iStent with Phacoemulsification versus Phacoemulsification Alone for Patients with Glaucoma and Cataract: A Meta-Analysis

Background Minimally invasive glaucoma surgeries (MIGS) have attracted significant attention, as they have been reported to lower intra-ocular pressure (IOP) and have an excellent safety profile. The iStent is an example of a minimally invasive glaucoma device that has received particular attention due to its early and wide spread utilization. There is a growing body of evidence supporting its use at the time of phacoemulsification to help lower IOP. However, it is still not clear how much of the IOP lowering effect can be attributed to the iStent, the crystalline lens extraction or both when inserted concurrently at the time of phacoemulsification. This has been an important issue in understanding its potential role in the glaucoma management paradigm. Purpose To conduct a systematic review and meta-analysis comparing the IOP lowering effect of iStent insertion at the time of phacoemulsification versus phacoemulsification alone for patients with glaucoma and cataracts. Methods A systematic review was conducted utilizing various databases. Studies examining the IOP lowering effect of iStent insertion in combination with phacoemulsification, as well as studies examining the IOP lowering effect of phacoemulsification alone were included. Thirty-seven studies, reporting on 2495 patients, met the inclusion criteria. The percentage reduction in IOP (IOPR%) and mean reduction in topical glaucoma medications after surgery were determined. The standardized mean difference (SMD) was computed as a measure of the treatment effect for continuous outcomes taking into account heterogeneity. Fixed-effect and random-effect models were applied. Results A 4% IOP reduction (IOPR%) from baseline occurred following phacoemulsification as a solo procedure compared to 9% following an iStent implant with phacoemulsification, and 27% following 2 iStents implants with phacoemulsification. Compared with cataract extraction alone, iStent with phacoemulsification resulted in significant reduction in the post-operative IOP (IOPR%) (SMD = -0.46, 95% CI: [-0.87, -0.06]). A weighted mean reduction in the number of glaucoma medications per patient was 1.01 following phacoemulsification alone compared to 1.33 after one iStent implant with phacoemulsification, and 1.1 after 2 iStent implants with phacoemulsification. Compared to cataract extraction alone, iStent with cataract extraction showed a significant decrease in the number of glaucoma medications (SMD = -0.65, 95% CI: [-1.18, -0.12]). Funnel plots suggested the absence of publication bias. Conclusion Both iStent implantation with concurrent phacoemulsification and phacoemulsification alone result in a significant decrease in IOP and topical glaucoma medications. In terms of both reductions, iStent with phacoemulsification significantly outperforms phacoemulsification alone.


Purpose
To conduct a systematic review and meta-analysis comparing the IOP lowering effect of iStent insertion at the time of phacoemulsification versus phacoemulsification alone for patients with glaucoma and cataracts.

Methods
A systematic review was conducted utilizing various databases. Studies examining the IOP lowering effect of iStent insertion in combination with phacoemulsification, as well as studies examining the IOP lowering effect of phacoemulsification alone were included. Thirtyseven studies, reporting on 2495 patients, met the inclusion criteria. The percentage reduction in IOP (IOPR%) and mean reduction in topical glaucoma medications after surgery were determined. The standardized mean difference (SMD) was computed as a measure of Introduction Open angle glaucoma (OAG) is the leading cause of irreversible vision loss [1,2] affecting 60.5 million people globally. The prevalence of OAG is projected to rise to nearly 80 million people by 2020. [3,4] OAG is an enormous public health burden, with the direct annual cost of treatment estimated at 2.86 billion (USD) in the U.S. [5] and 300 million (CAD) in Canada [6] including productivity loss. These costs are cumulative and can have a huge financial impact for patients, care givers, and for society at large. [7] Additionally, the social consequences of severe OAG includes difficulties in performing daily activities, increased risk of falls [8], injuries, depression [9], hip fractures [10], and increased risk of death [1]. Compared to people with normal vision, people with low vision are prematurely admitted to nursing homes. [11] At present, the only known treatment to lower the risk of visual loss secondary to glaucoma is lowering of the intraocular pressure (IOP). In the traditional treatment paradigm, topical medications are the first-line of therapy for OAG. Despite proven efficacy and a good safety profile, there are a number of issues with medications. Compliance is a major concern with any chronic medication use. [12][13][14] It has been estimated that only 67% of patients are compliant with topical glaucoma medications [15] and that rate decreases when multiple medications are utilized. [16] Side-effects such as ocular and peri-ocular erythema, transient blurred vision, itching, local irritation, and ocular surface toxicity [17][18][19], eye discomfort, photosensitivity, increased tears, burning and stinging, eyelash changes, hyperpigmentation of the eyelid and irides have been associated with poor compliance to topical medications. [20] Additionally, medications can be difficult to administer. [21,22] Micro-invasive glaucoma surgery (MIGS) and devices have been gaining increasing attention in recent years. One such device is the trabecular microbypass stent, or iStent (Glaukos Corporation, Laguna Hills, CA). It is the smallest implantable medical device which is inserted ab interno into Schlemm's canal [23], causing a lowering of IOP. [24,25] Part of the appeal of the iStent is that it circumvents issues related to patient compliance, and has a favorable safety profile. In addition, a cost-analysis suggests that the iStent may also be more cost effective than topical medications. [7] There are emerging data supporting the use of iStent as an adjunctive therapy at the time of phacoemulsification for patients who are on multiple glaucoma medications. [24,[26][27][28][29][30] Most of the studies examining the effect of iStent on IOP have been small or have lacked an adequate control arm. The lack of a control arm is a concern, as phacoemulsification alone has been found to lower IOP in both short and long term studies. [31][32][33] The primary objective of the current systematic review and meta-analysis is to compare the IOP lowering effect of iStent insertion at the time of phacoemulsification versus phacoemulsification alone for patients with glaucoma and cataract. A secondary aim is to examine visual acuity outcomes and topical glaucoma medication use for both groups.

Literature Retrieval
We adhered to the Preferred Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (S1 File). [34] Computer databases including MEDLINE (OVID and Pubmed), EMBASE (OVID), BIOSIS Previews (Thomson-Reuters), CINAHL (EBSCO), Health Economic Evaluations Database (HEED), ISI Web of Science (Thomson-Reuters) and the Cochrane Library (Wiley) were searched from the year 2000 to June 2014. The reference lists of all the included articles were hand searched to find potentially relevant studies. Grey literature was identified by searching the conference abstracts of various meetings including the Canadian Ophthalmology Society meeting (COS), American Academy of Ophthalmology annual meeting (AAO), European Society of Ophthalmology (SOE), American Glaucoma Society (AGS), and the Association for Research in Vision and Ophthalmology annual meeting (ARVO). The Proquest Dissertations and Theses databases and the Canadian Health Research Collection (Ebrary) were also searched for relevant content. Google and other internet search engines were used to search for additional web-based materials and information. OVID Auto-Alerts were set up to send monthly updates with any new literature.
Database specific subject headings and key words for "phacoemulsification", "iStent", "open angle glaucoma", "ocular hypertension" were employed in the search strategy. The searches were modified to accommodate the unique terminology and syntax of each database. Detailed search strategy for EMBASE has been provided in the supplementary material (S2 File). Searches were limited to human subjects and studies published between January 2000 and June 2014. Year 2000 was selected to be consistent with the way phacoemulsification surgeries are performed.

Inclusion and Exclusion Criteria
A minimum post-op period of two months was required for inclusion in order to limit the confounding effects of short-term fluctuations in IOP that can occur in the immediate post-operative period due factors such as retained viscoadaptive devices. Only studies published in the English language were included. Research studies such as journal articles, systematic reviews, meta-analysis, cost analysis, cost-utility analysis, cost-effectiveness analysis, multicenter studies, randomized controlled trials, quasi-randomized controlled trials, non-randomized studies including cohort studies (retrospective, prospective), clinical trials, and comparatives studies, were included for analysis. Editorial opinions, case reports, or reviews were excluded from this analysis. A minimum sample size of 20 patients was required for inclusion. Additionally, the study population included adults above the age of 18 with OAG or ocular hypertension. Studies examining subjects with other types of glaucoma were excluded.  All the records identified through database searching and grey literature search were imported to EPPI-Reviewer 4 gateway (by EPPI-Centre, Social Science Research Unit, the Institute of Education, University of London, UK). Duplicate records were removed and the included records were screened. Screening was done at three levels. Titles of the included records were screened at Level 1; abstracts of the records included in Level 1 were screened in Level 2; and in Level 3, full texts of the included articles were screened. Detailed screening questions are provided in the supplementary material (S3 File). Two reviewers independently screened the records at each level. Cohen's kappa (κ) coefficient was computed to assess the agreement of inclusion between the two reviewers after each level of screening. Articles were included for next level of screening if two reviewers agreed. In the case of disagreements, a third reviewer intervened to provide a decision.

Data Extraction
One reviewer extracted the data from eligible articles and a second reviewer verified the data extracted. Study data included study objective, design, location, inclusion and exclusion criteria, data collection technique, data collection period, follow-up, intervention performed such as phacoemulsification only or iStent with phacoemulsification or both, and number of iStents inserted. Additionally, data on total patients enrolled in the study, total patients completing the study, refusal to consent, and patient demographic characteristics such as mean age, standard deviation (SD), gender, race and ethnicity (%) were collected from each included article. Baseline and post-operative characteristics such as best visual acuity corrected (BCVA), uncorrected visual acuity, IOP, number of glaucoma medications, reported complications and their rates, percentage of patients off-medications, patients with 20/40 or better vision, patients with IOP 18 mmHg post-operatively, rate of iStent malposition, iStent occlusion, and allergic reaction to glaucoma medications were also obtained.
The available data on range, p-value, and confidence interval were utilized and converted to the common effect measure, SD. Further, means and measures of dispersion were approximated from figures if required. Two reviewers assessed the methodological quality of each included study using the Downs and Black checklist. [35]

Statistical Analysis
Meta-analysis was performed using STATA v. 13.0 (STATA Corporation, College Station, TX). The abstracted mean and standard error of the IOP at baseline and end-time point were used to compute the mean IOP reduction (IOPR), percentage of IOP reduction (IOPR%), within group standard error (SE IOPR ), and standard error of percentage of IOP reduction (SE IOPR% ) Standard deviation percentage of IOP reduction (SD IOPR% ) was then calculated by the formula SD IOPR% = SE IOPR% x n 1/2 .
For continuous scale outcomes including mean values, the standardized mean difference (SMD) was calculated as the treatment effect or effect size. This parameter represents the mean difference standardized for variances across all studies. To compute SMD for each study, the mean pre-and post-operative values for each outcome measure was divided by the SD for that same outcome measure. Weights were assigned to each SMD according to the inverse of its variance and then average was computed. The SMD in each study was pooled with a fixed-or random-effect model based on heterogeneity. To test heterogeneity, a Z-value was computed to test the null hypothesis which was a treatment effect of zero. Additionally, heterogeneity was determined using the I 2 value [37], which indicated the extent of variation across studies due to heterogeneity rather than chance. Heterogeneity between studies was examined using a chi-squared test, where the chi-squared test assessed whether the observed between-studies differences were due to chance only. [38] Low p-value and a large chi-squared statistic relative to its degree of freedom provided evidence of heterogeneity. Funnel plots were generated to check publication bias.
SMD of the mean post-operative number of topical glaucoma medications was computed to evaluate the impact of "iStent with phacoemulsification" versus "phacoemulsification only" on number of topical glaucoma medications. For the sub-group analysis, SMD of the mean preoperative and post-operative number of medications was computed for phacoemulsification only, one iStent with phacoemulsification, 2 iStents with phacoemulsification, and 3 iStents with phacoemulsification. Additionally, sub-group analysis was conducted by follow-up period for "phacoemulsification only" group and for "iStents with phacoemulsification" group.

Search Results
A total of 1151 records were identified through database searching and an additional 38 records were identified through grey literature search. A total of 1189 records were imported to EPPI-Reviewer 4 and duplicate records were removed. 933 records remained for screening after removing duplicates. 147 records were retrieved after three-levels of screening. After reviewing all the full text articles, a total of 32 articles (2143 subjects) met our inclusion criteria and were utilized in our quantitative and qualitative synthesis. See

Study Characteristics
Tables 1, 2, 3, 4, 5, 6, 7 and 8 summarize characteristics of studies included for meta-analysis. Table 1 lists baseline characteristics as well as pre-and post-operative IOP, IOPR%, and Blockage of the opening of the stent lumen in 8 eyes (15), hyphema (2), stent malposition (2) Caucasian (69), African American (15), South Asian (8) standard deviation of IOPR% (SD_IOPR%) of the 4 studies comparing iStent with and without concurrent phacoemulsification. Table 2 lists baseline characteristics including pre-and postoperative IOP, IOPR%, and SD_IOPR% of the 22 studies considering phacoemulsification as a solo procedure. Table 3 lists 6 studies evaluating iStent with phacoemulsification. The followup period varied from 2 months to 7 years. A total of 18 studies were conducted in the U.S., 4 in U.K., 3 in Italy, 1 in each country including Turkey, Spain, Germany, Finland, New Zealand, Japan, and Israel. Of the 32 studies, 21 were case series, 2 were case control, 4 were randomized control trials, and 5 were cohort studies (Table 1). Tables 4, 5, and 6 lists data on pre-and postoperative topical glaucoma medications, and mean reduction in topical glaucoma medications of the 4 studies comparing iStent with and without concurrent phacoemulsification, 22 studies considering phacoemulsification as a solo procedure, and 6 studies evaluating iStent with phacoemulsification, respectively. Table 7 lists data on pre-and post-operative best corrected visual acuity (BCVA) of the included studies. Table 8 lists data on post-operative characteristics of included studies such as complications, complication rates, percentage of patients with good vision, reduced number of medications, race and ethnicity of patients, and controlled IOP.

Publication Bias
Visual inspection of funnel plots by follow-up and number of iStents implanted for both preand post-operative IOPR% and topical medications did not reveal any asymmetry (Figs 2 and 3). Effect on IOPR% A 4% IOP reduction (IOPR%) from baseline occurred following phacoemulsification as a solo procedure compared to 9% following an iStent implant with phacoemulsification, and 27% following 2 iStents implants with phacoemulsification (Table 1) based on 4 studies comparing phacoemulsification only and iStent with concurrent phacoemulsification. Based on 22 studies considering phacoemulsification as a solo procedure in patients with glaucoma and cataract, a 31% IOP reduction from baseline occurred ( Table 2). This reflects significant between study variations due to numerous factors such as patient's income status, socio-economic status, previous ocular and non-ocular surgeries, family history, other ocular and non-ocular diseases, pre-operative and post-operative medications, number of medications, comorbidities, etc. influencing the estimates in the original studies. A 26% IOP reduction from baseline occurred following an iStent implant with phacoemulsification, 18% following 2 iStents implants with phacoemulsification, and 20% following 3 iStents implants with phacoemulsification based on 6 studies considering iStent implant with concurrent phacoemulsification (Table 3). Four studies examined the post-operative IOPR% for phacoemulsification with iStent compared to phacoemulsification alone for patients with glaucoma and cataract. Non-significant (p = 0.128) heterogeneity was found between these studies (I 2 = 47.3%). A significant decrease in the post-operative IOPR% was found in the phacoemulsification with iStent group compared to phacoemulsification only group (SMD = -0.46, 95% CI: [-0.87, -0.06]) (Fig 4).

Effect on Topical Glaucoma Medication Use
A weighted mean reduction in the number of glaucoma medications per patient was 1.01 following phacoemulsification alone compared to 1.33 after one iStent implant with phacoemulsification, and 1.1 after 2 iStent implants with phacoemulsification based on 4 studies comparing phacoemulsification only and iStent with concurrent cataract procedure (Table 4). Based on 22 studies considering phacoemulsification as a solo procedure, a weighted mean reduction in the number of glaucoma medications per patient was 0.23 (Table 5). Based on 6 studies considering iStent with concurrent phacoemulsification, a weighted mean reduction in the number of glaucoma medications per patient was 1.05 after an iStent implant with phacoemulsification, 1.46 after 2 iStents implants with phacoemulsification, and 2.2 after 3 iStents implants with phacoemulsification (Table 6).
Four heterogeneous studies (I 2 = 58.1%) performed a direct comparison of the number of post-operative topical glaucoma medications used after phacoemulsification with iStent insertion versus phacoemulsification alone. When compared to the phacoemulsification alone group, there was a significant decrease in the number of medications used post-operatively for the phacoemulsification with iStent group (SMD = -0.65, 95% CI: [-1.18, -0.12]) (Fig 5).

Discussion
The aggregated results presented in this meta-analysis indicate that both iStent implantation with concurrent phacoemulsification and phacoemulsification alone result in a significant A 4% IOP reduction (IOPR%) from baseline occurred following phacoemulsification as a solo procedure compared to 9% following an iStent implant with phacoemulsification, and 27% following 2 iStents implants with phacoemulsification. A weighted mean reduction in the number of glaucoma medications per patient was 1.01 following phacoemulsification alone compared to 1.33 after one iStent implant with phacoemulsification, and 1.1 after 2 iStent implants with phacoemulsification. The maximum reduction in the number of medications after phacoemulsification alone occurred at 12 months up to until 24 months. A significant reduction in the number of medications following phacoemulsification alone occurred at 24 months. Combined iStent with phacoemulsification significantly reduced the number of post-operative medications by 12 months of surgery which remained significantly reduced until 4 years of follow-up.
There was substantial heterogeneity among studies and therefore, a random-effect model was developed when required. A number of factors likely contributed to the heterogeneity including inconsistency in the differences in the study population, technical aspects of iStent implantation, technical differences in phacoemulsification, differential peri-operative management, differences in the study population, potential variability in facilities to perform surgeries, variable follow-up periods, rates of complications, year the surgeries were performed, attention to identification of collector channels as well as potential biases related to industry funding of studies. Although funding source is a potential bias it is recognized that peer-reviewed funding for device trials is exceedingly difficult to obtain. Recognizing and appreciating this bias is sufficient.
Limitations are inherent in any systematic review and meta-analysis due to the variability in the availability of published data. First, only limited information was available on pre-operative and post-operative contrast sensitivity, spherical equivalent, refractive status, stereopsis, and astigmatism and thus these parameters were excluded from the quantitative analysis. Second, meta-analysis of observational studies is influenced by inherent biases in the included articles. [38] For example, there could be other factors such as income status, socio-economic status, previous ocular and non-ocular surgeries, family history, other ocular and non-ocular diseases, pre-operative and post-operative medications, number of medications, comorbidities such as high blood pressure, diabetes, stroke, heart conditions, etc. influencing the estimates in the original studies. Third, some studies were excluded due to lack of necessary information. Data from these studies could have influenced the results.
As devices can be approved and brought to market with relatively few, and often smaller studies than medications, systematic reviews and meta-analyses provide the ability to analyze iStent with Phacoemulsification: A Meta-Analysis available data collectively with some degree of neutralization of the biases that may temper smaller studies. In chronic and prevalent diseases like glaucoma, for which the cause and cure remain unknown, innovations may represent a major step forward in management. However, without mechanisms to provide access and implementation, the potential benefits in patient care continue to remain unknown. Unfortunately criteria for approval at multiple levels, which are often influenced significantly by cost, represent potential barriers to access. Systematic reviews and meta-analyses may help provide an evidence-based foundation that can be used by health care systems and administrators to guide important decisions regarding access to innovation in a cost-driven era of health care management.