Subconjunctival THC:YAG (“Holmium”) Laser Thermal Sclerostomy Ab Externo

Subconjunctival THe: YAG ("Holmium") Laser Thermal Sclerostomy Ab Externo A One,year Report Andrew G. Iwach, MD,1,2 H. Dunbar Hoskins, Jr., MD,1.2 Michael V. Drake, MD, 2 Christopher J. Dickens, MD1 Background: Laser sclerostomy can be performed in a less-invasive manner than standard filtering surgery. Longer wavelengths in the infrared range have water-absorptive characteristics that facilitate perforation of the sclera. The goal was to perform laser sclerostomy ab externo to avoid intraocular instrumentation and minimize conjunctival trauma. Methods: A thulium, holmium, chromium-doped:YAG (THC:YAG) crystal laser was used to create thermal sclerostomies in 49 glaucomatous eyes of 46 patients. The laser is a long-pulsed (300-tlsecond), compact, self-contained, solid-state laser operating in the near infrared (2.1 tlm). Energy was delivered via a specially designed 22-gauge (712-tlm) optic probe that emits energy at a right angle to the long axis of the fiber. Pulse energies of 80 to 120 mJ were used. Total energy levels to produce full-thickness sclerostomies ranged from 1.4 to 7.2 J. Subconjunctival5-fluorouracil (5-FU) injections were administered in 46 eyes. Success was defined as an intraocular pressure (lOP) of less than or equal to 22 mmHg with or without medications. For eyes in which preoperative lOP was less than or equal to 22 mmHg, success was defined as a decrease in lOP of greater than or equal to 30%. Results: Estimated probability of success allowing for one retreatment was 0.75 at 6 months and 0.68 at 12 months. Mean lOP of successful cases was 13.3 mmHg at both 6 and 12 months. Twelve cases failed within the initial 6 months, and two additional cases failed by 12 months. Conclusion: THC:YAG ("holmium") laser thermal sclerostomy is an alternative to other full-thickness filtration procedures. Further evaluation and understanding will define its ultimate role in glaucoma management. Ophthalmology 1993; 100: 356-366

Originally received: October 13, 1991. Revision accepted: September 4, 1992. I

2

Foundation for Glaucoma Research, San Francisco. Department of Ophthalmology, University of California, San Francisco.

Presented at the American Academy of Ophthalmology Annual Meeting, Anaheim, October 1991. Supported by the Foundation for Glaucoma Research, San Francisco, California. Drs. Hoskins and Iwach are consultants to and have a financial interest in Sunrise Technologies, Inc., Fremont, California. Reprint requests to Andrew G. Iwach, MD, Foundation for Glaucoma Research, 490 POst St, Suite 830, San Francisco, CA 94102.

356

The use oflasers for glaucoma filtration surgery continues to receive much attention. Different laser systems and techniques with varying success rates have been reported. I - 19 Laser sclerostomy can be performed in a lessinvasive manner than standard filtering surgery. With some techniques, it can be performed in a clinic or minor surgery room, thus removing filtering surgery from the standard operating theater. Various wavelengths with different properties and different tissue interaction have been used to create filtering sclerostomies. Longer wavelengths in the infrared range (THC:Y AG ["holmium"], TC:YAG, Er:YAG, carbon

Iwach et al . Subconjunctival THC:YAG Laser Sclerostomy dioxide) have water-absorptive characteristics that facilitate perforation of sclera. Shorter wavelength lasers use scleral dyes to enhance energy absorption. 13 Sclerostomy by photodisruption can be performed with Q-switched neodymium:YAG (Nd:Y AG) lasers, 15.16 and sclerostomy by photoablation has been performed with excimer lasers. Laser energy may be delivered by mirrored contact lenses to the internal face of the filtration angle or by fiberoptic cables for ab interno or ab extemo sclerostomy formation. Our goal is to perform laser sclerostomy ab externo to avoid intraocular instrumentation and minimize conjunctival trauma. This article is an update of a previously published report of the 6-month results of an ongoing clinical study of subconjunctival THC:YAG laser sclerostomy.9

Materials and Methods The THC: YAG Laser and Fiberoptic Probe The Model gLase 210 holmium laser (Sunrise Technologies, Inc., Fremont, CA), achromium-sensitized thulium, holmium-doped YAG laser is a long-pulsed (300-Jlsecond), compact, self-contained, solid-state laser operating in the near infrared (2.1 Jlm). The laser delivers energy ranging from 50 to 500 mJ per pulse at a pulse rate of five pulses per second. A helium-neon aiming beam is used for orientation. Laser energy is delivered via a 200-Jlm-diameter quartz fiberoptic probe in a protective casing, giving a total outer diameter of 712 Jlm (22-gauge diameter). The tip of this fiberoptic probe is manufactured so that energy is transmitted at approximately 90 0 to the fiber axis. The fiberoptic cable inserts into a hand piece for delivery to the eye.

Patient Selection The patients selected for this study had uncontrolled glaucoma on maximum-tolerated medications, many associated with either aphakia, pseudophakia, congenital glaucoma, iridocorneal endothelial syndrome, inflammatory glaucoma, or elevated episcleral venous pressure. Fifty-eight procedures were performed in 49 eyes of 46 patients. Twenty-two phakic, 19 pseudophakic, and 8 aphakic eyes have been followed for up to 12 months. Patients ranged in age from 9 to 90 years (mean, 61 years). All patients gave informed consent. The consent form and clinical protocol were approved by the investigational review board of St. Mary's Hospital, San Francisco, California. This study is ongoing. The information in this report is the 3-, 6-, 9-, and 12-month follow-up data on all patients who have been followed 3, 6, 9, or 12 months since their last THC:YAG laser treatment. Estimated success rates were calculated by actuarial method. 2o The protocol provides for one retreatment if needed per the surgeon's judgment.

To better understand the influence of preoperative ocular risk factors for filtration surgery failure, each eye was assigned to a risk of failure category according to the following algorithm. One point was assigned for each of the following risk factors present: (I) previous filtration surgery; (2) aphakia or pseudophakia; (3) patient age less than 31 years; (4) ocular inflammation at time of surgery; and (5) especially high-risk diagnosis (e.g., iridocorneal endothelial syndrome). We compared the risk value of successful cases versus failures with a Wilcoxon test 21 to determine if successful cases had a lower risk. We examined the success rates of initial laser sclerostomy versus risk factor of the 6-month cohort. We separately examined success rates of eyes assigned either a 0 or I risk point, and estimated success rates were calculated by the actuarial method. 20 To determine whether additional clinical experience influenced our results, we compared results of initial laser sclerostomies performed the first 7 months versus those done thereafter. We examined eyes assigned a risk factor of 2 or 3 within the 3- and 6-month cohorts.

Success Criteria Eyes met criteria for success if lOP was less than or equal to 22 mmHg at the specified follow-up (3, 6, 9, or 12 months) from the initial or retreatment date, with or without medications. For eyes in which preoperative lOP was less than or equal to 22 mmHg, success was defined as a decrease in lOP of greater than or equal to 30%. Before laser sclerostomy, all patients were using antiglaucoma medications.

Procedure A detailed description of the surgical technique has been reported. 9 In brief, of the 49 eyes reported in this study, 31 were treated in an operating room and 18 were treated in a minor surgery room in an eye clinic. Of the most recent 20 procedures, 75% were performed in a minor surgery room of an eye clinic. Retrobulbar anesthesia was administered in all but one case using a 1.5- to 2.5-ml injection of a I: I mixture of 2.0% lidocaine and 0.75% bupivacaine (Marcaine). One patient was given only topical 0.5% proparacaine (Alcaine) and 4% tetracaine. After the preoperative preparation, the patient was draped in the usual sterile manner and positioned at the operating microscope. A 5-0 silk superior rectus traction suture was placed at the surgeon's discretion. A I-mm conjunctival stab incision was made 12 to 15 mm away from the intended sclerostomy site to allow entry of the fiberoptic probe. The probe was then gently advanced subconjunctivally and placed tangential to the limbus. Probe insertion produced minimal disturbance of the conjunctiva. Conjunctival bleeding was limited to the conjunctival incision site and readily controlled with cautery. The helium-neon aiming beam was directed perpendicular to the limbus and could be seen in the anterior chamber as the sclerostomy was created. Exposures were made with energy per pulse ranging from 80 to 120

357

Ophthalmology

Volume 100, Number 3, March 1993

mJ and a repetition rate of 5 pulses/second. The probe was gently retracted, and a large, diffuse bleb developed immediately in all treated eyes. When necessary, the conjunctival incision was closed with a 10-0 nylon or prolene suture. Total energy levels to produce full-thickness sclerostomies ranged from 1.4 to 7.2 J. Total energies required to create sclerostomies were influenced by factors not completely understood but include pre-existing limbal scarring, probe positioning and orientation, and surgeon's experience. In some aphakic and pseudophakic eyes, the laser energy was continued after perforation of the sclera to produce either a local iridoplasty or an iridectomy. Intraoperative viscoelastic material was introduced via the sclerostomy with a Rycroft needle to reduce iris prolapse in some phakic patients. No viscoelastic was used postoperatively. Two eyes sustained conjunctival burns after 40 pulses of 100 mJ/pulse output. These burns occurred in eyes of elderly patients with thin conjunctivae. A sclerostomy had been created beneath the conjunctival burn, and the conjunctiva was closed with a 10-0 prolene suture. In nine eyes, a second laser sclerostomy was created using the same technique as described above. The failed sclerostomy had been located at the superior limbus in all nine cases. Eight of these nine repeat sclerostomies were created immediately adjacent to the failed sclerostomy site. One repeat sclerostomy was created inferiorly. In seven initially successful patients, there was gradual shrinkage of the bleb with an accompanied increase in lOP. We were able to reverse the course and obtain sustained low lOPs in four patients by using Nd:YAG laser energy delivered via a gonioscopy lens to the internal aspect of the sclerostomy site. Power settings of 6 to 10 mJ with multiple applications were used. The energy was focused deep into the internal ostium and resulted in immediate elevation of the bleb and lowering of the lOP in four of these cases. Patient Follow-up Postoperatively, patients were treated with neomycin and polymyxin B sulfates and dexamethasone (Maxitrol), one drop every 4 hours. Postoperative incarcerated iris was dislodged by applying limbal indentation with either a Zeiss gonioprism (Ocular Instruments, Inc, Belleville, W A) or with an argon laser using no more than 30 applications of a 50-~m spot size and at a 600- to 1500-mW power setting (duration, 0.1-0.2 seconds). Two phakic patients with low lOPs, diffuse blebs, and shallow anterior chambers were given an extended-wear contact lens (Soft Mate II [Barnes-Hind, Sunnyvale, CAl; base curve, 9.0 mm; diameter, 14.8 mm; power, -3.00 diopters). Within 20 minutes of fitting, the anterior chamber deepened. The contact lens was replaced every 2 days and discontinued by postoperative day 7. Subconjunctival 5-fluorouracil (5-FU) injections were administered in 46 eyes. The total number of injections per procedure ranged from 1 to 21 and were administered according to tpe treating physician's judgment rather than

358

by a predetermined protocol. Eyes treated later in the study received more injections. 5-FU injections were terminated if there was significant corneal epithelial breakdown. Topical antibiotics were discontinued by 4 weeks. The corticosteroid was reduced to twice daily after 4 weeks. One phakic patient continues to receive tropicamide (Mydriacyl) twice daily at 6 months to prevent late iris reincarceration. Intraocular pressures and results of ocular examination were recorded at approximately 1, 3, 7, 14, 30, 42, 90, 180, 270, and 360 days after treatment.

Results Ab externo thermal laser sclerostomies were performed in 49 glaucomatous eyes of 46 patients (Table 1). A total of 58 procedures were performed. All sclerostomies were initially patent. Nine eyes subsequently required retreatments, and three patients had both eyes treated. Of the 49 treated eyes, 37 were successful at 3 months after one procedure. An additional four eyes were successful 3 months after retreatment. Eight of the 49 eyes had failed by 3 months (Table 2). The eight failed eyes either had lOPs greater than 22 mmHg (range, 28-38 mmHg) or underwent other forms of glaucoma surgery. Of the 40 eyes that had been followed for 6 months, 24 were successful at 6 months after one procedure. An additional four eyes were successful 6 months after retreatment. Four eyes that were successful at 3 months failed by 6 months. One eye that was successful at 3 months missed the 6-month visit and is not classified. Of the 31 eyes that had been followed for 9 months, 16 were successful at 6 months after one procedure. An additional four eyes were successful 9 months after retreatment. One eye that was successful at 6 months failed by 9 months. Of the 26 eyes that had been followed for 12 months, 12 were successful at 12 months after one procedure. An additional three eyes were successful 12 months after retreatment. One eye that was successful at 9 months failed by 12 months. Estimated success rates allowing one sclerostomy calculated by using the actuarial method are listed in Table 3 and plotted in Figure 1. Estimated success rates including both initial sclerostomy and eyes requiring a repeat sclerostomy are listed in Table 4 and plotted in Figure 1. Diffuse blebs were seen in all patients initially, and blebs were sustained in successful patients (Figures 2-5). Some eyes have been followed for as long as 18 months. The preoperative and postoperative mean lOPs for the Table 1. Individual Cohort Composition Profiles No. No. No. No. No.

of procedures of eyes of patients of retreatments of bilateral cases

3Mos

6Mos

9Mos

12 Mos

58 49 46 9 3

47 40 38 7 2

38 31 29 7 2

33 26 24 7 2

Iwach et al . Subconjunctival THC:YAG Laser Sclerostomy

Table 2. Success· and Failures of Eyes Maturing to Specific Three-Month Intervalst No. of eyes in cohort Successful Failure Missing data No. of retreatments

3 Mos

6 Mos

9 Mos

12 Mos

49 41

40

31 20 11

26 15 11

7

7

28 11 1

8

o

7

9

o

o

• Defined as an intraocular pressure of no more than 22 mmHg from the initial or retreatment date, with or without medications. Each eye is a defined data point. For eyes in which preoperative intraocular pressure was no more than 22 mmHg, success was defined as a decrease in intraocular pressure of at least 30%. One retreatment is allowed.

t Each time interval includes all eyes that have matured to that time interval.

cohorts are shown in Table 5. The mean lOP for successful eyes at 3 months was 13.7 ± 4.7 mmHg, at 6 months was 13.3 ± 4.8 mmHg, at 9 months was 12.8 ± 5.5 mmHg, and at 12 months was 13.3 ± 5.5 mmHg. Of eyes meeting the lOP success criteria, 12% were treated with antiglaucoma medications at 3 months, 21 % at 6 months, 40% at 9 months, and 47% at 12 months. Before laser sclerostomy, all eyes had been treated with antiglaucoma medications. Our study of the influence of preoperative ocular risk factors for initial laser sclerostomy failure is summarized in Table 6. The Wilcoxon test showed that failed patients had a statistically significant higher median risk factor as compared with successful patients (P = 0.035).1 1 Estimated success rates for initial laser sclerostomy in "lowrisk" eyes (either risk factor 0 or I) were calculated by using the actuarial method and are listed in Table 7 and plotted in Figure 6. Our study of the effect of the surgeon's holmium laser sclerostomy clinical experience suggests success rates increase as clinical experience increases (Table 8). However, no statistically significant difference (P = 0.13) was found using a Fisher's exact test. 21

Table 4. Estimated Success· Rates Allowing Two Sclerostomies Time Period (mos)

Estimated Success Rate

Standard Error

0-3 3-6 6-9 9-12

0.84 0.75 0.72 0.68

0.05 0.06 0.07 0.08

• Defined as an intraocular pressure of no more than 22 mmHg from the initial treatment or retreatment date, with or without medications. Each eye is a defined data point. For eyes in which preoperative intraocular pressure was no more than 22 mmHg, success was defined as a decrease in intraocular pressure of at least 30%. One retreatment is allowed.

A total of 14 eyes have failed. Eight eyes failed in the first 3 months, four additional eyes failed by 6 months one additional eye failed by 9 months, and one additionai e~e fail:d by 12 months. The circumstances concerning eIght faIlures have been previously described. 9 In one of these eight eyes (the eye associated with elevated episcleral venous pressure secondary to an arteriovenous shunt), !hough rep?rted as a failure by I-month postoperatively 10 the prevIOUS report, an obvious bleb and an lOP of 12 mmHg without medications developed by 3 months. Thus, at 3 months a total of eight eyes were designated as failures in this report (Table 2). This is an ongoing study and not all eyes have matured to later cohorts, thus explaining the apparent inconsistencies in Table 2. For example, of the eight eyes that failed by 3 months, only seven are 6 months postoperative and, therefore, included in the 6-month cohort. With these seven eyes and the four that were successful at 3 months but that failed by 6 months, there were II failures in the 6-month cohort. Similar situations exist for 9- and 12month cohorts. 1.0

•

CJ)

w

Table 3. Estimated Success· Rates Allowing Only One Sclerostomy Time Period (mos)

0-3 3-6 6-9 9-12

Estimated Success Rate

0.76 0.64 0.58 0.54

Standard Error

0.06 0.07 0.08 0.08

• Defined as an intraocular pressure of no more than 22 mmHg from the initial treatment date, with or without medications. Each procedure is a defined data point. For eyes in which preoperative intraocular pressure was no more than 22 mmHg, success was defined as a decrease in intraocular pressure of at least 30%.

!;( 0.8 a: CJ) CJ)

w

()

•

0.6

•

()

:::>

CJ)

0

•

0.4

w

!;( ~

i= CJ) w

0.2 0.0

0

3

6

9

12

MONTHS Figure 1. Estimated success rates for initial laser sclerostomy and initial laser sclerostomy (one procedure [squares with dot)) plus eyes requiring a repeat sclerostomy (2 procedures [solid diamonds)).

359

Ophthalmology

Volume 100, Number 3, March 1993

[;]

(/)

w

~

0.8

a: (/) (/)

w

0 0

[;]

[;]

9

12

0.6

=>

(/)

Cl

w

0.4

~

::!!

i= w

(/)

0.2

0.0 0

360

3

6

Iwach et al . Subconjunctival THC:YAG Laser Sclerostomy Top left, Figure 2. Bleb remains functioning 8 months after surgery. Center left, Figure 3. Bleb remains functioning 12 months after surgery. Bottom left, Figure 4. Bleb remains diffuse and functioning 18 months after surgery. Top right, Figure 5. Bleb remains functioning 18 months after surgery. Center right, Figure 6. Estimated success rates for initial laser sclerostomy in low-risk eyes (squares with dot

=

one procedure).

Bottom right, Figure 7. Example of a diffuse bleb in an eye having failed multiple filtration surgeries. A diffuse bleb remains 5 months after surgery.

Table 5. Preoperative and Postoperative Mean Intraocular Pressures (in mmHg) of Successful* Eyes in the Cohort Group

No. of eyes Preoperative lOP Postoperative lOP

3Mos

6Mos

9Mos

12Mos

41 33.6 ± 12.3 13.7 ± 4.7

28 34.6 ± 12.3 13.3 ± 4.8

20 34.5 ± 12.3 12.8 ± 5.5

15 35.0 ± 12.9 13.3 ± 5.5

lOP = intraocular pressure. • Success is defined as an intraocular pressure of no more than 22 mmHg from the initial treatment date, with or without medications. Each eye is a defined data point. For eyes in which preoperative intraocular pressure was no more than 22 mmHg, success was defined as a decrease in intraocular pressure of at least 30%. One retreatment is allowed.

Of the seven new failures since our previous report, one aphakic eye, after congenital cataract extraction in childhood, had failed previous filtration surgery. Low lOP associated with a large bleb was initially seen, but vitreous incarcerated into the stoma. Subsequently, the eye underwent a successful vitrectomy and Molteno implantation without change in visual acuity. An additional aphakic eye with pseudophakic bullous keratopathy had vitreous incarcerate into the stoma and failed. One phakic eye with neovascular glaucoma failed within 1 week. One inflamed phakic eye that had failed previous filtration surgery failed by 3 months. One pseudophakic eye was successful at 3 months but failed by 6 months with gradual shrinkage of the bleb. Endophthalmitis developed 8 months postoperatively in another pseudophakic eye with an obvious bleb at 6 months (see below). Although at 9 months it was classified as a failure, the eye subsequently Table 6. Success Rates* of the Six-Month Cohort Allowing Only One Sclerostomy Risk Factors Success Failure Missing data Success rate (%)

oor 1

2

3 or 4

12

8

8

1 1 9Z

10

7

0

0

44

53

• Success is defined as an intraocular pressure of no more than 22 mmHg from the initial treatment date, with or without medications. Each procedure is a defined data point. For eyes in which preoperative intraocular pressure was no more than 22 mmHg, success was defined as a decrease in intraocular pressure of at least 30%.

has an lOP of 15 mmHg with an obvious bleb. One phakic Axenfeld's syndrome eye, though successful at 9 months, had gradual shrinkage of the bleb with failure by 12 months. Iris incarceration into the sclerostomy site did not occur in aphakic eyes. Iris incarceration did occur in 9 of 22 phakic eyes (Table 9), albeit with easy postoperative iris dislodgement in all but one eye. Preoperative laser iridotomy, intraoperative local holmium iridoplasty, the use of viscoelastics, and creation of two adjacent sclerostomies prevented iris incarceration in some, but not all, of our phakic patients. (Iris incarceration has been seen in one pseudophakic eye. The internal ostium was in the posterior trabecular meshwork. The iris was dislodged postoperatively with an argon laser, Table 7. Estimated Success Rates* for Eyes with a Risk Factor of 0 or 1 Allowing Only One Sclerostomy Time Period (mos)

Estimated Success Rate

Standard Error

0-3 3-6 6-9 9-12

1.0 0.87 0.77 0.77

0.0 0.09 0.12 0.12

No. of Successful Procedurest 17 17

11 6

• Success is defined as an intraocular pressure of no more than 22 mmHg ftom the initial treatment date, with or without medications. Each procedure is a defined data point. For eyes in which preoperative intraocular pressure was no more than 22 mmHg, success was defined as a decrease in intraocular pressure of at least 30%.

t

Entered at the beginning of each time interval.

361

Ophthalmology

Volume 100, Number 3, March 1993

Table 8. Surgeons' Early-Experience Versus LateExperience Cases with a Risk Factor of 2 or 3 Allowing Only One Sclerostomy 3Mos

6Mos

Initial Clinical Experience Success' Failure Success rate (%)

8

11 10

13 38

52 Later Clinical Experience

Success Failure Success rate (%)

7

12 3

2

80

78

• Success is defined as an intraocular pressure of no more than 22 mmHg from the initial treatment date, with or without medications. Each procedure is a defined data point. For eyes in which preoperative intraocular pressure was no more than 22 mmHg, success was defined as a decrease in intraocular pressure of at least 30%.

but the patient was treated too recently to be included in this report.) Early and then late postoperative iris reincarceration occurred in one phakic eye. The iris was easily dislodged on the first postoperative day with an argon laser but reincarcerated twice during the next 2 days. The anterior chamber was deep in this eye, but the sclerostomy had been created through the posterior trabecular meshwork. We initially used pilocarpine to tighten the iris while dislodging the incarcerated iris with an argon laser. We then switched to Mydriacyl, and the patient did not have iris reincarceration for 3 months. Intraocular pressure averaged 9 mmHg (preoperative lOP was 60 mmHg) with no medications except tropicamide twice daily. Four months after surgery, the patient inadvertently ran out of tropicam ide and 2 days later had an irritated eye and an lOP of 60 mmHg. The iris again was dislodged with an argon laser, and tropicamide was restarted. The lOP was 9 mmHg 8 months postoperatively, but the patient continues to take tropicamide twice daily. Of the two eyes that sustained conjunctival burn, one failed after 3 months. The second eye is now 9 months postoperative with an lOP of 19 mmHg. Two phakic patients had focal iris contraction without iris perforation consistent with thermal laser damage. Results of dilated examination did not show evidence of associated early lenticular change. Two phakic eyes had intentional holmium iridectomy with associated lenticular change. One eye was found to have a focal iris adhesion to the lens capsule. This lesion appeared similar to lens changes typically seen after creation of an argon laser iridotomy. This eye has been followed for 6 months without progressive lenticular change. The second phakic eye had

362

an unremarkable postoperative course for 2 months. However, a mild iritis was noted soon after topical corticosteroids were discontinued. Results of examination showed a faint focal lens capsule opacity at the lens equator. This lesion could only be seen after the eye was dilated and visualized through a gonioprism. One eye, which had a pre-existing cataract, underwent a laser sclerostomy without intraoperative complication. However, 2 weeks postoperatively, the lOP was 23 mmHg, prompting Nd: YAG laser treatment to the internal stoma and external needling of the filtration bleb. These interventions were successful, resulting in an lOP of less than 5 mmHg associated with a shallow anterior chamber for the ensuing 2 weeks. During this period, the cataract worsened, eventually requiring cataract extraction with good postoperative vision and an lOP of 15 mmHg without antiglaucoma medications. Some shallowing of the anterior chamber was seen initially in all phakic patients. Anterior chamber shallowing was minimized in eyes that had intraoperative viscoelastic injected through the sclerostomy site. Two phakic eyes that had not received intraoperative viscoelastic deepened postoperatively after a contact lens was applied. In one phakic eye with elevated episcleral venous pressure due to a presumed arteriovenous shunt, a flat anterior chamber developed, requiring drainage of a choroidal effusion described in an earlier report on this series. 9 Surgical intervention for choroidal effusion was not needed except in one previously reported eye associated with elevated episcleral venous pressure. 9 No intraoperative choroidal hemorrhages were detected. In one eye with iridocorneal endothelial syndrome, an intraoperative I-mm hyphema developed, but it resolved spontaneously. Postoperatively, choroidal hemorrhage has been seen in one aphakic eye in this series but did not require drainage. Table 9. Incidence* of Complications in the Study Cohort Total no. of eyes Conjunctival bum Early hypotony Shallow anterior chamber Flat anterior chamber Early iris incarceration Late iris reincarceration Late lens changes Acute lens change Transient corneal changes Endophthalmitis Early Late Choroidal hemorrhage Intraoperative hyphema NA

=

Phakic

Pseudophakic

Aphakic

22 (100%)

19 (100%) 2 19

8 (100%)

0 0 0 0

0 0 0 0

2

NA NA

NA NA

5

0

0

0 0 0

0

1

0 0

0 0

0

0 22 22 1 9

not applicable.

• In eyes included in our center's series. See Discussion section.

0 8

1

Iwach et al . Subconjunctival THC:YAG Laser Sclerostomy This eye eventually underwent a vitrectomy with Molteno implantation and laser cyclophotocoagulation. Focal corneal changes were initially seen in five eyes. Contraction of scleral and corneal tissue adjacent to the laser sclerostomy was noted with associated Descemet's membrane striae emanating from the sclerostomy site. Within 6 weeks postoperatively, these changes were no longer visible. One eye with corneal edema due to Fuch's endothelial dystrophy had undergone a holmium sclerostomy in preparation for a penetrating corneal keratoplasty. During the laser treatment, focal worsening of preexisting corneal edema adjacent to the laser probe tip was seen. However, the entire cornea gradually cleared with sustained lower postoperative lOPs delaying the need for a cornea transplant. In one patient with bilateral holmium sclerostomies, late ocular infections developed. Three months postoperatively, central corneal infiltrates secondary to Moraxella non-liquefaciens developed in the pseudophakic right eye, which received 5-FU postoperatively. The infiltrates gradually resolved with topical polymyxin B sulfate (Polysporin) and ciprofloxacin (Ciloxan) though the corneal surface remains irregular. However, 2 weeks after the onset of corneal infiltrates in the right eye, a bleb infection and endophthalmitis developed in the left eye, which was 8 months postoperative holmium sclerostomy. No organisms were seen on Gram stain, nor was there any growth on cultures, including vitreous and aqueous specimens. The left eye underwent a vitrectomy with intravitreal antibiotics injection. Two months after the onset of infection, the right eye lost six lines of vision, which was attributed to irregular corneal surface, whereas the left eye lost one line of vision. Despite these infections, the patient maintains lOPs of 15 mmHg in both eyes while taking medication, with an obvious filtration bleb in the left eye.

Discussion Although the surgical procedure is straightforward, attention to detail is essential to minimize complications. As with any new procedure, changes in technique and new complications are encountered with additional experience. Various anesthetic methods have been used. Although topical anesthesia alone was adequate in one of our patients, retrobulbar anesthesia provides additional surgical control. One case of eye movement with resultant conjunctival injury associated with topical anesthesia has been reported at another center (Holmium Users' Group Meeting, Dallas, TX, June 1991). As with any surgical procedure, the choice of anesthesia should respect both the surgeon's and patient's needs. The positioning and orientation of the laser probe has proven to be of critical importance. If the probe is angled too posteriorly, the laser energy is directed into the ciliary body and/or lens. If the laser energy is directed only slightly posteriorly, the approximation of the sclerostomy's internal ostium to the iris insertion may lead to repeat iris incarceration in phakic patients. Excessive anterior misdirection of laser energy can result in significant cor-

neal thermal changes with potential endothelial cell loss or conjunctival burn. Inaccurate probe-to-sclera apposition pressure allows accumulation of subconjunctival fluid between the probe tip and sclera. The fluid absorbs the laser energy instead of it being absorbed by the underlying sclera. Thus, more energy is required to create a sclerostomy with greater thermal damage to the sclera and cornea. Excessive probe pressure against the sclera creates a conical sclerostomy with a smaller internal opening. Although the ostium is cylindrical while the probe is engaged, it becomes conical as soon as the probe is released. Less total energy was required when sclerostomies were created at sites where limbal tissue was thin due to previous surgery or inflammation. Higher energy pulses were used if a larger sclerostomy was desired (unpublished data). When creating a sclerostomy inferiorly, a more accurate perspective of the helium-neon aiming beam is achieved if the surgeon sits at the patient's side while performing the sclerostomy. Excessive laser energy can create conjunctival burns, iris changes, or even lens damage. Subtle modifications in surgical technique have reduced the likelihood of conjunctivallaser damage. In addition to cooling overlying conjunctiva with irrigation during laser emission, a 10second cooling period every 2 seconds (10 pulses) decreases the risk of conjunctival burns from excessive heat at the probe tip. The risk oflens injury can be minimized if, once anterior chamber shallowing is seen after sclerostomy creation, laser energy emission is reduced to 60 mJ/ pulse or less. If a larger or second sclerostomy is desired after the initial sclerostomy is created, intracameral viscoelastics can be injected through the patent sclerostomy using a blunt-tipped Rycroft cannula (Visitec, Sarasota, FL) to deepen the anterior chamber. Iris incarceration into the sclerostomy site in phakic patients may occur in some cases, despite various preventive measures. Preoperative iridoplasty may reduce the likelihood of iris incarceration (personal communication, Fathi EI Sayyad, Saudi Arabia). Other investigators have described iris incarceration occurring as late as 4 months postoperatively (personal communication, James McAllister, England; Howard Cohn, France). Pharmacologic manipulation of the iris in the postoperative period is usually avoided. It is unclear what role sclerostomy location and preoperative anterior chamber depth play in affecting postoperative use of cycloplegia or miosis to prevent iris incarceration. Both cycloplegia and miosis seemed to be effective in reducing iris incarceration in occasional cases. Careful preoperative ocular examination and assessment are critical. Findings carry important implications regarding the anticipated intraoperative and postoperative ocular response to full-thickness filtration surgery. If vitreous is seen preoperatively, vitrectomy at the time of filtration surgery should be considered. Corneal opacification restricts examination of the anterior chamber for the presence of vitreous and limits the use of postoperative argon laser in dislodging incarcerated iris in phakic pa-

363

Ophthalmology

Volume 100, Number 3, March 1993

tients. Low preoperative corneal endothelial cell count may alter the choice of laser probe orientation or even the type of procedure selected. Although the small size of the holmium sclerostomy may minimize complications offull-thickness procedures, it does not eliminate them. Shallow anterior chambers with hypotony are the rule in phakic patients. Although hypotony occurs in aphakic and pseudophakic eyes, clinically significant anterior chamber shallowing is rare. Flat anterior chamber requiring surgical intervention has occurred once in a previously reported case associated with preoperative elevated episcleral venous pressure. 9 Intraoperative viscoelastic materials and postoperative soft contact lenses have deepened shallow chambers and may minimize potential complications associated with prolonged extreme hypotony. Subsequent to the collection of these data, postoperative subchoroidal hemorrhages too recently treated to be included in this series developed in two eyes. One occurred in an 84-year-old patient with aphakia; the other was in a 70-year-old patient with pseudophakia. The aphakic eye was treated conservatively with spontaneous resolution of the hemorrhage, and 3 months postoperatively it has an lOP of 10 mmHg with no loss in visual acuity. Pain developed in the pseudophakic eye, which was drained on postoperative day 7 with no recurrence. This second eye, 4 months postoperatively, has an lOP of 13 mmHg. Subchoroidal hemorrhages are a known complication of glaucoma filtering surgery. In a pilot study using 5-FU with trabeculectomy, a 13% incidence of choroidal hemorrhage in aphakic patients was seen. 22 The small size of the sclerostomy created by the laser may prevent massive expulsion of intraocular contents, reducing the likelihood of catastrophic visual loss. Our lower incidence rate of choroidal hemorrhage may reflect the size of our series. Avoiding posterior laser emission with its potential risk of ciliary body injury may minimize incidence of hemorrhage. Even though this surgical technique can simplify filtration surgery itself, the technique does not lessen the need for precise operative technique and careful postoperative care. This new tool for an old procedure requires skills to manage anticipated postoperative complications of full-thickness filtration surgery as well as management of iris incarceration. The trend toward improvement in our results in the later cases (Table 8) suggests a learning curve both in operative and postoperative management. Holmium laser sclerostomy has been especially useful for eyes in which the disease state demands a full-thickness procedure. Some very challenging eyes have done exceptionally well. For example, a 32-year-old patient with buphthalmos secondary to congenital glaucoma, despite seven filtration procedures, two Nd:YAG cyclophotocoagulation treatments, and multiple medications, had poor lOP control. The patient underwent a holmium sclerostomy through scarred tissues with minimal postoperative inflammation. Five months postoperatively, the eye has a large, diffuse bleb and an lOP of 11 mmHg without medication (Fig 7). Another patient with longstanding scleritis and glaucoma and multiple failed filtration pro-

364

cedures was challenging due to extremely thin sclera and scarred conjunctiva. This eye is now 12 months after a second holmium sclerostomy and maintains a large bleb with an lOP of 12 mmHg without medication. Currently, we consider holmium laser thermal sclerostomy an effective alternative to other full-thickness filtration procedures. We believe that the technique is most useful when the following clinical conditions exist: (1) conjunctival scarring such that creation ofa conjunctival flap would be difficult yet advancement of the holmium probe is possible, and (2) when pressures in the low teens are required and a standard full-thickness procedure would be appropriate. Weare particularly encouraged by the successful control and low pressures attained in lowrisk eyes (Table 7, Fig 6). In ophthalmology, we are inclined to think of laser procedures as having a rather straightforward operative technique and postoperative care. Such is not the case with laser sclerostomy. Meticulous attention to the details of operative technique and careful postoperative management is required. As with all full-thickness procedures, postoperative hypotony is common and may be profound. Additionally, with laser sclerostomy, iris incarceration may occur early or late in the postoperative period. Bleb failure may be averted by timely laser or needling intervention. All of this requires careful postoperative followup. Unlike laser trabeculoplasty, laser sclerostomy requires the knowledge and skills necessary to manage the postoperative course of full-thickness filtration surgery. Surgeons who do not routinely perform filtration surgery need to acquire skills required to manage these patients postoperatively. Laser sclerostomy provides a simple way to perform full-thickness filtration surgery for glaucoma. It requires attention to operative details and a full knowledge of management requirements of postoperative full-thickness filtration surgery. It has not replaced trabeculectomy as an effective glaucoma procedure. Rather, the holmium laser appears to offer a new approach to filtration surgery that will require further evaluation and understanding to define its ultimate role in management of the glaucoma patient.

References 1. Berlin MS, Rajacich G, Duffy M, et ai. Excimer laser pho-

2. 3. 4. 5.

toablation in glaucoma filtering surgery. Am J Ophthalmol 1987; 103:713-14. Brown RH, Denham DB, Bruner WE, et ai. Internal sclerectomy for glaucoma filtering surgery with an automated trephine. Arch Ophthalmol 1987; 105:133-6. Brown RH, Lynch MG, Denham DB, et ai. Internal sclerectomy with an automated trephine for advanced glaucoma. Ophthalmology 1988; 95:728-34. Federman JL, Wilson RP, Ando F, Peyman GA. Contact laser: thermal sclerostomy ab interna. Ophthalmic Surg 1987; 18:726-7. Gaasteriand DE, Hennings DR, BoutacoffTA, Bilek C. Ab interno and ab externo filtering operations by laser contact surgery. Ophthalmic Surg 1987; 18:254-7.

Iwach et al . Subconjunctival THC:YAG Laser Sclerostomy 6. Gherezghiher T, March WF, Koss MC, Nordquist RE. Neodymium-YAG laser sclerostomy in primates. Arch Ophthalmol 1985; 103:1543-5. 7. Higginbotham EJ, Kao G, Peyman G. Internal sclerostomy with the Nd: Y AG contact laser versus thermal sclerostomy in rabbits. Ophthalmology 1988; 95:385-90. 8. Hoskins HD Jr, Iwach AG, Drake MV, et al. Subconjunctival THC:YAG laser Iimbal sclerostomy ab externo in the rabbit. Ophthalmic Surg 1990; 21 :589-92. 9. Hoskins HD Jr, Iwach AG, Vassiliadis A, et al. Subconjunctival THC:YAG laser thermal sclerostomy. Ophthalmology 1991; 98:1394-1400. 10. Jaffe GJ, Mieler WF, Radius RL, et al. Ab inferno sclerostomy with a high-powered argon endolaser: clinicopathologic correlation. Arch Ophthalmol 1989; 107: 1183-5. 11. Jaffe GJ, Williams GA, Mieler WF, Radius RL. Ab interno sclerostomy with a high-powered argon endolaser. Am J Ophthalmol 1988; 106:391-6. 12. Javitt JC, O'Connor SS, Wilson RP, Federman JL. Laser sclerostomy ab interno using a continuous wave Nd:YAG laser. Ophthalmic Surg 1989; 20:552-6. 13. Latina MA, Dobrogowski M, March WF, Birngruber R. Laser sclerostomy by pulsed-dye laser and goniolens. Arch Ophthalmol 1990; 108: 1745-50.

14. L'Esperance FA Jr, Mittl RN. Carbon dioxide laser trabeculostomy for the treatment of neovascular glaucoma. Trans Am Ophthalmol Soc 1982; 80:262-87. IS. March WF, Gherezghiher T, Koss MC, Nordquist RE. Experimental YAG laser sclerostomy. Arch Ophthalmol 1984; 102:1834-6. 16. March WF, Gherezghiher T, Koss MC, et al. Histologic study of a neodymium-VAG laser sclerostomy. Arch Ophthalmol 1985; 103:860-3. 17. Ozier SA, Hill RA, Andrews JJ, et al. Infrared laser sclerostomies. Invest Ophthalmol Vis Sci 1991 ; 32:2498-2503. 18. Shirato S, Adachi M, Yamashita H. Internal sclerostomy with argon contact laser-animal experiment using 5-fluorouracil. Jpn J Ophthalmol 1990; 34:381-7. 19. Wilson RP, Javitt Jc. Ab interno laser sclerostomy in aphakic patients with glaucoma and chronic inflammation. Am J Ophthalmol 1990; 110: 178-84. 20. Miller RG Jr. Survival Analysis. New York: John Wiley & Sons, 1981. 21. Bickel PJ, Doksum KA. Mathematical Statistics: Basic Ideas and Selected Topics. San Francisco: Holden-Day, 1977. 22. Heuer DK, Parrish RK II, Gressel MG, et al. 5-fluorouracil and glaucoma filtering surgery. III. Intermediate follow-up of a pilot study. Ophthalmology 1986; 93: 1537-46.

Discussion by Douglas E. Gaasterland, MD ]n 1990, this group presented an analysis of early results oflaser sclerostomy filtering surgery done with subconjunctival delivery of holmium laser energy. I Laser sclerostomy is a full-thickness procedure, easier to do than trabeculectomy. The surgeon can locate the filtering tract in unconventional or normally hard-toreach meridians. Adjusting the laser pulse energy during treatment affects the diameter of the scleral tract. In discussing this article last year, I noted that the short follow-up was inadequate to give a satisfactory indication of the effectiveness of the procedure. Also, the preliminary report raised several questions. Most importantly, we wondered whether the outcome of the new technique would be an improvement over the outcome after standard management. The investigators now report longerterm results. Follow-up has been extended to 1 year for 26 of 49 eyes enrolled in their study. During the year, follow-up measurements have been obtained every 3 months. Eligibility criteria for eyes enrolled in this study are lenient. As a result, approximately one third of enrolled eyes can be categorized as low risk for failure of glaucoma filtration surgery; the remaining eyes can be categorized as high risk. Three patients contributed two eyes to the study. There is no adjustment in the data analysis to account for this, but the number of bilateral participants is small enough that the effect on outcome assessment is probably small. In the study, the major criterion for success is reduction of intraocular pressure after surgery. The authors do not emphasize retention of visual function. This is acceptable in a pilot study, but most ophthalmologists do glaucoma surgery to preserve visual function . Because the method of treatment is comparatively simple, the protocol allows a secFrom the Center for Sight, Georgetown University Medical Center, Washington, DC. Presented at the American Academy of Ophthalmology Annual Meeting, October 1991, Anaheim, California.

ond operation to be done if the first one fails. This is similar to the usual clinical approach to surgical glaucoma management; ophthalmologists usually do a second filtering procedure after failure of the first. Although the protocol allowed a second sclerostomy, the protocol did not require it. Twelve of 49 eyes had failure of the first sclerostomy within 3 months. Nine of these had a second sclerostomy; it is noteworthy that only four of the second operations were successful for 3 more months. By the end of the study, 15 failures occurred among 49 treated eyes, despite a second operation in most of the first-failure eyes. Nine failures occurred within 3 months, and four more occurred during the second 3 months after the initial intervention. One eye with a "failure" recovered. Subgroup analysis shows that eyes with lower risk had fewer failures. Early in the study, the authors learned that administration of adjunctive 5-fluorouracil (5-FU) enhanced success. Most treated eyes received adjunctive 5-FU. Except for failures, most problems encountered by the authors during surgery and follow-up were minor. Endophthalmitis developed in an eye 8 months after surgery. The same patient had a bleb infection in the other eye, which also had been treated with laser sclerostomy. Iris incarceration in the sclerostomy site occurred in 9 of the 22 phakic eyes treated. Various interventions resolved these problems. Conjunctival burns occurred during treatment of two eyes; these required suturing. Other problems that occurred at the time of surgery include iris contraction and local corneal traction in a few eyes. Apre-existing cataract worsened during the early postoperative period in one eye. The important consideration for the reader is whether the outcome after laser sclerostomy is better than the outcome after alternative treatments. This study differs from others in the literature, and comparisons of outcome are potentially inaccurate. Because the investigators used adjunctive 5-FU for most of the eyes in this study and because their study group contains a large proportion (65%) of high-risk eyes, a comparison can be made

365

Ophthalmology

Volume 100, Number 3, March 1993

with the results of the Fluorouracil Filtering Surgery Study (FFSS).2 In the FFSS, 39% of 108 high-risk eyes had failure 1 year after standard trabeculectomy (without adjunctive 5-FU) for uncontrolled glaucoma. In contrast, 17% of 105 similar eyes randomly assigned in the FFSS to receive trabeculectomy and adjunctive 5-FU had a failure by 1 year. The I-year failure rate among the eyes enrolled in the current study is 42% after one laser sclerostomy and 32% after two operations, based on a Kaplan-Meier survival analysis provided by the authors. This suggests that eyes with a variety of risk levels receiving a first-laser sclerostomy with adjunctive 5-FU have approximately the same outcome as high-risk eyes after standard trabeculectomy (without adjunctive 5-FU). The rate of success for eyes in the current study, some of which had two laser sclerostomies, is considerably worse than success in the FFSS after a single trabeculectomy with adjunctive 5-FU.

366

In this study, the rate offailure oflaser sclerostomy is less in low-risk eyes than in high-risk eyes. Laser sclerostomy in lowrisk eyes may be an effective alternative to trabeculectomy. The comparative ease of doing this surgical treatment makes it attractive, particularly for patients unable to receive standard filtering surgery and with low-risk eyes. References I. Hoskins HD Jr, Iwach AG, Vassiliadis A, et al. Subcon-

junctival THC:YAG laser thermal sclerostomy. Ophthalmology 1991 ;98: 1394-1400. 2. The Fluorouracil Filtering Surgery Study Group. Fluorouracil study one-year follow-up. Am J Ophthalmol 1989; 108:625-35.