Ocular manifestations are observed in approximately 30% of systemic lupus erythematosus (SLE) patients, affecting various ocular structures. These ocular lesions, including lupus retinopathy and lupus-associated optic neuropathy, contribute to vision loss and serve as important indicators of SLE disease activity (1, 2). Early detection of preclinical retinal microvascular and microstructural alterations is crucial for preventing irreversible ocular and systemic complications and predicting prognosis (1, 3, 4).

Optical coherence tomography angiography (OCTA) is a non-invasive technology that enables high-resolution visualization of retinal vasculature, providing valuable insights into retinal vascular pathologies (5). Previous studies have shown that peripapillary retinal nerve fiber layer (RNFL) thickness and ganglion cell layer (GCL) thickness were vital parameters for glaucoma follow-up, reflecting prognosis and disease severity (6). Additionally, SLE patients exhibit thinning of the RNFL and GCL, suggesting neurodegeneration and early cognitive impairment (7).

OCTA has also revealed reductions in macular vessel density (VD) and foveal avascular zone (FAZ) area size in SLE patients without retinopathy, indicating microvasculature alterations even before ocular involvement. Furthermore, reduced retinal capillary vessel density has been observed in SLE patients with lupus nephritis (LN), suggesting potential early systemic vascular involvement (1, 8, 9). Recently, in our cross-sectional investigation (10), we utilized OCTA to image the microvasculature of the retina. Notable reductions in macular vessel density were observed, specifically within the superficial capillary plexus (SCP), among SLE patients without retinopathy compared to controls (10). These findings highlight the potential role of OCTA as a non-invasive tool for the early detection of systemic vascular changes and associated kidney impairment in lupus patients (3, 11).

It is well known that hydroxychloroquine (HCQ), a commonly used treatment for SLE, can lead to retinal toxicity by binding to melanin in retinal pigment epithelium (RPE) cells (12). This leads to disruptions in retinal cell metabolism and damage to the photoreceptors and outer nuclear layer of the retina (13, 14). Early detection of HCQ-induced retinopathy before RPE cell damage occurs is crucial for preserving vision, as the condition is irreversible and may lead to blindness.

As the number of biological treatment options continues to grow in the field of SLE therapy, there is a growing interest in investigating the effects of these medications on ocular health. Despite the growing body of research on ocular involvement in SLE, the effects of newer biological treatments, such as Belimumab, on ocular health remain largely unexplored.

Belimumab, a recombinant human IgG-1λ monoclonal antibody that inhibits B-cell activating factor, is approved for treating active autoantibody-positive SLE patients (15). It is part of the standard treatment for SLE and LN according to the European League Against Rheumatism (EULAR) 2023 guidelines (16). Recent research has demonstrated that Belimumab holds promise in the treatment of Graves’ Ophthalmopathy (GO) (17, 18). However, there is currently no research indicating its effects on the ocular vasculature of SLE patients.

The objective of this study is to conduct a detailed evaluation of ocular involvement in SLE and LN patients and investigate the influence of Belimumab on retinal microvascular density in these individuals.

SLE patients with LN experience higher morbidity and mortality rates compared to those without nephritis, leading to worse survival outcomes. Furthermore, lupus retinopathy can occur due to a vasculitic process affecting the microvasculature of the retina.The relationship between SLE and ocular vascular involvement is of particular clinical significance and requires comprehensive understanding. Previous studies found reduced retinal microvascular density in SLE patients, particularly those with renal involvement, compred to healthy controls (1, 3, 25). Recognizing the significance of ocular involvement is crucial for comprehensive management and improved outcomes in individuals with LN.

With the advent of steroids and immunosuppressive therapies, the incidence of retinal involvement in SLE patients ranges from 7-29%, which is associated with visual loss (26, 27). The main mechanisms leading to SLE retinopathy are the immune complex-mediated microangiopathic vasculopathy, secondary hypertension resulting in kidney involvement with activation of the renin-angiotensin-aldosterone system (RAAS), and the micro-thrombosis associated with either underlying endothelial injury or antiphospholipid (aPL) antibodies presence (28–30).In SLE patients, due to abnormal activity of the immune system, immune complexes may deposit in ocular tissues, which can lead to inflammation and damage to the retinal blood vessels, resulting in a decrease in retinal vascular density (28).

In SLE patients with early retinal lesions, the subjective perception of visual loss may be subtle due to the absence of prominent clinical manifestations. However, delayed recognition of visual impairment often coincides with the progression of retinal lesions, in concurrence with the advancing SLE pathology, rendering subsequent treatment interventions considerably more complex. Therefore, early detection and diagnosis of ocular involvement are crucial in SLE patients.

In recent years, there has been growing interest in the use of OCTA to detect early ocular vascular damage in systemic conditions, including arterial hypertension (31), diabetes (32), and chronic kidney disease (4). As a multisystemic disease, OCTA plays a particularly significant role in the early assessment of ocular vasculature in patients with SLE.

Wang et al. observed significant reductions in central macular thickness and in both the SCP-VD and DCP-VD of SLE patients compared to healthy controls as determined by OCTA. Additionally, no significant difference was found in retinal vascular densities between the LN and non-LN groups (33). Our research findings demonstrated that in LN patients, the loop area, superior, temporal, and whole image areas in the superficial capillaries exhibited significantly lower densities compared to the non-LN group (p <0.05), consistent with the findings of Conigliaro et al. Remarkably, we observed a significant negative correlation between vascular density and 24-hour urinary protein levels, further elucidating the association between ocular vascular density and renal impairment in patients. These patterns may reflect the cumulative burden of systemic inflammation and immunomodulation on the retinal microcirculation, supporting the utility of OCTA in capturing subclinical vascular damage.

The distinct retinal microvascular alterations in LN patients are strongly associated with shared pathophysiological mechanisms involving both ocular and renal microvasculature. Immune complex deposition induces endothelial activation and inflammation, leading to vascular damage (34). Elevated cytokines such as IL-6 and TNF-α exacerbate systemic inflammation, promoting endothelial dysfunction and microvascular remodeling (35).Furthermore, antiphospholipid antibodies contribute to thrombotic microangiopathy, resulting in vascular occlusion and ischemia (36).

Based on our findings of significantly reduced SCP-VD in LN patients, we further analyzed the associations between clinical parameters and vascular density. Taking into account the observed associations between age and CRP level changes with SCP-VD, therefore, we have considered that the reduction of SCP-VD is associated with vasculitis.

The retinal vasculature is one of the vessels that can be directly observed in SLE patients, and its performance often reflects the degree of systemic vascular damage. Similar to Conigliaro et al. and Ermurat et al. results, our study has revealed a negative correlation between the SLEDAI-2K scores and the retinal microvascular VD (1, 37).

While previous studies have reported enlargement in FAZ parameters in patients with SLE (38, 39), some other studies have failed to confirm this and reported no significant differences between SLE patients and control groups (40, 41). In our study, we didn’t observe an enlargement in the FAZ perimeter and FAZ area in LN patients. Based on our findings, we propose that the changes in vascular density in SLE are more prominent in the perifoveal area, while the parafovea and fovea are comparatively less affected.

In addition to the direct effects of SLE, drugs used in SLE therapy can potentially lead to retinal and visual impairment (42). HCQ was particularly associated with maculopathy and irreversible retinal damage (43). MTX can cause ischemic retinal complications, while cyclosporine can result in decreased visual acuity (44, 45).

To the best of our knowledge, this is the first investigation into the impact of Belimumab on retinal microvascular density in SLE patients. Our findings suggested that the use of Belimumab correlates with a decrease in vascular density.

Belimumab, as an emerging targeted therapy, is widely used in rheumatic diseases such as SLE, LN, and Sjögren’s syndrome, and has shown promising therapeutic outcomes (46–48). Therefore, it is crucial to closely monitor ocular involvement associated with the use of Belimumab to prevent severe visual impairment.

Notably, our research findings indicate a decrease in retinal microvascular vessel density in the patients treated with Belimumab, which has sparked great interest in us. Previous studies have shown promising efficacy of Belimumab in anti-neutrophil cytoplasmic antibody (ANCA)-associated vasculitis and cryoglobulinemic vasculitis (49, 50). However, our research results reveal a reduction in retinal microvascular vessel density in patients treated with Belimumab.

To investigate how Belimumab affects ocular vascular density, we compared cytokine and lymphocyte subset levels among patients before and after Belimumab treatment. We found that patients treated with Belimumab showed a significant decrease in CD19+ B cells and IL-10 levels. Previous studies showed that B-cell activating factor (BAFF) enhances angiogenesis and endothelial cell proliferation by upregulating vascular endothelial growth factor (VEGF) expression (51, 52). Furthermore, studies have indicated a correlation between CD19+ B cell numbers and VEGF expression levels. These cells are pivotal within the tumor microenvironment, influencing both VEGF expression and angiogenesis (53, 54). We hypothesize that the long-term use of Belimumab inhibits BAFF, subsequently altering B cell function and leading to a reduction in CD19+ cell counts, which ultimately affects VEGF expression and vascular density. Although prior immunological studies support this mechanistic pathway, it remains hypothetical due to the absence of direct VEGF measurements in our study. Additionally, our correlation regression analysis indicates that the use of Belimumab treatment negatively impacts DCP peripapillary VD. Our regression analysis, which adjusted for key confounders such as LN prevalence and disease duration ( Supplementary Table S6 ), showed that while these adjustments attenuated Belimumab’s effects, the consistent negative trend observed across both univariate (β = -2.29, P = 0.048) and multivariate analyses (β = -1.82, P = 0.135) suggests a potential role for Belimumab in influencing retinal vascular density. This may justify the inclusion of routine OCTA screening in clinical follow-up protocols for patients with SLE, particularly those receiving biologic agents. However, due to the limited sample size in our study, expanding the cohort is essential to validate the reliability of our findings.

Previous studies have evaluated the effect of HCQ on retinal microvascular structure and suggested its potential protective effect. However, these studies did not conduct subgroup analyses considering factors like cumulative dose and duration of use (1, 55). In contrast, Mihailovic et al. performed a study with low- and high-risk subgroups based on HCQ use duration and cumulative dose, revealing that the protective effect was only present in the low-risk group (38). Our findings indicate that patients treated with HCQ for more than 5 years showed reduced DCP-VD in the 1.5 mm circle, 2.5mm circle and loop area. These results are consistent with the result of Jelena et al (14). We hypothesize that these differences may be attributed to prolonged disease duration and long-term HCQ use. It is worth noting that we conducted further analysis and found a negative correlation between retinal vascular density and the cumulative dose of HCQ and GC in patients. Higher cumulative doses of HCQ and GC may contribute to a reduction in retinal vascular density, emphasizing the potential influence of long-term HCQ treatment on retinal vascular density in SLE patients. Additionally, longer disease duration may also negatively impact retinal vascular density. These findings underscore the importance of regular monitoring and assessment of ocular health in SLE patients, especially those receiving long-term HCQ treatment.

Our study also has several limitations. Our cross-sectional study design limits the ability to establish causal relationships between retinal vascular density changes and Belimumab treatment. While we adjusted for confounders such as LN status and disease duration, the multifactorial nature of retinal microvascular changes makes it challenging to disentangle the effects of treatment from those of disease severity. Additionally, the absence of a healthy control group for comparison limits the generalizability of our findings. The observed associations, while statistically significant, require further investigation to confirm their clinical relevance. To address the limitations of this study, we are conducting ongoing longitudinal cohort studies incorporating serial OCTA imaging to investigate the cumulative effects of Belimumab on microvascular health, aiming to validate these findings and establish their clinical relevance.

In conclusion, our study found lower parafoveal SCP-VD of SLE-LN patients without retinopathy compared to SLE patients. Notably, Belimumab treatment resulted in a reduction in peripapillary DCP-VD area. These findings enhance our understanding of treatment effects on retinal vasculature in SLE. Our results highlight the potential of OCTA for early detection of retinal vascular damage in SLE-LN patients without retinopathy. Future studies with larger cohorts can provide further insights into these observations.

These findings support the value of incorporating routine OCTA screening into clinical follow-up for SLE patients, especially those undergoing long-term biologic therapy, to detect subclinical retinal changes. Such integration may facilitate earlier therapeutic intervention and more individualized disease management.