Article Text

Download PDFPDF

Original article
Baerveldt shunts in the treatment of glaucoma secondary to anterior uveal melanoma and proton beam radiotherapy
  1. Eamon Sharkawi,
  2. Justyna D Oleszczuk,
  3. Ciara Bergin,
  4. Leonidas Zografos
  1. Jules-Gonin Eye Hospital, Lausanne, Switzerland
  1. Correspondence to Dr Eamon Sharkawi, Hôpital Ophtalmique Jules-Gonin, Université de Lausanne, Av. de France 15, Lausanne 1004, Switzerland; eamon.sharkawi{at}


Introduction Melanoma of the iris and ciliary body may be associated with secondary glaucoma. Treatment with proton beam radiotherapy (PBRT) to the anterior segment can also elevate intraocular pressure (IOP), resulting in uncontrolled glaucoma, often requiring enucleation. This is the first prospective study of Baerveldt aqueous shunts in irradiated eyes with anterior uveal melanoma (AUM; affecting the iris or ciliary body).

Methods Thirty-one eyes with uncontrolled IOP following anterior segment PBRT treatment for AUM were prospectively recruited to undergo Baerveldt shunt implantation. Postoperative examinations were performed on day 1; weeks 1, 3, 6, 9; months 3, 6, 9, 12 and annually thereafter. Surgical success was defined as IOP 21 mm Hg or less and 20% reduction from baseline. All complications were recorded.

Results Mean follow-up was 15.7 months (SD ±8.3 months). Mean interval from irradiation to shunt implantation was 2.5 years. Mean preoperative IOP was 31.0 (±10.3) mm Hg; mean IOP at last visit was 15.0 (±5.0) mm Hg; mean pre-operative glaucoma medications were 3.3 (±1.3); postoperatively 0.7 (±1.3) glaucoma medications. Surgical success rate was 86% using glaucoma medications. Four eyes had minor postoperative complications, none of which were sight threatening. There were no local tumour recurrences or systemic metastases. There were no enucleations caused by ocular hypertension.

Conclusions Baerveldt shunts were effective in lowering IOP, with few complications, in eyes treated with total anterior segment irradiation for AUM.

  • Field of vision
  • glaucoma
  • imaging
  • intraocular pressure
  • optic nerve
  • psychophysics
  • visual perception

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Uveal melanoma is the most common primary ocular malignancy in adults.1 It can develop within the iris (5–8%), ciliary body (5–6%), choroid (69–91%) or as a combination of these (16%).2 Secondary glaucoma occurs in 3–57% of patients with uveal melanoma.2–4 Underlying mechanisms include: tumour invasion of the drainage pathway (3–57%), seeding (16–50%), angle neovascularisation (7–47%) and mechanical blockage by melanin-laden macrophages (14–21%).2–4

Anterior uveal melanoma (AUM) affecting the iris or ciliary body is less common than choroidal melanoma.1 However, due to the close anatomical relationship of AUM to the trabecular meshwork, it is more commonly associated with elevated intraocular pressure (IOP).3

Treatment modalities of AUM include surgical tumour excision, irradiation and enucleation.1 Tumour irradiation has become the most frequently used treatment option.4–7 Proton beam radiotherapy (PBRT) provides a high degree of treatment precision, resulting in very low tumour recurrence rates.4–8 Recognised complications of PBRT include: cataract, chronic uveitis, hyphema, macula oedema, retinal detachment, corneal melt, scleral atrophy and secondary glaucoma.9 ,10 The latter is the principal cause of enucleation in eyes with AUM treated with PBRT.6 ,9 ,11–13

Ocular hypertension following PBRT is likely to be due to a direct effect on the iridocorneal angle, resulting in trabecular meshwork inflammation and scarring.2 ,14 This may compound ocular hypertension secondary to tumour infiltration. Glaucoma treatment modalities include anti-glaucoma medications, surgery and ciliary body ablation, which have limited success.3 ,9 ,13 ,14 Concerns that intraocular surgery may increase tumour seeding rates13 ,15 ,16 have traditionally restricted the role of glaucoma surgery in AUM patients to selected cases.7 ,9 ,15 ,17 Tan et al17 recently reported a single case of Baerveldt shunt implantation following PBRT for AUM.

This article describes the first prospective series of Baerveldt aqueous shunts in eyes with glaucoma secondary to AUM, treated by complete anterior segment irradiation (PBRT).


Patients with medically uncontrolled IOP following anterior segment PBRT for AUM were prospectively recruited, following informed consent, to undergo Baerveldt shunt implantation between January 2009 and September 2011 in the glaucoma unit of Jules-Gonin Eye Hospital. The study was conducted in accordance with the Declaration of Helsinki and ethics committee approval (ethical committee of the Canton of Vaud, Switzerland) was obtained. All eyes had AUM, treated with complete anterior segment PBRT (three eyes had tumour extending into the anterior choroid). Limbal stem cell autografts (3–6 clock hours) were harvested before PBRT, stored for up to 7 days in Eusol-C medium (Alchimia Srl, Padova, Italy) and re-planted following irradiation. Aqueous shunt surgery was then performed as part of a local multidisciplinary care package aimed at increasing ocular retention rates in AUM patients, including those with poor visual potential and ocular pain due to elevated IOP.

Data were recorded for age, sex, tumour location and diameter, irradiation dose, time between PBRT and tube surgery, pre and postoperative glaucoma medications, pre and postoperative IOP, visual acuity and all postoperative complications. Exclusion criteria were previous retinal detachment surgery (encircling buckle) and/or silicone-filled eyes.

Surgical procedure

Baerveldt shunts were implanted by a single surgeon (ES), under general anaesthesia. A 7.0 silk traction suture was placed in the superotemporal cornea. A 3-clock-hour conjunctival peritomy was performed in the superotemporal quadrant. The superior and lateral rectus muscles were identified and freed from the surrounding tenons. The tube plate was placed under the bellies of the recti and secured to the sclera with 9.0 prolene sutures. A Supramid 3.0 suture (S. Jackson Inc., Alexandria, Virginia, USA) was used as an intraluminal occluding stent and was inserted into the silicone tube until it reached the distal end, which was trimmed in a bevel-up configuration. The stented distal tube was inserted into the anterior chamber through a tight scleral canal, created using a 25-gauge needle, posterior to the limbus. The silicone portion of the tube was secured to the sclera with two 9.0 prolene sutures. The tube was generally left unligated. If aqueous egress was detected at the tube plate, then a 10.0 nylon ligature suture was tied around the proximal stented tube to halt aqueous flow. Viscoelastics were never injected into the anterior chamber. Processed human pericardium (Tutoplast pericardium; IOP Ophthalmics, Costa Mesa, California, USA) was placed over the extraocular portion of the tube and secured with one 10.0 nylon overlying mattress suture. A rectangular 4 mm window in the pericardial patch was created over the proximal portion of the silicone tube enabling argon laser suture lysis of the nylon ligature postoperatively. The conjunctiva was closed with 10.0 nylon sutures at the limbus and with 8.0 vicryl peripherally. Subconjunctival cefuroxime and beclamethasone were routinely administered.

Postoperative protocol

Follow-up visits were performed on day 1; weeks 1, 3, 6, 9; months 3, 6, 9, 12 and annually thereafter or more frequently if clinically indicated. Postoperative management included unpreserved topical dexamethasone (Dexafree UD 0.1%; Théa Pharma SA, Switzerland) eight times a day and ofloxacin (Floxal UD, Bausch and Lomb, Zug, Switzerland) four times a day. Topical steroids were tapered over 6–8 weeks. Argon laser suture lysis of the nylon ligature suture was performed using a Hoskins lens 1 month or more postoperatively if IOP lowering was clinically required. If sufficient IOP reduction was not obtained after 12 weeks, the stent was either retracted partially (5 mm) within the tube (partial stent removal) or completely removed (complete stent removal). Partial stent removal was performed as a first procedure when there were risk factors for hypotony such as iris neovascularisation or uveitis or when IOP had already dropped to the low teens but was clinically deemed to be too high. Complete stent removal was performed if the IOP dropped to the high teens yet a lower target IOP was required; or when partial stent removal did not provide adequate IOP lowering.

Stent removals

These were all performed under topical anaesthesia (Minims tetracain hydrochloride 1%; Chauvin Pharmaceuticals Ltd, Romford, UK). The conjunctiva was incised over the anterior portion of the stent, which was either retracted 5 mm and re-sutured to the sclera (partial stent removal) or completely removed (complete stent removal). Balanced salt solution was injected into the anterior chamber in eyes with all stent removals. Viscoelastic (0.05–0–1 ml) was routinely injected into the anterior chamber of eyes undergoing complete stent removal. The conjunctiva was closed using a single 8.0 vicryl suture and subconjunctival cefuroxime and beclamethasone were administered. Tobradex (Alcon, Switzerland) was prescribed four times a day and tapered over 3–4 weeks.


Macroscopic hyphema was considered present when over 1 mm of blood was seen in the anterior chamber. Transient hypotony was defined as IOP of 5 mm Hg or less on two consecutive visits more than 2 weeks apart, lasting 1 month or less and persistent hypotony lasting more than 1 month. The presence of choroidal effusions and maculopathy were assessed by routine dilated fundus examination.

Success criteria

Surgery was considered a complete success when IOP was 21 mm Hg or less and a minimum of 20% reduction from baseline was achieved without glaucoma medications. Partial success was defined by the same criteria, using glaucoma medications. Failure was defined by one or more of the following criteria: IOP over 21 mm Hg or reduced by less than 20% below baseline on two consecutive follow-up visits after 3 months, IOP of 5 mm Hg or less on two consecutive follow-up visits or additional glaucoma surgery.8

Statistical analysis

Analysis was performed using Matlab 2010 V.7.10. When a statistical test was used, the test type is specified.


Baerveldt shunts were implanted in 31 eyes. One eye was enucleated following a PBRT-related complication: painful pre-existing, non-healing corneal erosion, three eyes were lost to follow-up (international patients). The baseline pre and postoperative characteristics of the 31 eyes are summarised in table 1. Three eyes had previously undergone filtration surgery that failed due to excessive bleb scarring and preoperative IOP was elevated despite maximal medication.

Table 1

Baseline characteristics

There was no significant change in visual acuity following glaucoma surgery; mean preoperative visual acuity was 0.30 (±0.28) and postoperatively 0.25 (±0.25) (p=0.17, Wilcoxon paired t test). Eight patients had poor visual acuity (less than counting fingers) preoperatively and this remained unchanged postoperatively. Ten patients (32%) had a visual acuity of over 0.5 preoperatively and 11 patients (35%) postoperatively.

Mean preoperative IOP decreased after surgery by 44.4% (mean 15.0±5.0 mm Hg) at the last visit. The mean reduction in IOP was 30.5% (mean IOP 21.6±8.1 mm Hg) at 3 weeks, 47.4% (mean IOP 16.3±6.8 mm Hg) at 3 months, 52.1% (mean IOP 14.8±7.0 mm Hg) at 6 months, 59.8% (mean IOP 12.5±4.4 mm Hg) at 12 months (figure 1A, table 1). The reduction in IOP was statistically significant between the preoperative and all postoperative time points (p<0.05, Wilcoxon paired t test). Mean IOP decreased gradually following week 1 and remained between 10 and 20 mm Hg between week 3 and month 3 (figure 1). Following month 3, the intraluminal stent was removed in 26 of 31 study eyes (average time to removal 17.5 weeks), producing a further significant reduction of IOP (between week 9 and months 6, 9 and 12, p<0.05, Wilcoxon paired t test). IOP did not change significantly between months 6 and 9 (p=0.24); or 9 and 12 (p=0.87). At 12 months, the success rate was 86% using glaucoma medications and 50% without glaucoma medications, only one eye had IOP greater than 21 mm Hg (23 mm Hg) no other eyes met the failure criterion.

Figure 1

(A) Summary of postoperative intraocular pressure (IOP) measurements throughout follow-up. The diamond markers represent the mean IOP throughout follow-up. At each time point a box and whisker plot shows the range of IOP values recorded, the height of the box represents lower and upper IQR (50% of the range), the ends of the whiskers extend to 95% of the range (2.5–97.5%). The horizontal lines show the upper and lower limits associated with successful surgery (ie, IOP 21 mm Hg and 5 mm Hg). (B) Summary of the number of glaucoma medications used throughout the study period. Here the box and whisker plot shows the range of glaucoma medications prescribed, with the diamond markers indicating the mean number.

At the last visit, the mean reduction in the number of glaucoma medications was 76%, with a mean of 0.7 glaucoma medications per patient. Only 12 eyes (39%) required glaucoma medications. The median number of glaucoma medications prescribed preoperatively and at week 3 were equivalent. Following stent removal (at 17.5 weeks ±15.6 weeks) there was a significant reduction in glaucoma medications, with a median of zero thereafter.

There were no intraoperative complications, four eyes had postoperative complications, one shallow anterior chamber, one tube erosion and two transient hyphemas (in eyes with iris neovacularisation). None of the complications were visually significant. Hypotony (transient or persistent) did not occur at any time points.


In recent years, a rise in implantation rates of glaucoma aqueous shunts has been observed.18–21 In this study, the majority of Baerveldt shunts (90%) were performed as primary glaucoma surgery, achieving surgical success in 86% of eyes at 12 months. The IOP reduction was gradual and sustained, and a hypertensive phase was not observed following stent removals. Enucleation was never carried out for uncontrolled glaucoma, which has important implications, given that intractable glaucoma is the most common cause of enucleation following PBRT.4 ,5 ,8 ,12 ,13

PBRT is well established in the treatment of AUM achieving very low tumour recurrence rates (<1%).5 However, it often results in a poor ocular surface, chronically inflamed conjunctiva and dry eye symptoms.22 Study eyes therefore underwent limbal–conjunctival autografting to promote a healthy ocular surface following irradiation.

Aqueous shunts were preferred over perilimbal filtration surgery as they produce postequatorial blebs in tissues that are relatively spared from irradiation-induced damage. During shunt implantation, it was often observed that the irradiated perilimbal conjunctiva was markedly inflamed and the underlying sclera was thin. Double thickness pericardial patch grafts were therefore routinely used to protect the anterior tube within the tissues exposed to PBRT. There were few minor postoperative complications; tube erosion occurred in only one eye, there were no shunt plate exposures or tube retractions. Study eyes did not become hypotonous following stented Baerveldt 350 shunt implantation despite anterior segment irradiation involving the ciliary body. This is of particular importance as damage following irradiation22 is likely to make eyes more vulnerable to hypotony-related complications.

Concerns regarding tumour seeding following intraocular/glaucoma surgery still exist.15 ,16 ,23 As glaucoma surgery is often the only alternative to enucleation, it is performed in selected cases.9 ,15 ,17 Enucleation has not been associated with reduced death rates when compared with brachytherapy or PBRT,24 ,25 suggesting that metastases are largely independent of treatment modality, having occurred early and before treatment.25 In addition, total anterior segment irradiation using PBRT results in very low tumour recurrence rates,4–7 and approaches zero utilising local treatment protocols (L. Zografos, unpublished data, October 2008). These data suggest that glaucoma surgery performed in irradiated eyes will not affect systemic safety profiles. In study patients, there were no tumour recurrences or metastases. All patients received routine 6-monthly ocular anterior and posterior segment clinical examinations, photographs, ultrasound biomicroscopy scans with tumour measurements and systemic investigations including liver ultrasound scans and blood tests. Longer follow-up and greater numbers are required to determine whether systemic metastatic rates of shunted eyes differ from those of non-shunted eyes.


Baerveldt aqueous shunts were effective in lowering IOP in eyes that had received total anterior segment irradiation for AUM. Shunt-related complications were infrequent and visually insignificant. In contrast to previous studies of secondary glaucoma in AUM,3 there were no enucleations due to ocular hypertension/glaucoma. The high success rate (86%) suggests that tube implantation may also reduce glaucomatous visual loss in these relatively young patients with useful remaining vision.



  • Data presented in part at the Association for Research in Vision and Ophthalmology (ARVO) Annual Meeting, 2011.

  • Competing interests None.

  • Patient consent Obtained.

  • Ethics approval Ethics approval was provided by the ethical commitee of the Canton Vaud, Switzerland.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Linked Articles

  • At a glance
    Michelle English