Clinical outcomes after small-incision lenticule extraction versus femtosecond laser-assisted LASIK for high myopia: A meta-analysis

Authors: Fu Y, Yin Y, Wu X, Li Y, Xiang A, Lu Y, Fu Q, Hu T, Du K, Wen D.

Geographical coverage: Not reported

Sector: Biomedical

Sub-sector: Treatment

Equity focus: None

Study population: Patients with high myopia who have received SMILE or FS-LASIK treatment.

Review type: Other review

Quantitative synthesis method: Systematic review and meta-analysis

Qualitative synthesis method: Not applicable

Background: Small-incision lenticule extraction (SMILE) and femtosecond laser-assisted laser in-situ keratomileusis (FS-LASIK) are currently popular corneal refractive surgery options. FS-LASIK is effective, safe, and predictable for treating myopia, but its application is limited due to potential risks such as treatment regression, changes in corneal biomechanics, and flap complications. SMILE, a newer option, minimises these risks by replacing the production of a corneal flap with a smaller incision. While many studies have examined the clinical efficacy of both procedures, most focus on correcting low to moderate myopia without categorising by degree of myopia. Comparative studies specifically targeting high myopia populations are few.

Objectives: To compare postoperative clinical outcomes of high myopia after small-incision lenticule extraction (SMILE) and femtosecond laser-assisted laser in-situ keratomileusis (FS-LASIK).

Main findings:

To summarize, the authors discovered that both SMILE and FS-LASIK are safe, effective, and reliable for patients with high myopia. However, the SMILE procedure showed superior results in terms of postoperative corrected distance visual acuity (CDVA) compared to FS-LASIK. Additionally, SMILE resulted in fewer induced aberrations than FS-LASIK.

A total of 12 studies met the inclusion criteria and were included in this meta-analysis. Overall, these included studies had good quality (average NOS score: 6.5). The studies involved 766 patients (1,400 eyes: 748 receiving SMILE and 652 receiving FS-LASIK). Pooled results revealed no significant differences in the following outcomes: the logarithm of the mean angle of resolution (logMAR) of postoperative uncorrected distance visual acuity (weighted mean difference (WMD) = -0.01, 95% confidence interval (CI): -0.02 to 0.00, I² = 0%, p = 0.07 at 1 mo; WMD = -0.00, 95% CI: -0.01 to 0.01, I² = 0%, p = 0.83 at 3 mo; WMD = -0.00, 95% CI: -0.01 to 0.00, I² = 32%, p = 0.33 in the long term), and the postoperative mean refractive spherical equivalent (WMD = -0.03, 95% CI: -0.09 to 0.03, I² = 13%, p = 0.30). However, the SMILE group had significantly better postoperative corrected distance visual acuity (CDVA) than the FS-LASIK group (WMD = -0.04, 95% CI, -0.05 to -0.02, I² = 0%, p <0.00001)

Results of sensitivity analysis, conducted by the authors, showed that when the study by Li-kun Xia et al was excluded [17], the I² value of UDVA within the one month and three month subgroups reduced sharply, and the p value showed a stable significant difference. Publication bias was estimated by applying Egger’s test [25] (p = 0.207 to 1.000) and Begg’s test [26] (p = 0.246 to 1.000) to the 12 studies; these tests indicated no obvious publication bias.

Methodology:

The meta-analysis included randomised or non-randomised clinical trials involving patients with high myopia, defined as a preoperative spherical equivalent refractive error worse than -6.00 dioptres or a combination of spherical refractive error worse than -5.00 dioptres and cylindrical refractive error worse than -1.00. The interventions compared were SMILE and FS-LASIK, with outcomes including visual acuity, aberration, and other clinical parameters. Only original clinical articles with independent data were included, with no date restrictions. Exclusions were made for repeated publications, unpublished literature, abstracts, case reports, reviews, letters, comments, non-comparative studies, non-human investigations, and reports with incorrect or incomplete data.

Two reviewers independently searched several databases, including PubMed, EMBASE, the Cochrane Central Register of Controlled Trials, and three Chinese databases, from March 2018 to July 2020. They screened titles and abstracts, assessed potential reports, and selected articles based on inclusion criteria. Disagreements were resolved through discussion. Data was extracted using a customised form, and to reduce heterogeneity due to varying follow-up intervals, outcomes were presented in subgroups defined by follow-up time.The quality of the included studies was assessed with the Newcastle-Ottawa Scale (NOS), which was adopted to evaluate the cohorts.

Meta-analysis conducted in the statistical program RevMan 5.3, using weighted mean difference (WMD) and the corresponding 95% confidence interval (CI) to calculate the continuous outcomes. First, the I² statistic was used to test the heterogeneity of the included literature, and fixed-effect modelling was carried out when there was no statistical heterogeneity among studies (p >0.1, I² <50%). Conversely, random-effect modelling was used for analysis when the included literature bore significant evidence of statistical heterogeneity (p <0.1, I² >50%). In order to evaluate the robustness of the statistical model, authors conducted a sensitivity analysis.

Applicability/external validity: Authors note that the results of this meta-analysis should be interpreted in the context of several important limitations. First, the number of included clinical trials was relatively small, and only one randomised trial was included, which increased the risk of various types of bias. Second, the processed screening results showed that most of the included studies were performed in Asia, which may have caused publication bias. In addition, the extracted aberration data included various measurements from different wave-front analysers, which increased the methodological bias.

Geographic focus: Not stated on a study-by-study basis, although authors note that most of the included studies were conducted in Asia.

Summary of quality assessment:

The methods used for identifying, including, and evaluating studies were generally thorough, but the search was somewhat restricted as it did not review the references of included items, nor did it attempt to contact relevant experts or include unpublished material. The data analysis methods were fairly rigorous, but there was no separate analysis of studies based on their risk of bias, and unit of analysis errors were not discussed. Due to these reasons, the confidence in the review’s findings is considered low.