We sought to comprehensively evaluate the effectiveness of different types of laser trabeculoplasty (LT) in the treatment of open-angle glaucoma.
Systematic review and network meta-analysis.
Eligible randomized controlled trials were identified by searching PubMed, EMBASE, Cochrane Library, SCOPUS, China National Knowledge Infrastructure, and the Chinese Biomedical Literature Service System for studies published between January 1, 2000 and April 20, 2020. Eight interventions were evaluated, including argon LT (ALT), medications, 180-degree selective LT (SLT), 270-degree SLT, 360-degree SLT, new LT, transscleral 360-degree SLT with SLT performed without gonioscopy, and low-energy 360-degree SLT. The primary outcome was reduction of medicated and unmedicated intraocular pressure (IOP) at 6 months. Secondary outcomes included reduction of IOP at 12 months, incidences of complications, and change in number of medications. Head-to-head meta-analysis and network meta-analysis were performed using Stata and R software.
In total, 22 studies were included, involving 2859 eyes of 2704 patients. In terms of IOP reduction at 6 and 12 months, there were no statistically significant differences in both medicated and unmedicated IOP between any pairs of interventions considered herein, as determined based on both head-to-head and network meta-analyses (all P > .05). In terms of reduction of medications, the individuals treated with 180-degree SLT required fewer medications than those treated with ALT at 12 months (0.28 [95% confidence interval, 0.06-0.50]; P = .014). No severe adverse outcomes were reported for any of the interventions.
All the available types of LT are equally effective for decreasing IOP compared with medication-based therapy. The 180-degree SLT was slightly more effective than ALT in terms of reducing the number of medications needed. Additional well-performed randomized controlled trials with larger sample sizes are needed.
Glaucoma is the leading cause of irreversible blindness worldwide. It was estimated that in 2013, 64.3 million people in the world were affected by glaucoma; this number is expected to grow to 111.8 million by 2040. , Open-angle glaucoma (OAG) is a major type of glaucoma, and reduction of intraocular pressure (IOP) is the only clinically validated way to retard the progression of OAG. , Currently, medication, laser trabeculoplasty (LT), and surgery are the main treatment modalities for achieving this goal.
Diverse types of LT have been developed and used extensively over the past 40 years. Argon laser trabeculoplasty (ALT), which was introduced in 1979, was the first among such techniques; it can reduce IOP by 6.4 mm Hg-9.7 mm Hg. Selective laser trabeculoplasty (SLT), developed in 1995, was found to be as effective as ALT in decreasing IOP , and more repeatable than ALT. In recent years, several LT procedures have emerged, such as micropulse LT (MLT), titanium sapphire LT (TLT), pattern scanning LT (PSLT), and transscleral SLT without gonioscopy lens. These techniques are characterized by reduced trauma and reduced postoperative inflammation. , , By incorporating randomized controlled trial (RCT) studies, we and other researchers have conducted several traditional head-to-head meta-analyses to compare various laser procedures. , , However, these analyses have limitations: 1) they have been applied only to direct comparisons between traditional lasers; 2) they have not included the latest forms of laser procedures; and, more importantly, 3) they have not ranked the procedures in terms of efficacy, safety, and benefits, which is extremely valuable for clinical decision making.
Network meta-analysis (NMA), a relatively new technique and an extension of conventional meta-analysis, can be used to simultaneously compare multiple treatments in a single analysis by synthesizing both direct and indirect pieces of evidence. Unlike meta-analysis, which is only applicable when there is a direct treatment comparison and can compare just 2 interventions at a time, NMA can perform multiple indirect comparisons and provide valuable information for health care decision-making when no direct comparisons are available. When direct comparisons are available, NMA can synthesize the data of direct and indirect comparisons and increase the accuracy of the results. In addition, NMA allows for us to rank the relative efficacies of all available treatments, which is almost impossible when using meta-analysis. , Therefore, in this study, we aimed to determine the comparative efficacy of LT treatments for patients with OAG by following the NMA approach.
This systematic review and network meta-analysis has been reported according to the Preferred Reporting Items for Systematic reviews and Meta-analyses extension statement for NMA.
Pertinent studies were identified by systematically searching various electronic databases, including PubMed, EMBASE, SCOPUS, the Cochrane library, China National Knowledge Infrastructure, and the Chinese Biomedical Literature Service System for studies published between January 1, 2000 and April 20, 2020. In addition, the Chinese Clinical Trial Registry, World Health Organization International Clinical Trials Registry Platform, and ClinicalTrials.gov were searched for relevant trials. There was no restriction on study language. The reference lists of included studies and related reviews were screened for additional eligible studies. The following search terms were used and were modified accordingly: (“open-angle glaucoma” OR “primary glaucoma” OR “POAG” OR “chronic glaucoma” OR “low-tension glaucoma” OR “low-pressure glaucoma” OR “normal-tension glaucoma” OR “normal-pressure glaucoma” OR “pigmentary glaucoma” OR “exfoliation syndrome” OR “exfoliative glaucoma” OR “pseudoexfoliation syndrome” OR “pseudoexfoliation glaucoma”) AND (“laser trabeculoplasty” OR “SLT” OR “ALT” OR “MLT” OR “TSLT” OR “PLT”) AND (“randomized controlled trial” OR “controlled clinical trial” OR “randomized” OR “placebo” OR “randomly” OR “trial”). A complete description of the search strategies is given in the Supplemental Table 1 (Supplemental Material at AJO.com ).
The following inclusion criteria were used in this study: 1) design: RCTs with parallel group design; 2) patients: enrolled patients with OAG, including primary OAG (POAG), normal-tension glaucoma, ocular hypertension, pseudoexfoliative glaucoma, and pigmentary glaucoma; 3) interventions: including ≥1 LT intervention (SLT, ALT, MLT, TLT, PSLT, and transscleral SLT); 4) control subjects: antiglaucoma medications or another LT among those listed in 3); 5) follow-up: patients who followed up for ≥6 months after randomization; and 6) outcomes: reduction of IOP from baseline was reported or raw data were available for calculation. The exclusion criteria were as follows: 1) presence of other ocular or systematic diseases; 2) multiple interventions were combined simultaneously; 3) rate of lost to follow-up >20%; 4) reviews, letters, protocols, or conference abstracts; and 5) inadequate data for meta-analysis.
Two authors (R.Z., Y.S.) independently extracted the data using a standardized data extraction form and resolved discrepancies after mutual agreement and discussion with a third reviewer. The following data were extracted: author(s) of the study, year of publication, intervention, sample size, duration of follow-up period, diagnosis of included patients, previous history of LT, IOP and number of medications at baseline and at each time point of follow-up, laser parameters, complications, and items related to methodologic quality. In case there were multiple publications related to a study, the one with the longest follow-up duration was included. Data from the other reports were extracted when needed.
Primary efficacy outcomes were reduction of IOP at 6 months, defined as baseline IOP minus IOP at 6 months. If IOP values corresponding to the predetermined time points were not available, the IOP value at the time point closest to the predetermined time points was collected. Because of the confounding effect of use of medication on IOP measurement, the studies were separated into 2 subgroups according to whether medicated or unmedicated IOP was used. The outcome, IOP reduction, was analyzed for each subgroup. Secondary outcomes included reduction of IOP at 12 months, incidence rate of adverse events, and a reduction in the number of medications from baseline at 6 and 12 months.
Risk of Bias Assessment
The risk of bias tool outlined in the Cochrane Handbook for Systematic Reviews of Interventions (version 5.1.0) was used to assess the methodologic quality of the included studies. Each study was rated in terms of 6 different aspects: sequence generation; allocation concealment; blinding of patients, personnel, and outcome assessors; management of eventual incomplete outcome data; completeness of outcome reporting; and other potential threats to validity.
Data analyses were performed using Stata software (version 14; StataCorp LP, College Station, Texas, USA) and R software (version 3.6.1; Vienna, Austria, available at: www.R-project.org/ ). The intention-to-treat principle was followed. All participants were included in the analysis based on the initial randomization, regardless of which group they were assigned to or whether they were excluded from the analysis.
Head-to-head pairwise meta-analysis was performed first for each treatment comparison with ≥2 trials by using the fixed-effects model in cases where there was no significant heterogeneity. In cases with significant heterogeneity, the random-effect model was used. Studies brought together in a systematic review will inevitably differ. Some variance is caused by chance, while others may result from clinical and methodologic differences across studies, in which heterogeneity refers to all variance between the studies that is not related to chance. Heterogeneity among studies was assessed using the I 2 statistic and the Q test. I 2 denotes the percentage of the variability in effect estimates that is caused by heterogeneity. An I 2 of 25%-50% indicates a low degree of heterogeneity, 50%-75% a moderate degree of heterogeneity, and >75% a high degree of heterogeneity. The Q test uses the χ 2 test to determine the possibility that the variability is simply caused by chance, with a P < .10 indicating the presence of heterogeneity. The weighted mean difference and its 95% confidence interval (95% CI) were calculated. A positive weighted mean difference of IOP/reduction in the number of medications was considered an indicator that the first intervention was more efficacious in lowering IOP/reducing number of medications.
To incorporate indirect comparisons, random-effect network meta-analysis was performed using the frequentist framework. Consistency between direct and indirect pieces of evidence was tested with the node-splitting model, which classifies individual comparisons as direct and indirect pieces of evidence; any disagreement between them was considered an indicator of inconsistency. The hierarchy of the interventions was calculated using the surface under the cumulative ranking curve. The higher the surface under the cumulative ranking curve value, the higher the rank. By choosing alternative eligibility criteria for studies and repeating the primary analysis, sensitivity analysis allows for us to test the robustness of the original findings and exclude the possibility that these findings may depend on the study inclusion process. Additional sensitivity analyses were performed by following 5 approaches. The first 2 involved determining the impact of previous antiglaucoma treatment, including the history of LT and glaucoma surgery, on the treatment outcomes. The third involved excluding studies that could introduce significant heterogeneity in pairwise meta-analysis. The fourth and fifth involved excluding studies at high or unclear risk of bias in sequence generation or allocation concealment, respectively.
Figure 1 shows the literature selection procedure. In total, 1099 citations were identified after a systematic search by title and abstract screening. Of these, 1045 were excluded because they were duplicates in multiple databases or because of other reasons (unrelated to the topic, reviews, case reports, or case series), leaving 54 for full-text review. Thereafter, 31 of 54 studies were excluded because of the following reasons: 10 were not RCTs, 2 had a lost to follow-up rate >20%, , 13 were conference abstracts (Belkin M, et al. IOVS 2014; 55:ARVO E-abstract 819; Bovell AM, et al. IOVS 2003; 44:ARVO E-abstract 101; Bovell AM, et al. IOVS 2004; 45:ARVO E-abstract 5586; and Simon G, et al. IOVS 2006; 47:ARVO E-abstract 5458) or protocols, 1 applied multiple interventions simultaneously, 2 had inadequate follow-up durations, , 1 had no control group, and 2 lacked data on primary outcomes. , Finally, 22 studies of the aforementioned 23 citations were considered eligible for qualitative analysis, of which 21 studies reported IOP at the predefined time points and were included in the quantitative analysis , and 18 studies reported a reduction of medicated IOP at 6 months and were included in network meta-analysis.
Characteristics of Included Studies
The participants’ baseline characteristics are summarized in Table 1 . A total of 2859 eyes of 2704 patients were assigned to 8 interventions, including ALT, medications, 180-degree SLT, 270-degree SLT, 360-degree SLT, new LT (MLT, TLT, and PSLT), transscleral 360-degree SLT, and low-energy 360-degree SLT. The mean age of the participants ranged from 42.9-73.4 years. In total, except for the 2 studies without adequate data, females accounted for 25%-67.7% of the participants from the remaining studies. The mean follow-up durations ranged from 6-60 months. Sixteen studies recruited patients without a history of LT, 3 studies included patients with previous LT, and 3 studies were unclear about this matter. In 18 studies, the patients did not undergo previous glaucoma surgery. In 1 study, the patients with previous glaucoma surgery were included. In the other 3 studies, this is unclear. The baseline IOP ranged from 12.6 mm Hg-26.8 mm Hg. The baseline number of medications ranged from 0-2.9. Figure 2 shows the network of direct comparisons for reduction of medicated IOP at 6 months.
|Author (Year)||Treatment||Patients (Eyes)||Follow-Up (m)||Age (Year)||Female (%)||POAG (%) a||IOP (mm Hg)||No. of Medications||Medicated IOP (Yes/No)||Naive Laser||Glaucoma Surgery|
|Birt (2007)||180-degree SLT||30 (30)||12||64.0||50.0||83.3||22.9||2.9||Yes||Yes||No|
|180-degree SLT||27 (27)||72.4||48.1||81.5||21.5||2.8||No|
|180-degree ALT||39 (39)||70.0||46.2||87.2||22.0||2.8||Yes|
|Bovell (2011)||180-degree SLT||152 (176)||60||69.7||59.6||60.7||23.8||2.6||Yes||No||No|
|Liu (2012)||180-degree SLT||20 (20)||3||48.7||25.0||45.0||19.1||2.6||Yes||Yes||No|
|180-degree ALT||22 (22)||3||51.6||36.4||45.5||21.9||2.9||Yes|
|Rosenfeld (2012)||180-degree SLT||22 (22)||12||72.0||50.0||40.9||25.4||NA||Yes||Yes||No|
|180-degree ALT||30 (30)||71.9||53.3||46.7||25.1||NA||No|
|Casa (2004)||180-degree SLT||20 (20)||6||73.4||55.0||100||24.0||1.8||Yes||Yes||No|
|180-degree ALT||20 (20)||72.5||50.0||100||23.6||1.5||Yes|
|Kent (2015)||180-degree SLT||60 (76)||6||72.9||64.4||0||23.1||NA||Yes||Yes||No|
|Hutnik (2019)||180-degree SLT||69 (69)||12||65.0||46.9||71.9||21.6||1.3||Yes||No||No|
|180-degree ALT||70 (70)||67.1||39.4||71.2||21.5||1.1||No|
|Sreckovic (2011)||180-degree ALT||35 (50)||30||65.0||37.1||60.0||25.2||NA||Yes||NA||No|
|Goldenfeld (2009)||180-degree TLT||18 (18)||15||68.0||61.1||77.8||25.7||1.4||Yes||NA||NA|
|180-degree ALT||19 (19)||15||57.9||73.7||25.7||2.1||NA|
|Tufan (2017)||180-degree SLT||18 (18)||6||54.2||50.0||100||17.3||2.2||Yes||Yes||No|
|360-degree SLT||22 (22)||53.6||54.5||100||17.0||2.4||Yes|
|Mansouri (2017)||360-degree PSLT||29 (29)||6||54.1||27.6||86.2||17.3||0||No||Yes||No|
|Abramowitz (2018)||360-degree MLT||38 (38)||12||66.1||42.1||100||18.3||NA||Yes||No||Yes|
|360-degree SLT||31 (31)||67.6||61.3||100||16.9||No|
|Lai (2004)||360-degree SLT||29 (29)||60||51.9||55.2||58.6||26.8||NA||Yes||Yes||No|
|Nagar (2009)||360-degree SLT||20 (20)||6||66.4||47.5||42.5||26.1||0||Yes||Yes||No|
|Katz (2012)||360-degree SLT||38 (67)||12||25-82||57.9||NA||25.0||0||No||Yes||No|
|Lee (2014)||360-degree SLT||22 (22)||6||66.5||45.5||100||15.8||2.3||Yes||Yes||No|
|Kiddee (2017)||360-degree SLT||30 (30)||5||68.0||33.3||53.3||15.8||1.4||No||Yes||No|
|Keyser (2018)||360-degree SLT||67 (133)||12||68.1||50.4||85.7||14.0||1.5||Yes||Yes||No|
|Gazzard (2019)||360-degree SLT||356 (613)||36||63.4||43.8||76.7||24.5||NA||Yes||Yes||No|
|Zhang (2016)||360-degree SLT||26 (26)||12||42.9||NA||100||25.0||0||No||NA||NA|
|Low-energy 360-degree SLT||26 (26)||46.8||NA||100||25.7||0||NA|
|Zhang (2015)||270-degree SLT||24 (34)||9||44.0||NA||NA||15.0||2.0||NA||Yes||No|
|Low-energy 360-degree SLT||21 (33)||46.5||NA||15.5||2.0||Yes|
|Geffen (2017)||Transscleral 360-degree SLT||14 (14)||6||67.0||50.0||57.1||20.2||2.1||Yes||Yes||NA|
|360-degree SLT||16 (16)||6||66.6||35.7||57.1||21.1||1.5||Yes|