Corneal endothelial morphology in patients with herpetic keratitis attending tertiary eye centre

Abstract

Background

Herpetic keratitis is a common cause of infectious corneal disease and can result in long term visual morbidity. While its effects on the corneal epithelium and stroma are well recognized, its impact on the corneal endothelium remains less clearly understood. This study aimed to evaluate and compare corneal endothelial morphology, including endothelial cell density, cell size variability, and hexagonality, between eyes affected by herpetic keratitis and their contralateral unaffected eyes.

Methods

This hospital-based cross-sectional comparative study included 39 patients with unilateral herpetic keratitis, comprising 39 affected eyes and 39 contralateral unaffected eyes. All participants underwent corneal evaluation using non-contact specular microscopy (NIDEK CEM-530, NIDEK Co., Japan). Endothelial cell density, coefficient of variation, and percentage of hexagonal cells were recorded as measures of endothelial cell density and morphology. Central corneal thickness was also measured. Comparisons between affected and contralateral unaffected eyes were performed using paired statistical analysis.

Results

Mean endothelial cell density was significantly lower in eyes affected by herpetic keratitis compared with contralateral unaffected eyes (2555.38 ± 474.05 vs. 2879.00 ± 270.26 cells/mm², p < 0.001). Affected eyes demonstrated a significantly higher coefficient of variation (34.92 ± 6.35% vs. 29.77 ± 6.57%, p < 0.001), indicating greater variability in endothelial cell size. The percentage of hexagonal cells was significantly reduced in affected eyes compared with unaffected eyes (61.10 ± 5.04% vs. 66.82 ± 3.97%, p = 0.02). Central corneal thickness did not differ significantly between affected and contralateral unaffected eyes.

Conclusions

Eyes with unilateral herpetic keratitis demonstrated reduced corneal endothelial cell density and hexagonality, along with increased variability in cell size compared with contralateral unaffected eyes, indicating corneal endothelial alterations associated with the disease.

Background

Herpes simplex virus type 1, a member of the Orthoherpesviridae family, is an enveloped double-stranded DNA virus and the predominant cause of ocular herpes infections [1, 2]. Herpes simplex virus keratitis remains a leading infectious cause of corneal blindness, with clinical manifestations ranging from epithelial keratitis to stromal keratitis and endotheliitis. Recurrent inflammation, particularly with stromal or endothelial involvement, can result in progressive scarring, visual impairment, and irreversible corneal decompensation [3, 4]. Given the critical role of the corneal endothelium in maintaining corneal transparency, herpesvirus related endotheliitis represents a key determinant of long-term visual outcomes in herpetic keratitis [5–7].

In vivo confocal microscopy (IVCM) studies have demonstrated frequent endothelial involvement in herpetic keratitis, reporting reduced endothelial cell density (ECD), pseudoguttata, inflammatory cell infiltration, and an accelerated rate of endothelial cell loss compared with fellow eyes [5, 8]. Importantly, bilateral IVCM studies have shown reduced ECD even in clinically unaffected contralateral eyes, suggesting subclinical endothelial involvement in unilateral disease [9]. The specular microscope is an optical reflection microscope where a slit of light is focussed on the corneal endothelial surface and specularly (mirror-like) reflected light rays are focussed on the film plane for viewing on a real-time monitor [10]. Wide-field specular microscopy studies in herpetic stromal keratitis and keratouveitis have similarly reported polymegathism, pleomorphism, and ECD reductions in severely affected eyes relative to fellow eyes [6, 8].

Despite this evidence, most contemporary data rely on IVCM, and there is limited use of routine specular microscopy for systematic bilateral endothelial assessment across herpetic keratitis subtypes, particularly in real-world tertiary care setting [5, 11]. Given the high surgical burden and complex case mix encountered at tertiary eye centres, detailed specular microscopy based evaluation of ECD and endothelial morphology in both affected and contralateral eyes is clinically relevant for endothelial risk stratification and prevention of postoperative corneal decompensation [6, 12]. Accordingly, this study aims to assess corneal ECD and morphologic changes using specular microscopy in eyes with herpetic keratitis compared with healthy contralateral eyes.

Methods

This hospital-based cross-sectional study included patients clinically diagnosed with unilateral herpes simplex keratitis based on history and slit-lamp biomicroscopic examination, recruited from the cornea clinic of B.P. Koirala Lions Centre for Ophthalmic Studies (BPKLCOS). All consecutive eligible patients presenting during the study period were screened for inclusion. Participants were enrolled according to predefined inclusion and exclusion criteria using a non-probability consecutive sampling method. Inclusion criteria comprised unilateral herpes simplex epithelial or stromal keratitis in inactive state with no history of prior episodes of keratitis and no evidence of raised intraocular pressure, with the clinically unaffected fellow eye serving as the control. Exclusion criteria comprised prior intraocular surgery or ocular trauma; ocular or systemic disease known to influence corneal endothelial morphology (including glaucoma, increased intraocular pressure, diabetes mellitus, rheumatoid arthritis, and systemic lupus erythematosus); corneal pathologies such as scar, ectasia, or endothelial dystrophy; pterygium extending more than 2 mm onto the cornea; and poor endothelial image quality due to dense corneal haze, edema, or scarring. Cases with associated anterior uveitis/iritis with keratic precipitates were excluded. Written informed consent was obtained from all participants prior to enrollment. Each subject underwent a comprehensive ophthalmic examination including distance visual acuity measurement using a Snellen chart at 6 m, intraocular pressure assessment, slit-lamp evaluation for assessment of the type and severity of herpetic keratitis (epithelial or stromal), and dilated fundus examination (Fig. 1). Cases were categorized accordingly for subgroup analysis.

Fig. 1.

Flow diagram illustrating the study protocol, including patient recruitment, eligibility assessment, informed consent, comprehensive ophthalmic examination, and corneal endothelial evaluation using specular microscopy

Specular microscopy procedure

Corneal endothelial parameters were assessed using a non-contact specular microscope, the NIDEK CEM-530, following the manufacturer’s instructions. Images were acquired using the automated alignment and capture mode with central fixation, and endothelial images were obtained from the central cornea. Specular microscopy was performed within two weeks of clinical presentation or once the corneal media became sufficiently clear to obtain reliable endothelial counts, whichever occurred earlier. ECD, average cell area, coefficient of variation (CV), percentage of hexagonal cells, and central corneal thickness (CCT) were automatically calculated using the device software, with corneal thickness derived from the distance between epithelial and endothelial reflections (Fig. 2).

Fig. 2.

Representative specular microscopy images (NIDEK CEM-530) showing corneal endothelial morphology and quantitative parameters in an affected eye with herpetic keratitis and the contralateral unaffected eye

Statistical analysis

Statistical analysis was performed using SPSS software version 25. Data normality was assessed using visual inspection of histograms and Q–Q plots prior to parametric analysis. Continuous variables were compared between affected and contralateral unaffected eyes using the paired samples t-test. Results are presented as mean ± standard deviation, and a p-value < 0.05 was considered statistically significant with a 95% confidence interval.

Results

Demographics

Among the 39 patients with herpetic keratitis, epithelial keratitis was observed in 10 (26%) cases and stromal keratitis in 29 (74%). The cohort comprised 23 (59%) males and 16 (41%) females, with a mean age of 46.33 ± 18.90 years (range 18–77 years). In affected eyes, best-corrected visual acuity (BCVA) with a mean of 0.33 ± 0.34 logarithm of the minimum angle of resolution (logMAR), whereas contralateral unaffected eyes demonstrated a mean BCVA of 0.18 ± 0.31 logMAR (Table 1).

Table 1.

Demographic and baseline clinical characteristics of patients with herpetic keratitis

Variable	Value
Age (years), mean ± SD	46.33 ± 18.90
Gender, n (%)
Male	23 (59)
Female	16 (41)
Type of herpetic keratitis, n (%)
Epithelial keratitis	10 (26)
Stromal keratitis	29 (74)
BCVA (logMAR), mean ± SD
Affected eyes	0.33 ± 0.34
Contralateral unaffected eyes	0.18 ± 0.31

Specular microscopy findings

ECD was significantly lower in affected eyes compared with contralateral unaffected eyes (2555.38 ± 474.05 vs. 2879.00 ± 270.26 cells/mm², p < 0.001) (Fig. 3). The number of analyzed endothelial cells was also significantly reduced in affected eyes (88.00 ± 51.15 vs. 161.08 ± 40.26, p < 0.001). No significant difference was observed in average cell area between affected and unaffected eyes (385.92 ± 161.99 vs. 347.46 ± 95.46 μm², p = 0.21), and the standard deviation of cell area was likewise comparable between groups (120.15 ± 37.35 vs. 108.95 ± 36.49 μm², p = 0.08). The CV was significantly higher in affected eyes (34.92 ± 6.35% vs. 29.77 ± 6.57%, p < 0.001). Maximum and minimum endothelial cell areas did not differ significantly between the two groups (p = 0.73 and p = 0.18, respectively). The percentage of hexagonal cells was significantly lower in affected eyes compared with contralateral unaffected eyes (61.10 ± 5.04% vs. 66.82 ± 3.97%, p = 0.02) (Fig. 4). CCT was similar between affected and unaffected eyes (538.08 ± 40.41 vs. 546.26 ± 56.60 μm, p = 0.32) (Table 2). A sensitivity analysis excluding the two eyes with markedly reduced endothelial cell density demonstrated that the overall results remained consistent and statistically significant (p < 0.001), indicating that the observed differences were not solely driven by outlier cases.

Fig. 3.

Box-and-whisker plot comparing endothelial cell density (cells/mm²) between eyes affected by herpetic keratitis and contralateral unaffected eyes

Fig. 4.

Box-and-whisker plot comparing hexagonality (%) between eyes affected by herpetic keratitis and contralateral unaffected eyes

Table 2.

Comparison of specular microscopy parameters between affected and contralateral unaffected eyes in herpetic keratitis

Parameters	Affected eye (n = 39)			Contralateral unaffected eye (n = 39)			p-values
	Min	Max	Mean ± SD	Min	Max	Mean ± SD
ECD (cells/mm2)	1036	3263	2555.38 ± 474.05	2404	3491	2879.00 ± 270.26	< 0.001*
NUM	11	228	88.00 ± 51.15	34	231	161.08 ± 40.26	< 0.001*
AVG (µm2)	275	906	385.92 ± 161.99	273	901	347.46 ± 95.46	0.21
SD (µm2)	75	215	120.15 ± 37.35	69	283	108.95 ± 36.49	0.08
CV (%)	26	49	34.92 ± 6.35	13	48	29.77 ± 6.57	< 0.001*
MAX (µm2)	656	2695	1062.18 ± 347.03	624	2079	1088.69 ± 304.31	0.73
MIN (µm2)	45	653	172.15 ± 149.88	43	591	135.26 ± 78.42	0.18
HEX (%)	50	71	61.10 ± 5.04	60	75	66.82 ± 3.97	0.020*
CCT (µm)	469	650	538.08 ± 40.41	459	841	546.26 ± 56.60	0.32

* Indicates statistical significance (p < 0.05) NUM, number of analyzed cells; AVG, average cell area; SD, standard deviation of cell area; MAX, maximum cell area; MIN, minimum cell area; HEX, percentage of hexagonal cells

Subgroup analysis: epithelial vs. stromal keratitis

An independent samples t-test showed no significant difference in endothelial cell density (ECD) between epithelial and stromal herpetic keratitis groups (2459.80 ± 516.27 vs. 2588.34 ± 463.61 cells/mm², respectively; p = 0.47) (Fig. 5).

Fig. 5.

Box-and-whisker plot comparing endothelial cell density (cells/mm²) between eyes affected by epithelial and stromal herpetic keratitis

Discussion

In this hospital-based cross-sectional study employing specular microscopy, unilateral herpes simplex epithelial or stromal keratitis was associated with corneal endothelial involvement when compared with contralateral clinically unaffected eyes. This finding indicates that endothelial alterations may occur even in clinically unilateral disease, independent of prior keratitis episodes or elevated intraocular pressure. The detection of endothelial changes in this context highlights the vulnerability of the corneal endothelium to herpetic anterior segment inflammation and underscores the utility of specular microscopy for identifying early or subclinical endothelial compromise.

In this study, eyes affected by herpetic keratitis exhibited a significant reduction in corneal endothelial cell density compared with their contralateral clinically unaffected eyes. This finding is consistent with previous reports in unilateral herpes simplex keratitis, in which affected eyes demonstrated lower endothelial cell density than healthy controls, and contralateral eyes also showed evidence of subclinical endothelial involvement [13]. Similar severity dependent endothelial alterations have been described in herpetic keratouveitis, with minimal or non-significant changes in milder disease and more pronounced endothelial compromise in the presence of greater inflammatory activity [8]. Comparable endothelial vulnerability has also been reported in herpes zoster associated keratouveitis, further supporting the susceptibility of the corneal endothelium to herpetic anterior segment inflammation [14, 15]. In addition to endothelial cell loss, affected eyes in the present cohort demonstrated increased CV and reduced hexagonal cell percentage, indicative of polymegathism and pleomorphism. These morphologic changes likely reflect compensatory enlargement and shape distortion of surviving endothelial cells following endothelial cell loss, a well-recognized adaptive response of the corneal endothelium, as previously reported [16–19]. In summary, the reduced ECD, increased cellular size variability, and decreased hexagonality seen in herpetic keratitis eyes are consistent with findings from previous Herpes simplex virus (HSV) and Herpes zoster virus (HZV) keratouveitis studies and the broader anterior chamber inflammation literature. These patterns are suggestive of clinically relevant structural and functional endothelial compromise and emphasize the need for careful endothelial monitoring in this population [3, 8, 13–15].

Similarly, in this current study comparison of ECD between epithelial and stromal herpetic keratitis revealed no statistically significant difference, suggesting a similar comparable degree of endothelial compromise despite involvement of different corneal layers. This finding supports the concept that endothelial damage in herpes simplex virus-related disease is not strictly dependent on the depth of corneal involvement. Consistent with this observation, prior confocal and specular microscopy studies have demonstrated reduced ECD across the full clinical spectrum of herpetic keratitis including epithelial keratitis, stromal keratitis, endotheliitis, and keratouveitis without consistent evidence that stromal disease results in uniformly greater endothelial loss than more superficial forms [5, 6, 13, 20]. Reports of ECD reduction in affected eyes compared with healthy controls or fellow eyes further support the presence of diffuse endothelial vulnerability in herpetic disease [5–7, 13, 20]. Nevertheless, larger longitudinal studies are warranted to determine whether more subtle differences in endothelial loss or associated morphologic alterations, such as polymegathism and pleomorphism, emerge over time, particularly in stromal keratitis and endotheliitis.

The presence of polymegathism, pleomorphism, and reduced ECD in affected eyes supports direct involvement of the corneal endothelium in herpetic keratitis, as consistently demonstrated in prior specular and confocal microscopy studies across herpes simplex keratitis, endotheliitis, and keratouveitis [5, 6, 8]. These studies have reported lower ECD accompanied by increased CV and reduced hexagonality, reflecting endothelial cell loss with compensatory morphologic remodeling and functional compromise [5–7, 15]. The extent of endothelial damage appears to be influenced by disease severity and recurrence, with more pronounced endothelial alterations observed in severe or recurrent herpetic keratouveitis compared with milder disease, and recurrent stromal involvement contributing to progressive immune-mediated injury and cumulative endothelial loss over time [3, 6, 8, 15]. In the present study, two cases with markedly reduced endothelial cell density may reflect such more severe inflammatory presentations, although detailed clinical documentation was limited. These endothelial changes carry important clinical and surgical implications, as increased cellular size variability and loss of normal endothelial morphology indicate reduced endothelial reserve and greater susceptibility to decompensation during intraocular procedures, including cataract surgery and penetrating keratoplasty [15, 21–23]. Moreover, in eyes with herpetic corneal scarring, long-term graft survival is closely linked to endothelial health, with chronic endothelial loss recognized as a major contributor to late graft failure following penetrating keratoplasty [21–23]. Accordingly, patients with recurrent disease or stromal and endothelial involvement warrant close, long-term monitoring of endothelial parameters to identify progressive damage and to guide the timing and selection of future ocular interventions [5, 7, 8, 15, 23].

Most available evidence on corneal endothelial alterations in herpetic keratitis has been derived from IVCM rather than specular microscopy [5, 7, 8, 13, 24]. IVCM has been widely applied across the spectrum of herpetic eye disease, including epithelial and stromal keratitis, endotheliitis, and keratouveitis, and has facilitated bilateral comparisons between affected eyes, contralateral clinically unaffected eyes, and healthy controls, consistently demonstrating endothelial cell loss accompanied by polymegathism, pleomorphism, and inflammatory changes, even in clinically uninvolved fellow eyes [5, 7, 13]. In contrast, specular microscopy has been used less frequently and largely in studies focusing on stromal keratitis or keratouveitis, often without systematic bilateral evaluation across different herpetic subtypes, thereby limiting direct cross study comparison of endothelial parameters [6, 8, 14]. This heterogeneity in imaging modality and study design underscores the relative scarcity of quantitative bilateral endothelial assessment using specular microscopy and highlights the contribution of the present study in addressing this gap. Given the consistent documentation of endothelial cell loss and morphologic abnormalities across herpetic keratitis subtypes, routine endothelial evaluation using specular microscopy or IVCM appears warranted to inform clinical monitoring, therapeutic decision making, and surgical planning, with the aim of minimizing progressive endothelial decompensation and optimizing long-term visual outcomes [5–8, 13, 24].

Several limitations should be acknowledged. The sample size was relatively small, and endothelial assessment was limited to central measurements, which may affect generalizability. The cross-sectional design precludes assessment of longitudinal or progressive changes. Detailed clinical severity parameters, including symptom duration, corneal sensation, and antiviral history, were not evaluated, and cases with stromal edema, keratic precipitates, and corneal neovascularization were excluded, limiting further stratification. In addition, detailed documentation of subtle anterior chamber inflammation (e.g., endotheliitis or iritis) was not available, restricting clinicopathological correlation in cases with marked endothelial loss. The use of contralateral eyes as controls may introduce inter-eye correlation and limits comparison with independent healthy controls. Finally, a small number of cases with markedly reduced endothelial cell density may have influenced group estimates; however, sensitivity analysis confirmed the robustness of the findings.

Conclusion

In conclusion, eyes with unilateral herpetic keratitis demonstrated significantly reduced endothelial cell density, increased coefficient of variation, and decreased hexagonality compared with contralateral unaffected eyes, indicating corneal endothelial alterations associated with the disease. These findings highlight the importance of longitudinal studies to evaluate the progression and clinical implications of these endothelial changes.

Abbreviations

IVCM

In vivo confocal microscopy

ECD

Endothelial cell density

CV

Coefficient of variation

CCT

Central corneal thickness

BCVA

Best-corrected visual acuity

logMAR

Logarithm of the minimum angle of resolution

HSV

Herpes simplex virus

HZV

Herpes zoster virus

Author contributions

BKS, SS and SB conceptualized and designed the study. BKS and PJ collected the data. BKS and NDJ performed data analysis. BKS drafted the manuscript. SS, SB, NDJ, and PJ critically revised the manuscript. All authors read and approved the final manuscript.

Funding

This study received no external funding.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Declarations

Ethics approval and consent to participate

Ethical approval was obtained from the Institutional Review Committee of Maharajgunj Medical Campus, Institute of Medicine, Tribhuvan University, Kathmandu, Nepal (Ref. No.: 320/081/082). The study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. Written informed consent was obtained from all participants prior to enrollment.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.