Introduction
Diabetic macular edema (DME) is one of the main causes of visual deterioration and vision loss in patients with diabetes [
1]. The breakdown of the blood–retinal barrier is a key event in the progression of DME. This development encompasses a wide variety of cytokines under chronic hyperglycemia, in which the hypoxia-induced release of vascular endothelial growth factor (VEGF) plays an essential role [
2]. VEGF inhibitors have revolutionized the therapeutic management of several retinal ophthalmic diseases, surpassing laser photocoagulation in terms of the ability to limit visual deterioration by increasing visual acuity in patients with DME [
3]. Ranibizumab (0.3 mg or 0.5 mg, Lucentis; Genentech, South San Francisco, CA, USA; Novartis, Basel, Switzerland) is a humanized monoclonal antibody Fab fragment against VEGF-A [
4,
5]. Aflibercept (2 mg, Eylea; Regeneron Pharmaceuticals, Tarrytown, NY, USA) is a fusion dimer with the constant region Fc of human immunoglobulin G1 (IgG1) [
6]. Both intravitreal ranibizumab (IVR) and aflibercept (IVA) have shown anatomical and visual improvement in treating DME in several large randomized clinical trials [
4,
5,
7]. Nowadays, these anti-VEGF drugs are considered optimal first-line treatment for DME [
2].
Despite its harmful effects on diabetic eyes, VEGF plays an important role in renal pathophysiology, including glomerular podocytes, endothelial cells, and tubular epithelial cells [
8]. Adverse effects on the kidneys, including proteinuria, nephrotic syndrome, hypertension, and acute kidney injury, have been noted when systemic anti-VEGF agents are used for treating cancers [
9‐
11]. As for intravitreal use, the anti-VEGF agents may also enter the systemic circulation and result in inhibition of systemic VEGF [
12,
13]. Therefore, theoretically, intravitreal anti-VEGF treatment may also result in renal function impairment. Increased proteinuria after intravitreal bevacizumab has been noted in patients with preexisting diabetic nephropathy [
14,
15]. Cases of decreased estimated glomerular filtration rate (eGFR) have also been reported after intravitreal injection of anti-VEGF including ranibizumab, aflibercept, and bevacizumab [
16‐
19]. However, few studies have focused on the longitudinal changes in renal function after intravitreal anti-VEGF therapy. Since diabetic nephropathy is often comorbid with diabetic retinopathy, the longitudinal change in eGFR after intravitreal anti-VEGF treatment for DME should be an important consideration. In this study, we investigated the longitudinal change in renal function in patients receiving IVR or IVA for DME for up to 2 years and evaluated the possible risk factors for renal function deterioration and renal replacement therapy after intravitreal anti-VEGF treatment.
Methods
Study Population
This study retrospectively enrolled patients who started treatment with IVR or IVA for DME at the National Taiwan University Hospital between January 2013 and December 2018. The inclusion criteria were as follows: (1) diabetic retinopathy documented by fundus photography or fluorescein angiography; (2) macular edema with the presence of retinal thickening, intraretinal cysts, or subretinal fluid, and central subfield thickness of > 300 μm as documented by optical coherence tomography (OCT); and (3) available record of eGFR at baseline (within 3 months before the first anti-VEGF injection) and during the 2-year follow-up period (at least two records at different time points after the first anti-VEGF injection). The exclusion criteria were as follows: (1) patients who did not receive regular follow-up visits and necessary treatments during the 2-year follow-up period, (2) patients with an eGFR ≤ 15 at baseline, (3) patients who had already received renal replacement therapy at baseline, and (4) patients who had previously received systemic or intravitreal anti-VEGF therapy. After recruitment, 108 patients were enrolled in this study, which was larger than the estimated sample size (
n = 63) for a presumed eGFR change of −5 ± 10%. All patients received three consecutive monthly IVR or IVA injections as the loading treatment, which was aligned with the experts’ consensus in Taiwan [
20], and then received treatment as needed at the physicians’ discretion after 3 months. This study adhered to the tenets of the Declaration of Helsinki, and approval was obtained from the institutional review board of the National Taiwan University Hospital (no. 202005058RINA). The need for informed consent was waived because of the retrospective nature of the study. The rule of confidentiality was followed during the study process, and no identifiable personal information is shown in this article.
Data Collection
Data were collected on age, sex, duration of diabetes, hypertension, insulin use, serum hemoglobin A1c (HbA1c), systolic blood pressure (SBP), serum creatinine at baseline and during the follow-up period, and receipt of renal replacement therapy including hemodialysis, peritoneal dialysis, hemofiltration, and renal transplantation during the 2-year follow-up period. eGFR was calculated using the equation recommended by the National Kidney Foundation in the Modification of Diet in Renal Disease Study. The baseline eGFR levels were then divided into five groups for analysis: > 120 mL/min, 91–120 mL/min, 61–90 mL/min, 31–60 mL/min, and 16–30 mL/min. The cumulative injection doses of anti-VEGF during the 2-year follow-up period were also recorded. All patients received either ranibizumab or aflibercept throughout the 2-year treatment courses. Therefore, only one type of anti-VEGF drug was used for each patient, and simultaneous bilateral injection was counted as two injections. Best-corrected visual acuity (BCVA) and central foveal thickness (CFT) in OCT were collected at baseline and during the follow-up period, and OCT characteristics including intraretinal cyst (IRC), subretinal fluid (SRF), and ellipsoid zone disruption (EZD) at baseline were recorded.
Statistical Analysis
The eGFR data at months 3, 6, 12, 18, and 24 were recorded using the last observation carried forward method due to the individualized follow-up schedule. However, if the last observation time was longer than 3 months for month 6 or longer than 6 months for months 12, 18, and 24, the data were treated as missing. BCVA was converted to the logarithm of the minimum angle of resolution (logMAR) for analysis. The correlations between BCVA or OCT characteristics and eGFR groups were analyzed using t-tests or Fisher exact tests. Changes in eGFR and other systemic factors at different time points were calculated and compared with baseline using paired t-tests. For comparisons between the different baseline eGFR groups, t-tests were performed. Multiple linear regression analysis was used to evaluate the risk factors for eGFR decline. Multiple Cox regression models were used to evaluate risk factors for dialysis after anti-VEGF treatment. A P value of < 0.05 was considered statistically significant. SAS 9.4 software (SAS Institute Inc., Cary, NC, USA) was used for all statistical analyses.
Discussion
Ranibizumab and aflibercept have been approved by the U.S. Food and Drug Administration (FDA) for many years as intravitreal drugs for the treatment of DME. According to the data from clinical trials including RISE/RIDE and VIVID/VISTA, no significant adverse events of renal diseases were noted after IVR or IVA injections [
21,
22]. A post hoc analysis from DRCR.net Protocol T revealed changes in the urinary albumin–creatinine ratio after intravitreal anti-VEGF treatment for DME [
23]. However, no corresponding data regarding long-term changes in eGFR after such treatment have been reported in clinical trials. In this study, we found that after IVR or IVA treatment for DME, the eGFR decreased during the 2-year treatment period. The decline was most obvious in patients with a baseline eGFR of > 120 mL/min or 16–30 mL/min. Regression analysis revealed that the baseline eGFR was the only significant factor for retinal replacement therapy after intravitreal anti-VEGF treatment. To our knowledge, this is the first study to reveal long-term eGFR changes after intravitreal anti-VEGF treatment for DME.
One study in diabetic patients reported an average decline in eGFR per year of 0.3%, 1.5%, and 5.7% for those without albuminuria, with microalbuminuria, and with macroalbuminuria, respectively [
24]. In the present study, the mean eGFR decline during the first year was 10.4% in the study cohort, which was much larger than the previously reported data for diabetic patients. The mean duration of diabetes in the previous study (3.4 ± 5.8 years) was much shorter than that in the present study (16.6 ± 9.0 years). Although our study results showed that the duration of diabetes was not associated with eGFR decline after intravitreal anti-VEGF, we still cannot compare the study results of these two studies directly. However, this finding implies that IVR or IVA for DME may be related to eGFR decline. VEGF is important for protection of the glomerular filtration barrier, and the blockade of VEGF signaling results in proteinuria and renal toxic events [
25]. The relationship between systemic VEGF and diabetic nephropathy has been demonstrated in previous studies, which showed that systemic VEGF inhibition causes proteinuria and the progression of diabetic nephropathy [
26]. The doses of systemic anti-VEGF used for organic cancers were much higher than those used for intravitreal injection, so the concentration of anti-VEGF detected in the blood circulation should be even higher when administered systemically than that with intravitreal administration. On the other hand, previous studies have also revealed that intravitreal anti-VEGF may be related to deterioration of renal function and proteinuria [
27]. Different types of nephropathy have been reported, including minimal change disease, thrombotic microangiopathy, focal and segmental glomerulosclerosis, acute interstitial nephritis, and the progression of chronic kidney disease [
28]. However, causal relationships remain inconclusive. In the present study, we recruited DME patients receiving IVR or IVA because these patients were also at high risk for diabetic nephropathy.
After intravitreal injections, different anti-VEGF agents may enter the systemic circulation in different ways. According to the original data from FDA, intravitreal injections of ranibizumab would lead to a serum level of ∼0.05 nmol/L [
29] and ∼0.2 nmol/L for aflibercept [
30]. Such serum levels are 200-fold lower than the minimum concentration needed to “maximally inhibit systemic VEGF” and that are thought to be safe. However, Avery et al. [
12,
13] showed that bevacizumab, ranibizumab, and aflibercept could all be detected in the systemic bloodstream after intravitreal injection, but only ranibizumab was cleared very quickly. Bevacizumab and aflibercept showed greater systemic exposure, with delayed clearance from the bloodstream. For intravitreal injections of 2 mg of aflibercept, the maximum serum level of aflibercept within 1–3 days after injection was 0.309 nmol/L, which was much higher than the IC
50 of aflibercept for systemic VEGF inhibition for several weeks (up to 30 days) [
12,
13]. As for ranibizumab, its post-injection serum level was similar to the IC
50 of ranibizumab for systemic VEGF inhibition initially, but then quickly declined to below the IC
50 within days [
12,
13]. One randomized controlled trial also confirmed that aflibercept but not ranibizumab could inhibit systemic VEGF for up to 4 weeks after intravitreal injection [
31]. However, in the present study, we found no significant differences in the effects of aflibercept and ranibizumab on changes in renal function. Although the proportion of patients receiving dialysis was higher in the aflibercept group (8.9%) than in the ranibizumab group (1.9%), the difference was not significant (
P = 0.21). Another randomized controlled trial showed that for patients receiving long-term intravitreal anti-VEGF treatment for DME, decreases in plasma-free VEGF levels were greater after treatment with aflibercept than with ranibizumab at 4 weeks, while this difference was not significant at 52 weeks [
32]. Moreover, after multiple intravitreal injections, the serum levels of ranibizumab and aflibercept were both detectable (in some cases near the IC
50) for up to 3 months, although the levels were higher in aflibercept than in ranibizumab [
12,
13,
33]. This suggests that detectable serum drug levels, in some aspects, might influence renal function for both aflibercept and ranibizumab. We recommend cautious monitoring for changes in renal function for patients with poor renal function when receiving aflibercept or ranibizumab.
Interestingly, patients with extremely high (> 120 mL/min) or low (16–30 mL/min) baseline eGFR experienced the greatest eGFR decline after intravitreal anti-VEGF treatment for DME. It is reasonable that those who already have very poor renal function are prone to further renal function decline due to anti-VEGF-related renal injury. As for those with high eGFR at baseline, some previous studies found that patients with type 2 diabetes who had high eGFR (≥ 120 mL/min/1.73 m
2) were more likely to experience a subsequent rapid decline in eGFR during follow-up [
34], which was compatible with our results. However, these patients had rather well-preserved renal function, so the effect of intravitreal anti-VEGF therapy should be less concerning. For those with poor baseline renal function, a further decline in eGFR would result in the need for renal replacement therapy. Therefore, close monitoring of renal function may be necessary if these patients require long-term intravitreal anti-VEGF treatment.
The major limitation of this study is its retrospective nature. Since eGFR was not routinely checked for patients receiving intravitreal anti-VEGF for DME, there may have been selection bias because we enrolled only patients with eGFR data before and after their initial anti-VEGF treatments. Another limitation is that only a few patients received renal replacement therapy. Also, there was no control group in this study. However, this is the first study to investigate the long-term effects of intravitreal anti-VEGF on renal function.