Laser peripheral iridoplasty for chronic angle closure

In at least a third of primary angle closure cases, appositional angle closure persists after laser peripheral iridotomy, and further intervention may be considered. Laser peripheral iridoplasty (LPIp) can be used in treating chronic angle closure when angle closure persists after laser peripheral iridotomy. Previous reviews have found insufficient data to determine its clinical effectiveness, compared to other interventions. This is an update of a Cochrane Review first published in 2008 and updated in 2012. It examines all studies to date to establish whether LPIp shows any effectiveness over other available treatment options.


T A B L E O F C O N T E N T S
Cochrane Database of Systematic Reviews

Findings
Laser peripheral iridoplasty with iridotomy versus iridotomy alone as a primary procedure Two RCTs assessed the use of argon LPIp as a primary procedure with peripheral iridotomy, compared with peripheral iridotomy alone. However, neither study reported data for the primary outcome, disease progression. Argon LPIp showed no evidence of e ect on: final mean intraocular pressure (IOP) at 3 months and 12 months (mean di erence (MD) 0.39 mmHg, 95% confidence interval (CI) -1.07 to 1.85; I 2 = 38%; 2 studies, 174 participants; low-certainty evidence); further surgical or laser intervention at 12 months (risk ratio (RR) 1.21, 95% CI 0.66 to 2.21; 1 study, 126 participants; low-certainty evidence); or mean number of additional medications required at 12 months (MD 0.10, 95% CI -0.34 to 0.54; 1 study, 126 participants; low-certainty evidence). Complications were assessed at 3 to 12 months (2 studies, 206 participants; low-certainty evidence) and found to be mild and uncommon, with comparable levels between groups. The only severe complication encountered was one case of malignant glaucoma in one study's argon LPIp group. Quality of life measures were not assessed. In the other study, investigators found that argon LPIp showed no evidence of e ect on final mean anterior segment optical coherence tomography (AS-OCT) measurements, including anterior chamber depth (MD 0.00 mm, 95% CI -0.10 to 0.10; 24 participants, 48 eyes; very low-certainty evidence).
Laser peripheral iridoplasty as a secondary procedure versus no treatment One RCT assessed the use of argon LPIp as a secondary procedure compared with no further treatment in 22 participants over three months. Disease progression, additional medications required, complications, further surgical or laser intervention, and quality of life outcomes were not assessed. There was only very low-certainty evidence regarding final maximum IOP value (MD -1.81 mmHg, 95% CI -3.11 to -0.51; very low-certainty evidence), with no evidence of e ect on final minimum IOP values (MD -0.31 mmHg, 95% CI -1.93 to 1.31; very lowcertainty evidence). The evidence is very uncertain about the e ect of argon LPIp on AS-OCT parameters. The trial did not report AS-OCT measurements for the control group.

Laser peripheral iridoplasty as a secondary procedure versus medication
One RCT assessed the use of argon LPIp as a secondary procedure compared with travoprost 0.004% in 80 participants over 12 months. The primary outcome of disease progression was reported for this method: argon LPIp showed no evidence of e ect on mean final cup/ disk ratio (MD -0.03, 95% CI -0.11 to 0.05; low-certainty evidence). Argon LPIp showed no evidence of e ect for: mean change in IOP (MD -1.20 mmHg, 95% CI -2.87 to 0.47; low-certainty evidence) or mean number of additional medications (MD 0.42, 95% CI 0.23 to 0.61; lowcertainty evidence). Further surgical intervention was required by one participant in the intervention group alone, with none in the control group (low-certainty evidence). No serious adverse events were reported, with mild complications consisting of two cases of 'post-laser IOP spike' in the argon LPIp group. Quality of life measures were not assessed. The evidence is very uncertain about the e ect of argon LPIp on AS-OCT parameters. The trial did not report AS-OCT measurements for the control group.

Adverse events
Availability of data were limited for adverse e ects. Similar rates were observed in control and intervention groups, where reported. Serious adverse events were rare.

Authors' conclusions
A er reviewing the outcomes of four RCTs, argon LPIp as an intervention may be no more clinically e ective than comparators in the management of people with chronic angle closure. Despite a potential positive impact on anterior chamber morphology, its use in clinical practice in treating people with chronic angle closure is not supported by the results of trials published to date. Given these results, further research into LPIp is unlikely to be worthwhile.

Library
Trusted evidence. Informed decisions. Better health.

Cochrane Database of Systematic Reviews
First, we searched the medical literature for studies that compared the e ects of LPIp to other treatments or no treatment. We then compared the results, and summarised the evidence from all the studies. Finally, we rated our confidence in the evidence, based on factors such as study methods and sizes, and the consistency of findings across studies.

What did we find?
We found four studies on a total of 252 people, mostly from Asia. The studies followed participants for between 3 and 12 months, and compared: -LPIp plus peripheral iridotomy to peripheral iridotomy alone, as a primary treatment (that is, in people who had not received any other treatment for PAC before); -LPIp to no treatment, as a secondary treatment (that is, in people who had previously been treated for PAC, but not with LPIp); and -LPIp to eye drops (travoprost 0.004%), as a secondary treatment.

LPIp plus peripheral iridotomy compared to peripheral iridotomy alone, as a primary treatment
The evidence suggests that adding LPIp to peripheral iridotomy may make little or no di erence to: -eye pressure (2 studies, 174 people); -the need for medicines a er 12 months (1 study, 126 people); -the need for further laser or surgical treatment (1 study, 126 people); and -the shape of the front of the eye (1 study, 48 people).
Evidence provided by two studies suggests that: -unwanted e ects (such as bleeding inside the front of the eye) are uncommon; and -adding LPIp to peripheral iridotomy may make little or no di erence to the frequency of unwanted events.
We do not know if adding LPIp to peripheral iridotomy slows disease progression or improves quality of life because no study investigated this.

LPIp compared to no treatment, as a secondary treatment
We found one study on 22 people that compared the e ects of LPIp to no treatment on: -eye pressure; and -the shape of the front of the eye.
This study was not robust enough for us to determine which treatment is better.
The study did not investigate whether LPIp is better than no treatment to: -slow disease progression; -limit the need for medications; -avoid the need for more laser or surgical treatment; or -improve quality of life.
The study did not investigate unwanted events.
LPIp compared to travoprost 0.004% eye drops, as a secondary treatment The evidence from one study on 80 people suggests that there may be little to no di erence between the e ects of LPIp and travoprost 0.004% on: -disease progression; -eye pressure; -the need for medicines a er 12 months; and Library Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews -the need for further laser or surgical treatment.
The evidence further suggests that: -unwanted e ects are uncommon; and -there may be little or no di erence in the frequency of unwanted events between the two treatments.
The evidence was not robust enough for us to determine whether the treatments have di erent e ects on the shape of the front of the eye.
We do not know which treatment works better to improve quality of life because no study investigated this.

What does this mean?
The evidence suggests that: -LPIp may not be better than other treatments for chronic PAC; and -unwanted events may be as common with LPIp as with other treatments for chronic PAC.

How up-to-date is this review?
The evidence in this Cochrane Review is current to December 2020. The mean final ACD found in the control group was 2.10 mm.

Outcomes
The mean final ACD did not differ significantly between groups, with a mean final ACD of 0.00 mm greater in the intervention group (95% CI -0.10 to 0.10). --

Anterior chamber morphology
Assessed with: gonioscopy or AS-OCT Follow-up: 24 months -- Data reported at 12 months *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). ACD: anterior chamber depth; AS-OCT: anterior segment optical coherence tomography; CI: confidence interval; IOP: intraocular pressure; PAC: primary angle closure; PACG: primary angle-closure glaucoma; PACS: PAC suspect; RR: Risk ratio; OR: Odds ratio.

GRADE Working Group grades of evidence
High-certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate-certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low-certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low-certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
a We downgraded one level for risk of bias because participants, personnel and outcome assessors could not be masked. We downgraded a further level for a combination of imprecision (sample size was less than 400) and indirectness (no data were available at our pre-specified time point of 24 months and so we used available data at shorter followup which may not represent the situation at the longer follow-up). b We downgraded one level for risk of bias because participants, personnel and outcome assessors could not be masked. We downgraded a further level for imprecision due to sparse data. c We downgraded one level for risk of bias because participants, personnel and outcome assessors could not be masked. We downgraded a further level for a combination of imprecision (confidence intervals ranged from 0.66 to 1.21) and indirectness (no data were available at our pre-specified time point of 24 months and so we used available data at shorter follow-up which may not represent the situation at the longer follow-up Moderate-certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low-certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low-certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect. a We downgraded one level for risk of bias because participants, personnel and outcome assessors could not be masked. We downgraded a further level for imprecision (sample size was less than 400) and indirectness (no data were available at our pre-specified time point of 24 months and so we used available data at shorter follow-up which may not represent the situation at the longer follow-up). b We downgraded two levels for risk of bias because participants, personnel and outcome assessors could not be masked, no comparison was made to the control group, and control measurements were not reported. We downgraded a further level for a combination of imprecision (sample size was less than 400) and indirectness (no data were available at our pre-specified time point of 24 months and so we used available data at shorter follow-up which may not represent the situation at the longer follow-up). c Argon laser was used in all instances of LPIp.

Disease progression
The mean final vertical C/D ratio for the control group was 0.59.
The mean final vertical C/D ratio did not differ significantly between groups, with an average of 0.03 Assessed by: dilated assessment of the vertical C/D ratio using a 78 D lens Follow-up: 24 months Units: 0.00 to 1.00, higher indicating progression of the disease lower in the intervention group (95% CI -0.11 to 0.05).

IOP
Assessed with: Goldmann tonometry Follow-up: 24 months Units: mmHg, lower is better The mean change in IOP found in the control group was a reduction of 6.10 mmHg.
The mean change in IOP did not differ significantly between groups, with an average reduction of 1.20 mmHg lower in the intervention group (95% CI -2.87 to 0.47). Data reported at 12 months

Follow-up: 24 months
There was one case of surgical intervention required in the intervention group (n = 40), compared with no cases in the control group (n = 40).

Low b
Data reported at 12 months

Follow-up: 24 months
No data available

Anterior chamber morphology
Assessed with: AS-OCT images for AOD, TISA, ARA, and mean angle width

10
Units: mm for AS-OCT measurements AS-OCT: anterior segment optical coherence tomography;IOP: intraocular pressure; PAC: primary angle closure; PACS: suspected PAC; PI: peripheral iridotomy GRADE Working Group grades of evidence High-certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate-certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low-certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low-certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect.
a We downgraded one level for risk of bias because participants, personnel and outcome assessors could not be masked. We downgraded a further level for a combination of imprecision (sample size was less than 400) and indirectness (no data were available at our pre-specified time point of 24 months and so we used available data at shorter followup which may not represent the situation at the longer follow-up). b We downgraded one level for risk of bias because participants, personnel and outcome assessors could not be masked. We downgraded a further level for imprecision due to sparse data. c We downgraded two levels for risk of bias because participants, personnel and outcome assessors could not be masked, no comparison was made to the control group, and control measurements were not reported. We downgraded a further level for a combination of imprecision (sample size was less than 400) and indirectness (no data were available at our pre-specified time point of 24 months and so we used available data at shorter follow-up which may not represent the situation at the longer follow-up). d Argon laser was used in all instances of LPIp. Cochrane Database of Systematic Reviews

Description of the condition
'Angle closure' refers to the occlusion of the trabecular meshwork by the peripheral iris, preventing aqueous outflow (Kanski 2011). Chronic appositional contact between these two structures is thought to allow adherence, known as peripheral anterior synechiae (PAS), which is one of the hallmarks of primary angle-closure glaucoma (PACG) (Lee 2006). Other indicators, such as elevated intraocular pressure (IOP) are also used in the identification of the disease. In 2002, Foster and colleagues proposed a classification system for categorising cases of angle closure (Foster 2002). The American Academy of Ophthalmology adopted their system in 2015 (Prum 2016), and we use it throughout this review (see Table 1.) Glaucoma is a group of disorders which shows a distinctive and potentially progressive optic neuropathy, characterised by structural damage to the optic nerve and associated visual field loss (Kanski 2011). Glaucoma is the second leading cause of blindness worldwide. The number of people with glaucoma was estimated to be 76 million in 2020 and is expected to increase to 111.8 million by 2040 (Tham 2014). A quarter of cases are due to PACG. While open-angle glaucoma is more common than PACG, blindness is a more likely consequence of PACG (Tham 2014). Prevalence of PACG is higher in East Asians and Chinese populations and in women. In European populations, the prevalence of ACG is 0.4% of the adult population (Day 2012). PACG presents a serious health concern worldwide, and over 32 million cases are predicted by the year 2040 (Tham 2014).

Treatment of PAC
The main aims of treatment in people with chronic angle closure are to remove the underlying pathophysiological mechanism inducing closure and to reduce intraocular pressure (IOP) (Amerasinghe 2008). Currently, the first-line treatment for primary angle closure (PAC) is Nd:YAG laser peripheral iridotomy (NICE 2016), which involves the use of a laser to make a single hole in the peripheral iris (Le 2018). Peripheral iridotomy is thought to resolve the mechanism of 'pupillary block', whereby the pupillary iris makes a seal with the lens posteriorly, preventing the flow of aqueous humour into the anterior chamber (Gazzard 2003). Peripheral iridotomy was first developed in 1956. Since then, it has gone through a range of changes in modality (Khuri 1973;Pollack 1984).
In at least a third of cases of PAC, the anterior chamber angle remains appositionally closed a er an iridotomy. The use of secondary line treatment options to widen the anterior chamber angle, such as laser peripheral iridoplasty (LPIp) or lens extraction (NICE 2016), is then considered. LPIp, originally coined 'laser gonioplasty', was first introduced in the 1970s (Kimbrough 1979), and can potentially be done with other types of laser that have a thermal e ect. Its e icacy in the treatment of chronic angle closure is the focus of this review.

Description of the intervention
LPIp involves the placement of contraction burns in the extreme iris periphery. One or two burns are placed at each clock hour position circumferentially using an iridotomy lens (Kanski 2011). Precise parameters used appear to vary, with spot size ranging from 150 to 500 µm and power from 100 to 300 mW, for a duration of 0.4 to 0.5 s (Babighian 2018; Kanski 2011; Ramakrishnan 2016). Various laser modalities have been used, such as argon, krypton and diode lasers (Huang 2015;Muller 2016;Sassani 1993).

How the intervention might work
Several mechanisms have been proposed to explain the way in which LPIp opens the anterior chamber angle. The classic model suggests that the laser causes contraction of the iris stroma peripheral to the burn, between the burn and the trabecular meshwork, physically pulling open the angle (Ritch 1992;Ritch 2007). Histopathological reports further describe this as a result of the heat shrinkage of collagen, followed by contraction of a fibroblastic membrane formed at the site of injury (Sassani 1993). More recently, studies have placed emphasis on the cross-sectional thinning of the iris stroma at areas subjected to laser, which could contribute to a wider angle recess (Liu 2013). While peripheral iridotomy primarily acts by resolving pupillary block, LPIp could help to resolve chronic angle closure from other mechanisms including plateau iris syndrome (Gazzard 2003;Ritch 2007).

Why it is important to do this review
LPIp is typically used when a primary peripheral iridotomy does not successfully open an angle in PAC disease (Peng 2011). Previous systematic reviews have found inconclusive evidence regarding the use of LPIp in the treatment of chronic PAC cases. The previous version of this review was only able to analyse one randomised controlled trial (Ng 2012). This update aims to review the available literature to establish whether any further conclusions can be drawn.

O B J E C T I V E S
To assess the e ectiveness of laser peripheral iridoplasty in the treatment of people with chronic angle closure, when compared to laser peripheral iridotomy, medical therapy or no further treatment.

Types of studies
We included only randomised controlled trials (RCTs) in the review.

Inclusion criteria
Participants must have had either suspected or confirmed primary angle closure (PACS and PAC respectively), or primary chronic angle-closure glaucoma (PACG).
No restrictions were applied with regards to gender, age or ethnicity of participants.
Participants were not excluded due to a history of resolved acute angle closure or previous laser peripheral iridotomy.

Library
Trusted evidence. Informed decisions. Better health.

Types of interventions
We included RCTs which compared the use of iridoplasty with medical treatment, laser peripheral iridotomy, or no further treatment. We also included RCTs which assessed a combined intervention of both laser peripheral iridotomy and peripheral laser iridoplasty at a single sitting.
Included trials focused on the use of iridoplasty in either the primary or secondary management of chronic angle closure (defined as people diagnosed with PACS, PAC or PACG).

Comparators
The comparators used were laser iridotomy alone, pharmaceutical treatment or no treatment.

Primary outcomes
Percentage of cases found to have disease progression at two years, defined as either: a) change in stage of the disease (PACS to PAC to PACG); or b) worsening of glaucomatous damage among participants with PACG. As highlighted in Table 1, disease progression is based on clinical examination, including IOP and gonioscopy for PACS to PAC; and structural assessment of the optic disc and visual field testing for PAC to PACG.

Secondary outcomes
• IOP (intraocular pressure -measured using Goldmann applanation tonometry) • The number of IOP-lowering medications required • Complication rate and type (e.g. hyphema, persistently elevated IOP, change in corneal endothelial cell count (CECC), change in best corrected visual acuity (BCVA), etc.) • Further surgical or laser interventions (e.g. a filtering procedure or phacoemulsification) • Quality of life measures (e.g. National Eye Institute Visual Function Questionnaire (NEI-VFQ) or EuroQOL -5 Dimension (EQ-5D) instrument, as reported in the trial) • Anatomical change in the anterior chamber (measured clinically or using imaging techniques, such as anterior segment optical coherence tomography, AS-OCT)

Search methods for identification of studies Electronic searches
The following resources were searched. There were no restrictions to language or year of publication. The date of the search was 20 December 2020. Searches had no language restriction but only English language studies were included. The unrestricted search revealed one Korean language study as a possible inclusion.

Searching other resources
We did not carry out formal checking of topic-specific journals. However, on a cursory search of the journal in which the Korean language study was published, we also found another study. Both of the Korean studies have been noted in the Discussion section of this review, but are awaiting full assessment.

Data collection and analysis
Throughout data collection and analysis, we used standard methods proposed by Cochrane.

Selection of studies
One review author first removed duplicate papers, a er which two review authors independently assessed the titles and abstracts of the remaining discrete papers. We labelled each record as 'relevant', 'possibly relevant' or 'not relevant', and we put aside the 'relevant' and 'possibly relevant' papers for full-text review. Discrepancies were discussed between the two review authors where required, and a joint decision made.
The full-text copies of the papers which passed initial screening were then read independently by two review authors to ascertain whether they met the inclusion criteria. These were marked as 'Yes', or 'No', with discrepancies discussed and resolved between the two authors.
We generated a populated PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) flowchart (Moher 2009), which presents the study selection process (Figure 1).

Library
Trusted evidence. Informed decisions. Better health.
Cochrane Database of Systematic Reviews  Cochrane Database of Systematic Reviews

Data extraction and management
One review author extracted the relevant data from each included study using extraction tables created with guidance from Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2019, herea er referred to as the Cochrane Handbook). Outcome data were extracted as specified in the Methods section. A second review author subsequently checked the extracted data.

Assessment of risk of bias in included studies
Two review authors independently used Cochrane's Risk of Bias 1 tool (RoB 1) to assess the quality of the included RCTs (Higgins 2019). RCTs were rated as either 'low risk', 'unclear risk, or 'high risk', with details for each parameter included in the relevant Characteristics of included studies table. Any discrepancies were discussed between the two review authors and an agreement reached. RCTs were assessed for the following criteria: use of random sequence generation and allocation concealment (selection bias); blinding (masking) of participants and researchers (performance bias); masking of outcome assessment (detection bias); incomplete outcome data (attrition bias); and selective outcome reporting (reporting bias).

Measures of treatment e ect
Measurement of treatment e ect was determined using the risk ratio measure for dichotomous outcome variables, while the mean di erence measure was used for continuous outcome variables. All results were presented with 95% confidence intervals. It was only possible to calculate a pooled measurement of e ect estimate for one outcome (Figure 2).

Unit of analysis issues
The unit of analysis was the participant, and was not a major concern for this review as very little meta-analysis could be performed. Lee 2011 was the only study to use a within-person RCT design, randomising one eye to argon LPIp with peripheral iridotomy, and the other to peripheral iridotomy alone. The authors ensured that both eyes were classified as having a similar level of disease (both eyes were categorised as PACS as part of their inclusion criteria) before a random selection between the eyes was made to determine the intervention and control eyes. Baseline parameters for eyes from each group were given in their table 1, and groups were well matched. It is unclear, however, whether any statistical correction was performed to mitigate the risk of bias with taking repeated measurements.
In order to deal with cross-over studies, cluster trials or multiple treatment arms, we followed the recommendation given in Chapter 8 of the Cochrane Handbook (Higgins 2019).

Dealing with missing data
We planned to handle missing data in accordance with the Cochrane Handbook recommendations. Some data which investigators planned to collect in two of the included RCTs were not included in their results section. For example, Narayanaswamy 2016 did not include change in anterior chamber depth (ACD) at 12 months, and Bourne 2017 did not include P values for some of the trial's IOP results. However, we decided that this information was not key to the review. There was enough information available to reach our conclusions, and thus we decided that contacting trial authors about these data was unnecessary.

Assessment of heterogeneity
We assessed heterogeneity by examining the characteristics of the included studies, in order to determine whether a meta-analysis could be carried out or whether a narrative synthesis would be more appropriate. As many outcome measures could not be pooled (see Data synthesis), Chi 2 and I 2 values could not o en be calculated. A Chi 2 value of 1.62 and I 2 value of 38% (P = 0.20) was generated for the final mean IOP outcome of those studies which assessed LPIp with peripheral iridotomy, versus peripheral iridotomy alone as a primary procedure, suggesting a low level of heterogeneity between these studies. Therefore, a fixed-e ect model of meta-analysis was used.

Assessment of reporting biases
As the number of included studies was fewer than 10, we could not use a funnel plot using standard error values to assess the risk of publication bias.

Data synthesis
We could not perform a quantitative synthesis (meta-analysis) for many of the outcomes assessed, as there were not enough relevant high-quality RCTs available with su icient homogeneity between studies. Several distinct methods of assessing the intervention against a comparator were identified, and separated into appropriate subsections. We performed a narrative synthesis of the majority of outcomes.
A su icient level of homogeneity was present in the studies which assessed LPIp with peripheral iridotomy, versus peripheral Cochrane Database of Systematic Reviews iridotomy alone as a primary procedure, to allow a pooled estimate for the final mean IOP to be generated. Adverse e ects were also reported by these studies and pooled estimates for these can be found in Table 2.

Subgroup analysis and investigation of heterogeneity
Both a high level of clinical diversity and methodological diversity was evident between the included RCTs, resulting in an inability to perform meta-analysis on the majority of reported outcomes. In total, 15 discrete outcome variables were reported with a high disparity in those reported by each study (see Table 3). For example, many di ering variables were used to assess anterior chamber morphology, and did not lend themselves to the generation of pooled estimates.
In addition, as previously mentioned, several di erent comparators and methods of assessing the intervention were utilised, and we decided it was inappropriate to synthesise these results cumulatively. Instead, the results were separated into three groups and discussed individually. Subgroup analysis could not be performed due to a paucity of available data.

Sensitivity analysis
We planned for sensitivity analyses to be conducted to assess the influence of a high risk of reporting bias (found using the RoB1 tool), unpublished data, and industry-funded data (funding bias) on e ect sizes. However, this could not be done due to the small number of trials, with di erent comparators, making meta-analysis unfeasible for the majority of outcome measures.

Summary of findings and assessment of the certainty of the evidence
We used the GRADE system to assess the quality of the evidence (see Quality of the evidence). We followed the methods outlined in Section 8.5 and Chapter 12 of the Cochrane Handbook (Higgins 2019, Higgins 2020), regarding the use of GRADEpro so ware (GRADEpro GDT).
We included a 'Summary of findings' table for each of the following three methods of assessing the intervention.
• Laser peripheral iridoplasty with iridotomy versus iridotomy alone as a primary procedure. • Laser peripheral iridoplasty as a secondary procedure versus no treatment. • Laser peripheral iridoplasty as a secondary procedure versus medication.
We included the following outcomes, where measured.

R E S U L T S Description of studies
An overview of study attributes can be found in the Characteristics of included studies table.

Results of the search
For this update, a search of the databases on 20 December 2020 returned a total of 510 records (Figure 1). A er removing 273 duplicates, 237 distinct records remained to be screened. We excluded a further 232 records based on their titles and abstracts, leaving five papers for full-text assessment. Of these five, one was a case report and was thus excluded, while the remaining four were confirmed as inclusions.
We identified two Korean language studies, which are awaiting full assessment (Characteristics of studies awaiting classification).

Design of studies
The studies by Bourne 2017, Lee 2011 and Narayanaswamy 2016 were parallel group RCTs, including two separate groups for each arm. Lee 2011 was a within-person RCT: one eye underwent peripheral iridotomy alone, while the other eye underwent both laser peripheral iridotomy and argon LPIp.

Controls and Interventions
All studies used argon LPIp as their mode of delivering LPIp. Sun 2010 randomised each arm to either peripheral iridotomy alone or peripheral iridotomy with argon LPIp (argon LPIp was performed two to five days following peripheral iridotomy). Bourne 2017 randomised post-peripheral iridotomy participants with a persistently occludable angle on gonioscopy to either argon LPIp or no further treatment. Lee 2011 also compared peripheral iridotomy alone to peripheral iridotomy with argon LPIp (the intervention eye underwent peripheral iridotomy and argon LPIp at the same sitting). Narayanaswamy 2016 compared argon LPIp to medical therapy (travoprost) in post-peripheral iridotomy participants with a persistently occludable angle on gonioscopy.

Outcomes
The outcomes assessed have been listed in Table 3.

Excluded studies
We excluded one study because it was an observational study comparing a group which had undergone argon LPIp in one time period, with a matched group from an earlier period (Cho 2017).

Risk of bias in included studies
We assessed the risk of bias for the included studies, with the findings summarised in Figure 3 and Figure 4.

Figure 3. Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Random sequence generation (selection bias) Allocation concealment (selection bias) Blinding of participants and personnel (performance bias): All outcomes Blinding of outcome assessment (detection bias): All outcomes Incomplete outcome data (attrition bias): All outcomes Selective reporting (reporting bias) 0% 25% 50% 75% 100% Low risk of bias Unclear risk of bias High risk of bias

Allocation
Random sequence generation was used by Narayanaswamy 2016 via the use of pre-allocated codes, and by Sun 2010 using a random number generator. We assessed both of these as 'low risk'. Both Lee 2011 and Bourne 2017 commented on having randomised participants but did not specify their randomisation process and therefore were marked as 'unclear risk'. Concealment of allocations before assignment was ensured by Narayanaswamy 2016 using sealed envelopes opened at the point of assignment. Sun 2010 had very well-matched groups, and so were judged to be at 'low' risk of bias. Lee 2011 and Bourne 2017 did not include a comparison of the two arms, or mention concealment measures and so were marked as 'unclear risk'.

Blinding
None of the participants nor researchers from all four trials were masked to which intervention the participants had received. Thus, all trials were marked as high risk.
As mentioned, all studies encountered di iculty in removing bias due to the inability to mask personnel to the iridoplasty scars. Narayanaswamy 2016 masked the researchers to some of the results they were obtaining by concealing the scale of the tonometer when IOP measurements were taken; these were subsequently read by an assistant. Sun 2010 also reported masking of IOP measurements, as well as assessment of visual acuity and automated perimetry, but did not specify the methods used. Despite attempts to reduce bias by some trials, we assessed all studies as 'high risk' because concealment was not possible for other measurements, such as gonioscopy.

Incomplete outcome data
All participants in Lee 2011 and Bourne 2017 were followed to study completion. Losses to follow-up were similar in both groups in Sun 2010 (20.8% peripheral iridotomy, 19.8% argon LPIp with peripheral iridotomy; 32 lost in total of 158 enrolled) and were included in a separate 'last observation carried forward' analysis. Narayanaswamy 2016 only had one loss to follow-up, from the argon LPIp group, which was also included in a 'last observation carried forward' analysis. However, this does not seem to have been done separately from the main analysis. As all studies had a low number of losses to follow-up, or accounted for any significant number of losses, we assessed all as 'low' risk of bias.

Selective reporting
Sun 2010 and Narayanaswamy 2016 were both prospectively registered (on the Chinese Clinical Trial Register and ClinicalTrials.gov respectively), while Bourne 2017 and Lee 2011 did not appear to have registered their trials. Sun 2010 and Lee 2011 reported fully on all outcomes specified in their methods, and were marked as 'low' risk of bias. Narayanaswamy 2016 included the baseline values for most AS-OCT parameters (mean angle width, angle opening distance (AOD), trabecular-iris space area (TISA), angle recess area (ARA), and iris thickness). However, they did not report the final raw values, or P values for the travoprost group, if measured. The change in anterior chamber depth (ACD) was also not reported, despite measurement at baseline. Bourne 2017 did not report AS-OCT values at three months for the control group, only comparing pre-and post-argon LPIp values. Therefore, we assessed both Narayanaswamy 2016 and Bourne 2017 as 'high' risk of bias.

Other potential sources of bias
We identified no other potential sources of bias in the included studies.

E ects of interventions
See: Summary of findings 1 Summary of Findings; Summary of findings 2 Summary of Findings; Summary of findings 3 Summary of Findings

Laser peripheral iridoplasty with iridotomy versus iridotomy alone as a primary procedure
The use of argon LPIp with iridotomy as a primary procedure was used by two studies as their intervention (Lee 2011; Sun 2010).

Primary outcome
Neither study collected data for the primary outcome of interest for this review.

Intraocular pressure (IOP)
There was no di erence in IOP lowering e iciency between LPIp and control groups. Lee 2011 assessed IOP at one hour, day, week, and one month as well as three months post-intervention, but no evidence of e ect was observed in the argon LPIp group at three months (mean di erence (MD) 0.30 lower, 95% confidence interval (CI) -2.11 to 1.51). A later assessment at 12 months was made by Sun 2010, who also found no evidence of e ect in the argon LPIp group (MD 1.70 mmHg higher, 95% CI -0.79 to 4.19). Results from both studies were pooled, giving a combined mean di erence of 0.39 mmHg (Analysis 1.1, 95% CI -1.07 to 1.85; I 2 = 38%; 2 studies, 174 participants; low-certainty evidence).

Additional medications
Sun 2010 reported on the mean number of additional glaucoma medications prescribed at 12 months. No evidence of e ect was observed in the use of argon LPIp over the control (MD 0.10, 95% CI -0.34 to 0.54; 126 participants, low-certainty evidence).

Complications
Complications were reported by both studies, but were varied in nature. Both Lee 2011 and Sun 2010 noted a similar risk of bleeding between groups, with Lee 2011 reporting that 4% of participants in both groups developed 'hyphema'. Sun 2010 reported on 'iris haemorrhage', which was present in 11.7% of participants in the argon LPIp with peripheral iridotomy group, and 12.3% of participants in the peripheral iridotomy group.  Other complications reported by Lee 2011 included 'persistent uveitis' (8% in the peripheral iridotomy alone group versus 4% in the argon LPIp with peripheral iridotomy group) and 'transient atonic pupil' (occurred in a single participant of the argon LPIp with peripheral iridotomy group, 4%). Sun 2010 reported two cases of 'corneal endothelial burn' in the peripheral iridotomy alone group (2.6%), and one case of malignant glaucoma in the argon LPIp with peripheral iridotomy group (1.2%).

Extent of peripheral anterior synechiae (PAS)
Sun 2010 examined the extent of peripheral anterior synechiae (PAS) at baseline and 12 months. While finding a significant improvement in both groups (P < 0.001), the authors found no di erence between groups (P = 0.473).

Further surgical or laser interventions
The need for further surgery was assessed by Sun 2010 at 12 months, who found no evidence of e ect in the argon LPIp group over controls (risk ratio (RR) 1.21, 95% CI 0.66 to 2.21; 126 participants, low-certainty evidence).

Quality of life measures
Data regarding quality of life measures were not collected by the authors of either study.

Summary of e ects
No evidence of e ect was found for the use of argon LPIp alongside peripheral iridotomy over the control group for the included outcome measures. We created an 'Adverse e ects' table (Table 2) to better summarise any di erences in complication rate between groups. Overall complications were uncommon, and severe complications were rare.

Laser peripheral iridoplasty as a secondary procedure versus no treatment
The use of argon LPIp as a secondary procedure versus no treatment was examined by one included study (Bourne 2017).

Primary outcome
Bourne 2017 did not collect data for the primary outcome of interest for this review.

Intraocular pressure (IOP)
Bourne 2017 measured IOP hourly between 9:00 and 16:00 in order to obtain diurnal IOP (DIOP) results at baseline and three months. At three months, they found the mean di erence in maximum IOP in the intervention group to be 1.81 mmHg lower than the control (95% CI -3.11 to -0.51; (19.34 mmHg argon LPIp, 21.15 mmHg peripheral iridotomy alone)) suggesting a clinically significant reduction in participants post-argon LPIp. However, no evidence of e ect was observed for mean minimum values (MD -0.31, 95% CI -1.93 to 1.31), and P values were not included comparing the two groups. GRADE evidence for the final maximum and minimum IOP measurements was very low-certainty.

Additional medication
The authors did not collect data regarding the use of additional medications.

Complications
The authors did not collect data regarding adverse events.

Further surgical or laser interventions
The authors did not collect data regarding any requirement for further surgical or laser intervention.

Quality of life measures
The authors did not collect data regarding quality of life measures.

Anterior chamber morphology
Bourne 2017 reported AOD, TISA and ARA at three months, by the analysis of AS-OCT images. Measurements were taken at 500 µm and 750 µm from the scleral spur, for eight sectors. Statistically significant widening was found in the argon LPIp group for all AOD, ARA and TISA parameters at three months, excluding Inferior AOD500 and 750, Supero-nasal and Superotemporal ARA500 and Inferonasal ARA750 (5 of 48). Trabecular iris angle (TIA) measurements were also measured by Bourne 2017, and widening was found to be statistically significant in all sectors. However, the GRADE certainty of this evidence was very low, and no AS-OCT measurements for the control group (peripheral iridotomy alone) were reported.

Summary of e ects
While there was no di erence in final mean minimum IOP values between groups, a clinically significant reduction in maximum IOP values was observed in the LPIp group. While this evidence was of very low-certainty, it showed a reduced range of diurnal IOP in those having undergone the intervention. The evidence regarding change in AS-OCT parameters was very unclear. Although statistically significant increases were reported in the intervention group, the evidence was of very low-certainty, and no AS-OCT measurements for the control group were reported.

Laser peripheral iridoplasty as a secondary procedure versus medication
The use of argon LPIp as a secondary procedure versus travoprost 0.004% was examined by one included study (Narayanaswamy 2016).

Library
Trusted evidence. Informed decisions. Better health.

Primary outcome
Narayanaswamy 2016 assessed the primary outcome of progression of the disease state via measurement of vertical cup-to-disk ratio (C/D ratio). Argon LPIp showed no evidence of greater e ect in preventing disease progression when compared to travoprost (MD -0.03, 95% CI -0.11 to 0.05; 80 participants; lowcertainty evidence).

Intraocular pressure (IOP)
Narayanaswamy 2016 reported that participants from the travoprost group achieved a statistically significantly lower mean IOP at 12 months, when compared to those in the argon LPIp group (17.7 mmHg versus 19.2 mmHg, P = 0.02), despite no di erence at baseline (P = 0.61). However, when assessing the mean change in IOP at 12 months, no evidence of a greater e ect in the travoprost group was noted (MD -1.20, 95% CI -2.87 to 0.47; 80 participants; low-certainty evidence).

Additional medications
The mean number of glaucoma medications prescribed at 12 months was assessed by Narayanaswamy 2016. No evidence of e ect was shown for argon LPIp over the travoprost group: a mean di erence of 0.42 additional medications was required by the intervention group at 12 months (95% CI 0.23 to 0.61; 80 participants; low-certainty evidence).

Complications
No serious adverse events were reported. Narayanaswamy 2016 reported two cases of 'post-laser IOP spike' in their argon LPIp arm (5%). An IOP spike was classified as an increase of 5 mmHg or higher.

Extent of peripheral anterior synechiae (PAS)
The extent of peripheral anterior synechiae (PAS) was measured at 12 months. Narayanaswamy 2016 reported a decrease in the travoprost group (P = 0.03), while noting an increase in the extent of PAS in the argon LPIp group (mean progression by 0.9 clock hours, P = 0.03).

Further surgical or laser interventions
The need for further surgery was assessed by Narayanaswamy 2016 at 12 months, who reported one case in the intervention group alone, with none in the control group (80 participants; low-certainty evidence).

Quality of life measures
The authors did not collect data regarding quality of life measures.

Anterior chamber morphology
Narayanaswamy 2016 assessed anterior chamber morphology at 12 months, by the analysis of AS-OCT images. Statistically significant widening was reported in AOD500 and 750, TISA750 and ARA750 values (P < 0.001 for all) for the argon LPIp group. However, measurement of e ect could not be calculated due to a paucity of available data, and results were not included for control participants, despite measurement at baseline. Mean angle width was also reported to be significantly increased in the argon LPIp group versus the travoprost group (1.6 versus 2.0 Sha er angle, P = 0.001), while no change in iris thickness (750 µm and 2000 µm) was observed (P = 0.43 and P = 0.34 respectively). All evidence was of low GRADE certainty.

Summary of e ects
No evidence of greater e ect of LPIp was noted for the included outcomes. The evidence regarding change in AS-OCT parameters was unclear, Although statistically significant increases were reported in the intervention group, the evidence was of low certainty, and many control group measurements were not reported at 12 months. No serious adverse events were reported in either group, but a greater number of medications was required by the argon LPIp group at 12 months.

Summary of main results
In this review, the results of four RCTs assessing the e icacy of argon LPIp as an intervention have been presented. Overall, the included studies have produced insu icient evidence to support the use of argon LPIp in people with chronic angle closure. Only one study reported results pertaining to the primary outcome (by assessing vertical C/D ratio) and found no statistical di erence between groups (Narayanaswamy 2016). No data were reported regarding progression of visual field loss, patient views or quality of life. The IOP lowering e icacy of argon LPIp was similar or worse than other options explored.
Overall, the results suggest that argon LPIp may be safe, however no long term follow up data was reported beyond 12 months. The incidence of severe complications reported by three RCTs was rare for all participants across control and intervention groups, regardless of the specific intervention utilised. Two studies found no evidence of e ect of argon LPIp in the reduction of IOP when compared to control groups, whether used as a primary or secondary procedure (Lee 2011; Sun 2010).
Another study comparing argon LPIp with medication reported a greater reduction in IOP in their control participants; however, the mean change was not significantly di erent between groups (Narayanaswamy 2016). One study assessed diurnal IOP, and found a reduction in the range of diurnal IOP values in their argon LPIp group compared to that at baseline, and to that of the peripheral iridotomy control group (Bourne 2017). However, the clinical relevance of a reduction in IOP fluctuation is uncertain.
Anterior chamber morphology was assessed using a wide range of parameters. Widening of the majority of AOD, TISA and ARA measurements post-argon LPIp was reported by Bourne 2017 and Narayanaswamy 2016. However, control data were absent and evidence was of a low or very low certainty. TIA and mean angle width were examined by single studies and found to be significantly widened. However, the measure of e ect could also not be assessed Cochrane Database of Systematic Reviews for these outcomes. Finally, mean ACA, ACV and deepest ACD were assessed by Lee 2011 who found no di erence in measure of e ect between argon LPIp and peripheral iridotomy alone.
We identified two Korean language studies, and extracted information from their abstracts only (their characteristics are outlined in the Characteristics of studies awaiting classification table). Park 2011 compared laser peripheral iridotomy alone to argon LPIp with peripheral iridotomy. It is not clear whether this study is a RCT or just a comparison of two groups of unequal number (11 and 14). Baseline IOP was disparate between groups (15.0 mmHg versus 18.9 mmHg). However, no di erence in IOP reductions was noted between the peripheral iridotomy alone and the argon LPIp with peripheral iridotomy groups (1.5 mmHg and 1.1 mmHg respectively). The main outcome measure reported in the abstract was widening of anterior chamber angle. Change in IOP is not mentioned in the English language abstract.
The second study, by Kim 2003, is a RCT comparing peripheral iridotomy alone to argon LPIp with peripheral iridotomy, with 30 participants per group. The English language abstract and a graph of results showed an advantage of combined treatment at 2 and 4 months (when comparing 'success' defined as IOP < 21 mmHg without the use of medication), while only showing a just significant advantage at 6 and 8 months (both having P = 0.047) and no di erence at 12 months.
An observational study by Cho and colleagues was excluded (Cho 2017). It compared people who had undergone argon LPIp with peripheral iridotomy between April and August 2015, with an earlier group who had undergone peripheral iridotomy alone between October 2014 and March 2015. The participants in the earlier group were matched for age, gender and iridotrabecular contact index, but not IOP. The changes in IOP were not significantly di erent between groups.

Overall completeness and applicability of evidence
Of the four studies included, only Narayanaswamy 2016 measured direct progression of the disease, by assessing C/D ratio, while no studies assessed progression of visual field scotoma.
A varied range of secondary outcomes were reported by the included studies, which made it di icult to collate the evidence. Of the outcomes shared by the studies, high variation in findings was evident, which may be due in part to the small sample sizes of some of the included studies. Findings may also di er as a result of the di ering methods of study design (both regarding the methods of interventions and comparators used), as well as the short follow-up period of some of the studies (three months in half of the included studies). Meta-analysis was not feasible.
Most participants were recruited from an Asian population, with participants from China, Singapore and Korea (Lee 2011; Narayanaswamy 2016; Sun 2010), with only 7% of participant eyes from a European population (Bourne 2017). Di erences in anterior segment physiology between ethnic groups could hinder the applicability of the results found to non-Asian populations.

Quality of the evidence
The five GRADE domains (methodological limitations of the studies, inconsistency of e ect, imprecision, indirectness, and publication bias) were used to assess the quality of the evidence obtained from the included studies. Concerns were rated as 'not serious', 'serious' or 'very serious'.

Methodological limitations of the studies
All trials had a high risk of bias concerning the lack of ability to mask (blind) participants and personnel to the intervention received, and the ability to mask the assessor to the outcomes measured. Two trials showed small instances of selective reporting (Bourne 2017; Narayanaswamy 2016). There was a lack of specification of the randomisation and allocation process by two studies (Bourne 2017; Lee 2011). All studies included results of cases lost to follow-up where applicable. Therefore, we rated our concerns regarding this GRADE domain as 'serious'.

Indirectness
Three of the four included trials did not assess the primary outcome, disease progression. However, the participants, intervention and comparators used by all studies all provide direct evidence to the clinical question at hand. The parameters used to measure morphology of the anterior chamber varied between studies; despite this, no serious indirectness was found in the outcome measures assessed. Therefore, we rated our concerns regarding this GRADE domain as 'not serious'.

Imprecision
The total number of participants included in all the trials was 252 participants (276 eyes). Some outcomes were only measured by a single trial: C/D ratio by Narayanaswamy 2016; and BCVA and CECC by Sun 2010. However, these studies had larger sample sizes (80 and 126, respectively) and followed participants to 12 months. Therefore, we rated our concerns regarding this GRADE domain as 'not serious'.

Inconsistency
A significantly greater number of medications was required by the control group in one study (Narayanaswamy 2016), which also noted a decrease in the extent of PAS in the control group, yet an increase in the intervention group. This contrasts to another study, which found no significant di erence between intervention and control groups concerning both outcomes (Sun 2010). However, di erent comparators were used (travoprost and peripheral iridotomy alone, respectively).
IOP was measured by all four studies. Two found no significant di erence between groups at 3 months (Lee 2011), and 12 months (Sun 2010), while a third found a reduced range of DIOP in the intervention group (P values not given) (Bourne 2017). The last study, Narayanaswamy 2016, found a significantly lower mean IOP at 12 months in their comparator group. However, as mentioned, this trial used a di erent comparator. Therefore, we rated our concerns regarding this GRADE domain as 'serious'.

Likelihood of publication bias
No unpublished data were available for analysis. Of the published studies, no financial conflicts of interest were declared. Published studies did not exclusively show positive findings. Therefore, we rated our concerns regarding this GRADE domain as 'not serious'.

Library
Trusted evidence. Informed decisions. Better health.

Potential biases in the review process
In this review, we followed the steps for conducting a systematic review, as outlined by Cochrane, to prevent bias wherever possible. We expanded the search strategies used in previous iterations of this review in order to ensure all relevant studies could be identified, and we searched a wide variety of locations. As outlined in the methods section, several review authors independently searched for, and assessed the risk of bias of, included studies, and extracted data were checked by another author.
We identified two potentially relevant studies. However, as they were published in Korean and translation was not available, these could not be included in the synthesis of the review.

Agreements and disagreements with other studies or reviews
Not applicable, as there are no other current reviews.

Implications for practice
LPIp is currently used as a second-line treatment option in people with remaining appositional angle closure a er peripheral iridotomy. The results obtained from the four included RCTs do not provide enough evidence to suggest that argon LPIp confers any additional benefit to the use of peripheral iridotomy alone in reducing IOP and subsequently preventing the progression of the disease process, whether when used as a primary or secondary intervention. A low incidence of severe complications was reported by three RCTs across all groups. Findings suggest that anterior chamber morphology may be positively impacted by argon LPIp. However, there is no evidence of significant reduction in IOP a er argon LPIp and thus there is no evidence to support the use of LPIp in the management of chronic angle closure glaucoma.

Implications for research
The results of the included trials suggest argon LPIp provides little to no benefit over comparators in the management of people with chronic angle closure. Despite uncertainties, the existing evidence does not justify research e orts in further trials of argon LPIp as an intervention in cases of chronic angle closure.

A C K N O W L E D G E M E N T S
With thanks to Bourne 2017 for their response to a query regarding their paper.
We thank Anthony King and Kerry Dwan for their comments on this update and Anupa Shah and Jennifer Evans for their assistance/ guidance throughout the review process.