This study marks the first utilization of SLO for fundus photography to examine posterior pole HE pre- and post-surgery. SLO employs three different laser wavelengths (λ = 488; 518; 820 nm) to represent various retinal lesions across different retinal layers. Compared to traditional fundus photography, SLO is less affected by transparent media, offering enhanced contrast and resolution, thus offering clearer images of fundus structures [11, 12]. Besides, the combined use of Confocal Scanning Laser Ophthalmoscopy (CSLO) and Spectral-Domain Optical Coherence Tomography (SD-OCT) platforms enables the capture of multimodal images, facilitating more reliable examinations of changes in macular edema and HE pre- and post-surgery. Research by Gong et al. also found that MCI surpasses traditional CFP in detecting HEs and analyzing the correlation between HEs and lipid levels in DME [13].

The results of this study indicate that 1 month postoperatively, there was an increase in the area of HE and the ratio of HE area to optic disc area (HE/PD), while these changes were not statistically significant. However, a significant reduction in both HE and HE/PD was observed 6 months after treatment. To enhance the accuracy of HE area calculations, the HE/PD ratio was incorporated into this study to reduce discrepancies caused by factors such as axial length and refractive errors, which can magnify retinal images. It was also noted that one month postoperatively, there was an increase in both HE area and HE/PD compared to preoperative values, but these changes were not statistically significant. By 6 months postoperatively, both HE area and HE/PD had significantly decreased. One month postoperatively, the increase in HE area was attributed to a specific inflammatory response triggered by lipid exudation following the absorption of macular fluid. This inflammation is critical in the formation of DME and HE. During the development of DME, inflammatory markers such as CAM-1, VEGF, IL-1β, IL-6, and IL-8 are increased, disrupting the blood-retinal barrier and exacerbating leakage [14]. Lipoproteins leak from highly permeable capillaries into the extracellular space of the retina and accumulate extensively in the posterior pole, leading to the formation of HE. Histopathologically, HE is characterized by the accumulation of lipid-laden macrophages around microvascular aneurysms, primarily located in the outer plexiform layer (OPL) and the inner nuclear layer (INL) adjacent to the foveal depression. Due to the avascular nature of the outer plexiform layer, HE is challenging to absorb without treatment [15]. In the RIDE and RISE Phase III clinical trials, HE area increased in 3 months postoperatively, then decreased back to baseline levels, with HE areas at baseline and 3 months postoperatively being similar [16]. However, in this study, HE area began to increase one month postoperatively and significantly decreased by the 3 month, demonstrating that dexamethasone implants are more effective in absorbing HE in diabetic retinopathy than anti-VEGF medications. Dexamethasone implants, utilized to treat DME with HE, are administered as a second injection at the third month and can continuously release in the eye for three to 6 months. They reduce macular edema by inhibiting various inflammatory factors, reducing fibrin deposition, capillary leakage, and inflammatory cell migration [17]. Besides, they suppress macrophages in HE, prevent extracellular matrix remodeling, and induce macrophage differentiation, leading to a significant reduction in HE 6 months postoperatively. In addition, this study found that the area of HE located more than 1 mm from the foveal depression decreased more significantly after 6 months compared to HE in 1 mm of the foveal depression. This observation is attributed to non-proliferative HE primarily arising from leakage due to microvascular abnormalities, primarily distributed on the temporal side of the macula. As leakage reduces, HE begins to decrease in areas with higher microvascular density, eventually reaching near the foveal depression [18].

This study demonstrates that 3 months after the intravitreal implantation of dexamethasone, there was a significant reduction in macular edema and an initial improvement in visual acuity. 6 months post-treatment, there was a significant decrease in HE compared to preoperative levels, accompanied by significant enhancements in visual acuity [19]. The area of HE located more than 1 mm from the foveal center was positively correlated with CMT and BCVA. The significance of HE lies in their close association with macular edema and visual acuity. As indicators of past or present macular edema, hard exudates are included in the international clinical severity scales for diabetic macular edema, representing a critical indicator. While typically associated with retinal macular edema, an increase in HE alone can also heighten the risk of visual impairment. Studies both domestically and internationally have identified non-proliferative diabetic retinopathy HE as a primary factor affecting vision, with a significant correlation between the extent of HE and visual acuity [20, 21]. Besides, this research found that patients with HE located in 1mm of the foveal center experienced less improvement in postoperative visual acuity compared to those with HE located further from the foveal center. This may be due to the reduction in CMT leading to the gradual absorption of retinal cystic spaces, interlayer hyperreflective points, and subretinal fluid, but the migration of HE towards the foveal center affects the repair of the external limiting membrane and ellipsoid zone. In Sasaki's study [22], the location of HE significantly affected visual acuity, while the total area of HE had a minimal effect, aligning with the findings of this study. It was also observed that while macular edema decreased 1 month postoperatively, there was no significant improvement in visual acuity, potentially due to the choroid being thicker beneath the fovea than in other areas. The choroid's higher colloidal osmotic pressure facilitates fluid accumulation in the macular region, and lipid substances tend to migrate towards the foveal center. Research by Khairallah M et al.[23]. Suggests that the distribution of HE in the retina might be dynamic, with HE accumulating around the edema, forming HE. HE typically gathers on either side of the edematous OPL rather than at the center, indicating that changes in regional osmotic pressure cause HE to move and deposit in structurally denser tissues. 6 months into the study, both HE and CMT were significantly reduced compared to preoperative levels, and visual acuity was significant improved. This improvement is hypothesized to result from the prolonged macular edema increasing the concentration of inflammatory cytokines and vascular endothelial growth factors in the vitreous humor. The dexamethasone implant enhances tight junction proteins to stabilize the blood-retinal barrier and utilizes arachidonic acid for its anti-inflammatory effects, inducing the differentiation of specific anti-inflammatory macrophages, thus facilitating the absorption of HE [24].

The findings of this analysis further indicate that, following treatment, all patients exhibited a trend toward improvement in multifocal electroretinogram (mf-ERG) parameters over time. Typically, ischemia and hypoxia in the retinal macular region lead to edema, severely impairing visual function. Previous studies have demonstrated that patients with HE exhibit significantly lower mf-ERG amplitudes than those observed in the normal population [25]. However, post-treatment, these amplitudes tend to increase, suggesting a restoration and enhancement of macular function, thereby indicating that mf-ERG could represent a key indicator for assessing the functionality of DME associated with HE. The fundamental pathological changes in diabetic retinopathy (DR) involve disarray in the structure and function of the retinal microcirculation, with early disruptions to the blood-retinal barrier affecting the inner nuclear layer and the outer plexiform and nerve fiber layers. Ischemia and hypoxia exacerbate these effects, progressing toward proliferative stages. The P1 wave is likely derived from the inner retinal layers, while the N1 wave may originate from the outer retinal layers [26]. The selection of the P1 wave as a research focus is predicated on its amplitude density abnormalities appearing earlier than those of the N1 wave, highlighting its sensitivity as a diagnostic indicator for early-stage DR. This study also observes that HE located in 1 mm of the central fovea may lead to poorer vision and mf-ERG outcomes, potentially due to severe macular edema which, upon resolution, results in significant HE deposition beneath the macula, causing subretinal tissue fibrosis [27]. Therefore, the reduction of macular edema and the absorption of HE are equally crucial in the treatment of DME. Dexamethasone implants not only reduce inflammation but also reduce leakage from microaneurysms (MA) and downregulate VEGF production and expression, thereby exerting a vascular inhibitory effect which may continuously facilitate the absorption of HE [28].

This study demonstrates that serum total cholesterol and low-density lipoprotein cholesterol are independent risk factors for macular hard exudates in patients with type 2 diabetes. Cholesterol can lead to the formation of hard exudates in the retina. A prospective study by YS[29] et al. demonstrated that higher blood lipid levels, especially higher levels of low-density lipoprotein cholesterol and triglycerides, are independent risk factors for hard exudates in DR. While high blood lipids have limited effect on the progression of DR and the development of proliferative DR, they significantly increase the probability of CSME and macular HE. Lipid-lowering therapy can reduce these risks and prevent vision loss. The ETDRS study confirms that increased levels of triglycerides, low-density lipoprotein, and very low-density lipoprotein increase the risk of HE and vision decline, consistent with the findings of this study [30]. Long-term observations indicate that diabetic patients with hyperlipidemia experience an increase in intraocular HE. High lipid levels or excessive fat intake may induce endothelial dysfunction through lipid peroxidation reactions, triggering local inflammatory responses that release cytokines or growth factors, activate biomolecules on the vascular wall, and lead to an increase in oxidized low-density lipoprotein [31].

In addition, this research analyzed postoperative complications associated with dexamethasone implants: Three patients presented with increased intraocular pressure post-procedure, all of which were managed through topical administration of a single antiglaucoma medication, without necessitating antiglaucoma surgery. Thirteen individuals demonstrated cataract progression, which, beyond the intraocular steroid use, may also be attributable to the advanced age of the participants and the increased susceptibility to cataract development in diabetics. Specifically, none of the patients reported drug allergies, retinal detachment, vitreous hemorrhage, endophthalmitis, or other severe ocular complications or systemic adverse effects.

This study group is not without certain limitations: (1) As a retrospective case series, this study is subject to potential sampling bias; (2) The limited sample size and short follow-up duration may not be representative of the broader population, and stratification by HE location could increase the likelihood of Type II errors due to the small sample size; (3) The study did not differentiate patients based on varying types of macular edema, an aspect that warrants exploration in future research; (4) No control groups, such as those receiving anti-VEGF therapies, were established for comparative analysis; (5) Optimal parameters for mf-ERG recording require further ongoing analysis; (6) The study did not account for the duration of DME in participant selection; (7) Lack of comparison of systemic conditions before and after surgery; changes in blood glucose and lipids may cause changes in hard exudates; (8) As a clinic-based analysis, this study may not fully represent the entire DME population. Community-based epidemiological studies are needed to corroborate the findings of this research. Future studies should involve multicenter, large-sample, randomized controlled trials to confirm the effectiveness and safety of dexamethasone implants in the treatment of HE.