- Email: [email protected]
Alterations in corneal sensitivity and nerve morphology in patients with primary Sjögren's syndrome
Experimental Eye Research 86 (2008) 879–885
Contents lists available at ScienceDirect
Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer
Alterations in corneal sensitivity and nerve morphology in patients with primary Sjo¨gren’s syndromeq Ilpo S. Tuisku a, *, Yrjo¨ T. Konttinen b, Liisa M. Konttinen c, Timo M. Tervo a a
Helsinki University Eye Hospital, Helsinki, Finland ¨rtes Medicine, Helsinki University Hospital, Helsinki, Finland Department of Medicine, Inva c ORTON Orthopedic Hospital of the Invalid Foundation, Helsinki, Finland b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 August 2007 Accepted 4 March 2008 Available online 12 March 2008
The aim of the study was to assess subjective symptoms and objective clinical signs of dry eye in relation to corneal nerve morphology and sensitivity in primary Sjo¨gren’s syndrome. Twenty eyes of 20 primary Sjo¨gren’s syndrome patients and ten eyes of 10 healthy age- and sex-matched controls were included in the study. Ocular surface disease index (OSDI) questionnaire and visual analog scales were used to assess subjective symptoms. The mechanical sensitivity of the central cornea was measured using a modified Belmonte non-contact esthesiometer followed by an analysis of corneal nerve morphology using scanning slit confocal microscopy (ConfoScan 3). OSDI symptom scores were high in primary Sjo¨gren’s syndrome patients, compared with controls. Accordingly, the mean corneal detection threshold was low in patients implicating corneal mechanical hypersensitivity (54.5 40.1 ml/min vs. 85.0 24.6 ml/min, P ¼ 0.036). However, nerve densities were similar, and no correlation was present between corneal sensitivity and nerve density. In contrast, alterations in nerve morphology were found; stromal nerves appeared thicker, and nerve growth cone-like structures were seen in 20% of patients, often associated with dendritic antigen-presenting cells. Sjo¨gren’s syndrome patients presented with corneal mechanical hypersensitivity, although corneal nerve density did not differ from controls. However, alterations in corneal nerve morphology (nerve sprouting and thickened stromal nerves) and an increased amount of antigen-presenting cells, implicating the role of inflammation, were observed. These observations offer an explanation for the corneal mechanical hypersensitivity, or even hyperalgesia often observed in these patients. We hypothesize that patients with primary Sjo¨gren’s syndrome dry eye suffer from neuropathic corneal mechanical hypersensitivity induced by ocular surface inflammation. Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: cornea innervation nerve Sjo¨gren’s syndrome in vivo confocal microscopy non-contact esthesiometry
1. Introduction Dry eye is a common disorder associated with many different diseases and leading to a variety of symptoms. In many cases, dry eye seems to be associated with inflammatory changes both in the lacrimal glands and on the ocular surface. Consequently, antiinflammatory therapies, such as topical cyclosporin A and topical steroids, have been reported to be beneficial in the treatment of dry eye (Pflugfelder, 2004). Sjo¨gren’s syndrome (SS) is an important etiology of dry eye, in particular in middle aged women. SS is a chronic, generalized autoimmune disease. Its major clinical manifestations are dry eye and dry mouth. In primary SS, changes occur without an association
q A part of the data presented in: EVER 2005, Vilamoura, Portugal, November 2005 and ESCRS Winter Refractive meeting, Monte Carlo, Monaco, February 2006. * Corresponding author. Helsinki University Eye Hospital, P.O. Box 220, FIN00029 Helsinki, Finland. Tel.: þ358 9 4717 5197; fax: þ358 9 4717 5100. E-mail address: ilpo.tuisku@hus.fi (I.S. Tuisku). 0014-4835/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.exer.2008.03.002
with any other underlying autoimmune disease (Fox et al., 2000; Fox and Stern, 2002). Patients with primary Sjo¨gren’s syndrome may also show various neurological manifestations, including peripheral and autonomic neuropathies. The trigeminal nerve (cranial nerve V) is recognized as the most commonly affected cranial nerve in primary SS (Gemignani et al., 1994; Barendregt et al., 2001). Corneal afferent sensory neurons are derived from the Gasserian ganglion and enter the eye ball via long ciliary nerves, which are branches of the nasociliary portion of the ophthalmic division of the trigeminal nerve (Rozsa and Beuerman, 1982). In addition to afferent sensory function, the trigeminal nerve also seems to have an efferent trophic function for corneal epithelial cells. Epithelial wound healing is delayed in corneas with compromised innervation (Beuerman and Schimmelpfennig, 1980; Araki et al., 1994). Ocular surface, lacrimal glands and interconnecting nerves form a functional unit, and intact corneal innervation is required for blinking and tearing reflexes. Compromised function in one part of this reflex arch results in impaired ocular health (Stern et al., 1998; Dartt, 2004).
880
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
We revealed in our previous study that corneal nerves show morphological alterations in primary SS (Tuominen et al., 2003). In particular, altered keratocytes were found and they seemed to be associated with nerve growth cone-like structures present in the subbasal nerves. We hypothesized that altered keratocytes capable of producing neurotrophins lead to nerve sprouting. These alterations might contribute to corneal hyperalgesia or even chronic pain, often observed in these patients in a clinical context. A related hypothesis of neuropathic dry eye sensations has been suggested to play a role in LASIK associated dry eye: During the LASIK surgery corneal nerves are disrupted, and subsequently regenerated during the postoperative period. Abberrant or abnormal corneal nerve regeneration may lead to altered sensation or even phantom pain (Tuisku et al., 2007; Belmonte, 2007). The goal of our study was to investigate whether corneal neuronal alterations are related to corneal sensitivity in primary SS dry eye.
the lower conjunctival sac with a micropipette. The tear film was observed under cobalt-blue-filtered light. The average interval between the last complete blink and the first appearance of randomly distributed dry spots was calculated from triplicate measurements. 2.5. Subjective dry eye symptoms Subjective symptoms were evaluated using the Ocular surface disease index (OSDI), which is a validated 12-item questionnaire to assess subjective symptoms in ocular surface disease and its impact on visual functioning (Schiffman et al., 2000). Visual analog scale (VAS) was used to evaluate global severity of ocular discomfort. The subjects were asked to evaluate severity of ocular symptoms during the past month and to place a mark with a pencil on a 100-mm scale, with the left end (0 mm) representing minimal symptoms and the right end (100 mm) maximal symptoms. 2.6. Non-contact esthesiometer
2. Patients and methods 2.1. Patients Twenty primary Sjo¨gren’s syndrome patients 54.5 7.0 years old (19 females and 1 male) were compared with ten healthy controls 49.8 5.0 years old (9 females and 1 male). Diagnosis of primary Sjo¨gren’s syndrome was made according to the American – European consensus criteria (Vitali et al., 2002). The median duration of the disease symptoms was 17 years (interquartile range (IQR): 9–28 years), but the median time interval between the first symptoms and the diagnosis was five years (IQR: 2–19 years). All patients used topical ocular lubricants for symptomatic treatment on a daily basis. The majority of patients used non-preserved ocular lubricants whereas only 2 patients used preserved ocular lubricants consistently. The average number of drops of ocular lubricants used per day was 5.0 3.8 drops per eye. The study was approved by the Ethical Review Committee of Helsinki University Eye Hospital and performed according to The Declaration of Helsinki. Informed consent was obtained from all patients and control subjects. 2.2. Ophthalmologic examination All patients underwent evaluation for uncorrected (UCVA) and best spectacle-corrected (BCVA) visual acuity, and manifest refractions, as well as Goldmann tonometry, slit-lamp examination, and dilated funduscopic examination. Corneal fluorescein staining was graded from 0 to 15 according to the scale described by Lemp (1995). 2.3. Schirmer’s test Schirmer’s test was performed using Schirmer test strips (Clement Clarke International Ltd., Harlow, UK). Two drops of oxybuprocaine hydrochloride 4 mg/ml (OftanÒ Obucain, Santen Oy, Tampere, Finland) was administered to prevent reflex tearing. The strip was positioned below the lower lid between the temporal and middle thirds, and patients kept their eyes closed for five minutes, after which the strips were removed and the length of the moistened area was measured. 2.4. Tear break up time (tBUT) To measure tear break up time (tBUT), one drop of 3 mg/ml oxybuprocaine hydrochloride 1.25 mg/ml sodium fluorescein (OftanÒ Flurecain, Santen Oy, Tampere, Finland) dye was instilled in
The mechanical and chemical detection thresholds of the central cornea were assessed with a modified Belmonte non-contact esthesiometer. It was developed by the Cooperative Research Center for Eye Research and Technology, Sydney, Australia, based on an instrument previously designed by Dr. Carlos Belmonte, which has been extensively used in studies of corneal sensation (Belmonte et al., 1999; Acosta et al., 2001). The instrument, mounted on the frame of an air tonometer, consists of a box controller and two gas tanks, which contain 100% air and 100% CO2, respectively. The tip of the esthesiometer was adjusted to a distance of 4 mm from the surface of the cornea using a focusing mechanism. Subjects were seated in front of the gas esthesiometer, with their chin placed in a cup and their forehead against a band. An audible click produced by the opening of the gas valve identified the onset of the stimulus. Stimulation consisted of a series of pulses of warmed air (constant temperature of 42 C at the tip of the probe), applied to the surface of the central cornea. After each pulse, the subject was asked to report whether the stimulus was felt, independently of the sensation evoked. The subjects were asked to blink between stimuli. The intensity of the sensation caused by the mechanical stimulation was indicated using a visual analog scale. The first series of impulses consisted of eight 2-s pulses that were applied in a random order of magnitude using 20 ml/min intervals between 20 ml/min and 160 ml/min. The following stimulus was 10 ml/min lower than the lowest stimulus that elicited a positive response in the first series of impulses. Patients knew that impulses were in a random order. The reason why random order was used, and not ascending or descending method, was that patient could not know if they should feel something or not. The lowest airflow that elicited a response, even a weak one, was recorded as the mechanical detection threshold. Chemical detection threshold was assessed by stimulating the cornea with a mixture of air with different concentrations of CO2. To prevent mechanical stimulation, flow of the air/CO2 mixture was 10 ml/min below the previously established mechanical detection threshold. The esthesiometer was not re-calibrated after the initial installation made by the manufacturer, and the nominal units provided by the manufacturer were used. 2.7. In vivo confocal microscopy Corneal morphology was evaluated using an in vivo scanning slit confocal microscope (ConfoScan 3, software version 3.4, Nidek Technologies Slr, Vigonza, Italy) equipped with an Achroplan 40 objective (Carl Zeiss Meditec AG, Jena, Germany). A topical anesthetic was instilled in the lower conjunctival fornix of both eyes before examination (oxybuprocaine hydrochloride 4 mg/ml,
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
OftanÒ Obucain, Santen Oy, Tampere, Finland). The subject was seated in front of the microscope comfortably with the aid of chin and forehead rests and asked to look straight forward without any fixation target. A drop of ophthalmic lubricant gel (2 mg/g carbomere, ViscotearsÒ, CIBA Vision Europe Ltd., Southampton, UK) was applied on the objective tip to serve as a coupling medium. The microscope is supplied with automatic alignment software, which was used to center the objective. Automated mode was applied to obtain full thickness scans from the central cornea. During scanning the instrument recorded 25 frames/s as the focal plane advanced anteriorly, 4–8 mm between frames, until 350 frames were recorded. The best focused confocal microscopic images from the central cornea were used to calculate the nerve and antigen-presenting cell (APC) densities using software supplied by the manufacturer. The field size of the microscope was 422 mm 322 mm (0.136 mm2) according to calculations using a calibration slide. 2.8. Statistical analyses Statistical comparisons of the mean values between the groups were performed using t-test or Mann–Whitney U-test for normally distributed or skewed data, respectively, utilizing the SPSS for Windows program (version 11.0, SPSS Inc., Chicago, IL, USA). Normality of the data was tested using a Shaphiro–Wilk test. Bivariate correlations were examined using Spearman’s correlation test. Values are given as mean standard deviation or median, and interquartile range (IQR) for normally distributed or skewed data, respectively. Post hoc power calculations suggested a sample size of 15–17 when a-error level was set as 5%, and b-error level as 20%. P-values less than 0.05 were considered statistically significant. 3. Results 3.1. Clinical data The Schirmer test values with anesthesia (median 2.5 mm, IQR: 1.0–4.8 mm vs. median 9.0 mm, IQR: 3.0–10.0 mm, P ¼ 0.012, Mann–Whitney U-test) and tear fluid break up time (tBUT) (median 5.0 s, IQR: 3.0–6.0 s vs. 7.0 s, IQR: 5.8–12.0 s, P ¼ 0.006, Mann– Whitney U-test) were lower in primary SS patients than in controls. Median corneal fluorescein staining score was 1.5 (IQR: 0.0–4.0) in SS patients, while none of the controls showed corneal fluorescein staining. 3.2. Subjective symptoms OSDI and VAS scores were higher in primary SS patients than in controls (Table 1), with a strong positive correlation between these two measures of ocular discomfort (Table 2). 3.3. Esthesiometry The mean corneal mechanical detection threshold was low in SS patients (54.5 40.1 ml/min vs. 85.0 24.6 ml/min, P ¼ 0.036, t-test), implicating corneal mechanical hypersensitivity in this patient group. However, no difference was found in chemical detection thresholds (29.2 24.0 ml/min vs. 28.8 14.6 ml/min, P ¼ 0.960, t-test). 3.4. Confocal microscopy The normal subbasal nerve plexus of a healthy control subject is shown in Fig. 1. The nerve density was similar in controls and in SS patients (5.9 2.2 vs. 6.1 2.5, P ¼ 0.782, t-test). Abnormal nerve growth cone-like patterns (Fig. 2) were found in 20% (4/20) of patient corneas at the level of the subbasal nerve plexus, indicating
881
neural regeneration. Stromal nerves appeared significantly thicker in primary SS patients (Figs. 3 and 5) than in controls (Fig. 4, Table 1) possibly indicating neural regeneration. One patient had an extremely thick stromal nerve measuring 43 mm in diameter (Fig. 5). This patient also showed APCs among subbasal nerves in conjunction with a very low mechanical sensory threshold (10 ml/ min), implicating severe corneal mechanical hypersensitivity. Mature antigen-presenting cells (APC) with typical branching and long dendritic extensions were observed among the subbasal plexus in the central cornea in 35% (7/20) of the patients (Fig. 6), but in only 10% (1/10) of healthy controls. In patients in whom APCs were observed (n ¼ 7), the average density of the central cornea was 174 26 cells/mm2. The number of mature APCs in the central cornea in these patients displayed a relatively strong positive and statistically significant correlation with the subjective symptom score (r ¼ 0.847, P ¼ 0.016; Fig. 7). Otherwise patients with APCs and/or nerve growth cones did not differ from other patients with respect to subjective symptoms, objective signs, or other features of nerve morphology. Correlations between subjective symptoms, objective dry eye tests, and corneal thresholds are presented in Table 2. 4. Discussion 4.1. Corneal mechanical hypersensitivity Corneal detection thresholds to mechanical air jet stimuli were significantly decreased in patients with primary Sjo¨gren’s syndrome dry eye, implicating corneal mechanical hypersensitivity. Our finding is in line with earlier observations by De Paiva and Pflugfelder (2004), who also found decreased corneal detection thresholds in dry eye measured using a similar modified Belmonte non-contact gas esthesiometer. By contrast, Bourcier et al. (2005) and Benitez del Castillo et al. (2007) reported increased corneal detection thresholds in dry eye patients, suggesting corneal hypoesthesia. Many variables can influence the results, such as the selection of patients and controls and the severity of the disease. Corneal staining scores for our SS patient sample were slightly lower than those in the study by Bourcier et al. (2005) and Benitez del Castillo et al. (2007), which may in part explain the differences in the results. In the Bourcier study (2005), only one-third of patients had a diagnosis of primary or secondary Sjo¨gren’s syndrome. We included only patients with a diagnosis of primary Sjo¨gren’s syndrome and age- and sex-matched controls. In addition, variability among non-contact gas esthesiometers is inevitable and may explain part of the discrepancy. We and De Paiva and Pflugfelder
Table 1 Characteristics of primary Sjo¨gren’s syndrome patients and controls
Number of patients Mean age (years) Corneal fluorescein staining Schirmer’s test (mm) Tear break up time (s) Mechanical detection threshold (ml/min) Chemical detection threshold (% -CO2) Nerve count (nerves/frame) Stromal nerve thickness (mm) VAS (0–100 mm) OSDI (0–100%)
pSS
Controls
P value
20 54.5 7.0 1.5 (0.0–4.0) 2.5 (1.0–4.8) 5.0 (3.0–6.0) 54.5 40.1
10 49.8 5.0 – 9.0 (3.0–10.0) 7.0 (5.8–12.0) 85.0 24.6
– 0.073y 0.002**z 0.012*z 0.006**z 0.036*y
29.2 24.0
28.8 14.6
0.960y
5.9 2.2 7.9 (6.4–13.3) 61.0 (23.0–80.5) 37.5 (20.8–60.8)
6.1 2.5 5.7 (3.7–7.3) 2.5 (0.0–8.8) 5.3 (0.0–12.5)
0.782y 0.018*z 0.000**z 0.000**z
Data is presented as mean SD, or median and (interquartile range), OSDI ¼ ocular surface disease index, VAS ¼ visual analog scale, * ¼ P < 0.05, ** ¼ P < 0.01, y ¼ t-test, z ¼ Mann–Whitney U-test.
882
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
Table 2 Correlation coefficients between objective tests and subjective symptoms (n ¼ 30)
Mechanical threshold Chemical threshold Fluorescein staining Schirmer’s test tBUT VAS OSDI
Mechanical sensitivity
Chemical sensitivity
Fluorescein staining
Schirmer’s test
tBUT
VAS
OSDI
NA 0.21 0.51** 0.35 0.33 0.41* 0.46**
0.21 NA 0.20 0.14 0.15 0.29 0.18
0.51** 0.2 NA 0.53** 0.41* 0.54** 0.56**
0.35 0.14 0.53** NA 0.31 0.51** 0.47**
0.33 0.15 0.41* 0.31 NA 0.33 0.32
0.41* 0.29 0.54** 0.51** 0.33 NA 0.91**
0.46** 0.18 0.56** 0.47** 0.32 0.91** NA
Spearmann’s correlation test, * ¼ P < 0.05 and ** ¼ P < 0.01, Mechanical threshold ¼ mechanical detection threshold, Chemical threshold ¼ chemical detection threshold, tBUT ¼ tear break up time, VAS ¼ visual analog scale, OSDI ¼ ocular surface disease index.
(2004) used a modified Belmonte esthesiometer, while Bourcier et al. (2005) and Benitez del Castillo et al. (2007) utilized the original Belmonte esthesiometer. Differences in the size of the tip, the tip distance to the cornea, and the size of the corneal area stimulated by the air jet are important factors. Moreover, therapeutical approaches concerning the treatment of dry eye obviously differ. For example, different use of topical and/or systemic anti-inflammatory pharmaceuticals during the course of the disease may affect the results. Decreased corneal sensitivity in dry eye utilizing the Cochet– Bonnet esthesiometer was observed by Xu et al. (1996) and Villani et al. (2007). Different results in these cases might be explained by different research methodologies, as the Cochet–Bonnett esthesiometer uses mechanical contact probe stimulation with some limitations in sensitivity and reproducibility. In addition, a mechanical contact probe and a non-contact air jet differently stimulate corneal afferent sensory nerve endings. De Paiva and Pflugfelder (2004) suggested that corneal hypersensitivity observed in dry eye is due to compromised ocular surface barrier function. In our study, we found a positive correlation between corneal sensitivity and increased corneal fluorescein staining, supporting this part of their hypothesis. In addition, we observed a positive correlation between corneal morphology, corneal mechanical (hyper)sensitivity, and subjective symptoms, indicating that in patients with the most severe ocular symptoms weak stimuli are noted and may even elicit painful sensations. The apparent discrepancy between mechanical and chemical sensitivity is unexplained but it might relate to the type of stimulus delivered, or abnormal responsiveness of polymodal nociceptors. The aqueous deficient dry eye might inhibit conversion of CO2 to H2CO3 thus causing a diminished local Hþ concentration on the ocular surface,
which could result in falsely high chemical detection thresholds. Alternatively, degenerating or regenerating corneal nerve fibers in Sjo¨gren’s syndrome may exhibit an abnormal responsiveness to natural stimuli.
Fig. 1. Normal subbasal nerve plexus in a control subject. The size of the image is 422 mm 322 mm.
Fig. 2. Abnormal nerve growth cone-like structure at the level of the subbasal nerve plexus of a Sjo¨gren’s syndrome patient. The size of the image is 422 mm 322 mm.
4.2. Corneal nerve density In the current study, in line with our earlier observations (Tuominen et al., 2003), corneal subbasal nerve density was similar between primary SS patients and controls, suggesting that it may not per se explain the difference seen in corneal sensation. Previously published data concerning corneal nerve density in dry eye have been conflicting. Increased subbasal nerve counts in aqueous tear deficiency dry eye patients were observed by Zhang et al. (2005), whereas decreased nerve counts were reported by Benitez del Castillo et al. (2004, 2007) and Villani et al. (2007). However, in the Benitez del Castillo studies, a statistically significant difference was found between SS group and only the control group of younger individuals, not the control group of older subjects. Although corneal nerve density is easily calculated and quantified, other properties of the corneal nerves, such as morphological alterations and inflammatory findings, seem to play a more important role than nerve density, which may be increased, decreased, or similar compared with controls. 4.3. Corneal nerve morphology Although corneal nerve density showed no difference, we observed several morphological alterations in corneal nerves. Stromal nerves appeared significantly thicker in SS patients than in controls. A similar finding was reported by Benitez del Castillo et al.
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
883
Fig. 3. Abnormally thick (15 um) stromal nerve in a 58-year-old Sjo¨gren’s syndrome patient, with a 13-year history of dry eye symptoms. The size of the image is 422 mm 322 mm.
Fig. 5. Extremely thick stromal nerve, measuring in diameter 43 mm in a 53-year-old female, with a 34-year history of dry eye symptoms. The size of the image is 422 mm 322 mm.
(2004). The implication of thickening of stromal nerves remains unclear. However, we have earlier shown using immunoelectron microscopy and morphometry that nerve fibers regenerating after an inflammatory insult are significantly thicker than normal healthy nerve fibers (Imai et al., 1997; Niissalo et al., 2002) Possibly, stromal nerves are thickened secondary to chronic ocular surface inflammation, or this finding may present a form of neural regeneration. Nerve growth cone-like structures in subbasal nerves were found in 20% of our patients (Fig. 2); this finding is in line with our previous observations (Tuominen et al., 2003). These alterations may implicate ongoing nerve sprouting, and may result from attempts of the inflammation-injured nerve fibers to regenerate. Inflammation is recognized as a critical factor in the pathogenesis of dry eye, and the upregulation of neurotrophins, e.g. nerve growth factor (NGF), during inflammation has been shown in many studies. NGF is known to induce axonal regeneration and nerve sprouting (Albers et al., 1994; Streppel et al., 2002). It is found in alphasmooth muscle actin-positive dermal myofibroblasts (Hasan et al., 2000), which are comparable with corneal altered keratocytes.
Altered keratocytes develop during the corneal wound healing process (Jester et al., 1999), in various inflammatory conditions (Petroll et al., 1998; Rosenberg et al., 2002), and in SS (Tuominen et al., 2003). Presumably, these altered keratocytes are a source of NGF in SS dry eye. Moreover, patients with dry eye present with elevated tear fluid levels of NGF, and topical anti-inflammatory treatment with prednisolone seems to diminish tear fluid NGF levels and markedly alleviate dry eye symptoms (Lee et al., 2006).
Antigen-presenting cells (APC) play a critical role in corneal immunology in health and disease (Rosenberg et al., 2000, 2002; Hamrah et al., 2002, 2003; Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). APCs have been observed in living healthy corneas by modern in vivo confocal microscopy (Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). The density of APCs declines from the limbus to the center in healthy cornea (Hamrah et al., 2002). In the corneal limbal epithelium, dendritic cells are present in almost
Fig. 4. Normal stromal nerve in a control subject. The size of the image is 422 mm 322 mm.
Fig. 6. Antigen-presenting cells in the subbasal nerve plexus in a patient with primary Sjo¨gren’s syndrome. The size of the image is 422 mm 322 mm.
4.4. Antigen-presenting cells (APC)
OSDI score (0-100 %)
884
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
50 40 30 20 10 0
0
50
100
150
200
250
Antigen presenting cell density (cells/mm2) Fig. 7. Sjo¨gren’s syndrome patients with antigen presenting cells in the central cornea showed a positive correlation between antigen-presenting cell density and subjective dry eye symptoms as measured by the Ocular surface disease Index (rho ¼ 0.847, P ¼ 0.016; Spearman’s correlation test).
every healthy subject (Mastropasqua et al., 2006), while in the central cornea only some 20–30% of healthy controls show APCs (Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). More often APCs are found in the peripheral cornea, where they show signs of having a mature phenotype with long slender dendritic processes, while immature APCs without dendrites typically predominate in the central cornea (Zhivov et al., 2005). We focused on the central cornea and observed APCs in 35% of patients and in 10% of controls. The density of the APCs in the central cornea in patients with Sjo¨gren’s syndrome was significantly higher that reported in healthy subjects (Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). The high APC density together with the mature phenotype of these cells in the central cornea (Fig. 6) suggests that these cells are actively involved in the local pathomechanism as antigen processing and presenting cells. This may contribute to the perpetuation of dry eye and even lacrimal gland adenitis, as it has recently been shown that desiccating stress exposes antigenic epitopes shared by the ocular surface and lacrimal glands (Niederkorn et al., 2006). Primary Sjo¨gren’s syndrome patients presented with corneal mechanical hypersensitivity, although their corneal nerve density did not differ from that of controls. Corneal mechanical sensitivity appeared to correlate with subjective symptoms and objective ocular signs, but not with corneal nerve density. However, alterations in corneal nerve morphology (nerve sprouting and thickened stromal nerves), and an increased number of antigen-presenting cells, implicating the role of inflammation, were observed. We suggest that these observations offer an explanation for the corneal mechanical hypersensitivity or even ocular hyperalgesia often observed in these patients. Acknowledgements Research grants form the following organizations are gratefully acknowledged: Mary and Georg C. Ehnrooth’s foundation, Finnish Eye Foundation, De Blindas va¨nner – Sokeiden ysta¨va¨t, Finska La¨karesa¨llskapet, Victoriastiftelsen, Finnish Eye and Tissue Bank Foundation, Paulo Foundation, EVO grant/Helsinki University Eye Hospital, Finland. References Acosta, M.C., Tan, M.E., Belmonte, C., Gallar, J., 2001. Sensations evoked by selective mechanical, chemical, and thermal stimulation of the conjunctiva and cornea. Invest. Ophthalmol. Vis. Sci. 42, 2063–2067. Albers, K.M., Wright, D.E., Davis, B.M., 1994. Overexpression of nerve growth factor in epidermis of transgenic mice causes hypertrophy of the peripheral nervous system. J. Neurosci. 14, 1422–1432. Araki, K., Ohashi, Y., Kinoshita, S., Hayashi, K., Kuwayama, Y., Tano, Y., 1994. Epithelial wound healing in the denervated cornea. Curr. Eye Res. 13, 203–211.
Barendregt, P.J., van den Bent, M.J., van Raaij-van den Aarssen, V.J., van den Meiracker, A.H., Vecht, C.J., van der Heijde, G.L., Markusse, H.M., 2001. Involvement of the peripheral nervous system in primary Sjo¨gren’s syndrome. Ann. Rheum. Dis 60, 876–881. Belmonte, C., 2007. Eye dryness sensations after refractive surgery: impaired tear secretion or ‘‘phantom’’ cornea? J. Refract. Surg. 23, 598–602. Belmonte, C., Acosta, M.C., Schmelz, M., Gallar, J., 1999. Measurement of corneal sensitivity to mechanical and chemical stimulation with a CO2 esthesiometer. Invest. Ophthalmol. Vis. Sci. 40, 513–519. Benitez del Castillo, J.M., Wasfy, M.A., Fernandez, C., Garcia-Sanchez, J., 2004. An in vivo confocal masked study on corneal epithelium and subbasal nerves in patients with dry eye. Invest. Ophthalmol. Vis. Sci. 45, 3030–3035. Benitez del Castillo, J.M., Acosta, M.C., Wassfi, M.A., az-Valle, D., Gegundez, J.A., Fernandez, C., Garcia-Sanchez, J., 2007. Relation between corneal innervation with confocal microscopy and corneal sensitivity with noncontact esthesiometry in patients with dry eye. Invest. Ophthalmol. Vis. Sci. 48, 173–181. Beuerman, R.W., Schimmelpfennig, B., 1980. Sensory denervation of the rabbit cornea affects epithelial properties. Exp. Neurol 69, 196–201. Bourcier, T., Acosta, M.C., Borderie, V., Borras, F., Gallar, J., Bury, T., Laroche, L., Belmonte, C., 2005. Decreased corneal sensitivity in patients with dry eye. Invest. Ophthalmol. Vis. Sci. 46, 2341–2345. Dartt, D.A., 2004. Dysfunctional neural regulation of lacrimal gland secretion and its role in the pathogenesis of dry eye syndromes. Ocul. Surf 2, 76–91. De Paiva, C.S., Pflugfelder, S.C., 2004. Corneal epitheliopathy of dry eye induces hyperesthesia to mechanical air jet stimulation. Am. J. Ophthalmol. 137, 109–115. Fox, R.I., Stern, M., 2002. Sjo¨gren’s syndrome: mechanisms of pathogenesis involve interaction of immune and neurosecretory systems. Scand. J. Rheumatol. (Suppl. (116)), 3–13. Fox, R.I., Stern, M., Michelson, P., 2000. Update in Sjo¨gren syndrome. Curr. Opin. Rheumatol 12, 391–398. Gemignani, F., Marbini, A., Pavesi, G., Di, V.S., Manganelli, P., Cenacchi, G., Mancia, D., 1994. Peripheral neuropathy associated with primary Sjo¨gren’s syndrome. J. Neurol. Neurosurg. Psychiatry 57, 983–986. Hamrah, P., Zhang, Q., Liu, Y., Dana, M.R., 2002. Novel characterization of MHC class II-negative population of resident corneal Langerhans cell-type dendritic cells. Invest. Ophthalmol. Vis. Sci. 43, 639–646. Hamrah, P., Liu, Y., Zhang, Q., Dana, M.R., 2003. Alterations in corneal stromal dendritic cell phenotype and distribution in inflammation. Arch. Ophthalmol. 121, 1132–1140. Hasan, W., Zhang, R., Liu, M., Warn, J.D., Smith, P.G., 2000. Coordinate expression of NGF and alpha-smooth muscle actin mRNA and protein in cutaneous wound tissue of developing and adult rats. Cell Tissue Res. 300, 97–109. Imai, S., Tokunaga, Y., Konttinen, Y.T., Maeda, T., Hukuda, S., Santavirta, S., 1997. Ultrastructure of the synovial sensory peptidergic fibers is distinctively altered in different phases of adjuvant induced arthritis in rats: ultramorphological characterization combined with morphometric and immunohistochemical study for substance P, calcitonin gene related peptide, and protein gene product 9.5. J. Rheumatol. 24, 2177–2187. Jester, J.V., Huang, J., Barry-Lane, P.A., Kao, W.W., Petroll, W.M., Cavanagh, H.D., 1999. Transforming growth factor(beta)-mediated corneal myofibroblast differentiation requires actin and fibronectin assembly. Invest. Ophthalmol. Vis. Sci. 40, 1959–1967. Lee, H.K., Ryu, I.H., Seo, K.Y., Hong, S., Kim, H.C., Kim, E.K., 2006. Topical 0.1% prednisolone lowers nerve growth factor expression in keratoconjunctivitis sicca patients. Ophthalmology 113, 198–205. Lemp, M.A., 1995. Report of the National Eye Institute/Industry workshop on Clinical Trials in Dry Eyes. CLAO J. 21, 221–232. Mastropasqua, L., Nubile, M., Lanzini, M., Carpineto, P., Ciancaglini, M., Pannellini, T., Di, N.M., Dua, H.S., 2006. Epithelial dendritic cell distribution in normal and inflamed human cornea: in vivo confocal microscopy study. Am. J. Ophthalmol. 142, 736–744. Niederkorn, J.Y., Stern, M.E., Pflugfelder, S.C., De Paiva, C.S., Corrales, R.M., Gao, J., Siemasko, K., 2006. Desiccating stress induces T cell-mediated Sjo¨gren’s syndrome-like lacrimal keratoconjunctivitis. J. Immunol. 176, 3950–3957. Niissalo, S., Hukkanen, M., Imai, S., Tornwall, J., Konttinen, Y.T., 2002. Neuropeptides in experimental and degenerative arthritis. Ann. N.Y. Acad. Sci. 966, 384–399. Petroll, W.M., Cavanagh, H.D., Jester, J.V., 1998. Clinical confocal microscopy. Curr. Opin. Ophthalmol 9, 59–65. Pflugfelder, S.C., 2004. Antiinflammatory therapy for dry eye. Am. J. Ophthalmol. 137, 337–342. Rosenberg, M.E., Tervo, T.M., Petroll, W.M., Vesaluoma, M.H., 2000. In vivo confocal microscopy of patients with corneal recurrent erosion syndrome or epithelial basement membrane dystrophy. Ophthalmology 107, 565–573. Rosenberg, M.E., Tervo, T.M., Muller, L.J., Moilanen, J.A., Vesaluoma, M.H., 2002. In vivo confocal microscopy after herpes keratitis. Cornea 21, 265–269. Rozsa, A.J., Beuerman, R.W., 1982. Density and organization of free nerve endings in the corneal epithelium of the rabbit. Pain 14, 105–120. Schiffman, R.M., Christianson, M.D., Jacobsen, G., Hirsch, J.D., Reis, B.L., 2000. Reliability and validity of the. Ocular Surface Disease Index. Arch. Ophthalmol. 118, 615–621. Stern, M.E., Beuerman, R.W., Fox, R.I., Gao, J., Mircheff, A.K., Pflugfelder, S.C., 1998. The pathology of dry eye: the interaction between the ocular surface and lacrimal glands. Cornea 17, 584–589. Streppel, M., Azzolin, N., Dohm, S., Guntinas-Lichius, O., Haas, C., Grothe, C., Wevers, A., Neiss, W.F., Angelov, D.N., 2002. Focal application of neutralizing
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885 antibodies to soluble neurotrophic factors reduces collateral axonal branching after peripheral nerve lesion. Eur. J. Neurosci. 15, 1327–1342. Tuisku, I.S., Lindbohm, N., Wilson, S.E., Tervo, T.M., 2007. Dry eye and corneal sensitivity after high myopic LASIK. J. Refract. Surg. 23, 338–342. Tuominen, I.S., Konttinen, Y.T., Vesaluoma, M.H., Moilanen, J.A., Helinto, M., Tervo, T. M., 2003. Corneal innervation and morphology in primary Sjo¨gren’s syndrome. Invest. Ophthalmol. Vis. Sci. 44, 2545–2549. Villani, E., Galimberti, D., Viola, F., Mapelli, C., Ratiglia, R., 2007. The cornea in Sjo¨gren’s syndrome: an in vivo confocal study. Invest. Ophthalmol. Vis. Sci. 48, 2017–2022. Vitali, C., Bombardieri, S., Jonsson, R., Moutsopoulos, H.M., Alexander, E.L., Carsons, S.E., Daniels, T.E., Fox, P.C., Fox, R.I., Kassan, S.S., Pillemer, S.R., Talal, N., Weisman, M.H., 2002. Classification criteria for Sjo¨gren’s syndrome: a revised
885
version of the European criteria proposed by the American-European Consensus Group. Ann. Rheum. Dis. 61, 554–558. Xu, K.P., Yagi, Y., Tsubota, K., 1996. Decrease in corneal sensitivity and change in tear function in dry eye. Cornea 15, 235–239. Zhang, M., Chen, J., Luo, L., Xiao, Q., Sun, M., Liu, Z., 2005. Altered corneal nerves in aqueous tear deficiency viewed by in vivo confocal microscopy. Cornea 24, 818– 824. Zhivov, A., Stave, J., Vollmar, B., Guthoff, R., 2005. In vivo confocal microscopic evaluation of Langerhans cell density and distribution in the normal human corneal epithelium. Graefes Arch. Clin. Exp. Ophthalmol. 243, 1056–1061. Zhivov, A., Stave, J., Vollmar, B., Guthoff, R., 2007. In vivo confocal microscopic evaluation of langerhans cell density and distribution in the corneal epithelium of healthy volunteers and contact lens wearers. Cornea 26, 47–54.
Contents lists available at ScienceDirect
Experimental Eye Research journal homepage: www.elsevier.com/locate/yexer
Alterations in corneal sensitivity and nerve morphology in patients with primary Sjo¨gren’s syndromeq Ilpo S. Tuisku a, *, Yrjo¨ T. Konttinen b, Liisa M. Konttinen c, Timo M. Tervo a a
Helsinki University Eye Hospital, Helsinki, Finland ¨rtes Medicine, Helsinki University Hospital, Helsinki, Finland Department of Medicine, Inva c ORTON Orthopedic Hospital of the Invalid Foundation, Helsinki, Finland b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 31 August 2007 Accepted 4 March 2008 Available online 12 March 2008
The aim of the study was to assess subjective symptoms and objective clinical signs of dry eye in relation to corneal nerve morphology and sensitivity in primary Sjo¨gren’s syndrome. Twenty eyes of 20 primary Sjo¨gren’s syndrome patients and ten eyes of 10 healthy age- and sex-matched controls were included in the study. Ocular surface disease index (OSDI) questionnaire and visual analog scales were used to assess subjective symptoms. The mechanical sensitivity of the central cornea was measured using a modified Belmonte non-contact esthesiometer followed by an analysis of corneal nerve morphology using scanning slit confocal microscopy (ConfoScan 3). OSDI symptom scores were high in primary Sjo¨gren’s syndrome patients, compared with controls. Accordingly, the mean corneal detection threshold was low in patients implicating corneal mechanical hypersensitivity (54.5 40.1 ml/min vs. 85.0 24.6 ml/min, P ¼ 0.036). However, nerve densities were similar, and no correlation was present between corneal sensitivity and nerve density. In contrast, alterations in nerve morphology were found; stromal nerves appeared thicker, and nerve growth cone-like structures were seen in 20% of patients, often associated with dendritic antigen-presenting cells. Sjo¨gren’s syndrome patients presented with corneal mechanical hypersensitivity, although corneal nerve density did not differ from controls. However, alterations in corneal nerve morphology (nerve sprouting and thickened stromal nerves) and an increased amount of antigen-presenting cells, implicating the role of inflammation, were observed. These observations offer an explanation for the corneal mechanical hypersensitivity, or even hyperalgesia often observed in these patients. We hypothesize that patients with primary Sjo¨gren’s syndrome dry eye suffer from neuropathic corneal mechanical hypersensitivity induced by ocular surface inflammation. Ó 2008 Elsevier Ltd. All rights reserved.
Keywords: cornea innervation nerve Sjo¨gren’s syndrome in vivo confocal microscopy non-contact esthesiometry
1. Introduction Dry eye is a common disorder associated with many different diseases and leading to a variety of symptoms. In many cases, dry eye seems to be associated with inflammatory changes both in the lacrimal glands and on the ocular surface. Consequently, antiinflammatory therapies, such as topical cyclosporin A and topical steroids, have been reported to be beneficial in the treatment of dry eye (Pflugfelder, 2004). Sjo¨gren’s syndrome (SS) is an important etiology of dry eye, in particular in middle aged women. SS is a chronic, generalized autoimmune disease. Its major clinical manifestations are dry eye and dry mouth. In primary SS, changes occur without an association
q A part of the data presented in: EVER 2005, Vilamoura, Portugal, November 2005 and ESCRS Winter Refractive meeting, Monte Carlo, Monaco, February 2006. * Corresponding author. Helsinki University Eye Hospital, P.O. Box 220, FIN00029 Helsinki, Finland. Tel.: þ358 9 4717 5197; fax: þ358 9 4717 5100. E-mail address: ilpo.tuisku@hus.fi (I.S. Tuisku). 0014-4835/$ – see front matter Ó 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.exer.2008.03.002
with any other underlying autoimmune disease (Fox et al., 2000; Fox and Stern, 2002). Patients with primary Sjo¨gren’s syndrome may also show various neurological manifestations, including peripheral and autonomic neuropathies. The trigeminal nerve (cranial nerve V) is recognized as the most commonly affected cranial nerve in primary SS (Gemignani et al., 1994; Barendregt et al., 2001). Corneal afferent sensory neurons are derived from the Gasserian ganglion and enter the eye ball via long ciliary nerves, which are branches of the nasociliary portion of the ophthalmic division of the trigeminal nerve (Rozsa and Beuerman, 1982). In addition to afferent sensory function, the trigeminal nerve also seems to have an efferent trophic function for corneal epithelial cells. Epithelial wound healing is delayed in corneas with compromised innervation (Beuerman and Schimmelpfennig, 1980; Araki et al., 1994). Ocular surface, lacrimal glands and interconnecting nerves form a functional unit, and intact corneal innervation is required for blinking and tearing reflexes. Compromised function in one part of this reflex arch results in impaired ocular health (Stern et al., 1998; Dartt, 2004).
880
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
We revealed in our previous study that corneal nerves show morphological alterations in primary SS (Tuominen et al., 2003). In particular, altered keratocytes were found and they seemed to be associated with nerve growth cone-like structures present in the subbasal nerves. We hypothesized that altered keratocytes capable of producing neurotrophins lead to nerve sprouting. These alterations might contribute to corneal hyperalgesia or even chronic pain, often observed in these patients in a clinical context. A related hypothesis of neuropathic dry eye sensations has been suggested to play a role in LASIK associated dry eye: During the LASIK surgery corneal nerves are disrupted, and subsequently regenerated during the postoperative period. Abberrant or abnormal corneal nerve regeneration may lead to altered sensation or even phantom pain (Tuisku et al., 2007; Belmonte, 2007). The goal of our study was to investigate whether corneal neuronal alterations are related to corneal sensitivity in primary SS dry eye.
the lower conjunctival sac with a micropipette. The tear film was observed under cobalt-blue-filtered light. The average interval between the last complete blink and the first appearance of randomly distributed dry spots was calculated from triplicate measurements. 2.5. Subjective dry eye symptoms Subjective symptoms were evaluated using the Ocular surface disease index (OSDI), which is a validated 12-item questionnaire to assess subjective symptoms in ocular surface disease and its impact on visual functioning (Schiffman et al., 2000). Visual analog scale (VAS) was used to evaluate global severity of ocular discomfort. The subjects were asked to evaluate severity of ocular symptoms during the past month and to place a mark with a pencil on a 100-mm scale, with the left end (0 mm) representing minimal symptoms and the right end (100 mm) maximal symptoms. 2.6. Non-contact esthesiometer
2. Patients and methods 2.1. Patients Twenty primary Sjo¨gren’s syndrome patients 54.5 7.0 years old (19 females and 1 male) were compared with ten healthy controls 49.8 5.0 years old (9 females and 1 male). Diagnosis of primary Sjo¨gren’s syndrome was made according to the American – European consensus criteria (Vitali et al., 2002). The median duration of the disease symptoms was 17 years (interquartile range (IQR): 9–28 years), but the median time interval between the first symptoms and the diagnosis was five years (IQR: 2–19 years). All patients used topical ocular lubricants for symptomatic treatment on a daily basis. The majority of patients used non-preserved ocular lubricants whereas only 2 patients used preserved ocular lubricants consistently. The average number of drops of ocular lubricants used per day was 5.0 3.8 drops per eye. The study was approved by the Ethical Review Committee of Helsinki University Eye Hospital and performed according to The Declaration of Helsinki. Informed consent was obtained from all patients and control subjects. 2.2. Ophthalmologic examination All patients underwent evaluation for uncorrected (UCVA) and best spectacle-corrected (BCVA) visual acuity, and manifest refractions, as well as Goldmann tonometry, slit-lamp examination, and dilated funduscopic examination. Corneal fluorescein staining was graded from 0 to 15 according to the scale described by Lemp (1995). 2.3. Schirmer’s test Schirmer’s test was performed using Schirmer test strips (Clement Clarke International Ltd., Harlow, UK). Two drops of oxybuprocaine hydrochloride 4 mg/ml (OftanÒ Obucain, Santen Oy, Tampere, Finland) was administered to prevent reflex tearing. The strip was positioned below the lower lid between the temporal and middle thirds, and patients kept their eyes closed for five minutes, after which the strips were removed and the length of the moistened area was measured. 2.4. Tear break up time (tBUT) To measure tear break up time (tBUT), one drop of 3 mg/ml oxybuprocaine hydrochloride 1.25 mg/ml sodium fluorescein (OftanÒ Flurecain, Santen Oy, Tampere, Finland) dye was instilled in
The mechanical and chemical detection thresholds of the central cornea were assessed with a modified Belmonte non-contact esthesiometer. It was developed by the Cooperative Research Center for Eye Research and Technology, Sydney, Australia, based on an instrument previously designed by Dr. Carlos Belmonte, which has been extensively used in studies of corneal sensation (Belmonte et al., 1999; Acosta et al., 2001). The instrument, mounted on the frame of an air tonometer, consists of a box controller and two gas tanks, which contain 100% air and 100% CO2, respectively. The tip of the esthesiometer was adjusted to a distance of 4 mm from the surface of the cornea using a focusing mechanism. Subjects were seated in front of the gas esthesiometer, with their chin placed in a cup and their forehead against a band. An audible click produced by the opening of the gas valve identified the onset of the stimulus. Stimulation consisted of a series of pulses of warmed air (constant temperature of 42 C at the tip of the probe), applied to the surface of the central cornea. After each pulse, the subject was asked to report whether the stimulus was felt, independently of the sensation evoked. The subjects were asked to blink between stimuli. The intensity of the sensation caused by the mechanical stimulation was indicated using a visual analog scale. The first series of impulses consisted of eight 2-s pulses that were applied in a random order of magnitude using 20 ml/min intervals between 20 ml/min and 160 ml/min. The following stimulus was 10 ml/min lower than the lowest stimulus that elicited a positive response in the first series of impulses. Patients knew that impulses were in a random order. The reason why random order was used, and not ascending or descending method, was that patient could not know if they should feel something or not. The lowest airflow that elicited a response, even a weak one, was recorded as the mechanical detection threshold. Chemical detection threshold was assessed by stimulating the cornea with a mixture of air with different concentrations of CO2. To prevent mechanical stimulation, flow of the air/CO2 mixture was 10 ml/min below the previously established mechanical detection threshold. The esthesiometer was not re-calibrated after the initial installation made by the manufacturer, and the nominal units provided by the manufacturer were used. 2.7. In vivo confocal microscopy Corneal morphology was evaluated using an in vivo scanning slit confocal microscope (ConfoScan 3, software version 3.4, Nidek Technologies Slr, Vigonza, Italy) equipped with an Achroplan 40 objective (Carl Zeiss Meditec AG, Jena, Germany). A topical anesthetic was instilled in the lower conjunctival fornix of both eyes before examination (oxybuprocaine hydrochloride 4 mg/ml,
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
OftanÒ Obucain, Santen Oy, Tampere, Finland). The subject was seated in front of the microscope comfortably with the aid of chin and forehead rests and asked to look straight forward without any fixation target. A drop of ophthalmic lubricant gel (2 mg/g carbomere, ViscotearsÒ, CIBA Vision Europe Ltd., Southampton, UK) was applied on the objective tip to serve as a coupling medium. The microscope is supplied with automatic alignment software, which was used to center the objective. Automated mode was applied to obtain full thickness scans from the central cornea. During scanning the instrument recorded 25 frames/s as the focal plane advanced anteriorly, 4–8 mm between frames, until 350 frames were recorded. The best focused confocal microscopic images from the central cornea were used to calculate the nerve and antigen-presenting cell (APC) densities using software supplied by the manufacturer. The field size of the microscope was 422 mm 322 mm (0.136 mm2) according to calculations using a calibration slide. 2.8. Statistical analyses Statistical comparisons of the mean values between the groups were performed using t-test or Mann–Whitney U-test for normally distributed or skewed data, respectively, utilizing the SPSS for Windows program (version 11.0, SPSS Inc., Chicago, IL, USA). Normality of the data was tested using a Shaphiro–Wilk test. Bivariate correlations were examined using Spearman’s correlation test. Values are given as mean standard deviation or median, and interquartile range (IQR) for normally distributed or skewed data, respectively. Post hoc power calculations suggested a sample size of 15–17 when a-error level was set as 5%, and b-error level as 20%. P-values less than 0.05 were considered statistically significant. 3. Results 3.1. Clinical data The Schirmer test values with anesthesia (median 2.5 mm, IQR: 1.0–4.8 mm vs. median 9.0 mm, IQR: 3.0–10.0 mm, P ¼ 0.012, Mann–Whitney U-test) and tear fluid break up time (tBUT) (median 5.0 s, IQR: 3.0–6.0 s vs. 7.0 s, IQR: 5.8–12.0 s, P ¼ 0.006, Mann– Whitney U-test) were lower in primary SS patients than in controls. Median corneal fluorescein staining score was 1.5 (IQR: 0.0–4.0) in SS patients, while none of the controls showed corneal fluorescein staining. 3.2. Subjective symptoms OSDI and VAS scores were higher in primary SS patients than in controls (Table 1), with a strong positive correlation between these two measures of ocular discomfort (Table 2). 3.3. Esthesiometry The mean corneal mechanical detection threshold was low in SS patients (54.5 40.1 ml/min vs. 85.0 24.6 ml/min, P ¼ 0.036, t-test), implicating corneal mechanical hypersensitivity in this patient group. However, no difference was found in chemical detection thresholds (29.2 24.0 ml/min vs. 28.8 14.6 ml/min, P ¼ 0.960, t-test). 3.4. Confocal microscopy The normal subbasal nerve plexus of a healthy control subject is shown in Fig. 1. The nerve density was similar in controls and in SS patients (5.9 2.2 vs. 6.1 2.5, P ¼ 0.782, t-test). Abnormal nerve growth cone-like patterns (Fig. 2) were found in 20% (4/20) of patient corneas at the level of the subbasal nerve plexus, indicating
881
neural regeneration. Stromal nerves appeared significantly thicker in primary SS patients (Figs. 3 and 5) than in controls (Fig. 4, Table 1) possibly indicating neural regeneration. One patient had an extremely thick stromal nerve measuring 43 mm in diameter (Fig. 5). This patient also showed APCs among subbasal nerves in conjunction with a very low mechanical sensory threshold (10 ml/ min), implicating severe corneal mechanical hypersensitivity. Mature antigen-presenting cells (APC) with typical branching and long dendritic extensions were observed among the subbasal plexus in the central cornea in 35% (7/20) of the patients (Fig. 6), but in only 10% (1/10) of healthy controls. In patients in whom APCs were observed (n ¼ 7), the average density of the central cornea was 174 26 cells/mm2. The number of mature APCs in the central cornea in these patients displayed a relatively strong positive and statistically significant correlation with the subjective symptom score (r ¼ 0.847, P ¼ 0.016; Fig. 7). Otherwise patients with APCs and/or nerve growth cones did not differ from other patients with respect to subjective symptoms, objective signs, or other features of nerve morphology. Correlations between subjective symptoms, objective dry eye tests, and corneal thresholds are presented in Table 2. 4. Discussion 4.1. Corneal mechanical hypersensitivity Corneal detection thresholds to mechanical air jet stimuli were significantly decreased in patients with primary Sjo¨gren’s syndrome dry eye, implicating corneal mechanical hypersensitivity. Our finding is in line with earlier observations by De Paiva and Pflugfelder (2004), who also found decreased corneal detection thresholds in dry eye measured using a similar modified Belmonte non-contact gas esthesiometer. By contrast, Bourcier et al. (2005) and Benitez del Castillo et al. (2007) reported increased corneal detection thresholds in dry eye patients, suggesting corneal hypoesthesia. Many variables can influence the results, such as the selection of patients and controls and the severity of the disease. Corneal staining scores for our SS patient sample were slightly lower than those in the study by Bourcier et al. (2005) and Benitez del Castillo et al. (2007), which may in part explain the differences in the results. In the Bourcier study (2005), only one-third of patients had a diagnosis of primary or secondary Sjo¨gren’s syndrome. We included only patients with a diagnosis of primary Sjo¨gren’s syndrome and age- and sex-matched controls. In addition, variability among non-contact gas esthesiometers is inevitable and may explain part of the discrepancy. We and De Paiva and Pflugfelder
Table 1 Characteristics of primary Sjo¨gren’s syndrome patients and controls
Number of patients Mean age (years) Corneal fluorescein staining Schirmer’s test (mm) Tear break up time (s) Mechanical detection threshold (ml/min) Chemical detection threshold (% -CO2) Nerve count (nerves/frame) Stromal nerve thickness (mm) VAS (0–100 mm) OSDI (0–100%)
pSS
Controls
P value
20 54.5 7.0 1.5 (0.0–4.0) 2.5 (1.0–4.8) 5.0 (3.0–6.0) 54.5 40.1
10 49.8 5.0 – 9.0 (3.0–10.0) 7.0 (5.8–12.0) 85.0 24.6
– 0.073y 0.002**z 0.012*z 0.006**z 0.036*y
29.2 24.0
28.8 14.6
0.960y
5.9 2.2 7.9 (6.4–13.3) 61.0 (23.0–80.5) 37.5 (20.8–60.8)
6.1 2.5 5.7 (3.7–7.3) 2.5 (0.0–8.8) 5.3 (0.0–12.5)
0.782y 0.018*z 0.000**z 0.000**z
Data is presented as mean SD, or median and (interquartile range), OSDI ¼ ocular surface disease index, VAS ¼ visual analog scale, * ¼ P < 0.05, ** ¼ P < 0.01, y ¼ t-test, z ¼ Mann–Whitney U-test.
882
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
Table 2 Correlation coefficients between objective tests and subjective symptoms (n ¼ 30)
Mechanical threshold Chemical threshold Fluorescein staining Schirmer’s test tBUT VAS OSDI
Mechanical sensitivity
Chemical sensitivity
Fluorescein staining
Schirmer’s test
tBUT
VAS
OSDI
NA 0.21 0.51** 0.35 0.33 0.41* 0.46**
0.21 NA 0.20 0.14 0.15 0.29 0.18
0.51** 0.2 NA 0.53** 0.41* 0.54** 0.56**
0.35 0.14 0.53** NA 0.31 0.51** 0.47**
0.33 0.15 0.41* 0.31 NA 0.33 0.32
0.41* 0.29 0.54** 0.51** 0.33 NA 0.91**
0.46** 0.18 0.56** 0.47** 0.32 0.91** NA
Spearmann’s correlation test, * ¼ P < 0.05 and ** ¼ P < 0.01, Mechanical threshold ¼ mechanical detection threshold, Chemical threshold ¼ chemical detection threshold, tBUT ¼ tear break up time, VAS ¼ visual analog scale, OSDI ¼ ocular surface disease index.
(2004) used a modified Belmonte esthesiometer, while Bourcier et al. (2005) and Benitez del Castillo et al. (2007) utilized the original Belmonte esthesiometer. Differences in the size of the tip, the tip distance to the cornea, and the size of the corneal area stimulated by the air jet are important factors. Moreover, therapeutical approaches concerning the treatment of dry eye obviously differ. For example, different use of topical and/or systemic anti-inflammatory pharmaceuticals during the course of the disease may affect the results. Decreased corneal sensitivity in dry eye utilizing the Cochet– Bonnet esthesiometer was observed by Xu et al. (1996) and Villani et al. (2007). Different results in these cases might be explained by different research methodologies, as the Cochet–Bonnett esthesiometer uses mechanical contact probe stimulation with some limitations in sensitivity and reproducibility. In addition, a mechanical contact probe and a non-contact air jet differently stimulate corneal afferent sensory nerve endings. De Paiva and Pflugfelder (2004) suggested that corneal hypersensitivity observed in dry eye is due to compromised ocular surface barrier function. In our study, we found a positive correlation between corneal sensitivity and increased corneal fluorescein staining, supporting this part of their hypothesis. In addition, we observed a positive correlation between corneal morphology, corneal mechanical (hyper)sensitivity, and subjective symptoms, indicating that in patients with the most severe ocular symptoms weak stimuli are noted and may even elicit painful sensations. The apparent discrepancy between mechanical and chemical sensitivity is unexplained but it might relate to the type of stimulus delivered, or abnormal responsiveness of polymodal nociceptors. The aqueous deficient dry eye might inhibit conversion of CO2 to H2CO3 thus causing a diminished local Hþ concentration on the ocular surface,
which could result in falsely high chemical detection thresholds. Alternatively, degenerating or regenerating corneal nerve fibers in Sjo¨gren’s syndrome may exhibit an abnormal responsiveness to natural stimuli.
Fig. 1. Normal subbasal nerve plexus in a control subject. The size of the image is 422 mm 322 mm.
Fig. 2. Abnormal nerve growth cone-like structure at the level of the subbasal nerve plexus of a Sjo¨gren’s syndrome patient. The size of the image is 422 mm 322 mm.
4.2. Corneal nerve density In the current study, in line with our earlier observations (Tuominen et al., 2003), corneal subbasal nerve density was similar between primary SS patients and controls, suggesting that it may not per se explain the difference seen in corneal sensation. Previously published data concerning corneal nerve density in dry eye have been conflicting. Increased subbasal nerve counts in aqueous tear deficiency dry eye patients were observed by Zhang et al. (2005), whereas decreased nerve counts were reported by Benitez del Castillo et al. (2004, 2007) and Villani et al. (2007). However, in the Benitez del Castillo studies, a statistically significant difference was found between SS group and only the control group of younger individuals, not the control group of older subjects. Although corneal nerve density is easily calculated and quantified, other properties of the corneal nerves, such as morphological alterations and inflammatory findings, seem to play a more important role than nerve density, which may be increased, decreased, or similar compared with controls. 4.3. Corneal nerve morphology Although corneal nerve density showed no difference, we observed several morphological alterations in corneal nerves. Stromal nerves appeared significantly thicker in SS patients than in controls. A similar finding was reported by Benitez del Castillo et al.
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
883
Fig. 3. Abnormally thick (15 um) stromal nerve in a 58-year-old Sjo¨gren’s syndrome patient, with a 13-year history of dry eye symptoms. The size of the image is 422 mm 322 mm.
Fig. 5. Extremely thick stromal nerve, measuring in diameter 43 mm in a 53-year-old female, with a 34-year history of dry eye symptoms. The size of the image is 422 mm 322 mm.
(2004). The implication of thickening of stromal nerves remains unclear. However, we have earlier shown using immunoelectron microscopy and morphometry that nerve fibers regenerating after an inflammatory insult are significantly thicker than normal healthy nerve fibers (Imai et al., 1997; Niissalo et al., 2002) Possibly, stromal nerves are thickened secondary to chronic ocular surface inflammation, or this finding may present a form of neural regeneration. Nerve growth cone-like structures in subbasal nerves were found in 20% of our patients (Fig. 2); this finding is in line with our previous observations (Tuominen et al., 2003). These alterations may implicate ongoing nerve sprouting, and may result from attempts of the inflammation-injured nerve fibers to regenerate. Inflammation is recognized as a critical factor in the pathogenesis of dry eye, and the upregulation of neurotrophins, e.g. nerve growth factor (NGF), during inflammation has been shown in many studies. NGF is known to induce axonal regeneration and nerve sprouting (Albers et al., 1994; Streppel et al., 2002). It is found in alphasmooth muscle actin-positive dermal myofibroblasts (Hasan et al., 2000), which are comparable with corneal altered keratocytes.
Altered keratocytes develop during the corneal wound healing process (Jester et al., 1999), in various inflammatory conditions (Petroll et al., 1998; Rosenberg et al., 2002), and in SS (Tuominen et al., 2003). Presumably, these altered keratocytes are a source of NGF in SS dry eye. Moreover, patients with dry eye present with elevated tear fluid levels of NGF, and topical anti-inflammatory treatment with prednisolone seems to diminish tear fluid NGF levels and markedly alleviate dry eye symptoms (Lee et al., 2006).
Antigen-presenting cells (APC) play a critical role in corneal immunology in health and disease (Rosenberg et al., 2000, 2002; Hamrah et al., 2002, 2003; Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). APCs have been observed in living healthy corneas by modern in vivo confocal microscopy (Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). The density of APCs declines from the limbus to the center in healthy cornea (Hamrah et al., 2002). In the corneal limbal epithelium, dendritic cells are present in almost
Fig. 4. Normal stromal nerve in a control subject. The size of the image is 422 mm 322 mm.
Fig. 6. Antigen-presenting cells in the subbasal nerve plexus in a patient with primary Sjo¨gren’s syndrome. The size of the image is 422 mm 322 mm.
4.4. Antigen-presenting cells (APC)
OSDI score (0-100 %)
884
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885
50 40 30 20 10 0
0
50
100
150
200
250
Antigen presenting cell density (cells/mm2) Fig. 7. Sjo¨gren’s syndrome patients with antigen presenting cells in the central cornea showed a positive correlation between antigen-presenting cell density and subjective dry eye symptoms as measured by the Ocular surface disease Index (rho ¼ 0.847, P ¼ 0.016; Spearman’s correlation test).
every healthy subject (Mastropasqua et al., 2006), while in the central cornea only some 20–30% of healthy controls show APCs (Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). More often APCs are found in the peripheral cornea, where they show signs of having a mature phenotype with long slender dendritic processes, while immature APCs without dendrites typically predominate in the central cornea (Zhivov et al., 2005). We focused on the central cornea and observed APCs in 35% of patients and in 10% of controls. The density of the APCs in the central cornea in patients with Sjo¨gren’s syndrome was significantly higher that reported in healthy subjects (Zhivov et al., 2005, 2007; Mastropasqua et al., 2006). The high APC density together with the mature phenotype of these cells in the central cornea (Fig. 6) suggests that these cells are actively involved in the local pathomechanism as antigen processing and presenting cells. This may contribute to the perpetuation of dry eye and even lacrimal gland adenitis, as it has recently been shown that desiccating stress exposes antigenic epitopes shared by the ocular surface and lacrimal glands (Niederkorn et al., 2006). Primary Sjo¨gren’s syndrome patients presented with corneal mechanical hypersensitivity, although their corneal nerve density did not differ from that of controls. Corneal mechanical sensitivity appeared to correlate with subjective symptoms and objective ocular signs, but not with corneal nerve density. However, alterations in corneal nerve morphology (nerve sprouting and thickened stromal nerves), and an increased number of antigen-presenting cells, implicating the role of inflammation, were observed. We suggest that these observations offer an explanation for the corneal mechanical hypersensitivity or even ocular hyperalgesia often observed in these patients. Acknowledgements Research grants form the following organizations are gratefully acknowledged: Mary and Georg C. Ehnrooth’s foundation, Finnish Eye Foundation, De Blindas va¨nner – Sokeiden ysta¨va¨t, Finska La¨karesa¨llskapet, Victoriastiftelsen, Finnish Eye and Tissue Bank Foundation, Paulo Foundation, EVO grant/Helsinki University Eye Hospital, Finland. References Acosta, M.C., Tan, M.E., Belmonte, C., Gallar, J., 2001. Sensations evoked by selective mechanical, chemical, and thermal stimulation of the conjunctiva and cornea. Invest. Ophthalmol. Vis. Sci. 42, 2063–2067. Albers, K.M., Wright, D.E., Davis, B.M., 1994. Overexpression of nerve growth factor in epidermis of transgenic mice causes hypertrophy of the peripheral nervous system. J. Neurosci. 14, 1422–1432. Araki, K., Ohashi, Y., Kinoshita, S., Hayashi, K., Kuwayama, Y., Tano, Y., 1994. Epithelial wound healing in the denervated cornea. Curr. Eye Res. 13, 203–211.
Barendregt, P.J., van den Bent, M.J., van Raaij-van den Aarssen, V.J., van den Meiracker, A.H., Vecht, C.J., van der Heijde, G.L., Markusse, H.M., 2001. Involvement of the peripheral nervous system in primary Sjo¨gren’s syndrome. Ann. Rheum. Dis 60, 876–881. Belmonte, C., 2007. Eye dryness sensations after refractive surgery: impaired tear secretion or ‘‘phantom’’ cornea? J. Refract. Surg. 23, 598–602. Belmonte, C., Acosta, M.C., Schmelz, M., Gallar, J., 1999. Measurement of corneal sensitivity to mechanical and chemical stimulation with a CO2 esthesiometer. Invest. Ophthalmol. Vis. Sci. 40, 513–519. Benitez del Castillo, J.M., Wasfy, M.A., Fernandez, C., Garcia-Sanchez, J., 2004. An in vivo confocal masked study on corneal epithelium and subbasal nerves in patients with dry eye. Invest. Ophthalmol. Vis. Sci. 45, 3030–3035. Benitez del Castillo, J.M., Acosta, M.C., Wassfi, M.A., az-Valle, D., Gegundez, J.A., Fernandez, C., Garcia-Sanchez, J., 2007. Relation between corneal innervation with confocal microscopy and corneal sensitivity with noncontact esthesiometry in patients with dry eye. Invest. Ophthalmol. Vis. Sci. 48, 173–181. Beuerman, R.W., Schimmelpfennig, B., 1980. Sensory denervation of the rabbit cornea affects epithelial properties. Exp. Neurol 69, 196–201. Bourcier, T., Acosta, M.C., Borderie, V., Borras, F., Gallar, J., Bury, T., Laroche, L., Belmonte, C., 2005. Decreased corneal sensitivity in patients with dry eye. Invest. Ophthalmol. Vis. Sci. 46, 2341–2345. Dartt, D.A., 2004. Dysfunctional neural regulation of lacrimal gland secretion and its role in the pathogenesis of dry eye syndromes. Ocul. Surf 2, 76–91. De Paiva, C.S., Pflugfelder, S.C., 2004. Corneal epitheliopathy of dry eye induces hyperesthesia to mechanical air jet stimulation. Am. J. Ophthalmol. 137, 109–115. Fox, R.I., Stern, M., 2002. Sjo¨gren’s syndrome: mechanisms of pathogenesis involve interaction of immune and neurosecretory systems. Scand. J. Rheumatol. (Suppl. (116)), 3–13. Fox, R.I., Stern, M., Michelson, P., 2000. Update in Sjo¨gren syndrome. Curr. Opin. Rheumatol 12, 391–398. Gemignani, F., Marbini, A., Pavesi, G., Di, V.S., Manganelli, P., Cenacchi, G., Mancia, D., 1994. Peripheral neuropathy associated with primary Sjo¨gren’s syndrome. J. Neurol. Neurosurg. Psychiatry 57, 983–986. Hamrah, P., Zhang, Q., Liu, Y., Dana, M.R., 2002. Novel characterization of MHC class II-negative population of resident corneal Langerhans cell-type dendritic cells. Invest. Ophthalmol. Vis. Sci. 43, 639–646. Hamrah, P., Liu, Y., Zhang, Q., Dana, M.R., 2003. Alterations in corneal stromal dendritic cell phenotype and distribution in inflammation. Arch. Ophthalmol. 121, 1132–1140. Hasan, W., Zhang, R., Liu, M., Warn, J.D., Smith, P.G., 2000. Coordinate expression of NGF and alpha-smooth muscle actin mRNA and protein in cutaneous wound tissue of developing and adult rats. Cell Tissue Res. 300, 97–109. Imai, S., Tokunaga, Y., Konttinen, Y.T., Maeda, T., Hukuda, S., Santavirta, S., 1997. Ultrastructure of the synovial sensory peptidergic fibers is distinctively altered in different phases of adjuvant induced arthritis in rats: ultramorphological characterization combined with morphometric and immunohistochemical study for substance P, calcitonin gene related peptide, and protein gene product 9.5. J. Rheumatol. 24, 2177–2187. Jester, J.V., Huang, J., Barry-Lane, P.A., Kao, W.W., Petroll, W.M., Cavanagh, H.D., 1999. Transforming growth factor(beta)-mediated corneal myofibroblast differentiation requires actin and fibronectin assembly. Invest. Ophthalmol. Vis. Sci. 40, 1959–1967. Lee, H.K., Ryu, I.H., Seo, K.Y., Hong, S., Kim, H.C., Kim, E.K., 2006. Topical 0.1% prednisolone lowers nerve growth factor expression in keratoconjunctivitis sicca patients. Ophthalmology 113, 198–205. Lemp, M.A., 1995. Report of the National Eye Institute/Industry workshop on Clinical Trials in Dry Eyes. CLAO J. 21, 221–232. Mastropasqua, L., Nubile, M., Lanzini, M., Carpineto, P., Ciancaglini, M., Pannellini, T., Di, N.M., Dua, H.S., 2006. Epithelial dendritic cell distribution in normal and inflamed human cornea: in vivo confocal microscopy study. Am. J. Ophthalmol. 142, 736–744. Niederkorn, J.Y., Stern, M.E., Pflugfelder, S.C., De Paiva, C.S., Corrales, R.M., Gao, J., Siemasko, K., 2006. Desiccating stress induces T cell-mediated Sjo¨gren’s syndrome-like lacrimal keratoconjunctivitis. J. Immunol. 176, 3950–3957. Niissalo, S., Hukkanen, M., Imai, S., Tornwall, J., Konttinen, Y.T., 2002. Neuropeptides in experimental and degenerative arthritis. Ann. N.Y. Acad. Sci. 966, 384–399. Petroll, W.M., Cavanagh, H.D., Jester, J.V., 1998. Clinical confocal microscopy. Curr. Opin. Ophthalmol 9, 59–65. Pflugfelder, S.C., 2004. Antiinflammatory therapy for dry eye. Am. J. Ophthalmol. 137, 337–342. Rosenberg, M.E., Tervo, T.M., Petroll, W.M., Vesaluoma, M.H., 2000. In vivo confocal microscopy of patients with corneal recurrent erosion syndrome or epithelial basement membrane dystrophy. Ophthalmology 107, 565–573. Rosenberg, M.E., Tervo, T.M., Muller, L.J., Moilanen, J.A., Vesaluoma, M.H., 2002. In vivo confocal microscopy after herpes keratitis. Cornea 21, 265–269. Rozsa, A.J., Beuerman, R.W., 1982. Density and organization of free nerve endings in the corneal epithelium of the rabbit. Pain 14, 105–120. Schiffman, R.M., Christianson, M.D., Jacobsen, G., Hirsch, J.D., Reis, B.L., 2000. Reliability and validity of the. Ocular Surface Disease Index. Arch. Ophthalmol. 118, 615–621. Stern, M.E., Beuerman, R.W., Fox, R.I., Gao, J., Mircheff, A.K., Pflugfelder, S.C., 1998. The pathology of dry eye: the interaction between the ocular surface and lacrimal glands. Cornea 17, 584–589. Streppel, M., Azzolin, N., Dohm, S., Guntinas-Lichius, O., Haas, C., Grothe, C., Wevers, A., Neiss, W.F., Angelov, D.N., 2002. Focal application of neutralizing
I.S. Tuisku et al. / Experimental Eye Research 86 (2008) 879–885 antibodies to soluble neurotrophic factors reduces collateral axonal branching after peripheral nerve lesion. Eur. J. Neurosci. 15, 1327–1342. Tuisku, I.S., Lindbohm, N., Wilson, S.E., Tervo, T.M., 2007. Dry eye and corneal sensitivity after high myopic LASIK. J. Refract. Surg. 23, 338–342. Tuominen, I.S., Konttinen, Y.T., Vesaluoma, M.H., Moilanen, J.A., Helinto, M., Tervo, T. M., 2003. Corneal innervation and morphology in primary Sjo¨gren’s syndrome. Invest. Ophthalmol. Vis. Sci. 44, 2545–2549. Villani, E., Galimberti, D., Viola, F., Mapelli, C., Ratiglia, R., 2007. The cornea in Sjo¨gren’s syndrome: an in vivo confocal study. Invest. Ophthalmol. Vis. Sci. 48, 2017–2022. Vitali, C., Bombardieri, S., Jonsson, R., Moutsopoulos, H.M., Alexander, E.L., Carsons, S.E., Daniels, T.E., Fox, P.C., Fox, R.I., Kassan, S.S., Pillemer, S.R., Talal, N., Weisman, M.H., 2002. Classification criteria for Sjo¨gren’s syndrome: a revised
885
version of the European criteria proposed by the American-European Consensus Group. Ann. Rheum. Dis. 61, 554–558. Xu, K.P., Yagi, Y., Tsubota, K., 1996. Decrease in corneal sensitivity and change in tear function in dry eye. Cornea 15, 235–239. Zhang, M., Chen, J., Luo, L., Xiao, Q., Sun, M., Liu, Z., 2005. Altered corneal nerves in aqueous tear deficiency viewed by in vivo confocal microscopy. Cornea 24, 818– 824. Zhivov, A., Stave, J., Vollmar, B., Guthoff, R., 2005. In vivo confocal microscopic evaluation of Langerhans cell density and distribution in the normal human corneal epithelium. Graefes Arch. Clin. Exp. Ophthalmol. 243, 1056–1061. Zhivov, A., Stave, J., Vollmar, B., Guthoff, R., 2007. In vivo confocal microscopic evaluation of langerhans cell density and distribution in the corneal epithelium of healthy volunteers and contact lens wearers. Cornea 26, 47–54.