Long‐term limitations and complications of trans‐scleral diode laser cycloablation for refractory glaucoma

Trans‐scleral diode laser cycloablation (cyclodiode) is effective in the short‐term management of refractory glaucoma where alternative treatments are not feasible. Long‐term outcomes of 5‐years or more are not well‐documented, particularly in relation to intraocular pressure (IOP) control, need for further procedures and complications such as hypotony and phthisis.


| INTRODUCTION
Trans-scleral diode laser cycloablation (cyclodiode) is an established therapeutic modality used in the acute treatment of refractory glaucoma. Cyclodiode was first conceptualised and described by Weekers et al. in 1961 and has been used in clinical practice since 1969. 1,2 Other methods of laser cycloablative procedures include transpupillary cyclophotocoagulation, non-contact and contact ND:YAG laser, micropulse transscleral cyclophotocoagulation and endoscopic cyclophotocoagulation. 1,3 Cyclodiode lowers the IOP through ciliary body destruction as infrared energy is absorbed by melanin pigment in the ciliary epithelium with subsequent reduction of aqueous humour production.
Cyclodiode is often employed in the treatment of primary open angle glaucoma (POAG), neovascular glaucoma (NVG) and inflammatory glaucoma, where patients have persistently high IOP despite treatment with maximal medical therapy and experience pain. It is generally reserved for patients with limited visual prognosis, however, may be useful in patients with some remaining vision, who are poor surgical candidates. Potential complications that have been reported include hypotony, phthisis, cystoid macular oedema, uveitis, hyphaema, vitreous haemorrhage, scleral perforation and visual loss. 1,[4][5][6] Few studies have followed patients who have undergone cyclodiode for 5 years or longer. [7][8][9][10][11][12][13] The purpose of this study is to evaluate the long-term limitations and complications of cyclodiode so that clinicians may have a better guide on potential outcomes.

| METHODS
A retrospective cohort study was conducted investigating the medical records of all patients who underwent cyclodiode for refractory glaucoma at City Eye Centre in Brisbane, Australia from 2012 to 2016. The procedures were performed by a single glaucoma specialist (GL) and all patients had a minimum follow-up period of 5 years. Exclusion criteria included patients who were lost to follow-up, followed up elsewhere, inadequate follow-up period of less than 5 years, deceased within the follow-up period or underwent evisceration prior to the end of the follow-up period. Ethics approval was granted from the University of Queensland's Hospital Human Research and Ethics Committee (2020/HE001035) and the study was carried out in adherence to the tenets of the Declaration of Helsinki.
The indication for cyclodiode treatment was uncontrolled IOP despite maximally tolerated medication, failure of previous filtering or tube surgery, or attempting to wean off acetazolamide. The parameters studied were the sex, age, laterality, type of glaucoma, cyclodiode parameters, number of glaucoma medications including acetazolamide, IOP and visual acuity. Data were recorded before, and at yearly intervals after cyclodiode treatment.
The procedure was performed under a peribulbar block (0.5% bupivacaine hydrochloride) and cyclodiode was performed using a transscleral contact fiberoptic G probe (OcuLight SLx, Iris Medical, Mountain View). The probe was positioned at, or slightly behind, the limbus. Prior to the cyclodiode, the position and condition of the ciliary body was ascertained at the slit lamp by transpupillary retro-illumination ( Figure 1).
Each laser burn was spaced by half the width of the G-probe. Laser burns were not applied to the three and nine o'clock positions to avoid damage to the ciliary vessels and nerve. The number of laser burns varied, depending on the patient's IOP and whether they were on acetazolamide.
Following the procedure, chloramphenicol ointment and a double eye pad was applied. The next day the patient commenced Prednisolone acetate 1%/phenylephrine HCl 0.12% (Prednefrin Forte, Allergan) four times a day for a period of 2 weeks. If the patient was on acetazolamide prior to treatment, this was continued for a further week and ceased if IOP was <25 mmHg. Patients were reviewed at 1 week, 1 month and every 6 to 12 months thereafter post cyclodiode treatment. Complications such as hypotony was defined as an IOP of ≤5 mmHg at final follow-up and phthisis was defined as a visible shrinkage of the globe associated with hypotony.
The outcome of cyclodiode intervention was analysed and considered successful if the final IOP was >5 and <25 mmHg, or if a patient experienced a >30% reduction from pre-treatment IOP or with a reduction in the number of glaucoma medications and cessation of acetazolamide. Data were tabulated using Microsoft Excel 2011 (Microsoft Corporation) and statistical analysis performed using IBM SPSS statistics for Windows (version 26.0 IBM Corporation). Pre-treatment and posttreatment mean IOP and mean number of glaucoma medications were compared with paired sample Student t tests. Kaplan-Meier analysis with the Log-rank (Mantel-Cox) test of equality was used to compare survival distributions with respect to age, sex, glaucoma type, number of treatments, and power. Generalised linear models were used to examine the association between final follow-up visit outcome measures (IOP, percentage reduction from pre-treatment IOP, number of medications, reduction from pre-treatment number of medications) and a range of predictor variables including age, sex, glaucoma type, power, number of treatments, pre-treatment IOP and the number of pre-treatment medications.

| RESULTS
Cyclodiode was performed on 98 patients with refractory glaucoma at City Eye Centre, Brisbane from 2012 to 2016. There were 44 (44.9%) patients excluded from this study with the reasons including follow-up in other centres closer to their geographic locations 20 (20.4%), did not have an adequate follow-up period 18 (18.4%), deceased 5 (5.1%) or had an evisceration before the end of the follow-up period 1 (1.0%).
The mean pre-treatment IOP was 31.5 mmHg (range 17-56 mmHg), and the mean IOP at 5 years posttreatment was 16.1 mmHg (2-42 mmHg). This represented a mean decrease in IOP of 15.1 mmHg at the end of the 5-year period. The mean number of pre-treatment medications was 3.6 (range 1-6), including 9 (16.7%) on oral acetazolamide. At 5 years post-treatment, the mean number of medications was 2.7 (range 0-5), including five patients (8.3%) remaining on oral acetazolamide.
The mean baseline best-corrected visual acuity (VA) pre-treatment was 6/60. At the final follow-up visit, the mean post-treatment VA was 6/240. A reduction in VA was experienced in 37 (69%), stable VA in 12 (22%) and 5 (9%) displayed improved VA post-treatment. Other complications of the cyclodiode were hypotony 3 (5.6%), phthisis 2 (3.7%), uveitis 1 (1.9%) and cystoid macular oedema 1 (1.9%). Of the five patients with hyptony and phthisis, three were neovascular glaucoma and over 78 years of age. Complications related to advanced disease were evisceration 1 (1.9%) and enucleation 1 (1.9%) secondary to an ocular surface squamous neoplasia. Five patients deceased within the 5 year follow up period and not included in this study. Table 1 displays the percentage of patients who achieved treatment success based on the various criteria at each annual follow-up visit (non-cumulative data) and the final follow-up visit. At the final follow-up visit 43 (79.6%) had an IOP >5 and <25 mmHg, or a >30% reduction from the pre-treatment IOP, 36 (66.7%) had a reduction in the number of pre-treatment medications, and 25 (46.2%) of patients met both success criteria. Figure 1 displays the cumulative success over time for each success criterion.
Kaplan-Meier analysis revealed there were no significant differences in the survival distribution curves (i.e., time to failure or duration of success) with respect to age, sex, glaucoma type, number of treatments, or power (all p > 0.05 for each success criterion) ( Table 2). These findings would indicate cyclodiode is equally efficacious over the age range of this study, in males and females, in primary and secondary glaucomas using the laser parameters according to this study's protocol ( Figure 2). Generalised linear modelling (Table 3) revealed significant associations between the final follow-up visit IOP and the percentage reduction in pre-treatment IOP with age, when controlling for other predictor variables. On average, the final IOP was 2.2 mmHg less per decade increase in age (95% CI 0.7 to 3.6 mmHg less) (p = 0.003) and the percentage reduction in pre-treatment IOP was 6.9% greater per decade increase in age (95% CI 2.2 to 11.6% greater reduction) (p = 0.004).

| DISCUSSION
The results of this study provide data for the use of cyclodiode laser treatment in Australian patients with refractory glaucoma, specifically where medical management has been maximised and the patient is unable to undergo surgery. Many patients with advanced glaucoma have persistently high IOP (despite multiple medications), poor visual acuity and ocular pain. Cyclodiode laser can provide a long-term management option for reducing the IOP, medication burden and symptoms, however there are significant potential complications.
There is variable long-term worldwide data with similar patient cohorts, demonstrating mean follow-up durations of 12.5 to 42 months. 7-13 A recent UK study followed patients for 5 years post-treatment, however, the patient cohort was focused on glaucoma patients following penetrating keratoplasty (PK). 14 In the study by Aygun et al in Turkey, the success rate was considered a final IOP of <21 mmHg, or greater than 20% reduction in IOP post-treatment. 8 Based on these criteria the success rate was reported to be 66.67%, increasing to 86.7% after repeat treatment, with over half of the cohort requiring repeat treatment. 8 There appears to be a potential loss of efficacy over time, whereby the greater number of years from the cyclodiode treatment, the less efficacious is the treatment. This was seen in Zekhov et al.'s patients with 6.3% of patients not maintaining the IOP reduction achieved at 6 weeks, at the end of the 3-year review period. 13 A loss of efficacy over time is contributed by the progressive reduction in outflow facility.   Numerous protocols exist for the use of cyclodiode laser, although standardised treatment protocols have been advocated. 15 In 2011, a survey of UK ophthalmologists performing cyclodiode laser found great variability among surgeons. 16 Overall, applications used included an average power of 1500 mW and 2000 ms over 25.5 applications. 16 In addition to protocol variation, whether the use of one, or multiple, treatments to obtain significant reduction and longevity of reduced IOP is variable among the literature. Repeat cyclodiode treatment was required in 53.3% of participants in Turkish study, while 76% to 85% required additional sessions in a Swiss population. 8,11 Significant complications, such as uveitis, hypotony and phthisis have been reported following cyclodiode laser. In this review, 11.1% of patients suffered adverse outcomes. Other studies demonstrated similar or higher complication rates of earlier research, from 5.3% to 18%. 6,[8][9][10]12,13,17 Reports of complication rates being associated with neovascular glaucoma and increased total energy used for laser have been reported. 6,10,13,17 All patients considered for cyclodiode treatment must be assessed from a risk perspective. It is inherent that any medical or surgical intervention has risks and benefits, however, with careful patient selection these can potentially be minimised. An interesting finding in this study was cyclodiode in older patients appears to be more efficacious. This may also be related to longer-term failure of IOP control in younger patients, as there may be incomplete loss of the ciliary processes, with sparing within the field of treatment. 18 A reduction in VA was experienced in the majority of patients 37 (69%). Unfortunately this is related to the end stage nature of this type of glaucoma, especially when following patients over a longer period. The aim of the cyclodiode is not necessarily to save the vision, but often to keep the cornea stable and avoid bullous keratopathy as well as prevent ocular pain. Given these findings, patient selection for cyclodiode favours older patients, patients with non-neovascular glaucoma and those with an understanding that vision in the longer term is likely to worsen due to the advanced glaucoma disease. It is also difficult to formulate definitive conclusions on the subtypes of glaucoma due to small sample size of each category.
Unfortunately, due to the retrospective nature of this review, there was a significant proportion of patients (44.9%) lost to follow-up. Almost half of these patients (20.4%) were followed by their local ophthalmologists, however, generally would have been re-referred if there was persisting inadequate intraocular pressure control or complications such as hypotony. The patients who did not have adequate follow up period (18.4%) similarly would likely re-present if they were having ongoing issues. Both these factors would tend to skew the results more in favour of cyclodiode.
Cyclodiode is often utilised for end-stage glaucoma when the IOP is uncontrolled on medical treatment and drainage surgery is not indicated, such as in the presence of conjunctival scarring or significant thinning. It can provide long-term reduction of IOP and reduce the number of medications including acetazolamide. However, there are significant complications including phthisis, that need to be considered and an integral part of the patient informed consent. Standardised treatment protocols are difficult to formulate due to the variable nature of glaucoma aetiology, so laser parameters need to be tailored to each case.