Comparative Efficacy of Subthreshold Micropulse Laser Photocoagulation versus Conventional Laser Photocoagulation for Diabetic Macular Edema: A Meta-Analysis

Diabetic macular edema (DME) is defined as an abnormal thickening of the macula associated with the accumulation of excess fluid in the extracellular space of the neurosensory retina that causes visual deterioration [1]. In addition to optimizing the therapy in internal medicine, anti-vascular endothelial growth factor (anti-VEGF) drugs, steroids, macular laser photocoagulation, and surgery are the mainstays of therapy for DME. Although anti-VEGF therapy can reduce macular edema and improve vision in a short term, its costs (expensive drugs and repeated surgeries) cause a heavy economic burden on patients. Macular laser photocoagulation remains an irreplaceable treatment for DME. The Early Treatment of Diabetic Retinopathy Study (ETDRS) confirmed, for clinically significant DME, macular laser photocoagulation reduced the incidence of visual loss by approximately 50% at 3 years [2, 3].

Conventional laser photocoagulation (CLP) is performed using a continuous wave (CW) laser that produces a visible burn on the area of diffuse leakage or focal non-perfusion. The laser energy is predominantly absorbed by one of the layers of the retina, the retinal pigment epithelium, and converted into heat [4]. Although CLP is proven effective in reducing the risk of visual loss, its beneficial effects are accompanied by the destruction of retinal photoreceptors, progressive enlargement of laser retinal scars, and development of choroidal neovascularization and subretinal fibrosis [5].

Reducing laser exposure duration and using a subvisible clinical endpoint for therapy can decrease the complications above [6, 7]. Subthreshold micropulse laser photocoagulation (SMLP), which no visible scar occurs and the burns remain below the observability limit [8, 9], has been introduced in the management of DME. Unlike CLP, SMLP is a laser emission technique in which the standard CW emission is chopped in a series of repetitive micropulses (short laser pulses), each one separated by a relatively long off-time which reduces the increased temperature in the tissue that follows CW laser [2, 4]. Repetitive micropulses can produce higher peak tissue temperatures and more abrupt tissue temperature changes compared to CW laser, whereas pulse spacing permits the tissue to cool between pulses, minimizing heat buildup [10].

Recently, some scholars consider SMLP appears to be the ideal first-choice therapy for DME [4, 11, 12]. Despite the proven effectiveness of SMLP, lack of standardization has limited its clinical application and usefulness [13]. Therefore, we undertook a meta-analysis to assess the efficacy and safety of SMLP compared with CLP for the management of DME. On this basis, comparisons among different parameters of SMLP were also performed to obtain the optimal laser therapy for DME.

This study was a systematic review and meta-analysis designed according to the considerations and items of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis Protocols (PRISMA-P) [14]. Registration number: CRD42022342178. The PRISMA 2020 Checklist is presented as online Supplementary Table S1 (for all online suppl. material, see www.karger.com/doi/10.1159/000529224).

A literature review was performed to identify all relevant studies that compared the outcomes between SMLP and CLP for DME. The PubMed, Embase, Web of Science, Cochrane, SinoMed, ClinicalTrials.gov, Wanfang, and China National Knowledge Infrastructure (CNKI) databases were searched systematically for all articles published before Dec 2021. All English and Chinese articles were screened. The following terms were used for the search: (“micropulse” OR “subthreshold” OR “pulse” OR “pulsed”) and (“laser” OR “photocoagulation”) and (“macular edema” OR “diabetic retinopathy” OR “DME”). Reference lists of all retrieved articles were searched manually to broaden the search. All articles are managed with Endnote X7. Take PubMed as an example: (((((micropulse[Title/Abstract]) OR (subthreshold[Title/Abstract])) OR (pulse[Title/Abstract])) OR (pulsed[Title/Abstract])) AND ((laser[Title/Abstract]) OR (photocoagulation[Title/Abstract]))) AND (((macular edema[Title/Abstract]) OR (diabetic retinopathy[Title/Abstract])) OR (DME[Title/Abstract])).

The inclusion criteria were as follows: (1) Study design: Randomized controlled trials (RCTs) that comparing the outcomes between SMLP and CLP for DME; (2) Study object: Patients have been diagnosed with DME without age, sex, race limited; (3) Intervention: Studies included SMLP or CLP; (4) Follow-up time: Studies with a minimum follow-up period of 3 months; (5) Outcome evaluation index: The outcomes were evaluated with one or more comparisons as follows: improvement of best corrected visual acuity (BCVA), reduction of central macular thickness (CMT), and change in contrast sensitivity (CS) at the end of follow-up. The exclusion criteria were as follows: (1) patients with previous treatment in the study eyes; (2) noncomparative studies, abstracts, letters, editorials and experts opinions and reviews without original data; (3) the outcomes and parameters of patients were not clearly reported and it was impossible to extract or calculate appropriate data from the published results; (4) only title related to SMLP while content was not mentioned; (5) articles repeated published or with duplicated contents.

Two reviewers independently reviewed study titles and abstracts, and studies that satisfied the inclusion criteria were selected. For each selected study, we extracted the following data: authors, year of publication, country, sample size, average age, duration of the study, laser parameters (wavelength, duty cycle, spot diameter, etc.), preoperative BCVA, preoperative CMT, preoperative CS and their changes after laser photocoagulation. Discrepancies between the two reviewers were resolved by discussion or a third reviewer.

All selected studies were RCTs. Two independent reviewers assessed the quality and discrepancies were resolved by discussion. The quality of each study was assessed by using a 0–5 grade of Jadad scores [15]. Each study was evaluated in three main aspects of its trial design: randomization, masking, and participant withdrawals/dropouts. Studies with a score of 3 or higher were considered high quality.

The meta-analysis was conducted using Review Manager (V.5.3, the Cochrane Collaboration, Oxford, UK) and Stata software (version 12.0; Stata Corp, College Station, TX, USA). We evaluated the efficacy of SMLP on three outcomes: improvement of BCVA, presented as logarithm of the minimal angle of resolution (logMAR); reduction of CMT, measured with optical coherence tomography; change of CS, presented as decibels. As all these outcomes were continuous data, the means and standard deviations were used to calculate the estimated mean difference (MD) between SMLP group and CLP group, reported with a 95% confidence interval (CI). p < 0.05 was considered statistically significant on the test for overall effect. Heterogeneity was assessed by calculating the I² statistic and performing the ?² test (to assess the p value). The I² statistic refers to the proportion of total variation observed between the trials rather than the sampling error. Higgins [16] reported that the I² ranges were from 0% to 100% and a greater I² value meant a higher chance of heterogeneity. An I² value larger than 50%, indicated a moderate to high heterogeneity. A fixed-effects model was used when no heterogeneity was detected, which meant that there were no variances among all studies. If any heterogeneity existed, a random-effects model was used for the meta-analysis and some possible moderators were tested to explore the heterogeneity. In addition, subgroup analysis with different laser parameters was used for the study of heterogeneity. Publication bias was explored by searching for asymmetry in the funnel plot.

A total of 891 articles were initially identified. We screened titles and abstracts of these for potentially relevant articles, 883 publications were excluded according to the selection criteria and a total of 8 studies [17‒24] were eligible for this meta-analysis. The search process is illustrated in Figure 1.

In total 8 studies [17‒24], 546 eyes with DME (275 eyes treated with SMLP, 271 eyes treated with CLP) were included. Two studies [22, 23] were conducted in China, 1 study [18] was from Portugal and England, the remaining one of each was from India [17], Denmark [19], Brazil [20], Italy [21], and Iran [24], respectively. There were no statistically significant differences in outcome evaluation index (BCVA, CMT, and CS) between the SMLP and CLP groups at baseline for all selected studies. Jadad scores were used to assess the quality of each study and all studies fulfilled the quality criteria (3 points or more). The main characteristics and quality scores of the included studies are shown in Table 1.

Five studies [19‒21, 23, 24] including 343 eyes reported the improvement of BCVA (logMAR) after laser photocoagulation (SMLP vs. CLP). One study [23] used 577 nm micropulse laser photocoagulation to treat DME and the remaining 4 studies [19‒21, 24] employed micropulse laser photocoagulation with the wavelength of 810 nm in SMLP group. Moreover, 2 studies [19, 21] in which eyes randomized to receive SMLP were treated with micropulse laser photocoagulation at 5% duty cycle (the frequency of the train of micropulses) and the others [20, 23, 24] employed 15% duty cycle micropulse laser photocoagulation. There was high statistical heterogeneity among the studies (heterogeneity p = 0.00, I² = 98%) which might result from the variations in the parameters (wavelength, duty cycle) of SMLP. By using subgroup analysis according to different parameters, no significant heterogeneity was identified among studies in subgroups (p > 0.05). Figure 2 showed the results of this meta-analysis which compared the improvement of BCVA between SMLP and CLP groups. In terms of BCVA, both 577 nm 15% duty cycle and 810 nm 15% duty cycle SMLP had a statistically significant higher efficacy than CLP (MD = -0.02, 95% CI: -0.03 to -0.01, p < 0.01; MD = -0.09, 95% CI: -0.09 to -0.09, p < 0.01), whereas the efficacy of 810 nm 5% duty cycle SMLP was numerically, but nonsignificantly, superior to CLP (MD = -0.01, 95% CI: -0.13 to 0.10, p = 0.82).

Changes in CMT after laser photocoagulation were reported in 8 studies [17‒24] including 546 eyes. Among the 8 studies, SMLP of different wavelengths (577 nm [17, 23] and 810 nm [18‒22, 24]) and duty cycles (5% [17, 19, 21, 22] and 15% [18, 20, 23, 24]) were applied. High statistical heterogeneity was also detected among the studies (heterogeneity p = 0.00, I² = 93%). By using subgroup analysis according to different wavelengths and duty cycles, no significant heterogeneity was identified among studies in subgroups (p > 0.05). Meta-analysis of changes in CMT between SMLP and CLP groups is shown in Figure 3, which indicated that patients receiving 577 nm or 810 nm 15% duty cycle SMLP would achieve more reduction in CMT than patients receiving CLP (MD = -32.87, 95% CI: -37.61 to -28.13, p < 0.01; MD = -8.01, 95% CI: -9.06 to -6.96, p < 0.01), while the efficacy of 577 nm and 810 nm 5% duty cycle SMLP remained numerically superior to CLP, but nonsignificantly (MD = -24.70, 95% CI: -58.86–9.46, p = 0.16; MD = -3.19, 95% CI: -25.31–18.94, p = 0.78, respectively).

Three studies [17, 21, 23] reported changes in CS after laser photocoagulation between SMLP and CLP groups including 188 eyes. It was important to note that heterogeneity testing indicated moderate heterogeneity between the two groups (heterogeneity p = 0.29, I² = 19%) and a random-effects model was used. The results of changes in CS are shown in Figure 4, CS increased in SMLP group (no matter parameter settings), whereas it decreased in CLP group and the difference was significant between the two groups (MD = 1.96, 95% CI: 1.47–2.46, p < 0.01).

Funnel plots adopted for outcome of BCVA, CMT, and CS are shown in Figure 5a–c, respectively. Based on a visual analysis of the funnel plot, the approximate symmetry indicated low publication bias.

The prevalence of diabetes mellitus is increasing worldwide, with estimates indicating that DM affected 285 million adults (6.4% of the worldwide population) in 2010 [25]. This figure is projected an increase up to 439 million (7.7%) by 2030 [25], thus the prevalence of DME and visual impairment due to DME is expected to increase substantially over time. In the past years, many therapies have been proposed for the management of DME including laser photocoagulation, ocular steroids, intravitreal anti-VEGF drugs, and vitreoretinal surgery [26]. Once macular edema appears, laser photocoagulation may be beneficial, but irreversible damage of retina has also been observed and reported. SMLP is designed to target the retinal pigment epithelium with minimal effect on the neurosensory retina and choroid. In this modality, the laser energy is transmitted as micropulses, which has “on” and “off” cycles rather than as a CW with hundreds of milliseconds cycle duration as with the CL. Heat is distributed by long “off” cycles and energy transfer to the tissue is limited by maximum absorption of the laser energy [8, 27]. The main determinant of average tissue temperature rise is duty cycle and clinical studies have demonstrated that treatment safety is maximized by the use of 5% or lower duty cycles [10, 28]. In addition, SMLP is available in different wavelengths (577 nm and 810 nm), which may result in different efficacy of micropulse laser photocoagulation [29]. This meta-analysis aims to assess the efficacy and safety of SMLP versus CLP for the management of DME and compare the different parameters of SMLP to evaluate the optimal one.

We reviewed 8 RCTs [17‒24] involving a total of 546 eyes (275 eyes in SMLP group and 271 eyes in CLP group). The pooled outcomes from our meta-analysis indicated that SMLP, especially 577 nm and 810 nm 15% duty cycle parameter settings, had a statistically significant higher efficacy than CLP in terms of BCVA (MD = -0.02, 95% CI: -0.03 to -0.01, p < 0.01; MD = -0.09, 95% CI: -0.09 to -0.09, p < 0.01) and showed more significant advantages than CLP in resolving macular edema assessed by reduction of CMT (MD = -32.87, 95% CI: -37.61 to -28.13, p < 0.01; MD = -8.01, 95% CI: -9.06 to -6.96, p < 0.01), whereas the efficacy of 577 nm and 810 nm 5% duty cycle SMLP remained numerically superior to CLP, but nonsignificantly (p > 0.05). In the field of CS, SMLP (no matter parameter settings) resulted in better preservation of CS compared to CLP (MD = 1.96, 95% CI: 1.47–2.46, p < 0.01). This study provides important findings that may be helpful in the selection of laser photocoagulation for the treatment of DME.

Our meta-analysis is the third, as per our knowledge, to provide statistical results by comparing the anatomical and functional outcomes to evaluate the effect of SMLP versus CLP for DME. Different from the previous meta-analysis, comparisons among different parameters of SMLP were also performed in this study, which was the characteristic of this study. Our findings showed that SMLP got superior efficacy and safety on improvement of BCVA, reduction of CMT, and preservation of CS, which were consistent with Chen et al. [30]. However, Qiao et al. [31] suggested that SMLP showed an equally good effect on visual acuity, CS, and CMT compared to CLP. We considered the reasons leading to the different results with Qiao et al. [31] may be the different indexes to evaluate the efficacy of SMLP. Additionally, our study also reported anatomical and functional outcomes using SMLP of two different wavelengths (577 nm and 810 nm) and duty cycles (5% and 15%) among patients with DME. Figures 2,,3 showed that the differences in wavelength and duty cycle came to different effects. However, due to the lack of parameter and protocol detail information, the results we obtained were relatively limited. In the future, further clarification in terms of different parameter settings with larger sample size is essential.

It is worthy to note the relatively limited powers of our meta-analysis when considering the results. First, resource use of the different laser treatments, including staff time, equipment required, overheads, consumables, and any rescue treatments (anti-VEGF therapy) results in huge expenditure and lack of financially motivated study sponsors, which causes the notably small number of comparative RCTs available for this study. Second, in included studies, outcomes are measured at different follow-up times and this may induce heterogeneity. Third, the parameters and settings of the laser photocoagulation in each study are different, such as duty cycle, spot size, exposure time, and so on. Although we do our best to control these differences among studies, different instruments, parameters and operators could still affect the results. Furthermore, Citirik [27] reported that anatomical severity of DME may affect the treatment response to SMLP, which may also affect the results. Finally, this meta-analysis is limited to use published index, and the papers published in languages other than English and Chinese may have failed to be included.

Leaving the limitations aside, we believe that the results of this meta-analysis are credible. Compared with CLP, SMLP may get superior efficacy and safety on improvement of BCVA, reduction of CMT, and preservation of CS. SMLP can be considered as a safe and effective therapy in the management of DME.

The authors thank all the colleagues in Jiangyin People’s Hospital who once offered the valuable courses and advice during my study.

The research was conducted ethically in accordance with the World Medical Association Declaration of Helsinki. Our research did not include original overview of clinical trials, and this article is a summary analysis of published data. Therefore, ethical approval was not required.

The authors have no conflicts of interest to declare.

This work was supported by Postgraduate Research and Practice Innovation Program of Jiangsu Province (SJCX18_0063).

All authors contributed to the study conception and design. Literature search and screening, data collection and analysis, generation of the figures, data interpretation, and preparation of the manuscript were performed by Xin-ying Hu, Li Cao, and Jie Luan. Study design, study analysis, writing of the discussion, and revision of the manuscript were performed by Ye Gao and Xue-dong Xu. All authors read and approved the final version of the manuscript.

All data generated or analyzed during this study are included in this article and its online supplementary material files. Further inquiries can be directed to the corresponding author.