesionby tissue freezing, has supplanted diathermy inthe transscleral treatment of retinal breaks. Itproduces little scleral damage" but a similarstrength of chorioretinal adhesion. Cryo-therapy, however, does have a number ofpotentially serious side effects, which include amarked breakdown in the integrity of theblood-retinal barrier and dispersion of retinalpigment epithelial cells into the vitreous cavity.The combination of serum derived factors andpigment epithelial cells may increase the risk ofproliferative vitreoretinopathy (PVR) and offailure of retinal detachment surgery.'2The essential mechanism of action of retinal

laser photocoagulation is the absorption ofradiant energy by a chromophore, melanin,within the retinal pigment epithelium andchoroid and its conversion to heat, with sub-sequent tissue coagulation. The initial coagu-lation, followed by secondary gliosis, accountsfor the adhesive effect of lesions. The majorityof lasers that are currently employed for thispurpose emit radiation in the visible part ofthe electromagnetic spectrum - for example,the argon laser (emission wavelength488-514-5 nm, blue-green). They have poortransmission properties through the sclera andtheir use is confined to transpupillary orintraocular systems of delivery. They are,therefore, of more limited effectiveness in thepresence of media opacities or very peripheralpathology.The recent introduction of lasers which emit

in the near infrared region of the spectrum hasallowed consideration of alternative deliverysystems. The radiant emissions ofthe Nd:YAGlaser (1064 nm) and the semiconductor diodelaser (810 nm) have a combination of highscleral transmission and significant absorptionwithin melanin. In contrast, several studieshave demonstrated that argon laser irradiationis poorly transmitted through the sclera, evenwith indentation. Therefore, despite its higherabsorption within melanin than the infrared

wavelengths, it has not been considered appro-priate for transscleral applications (Table 1,refs 13-15). In relation to the diode laser, his-tological studies of transscleral irradiationdemonstrated its ability to produce both ciliarybody and chorioretinal lesions, but with noevidence of thermal damage to the sclera.7 16High transmission through silicone explantsmay also be of advantage in the retreatment ofeyes with retinal breaks (Nanda et al; posterpresentation, ARVO, 1993).

There is also evidence that diode laserretinopexy induces less pigment epithelial celldispersion than cryotherapy, with importantimplications for the relative risks of induction ofPVR following each form of therapy (Nissenet al; poster presentation, ARVO, 1993).Another study specifically compared blood-retinal barrier breakdown in rabbits induced bytransscleral diode laser irradiation and retinalcryotherapy. Magnetic resonance imagingwith gadolinium-DPTA enhancement of equalareas of treated retina demonstrated a twofoldincrease in signal enhancement in eyes treatedwith cryotherapy, compared with diode lasertreated eyes. The difference in signal enhance-ment was correlated directly with change in per-meability of the blood-retinal barrier (Arrindellet al; poster presentation, ARVO, 1992).

In a pilot study, Haller and colleaguesreported their success in employing transscleraldiode laser retinopexy in retinal detachmentsurgery.8 Lesions were produced using energieswhich ranged between 234 and 1492 mJ (whitepatients) and 74 and 251 mJ (black patients),the differential energies being explained bypigmentary variation. Although generallyhigher radiant energies were employed in thepresent series, this may be explained by individ-ual differences in fundal pigmentation, or thepresence of retinal oedema, or haemorrhage.The complication rate reported by Haller

was low, being confined to pinpoint choroidalhaemorrhages within lesions in three eyes andfocal scleral discoloration in two eyes. Theseincidents were both attributed to employmentof an excessively higher power. Despite thelimited sample sizes in that and the currentstudy, the low incidence of PVR followingtransscleral diode laser therapy is encourag-ing. In particular, the management of giantretinal tears requires relatively extensive areasof treatment and therefore the employment ofa transscleral laser in preference to cryo-therapy may minimise intravitreal dispersionof the cellular and humoral precursors ofPVR.The use of suture needle drainage of subreti-

nal fluid obviates the risk of retinal incarcera-tion that may follow scleral dissection andchoroidal perforation. The inability to directlycauterise the choroid does introduce the risk ofinducing a choroidal haemorrhage. A recentanimal study demonstrated the efficacy oftransscleral diode laser retinopexy in achievingchoroidal vascular stasis sufficient to preventhaemorrhage following passage of a 20 gaugeneedle through the irradiated areas (Kaplanet al; poster presentation, ARVO, 1993). Thelow incidence of haemorrhagic complications

1086

on June 10, 2021 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.79.12.1083 on 1 Decem

ber 1995. Dow

nloaded from

http://bjo.bmj.com/

Diode laser contact transscleral retinal photocoagulation: a clinical study

of drainage in the present study tends to con-firm these findings. The procedure is easy toperform, the only potential difficulty being invisualisation of choroidal lesions in thepresence of deep or turbid subretinal fluid.The successful deployment of the transscle-

ral probe requires awareness of its differencesfrom retinal cryotherapy. It is important tomaintain the probe tip perpendicular to thesclera while indenting in order to maximisescleral transmission, and the aiming beam mustbe in sharp focus. In contrast with cryotherapy,laser photocoagulation is sensitive to variationsin fundal pigmentation and the smaller areas oflaser lesions does increase the treatment timecompared with cryotherapy. Localisation ofsmall retinal breaks with cryotherapy isenhanced by the contrast of retinal whiteningseen around the hole, an effect more difficult toachieve with a laser. However, although theimmediate cryotherapy reaction tends to fade, alaser lesion remains clearly visible, and thus therisks of overtreatment are reduced.

This study has given an indication of thepotential role of transscleral diode laserretinopexy in the treatment of surgical retinalconditions. The low complication rate and theearly indications of clinical efficacy providejustification for the establishment of formaltrials of its use in individual conditions.Each author states that he has no proprietary interest in thedevelopment or marketing of this or a competing instrument.

1 McHugh JDA, Marshall J, ffytche TJ, Hamilton PAM,Raven A, Keeler CR. Initial clinical experience using a

diode laser in the treatment of retinal vascular disease. Eye1989; 3: 516-27.

2 Balles MW, Puliafito CA, D'Amico DJ, Jacobson JJ,Bimgruber P. Semiconductor diode laser photocoagula-tion in retinal vascular disease. Ophthalmology 1990; 97:1553-61.

3 McHugh D, Marshall J, ffytche TJ, Hamilton PAM, RavenA. Diode laser trabeculoplasty (DLT) for primary openangle glaucoma and ocular hypertension. BrJy Ophthalmol1990; 74: 743-7.

4 Horeckler BL. The absorption spectrum of haemoglobinand its derivatives in the visible and near infrared regions.Y Biol Chem 1943; 148: 173-8.

5 Nussbaum JJ, Pruett RC, Delori RC. Macular yellow pig-ment: the first 200 years. Retina 1981; 1: 296-310.

6 Gaasterland DE, Pollack IP. Initial experience with a newmethod of laser transscleral cyclophotocoagulation for cil-iary ablation in severe glaucoma. Trans Am OphthalmolSoc 1992; 90: 225-43.

7 Jennings T, Fuller T, Vulich J, Lam T, Joondeph BC,Ticho B, et al. Transscleral contact retinal photocoagula-tion with an 810 nm semiconductor diode laser.Ophthalmic Surg 1990; 21: 492-6.

8 Haller J, Lirn J, Goldberg M. Pilot trial of transscleral diodelaser retinopexy in retinal detachment surgery. ArchOphthalmol 1993; 111: 952-6.

9 Raymond GL, Lavin MJ, Dodd CL, McLeod D. Sutureneedle drainage of subretinal fluid. Bry Ophthalmol 1993;77: 428-9.

10 Schepens CL. Retinal detachment and aUlied diseases.Philadelphia: WB Saunders Co, 1983: 289-99.

11 Steinmetz CG, Im. Reaction of the sclera to cryosurgery. IntOphthalmol Clin 1967; 7: 395.

12 Glaser BM, Vidaurri-LealJ, Michels RG, Campochiaro PA.Cryotherapy during surgery for giant retinal tears andintravitreal dispersion of viable RPE cells. Ophthalmology1993; 100: 466-70.

13 Rol P, Niederer P, Duff U, Henchoz PD, Fanlkhauser F.Experimental investigations on the light scattering proper-ties of the human sclera. Lasers Light Ophthalmol 1990; 3:201-12.

14 Vogel A, Dlugos C, Nuffer R, Birngruber R. Opticalproperties of human sclera and their consequences fortransscleral laser applications. Lasers Surg Med 1991; 11:331-40.

15 Geeraets WJ, Berry BS. Ocular spectral characteristics asrelated to hazards from lasers and other light sources. AmJ Ophthalmol 1968; 66: 15-20.

16 Hennis HL, Assia E, Stewart WC, Legler UFC, Apple DJ.Transscleral cyclophotocoagulation using a semiconduc-tor diode laser in cadaver eyes. Ophthalmology 1991; 21:274-8.

1087

on June 10, 2021 by guest. Protected by copyright.

http://bjo.bmj.com

/B

r J Ophthalm

ol: first published as 10.1136/bjo.79.12.1083 on 1 Decem

ber 1995. Dow

nloaded from

http://bjo.bmj.com/