Femtosecond laser-assisted astigmatic keratotomy versus toric IOL implantation for correcting astigmatism in cataract patients: a systematic review and meta-analysis with trial sequential analysis

Authors: Yen WT, Weng TH, Lin TY, Tai MC, Chen YH, Chang YM.

Geographical coverage: Korea, Egypt, Spain, Taiwan and China

Sector: Cataract surgery

Sub-sector: Treatment comparison

Equity focus: Not reported

Study population: Adult patients with cataract and astigmatism

Review type: Effectiveness review

Quantitative synthesis method: Meta-analysis

Qualitative synthesis method: Not applicable

Background

Cataract surgery aims to restore vision by removing cataracts, with emmetropia (no refractive error) being the goal. Factors such as residual astigmatism and errors in intraocular lens (IOL) power calculation can affect outcomes. Preoperatively, many patients have corneal astigmatism. Postoperatively, 15%–56% experience astigmatism of 1 dioptre (D) or more, which impacts vision. To address astigmatism, methods such as limbal relaxing incisions, arcuate keratotomy (AK), and toric IOLs are used. Toric IOLs are popular for their predictability and improved visual outcomes, but they can rotate after surgery, affecting accuracy. In such cases, AK, especially with femtosecond laser assistance, can be more suitable.

Objective

To evaluate the refractive and visual effectiveness of femtosecond laser-assisted astigmatic keratotomy (FSAK) compared to toric intraocular lens (IOL) implantation in the correction of astigmatism among patients with cataracts.

Main findings:

Overall, authors found that toric intraocular lenses (IOLs) are generally more effective than femtosecond laser-assisted astigmatic keratotomy (FSAK) in reducing postoperative refractive cylinder in cataract patients. The review included nine studies: two randomised controlled trials (RCTs) and seven observational studies, with a total of 590 participants. These were conducted in China (n=3), Korea (n=3), Egypt (n=1), Spain (n=1) and Taiwan refractive cylinder compared to FSAK (weighted mean difference (WMD) = 0.17; 95% CI 0.05 to 0.29). Additionally, toric IOLs were more likely to achieve a residual refractive cylinder of 1.00 dioptre (D) or less (OR = 0.23; 95% CI 0.09 to 0.63). However, FSAK was associated with a statistically significant lower mean target-induced astigmatism (TIA) (WMD = -0.48 D; 95% CI -0.82 to -0.14) and surgically induced astigmatism (SIA) (WMD = -0.46; 95% CI -0.85 to -0.07).

Future research should address the moderate risk of bias identified in the observational studies and the concerns in the randomisation process of the RCTs. Further studies should also explore the long-term outcomes and potential complications associated with both interventions to provide a more comprehensive understanding of their effectiveness and safety. Additionally, research should focus on improving study designs to minimise bias and enhance the reliability of findings.

Methodology: 

Review authors searched in EMBASE, Ovid-Medline, Cochrane Central Register of Controlled Trials, and Scopus databases using key terms like ‘femtosecond laser’, ‘toric IOL’, and ‘astigmatic keratotomy’. The final search was performed on 10 December 2023. Inclusion criteria encompassed randomised controlled trials (RCTs) and observational studies involving adult cataract patients with astigmatism, comparing FSAK and toric IOL for astigmatism correction, and reporting at least one clinical outcome such as postoperative refractive cylinder, correction index, UDVA, TIA, and SIA. Exclusion criteria included phase I and II clinical trials and studies where AK was not performed using a femtosecond laser.

Article selection was independently conducted by two review authors. Data extraction was performed by two authors using Microsoft Excel, tabulating parameters like author, publication year, study method, country, FSAK process, femtosecond laser system, types of IOL, sample size, and follow-up duration. Furthermore, quality assessment of RCTs was done by two reviewers independently using the Cochrane Risk-of-Bias Tool for Randomized Trials (RoB 2), while observational studies were evaluated using the Risk Of Bias In Non-randomized Studies of Interventions (ROBINS-I) tool.

Data synthesis and analysis were conducted using Review Manager 5.4 software. Continuous variables like postoperative refractive cylinder, correction index, UDVA, TIA, and SIA were reported as mean differences (MD) with a 95% CI. Dichotomous variables, such as the proportion of patients achieving a residual refractive cylinder of 1.00 D or less, were expressed as ORs with a 95% CI. Statistical significance was set at P<0.05. Heterogeneity was assessed using I² and Q test analyses, with significant heterogeneity identified at p<0.10 and I²>50%. Publication bias was examined through funnel plots. The inverse variance method and the Mantel-Haenszel method were used for continuous and dichotomous variables, respectively. Trial sequential analysis (TSA) was performed using TSA software V.0.9.5.10 beta

Applicability/external validity:

The included studies spanned multiple countries, increasing generalisability. However, varied follow-up periods could limit applicability. No theoretical framework was used, but trial sequential analysis supported statistical reliability.

Geographic focus:

Korea, Egypt, Spain, Taiwan and China.

Summary of quality assessment:

Overall, confidence in the review conclusions is low due to limitations in search comprehensiveness, absence of subgroup analysis by risk of bias, and insufficient methodological detail regarding data extraction.