This study explored corneal inflammation in KC by detecting ICs using laser IVCM. The investigations revealed a significant rise in ICs in keratoconic eyes relative to healthy corneas. An increase in the IC density of patients with KC is consistent with a decrease in the number of subbasal nerves. The alterations were more noticeable in high-grade KC than in low-grade KC. Our results confirm the presence of intracorneal inflammation and some degree of neurotrophic keratopathy in patients with KC.

In general, there are two main pathogeneses of corneal tissue degradation: degeneration and inflammation [34]. Theoretically, the etiology of progressive stromal loss in KC is presumed to arise from a degenerative process rather than an inflammatory response, as keratoconic eyes appear clinically uninflamed. However, several researchers have proposed that KC’s pathogenesis may involve chronic inflammatory pathways [35,36,37,38]. Elevated levels of diverse serum markers (eg. neutrophil/lymphocyte ratio, monocyte/high-density lipoprotein cholesterol ratio), indicating systemic immune responses, were also found in patients with KC [36,37,38].

Our study results concur with the inflammation hypothesis, as we found a significant rise in ICs in the subbasal layer of keratoconic corneas visualized by IVCM, relative to normal corneas. Moreover, the IC density was greater in high-grade KC than in low-grade KC. These findings demonstrate the existence of intracorneal inflammation in KC, with the degree of inflammation depending upon the severity of the disease. Mandathara et al. also reported concordant results, as they found a significantly high number of Langerhans cells (LCs) in their series of KC patients [39].

Typically, dendritiform LCs or ICs are rarely detected in the central cornea in the quiescent state. The migration of ICs from the periphery toward the central cornea indicates a trafficking pattern of ICs mediated by the immune response [21]. An influx of corneal ICs has been shown in various conditions, such as those found in infectious keratitis, corneal graft rejection, and vernal keratoconjunctivitis [20, 23, 29, 40, 41]. However, some indiscernible inflammation resulting from dry eye disease or prolonged contact lens wear can also induce a rise in IC numbers [20, 21, 42].

As we know, the pathophysiology of KC is multifactorial. Several KC-related factors are constituents and play a role in destroying corneal collagen. Ocular allergy and systemic atopy, often present in patients with KC, render high levels of pro-inflammatory cytokines in tears and serum [16,17,18,19, 36,37,38]. Floppy eyelid syndrome (FES), recognized as a common coexisting disorder of KC. Abnormalities in the tear film dynamics in FES generate a rise in MMP-9, which produces chronic ocular surface inflammation [43, 44]. Eye rubbing has also been documented as a cause of keratoconjunctivitis [45]. Based on the physiological role of the corneal epithelial basement membrane in controlling corneal homeostasis and wound healing, the inciting signal of epithelial trauma mediates an inflammatory cascade through the healing process. From scientific rationale mentioned above, we presumed that repetitive mechanical trauma from chronic eye rubbing in KC patients activates corneal inflammation.

Regarding contact lens use, studies have analyzed the effects of various contact lens types on the cornea [46, 47]. The prolonged use of contact lenses leads to mechanical corneal epithelial injury, which releases apoptotic cytokines [48]. Moreover, confocal microscopic studies have revealed a high corneal epithelial dendritic cell density in the central cornea of contact lens wearers. This finding suggests that a centripetal migration of ICs occurred in response to an inflammatory stimulus [42, 49].

In the current study, ocular allergies and systemic atopy were the most common KC-related factors, similar to reports in the literature [3, 4, 8]. One may think that the increased number of ICs detected in keratoconic eyes might be influenced by allergies. From our study, there was not associated between ocular allergies and IC density. Moreover, none of the KC-related factors were significantly correlated with the IC numbers detected by IVCM.

In actuality, any KC-related factor is a plausible cause of corneal inflammation. We were somewhat surprised not to find a correlation between KC-associated factors and IC density unlike the results from previous studies. However, those prior studies analyzed ocular surface inflammation in patients with specific conditions, and many of their cases exhibited similar features with clinically inflamed eyes, resulting in a positive correlation with ICs. In contrast, our KC patients had a variety of associated factors with different severities, and their eyes were clinically not inflamed. Therefore, it is possible that the relationship between each KC-related factor and the ICs may not have been detected. Nonetheless, most KC patients presented with composite factors. A cumulative level of inflammation could contribute to a significant increase in ICs, as we observed in confocal microscopy. In other word, an immune-inflammatory response may result from not just a single factor but several factors’ combined effect, yielding the distinctly high IC density of keratoconic corneas. Additionally, many patients with KC need to wear contact lenses to improve their vision. Thirty-eight percent of our patients had a history of RGP use. These patients were requested to refrain from wearing the lenses for at least a month before enrollment to obtain the decent corneal topographic evaluation. Nevertheless, no data indicate the duration needed to ensure that eyes are free of inflammation after discontinuing contact lenses.

The present study revealed that the density of ICs was greater in patients with high-grade KC than in those with low-grade KC, and the number of ICs showed a strong relationship with the severity of KC. It could be speculated that sustained exposure to inflammatory-precipitating factors could engender inflammation accumulating in the patients’ corneas, eventually progressing to an advanced KC stage.

Regarding corneal innervation, it has been widely demonstrated the structural changes of subbasal nerve plexus in KC. Many researchers have shown a reduced nerve length and density in confocal images, while some have observed increased nerve tortuosity in the cornea’s apex [50,51,52,53]. Our results also found subbasal nerve decline in both the number and density, with the decline correlating with a slight loss of corneal sensitivity. In our subgroup analysis, corneal nerves were significantly reduced in high-grade KC compared with low-grade KC. Several studies have shown that inflammatory cell infiltrates presented around corneal nerves were associated with various corneal diseases. Cruzat et al. demonstrated that perineural dendritic cells increase in parallel with a decrease of corneal nerves in patients with different types of infectious keratitis [54]. Hamrah et al. also revealed similar findings in herpes simplex virus (HSV) keratitis [26]. In addition, other research groups observed perineural infiltration corresponding to subbasal nerve changes in conditions such as aqueous tear deficiency dry eye [55] and vernal keratoconjunctivitis [32]. All this research evidently supports our findings that inflammatory cells located near subbasal nerves are implicated in corneal nerve alterations.

Interestingly, we observed an increase in nerve tortuosity only in low-grade KC. Kawabuchi et al. explained this morphological change in which nerve tortuosity was deemed to be a biomarker in representing nerve regeneration [56]. The researchers presumed that mild inflammation in the early stage of KC may cause minimal nerve damage; therefore, the remaining nerves can renew. As the inflammation persists and the disease progresses to a more advanced stage, the subsequent nerve destruction might be too severe to restore. Although significant nerve diminishment was detected in overall KC patients, we did not find an association between subbasal nerve numbers and IC density. We assume that the decreased nerve number in KC may derive from chronic subtle inflammation and also other factors, such as disorganized corneal stroma, which disallow proper nerve regeneration.

There are some limitations of this study: 1) We performed confocal analysis in both eyes of some patients if those eyes were eligible. Taking into account the same KC-related factors in both eyes of some patients, we modified the statistical analysis on a dependent-eye basis. However, a larger number of cases would yield more valid outcomes, and 2) we did not include patients with other inflammatory cornea and ocular surface diseases to compare with KC patients. The different confocal findings between KC and clinically inflamed eyes may offer new insights into the evolution and progression of KC.