Aqueous Humor Analyses in Patients with Diabetic Retinopathy Who Had Undergone Panretinal Photocoagulation

Purpose To determine the associations between aqueous humor cytokine levels and the severity of diabetic retinopathy and the prior panretinal photocoagulation (PRP) status of patients with diabetic macular edema (DME). Methods We divided 98 DME patients into those with nonproliferative diabetic retinopathy (NPDR), proliferative diabetic retinopathy (PDR), and PRP patients. We compared the concentrations of interleukin- (IL-) 1β, IL-6, IL-8, IL-10, and IL-17; placental growth factor (PlGF); and vascular endothelial growth factor (VEGF) in the aqueous humors. We subclassified PRP patients by the interval between PRP and aqueous sampling and analyzed the associations between aqueous cytokine levels and this interval. Results The aqueous humor levels of IL-6, IL-8, VEGF, and PlGF were significantly higher in the PDR group than in the NPDR group. The PlGF and VEGF levels in the PDR group were significantly higher than those in the PRP group. On PRP subgroup analyses, patients who had undergone PRP within 6 months prior exhibited higher levels of VEGF, PlGF, and TNF-α than did those who had undergone PRP more than 12 months prior. The TNF-α level of the PRP subgroup treated within 6 months prior was significantly higher than that of the PDR group. Regression analyses showed that the levels of VEGF, PlGF, and TNF-α decreased significantly as the interval between PRP and aqueous sampling became longer. Conclusions PDR patients exhibited higher concentrations of VEGF and certain inflammatory cytokines than did NPDR and PRP patients. In the latter patients, the intraocular VEGF and inflammatory cytokine levels fell gradually over time.


Introduction
Diabetic macular edema (DME) is one of the most significant causes of visual disturbance in patients with diabetic retinopathy (DR) [1,2]. One meta-analysis reported that the overall prevalence of DR was 34.6%, that of proliferative DR (PDR) 6.96%, that of DME 6.81%, and that of visionthreatening DR 10.2% in patients with diabetes [3].
Of the various causes of DME, breakdown of the bloodretina barrier is characterized by a loss of pericytes and disruption of endothelial tight junctions induced by metabolic changes and inflammation [4,5]. Many cytokines and cell types affect the neurovascular unit [4].
When the crucial role of vascular endothelial growth factor (VEGF) in this context was discovered and the effectiveness of anti-VEGF therapy in DR patients estab-lished, such therapy became the first-line treatment option for DME [6,7]. Intravitreal steroids have also been widely used for several decades [8,9].
Most studies of DME patients selected treatment options depending on the responsiveness to anti-VEGF preparations; steroids were prescribed for those with VEGF-refractory or chronic DME [10]. Optical coherence tomography (OCT) usefully predicts responsiveness and prognosis [11]. The use of aqueous humor biomarkers remains controversial; however, the levels thereof may reflect the pathogenesis and condition of DME patients [12,13]. Some studies were of small sample size and enrolled very heterogeneous patients [14,15]. More work is required; aqueous humor biomarkers might predict prognosis or allow customized treatment of selected DME patients. Therefore, we enrolled a relatively large number of patients when exploring changes in aqueous humor cytokine levels, and we determined whether such levels were affected by prior laser treatment.

Methods
The study protocol adhered to all relevant tenets of the Declaration of Helsinki and was approved by the institutional review/ethics board of the Catholic University of Korea. All participants gave written informed consent for use of their clinical records.
We enrolled treatment-naïve center-involving DME (ciDME) eyes of central subfield thickness ðCSTÞ ≥ 300 μm from 2017 to 2020. Study participants were at least 18 years of age, had type II diabetes, and had received no anti-VEGF treatment or steroid treatments previously. The exclusion criteria included macular edema attributable to other causes. We also excluded eyes with any history of uveitis or intraocular surgery including cataract surgery and/or any laser treatments except panretinal photocoagulation (PRP). In patients who received PRP, we included eyes showing regression of NVE or NVD and no recurrent vitreous hemorrhage after the treatment. The PRP treatment used 20 ms pulse Pascal (Topcon Medical Laser Systems, Santa Clara, CA, USA) pattern-scanning laser photocoagulation with 3 × 3 multispot arrays and one burn width apart. Laser power was titrated (ranged from 200 to 500 mW) to apply a mild white burn with 200 μm spot size using SuperQuad 160 lens according to ETDRS guidelines [16]. Burn distribution involved no closer than 2 disc diameters temporal to the fovea and 500 μm nasal to disc and no further posterior than 1 burn within the temporal arcades. Numbers of final burns were ranged from 1800 to 2400.
We measured glycated hemoglobin levels and subjected all patients to ophthalmic examinations, including measurement of the best-corrected visual acuity (BCVA) and fundus examination. In addition, ultra-wide-field angiography was performed if necessary for DR classification. CST was measured using a Cirrus High-Definition OCT platform (Carl Zeiss Meditec, Dublin, CA, USA). Ellipsoid zone (EZ) disruptions were measured within 1,000 μm in horizontal scans

Results
We enrolled 98 treatment-naïve ciDME eyes of 98 patients of mean age 56:79 ± 9:84 years (6 males and 42 females   Table 1 by DR severity and PRP status. On post hoc analyses, the levels of IL-6, IL-8, VEGF, and PlGF in the PDR group were significantly higher than those in the NPDR group (P = 0:005, P = 0:013, P < 0:001, and P < 0:001, respectively), and the levels of VEGF and PlGF in the PDR group were higher than those in the PRP group (P < 0:001, P = 0:001) (Figure 1). We subdivided the PRP group by the interval between PRP and aqueous sampling; the systemic and ocular characteristics of the patients are summarized in Table 2. On post hoc analyses, the subgroup who had undergone PRP within 6 months prior to sampling exhibited significantly higher levels of VEGF, PlGF, and TNF-α than those who had undergone PRP more than 12 months prior (P < 0:001, P = 0:001, and P = 0:005, respectively).

Discussion
In this cross-sectional study, we found that the aqueous levels of certain proinflammatory interleukins and VEGF of PDR patients were significantly higher than those of NPDR patients and those who had undergone PRP. The TNF-α level early after PRP was significantly higher than that in the PDR group. PRP was associated with gradual decreases in the levels of aqueous VEGF and inflammatory cytokines over time. Values are expressed as the mean ± standard deviation or median and interquartile range, as appropriate. PRP: panretinal photocoagulation; HbA1c: glycated hemoglobin; ESRD: end-stage renal disease; IL: interleukin; TNF: tumor necrosis factor; VEGF: vascular endothelial growth factor; PlGF: placental growth factor; OCT: optical coherence tomography; CST: central subfield thickness; DME: diabetic macular edema; DRT: diffuse retinal thickening; CME: cystoid macular edema; EZD: ellipsoid zone disruption.

Journal of Diabetes Research
The cytokine and VEGF levels well reflected the retinal and DR status. We assumed that PDR patients might exhibit higher levels of inflammatory cytokines and VEGF and that PRP would reduce these levels. Indeed, the PDR group evidenced the highest levels, and PRP gradually reduced these levels. Thus, aqueous humor measures may be valuable biomarkers of ocular status. However, aqueous humor parameters are affected not only by ocular status but also by intraocular injection or surgery; enrollment and grouping must be well controlled [18,19]. We excluded patients with prior ocular surgery; we enrolled only treatment-naïve DME patients. We subgrouped PRP patients by the interval between PRP and aqueous sampling. We thus derived meaningful results.
PRP affects retinal and choroidal status [20,21]. Few studies have explored how PRP influences the levels of intraocular cytokines or VEGF. A previous study reported that a PRP group exhibited a higher concentration of aqueous humor matrix metalloproteinase than controls [22] and another that PRP induced the synthesis of proinflammatory cytokines [23]. We found that, compared to the PDR group, the PRP group exhibited significantly lower levels of VEGF and PlGF. However, the more recent PRP group evidenced a significantly higher TNF-α level than the earlier PDR group. TNF-α is a potent proinflammatory cytokine that plays a key role in ocular inflammation [24]. Some studies have suggested that the aqueous TNF-α level is associated with uveitis and glaucoma [25,26], but it is unclear whether this reflects DR severity [27]. An increased TNF-α level appeared to be associated with the extent of the inflammatory response after PRP, but further work is required.
Some reports found that PRP reduced the levels of intraocular VEGF and inflammatory cytokines and stabilized retinal status in DR patients [7,28]. Other reports used fundus photography or OCT to evaluate serial retinal changes after PRP [29,30]. However, no study has yet evaluated how long it might take to significantly reduce the levels of intraocular cytokines and VEGF. We found that patients who had undergone PRP more than 12 months prior to sampling exhibited PlGF, VEGF, and TNF-α levels that differed significantly from those in patients who had undergone PRP within 6 months prior to sampling. Linear regression analyses supported these findings. Thus, PRP affects intraocular status for a longer time than might be expected.

Journal of Diabetes Research
Our study has several limitations. Although we enrolled more patients than previous studies, the subgroup sizes were relatively small. Aqueous level of IL-6 in the subgroup that received PRP within 6 months prior to sampling was not the highest, unlike other inflammatory cytokines of VEGFs. Additionally, aqueous levels of other ILs were the highest in this subgroup but were not statistically significant. A larger sample size might have yielded different results. Also, this was a retrospective study; patients had been treated differently and the follow-up periods varied. In addition, various conditions in DME or PVD (posterior vitreous detachment) status could affect the aqueous profile of each patient [31]. We should have considered PVD status before planning this study using B-scans. We cannot make prognostic comments or analyze the responsiveness to DME treatments. In the future, we are planning to conduct a prospective, wellcontrolled study and animal study to compare aqueous profiles between before and after PRP treatment.
In conclusion, aqueous humor status well reflected retinal status. PDR patients exhibited higher concentrations of VEGF and certain inflammatory cytokines than did NPDR patients and those who had undergone PRP. The aqueous TNF-α level increased early after PRP. PRP was associated with gradual reductions in intraocular VEGF and inflammatory cytokine levels over time.

Data Availability
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.