Introduction

Cerebrovascular disease is the leading cause of acquired disability and the second leading cause of death in the world1. In Qatar, patients with acute stroke are younger than other parts of the world and do not necessarily have the traditional risk factors for stroke2,3. There is a need to establish biomarkers that identify people at increased risk of stroke.

Inflammatory cytokines are elevated in acute ischemic stroke4. Elevated circulating interleukin-6 (IL-6) is associated with a poorer outcome in patients with acute ischemic stroke (AIS)5,6 and predicts recurrent stroke in patients with small vessel disease7. Furthermore, elevated levels of matrix metalloproteinase (MMP)-9 are associated with infarct growth and hemorrhagic transformation in patients with acute ischemic stroke8, whilst lower levels are associated with a better pre-stroke collateral status9. Whilst, vascular endothelial growth factor (VEGF) levels were lower in patients with TIA or AIS10, elevated VEGF levels have been associated with larger infarct volume, post stroke cognitive impairment11 and poorer outcome12. However, a recent systematic review and meta-analysis has shown no association between serum VEGF levels and outcomes in ischemic stroke13. Whilst elevated E-selectin levels were independently associated with poorer outcomes in patients with acute ischemic stroke14, a recent prospective analysis of the EPIC-Heidelberg study showed no association between E-selectin or P-selectin levels and the risk of incident stroke15. These observations suggest a complex temporal relationship between changing levels of circulating markers of inflammation and vascular integrity with outcomes in acute ischemic stroke.

Corneal confocal microscopy (CCM) is a non-invasive ophthalmic imaging technique that has identified corneal nerve loss in patients with diabetic neuropathy16,17,18, other peripheral neuropathies19, multiple sclerosis20,21 and dementia22. Recently, we have shown corneal nerve loss in patients with TIA23, acute ischemic stroke24,25,26,27 and recurrent stroke28. Increased plasma adhesion molecules including P-selectin and ICAM-1 predict the development of diabetic neuropathy29 and elevated IL-6 has been associated with distal neuropathy in subjects with impaired glucose tolerance30.

We have assessed the association between corneal nerve loss and circulating vascular and inflammatory markers in patients with TIA and stroke.

Results

Clinical, metabolic and corneal confocal microscopy measures

Clinical, metabolic and CCM parameters in the study participants are given in Table 1.

Table 1 Clinical, metabolic, corneal nerve and circulatory markers in study participants.
Full size table

Eighty-one patients with acute ischemic stroke and 24 patients with TIA were compared with 56 age-matched healthy controls. Age (P = 0.634, P = 1.000), BMI (P = 1.000, P = 1.000), triglycerides (P = 1.000, P = 1.000), total cholesterol (P = 1.000, P = 1.000) and LDL (P = 1.000, P = 1.000) in patients with AIS and TIA were comparable to healthy controls. HDL (P = 0.000, P = 0.009), CNFL (P = 0.000, P = 0.000), CNFD (P = 0.000, P = 0.122) and CNBD (P = 0.000, P = 0.002) were lower and systolic BP (P = 0.000, P = 0.000) and HbA1c (P = 0.000, P = 0.002) were higher in patients with AIS and TIA compared to controls with no difference between patients with AIS and TIA (Fig. 1).

Figure 1
figure 1

CCM image of the sub-basal nerve plexus in a control participant (A) and a patient with TIA (B) and an acute ischemic stroke (C).

Full size image

The NIHSS Score (P = 0.000), IL-6 (P = 0.011) and E-Selectin (P = 0.032) were higher with no difference in MMP-2 (P = 0.636), MMP-9 (P = 0.098), P-Selectin (P = 0.395) and VEGF (P = 0.831) in patients with AIS compared to TIA.

Correlation of CCM parameters with inflammatory markers

IL-6 correlated with CNFD (r = 0.230, P = 0.019) and CNFL (r = 0.218, P = 0.026) with no association between corneal nerve parameters and other circulating biomarkers (Table 2).

Table 2 Correlation between corneal nerve parameters and circulating markers.
Full size table

Multiple linear regression

Multiple regression analysis showed a positive association between CNFL (Table 3) and CNFD (Table 4) with IL-6 and a negative association with age. The CNBD distribution was skewed and was not included in the regression analysis.

Table 3 Independent factors for altered CNFL in patients with TIA and stroke.
Full size table
Table 4 Independent factors for altered CNFD in patients with TIA and stroke.
Full size table

Discussion

In the present study, we show that patients admitted with TIA or acute ischemic stroke have corneal nerve loss. These data support our previous smaller studies showing a loss of corneal nerves in patients with TIA23 and acute ischemic stroke24. Moreover, we now show that the severity of corneal nerve loss is comparable between patients with TIA and major stroke. Recently, we have shown greater corneal nerve loss in patients with recurrent stroke, despite comparable vascular risk factors suggesting that assessment of corneal nerve integrity may be a more robust measure of risk of stroke and recurrent stroke compared to the evaluation of conventional risk factors for stroke28. Recent skin biopsy studies have also shown a loss of intraepidermal nerve fibres in patients with stroke, particularly those with central post stroke pain31,32. Given that we have previously shown comparable reductions in intraepidermal nerves and corneal nerves in diabetic neuropathy33, our observations suggest that our finding of corneal nerve loss in TIA and stroke represents more global evidence of nerve loss in stroke and TIA.

Vascular factors play an integral role in the development and progression of ischemic stroke. Mounting evidence shows that circulating biomarkers are associated with incident stroke7 and poorer outcomes after stroke5,6,12. Endothelial extracellular vesicles have prothrombotic properties and we have recently shown an increase in patients with TIA and ischemic stroke34. The current study did not find an elevation in E-selectin, P-selectin or VEGF or a relationship with corneal nerve loss in patients with TIA or stroke. Although, previously, we have shown that corneal nerve fibre loss is independently associated with the presence of white matter hyperintensities in patients with ischemic stroke26. We have also shown a reduction in corneal nerves and endothelial cell density in patients with TIA and minor ischemic stroke23 and an independent association between corneal endothelial cell density, area and perimeter and acute ischemic stroke25.

In the present study, we show that patients with AIS have greater IL-6 and MMP-2 levels compared to TIA which may reflect the severity of neurodegeneration in AIS. The level of IL-6 at admission has been associated with the severity of infarct and neurological outcomes on day 2835. Furthermore, the level of IL-6, fibrinogen and white blood cells had a comparable predictive ability to admission NIHSS and infarct size on recovery from stroke at 6-months36. It is important to note that the time course for appearance of various biomarkers may vary. Thus, monocyte chemotactic protein-1 (MCP-1), matrix metalloproteinase-9 (MMP-9) and interleukin-6 (IL-6) appear within hours of a stroke, but tissue inhibitor of matrix metalloproteinase-1 (TIMP-1), C-reactive protein (CRP), and S100B appear after 12–24 h and each biomarker had differing abilities to predict 90-day outcomes of stroke37. More recently, IL-6 levels correlated with the severity of stroke at admission based on the NIHSS and mRS and also predicted recurrence of stroke38. The Reasons for Geographic and Racial Differences in Stroke (REGARDS) study enrolled 30,237 participants and showed that IL-6, but not IL-8 or IL-10 was strongly associated with risk of incident stroke over 5.4 years39.

IL-6 has been traditionally considered to be a pro-inflammatory cytokine integral to initiating the acute phase response of the immune system40,41. However, it is increasingly recognized as a multifunctional cytokine capable of eliciting both pro- and anti-inflammatory effects42. Indeed, in the present study, higher levels of IL-6 were independently associated with a higher CNFD and CNFL, despite a loss of corneal nerves in patients with TIA and stroke. In this context, IL-6 has been associated with axonal regeneration following inflammation43 and intrathecal delivery of IL-6 leads to activation of pyramidal cells in the sensory motor cortex following spinal cord injury44. In the context of peripheral nerves, IL-6 receptors are expressed on Schwann cells and following nerve transection the administration of IL-6/IL-6R fusion protein resulted in a four-fold increase in myelinated nerve fiber regrowth45. IL-6 is also intimately involved in axonal regeneration as it induces increased expression of growth associated protein-43 (GAP-43)46. IL-6 deficient mice display reduced amplitude of sensory action potentials and temperature sensitivity and impaired axonal regeneration following sciatic nerve crush injury47.

This study has limitations. CCM was not performed in patients with severe disability due to their inability to cooperate during the CCM procedure. This may have biased the outcomes as the results may have been even more pronounced in those with more severe stroke. Second, circulatory markers and neurological outcomes were only measured at one time point.

In conclusion, we show greater corneal nerve loss in patients with stroke and TIA compared to healthy controls and contrary to our initial hypothesis we show that elevated IL-6 levels were independently associated with greater corneal nerve measures. However, to establish a mechanistic link between circulating markers of inflammation and vascular integrity in relation to corneal nerve integrity and outcomes in TIA and stroke, longitudinal studies are required.

Materials and methods

One hundred and five patients-admitted with AIS (n = 24) and acute ischemic stroke (n = 81) were recruited from the stroke ward at Hamad General Hospital. The diagnosis of TIA and stroke were confirmed clinically and radiologically using AHA criteria48. Exclusion criteria included patients with a known history of ocular trauma or surgery, glaucoma, dry eye and corneal dystrophy49. Demographic (age, gender, ethnicity) and clinical (blood pressure, HbA1c, lipid profile) data were obtained from the admission patients’ electronic health records. All patients underwent assessment of the National Institutes of Health Stroke Scale (NIHSS) at presentation. This study adhered to the tenets of the declaration of Helsinki and was approved by the Institutional Review Board of Weill Cornell Medicine (15-00021) and Hamad General Hospital (15304/15). Informed, written consent was obtained from all patients/guardians before participation in the study.

Corneal confocal microscopy

All patients underwent CCM (Heidelberg Retinal Tomograph III Rostock Cornea Module; Heidelberg Engineering GmbH, Heidelberg, Germany). CCM uses a 670 nm wavelength helium neon diode laser, which is a class I laser and therefore does not pose any ocular safety hazard. A × 63 objective lens with a numeric aperture of 0.9 and a working distance, relative to the applanating cap (TomoCap; Heidelberg Engineering GmbH) of 0.0 to 3.0 mm, is used. The size of each 2-dimensional image produced is 384 × 384 pixels with a 15° × 15° field of view and 10 μm/pixel transverse optical resolutions. To perform the CCM examination, local anesthetic (0.4% benoxinate hydrochloride; Chauvin Pharmaceuticals, Chefaro, United Kingdom) was used to anesthetize both eyes, and Viscotears (Carbomer 980, 0.2%, Novartis, United Kingdom) was used as the coupling agent between the cornea and the cap. Patients were asked to fixate on an outer fixation light throughout the CCM scan and a CCD camera was used to correctly position the cap onto the cornea50. The examination took approximately 10 min for both eyes. The examiners captured images of the central sub-basal nerve plexus using the section mode. Six images per participant were selected based on our established protocol taking into account the position, depth, contrast and focus51. All CCM images were manually analyzed using validated, purpose-written software. Corneal nerve fiber density (CNFD), corneal nerve branch density (CNBD) and corneal nerve fiber length (CNFL) were analyzed using CCMetrics (M. A. Dabbah, ISBE, University of Manchester, Manchester, United Kingdom)16.

Blood collection

Following consent, within 48 h of admission, peripheral venous blood (5 mL) from patients was collected in ethylenediaminetetraacetic acid (EDTA) tubes (Vacutainers; Becton Dickinson) by using a 21-gauge needle to reduce platelet stimulation. The tubes containing the blood were immediately transferred on ice to the stroke research laboratory and centrifuged within 60 min at 1500g for 15 min at 4 °C to obtain platelet poor plasma. The Plasma was immediately snap frozen and stored at − 80 °C until biomarker quantification. MMP-9, MMP-2, IL-6, E-Selectin, P-Selectin and VEGF-A levels in the plasma were measured by enzyme-linked immunosorbent assay (ELISA) using MMP-9 Human ELISA Kit (Cat: DMP900), MMP-2 Human ELISA Kit (Cat: MMP200), IL-6 Human ELISA Kit (Cat: D6050), E-Selectin Human ELISA Kit (Cat: DSLE00), P-Selectin Human ELISA Kit (Cat: DPSE00) and VEGF Human ELISA Kit (Cat: DVE00). All the ELISA kits were procured from R&D Systems, Inc. Minneapolis MN, USA. The ELISA was performed according to the manufacturer’s instructions.

Statistical analysis

Statistical justification for the number of participants was based on a power analysis using the freeware program G*Power version 3.0.10 for α (type 1 error) of 0.05 and power (1 − type 2 error) of 0.80 using corneal nerve fibre density mean (37.12 vs 29.18) and standard deviation (8.35 and 7.16) in healthy controls and patients with stroke from our previous study24 which established that we needed a minimum of 64 patients with stroke. All statistical analyses were performed using IBM SPSS Statistics software Version 25. Normality of the data was assessed using the Shapiro–Wilk test and by visual inspection of the histogram and a normal Q–Q plot. Data are expressed as mean ± standard deviation (SD). Mann Whitney test (for non-normally distributed variables) and t-test (for normally distributed variables) were performed to find the differences between two groups, and one-way ANOVA was performed to find the differences between three groups and Bonferroni correction was performed. Pearson correlation was performed to find association between corneal nerves and circulatory biomarkers. Multiple linear regression analysis was conducted to evaluate the independent association between CNFL and CNFD and covariates.

Ethics approval

This study adhered to the tenets of the declaration of Helsinki and was approved by the Institutional Review Board of Weill Cornell Medicine (15-00021) and Hamad General Hospital (15304/15).

Consent to participate

Informed, written consent was obtained from all patients/guardians before participation in the study.

Consent for publication

Written consent was obtained from all patients/guardians for publications.