Medical interventions for treating primary angle‐closure glaucoma
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To assess the comparative effectiveness of medical interventions for primary angle‐closure glaucoma in people with poorly controlled IOP following LPI (with or without iridoplasty).
Background
Description of the condition
Angle‐closure is a condition characterized by appositional or synechial closure of the eye's anterior chamber angle, potentially leading to outflow obstruction with consequent increase in intraocular pressure (IOP) (Emanuel 2014). When obstruction is due to an anatomical predisposition of the eye, angle‐closure is defined as primary angle‐closure (PAC). When obstruction is due to a pathophysiological process related to a concomitant ocular disease (eg, neovascular glaucoma, inflammation, iridocorneal endothelial syndrome, uveal effusion), the preferred term is secondary angle‐closure.
Different mechanisms can be involved in the pathogenesis of PAC. Most cases are related to pupillary block, in which aqueous flow from the posterior chamber to the anterior chamber of the eye is hampered by a tight clinch between the lens and the posterior surface of the iris. Resulting higher pressure in the posterior chamber makes the iris bowed forward and brings apposition to the trabecular meshwork (Mapstone 1968; Salmon 1999; Silver 2004). Other mechanisms can play a role in PAC by crowding the anterior chamber angle. These mechanisms can co‐exist with pupillary block or can act alone and are related to volume of the iris and its insertion into the ciliary body, and to position and thickness of the ciliary body (Wand 1977). Anterior positioning of the lens or increased lens thickness, which can contribute to shallowing of the anterior chamber, represents a further important anatomical risk factor for development of PAC (Lee 2012; Lowe 1969),
When the anterior chamber angle is closed, physiological drainage of the aqueous humor is blocked, and prolonged iridotrabecular contact (ITC) may result in scarring and functional damage of the trabecular meshwork, or may induce the formation of peripheral anterior synechiae (PAS) (Bonomi 2000). Thus, PAC can be associated with elevated IOP and glaucomatous optic neuropathy. Clinically, ITC extended by at least 180° with no sign of permanent aqueous flow obstruction (PAS, or raised IOP) is defined as primary angle‐closure suspect (PACS). Patients with PAC present with ITC of at least 180° associated with PAS or increased IOP (AAO 2010; EGS 2014). Primary angle‐closure glaucoma (PACG) is characterized by a PAC associated with glaucomatous optic neuropathy (Foster 2002).
Most often, PAC and PACG develop insidiously and chronically without symptoms. However, angle‐closure develops suddenly in a minority of cases, with consequent rapid elevation of IOP when a large portion of the trabecular meshwork is obstructed. Such a condition, referred to as acute angle‐closure crisis (AACC), is characterized by acute symptoms and may or may not be self resolving (Saw 2003b).
Epidemiology
Glaucoma is the leading cause of irreversible blindness globally. Among different types of glaucoma, PACG accounts for approximately 31% of cases, and it has been estimated that about 21 million people worldwide will have angle‐closure glaucoma by 2020 (Quigley 2006). According to a World Health Organization report, almost two‐thirds of all PACG cases have been reported in China and other Asia‐Pacific countries, reflecting the known higher prevalence of angle‐closure in Asian populations. On the basis of gonioscopic examination, a systematic review estimated age‐specific prevalence in the adult Asian population at 0.75%, with an increasing trend in the age group from 40 to 49 to 70 years and older (Cheng 2014).
Ethnicity‐specific prevalence within Asian countries, including the Middle East, South East Asia, India, China, and Japan, is reported to range from 0.46% to 1.10% (Cheng 2014). Although studies of Caucasian populations have revealed that angle‐closure glaucoma may not be as common as open‐angle glaucoma, angle‐closure glaucoma represents a significant percentage of cases of glaucoma (Bonomi 2000). Among native populations of Canada and Greenland, studies have shown higher prevalence of angle‐closure glaucoma compared with Caucasian populations in Europe (Bankes1968; Clemmesen 1971; Drance 1973; Hollows 1966).
In addition, the prevalence of angle‐closure glaucoma increases with age. As of 2013, among people from 40 to 80 years of age, an estimated 20 million people worldwide had angle‐closure glaucoma; 76% lived in Asia and 7% in Europe (Tham 2014). Two studies investigated European populations and estimated the pooled age‐adjusted prevalence of angle‐closure glaucoma at around 0.4% (Day 2012). Other studies have reported age‐specific prevalences in European populations of 0.02% among those 40 to 49 years of age, and 0.94% for individuals 70 years of age and older (Bonomi 2000; Cedrone 1997).
Presentation
In contrast with patients experiencing an acute angle‐closure crisis, who present with sudden onset of characteristic symptoms such as blurred vision, eye pain, headache, nausea, vomiting, mid‐dilated pupil, and corneal edema, individuals with PACG usually do not experience symptoms at all. In a minority of cases, patients report experiencing transient, slight, and fast‐resolving disturbances, such as blurred vision, eye redness, eye pain, headache, and halos around light, which are suggestive of intermittent and self contained angle‐closure attacks. Careful investigation of these symptoms in patients suspected of having a primary angle‐closure condition is heightened when familial history of disease is reported. As a result of the frequent absence of symptoms, PACG may proceed silently toward vision loss and blindness ‐ very similar to the process of open‐angle glaucoma. Therefore, angle‐closure glaucoma may be diagnosed only on the basis of optic nerve damage and visual field changes, along with evidence of a narrow angle observed on gonioscopy examination. As the presentation of PACG varies among cases, it is recommended that epidemiology, pathophysiology, and clinical manifestations should be taken into consideration when PACG is identified (http://www.aoa.org/).
Description of the intervention
Preliminary treatment of patients with PACG aims to relieve pupillary block components (Saw 2003a). Laser peripheral iridotomy (LPI) is the current standard of care as primary treatment for individuals with PACG (AAO 2010). In LPI, a small opening is made in the iris to facilitate the outflow of aqueous humor from the posterior chamber to the anterior chamber of the eye, and to reduce IOP. In patients with persistent closure of the angle, an adjunctive laser peripheral iridoplasty can pull the iris tissue away from the trabecular meshwork by shrinking peripheral iris tissue (Ng 2012).
However, after laser treatment, patients may continue to have poorly controlled IOP despite an open LPI (Alsagoff 2000). In fact, a patent LPI is capable of relieving pupillary block components, but other mechanisms, including trabecular meshwork damage, PAS formation, or plateau iris configuration, may contribute to maintenance of elevated IOP (Hung 1979). As a result, medical topical treatment often is necessary to reduce residual uncontrolled IOP among patients with PACG treated with LPI. When first‐line treatment (LPI plus medical therapy) fails to control IOP, or when signs suggest progression of glaucomatous damage, a filtering incisional surgery is indicated (Quigley 2011; Thomas 2013; Weinreb 2014). In PACG, trabeculectomy is associated with higher rates of failure and postoperative complications, such as flattening anterior chamber, malignant glaucoma, cataract formation, and other glaucoma subtypes (Aung 2000).
Anatomical characteristics of the lens are involved in both pupillary block and angle crowding mechanisms; therefore, lens extraction, performed alone or in combination with other procedures, has been suggested as potential treatment for patients with PACG (Thomas 2011). A Cochrane review has addressed the effectiveness of lens extraction for PACG versus other interventions (Friedman 2006). No randomized controlled trials (RCTs) but two comparative studies were found eligible, and study authors concluded that no adequate evidence showed effects of lens extraction as therapy for PACG. Another RCT was conducted to evaluate the clinical effectiveness of early lens extraction surgery versus standard care consisting of LPI, followed by topical anti‐glaucoma medication (Azuara‐Blanco 2011). Results of this trial are not yet available.
How the intervention might work
Medical treatments for PACG aim to decrease IOP by increasing aqueous outflow or by decreasing production of aqueous humor. Drugs that decrease aqueous humor production include beta‐blockers, alpha‐2 agonists, and carbonic anhydrase inhibitors (Doyle 1999; Gandolfi 2001). Therapies that increase aqueous outflow include miotics, adrenergics, and prostaglandin analogs (Brubaker 2003).
Although prostaglandins and miotics increase aqueous outflow, they act through different mechanisms. Miotics such as pilocarpine (a cholinergic agonist) stimulate contractions of the ciliary muscle, which pull on the scleral spur. By opening fluid channels, miotics reduce resistance in the trabecular meshwork, thereby facilitating trabecular outflow (Diestelhorst 2002). Pilocarpine also acts on iris sphincter muscle, causing pupil contraction (miosis) and reducing crowding of the anterior chamber angle (Kobayashi 1999).
Prostaglandin analogs increase uveoscleral outflow mainly by reducing extracellular matrix components within the ciliary muscle, iris root, and sclera, with little or no effect on trabecular outflow facility (Weinreb 2002). They have been demonstrated to be the most effective ocular topical hypotensive drugs for lowering IOP in primary open‐angle glaucoma and ocular hypertension (van der Valk 2005), but beta‐blockers and pilocarpine have been longer preferred for medical treatment of PACG (Hoh 2002). When the anterior chamber angle is closed fully or partially, the effectiveness of prostaglandin analogs is unknown.
Why it is important to do this review
Pupillary block is the main mechanism involved in the development of angle‐closure glaucoma. Although LPI represents the preferred and well‐recognized initial treatment to relieve the pupillary block component, it may fail in controlling IOP in the long term (Rosman 2002). Medical therapy is crucial in reducing IOP when LPI is insufficient, and it usually supplements LPI to control IOP postoperatively. The comparative effectiveness of different hypotensive ocular agents in patients with PACG is unknown and was identified by members of the Asia‐Pacific Joint Glaucoma Congress as a priority topic for a systematic review (Yu 2013).
Objectives
To assess the comparative effectiveness of medical interventions for primary angle‐closure glaucoma in people with poorly controlled IOP following LPI (with or without iridoplasty).
Methods
Criteria for considering studies for this review
Types of studies
We will include randomized controlled trials (RCTs).
Types of participants
We will include studies in which participants were diagnosed with PACG (as defined by the study investigator) and presented with uncontrolled IOP after a patent LPI. We will not include participants with a history of an acute attack of angle‐closure crisis. We will apply no restriction with regard to participant age, gender, race, baseline IOP, or baseline anatomical configuration of the angle. We will not include participants with uncontrolled IOP after lens extraction because treatment guidelines still consider LPI as the first treatment option.
Types of interventions
We will include studies in which one type of medical intervention (beta‐blockers, prostaglandin analogs, or miotics) is compared with another type of medical intervention. Specifically, we will look at three types of comparisons.
-
Prostaglandins versus pilocarpine.
-
Beta‐blockers versus pilocarpine.
-
Prostaglandins versus beta‐blockers.
Types of outcome measures
Primary outcomes
-
Progression of visual field loss, defined per criteria described in the included studies. The primary time point will be three years from baseline.
Secondary outcomes
-
Decrease in IOP: assessed as mean decrease from baseline to one year of follow‐up. We also will assess longer follow‐up time points as reported in the included studies. If we are unable to procure data on the mean decrease, we will analyze the mean IOP to estimate the treatment effect. If participants received medical interventions before study enrollment, baseline is seen as time after the wash‐out period. If no washout period is mentioned, we would still include data with caution and would discuss the effects, although this rarely happens in RCTs.
-
Progression of optic disc damage or nerve fiber layer loss, defined per criteria described in the included studies.
-
Adverse effects: increased IOP, corneal edema, posterior synechiae, and discontinuation of medications due to toxicity, allergy, visually significant cataract, or systemic side effects.
-
Quality of life: measured by the National Eye Institute Visual Function Questionnaire (NEI VFQ) or other available measurements used in individual studies.
-
Cost‐effectiveness data.
Definitions and measurements of progression of visual field loss and optic nerve damage may vary among included studies. These variations in definitions will be recorded and will be reported in the ‘Characteristics of included studies’ table as whether an outcome was measured by a validated technique.
Search methods for identification of studies
Electronic searches
We will search CENTRAL (which contains the Cochrane Eyes and Vision Trials Register) (latest issue), Ovid MEDLINE, Ovid MEDLINE In‐Process and Other Non‐Indexed Citations, Ovid MEDLINE Daily, Ovid OLDMEDLINE (January 1946 to present), EMBASE (January 1980 to present), PubMed (1948 to present), Latin American and Caribbean Health Sciences Literature Database (LILACS) (1982 to present), ClinicalTrials.gov (www.clinicaltrials.gov) and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) (www.who.int/ictrp/search/en). We will not use any date or language restrictions in the electronic search for studies. See: Appendices for details of search strategies for CENTRAL (Appendix 1), MEDLINE (Appendix 2), EMBASE (Appendix 3), PubMed (Appendix 4), LILACS (Appendix 5), ClinicalTrials.gov (Appendix 6), and ICTRP (Appendix 7).
Searching other resources
We will search reference lists of included studies to identify additional eligible studies. We will use the Web of Science to identify additional studies that may have cited studies included in this systematic review. We will not manually search conference proceedings or journals specifically for this review.
Data collection and analysis
Selection of studies
Two review authors (XH and MM) will independently review titles and abstracts of records identified via electronic searches using Criteria for considering studies for this review. Each review author will classify each record as "definitely relevant," "possibly relevant," or "definitely not relevant." We will resolve disagreements through discussion. After reaching consensus, we will retrieve full‐text reports for all records classified by both review authors as "definitely relevant" or "possibly relevant." We will group reports by study, and each review author will independently assess each study for inclusion and will classify each as "include," "unsure," or "exclude." We will contact investigators of studies for which eligibility is uncertain to request further clarification after we have assessed the full‐text reports. We will resolve discrepancies through discussion. We will document the reasons for exclusion of studies after reviewing the full‐text reports. We will further classify included studies as "ongoing" when studies were eligible but were not yet completed, or when study results were not available. For reports written in languages not read by the review authors, we will use Google Translate or will consult with a translator to assess studies for eligibility.
Data extraction and management
When independently extracting data related to study methods, participant characteristics, and outcomes, two review authors used data extraction forms developed by Cochrane Eyes and Vision. One review author will enter data into Review Manager 5 (RevMan 2014), and a second review author will verify the data entered. We will resolve disagreements after data extraction or data entry through discussion. We will contact study investigators to provide missing information or to clarify data, and we will allow two weeks for a response. If we do not receive a response during this time, we will proceed by using available information as provided in published reports.
Assessment of risk of bias in included studies
Two review authors will independently assess included studies for sources of potential bias, according to guidelines provided in Chapter 8 of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We will evaluate the following risk of bias domains: random sequence generation and allocation concealment (selection bias), masking of participants and personnel (performance bias), masking of outcome assessors (detection bias), missing data and intention‐to‐treat analysis (attrition bias), selective outcome reporting (reporting bias), and other potential sources of bias. We will assess each risk of bias parameter as having a "low risk," "high risk," or "unclear risk" of bias (information insufficient for assessment). We will resolve disagreements through discussion and will contact study investigators when methods are unclear or when additional information is required for risk of bias assessment. If investigators do not respond within two weeks, we will assess risk of bias on the basis of descriptions provided in published reports.
Measures of treatment effect
For continuous outcomes (eg, decreased IOP), we will use mean differences (MDs) to measure treatment effects and to determine 95% confidence intervals (CIs). For dichotomous outcomes, such as presence of corneal edema, we will use risk ratios (RRs) with corresponding 95% CIs to measure treatment effects. We will treat ordinal outcomes and measurement scales, such as vision‐related quality of life, as continuous data, or we will choose a cutoff and will handle it as for binary data.
Unit of analysis issues
The primary unit of analysis will be one eye per participant; except for the quality of life outcome, the unit of analysis is the participant. We also will consider studies that included two eyes per participant. If two eyes were included per participants who received the same treatment, but they are treated as a single unit (eg, average values, selection of one eye for analysis), we will consider the unit of analysis to be the participant. If the study explicitly reported binocular visual outcomes (eg, binocular visual fields, binocular visual acuity), we will consider the unit to be the participant. If two eyes are included per participants who received the same treatment, and they are analyzed as two units, we will consider the study as designed to be comparable with a cluster‐randomized study, meaning that the outcome in one eye is more likely to be similar to the outcome in its fellow eye than to the outcome in the eye of a different participant. If two eyes of a participant receive different treatments (ie, a within‐person or paired‐eye design), we will compare outcomes between the two eyes and will analyze within‐person differences when data are available.
Dealing with missing data
We will contact study authors to ask for missing or unclear information regarding study methods, outcome data, standard deviations for means, and any other desired information that had not been reported or had been reported unclearly in study reports. We will allow two weeks for study authors to respond and will move forward with best available information if we do not hear back from them. We do not plan to impute data.
Assessment of heterogeneity
We will assess clinical and methodological heterogeneity by evaluating participant characteristics, inclusion and exclusion criteria, and assessments of primary and secondary outcomes of included studies. We will use the I2 statistic (expressed as a percentage) to estimate the proportion of variation due to statistical heterogeneity that is not attributable to random error. We will consider a value above 50% as suggesting substantial statistical heterogeneity (Higgins 2011; Turner 2012). We will examine results of the Chi2 test for heterogeneity and the degree of overlap in CIs to determine effect estimates from included studies. Poor overlap of CIs may suggest the presence of heterogeneity.
Assessment of reporting biases
We will examine selective outcome reporting by comparing outcomes reported in included studies versus outcomes listed in study protocols, when available. When we have a sufficient number of studies (n ≥ 10), we will examine funnel plots of intervention effect estimates for evidence of asymmetry. A symmetrical funnel plot is expected in the absence of publication bias. An asymmetrical funnel plot may imply possible publication bias, poor reporting of small studies, true heterogeneity, or chance.
Data synthesis
When clinical or methodological heterogeneity is observed, we will not combine studies in a meta‐analysis but will present a narrative summary instead. When the I2 statistic and inspection of the forest plot do not suggest substantial statistical heterogeneity, we will combine the results of included studies in a meta‐analysis by using a random‐effects model. We will use a fixed‐effect model when fewer than three studies are included in a meta‐analysis, and when no evidence suggests statistical, clinical or methodological heterogeneity.
Subgroup analysis and investigation of heterogeneity
We plan to conduct subgroup analyses for race, primary angle‐closure subtype (PAC, PACG), and disease severity, when data are provided in included studies.
Sensitivity analysis
We will conduct sensitivity analysis to evaluate the effects of excluding studies judged as having high risk of bias for incomplete outcome data or selective outcome reporting. We also will conduct a sensitivity analysis after excluding studies sponsored by drug companies.
Summary of findings
We will summarize the main findings, including strengths and limitations of evidence for both primary and secondary outcomes, using the Grades of Recommendation, Assessment, Development and Evaluation (GRADE) approach (GRADEpro 2015). We will provide a summary of the effectiveness of the intervention. We will provide a general interpretation of the evidence in the context of other evidence and implications for practice and future research. We will present a 'Summary of findings' table when appropriate.
