Comparative Efficacy of Phacotrabeculectomy versus Trabeculectomy with or without Later Phacoemulsification: A Systematic Review with Meta-analyses

Author: Ahmadzadeh A, Kessel L, Subhi Y, Bach-Holm D.

Geographical coverage: USA, Canada, Chile, Australia, Europe and Asia

Sector: Cataract surgery

Sub-sector: Efficacy

Equity focus: Not reported

Study population: Patients with glaucoma and co-existing cataract

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background:

Cataract and glaucoma are the two leading causes of global blindness and frequently coexist, especially in older adults. Up to 10 % of patients presenting for cataract surgery have glaucoma or ocular hypertension. Determining the optimal surgical sequence is challenging: trabeculectomy lowers intra-ocular pressure (IOP) but can hasten cataract formation or fail if followed by cataract extraction, whereas cataract surgery first may trigger IOP spikes that endanger the optic nerve. Phacotrabeculectomy—combined cataract extraction and trabeculectomy—was developed to address these issues, yet concerns over complications have limited its uptake.

Objective:

To compare the efficacy and safety of phacotrabeculectomy with those of trabeculectomy, with or without subsequent phacoemulsification.

Main findings:

The review included 25 comparative studies (4 749 eyes): six prospective and 19 retrospective. Studies were conducted in North and South America (USA, n=3; Canada, n=2; Chile, n=1), Europe (UK, n=2; Italy, n=2; Switzerland, n=1; Belgium, n=1; Turkey, n=1), and Asia (China, n=3; Singapore, n=1; Japan, n=1; Hong-Kong, n=1; South Korea, n=2; Iran, n=1; Israel, n=1; Saudi Arabia, n=1and Australia.

Authors reported no statistically significant difference between phacotrabeculectomy and trabeculectomy, either without later phaco (MD = 0.63 mmHg; 95 % CI –0.32 to 1.59) or with later phaco (MD = –0.52 mmHg; 95 % CI –1.45 to 0.40). Phacotrabeculectomy yielded better logMAR acuity (MD = –0.14; 95 % CI –0.27 to –0.01), equivalent to ~1.4 lines of improvement.

Authors found fewer overall complications with phacotrabeculectomy than with trabeculectomy alone (RR = 0.80; 95 % CI 0.67 to 0.95). Rates were comparable to the sequential procedure. In addition, the review findings revealed no material differences in visual-field progression, needling or revision procedures, glaucoma-medication burden, or surgical success.

Overall, combined surgery achieves IOP control similar to trabeculectomy while improving vision and reducing complications. Evidence quality is, however, low to moderate because most studies were observational.

Methodology:

MEDLINE, Embase and CENTRAL were searched (1997 – 20 January 2020) for English-language studies comparing phacotrabeculectomy with trabeculectomy (with or without later phacoemulsification). Two reviewers independently screened records, extracted data and assessed quality with GRADE; disagreements were resolved by a third reviewer. Random-effects meta-analyses produced pooled mean differences or risk ratios; heterogeneity was explored with the I² statistic. Reference lists of included studies were hand-searched for additional reports.

Applicability / external validity:

Most evidence derives from non-randomised studies with varied surgical techniques, patient characteristics and use of antimetabolites, limiting generalisability. Findings should therefore be interpreted cautiously and may not translate directly to all clinical settings.

Geographic focus:

Studies were undertaken in the USA, Canada, Chile, Australia, several European countries, and across Asia; no geographical restrictions were applied.

Summary of quality assessment:

Overall, there is medium confidence in the review’s conclusion. Searches were appropriate, eligibility criteria clear, and dual-reviewer processes employed. Lists of included and excluded studies were provided and evidence quality graded. Limitations include English-language restriction, absence of contact with study authors, and synthesis not stratified by risk-of-bias level.