Corneal Nerve Structure and Function in Patients With Non-Sjögren Dry Eye: Clinical Correlations

Abstract

Purpose.: To evaluate the relationship between the in vivo confocal microscopic (IVCM) morphology of subbasal corneal nerves, corneal sensitivity, and clinical parameters in patients with non-Sjögren dry eye disease (NSDD).

Methods.: Forty-three patients with NSDD and 14 healthy age- and sex-matched control subjects were included. Each patient underwent an evaluation of ocular surface disease symptoms using the Ocular Surface Disease Index (OSDI); tear film break-up time (TBUT); corneal and conjunctival staining (Oxford scale); the Schirmer test; corneal sensation testing using the Cochet-Bonnet esthesiometer; and corneal subbasal nerve analysis with IVCM. One eye of each subject was included in the study.

Results.: Mean corneal sensitivity was significantly lower in the NSDD group as compared with the control group (P = 0.014). Corneal subbasal nerves showed significant changes in NSDD patients as compared with normal subjects: lower density (P < 0.001); increased tortuosity (P < 0.001); number of beadings (P < 0.001); and width (P = 0.041). In patients with NSDD, corneal subbasal nerve density was correlated to age (r = −0.352, P = 0.021); the Oxford scale (r = −0.486, P = 0.01); and central corneal sensitivity (r = 0.383, P = 0.041). The maximum length of nerve fibers within a frame (MaxL) was correlated to the OSDI (r = −0.265, P = 0.019) and the Oxford scale (r = −0.307, P = 0.039). In multivariate analysis after adjustment for age, corneal subbasal nerve density remained correlated to the Oxford score (P < 0.001). The Oxford score was also correlated to central corneal sensitivity (r = −0.304, P = 0.042).

Conclusions.: NSDD patients have both structural and functional alterations of subbasal corneal nerves and these changes are related to the severity of dry eye.

Introduction

Dry eye disease (DED) is a multifactorial disease of the tears and ocular surface resulting in symptoms of discomfort, visual disturbance, and tear film instability with potential damage to the ocular surface. ¹ DED is also recognized as a disturbance of the lacrimal functional unit that comprises the lacrimal glands, the ocular surface and lids, and the sensory and motor nerves that connect them. ^1,2 The innervation of the cornea and bulbar conjunctiva is supplied mainly by the sensory fibers of the ophthalmic branch of the trigeminal nerve and by the less numerous sympathetic and parasympathetic nerve fibers. ³ In addition to their important sensory function, corneal nerves provide protective and trophic functions and also regulate corneal epithelial integrity, proliferation, and wound healing. ⁴ A stimulation of corneal nerves followed by nerve alterations has been postulated as one of the core pathophysiological mechanisms in DED. ¹ Moreover, although the basis for symptoms in DED is not fully understood, it may imply the stimulation of nociceptive ocular surface nerves. ¹  

The well-defined in vivo confocal microscopy (IVCM) appearance of subbasal corneal nerves has facilitated their qualitative and quantitative in vivo analysis in health and disease states and following corneal surgery. ⁵ However, there are still conflicting results in the literature concerning corneal subbasal nerve structure in DED. ⁶ Similarly, studies evaluating corneal sensation in patients with DED have obtained controversial findings. ⁶ This apparent variability may be attributed to the different methodologies used, the stages and etiologies of DED, especially the proportion of Sjögren-syndrome patients enrolled. ^6,7  

In a preliminary study, we used semiautomatic software to quantify corneal subbasal nerves with IVCM images, a Java-based image processing software (ImageJ; National Institutes of Health, Bethesda, MD), and a plug-in (NeuronJ; Biomedical Imaging Group, Lausanne, Switzerland) dedicated to the tracing and quantification of elongated image structures such as dendrites and neurons. ⁶ By analyzing patients with different types of ocular surface disease, we demonstrated that the correlation between structure and function of corneal nerves might vary according to ocular surface disease etiology. The purpose of the present study was therefore to evaluate the relationship between IVCM morphology of subbasal corneal nerves, corneal sensitivity, and dry eye clinical parameters in a large number of patients with non-Sjögren dry eye disease (NSDD). 

Patients and Methods

Patients

The study was conducted at the Beijing Institute of Ophthalmology with the approval of the Medical Ethics Committee of Beijing Tongren Hospital. All patients were informed of the aims of the study and their consent was obtained according to the Declaration of Helsinki. A total of 43 patients with DED not associated with Sjögren syndrome (29 women and 14 men; mean age, 46.23 ± 9.74 years; range, 28–68 years) were consecutively recruited from the Cornea Unit of Beijing Tongren Hospital from October 2012 to February 2013 (NSDD group). NSDD was defined as Schirmer 1 testing less than 5 mm and/or tear film break-up time (TBUT) less than 10 seconds, accompanied by complaints of ocular irritation in the absence of other ocular or systemic diseases. ⁸ Fourteen age- and sex-matched control subjects (eight women and six men; mean age, 45.40 ± 9.20 years; range, 30–56 years) were also recruited (control group). All control subjects had no complaint of ocular surface irritation and no anterior segment abnormality on biomicroscopic examination and ocular surface tests. Exclusion criteria for both groups were: aged under 18 years; subjects unable to complete the questionnaire or understand the procedures; the presence of ocular or systemic disease or the use of topical or systemic medications that may affect the cornea and the ocular surface (except the use of nonpreserved tear substitutes in the DED group); previous eye surgery; or contact lens wear. 

Clinical Evaluation

The Ocular Surface Disease Index (OSDI) questionnaire was used for the grading of subjective dry eye symptoms. ⁸ Then all patients underwent a complete examination of the ocular surface of both eyes in the following order: TBUT, corneal and conjunctival fluorescein staining, the Schirmer test, corneal sensitivity evaluation, and IVCM analysis of the central cornea subbasal nerves. The Schirmer test was performed without anesthesia for 5 minutes with the patient's eyes closed. TBUT was measured by instilling fluorescein into the inferior cul-de-sac and calculating the average of two consecutive breakup times. Corneal and conjunctival staining was evaluated using the Oxford scale after instillation of fluorescein under a yellow filter. 

Corneal Sensation

Corneal sensation was measured using a contact nylon thread esthesiometer (Luneau 12/100-mm Cochet-Bonnet; Luneau, Prunay-le-Gillon, France). The nylon filament, which mechanically stimulates corneal nerves, was applied with a low pressure perpendicular to the cornea. Starting from 6 cm, the filament length was progressively reduced in 5-mm steps until the first response occurred. The longest filament length (cm) resulting in a positive response was verified twice and recorded as the indicator of corneal sensitivity. ^6,9 Corneal sensitivity was measured in the central cornea and in the superior, inferior, nasal, and temporal quadrants. Mean corneal sensitivity was defined as the mean of the measures obtained in the five different areas. 

In Vivo Confocal Microscopy

In vivo laser scanning confocal microscopy of the cornea was performed using confocal microscopy (Rostock Cornea Module of the Heidelberg Retina Tomograph [HRT/RCM]; Heidelberg Engineering GmbH, Heidelberg, Germany). ⁶ The images comprised 384 × 384 pixels covering an area of 400 × 400 μm with a transversal optical resolution of 2 μm, an axial optical resolution of 4 μm, and an acquisition time of 0.024 seconds (Heidelberg Engineering GmbH). Images of subbasal nerves of the central cornea were acquired using the same illumination intensity (manual mode) and by focusing the microscope beneath the basal epithelium. Approximately 20 images of the corneal subbasal nerve layer were acquired in the central cornea for each eye and the five images (400 × 400 μm) with most of the subbasal nerve fibers were selected for quantitative analysis. As described previously, images of corneal subbasal nerves were analyzed retrospectively using NeuronJ (Biomedical Imaging Group) by a single researcher (AL), who was masked regarding patient identity and the results of ocular surface investigations. ⁶ The different parameters evaluating the subbasal corneal nerves were the following: the density of corneal nerves; the mean (MeanL), maximum (MaxL), and minimum length (MinL) of corneal nerves; the number of corneal nerves; the width of corneal nerves; the number of beadings; the number of branchings; and the tortuosity and reflectivity classified according to a semiquantitative scale. ^4,6 The definition of each corneal subbasal nerve parameter is shown in Table 1. For each eye and each parameter, the results were the mean of the analysis of five images. 

Table 1

 

Definition of IVCM Subbasal Nerve Parameters

Parameter	Definition	Expressed In
Nerve density	Total length of the nerves visible within a frame	mm/mm2
MeanL	Mean length of nerves and nerve fragments within a frame	μm
MaxL	Length of the longest nerve or nerve fragment observed within a frame	μm
MinL	The length of the shortest nerve or nerve fragment observed within a frame	μm
Number of nerves	Sum of the long nerve fiber bundles observed within a frame	Number/mm2
Width of nerves	Mean of the width of all long nerve fibers within a frame (five measurements per long nerve fiber)	μm
Number of beadings	Number of beadings measured on all long nerve fibers within a frame	Number/100 μm of nerve length
Number of branchings	Number of fine fibers connecting to long nerve fibers within a frame	Number/mm2
Tortuosity	Classified according to a semiquantitative scale	Arbitrary unit (0–4)
Reflectivity	Classified according to a semiquantitative scale	Arbitrary unit (0–4)

Statistical Analysis

Statistical analysis was performed using a commercially available statistical software package (SPSS for Windows, version 20.0; SPSS, Chicago, IL). For each patient, one eye was chosen randomly for statistical analysis. Variables were compared using the nonparametric Mann-Whitney Test. The association between variables was examined using the Spearman correlation test. Multivariate logistic regression analysis was used to investigate the relation between the different associated parameters. All P values were considered statistically significant when the values were less than 0.05. 

Results

There was no difference in terms of sex (P = 0.658) and age (P = 0.9) between the DED group and the control group. Concerning ocular surface clinical evaluation, DED patients had significantly more symptoms: OSDI (P < 0.001); lower TBUT (P < 0.001); lower Schirmer test results (P = 0.001); and a higher Oxford score (P < 0.001) compared with the control group. Mean corneal sensitivity was significantly lower in the DED group (P = 0.014), while central sensitivity did not differ between the two groups (P = 0.238). The results of the clinical data are presented in Table 2. 

Table 2

 

Demographic and Clinical Test Results

Parameters	Control Group	DED Group	P Value
Patients, n	14	43
Sex
Female, n (%)	8 (57.1)	29 (67.4)	0.658
Male, n (%)	6 (42.9)	14 (32.6)
Age, y	45.40 ± 9.20	46.23 ± 9.74	0.900
ODSI, arbitrary units	1.50 ± 4.09	36.93 ± 15.09	<0.001
Schirmer 1 test, mm	12.20 ± 6.03	5.53 ± 7.00	0.001
TBUT, s	14.00 ± 2.78	4.52 ± 2.03	<0.001
Oxford scale, arbitrary units	0.00 ± 0.00	1.16 ± 1.77	<0.001
MCS, mm	5.98 ± 0.04	5.79 ± 0.34	0.014
CCS, mm	5.95 ± 0.158	5.84 ± 0.42	0.238

 CCS, central cornea sensitivity; MCS, mean corneal sensitivity.

DED patients had significant corneal subbasal nerve changes compared with control subjects. The density and number of nerves were significantly lower (P < 0.001 for both comparisons). The MaxL of the subbasal nerves was shorter in the DED group (P = 0.019). Greater width (P = 0.041); number of beadings (P < 0.001); and tortuosity (P < 0.001) were also observed (Fig.). The IVCM analysis of subbasal corneal nerves is presented in Table 3. 

Figure

 

IVCM images (400 × 400 μm) of corneal subbasal nerves in the control group (A) and in the dry eye group (C). Images (B) and (D) are their respective subbasal nerve tracings using the semiautomatic NeuronJ software.

Table 3

 

IVCM Analysis of Subbasal Nerve Parameters

Parameters	Control Group	DED Group	P Value
Patients, n	14	43
Density of subbasal nerves, mm/mm2	21,856 ± 2138	16,256 ± 3,478	<0.001
MeanL, μm	197.54 ± 27.97	187.44 ± 24.41	0.340
MaxL, μm	440.34 ± 24.31	412.35 ± 41.68	0.019
MinL, μm	40.80 ± 10.64	44.09 ± 13.61	0.517
Number of subbasal nerves, number/mm2	45.87 ± 4.21	34.91 ± 8.08	<0.001
Width, μm	2.30 ± 0.41	2.75 ± 0.68	0.041
Number of beadings, number/100 μm of nerve length	13.44 ± 2.45	18.62 ± 4.04	<0.001
Number of branchings, number/mm2	15.40 ± 6.09	16.08 ± 6.57	0.699
Tortuosity, arbitrary units	1.94 ± 0.46	3.01 ± 0.49	<0.001
Reflectivity, arbitrary units	3.20 ± 0.55	3.06 ± 0.75	0.552

Within the DED group, there were significant correlations between age and symptoms (r = −0.384; P = 0.011) and between TBUT and the Oxford scale (r = −0.553, P < 0.001). The mean corneal sensitivity and central sensitivity were both correlated with age (r = −0.353, P = 0.02 and r = −0.459, P = 0.002, respectively); and central sensitivity was correlated with the Oxford score (r = −0.304, P = 0.042; Table 4). In univariate analysis, when evaluating the relationship between dry eye clinical tests and IVCM subbasal nerve parameters, the corneal subbasal nerve density evaluated was correlated with age (r = −0.352, P = 0.021); the Oxford scale (r = −0.486, P = 0.01); and central sensitivity (r = 0.383, P = 0.041). Similarly, the number of nerves was correlated with age (r = −0.397, P = 0.008) and the Oxford score (r = −0.357, P = 0.023), but not with central sensitivity (r = 0.243, P = 0.117). The Oxford score was also correlated with the MaxL (r = −0.307, P = 0.039); the MinL (r = 0.324, P = 0.028); the number of branchings (r = −0.373, P = 0.014); and corneal nerve reflectivity (r = −0.339, P = 0.026). The MeanL and MaxL of subbasal corneal nerves were significantly correlated with the OSDI (r = −0.393, P = 0.009 and r = −0.265, P = 0.019, respectively; Table 5). In multivariate analysis, the density and the number of subbasal corneal nerves remained correlated with the Oxford score (r = −0.570, P < 0.001 and r = −0.431, P = 0.002, respectively) and age (r = −0.347, P = 0.018 and r = −0.392, P = 0.004, respectively). 

Table 4

 

Statistical Results of Correlations Between Dry Eye Clinical Tests

Parameters	Sex	Age	OSDI	Schirmer I Test	TBUT	Oxford Scale	MCS
Age
r	−0.260
P	0.092
OSDI
r	−0.054	−0.384
P	0.730	0.011
Schirmer I test
r	0.117	−0.185	−0.035
P	0.454	0.234	0.826
TBUT
r	0.072	−0.019	−0.212	−0.057
P	0.645	0.902	0.172	0.719
Oxford scale
r	−0.102	−0.054	0.079	−0.184	−0.553
P	0.517	0.733	0.615	0.236	<0.001
MCS
r	0.194	−0.353	0.125	0.036	0.035	−0.038
P	0.213	0.02	0.424	0.817	0.822	0.807
CCS
r	0.212	−0.459	0.231	0.171	0.288	−0.304	0.652
P	0.171	0.002	0.137	0.273	0.062	0.042	<0.001

Table 5

 

Statistical Results of Correlations Between Dry Eye Clinical Tests and IVCM Subbasal Parameters in the DED Group

Parameters	Sex	Age	OSDI	Schirmer I Test	TBUT	Oxford Scale	MCS	CCS
Density of subbasal nerves
r	0.096	−0.352	0.470	0.129	0.132	−0.486	−0.087	0.383
P	0.540	0.021	0.763	0.409	0.398	0.010	0.577	0.041
MeanL
r	−0.126	0.225	−0.393	0.162	0.072	0.069	−0.237	−0.252
P	0.421	0.147	0.009	0.300	0.251	0.659	0.126	0.103
MaxL
r	−0.062	−0.037	−0.265	0.228	0.219	−0.307	0.109	0.205
P	0.693	0.813	0.019	0.142	0.157	0.039	0.487	0.187
MinL
r	−0.036	0.195	−0.180	−0.018	−0.182	0.324	−0.212	−0.260
P	0.819	0.210	0.249	0.908	0.243	0.028	0.173	0.092
Number of subbasal nerves
r	0.094	−0.397	0.185	0.026	0.088	−0.357	−0.102	0.243
P	0.548	0.008	0.236	0.867	0.574	0.023	0.514	0.117
Width
r	−0.062	0.055	0.052	−0.124	0.112	−0.058	−0.137	0.062
P	0.693	0.727	0.740	0.428	0.475	0.713	0.381	0.693
Number of beadings
r	−0.016	−0.131	0.077	0.105	0.036	−0.027	0.223	0.266
P	0.919	0.401	0.624	0.501	0.817	0.863	0.151	0.085
Number of branchings
r	0.072	−0.278	0.089	0.39	0.129	−0.373	0.179	0.248
P	0.646	0.058	0.569	0.806	0.408	0.014	0.252	0.108
Tortuosity
r	0.099	0.013	0.079	−0.048	−0.273	0.231	0.282	0.164
P	0.528	0.935	0.613	0.760	0.076	0.136	0.067	0.293
Reflectivity
r	−0.089	0.013	−0.159	0.158	0.210	−0.339	−0.166	0.060
P	0.572	0.934	0.308	0.312	0.176	0.026	0.287	0.702

In DED patients, there were correlations between the density of subbasal nerves and the number of nerves (r = 0.719, P < 0.001); the number of branchings (r = 0.576, P < 0.001); MaxL (r = 0.354, P = 0.02); and tortuosity (r = −0.384, P = 0.011). The number of subbasal nerves was correlated with the tortuosity (r = −0.295, P = 0.047). The number of branchings was correlated with the MeanL (r = −0.382, P = 0.011); MaxL (r = 0.286, P = 0.043); MinL (r = −0.340, P = 0.026); and reflectivity (r = 0.346, P = 0.023). The reflectivity of subbasal nerves was also correlated to their width (r = 0.398, P = 0.008). The results of subbasal nerve correlations are presented in Table 6. 

Table 6

 

Statistical Results of Correlations Between IVCM Subbasal Parameters in the DED Group

Parameters	Density of Subbasal Nerves	MeanL	MaxL	MinL	Number of Subbasal Nerves	Width	Number of Beadings	Number of Branchings	Tortuosity
MeanL
r	0.038
P	0.810
MaxL
r	0.354	0.243
P	0.020	0.116
MinL
r	−0.117	0.588	−0.073
P	0.455	<0.001	0.641
Number of subbasal nerves
r	0.719	−0.112	0.116	−0.236
P	<0.001	0.476	0.458	0.127
Width
r	−0.253	−0.122	−0.259	−0.323	−0.069
P	0.101	0.435	0.094	0.035	0.662
Number of beadings
r	−0.074	−0.083	0.01	−0.112	0.241	0.074
P	0.636	0.598	0.949	0.476	0.119	0.638
Number of branchings
r	0.576	−0.382	0.286	−0.340	0.233	−0.169	0.086
P	<0.001	0.011	0.043	0.026	0.132	0.280	0.585
Tortuosity
r	−0.384	−0.160	−0.164	0.211	−0.295	−0.029	0.03	−0.002
P	0.011	0.307	0.294	0.174	0.047	0.853	0.847	0.988
Reflectivity
r	0.186	−0.227	0.059	−0.640	0.160	0.398	−0.027	0.346	−0.147
P	0.233	0.144	0.709	<0.001	0.305	0.008	0.863	0.023	0.348

Discussion

In the present study, we found that patients with NSDD had both structural and functional alterations of subbasal corneal nerves, and these changes were related to the severity of dry eye. Compared with an aged-matched control group, corneal subbasal nerves in NSDD patients showed lower density and number, shorter long nerve fibers and greater tortuosity, number of beadings and width. Subbasal corneal nerves have been previously analyzed in DED using IVCM with conflicting results. When evaluating subbasal nerve density, some authors including our group ⁶ observed a decrease, ^7,10,11 while others found no change ^12–15 or even an increase in density in patients with SSDD. ¹⁶ However, most of these studies were conducted on SSDD patients or groups of patients composed of different proportions of both SSDD and NSDD patients, making the comparison between studies difficult. To our knowledge, considering specifically NSDD, which is a much more prevalent condition than SSDD, ¹⁷ the present study is the first to find significantly lower corneal subbasal nerve density and number compared with age-matched normal subjects. ¹⁶ Although in a similar study Benitez del Castillo et al. observed lower density and nerve number in both NSDD and SSDD patients, a statistically significant difference was only reached when comparing SSDD with the control group of young subjects. ⁷ The effect of a negative correlation between corneal nerve density and age also observed in the present study may explain these results. ^10,16 The smaller sample size of previous studies evaluating NSDD, the lower resolution of the confocal microscope used (slit-scanning versus laser scanning confocal microscope), as well as the nerve density quantification method may also account for these differences. The correlation between disease severity (Oxford score) and nerve density could also explain why nerve alterations might be more easily detected in SSDD patients, who are often more severely injured than NSDD patients. As previously shown for both SSDD and NSDD patients, ^5,7,10,15,16 we observed that subbasal nerves in NSDD patients tend to have greater tortuosity and number of beadings. These changes are thought to be characteristic of metabolically active subbasal nerves in response to tissue damage. ⁷ This chronic inflammatory stimulation of corneal nerves could also lead to the increased subbasal nerve width detected in NSDD patients. In DED, the long-lasting inflammation of the ocular surface stimulates the release of various cytokines and neurotrophic factor release such as the nerve growth factor (NGF), which is known to induce peripheral nervous system hypertrophy. ^18,19 Although it is considered that the pathophysiology of SSDD is different from NSDD, these results emphasize that both diseases share common features, in particular regarding corneal nerve alterations. 

Nerve structural changes were associated with functional alterations in patients with NSDD with a weak but significantly lower mean sensitivity as compared with age-matched subjects. Similar to our preliminary study that included a small group of SSDD and NSDD patients, ⁶ a correlation was found between central subbasal nerve density and central corneal sensitivity in patients with NSDD. However, both parameters were associated with age and after multivariate logistic regression, the density of subbasal nerves was no longer associated with corneal sensitivity. Corneal sensation in DED and its correlation to IVCM subbasal nerve alterations is a subject of debate. ^6,7,12,13 Although most studies described decreased corneal sensation in patients with DED, ^{6,7,9,11,12,20} no change or even an increased corneal sensation has been described in either SSDD ¹³ or NSDD. ²¹ The Cochet-Bonnet esthesiometer explores only mechanical sensitivity and is less sensitive than the noncontact pneumo-esthesiometer. ^7,20 However, despite the differences in corneal sensitivity measurements, the pathophysiology of nerve alterations in dry eye itself may explain such conflicting results. Although a lower number and density of nerves can account for the lower central corneal sensation observed in DED patients, the altered excitability of corneal nerves also interferes with this correlation and makes its evaluation difficult. De Paiva et al. described mechanical corneal hypersensitivity in patients with NSDD as compared with normal subjects, probably due to altered corneal epithelial barrier functions. ²¹ As suggested by these authors, corneal epitheliopathy exposes subapical nerve endings in patients with DED and may modify nerve response by lowering the stimulation threshold. Altered ocular surface epithelial cells also release inflammatory mediators that can locally sensitize corneal nerves. ¹⁹ Both corneal nerve exposure and sensitization may lead to structural alterations with abnormal enhanced but also lowered response to stimulation, as previously described. ^6,7 Unfortunately, the actual resolution of IVCM (1–2 μm) does not allow the visualization of nerve epithelial ends that may be primarily injured in DED and also directly responsible for corneal surface sensitivity. ^5,6,22 Nevertheless, the present correlation between the Oxford score and both functional and morphological alterations of corneal subbasal nerves also supports this hypothesis of a direct relation between ocular surface epithelium and nerve alterations in NSDD. 

Subjective ocular symptoms often do not correlate with ocular clinical tests in DED. ¹ Although corneal nerves regenerate after injury, they exhibit abnormal responsiveness and spontaneous discharges ¹⁹ and possibly new connections between them. ²³ This altered excitability may lead to the abnormal sensitivity observed in DED, but also to spontaneous pain or hyperesthesia not correlated to clinical DED test findings. The modifications of corneal nerves may account for this discrepancy, but to date no study has observed a direct relationship between ocular symptoms and IVCM subbasal nerve changes in DED. Interestingly, an original IVCM subbasal nerve parameter, MaxL, was negatively correlated to both symptoms (OSDI) and the Oxford score. This parameter was also associated with the density of nerves, which remains the main quantifying parameter for the study of subbasal corneal nerves with IVCM. ^6,22 As IVCM images are 2D images, this parameter represents the presence of long nerve fibers within a focal plane parallel to the corneal surface. Therefore, it may combine a density analysis and the axial tortuosity of subbasal nerves. This increased subbasal nerve tortuosity has often been associated with dry eye nerve alterations. ⁵ However, despite the development of new software allowing for a more precise quantification of tortuosity, ²⁴ this interesting parameter has always been evaluated semiquantitatively by comparing nerve images to a validated scale. ⁴ A new parameter that can reflect tortuosity but is easy to quantify and is correlated to both symptoms and ocular surface alterations may be very useful for further IVCM studies on subbasal nerves in DED. 

Most IVCM studies have analyzed central subbasal corneal nerves because of their particularly well-defined appearance, but they represent only a limited segment of corneal nerve structures. ^6,22 However, it is actually the only imaging device that allows in vivo direct visualization and quantification of human corneal nerves in normal and pathological conditions. As reviewed recently by Zhivov et al., various approaches have been developed in the last few years to better characterize subbasal nerve parameters on IVCM images. ²⁵ Software such as the NeuronJ plug-in (Biomedical Imaging Group) used in this study offers new parameters that may better evaluate the corneal nerve structure changes in the eye, but also systemic diseases such as diabetes mellitus. ²⁵ Composite images covering larger corneal areas made automatically by computer software ²⁶ and three-dimensional reconstruction ²⁷ will certainly provide more reliable evaluation of corneal subbasal nerve structures in the near future. 

In the present study, NSDD patients showed subbasal corneal nerve alterations that were associated with dry eye severity, lower corneal sensation, and ocular surface symptoms. Ocular surface nerve alterations are involved in the pathophysiological mechanism leading to DED and are also responsible for symptoms. Consequently, new neuroprotective or neurotrophic approaches are currently under evaluation or have been evaluated for the treatment of DED. ^7,28,29 Therefore, a better understanding of the functional and structural alterations of nerves in DED and their correlation with clinical symptoms and signs is an important step for the further development of such treatments. 

Acknowledgments

Supported by an unrestricted grant from “La Fondation de France” (AL). The authors alone are responsible for the content and writing of the paper. 

Disclosure: A. Labbé, None; Q. Liang, None; Z. Wang, None; Y. Zhang, None; L. Xu, None; C. Baudouin, None; X. Sun, None