The role and efficacy of vitrectomy for the management of refractory diabetic macular edema: Systematic review and meta-analysis

INTRODUCTION

Diabetic macular edema (DME), a manifestation of diabetic retinopathy (DR), is a major cause of severe visual loss in diabetic patients. Approximately 40% of patients with diabetes mellitus develop DME during their lifetime.^[1] The incidence of DME increases with the severity and duration of diabetes. The Wisconsin epidemiologic study of DR found that after 10 years of follow-up, 20% and 14–25% of patients with type 1 and type 2 diabetes, respectively, developed DME.^[2] The major pathophysiological cause involves the rupture of the blood-retinal barrier due to elevated vascular endothelial growth factor (VEGF) and pro-inflammatory cytokine expression. This breakdown leads to hyperpermeability, angiogenesis, and vascular leakage, ultimately resulting in DME.^[3] Vascular, neurodegenerative, and inflammatory components have been implicated as causes of DME.^[4]

In refractory cases, with the financial burden and treatment complications related to anti-VEGF therapy, there is a general need for repetitive injections to maintain their therapeutic effect due to the prolonged clinical course of DME. In chronic refractory DME, the goal of developing a combination therapy is to prolong the effectiveness of these compounds acting as angiogenesis inhibitors, thus eliminating the need for frequent injections that can have serious side effects along with cost-effectiveness.^[5] Optical coherence tomography (OCT) is considered the most important test for diagnosing and monitoring DME progression.^[5]

Controlling systemic factors such as hyperglycemia, hypertension, and hyperlipidemia is advantageous for reducing the incidence of retinopathy in patients with type 1 and type 2 diabetes mellitus.^[6,7] However, despite these interventions, a significant number of patients still experience vision loss, which has ultimately encouraged the development of pharmacological treatments for DME.^[8] The first-line treatment is anti-VEGF agents, such as aflibercept or ranibizumab, whereas corticosteroids, such as dexamethasone or fluocinolone acetonide implants, are only used as second-line treatment.^[9]

Patients with refractory macular edema, which is a persistent state of DME that is unresponsive to the current standard of care, including anti-VEGF and intravitreal corticosteroid agents or laser therapy, experience no visual acuity (VA) gain, a reduced anatomical response, and frequent injections therapy.^[9-11] The best-known interval between treatment with anti-VEGF and dexamethasone implants has been described for persistent DME in the BEVORDEX trial.^[12]

The rate of refractory DME varies among studies, with reported rates ranging from 25% to 64% in eyes with chronic DME. Understanding the prevalence and variability of refractory DME is crucial for identifying appropriate treatment strategies and may also serve as an indication for vitrectomy.^[13] New strategies are being explored to manage refractory DME, including newer anti-VEGF agents, and combination therapies such as intravitreal dexamethasone implants combined with navigated 577 nm subthreshold micropulse laser.^[4,14,15]

Corticosteroids have become increasingly used in DME management due to the limitations of anti-VEGF treatments.^[16] These molecules are powerful non-specific anti-inflammatory agents that inhibit adhesion, leukostasis, and transmigration of leukocytes, downregulation of cytokine and prostaglandin expression, and growth factors, especially VEGF, both in vivo and in vitro.^[17,18] The mechanism of action of corticosteroids in DME treatment is thought to be multifactorial and is considered the most effective against DME when delivered intravitreally.^[17,19] According to the DR clinical research (DRCR) Network, randomized controlled trials (RCTs) have shown that peribulbar triamcinolone acetonide is not beneficial in patients with mild DME.^[20] Three potent synthetic corticosteroids, triamcinolone acetonide, fluocinolone acetonide, and dexamethasone, have been investigated as intravitreal treatments for DME. The elimination half-life of these agents has been reported to be 2–3 h in the vitreous humor of animal models.^[21] Chemicals (such as triamcinolone acetonide) may be slowly dissolved from a crystal structure to enhance the duration of action, or a specialized slow-release device is in development (e.g., dexamethasone and fluocinolone acetonide).^[21]

In selected cases, additional therapeutic options include pars plana vitrectomy. Despite not being a routine surgery for patients with DME, vitrectomy not only relieves traction but also raises oxygen levels and lowers VEGF levels close to the fovea.^[1]

There are several indications for vitrectomy, including coexisting epiretinal membranes and traction.^[13] It has been suggested that diffuse non-tractional DME in patients with subretinal fluid may also benefit from vitrectomy.^[23] Although some researchers have confirmed that vitrectomy results in reduced central retinal thickness and favorable anatomical outcomes, others have claimed that it has limited visual improvement. This may be due to several reasons, including the fact that surgeons typically decide to perform vitrectomy in patients with refractory DME, which has been previously reported to be associated with an increased risk of irreversible photoreceptor loss,^[1,22] but correlated with continuous and persistent VA gain due to photoreceptors integrity in a significant percentage of patients who showed inner segment/outer segment restoration coexisting with vessel density improvement on long-term evaluations.^[1] However, Michalewska et al.^[23] reported that internal limiting membrane (ILM) peeling in patients with severe diabetic complications, such as tractional retinal detachment or vitreous hemorrhage, reduces the rate of DME in the long-term. With respect to these findings, in 2018, it was reported that patients who underwent surgery for treatment-naïve DME had good functional outcomes with a gain of over one logarithm of the minimum angle of resolution (LogMAR) in 60% of the eyes.^[24]

To compare the efficacy and the role of vitrectomy versus anti-VEGF and intravitreal corticosteroid agents in treating refractory DME, we performed a systematic review and meta-analysis to quantify the efficacy and safety of these two treatments.

MATERIAL AND METHODS

RESULTS

Study selection

In this meta-analysis and systematic review, a comprehensive search was conducted across various databases and other sources (gray literature) to identify potential studies for inclusion. Initially, 883 studies were identified, of which 839 were obtained from database searches and 44 from alternative sources. Duplicates (n = 74) were removed, resulting in a reduced set of 809 unique studies. A subsequent exclusion process was applied, leading to the removal of 391 non-comparative studies and studies lacking treatment interventions for refractory DME. Consequently, a final selection of 418 articles were screened based on their titles and abstracts. Following a thorough assessment of 103 potentially relevant studies, the full texts were downloaded for further screening. Ultimately, eight studies^[27-34] were deemed eligible for inclusion in the present study. A detailed Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart depicting the study selection process is shown in [Figure 1].

Close

Figure 1:

The Preferred Reporting Items for Systematic Reviews and Meta-Analyses flow chart shows the sequential extraction and screening of the literature.

Export to PPT

Meta-analysis of efficacy

Analysis of the improvement of BCVA

In this study, we analyzed three studies^[27,32,33] that evaluated the changes in BCVA in 79 eyes. Pre-operative heterogeneity was not significant among the included studies (I² = 0%, P = 0.79), as shown in [Figure 2a], using a fixed-effects model for the meta-analysis. The mean difference (MD) between the pre-operative and post-operative BCVA values in the intravitreal anti-VEGF therapy and vitrectomy groups indicated significant heterogeneity (I² = 79%) [Figure 2b]. However, no significant difference was found between the two groups (standardized MD = −0.31, 95% CI: −0.76, 0.14, P = 0.18). A fixed-effects model analysis revealed that the improvement in BCVA was not significantly different; however, there may be some variability in the outcomes between the two groups, for which further research is necessary to determine the significance of these differences.

Close

Figure 2:

Analysis of the improvement of best-corrected visual acuity (BCVA) and comparison of pre-operative/post-operative central macular thickness (CMT). (a) Comparison of pre-operative BCVA between the two groups. (b) Comparison of post-operative BCVA between the two groups. (c) Comparison of pre-operative CMT between the two groups through a fixed-effects model. (d) Comparison of post-operative CMT between the two groups using a fixed-effects model. (e) Comparison of pre-operative CMT between the two groups using a random effects model. (f) The funnel plot shows that all the studies resided inside the funnel.

Export to PPT

CMT

Three studies^[29-31] assessed the changes in CMT before and after surgical intervention, as shown in [Figures 2c and d]. The studies revealed a consistent decrease in CMT both preoperatively (I² = 29% and MD= −58.21, 95% CI: −81.50, −34.92, P < 0.00001) and postoperatively (I² = 44% and MD= −62.77, 95% CI: −96.77, −28.78, P = 0.0003). Notably, the heterogeneity across studies was significant, likely due to variations in the treatment criteria for macular edema. The results of the random-effects model analysis did not reveal a significant difference between the effects of intravitreal corticosteroid injections and vitrectomy on the changes in CMT (MD = −49.92, 95% CI: −103.72, 3.88, P = 0.07), as shown in [Figure 2e]. However, vitrectomy was more effective than intravitreal corticosteroid injection, as evidenced by a significant decrease in CMT.

Publication bias and heterogeneity

The present study used Begg’s funnel plots to assess the symmetry of mean and risk differences in post-operative BCVA. The results showed satisfactory symmetry, as shown in [Figure 2f].

COMPLICATIONS ASSOCIATED WITH INTRAVITREAL ANTI-VEGF, INTRAVITREAL CORTICOSTEROID, AND VITRECTOMY

This study provides a detailed account of the complications arising from intravitreal anti-angiogenic (IVA) and intravitreal corticosteroid (IVC) injections as well as vitrectomy. The results are presented in [Tables 1 and 2], in which the different types of complications are described. In the IVA and IVC groups, the major complications observed were serious retinal detachment under the fovea, foveal cystoid spaces, and subretinal hemorrhage. No serious complications occurred in the vitrectomy group. However, some studies^{[28,30-32,34]} have indicated that neovascular glaucoma and the progression of lens opacity are common complications of vitrectomy.

DISCUSSION

DME is a major complication of DR that can lead to a significant vision loss.^[35-37] Despite advances in the treatment of DME, refractory DME remains difficult to treat. Vitrectomy has been suggested as the treatment option for refractory DME. However, its role remains controversial. This systematic review and meta-analysis aimed to evaluate the role of vitrectomy in the treatment of refractory DME. This study included three non-randomized studies with a total of 203 eyes. The primary outcome measures were VA and CMT measured using OCT. The results of the meta-analysis showed that vitrectomy did not significantly improve the VA in patients with refractory DME. However, there was a significant reduction in CMT after vitrectomy compared with previous treatments with IVA and IVC. The mean reduction in CMT was MD = −62.77, 95% CI: −96.77, −28.78, P = 0.0003.

Our results indicated that vitrectomy could lead to a significant reduction in CMT in patients with refractory DME. Our findings are consistent with those of previous studies,^[32] in that vitrectomy helps reduce macular thickness in patients with refractory DME; however, whether VA improves after vitrectomy remains unclear. Several long-term studies have suggested that vitrectomy is beneficial for patients with refractory DME because CMT is reduced, and VA is improved.^[36-40] In contrast, in a large prospective study from the DRCR Network, vitrectomy was beneficial in reducing CMT; however, its efficacy in improving VA was limited.^[39] However, we did not observe a significant improvement in the VA after vitrectomy. This may be because most of the included studies were non-randomized and had small sample sizes due to the severity and duration of DME, the presence of comorbidities such as DR and macular ischemia, the surgical technique, and the surgeon’s experience. Furthermore, the absence of a significant improvement in VA may also be because VA is a subjective measure influenced by factors such as cataracts, macular ischemia, and variability in the testing conditions.

One of the major concerns associated with IVA and IVC is the risk of complications.^[41-43] In our study, due to the limitations of the single report form of each article and the limited sample size, we acknowledge that a direct comparison of complications among the three variables was not conducted; rather, the complication analysis was more an estimation than a direct analysis. We found that the incidence of complications was higher in the control group than in the vitrectomy group and that the types of complications differed between the two groups. The most common complications in the vitrectomy group were neovascular glaucoma formation and intraocular pressure elevation, which is consistent with previous reports in the literature.^[44] In the IVA and IVC groups, multiple complications were reported in the included studies; however, the major complications were serious retinal detachment under the fovea, foveal cystoid spaces, and subretinal hemorrhage, consistent with previous studies.^[45-48] These findings suggest that vitrectomy may be a viable treatment option for refractory DME in cases where other treatments have failed to adequately reduce CMT. The reduction in CMT observed in this meta-analysis may have had a positive impact on long-term visual outcomes because chronic refractory macular edema is a major contributor to vision loss in DME.

Our study has several limitations that need to be considered. First, most of the included studies were non-randomized and had small sample sizes, which limited the strength of our conclusions. Second, it is important to note that the studies included in this meta-analysis had a relatively short follow-up period, ranging from 6 months to 2 years. Therefore, the long-term effects of vitrectomy on visual outcomes and the need for additional treatments require further investigation. Third, the included studies used different vitrectomy techniques, which may have influenced their outcomes. Fourth, the analysis of the complications was more an estimate rather than a direct statistical analysis of the complications as a variable.

CONCLUSION

Vitrectomy appears to be a promising treatment option for refractory DME and significantly reduces the incidence of CMT. However, its role in the improvement of VA remains unclear. Future studies with longer follow-up periods are needed to determine the optimal timing, patient selection criteria, and long-term outcomes of vitrectomy in patients with chronic refractory DME.