Refractory glaucoma – tube or diode?

$Page 1: Refractory glaucoma – tube or diode?$
Clinical and Experimental Ophthalmology 2006; 34: 771–777doi:10.1111/j.1442-9071.2006.01339.x

© 2006 Royal Australian and New Zealand College of Ophthalmologists

� Correspondence: Dr Rizwan Malik, Glaucoma Research Unit, Moorfields Eye Hospital, London, UK. Email: [email protected]

Received 23 January 2006; accepted 9 June 2006.

Original Article

Refractory glaucoma – tube or diode?Rizwan Malik MRCOphth,1 Roger B Ellingham PhD FRCOphth,2 Hanif Suleman MRCOphth3 and William H Morgan PhD FRANZO2

1Bristol Eye Hospital, Bristol, 3Taunton & Somerset Hospital, Taunton, UK; and 2Lion’s Eye Institute, Perth, Western Australia, Australia

INTRODUCTION

Both tube shunt surgery and trans-scleral (diode) laser cyclo-photocoagulation are usually reserved for cases of refractoryglaucoma which have failed, or are deemed at high risk offailure, from trabeculectomy. The choice of one of thesetreatment modalities over the other is not an infrequentclinical dilemma. However, despite similar indications, theclinical evidence to support the preference of clinical inter-vention is scarce.

Previous studies have compared tube-shunt surgery withNeodymium:YAG (Nd:YAG) cyclophotocoagulation.1,2

However, cyclodiode transscleral laser has largely replacedNd:YAG cyclophotocoagulation because of its technicalease, equivalent IOP effects3 and lower incidence of vision-threatening complications,4,5 presumably owing to sparingtissue damage in the treatment region.6 In one animal study,diode laser was found to be of similar IOP-lowering efficacyto Mitomycin C augmented filtration surgery.7

Given that cyclodiode laser is less invasive and techni-cally simpler to perform, and the postoperative recovery isusually less prolonged than tube surgery, the relative IOP-lowering efficacies of these two methods needs to beassessed.

The purpose of this study was to compare the long-termresults of cyclodiode laser and Molteno tube implantationfor the management of refractory glaucoma.

METHODS

Study design

A retrospective cohort study was undertaken to compare theoutcome of double plate Molteno tube surgery with transs-cleral diode laser. Case notes from two centres (Bristol EyeHospital, Bristol, UK and Royal Perth Hospital, Perth,Australia) were reviewed. Consecutive patients who under-went Molteno tube implantation and trans-scleral diode laserbetween 1996 and 2001 were included. Demographic

ABSTRACT

Background: The aim of the present study was to assessthe relative effectiveness of tube surgery and cyclodiodelaser in terms of achieving intraocular pressure control.

Methods: A retrospective study was undertaken to comparepatients undergoing double plate Molteno tube implantationwith patients undergoing diode cyclophotocoagulation.Intraocular pressure (IOP) was documented at 7 days priorto surgery and postoperatively at various time points. Sur-gical success was defined as a final IOP between 6 (inclusive)and 21 mmHg (inclusive), without the use of topical medi-cation, while ‘qualified’ success was defined as IOP within thesame range with the use of topical medication.

Results: Twenty-eight diode patients and 26 tube patientswere included for the study. An average follow up of150 weeks (range = 21–322 weeks) was available. Meanpreoperative IOP was 37 ± 12 mmHg for the tube groupand 39 ± 16 mmHg for the diode group (t = 0.51,P = 0.61). The final IOP was 17 ± 12 mmHg for the tubegroup and 21 ± 13 mmHg for the diode group (t = 0.35,P = 0.73). Surgical success was achieved in 46% of tube eyesand 11% of diode eyes, while qualified success was achievedin 81% of tube eyes compared with 64% of eyes in thediode group. Two eyes which underwent diode becamephthisical.

Conclusions: IOP control may be achieved in a greaternumber of patients with tube surgery. The possible benefitsof IOP control in diode patients need to be weighed againstthe risks of long-term visual loss and the need for multiplere-treatments in this group.

Key words: Molteno tube surgery, refractory glaucoma,trans-scleral cyclodiode laser.

772 Malik et al.


details, past ocular history and glaucoma subtype were doc-umented. Visual acuity, IOP and number of ocular medica-tions were recorded preoperatively and at each successivepostoperative visit. None of the tube subjects had receivedany form of cyclodestructive surgery (laser or cryotherapy)and diode patients who had previously received other formsof cyclodestructive therapy were excluded from the study.

Intervention

Molteno tube surgery

This surgical technique has been previously described indetail8 and is mentioned here briefly.

All patients required general anaesthesia. Following a 220degree conjunctival limbal peritomy, the Molteno double-plate implant was sutured (with the interconnecting tubepassing above the superior rectus muscle). A partial thicknessscleral flap, with a limbal base, was prepared. The primarytube was cut with a forward-facing bevel tip extendingbeyond the limbus. The tube was passed through a 23-gaugeneedle track entering the anterior chamber parallel with theiris plane and protected externally by a donor scleral graftbeneath the host scleral flap. The primary tube was tied witha 5/0 absorbable suture around a 3/0 nylon pull string and aSherwood slit made to prevent the IOP rising in the imme-diate postoperative period. The conjunctiva and Tenon’s cap-sule was closed in one layer with 8/0 vicryl.

Postoperatively, patients were given guttae 0.1% dexam-ethasone and 0.5% chloramphenicol (both four times daily)until the conjunctiva healed. The steroid was continued forat least 2 months. The 3/0 nylon ligature was released if theIOP rose above 20 mmHg postoperatively, although it wasnever pulled before the second postoperative week to allowconjunctival healing. During this period, IOP elevation wascontrolled with the use of topical medication and, if needed,oral acetozolamide tablets. Following the release of the 3/0nylon suture, if the IOP was rising to a level perceived toohigh for the level of optic disc damage, a medical antifibrosisregimen was prescribed orally. This consisted of 40–60 mgprednisolone daily, 500 mg colchicine twice daily and piroxi-cam 10 mg daily. This was generally commenced 4 monthspost surgery. Oral steroids were also used to protect cornealgrafts from the short-term inflammatory effects of surgery.

Trans-scleral diode laser

Diode laser was performed with the patient supine underlocal anaesthetic whenever possible; patient preference andanxiety were common reasons for the use of generalanaesthesia.

For local anaesthesia, a peribulbar or retrobulbar injectionof 2 mL to 8 mL of lignocaine hydrochloride (with bupivic-aine hydrochloride if necessary) was given. Prior to treat-ment, a lid speculum was placed. Laser energy was deliveredusing a diode laser system with a 600 micrometer quartz fibre‘G’ probe (IRIS Oculight SLx, IRIS Medical Inc., Mountain

View, CA). The ciliary body was identified by transillumina-tion. The duration of the laser system was set to 2000 msand the power to 1750 mW. Power increments of 250 mWwere used (to a maximum of 2500 mW) until audible ‘pops’were heard to signify tissue disruption in the ciliary process.At this point, the power was reduced by 250 mW, and thetreatment completed at this power. Typically, six applica-tions per quadrant were used, with burns spaced about 2 mmapart. The number of burns and total energy used wasdocumented.

Data analysis

In accordance with a previous study,9 preoperative and post-operative visual acuities were converted to LogMar equiva-lents from Snellen acuity to aid simple comparison betweengroups.

Intraocular pressure measurements for each patient wereaveraged for the following time periods: 7 days prior to sur-gery and postoperatively at 1–30, 31–90, 91–180, 181–270,271–360 and 361–720 days. The final available IOP readingwas also recorded. The use of postoperative hypotensiveagents and complications of each of the two procedures wasnoted. Surgical success was defined as an IOP between6 mmHg (inclusive) and 21 mmHg (inclusive) without ocu-lar antihypertensives and qualified success was defined as an IOPbetween 6 and 21 mmHg with the use of medication.

Kaplan–Meir survival analyses were plotted with the def-inition of qualified IOP success as stated above. As IOPfollowing tube surgery is often labile in the postoperativeperiod, IOP values within 3 months postoperatively wereexcluded from this analysis.

Data analysis was performed with Microsoft Excel 2003(Microsoft, Redmond, WA) and StatView 4.53 (StatView,Cary, NC, USA). Unpaired t-tests were used for comparisonof continuous variables and chi-squared tests were used forcomparison of categorical data. A significance level ofP < 0.05 was used.

RESULTS

Twenty-eight diodes and 26 tubes were included, performedduring the period between 1996 and 2001. Sex characteris-tics of the ‘diode’ and ‘tube’ groups were similar (χ2 = 0.12,P = 0.73). The average age of diode patients was50 ± 26 years compared with 64 ± 21 years for the tubegroup (t = 2.2, P = 0.03).

Overall, a longer postoperative final follow up was avail-able for the diode group (204 ± 78 weeks compared with98 ± 65 weeks for the tube group). The range of diagnosesfor the two groups was fairly similar (Table 1), although alarger proportion of diode patients had open angle andneovascular glaucoma, but less uveitic glaucoma. Previoustreatments are shown in Table 2.

Eyes which underwent diode laser mostly received 20–25burns per session (range 14–50 burns). An average totalpower of 206 mW (SD 212 mW, range = 80–1047 mW) was

Tube or diode? 773


used over the course of diode treatments. Eleven eyes (39%)required only a single diode treatment to achieve IOP con-trol. Seventeen eyes (61%) needed diode re-treatment, withtwo eyes (7%) requiring seven separate treatments.

Mean preoperative IOP was similar for both groups(37 ± 12 mmHg for tubes and 39 ± 16 mmHg for diodes(t = 0.51, P = 0.61). Figure 1 shows the mean IOP at eachtime interval. There was no statistical difference in IOPbetween groups at any time point (taking Bonferroni-adjusted P = 0.006 for eight comparisons), but the scatter(range) was greater for diodes at all time intervals. The finalIOP (17 ± 9 mmHg for tube group and 17 ± 12 mmHg fordiodes) was similar (t = 0.03, P = 0.98).

At the final visit, 12 of the patients who had tube surgeryand 19 patients who received diode laser were receiving

ocular hypotensive treatment. For the patients on hypoten-sive treatment at final follow up, the mean number of medi-cations was similar for patients in the tube and diode group(t = 0.92, P = 0.36).

Inspection of survival analyses suggested that qualifiedIOP success was more consistently achieved in the tubegroup compared with the diode group, log-rank χ2 = 7.49,

Table 1. Diagnoses for each treatment group

Diagnosis Diode Tube Total

POAG/NTG 10 5 15Rubeosis 9 5 14Aphakia 3 4 7Congenital glaucoma 1 5 6Uveitis/FHC 1 4 5Post PKP 1 2 3ICE/Rieger 0 3 3Retinal detachment 1 0 1Aniridia 1 0 1Sturge Weber 0 1 1Chemical injury 0 1 1Blunt trauma 0 1 1Penetrating trauma 1 2 3Epithelial down-growth 1 0 1

Totals do not agree with totals in each diagnosis group as someeyes had more than one diagnosis. FHC, Fuch’s heterochromiccyclitis; ICE, iridocorneal endothelial syndrome; NTG, normal ten-sion glaucoma; PKP, penetrating keratoplasty; POAG, primaryopen angle glaucoma.

Table 2. Previous treatments

Previous treatments Diode Tube Total

None 8 1 9Trabeculectomy 5 3 8Trabeculectomy with antimetabolite 1 7 8Multiple trabeculectomy 3 5 8ECCE IOL 4 5 9ICCE/lensectomy 1 3 4Penetrating Keratoplasty 3 3 6PRP laser 6 3 9Trauma repair 1 2 3Keratoprosthesis 0 1 1Vitrectomy 1 1 2

Totals do not agree with totals in each group as some eyes hadmore than one previous treatment. EECE IOL, extracapsular cata-ract extraction with intraocular lens implantation; ICCE, intracap-sular cataract extraction; PRP, panretinal photocoagulation.

Figure 1. Mean intraocular pressure (IOP) versus time for diodeand tube groups.

45

40Tube

Diode

30

20

15

10

Time (days postoperative)

Mea

n IO

P (

mm

Hg)

Preop

erat

ive

1st V

isit

1–30

31–9

0

91–1

80

181–

270

271–

360

361–

720

Final IO

P

35

25

Figure 2. Survival analyses for diode and tube groups, withintraocular pressure qualified success defined as intraocular pressurebetween 6 and 21 mmHg (inclusive), for (a) all cases (log-rankχ2 = 7.49, P = 0.006), and (b) with neovascular cases excluded (log-rank χ2 = 2.57, P = 0.11).

100

(a)

(b)

80

60

40

20

0

0 200

DiodeTube

400 600Time (days postoperative)

Sur

viva

l pro

babi

lity

(%)

800 1000 1200 1400

0 200

DiodeTube

400 600

Time (days postoperative)

Sur

viva

l pro

babi

lity

(%)

800 1000 1200 1400

100

80

60

40

20

0

774 Malik et al.


P = 0.006 (Fig. 2a). A higher proportion of eyes whichunderwent diode had neovascular glaucoma compared witheyes in the tube group (Table 1). As this diagnosis is oftenassociated with a poor IOP control, the Kaplan–Meir anal-ysis was recomputed with all cases of neovascular glaucomaexcluded from both groups. Survival for the diode and tubeeyes were similar once cases of neovascular glaucoma hadbeen excluded, log-rank χ2 = 2.57, P = 0.11 (Fig. 2b).

The mean IOP drop (preoperative to last postoperativevisit) was 20 ± 14 mmHg for the tube group and19 ± 23 mmHg for the diode group and therefore was equiv-alent (t < 0.001, P = 1.00). This represented a 46% reductionfrom baseline for the tube group compared with a 50%reduction in the diode group. Surgical success (IOP between6 and 21 mmHg without medication) was achieved in 12eyes (46%) in the tube group compared with only three(11%) eyes in the diode group. Qualified success, that is, afinal IOP (with or without treatment) in the range 6–21 mmHg was achieved in 18/28 (64%) of diode patientscompared with 21/26 (81%) of tube patients. When cases ofneovascular glaucoma were excluded from this analysis, qual-ified success rates were 81% for tube and 74% for diode eyes,respectively. The proportion of eyes achieving qualified suc-cess was similar for tube and diode eyes (χ2 = 1.10, P = 0.30).Excluding neovascular cases in each group did not affect thisresult (χ2 = 0.03, P = 0.86). Figure 3 shows the final IOP ineach group.

Both preoperative and postoperatively, visual acuitieswere worse in the ‘diode’ group (Fig. 4). Vision deterioratedin 14 of the tube eyes and 13 in the diode eyes, althoughthe mean deterioration in log visual acuity was worse for thediode group (t = 4.51, P < 0.0001). Deterioration of visionwas due to the complications outlined in Table 3. In threeinstances, visual acuity in the tube group improved. Of these,one experienced an improvement beyond 1-line Snellenequivalent owing to clearing of corneal graft oedema. In theother two cases, a 1-line improvement of visual acuityoccurred.

Nearly half (13/28, 46%) of the diode procedures werefree of complications compared with 31% for the tube group.However, two eyes that underwent diode became phthisical(Table 3). The number with persistent postoperative uveitis

Figure 3. Final intraocular pressure (IOP) (with and without top-ical treatment) by treatment group (tube/diode) for (a) all cases,and (b) eyes with neovascular glaucoma excluded.

40

(a)

(b)

30

20

10

Pro

port

ion

of e

yes

(%)

00–3 4–7 8–11 12–15 16–19 20–23 24–27

Tube

Diode

28–31 >31

Final IOP (mmHg)

40

30

20

10

Pro

port

ion

of e

yes

(%)

00–3 4–7 8–11 12–15 16–19 20–23 24–27 28–31 >31

Final IOP (mmHg)

TubeDiode

Figure 4. Preoperative and postoperative log visual acuities (VA).

TubeDiode

4

2

0

0 1 2

Log final VA

3 4

Log

preo

pera

tive

VA

Table 3. Postoperative complications

Complication Tube Diode Total

None 8 13 21Corneal oedema 3 1 4Flat AC 2 0 2Hyphaema 3 2 5Worsening graft oedema/failure 2 1 3Choroidal effusion/detachment 3 3 6Rubeosis 0 1 1Mirobial keratitis 0 1 1Late ptosis 1 0 1Bullous keratopathy 0 1 1Uveitis 1 5 6Cataract 1 0 1CMO 0 1 1Tube repositioning 2 0 2Steroid psychosis 1 0 1Phthisis 0 2 2Pain 0 1 1

Totals do not agree with totals in each group as some eyes hadmore than one complication. AC, anterior chamber; CMO, cystoidmacular oedema.

Tube or diode? 775


was also higher in the diode group. Three eyes from the tubegroup developed corneal oedema and a further two devel-oped graft oedema/failure. In three of these eyes, this wasthe cause of poor visual acuity (6/60 or worse) at final follow-up. The primary diagnoses in the three eyes which devel-oped persistent corneal oedema after tube surgery wereFuch’s Heterochromic Cyclitis, aphakic glaucoma and ICEsyndrome. Direct tube-cornea touch was related to one caseof PKP failure. The second patient with post-tube PKP fail-ure had developed refractory glaucoma after a previous PKP.The two patients who developed corneal decompensationafter diode laser had rubeotic glaucoma and Fuch’s endothe-lial dystrophy.

DISCUSSION

Inspection of the results of this study suggest that the aver-age IOP-lowering efficacy of diode and Molteno tube shuntsurgery are fairly comparable, although a greater proportionof eyes may achieve final surgical success with tube implantsurgery. However, the nature of complications and rate ofvisual loss in these two groups may be different.

Both cyclodiode laser10–18 and tube surgery1,2,19–22 areeffective methods of achieving IOP control. The findings ofthe current study are in agreement with previous studieswhich have compared other cyclodestructive procedureswith drainage implant surgery,1,2,23 although direct compari-sons between studies are difficult owing to the differingdiagnoses, definitions of ‘success’ and thresholds for re-treat-ment. Ayyala et al.2 reported a final IOP of between 6 and20 mmHg in 80% of eyes which underwent glaucoma devicesurgery compared with 63% in eyes which received neody-mium:YAG cyclophotocoagulation for patients with intrac-table glaucoma after penetrating keratoplasty.2 These areconsistent with our figures of 64% for diode and 81% fortube surgery. Noureddin et al. found tube surgery maintainedan IOP of less than 21 mmHg in 87% of patients comparedwith 69% of Nd:YAG treated patients at 1-year follow up(with or without medication), with fewer patients in the tubegroup needing medication to control the IOP.23 Anotherstudy found success (defined as IOP between 6 and25 mmHg) rates at the final postoperative visit,1 of about40% success following Nd:YAG cyclophotocoagulation andnearly 70% success following tube surgery. The latter studyexamined only patients with neovascular glaucoma and thismay explain the lower success rates, as both transscleraldiode laser4 and tube implants may be less successful in thisgroup of patients.10,22,24

This, to our knowledge, is the first study which hasdirectly compared the IOP-lowering effects of transscleraldiode laser with tube shunt surgery.

To allow direct comparison of our results with previousstudies, results of survival analyses have been presented(Fig. 2). A Kaplan–Meier analysis of our data, excluding IOPvalues within 3 months of the surgical intervention, revealed1-year survival rates of around 81% for the tube group versus68% for the diode group and respective 2-year survival rates

of 73% and 54% when a (qualified) ‘success’ definition ofIOP between 6 and 21 mmHg (inclusive) was applied. Thesesurvival rates are consistent with values reported in the liter-ature.1,5,13,21,24 While this analysis suggested higher survivalfor eyes in the tube group (Fig. 2a), there was no differenceof survival between groups when cases of neovascular glau-coma were excluded (Fig. 2b). Although several studies haveused survival curves and cumulative probabilities to assessoutcome from drainage surgery,7,10,23,24 such analyses may notbe ideal for comparing the two treatment groups. Survivalanalyses are unlikely to indicate long-term IOP success rates:IOP is often labile in the first postoperative months follow-ing drainage implant surgery, pushing IOP beyond thedefined endpoints for failure in the early postoperativephase, despite achieving long-term success. An analysis ofthe distribution of ‘final IOP’ for the two groups, even whencases of neovascular glaucoma had been excluded (Fig. 2b),suggested that a greater proportion of diode patients hadextremes of IOP in the longer term.

Complications of cyclodiode laser include phthisis,chronic hypotony, corneal graft decompensation, hyphaemaand vitreous haemorrhage.10,14,17,23 Use of high total diodeenergy may have contributed to the phthisis rate in our diodegroup. Three of the patients needed five or more diodesessions for IOP control with total powers of 408 mW,764 mW and 1047 mW over the course of their treatment.Complications of tube surgery relate to over-draining of thetube (choroidal effusion, hypotony, flat anterior chamber);tube malposition, corneal touch, tube block; and other com-plications of ocular surgery: (corneal oedema, corneal graftdecompensation, retinal detachment, vitreous haemorrhage,cataract).25 Failure of tube surgery may be due to prolifera-tion of fibrous tissue around the implant plates.26

Although total complication rates were similar in thediode and tube groups, there were two cases (10%) of phthi-sis in the diode group, which contributes to the higher meandeterioration in visual acuity in the diode group. Whilst thisis complicated by the initial poorer visual acuity in thisgroup of patients, visual loss is a well-recognized complica-tion of cyclodiode laser. The Diode Laser Ciliary AblationStudy Group reported visual loss in 30% of eyes.13 Reportedrates of visual loss from cyclodiode laser vary from approxi-mately 20% to over 50%.3,10,13,27 Many clinicians may delib-erately under-treat with diode laser to avoid eventualphthisis.

It is well-recognized that, on average, patients need mul-tiple diode treatments to attain IOP control. Diode treat-ment should therefore be regarded as a course of treatmentand patients should be consented with this understanding.

Corneal or graft decompensation occurred in five cases oftube surgery compared with only two eyes which receiveddiode. Underlying diagnoses in at least three eyes whichdeveloped irreversible corneal oedema following tubeimplant may have contributed (ICE syndrome, Fuch’s Het-erochromic cyclitis and post-PKP glaucoma). One case ofgraft failure was caused by direct tube touch. The aetiologyof persistent post-tube corneal oedema is not always clear.

776 Malik et al.


Reduced endothelial cell counts in eyes undergoing unevent-ful Molteno implant surgery may offer a possible explana-tion.28 Corneal decompensation rates as high as 50% havebeen reported following Molteno tube implantation.29

The results of this study imply that IOP control may beachieved in a greater number of patients with tube surgery.The possible benefits of IOP control in diode patients needto be weighed against the risks of long-term visual loss andthe need for multiple re-treatments in this group. In addition,the treatment needs to be tailored to the individual patient.

The findings of this study may be somewhat confoundedby age and diagnoses differences between the two groups.Patients in the tube group were, on average, younger thanpatients in the diode group. Success of cyclodiode laser maybe lower in younger eyes.4,12

A retrospective study such as this has intrinsic limitations,including the differing length of follow up and varyingdegrees of documentation of specific outcomes in case notesat each visit. Inherent selection bias is also possible: eyesconsidered to have a poor prognosis might have beenselected for diode laser rather than tube surgery. The higherproportion of neovasular glaucoma and poorer preoperativevisual acuity in the diode group is consistent with this argu-ment. However, this study illustrates that the relative IOP-lowering efficacies of the two treatment modalities are com-parable and highlights the need for a randomized controlledtrial in a larger population to fully evaluate the optimumintervention for the management of the different refractoryglaucomas.

REFERENCES

1. Eid TE, Katz LJ, Spaeth GL, Augsberger JJ. Tube-shunt surgeryversus Neodymium:YAG cyclophotocoagulation in the man-agement of neovascular glaucoma. Ophthalmology 1997; 104:1692–700.

2. Ayyala RS, Pieroth L, Vinals AF et al. Comparison of mitomycinC trabeculectomy, glaucoma drainage device implantation, andlaser neodymium: YAG cyclophotocoagulation in the manage-ment of intractable glaucoma after penetrating keratoplasty.Ophthalmology 1998; 105: 1550–6.

3. Youn J, Cox TA, Herndon LW, Allingham RR, Shields MB. Aclinical comparison of transscleral cyclophotocoagulation withneodymium: YAG and semiconductor diode lasers. Am J Oph-thalmol 1998; 126: 640–7.

4. Kramp K, Hans-Peter V, Guthoff R. Transscleral diode lasercontact cyclophotocoagulation in the treatment of differentglaucomas, also as primary surgery. Graefes Arch Clin Exp Oph-thalmol 2002; 240: 698–703.

5. Oguri A, Takahashi E, Tomita G, Yamamoto T, Jikihara S,Kitazawa Y. Transscleral cyclophotocoagulation with thediode laser for neovascular glaucoma. Ophthalmic Surg Lasers1998; 29: 722–7.

6. McKelvie PA, Walland MJ. Pathology of cyclodiode laser: aseries of nine enucleated eyes. Br J Ophthalmol 2002; 86: 381–6.

7. Pablo LE, Gomez ML, Pueyo M et al. Semiconductor diodelaser transscleral cyclophotocoagulation versus filtration sur-gery with Mitomycin-C. Int Ophthalmol 1996; 20: 11–4.

8. Morgan WH. Molteno tube insertion avoiding complications.Asia-Pacific J Ophthalmol 1996; 8: 7–9.

9. Ellingham RB, Morgan WH, Westlake W, House PH. Mito-mycin C eliminates the short-term intraocular pressure risefollowing Molteno tube implantation. Clin Exp Ophthalmol 2003;31: 191–8.

10. Bloom PA, Tsai JC, Sharma K et al. ‘Cyclodiode’. Trans-scleraldiode laser cyclophotocoagulation in the treatment ofadvanced refractory glaucoma. Ophthalmology 1996; 104:1508–20.

11. Bock CJ, Freedman SF, Buckley EG, Shields MB. Transscleraldiode laser cyclophotocoagulation for refractory paediatricglaucomas. J Paediatr Strabismus 1997; 34: 235–9.

12. Kirwan JF, Shah P, Khaw PT. Diode laser cyclophotocoagula-tion: role in the management of refractory paediatric glauco-mas. Ophthalmology 2002; 109: 316–23.

13. Kosoko O, Gaasterland DE, Pollack IP, Enger CL, the DiodeLaser Ciliary Ablation Group. Long-term outcome of initialciliary ablation with contact diode transscleral cyclophotoco-agulation for severe glaucoma. Ophthalmology 1995; 103: 1294–302.

14. Schlote T, Derse M, Rassmann K, Nicaeus T, Dietz K, ThielHJ. Efficacy and safety of contact transscleral diode laser cyco-photocoagulation for advanced glaucoma. J Glaucoma 2002; 10:294–301.

15. Shah P, Lee GA, Kirwan JK et al. Cyclodiode photocoagulationfor refractory glaucoma after penetrating keratoplasty. Ophthal-mology 2001; 108: 1986–91.

16. Spencer AF, Vernon SA. ‘Cyclodiode’: results of a standardprotocol. Br J Ophthalmol 1999; 83: 311–6.

17. Yap-Veloso MIR, Simmons RB, Echelman DA, Gonzales TK,Veira WJ, Simmons RJ. Intraocular pressure control after con-tact transscleral diode cyclophotocoagulation in eyes withintractable glaucoma. J Glaucoma 1997; 7: 319–28.

18. Hauber FA, Scherer WJ. Influence of total energy delivery onsuccess rate after contact diode laser transscleral cyclophoto-coagulation: a retrospective case review and meta-analysis. JGlaucoma 2002; 11: 329–33.

19. Broadway DC, Lester M, Schulzer M et al. Survival analysis forsuccess of Molteno tube implants. Br J Ophthalmol 2001; 85:689.

20. Mermoud A, Salmon JF, Alexander P, Straker C, Murray AD.Molteno tube implantation for neovascular glaucoma: long-term results and factors influencing outcome. Ophthalmology1992; 100: 897–902.

21. Mills RP, Reynolds A, Emond MJ, Barlow WE, Leen MM.Long-term survival of Molteno glaucoma drainage devices.Ophthalmology 1996; 103: 299–305.

22. Minckler DS, Heuer DK, Hasty B, Baerveldt G, Cutting RC,Barlow WE. Clinical Experience with the single-plate Moltenoimplant in complicated glaucomas. Ophthalmology 1988; 95:1181–8.

23. Noureddin BN, Wilson-Holt N, Lavin M, Jeffrey M, HitchingsRA. Advanced uncontrolled glaucoma. Nd:YAG cyclophoto-coagulation or tube surgery. Ophthalmology 1992; 99: 430–6.

24. Lloyd MA, Sedlak T, Heuer DK et al. Clinical experience withthe single-plate Molteno implant in complicated glaucomas.Update of a pilot study. Ophthalmology 1992; 99: 679–87.

25. Molteno ACB. Use of Molteno implants to treat secondaryglaucoma. In: Cairns JE, ed. Glaucoma. London: Grune & Strat-ton, 1996; 221–38.

Tube or diode? 777


26. Classen L, Kivela T, Tarkkanen A. Histolopathologic andimmunohistochemical analysis of the filtration bleb afterunsuccessful glaucoma Seton implantation. Am J Ophthalmol1996; 122: 205–12.

27. Brancato R, Carassa RG, Bettin P, Fiori M, Trabucchi G. Con-tact transscleral cyclophotocoagulation with diode laser inrefractory glaucoma. Eur J Ophthalmol 1995; 5: 32–9.

28. McDermott ML, Swendris RP, Shin DH, Juzych MS, CowdenJW. Corneal endothelial counts after Molteno implantation.Am J Ophthalmol 1993; 115: 93–6.

29. Zalloum JN, Ahuja RM, Shin D, Weiss JS. Assessment of cor-neal decompensation in eyes having undergone Molteno shuntprocedures compared to eyes having undergone trabeculec-tomy. CLAO J 1999; 25: 57–60.

Documents

Refractory glaucoma – tube or diode?