Serum heparan sulfate and chondroitin sulfate concentrations in patients with newly diagnosed exfoliative glaucoma

Introduction

Glaucoma is a term describing a group of ocular disorders with multi-factorial etiology and it is a form of optic neuropathy related to elevated intraocular pressure as the main risk factor (Weinreb, Aung & Medeiros, 2014). It is the second leading cause of blindness worldwide (Quigley & Broman, 2006), with primary open-angle glaucoma (POAG) and exfoliative glaucoma (XFG) as its most prevalent types in developed countries (Prum et al., 2016; Hollo, Katsanos & Konstas, 2015).

Exfoliation syndrome (XFS) is an age-related, generalized disorder of the extracellular matrix (ECM) characterized by production and progressive accumulation of a fibrillar exfoliation material (XFM) deposits in tissues throughout the anterior segment, as well as in connective tissues comprising various visceral organs (Ritch, Schlötzer-Schrehardt & Konstas, 2003; Ritch & Schlötzer-Schrehardt, 2001). The most widely recognized disease manifestation is XFG, while glaucoma develops in only about 30% of XFS eyes in the patients’ lifetime (Ritch, 2015).

Exfoliative glaucoma is typically classified as a high-pressure type of secondary open-angle glaucoma (Puska, 2015). Compared with POAG, more adverse clinical progression and poorer prognosis is usually associated with XFG (Ritch & Schlötzer-Schrehardt, 2001).

According to the findings yielded by immunohistochemical and biochemical studies, XFM represents a highly glycosylated, cross-linked and enzymatically resistant glycoprotein/proteoglycan complex (Zenkel & Schlötzer-Schrehardt, 2014; Schlötzer-Schrehardt & Naumann, 2015). Proteoglycans consist of a core protein to which glycosaminoglycan (GAG) side chains are attached. Glycosaminoglycans are an important component of the XFM. They are linear polysaccharides comprised of glucosamine-/ galactosamine-containing repeating disaccharides (Yue, 2014). There are four GAG groups: hyaluronic acid, keratan sulfate, chondroitin/dermatan sulfate and heparan sulfate (HS), including heparin (Huflejt et al., 2014).

In both XFS and XFG, the progressively accumulating XFM successively detaches, destroys and replaces the normal ECM, e.g., the basement membranes of the cells, and ultimately results in degeneration of the cells involved (Schlötzer-Schrehardt & Naumann, 2006; Schlötzer-Schrehardt, 2012). Thus, overproduction and abnormal metabolism of GAGs has been suggested as one of the key changes in XFS (Baba, 1983; Schlötzer-Schrehardt, Dorfler & Naumann, 1992). As noted above, HS and chondroitin sulfate (CS) are two major GAG types (Lu et al., 2010; Li et al., 2017). Therefore, ability to quantify changes in blood GAG structures will advance the understanding and diagnosis of human diseases (Lu et al., 2010).

The purpose of the present study was to investigate HS and CS concentrations in serum samples of patients with newly diagnosed untreated XFG and compare the obtained values with those pertaining to newly diagnosed untreated POAG, normal controls (NC) and subjects with XFS. The second goal was to investigate the relationship between the HS and CS serum concentrations and clinical parameters in XFG, POAG and XFS groups.

Materials and Methods

Blood Sampling

After eye examinations, blood samples were collected from each subject. Samples were collected into a serum separator tube. Blood was coagulated at room temperature for two hours and was subsequently centrifuged at approximately 1,000× g for 15 min. Serum was separated into aliquots and was stored at −70 °C until required for analyses.

Results

Heparan Sulfate and Chondroitin Sulfate Concentrations

Serum human HS concentrations (mean and standard deviation) in the XFG, POAG, NC and XFS groups were 3,189.0 ± 1,473.8 ng/mL, 2,091.5 ± 940.9 ng/mL, 2,543.1 ± 1,397.3 ng/mL and 2,658.2 ± 1,426.8 ng/mL, respectively, whereby only the differences between XFG vs. POAG (p = 0.001) and XFG vs. NC (p = 0.032) were statistically significant, while that between XFG and XFS groups was not (p = 0.244), as shown in Fig. 1.

Serum CS concentrations (mean and standard deviation) in the XFG, POAG, NC and XFS group were 43.9 ± 20.7 ng/mL, 38.5 ± 22.0 ng/mL, 35.8 ± 16.4 ng/mL and 43.3 ± 21.8 ng/mL respectively. However, as shown in Fig. 2, only the difference between XFG and NC was statistically significant (p = 0.041), while the difference between XFG and POAG (p = 0.099) as well as the XFG and the XFS group was not statistically significant (p = 0.857).

When the participants in the XFM (XFG and XFS) group were compared to subjects without XFM (those assigned to the POAG and NC groups), their serum HS concentrations were 3,019.8 ± 1,469.8 ng/mL vs. 2,338.7 ± 1,227.2 ng/mL (p = 0.006) (Fig. 3) and CS concentrations were 43.7 ± 20.9 ng/mL vs. 37.1 ± 19.0 ng/mL (p = 0.026) (Fig. 4).

In Table 3 we report HS and CS concentrations segregated by gender for the XFG, POAG, XFS and NC groups.

Factors associated with XFG that, according to multivariate logistic regression analysis, distinguish it from POAG, NC and XFS are shown in Table 4.

Moreover, as shown in Table 5, the Spearman’s correlation coefficient values indicated absence of correlation between either ophthalmic clinical parameters or indices of glaucoma severity in XFG, POAG and XFS eyes at presentation and serum HS and CS concentrations.

Correlation between HbA1c and serum glucose and HS and CS concentrations, as measured by the Spearman’s correlation coefficient, is reported in Table 6. In the present study, we evaluated the clinical diagnostic value of HS and CS, in terms of their sensitivity and specificity, in patients with XFG compared to POAG patients, NC and subjects with XFS. When the XFG group was compared to NC, HS had a sensitivity of 60% and specificity of 73% at a cut-off value of 2,688.3 ng/mL. Area under the ROC curve (AUC) was 0.62 (95% CI [0.510−0.741], p = 0.032) (Fig. 5). In the comparison between XFG and POAG, HS had a sensitivity of 60% and specificity of 86% at a cut-off value of 2,642.9 ng/mL, and the AUC was 0.70 (95% CI [0.590−0.814], p = 0.001). In the comparison of XFG to all remaining groups, HS had a sensitivity of 64% and specificity of 71% at a cut-off value of 2,495.4 ng/mL, with the AUC of 0.65 (p = 0.003), and if the group with XFM was compared to the one without XFM, HS had a sensitivity of 59% and specificity of 74% at a cut-off value of 2,495.4 ng/mL, with the AUC of 0.63 (95% CI [0.538−0.721], p = 0.006).

With respect to the diagnostic value of CS, if XFG was compared to NC, its sensitivity and specificity was 60% and 62%, respectively, at a cut-off value of 3.92 ng/mL, while the AUC was 0.62 (95% CI [0.508−0.730], p = 0.041) (Fig. 6). When the group with XFM was compared to the one without XFM, CS had a sensitivity of 64% and specificity of 61% at a cut-off value of 3.62 ng/mL, with the AUC of 0.60 (95% CI [0.513−0.690], p = 0.026).

If serum HS and CS concentrations were measured simultaneously, the sensitivity and specificity values were 84% and 45%, respectively, when the XFG group was compared to NC, whereas if the group with XFM was compared to that without XFM, a sensitivity of 85% and a specificity of 44% were obtained.

Discussion

Exfoliation syndrome is an age-related disorder characterized by the pathological accumulation deposition of fibrillar material in multiple tissues, frequently associated with severe secondary open-angle glaucoma (Puska, 2015; Ritch & Schlötzer-Schrehardt, 2001).

Exfoliative glaucoma typically develops after 60 years of age and, in most cases, significant optic nerve head and visual field damage in at least one eye is already present at the time of diagnosis (Ritch & Schlötzer-Schrehardt, 2001). XFG clinical features are distinct from those characterizing POAG (Gonzalez-Iglesias et al., 2014).

In the present study, XFG patients were statistically significantly older than those diagnosed with POAG, as their mean age was 7 years greater, but they were of similar age as the XFS participants.

The VA findings yielded by our study indicate a more severe visual loss in newly diagnosed XFG eyes relative to that noted in the POAG group. Compared with the POAG group, higher mean IOP was measured in the XFG eyes that took part in the present study.

Moreover, a significantly greater severity of optic nerve damage was noted in newly diagnosed XFG compared to newly diagnosed POAG, as indicated by vC/D. In addition, the newly diagnosed XFG patients also had significantly more advanced visual field changes than those newly diagnosed with POAG.

As early diagnosis is key to preventing visual impairment in glaucoma, especially in XFG which has aggressive course, extensive research efforts have recently been dedicated to glaucoma biomarker discovery in bodily fluids that would permit disease identification.

In principle, a biomarker is an indicator of a biochemical feature or facet that can be used to diagnose or monitor the progress of a disease (Ross et al., 2005). Thus, an agent of measurable entity with high sensitivity and specificity that accurately predicts the presence, progression or absence of a disease would be an “ideal” biomarker. Hence, in XFS subjects, it is critical that a biomarker can accurately and reliably predict progression from syndrome to glaucoma, as well as allow identification of individuals at high risk of progressive damage. Given that XFS is a systemic disease, serum or plasma would be ideal bodily fluids for identification of potential biomarkers. Thus far, systemic XFM has never been diagnosed in the absence of intraocular manifestations (Ritch & Schlötzer-Schrehardt, 2001).

Human serum or plasma serve as typical clinical specimens, since both are accessible and convenient for use in clinical trials. It has been demonstrated that the low molecular weight fraction of human serum or plasma provides a rich source of potential biomarkers pertaining to diseases generated through enzymatic cleavage (Anderson et al., 2004; Adkins et al., 2002). The aim of the present study was to investigate with relatively simple method, patients’ serum samples, in order to examine potential biomarkers (GAG concentrations, CS and HS) in newly diagnosed untreated XFG and compare their levels with those pertaining to newly diagnosed untreated POAG, as well as NC and subjects with XFS. To the best of our knowledge, no prior studies on this topic have been conducted to date.

Chondroitin sulfate and HS are two distinct classes of GAGs. Both are distributed on the surface of virtually all cells and throughout most extracellular matrices. Consequently, CS and HS are the major GAGs in the blood, whereby other serum GAGs include keratin sulfate and hyaluronan (Lu et al., 2010). GAGs such as CS and HS in human serum or plasma carry important biological information and their variations in human serum or plasma are vital for investigating and monitoring certain disease conditions, such as cancer, rheumatoid arthritis and DM (Li et al., 2017; Pothacharoen et al., 2006; Lamari et al., 2006; Komosińska-Vassev et al., 2005). Therefore, evaluating GAG presence and quantity variations has a great potential for disease diagnosis and prognosis (Wei et al., 2011). Nevertheless, no standardized methods exist for serum/plasma GAG isolation and quantification (Lu et al., 2010). The approach based on a sandwich ELISA adopted in the present study was easy to perform, and it yielded reliable and reproducible quantitative results.

In the current study, HS serum concentration in newly diagnosed XFG patients was significantly higher relative to that in newly diagnosed POAG patients and NC. While elevated concentrations were also noted for the XFG group relative to the subjects with XFS, the difference was not statistically significant. Similarly, serum CS concentrations in newly diagnosed XFG patients were elevated compared to the NC, as well as POAG and XFS subjects; however, only the XFG vs. NC difference was statistically significant.

As already mentioned, XFS is a generalized disorder of the ECM characterized by production and progressive accumulation of XFM deposits in tissues of the anterior segment, as well a systemic disorder. These findings suggest that ocular XFS is part of a general disorder of the ECM and that patients with XFM may suffer from increased comorbidity. As XFS/XFG results in excessive pathological ECM deposition, this increases GAG, as well as heparan and CS levels.

When the CS values were examined with respect to gender, they were higher in women relative to men in the XFG, POAG and NC groups, but the difference was statistically significant in the POAG group only. Exfoliation syndrome was the only group in which male subjects had negligibly higher CS values. On the other hand, HC concentration values were higher among men in XFG, NC and XFS groups, but the difference relative to women was not statistically significant. Slightly higher HC values were obtained for women in the POAG group.

The multivariate analysis findings revealed that, GAG concentrations (especially HS) were associated with XFG. In addition, statistically significant differences in the serum concentrations of GAGs, HS and CS were noted between XFM patients (XFG and XFS group) and subjects without XFM (those assigned to the POAG and NC groups).

The total amount of GAGs in serum was previously found to be closely related to diabetic pathology, advanced diabetic complications in particular (Komosińska-Vassev et al., 2005). However, in our study, presence of DM did not influence the results, and the differences among groups were not statistically significant. Moreover, in line with the results reported by Shingleton, Heltzer & O’Donoghue (2003), DM was slightly less frequent in the exfoliation group relative to the non-exfoliation group. Similarly, Musch et al. (2012) found diabetes significantly more frequent among patients with POAG (18.2%) vs. patients with XFG (3.5%). In an earlier study, Konstas et al. (1998) noted a lower prevalence of diabetes in patients with XFG requiring surgery than in those with POAG. Thus, in multivariate logistic regression analyses, we adjusted for presence of DM. On the other hand, blood glucose and HbA1c values were not statistically significantly different across groups. No link between blood glucose and HbA1c and HS and CS values was noted in the XFG group, while in the NC group HS and blood glucose were in the moderately positive statistically significant correlation, whereas in the XFS group HbA1c was in a negative moderate correlation with CS.

Identification of potential XFG biomarkers from bodily fluids has been the subject of extensive research. For example, McNally & O’Brien (2014) highlighted some important findings yielded by employing metabolomics and proteomics strategies in the search for possible biomarkers of XFG from anterior lens capsule, blood or aqueous humor in exfoliation glaucoma patients. Similarly, Kamel, Bourke & O’Brien (2018) conducted a literature review in order to present an up-to-date list of clinical and laboratory-based biomarkers relating to XFS and XFG.

In prior studies in this field, analyses were time consuming and complex in most cases, involving analysis possible biomarkers from the anterior lens capsules, TM and the aqueous humor. Ghanem, Arafa & El-Baz (2011) analyzed the aqueous humor composition in XFS/XFG patients because all ocular tissues involved are bathed by the aqueous humor and should therefore be influenced by the factors contained therein. However, the practical utility of these findings is limited, as aqueous humor cannot be easily obtained in routine clinical practice.

Gonzalez-Iglesias et al. (2014) conducted a comparative proteomic study involving serum of patients with POAG and XFG to identify a candidate panel of glaucoma biomarkers for the clinical prediction, prognosis, diagnosis and monitoring of POAG and XFG cases.

The novel contribution of the present study stems from attempting to establish a link between HS and CS serum concentrations with ophthalmic clinical parameters in XFG, XFS and POAG eyes, which was not supported by the research findings. Moreover, no correlation was noted between HS and CS serum concentrations and glaucomatous damage in newly diagnosed XFG and POAG cases, expressed via glaucoma severity indices, such as vC/D, Hodapp Classification and MD.

Our aim was to identify a biomarker that would be specific to patients with XFG. Our results as well as findings reported by other authors indicate that, in XFG, significant optic nerve head and visual field damage is already present at the time of diagnosis (Ritch & Schlötzer-Schrehardt, 2001; Konstas et al., 2006). Consequently, in the involved eye(s), optic nerve head damage and visual filed deterioration are frequently more severe in XFG compared to POAG patients (Ritch, 2001). Owing to these facts, we attempted to discover serum biomarkers that would prompt a patient to seek ophthalmologist consultation, if indicated by the blood analysis results, in order to obtain a timely XFG diagnosis and start appropriate therapy. This is highly relevant, given that older individuals tend to visit general practitioners more frequently than ophthalmologists. As a part of the present investigation, we evaluated the clinical diagnostic value of HS and CS and, according to our findings, we established that HS is an adequate diagnostic test for comparing XFG to NC and all other control groups, whereas CS is an adequate diagnostic test for comparing XFG with NC. Moreover, these tests, when performed separately, can adequately distinguish subjects with XFM from those without XFM. Clearly, performing both tests would be advantageous, as this would increase sensitivity. It should be noted that the results reported here were affected by a small sample size, as noted in study limitations.

When interpreting the results yielded by our investigation, it is important to note some study limitations. Specifically, as the study aim was to determine the time of the XFG and POAG occurrence, we did not select POAG patients that were age- and sex-matched to the XFG cohort. The existence of systemic diseases was established through interviews, as well as via a detailed review of medical documentation. Thus, it needs to be emphasized that systemic diseases were not clinically determined in our study sample. It is also noteworthy that we might have failed to detect some subclinical cases of XFS, which would be diagnosable by histological methods only. Moreover, DM patients were not excluded from the sample, as already explained, and we adjusted for presence of DM in multivariate logistic regression analyses. Our findings confirmed absence of statistically significant differences between groups, indicating that inclusion of DM patients should not have influenced our results. A further limitation of the study is a small sample size.

Conclusions

In conclusion, our findings revealed greater HS and CS concentrations in newly diagnosed XFG patients and XFS subjects compared to those without XFM. Heparan sulfate is an adequate diagnostic test for comparing XFG to NC and all other control groups, whereas CS is an adequate diagnostic test for comparing XFG with NC.

As this work, to the best of our knowledge, marks the first attempt to evaluate HS and CS serum concentrations in both XFG and XFS subjects, this data and ongoing research in this field will continue to improve our understanding of XFS/XFG as a systemic disorder. Implications of HS and CS in the pathophysiology of XFS and glaucoma should be studied further. Serum or plasma is easily accessible and should thus be explored as rich sources of potential disease biomarkers. Further research should aim to identify XFG biomarkers that could be utilized in routine blood analysis tests, aiding in timely disease diagnosis.

Supplemental Information