Purpose
To analyze the density and morphology of corneal epithelial cells and keratocytes by in vivo confocal microscopy (IVCM) in patients with herpes zoster ophthalmicus (HZO) as associated with corneal innervation.
Design
Prospective, controlled and masked cross-sectional study.
Methods
setting : Single-center study. patients : Thirty eyes with the diagnosis HZO and their contralateral clinically unaffected eyes, 15 eyes of 15 normal controls. intervention procedures : In vivo confocal microscopy and corneal esthesiometry of the central cornea. main outcome measures : Changes in morphology and density of the superficial and basal epithelial cells and stromal keratocytes, and correlation with corneal sensation.
Results
The density of superficial epithelial cells in HZO eyes with severe sensation loss (766.5 ± 25.2 cells/mm 2 ) was significantly lower than both healthy control eyes (1450.23 ± 150.83 cells/mm 2 ) and contralateral unaffected eyes (1974.13 ± 298.24 cells/mm 2 ) ( P = .003). Superficial epithelial cell size (1162.5 μm 2 ) was significantly larger in HZO eyes with severe loss of sensation, as compared to contralateral (441.46 ± 298.14) or healthy eyes (407.4 ± 47.2μm 2 ; all P < .05). The density of basal epithelial cells, anterior keratocytes, and posterior keratocytes did not show statistical significance between patients, controls, and contralateral unaffected eyes. Changes in superficial epithelial cell density and morphology correlated strongly with corneal sensation.
Conclusions
In vivo confocal microscopy reveals profound HZO-induced changes in the superficial epithelium, as demonstrated by increase in cell size, decrease in cell density, and squamous metaplasia. We demonstrate that these changes strongly correlate with changes in corneal innervation in eyes affected by HZO.
Herpes zoster ophthalmicus (HZO) has numerous corneal presentations, such as epithelial keratitis, necrotizing stromal and immune keratitis, endotheliitis, and neurotrophic keratopathy. Neurotrophic keratopathy is thought to result from loss of corneal sensation owing to denervation, damaged epithelial basement membrane, stromal inflammation, and toxicity from topical medications. Neurotrophic keratopathy may present clinically with decreased visual acuity, persistent corneal epithelial defects, stromal opacification, corneal neovascularization, and stromal melts.
Slit-lamp examination has been the gold standard for detecting epithelial defects, stromal edema and infiltration, keratic precipitates, and iritis attributable to HZO, allowing for diagnosis and monitoring of patients. Recently, in vivo confocal microscopy (IVCM) has become an increasingly popular noninvasive device to image the cornea at the cellular and microstructural level in healthy as well as diseased corneas. Several studies have demonstrated the feasibility of this technology and have analyzed corneas in ocular rosacea, Sjögren’s syndrome, meibomian gland dysfunction, dry eye disease, Stevens-Johnson syndrome, and keratitis. More recently, our group has demonstrated that patients with unilateral HZO demonstrated bilateral subbasal corneal nerve alterations by IVCM. Further, the subbasal corneal nerve changes correlated strongly with loss of corneal sensation in the affected eyes of these patients. However, the consequences of altered sensory nerve function on other corneal layers still remain unknown in patients with HZO. Other studies have demonstrated that epithelial turnover is regulated by corneal nerves in several animal models. Thus, we hypothesized that corneal nerve loss would lead to loss of epithelial cells and keratocytes in patients with HZO. To test this hypothesis, we analyzed the morphology and density of corneal epithelial cells and keratocytes by IVCM in patients with HZO and correlated these findings with corneal innervation in this group.
Methods
This is a prospective, controlled, and masked cross-sectional study. This study was Health Insurance Portability and Accountability Act (HIPAA) compliant and adhered to the tenets of the Declaration of Helsinki. The Massachusetts Eye and Ear Infirmary institutional review board approved the study. All patients gave informed consent for this research after a detailed explanation of the nature of the study.
Subjects were recruited from the Cornea Service of the Department of Ophthalmology of the Massachusetts Eye & Ear Infirmary, Boston, Massachusetts, between 2006 and 2008. None of the patients was immunocompromised and all eyes were inactive, with no signs of active keratitis. Subjects with a history of ocular trauma, ocular surgery, contact lens use, or diabetes were excluded from the study. Prior to recruitment into the study, all patients and healthy subjects underwent a complete baseline ophthalmologic examination, including visual acuity measurement, anterior segment evaluation with a slit-lamp biomicroscope, fundus examination, and intraocular pressure (IOP) measurement by applanation tonometry.
Central corneal sensation was measured bilaterally with a Cochet-Bonnet esthesiometer (Luneau Ophthalmologie, Chartres, France). The corneal sensation in the unaffected eye was measured first. The esthesiometer stimulates the cornea by a retractable monofilament nylon (6 cm length, 0.12 mm diameter). It was shortened in steps of 1.0 cm if a positive response was not obtained, or advanced by 0.5 cm if a positive response was obtained. The longest final filament length resulting in a positive response was considered the corneal sensitivity threshold. This threshold was verified twice in our study.
A total of 30 eyes of 30 patients with a diagnosis of HZO, as well as the contralateral, clinically unaffected eyes, were included in this study. All patients have had a history of epithelial herpetic keratitis with no stromal involvement. Fifteen eyes of 15 normal volunteers comprised the control group. HZO keratitis patients were grouped into normal (>5.5 cm), mild (>2.5–5.5 cm), and severe (<2.5 cm) loss of corneal sensation, according to the corneal sensitivity threshold measurements, and results were correlated to confocal microscopic findings. Because contralateral eyes did not show significant differences between HZO subgroups, they were kept together as 1 combined group. No fluorescein was used prior to IVCM during the study. All subjects underwent bilateral examination of the central cornea with a slit-scanning confocal microscope (Confoscan 4; Nidek Technologies, Gamagori, Japan). In order to ensure masking, a randomly generated code was assigned to each patient at the time of scanning. The coding key was kept in a separate file by a technician, who was masked. Therefore, during image analysis, the observers (P.H. and M.H.D.) were masked to patient name, diagnosis, and severity. The microscope was equipped with a 40×/0.75 Zeiss Acroplan numerical aperture immersion objective lens. Topical 0.5% proparacaine eyedrops (Alcaine; Alcon, Fort Worth, Texas, USA) were instilled in both eyes prior to confocal scans, and 0.3% hypromellose (GenTeal gel; Novartis, East Hanover, New Jersey, USA) was used as a coupling medium for the objective lens. The lens was manually advanced until the gel contacted the central surface of the cornea. Three hundred fifty images per scan in every 7 μm were obtained at a speed of 25 frames per second. Moreover, a second scan was obtained for the anterior cornea, obtaining sections every 3 μm. Each image represented a coronal section of 460 × 345 μm, with a minimum axial step of 1 μm, magnification of ×500, and lateral resolution of 1 μm/pixel. A total of 4–8 scans were obtained for each cornea by the same experienced operator (P.H.) in all subjects, depending on full-thickness or anterior scan mode.
The mean of the 3 representative images per layer was used for analysis for each eye. The superficial epithelial layer, basal epithelial layer, subbasal corneal nerves, anterior stromal keratocytes, and posterior stromal keratocytes were included in the analysis. Cell counts were performed manually by Confoscan 4 NAVIS analysis software (Nidek Technologies). This was achieved by marking each clearly defined cell or nucleus in a predefined rectangular frame using the full-frame size on the computer screen. Two independent masked experienced observers counted the cells to determine superficial and basal epithelial cell density, as well as anterior and posterior keratocyte density. Superficial epithelial cells become highly reflective before desquamation. To assess epithelial damage, the percentage of highly reflective superficial epithelial cells was calculated by dividing highly reflective cells per frame by total number of superficial epithelial cells per frame and multiplying by 100. For epithelial cell size, 2 masked observers defined the borders of superficial epithelial cells, with Confoscan NAVIS software (Nidek Technologies) calculating epithelial cell area. The areas of superficial epithelial cells were averaged per frame. Cellular changes of the epithelium and stroma were correlated to corneal sensation.
Statistics
Statistical analysis was performed by Student’s t test, analysis of variance (ANOVA), and Pearson correlation coefficient. P values less than .05 were considered statistically significant. Analyses were performed with SAS software version 9.2 (SAS Institute Inc, Cary, North Carolina, USA).
Results
Patient Characteristics and Study Groups
Thirty eyes of 30 patients with HZO and 30 contralateral clinically unaffected eyes were included in the study. Fifteen eyes of 15 normal volunteers served as control group. Patient demographics and healthy subjects’ characteristics are shown in Table 1 . All eyes with HZO were subcategorized into 3 groups based on central corneal sensation as follows: normal sensation, mild sensation loss, and severe sensation loss (described in the methods section). These groups were used in order to correlate corneal sensation to IVCM findings. The correlation of corneal sensation loss to IVCM parameters of superficial epithelium, basal epithelium, and anterior and posterior keratocytes for all groups is presented in Table 2 . Demonstrative IVCM images for all layers analyzed for patients with HZO and controls are shown in Figure 1 .
| Controls | Herpes Zoster Ophthalmicus | |
|---|---|---|
| No. of patients | 15 | 30 |
| Age (mean ± SD; y) | 59 ± 17 | 58 ± 17 |
| Sex (male/female) | 8/7 | 14/16 |
| Corneal sensation (mean ± SD; cm) | 6.0 ± 0 | 3.4 ± 2.3 |
| Disease duration (mean ± SD; y) | – | 5.8 ± 6.7 |
| No. of episodes a (mean ± SD; n) | – | 2.1 ± 1.3 |
a Episodes reflect both primary and secondary immune keratitis.
| Superficial Epithelial Cell Density | Basal Epithelium Cell Density | Anterior Keratocyte Density | Posterior Keratocyte Density | Superficial Epithelial Cell Size | Hyperreflective Superficial Epithelial Cells | |
|---|---|---|---|---|---|---|
| Normal corneal sensation | 1754.23 ± 238.46 cells/mm 2 | 4950.2 ± 250.3 cells/mm 2 | 629.3 ± 381.4 cells/mm 2 | 827.5 ± 227.5 cells/mm 2 | 466.68 ± 328.86 μm 2 | 30.2% ± 30.55% |
| Mild corneal sensation loss | 946.67 ± 73.18 cells/mm 2 a | 4521.1 ± 201.02 cells/mm 2 | 782.2 ± 58.88 cells/mm 2 | 641.22 ± 13.59 cells/mm 2 | 902.2 ± 216.02 μm 2 a | 58.3% ± 11.79% a |
| Severe corneal sensation loss | 766.5 ± 25.2 cells/mm 2 a | 4412.5 ± 120.2 cells/mm 2 | 516.5 ± 221.2 cells/mm 2 | 586.01 ± 192.84 cells/mm 2 | 1162.5 ± 198.04 μm 2 a | 62.5% ± 12.99% a |
| Contralateral eyes | 1974.13 ± 298.24 cells/mm 2 | 5278.32 ± 432.96 cells/mm 2 | 822.68 ± 354.84 cells/mm 2 | 758.9 ± 276.32 cells/mm 2 | 441.46 ± 298.14 μm 2 a | 10.72% ± 10.65% |
| Controls | 2435.23 ± 224.3 cells/mm 2 | 5962.5 ± 83.21 cells/mm 2 | 901.2 ± 75.56 cells/mm 2 | 603.32 ± 55.34 cells/mm 2 | 407.4 ± 47.2 μm 2 | 0% |
a Statistically significant ( P < .05) compared to controls.
Superficial Epithelial Cell Density
We found a significant and gradual decrease in the density of superficial epithelial cells in HZO eyes, with 1754.23 ± 238.46 cells/mm 2 in HZO eyes with normal sensation, 946.67 ± 73.17 cells/mm 2 in HZO eyes with mild sensation loss, and 766.5 ± 25.2 cells/mm 2 in eyes with severe sensation loss as compared to 1974.13 ± 298.24 cells/mm 2 in contralateral eyes ( P = .01) and 2435.23 ± 224.3 cells/mm 2 in control eyes ( P = .003) ( Table 2 and Figure 2 ). Clinically unaffected contralateral eyes did not show any significant epithelial or stromal changes as compared to controls.
Superficial Epithelial Cell Size
There was a gradual increase in superficial epithelial cell size with loss of corneal sensation in HZO eyes. Superficial epithelial cell size was almost 3.0-fold larger in HZO eyes with severe loss of sensation (1162.5 ± 198.04 μm 2 ) as compared to contralateral (441.46 ± 298.14; P = .012) or normal eyes (407.4 ± 47.2 μm 2 ; P = .008) ( Table 2 and Figure 3 ). The IVCM scans in the mild sensation loss group also revealed a significant increase in mean superficial epithelial cell size (902.2 ± 216.02 μm 2 ), as compared with the contralateral eyes (441.46 ± 298.14; P = .03) and normal control group (407.4 ± 47.2 μm 2 ; P = .001), while HZO eyes with normal sensation did not demonstrate statistically significant changes from normal controls and contralateral eyes (both P > .05) ( Table 2 ).
Hyperreflective Superficial Epithelial Cells
Apart from cell density and size, we looked for other morphologic epithelial characteristics such as reflectivity of cell nucleoli. There was minimal reflectivity of cell nucleoli in controls or contralateral eyes. However, in eyes with HZO, we consistently found increased reflectivity of superficial epithelial cell nucleoli. Another striking finding was the presence of hyporeflective rings around the hyperreflective cell nucleoli in patients with severe sensation loss. Moreover, we observed bright cell borders and increased cell hyperreflectivity in HZO eyes with normal sensation. Interestingly, these findings were more pronounced in patients with mild and severe loss of sensation. A significant number of hyperreflective desquamating superficial epithelial cells were present in HZO eyes with normal sensation (30.2% ± 30.55%, P = .09) and mild (58.3% ± 11.79%, P = .002) and severe (62.5% ± 12.99%, P = .0003) loss of sensation but were absent in controls or minimal (10.72% ± 10.65%) in contralateral eyes ( Table 2 and Figure 4 ).
