The Role of Hi-Tech Devices in Assessment of Corneal Healing in Patients with Neurotrophic Keratopathy

To prove the role of high-tech investigation in monitoring corneal morphological changes in patients with neurotrophic keratopathy (NK) using Keratograph 5M (K5M) and anterior segment OCT (AS-OCT), corneal healing was monitored with Keratograph 5M (Oculus, Wetzlar, Germany) and AS-OCT (DRI, Triton, Topcon, Tokyo, Japan) in 13 patients (8F and 5M), aged from 24 to 93 years (67.8 ± 19) with severe NK, who were treated with Cenegermin 0.002% (20 μg/mL) (Oxervate®, Dompè, Farmaceutici Spa, Milan, Italy). The surface defects were evaluated on Keratograph 5M with ImageJ software and the corneal thickness variations were measured using DRI-Triton OCT software. Instrumental procedures were performed at baseline, and after 4 and 8 weeks of the treatment, respectively. The main outcome measures were reduction of the ulcers’ area and corneal thickness recovery. The mean area of the corneal ulcers was reduced between baseline and 4 weeks examination in all patients, and at 8 weeks all ulcers were completely healed. An increase of the corneal thickness was evidenced between the baseline visit and after the 4- and 8-week follow-up, respectively. Additionally, only in collaborating subjects the In Vivo Confocal Microscopy (IVCM) was performed with HRT Rostock Cornea Module (Heidelberg Eng GmbH) to study the corneal nerves fibres. High-tech diagnostics with K5M, AS-OCT and IVCM proved useful in the assessment of corneal morphology and the healing process in patients with NK and could be extended to assess other corneal pathologies.


Introduction
Unstoppable technological advances during previous decades have introduced new diagnostic tools in ophthalmology that provide high-level information about ocular structure and function. These cutting-edge tools for cornea and ocular surface investigation 2 of 9 permit assessment and quantification of corneal morphological changes during the healing processes and in relation to the adopted therapies.
Anterior segment optical coherence tomography (AS-OCT), based on swept-source laser technology, is an imaging technique for investigation of the anterior segment of the eye, allowing better evaluation, diagnosis, and management of different corneal diseases [1][2][3].
AS-OCT measures the entire corneal and epithelial thickness, analyzes corneal deposits and infiltrates, providing information as to their location and stromal depth, and enables the evaluation of the corneal flap following LASIK. It is also crucial in corneal lamellar surgery and in the study of ectatic pathologies, and a proven utility in the evaluation of pterygium, pinguecula, and scleromalacia [2,[4][5][6][7][8]. Different studies have highlighted the role of AS-OCT in the management of corneal infections such as keratitis and peripheral ulcers [9][10][11][12].
Keratograph 5M (OCULUS Optikgeräte GmbH, Wetzlar, Germany) is a video topographer and a non-invasive, repeatable, and reproducible tool for the assessment of the ocular surface, allowing the analysis of different morpho-functional aspects. It combines corneal topography and dry eye study in a single tool and represents the gold standard for ocular surface investigation [13][14][15][16][17][18]. A series of studies has demonstrated the importance and effectiveness of Keratograph 5M (K5M) in the evaluation of ocular surface diseases, such as Dry Eye Disease (DED) and Meibomian Gland Dysfunction (MGD) [13,15,16].
In vivo corneal confocal microscopy (IVCM) represents the main tool in the clinical investigation of the cornea in health and disease. It allows visualization of all corneal layers, cells, and nerves with extreme precision and high resolution that can be compared with histological imaging.
The normal SBNP fibers can be easily detected and appear as straight, well-defined, and clearly identified reflective structures that run in parallel and exhibit different grades of branching, tortuosity, and beadings [19].
The opportunity to examine, in vivo, the corneal sub-basal nerve plexus with a fast and repeatable method, and to perform a qualitative and quantitative analysis, makes the IVCM an important diagnostic tool to investigate the nerve fiber's status in corneal and systemic diseases [20,21].
Neurotrophic keratitis (NK) is caused by damage to the corneal innervation at any level (from the brain to the nerve endings of the sub-basal corneal nerve plexus), resulting in a reduction or abolition of corneal sensitivity, and alteration of the corneal trophism ensured by the proper innervation [22,23]. If untreated, it can compromise corneal healing, resulting in recurrent corneal erosions and, consequently, an increased risk of corneal over-infection, melting, and perforation [5].
There are many causes that can lead to neurotrophic keratitis (NK), including local or systemic diseases, surgery, trauma, infectious, and metabolic diseases; however, herpetic keratitis is the most frequently encountered cause [24,25].
In 1995, Mackie proposed the first classification of NK, with stages I, II, and III. Stage I is characterized by epithelial irregularity, such as punctate keratopathy; stage II is defined by recurrent or persistent epithelial defects (PED) without stromal involvement; and stage III is characterized by stromal involvement with ulcer, melting, and perforation [26]. Dua et al. classified the pathology into mild, involving only slight epithelial changes; moderate, with persisting epithelial defects (PED); and severe, with different grades of stromal involvement [23,24,27].
There is only one report highlighting the role of AS-OCT in NK assessment, together with confocal microscopy findings and used for classification purposes. Nevertheless, to our knowledge, there are no studies on the utility of AS-OCT and K5M in the monitoring of corneal healing processes in patients with corneal degenerations, particularly in severe NK, where the important corneal defects might occur.
The aim of this study is to analyze the morphological changes occurring during the healing of the corneal ulcers in patients with severe NK treated with rh-NGF using AS-OCT and K5M.

Materials and Methods
We analyzed instrumental data of patients with severe NK treated in the Ocular Surface diseases section of the Ophthalmology Clinic of the University Hospital of Messina. Thirteen patients (8 F and 5 M), aged from 24 to 93 years (67.8 ± 19) were considered.
The diagnosis and grade of severity of NK were made with slit-lamp biomicroscopy and Cochet-Bonnet aesthesiometer, and the ocular and/or systemic pathologies that caused NK were identified. The patients received Cenegermin drops (20 µg/mL) (Oxervate ® , Dompè, Farmaceutici Spa, Milan, Italy) 1 drop 6 times daily for 8 weeks, accordingly to the standardized protocol (https://www.ema.europa.eu/en/documents/product-information/ oxervate-epar-product-information_it.pdf). The study was conducted with respect to tenets of the Declaration of Helsinki and obtained approval from the Ethical Committee of the University Hospital of Messina.
The ocular surface and corneal morphology were examined by the same physician (LI) with Keratograph 5M (OCULUS Optikgeräte GmbH, Wetzlar, Germany) and the anterior segment module of the Swept Source OCT (AS-OCT, DRI Triton, Topcon, Japan); exams were performed at baseline and repeated after 4 weeks and after 8 weeks (end of treatment). Additionally, the IVCM was performed in cooperating patients with the Laser Scanning Confocal Microscope represented by Heidelberg retina tomograph (HRT) Rostock Cornea Module (Heidelberg Eng GmbH) aimed to detect the corneal nerve fibers.
The area of corneal defect was evaluated on K5M images through fluorescein staining with ImageJ software (imagej.nih.gov/ij/index.html) and expressed in microns squared. The corneal thickness was measured on corneal images acquired with AS-OCT and expressed in microns. The corneal thickness was calculated in the thinnest corneal point, corresponding to the ulcer bed using the ruler included in the software. The mean values of three consecutive measurements were used for analysis.
The main outcome measures were related to the healing process and comprised reduction of the area of ulcer and recovery of the corneal thickness after 4 and 8 weeks of the treatment.
Numerical data were expressed as median, first and third quartile and categorical variables as numbers.
A non-parametric approach was adopted for the data analysis using the exact significance of the p value at the 0.050 level. This was due to the small sample size and the results of the Kolmogorov test, which revealed that the data examined were not normally distributed.
The Friedman test was applied to compare measurements at different times in each group. Wilcoxon's test was used for pair comparisons in each subsequent session.
The statistical analyses were performed using SPSS 26.0 for Windows statistical software (SPSS Inc., Chicago, IL, USA).

Area of Ulcers
This group differs in the area levels recorded in the three study visits (χ2(2) = 14.89; p < 0.001). At the eighth week, all corneal defects healed. In the Median (IRQ) area, levels on the first day and the fourth week were 1,694,333 (628,218 to 5,214,884), 0 (0 to 1,370,276), respectively.
Post hoc analysis showed there were not statistically significant differences between week 4 and week 8 follow-up for area levels (p = 0.091). However, there was a statistically significant decrease in area levels in 4 and 8 weeks in comparison to baseline (p = 0.012, p < 0.011, respectively).
According to K5M data, after 4 weeks of the treatment, the ulcers healed completely in 69% (9 of 13) of eyes and at 8 weeks a complete recovery was registered in 100% of eyes (13 of 13) ( Figure 1). This group differs in the area levels recorded in the three study visits (χ2(2) = 14,89; p < 0.001). At the eighth week, all corneal defects healed. In the Median (IRQ) area, levels on the first day and the fourth week were 1,694,333 (628,218 to 5,214,884), 0 (0 to 1,370,276), respectively.
Post hoc analysis showed there were not statistically significant differences between week 4 and week 8 follow-up for area levels (p = 0.091). However, there was a statistically significant decrease in area levels in 4 and 8 weeks in comparison to baseline (p = 0.012, p < 0.011, respectively).
According to K5M data, after 4 weeks of the treatment, the ulcers healed completely in 69% (9 of 13) of eyes and at 8 weeks a complete recovery was registered in 100% of eyes (13 of 13). (Figure 1).

Corneal Thickness
This group differs in the OCT levels recorded in the three times (χ2 (2)

Corneal Nerves
The confocal examination was performed only in three cooperating patients at baseline and in five subjects at 8 weeks at the end of the treatment. The examination was unexecutable in other participants.
The total absence of nerves was observed at baseline examination, whereas after 8 weeks in four eyes the presence of sporadic fibers was detected. (Figure 3).

Corneal Thickness
This group differs in the OCT levels recorded in the three times (χ2 (2)  This group differs in the area levels recorded in the three study visits (χ2(2) = 14,89; p < 0.001). At the eighth week, all corneal defects healed. In the Median (IRQ) area, levels on the first day and the fourth week were 1,694,333 (628,218 to 5,214,884), 0 (0 to 1,370,276), respectively.
Post hoc analysis showed there were not statistically significant differences between week 4 and week 8 follow-up for area levels (p = 0.091). However, there was a statistically significant decrease in area levels in 4 and 8 weeks in comparison to baseline (p = 0.012, p < 0.011, respectively).
According to K5M data, after 4 weeks of the treatment, the ulcers healed completely in 69% (9 of 13) of eyes and at 8 weeks a complete recovery was registered in 100% of eyes (13 of 13). (Figure 1).

Corneal Thickness
This group differs in the OCT levels recorded in the three times (χ2 (2)

Corneal Nerves
The confocal examination was performed only in three cooperating patients at baseline and in five subjects at 8 weeks at the end of the treatment. The examination was unexecutable in other participants.
The total absence of nerves was observed at baseline examination, whereas after 8 weeks in four eyes the presence of sporadic fibers was detected. (Figure 3).

Corneal Nerves
The confocal examination was performed only in three cooperating patients at baseline and in five subjects at 8 weeks at the end of the treatment. The examination was unexecutable in other participants.
The total absence of nerves was observed at baseline examination, whereas after 8 weeks in four eyes the presence of sporadic fibers was detected (Figure 3).

Discussion
Several studies reported the utility of high-tech instruments in the assessment of corneal and ocular surface pathologies. However, to our knowledge, there are no reports that demonstrate the benefit of these instruments in monitoring morphological changes occurring in corneal tissue in patients with NK.
Optical coherence tomography (OCT) is a non-contact, in vivo imaging technique that uses low coherence interferometry to obtain in vivo imaging of different ocular structures [28].
As applied to cornea with an anterior segment module, the OCT imaging confirmed its crucial role in the clinical examination of the cornea and anterior segment of the eye.
Several studies demonstrated the usefulness of the AS-OCT in the clinical evaluation of a large number of corneal pathologies and in assessing the differences between normal and abnormal epithelium in the ocular surface neoplasia [29][30][31]. The undisputed role of AS-OCT in corneal assessment was reported in cases of edema, corneal infiltrates in microbial keratitis, and in peripheral ulcers [9][10][11][12]22]. AS-OCT also became a fundamental tool in the study of stromal deposits and structural changes in corneal dystrophies and degenerations and in the evaluation of the progression of keratoconus [32][33][34][35][36][37][38][39]. However, to our knowledge, the present study represents an initial report of the meaningful role of AS-OCT in the monitoring of corneal structural changes in the eyes with severe NK as the response to the therapy [40].
In the present study, we used AS-OCT to analyze the corneal thickness changes in the response to the treatment with Cenegermin in patients with severe NK. The results showed a significant increase of the corneal thickness that permitted us to consider the importance of imaging in monitoring the healing process.
K5M is a video topographer that allows analysis of tear meniscus height, lipid layer interferometry, NIBUT, bulbar redness, and/or Meibomian glands through Meibo-Scan in an automated, non-invasive, and objective manner [13]. Several studies highlighted the importance of K5M in evaluation and follow-up of patients with Dry Eye disease, but no studies reported the utility of K5M in monitoring the healing of corneal defects in patients with NK [14][15][16][17].
IVCM is the main tool in the structural investigation of the cornea. It allows visualization of all corneal layers, with cells and nerves, with a high resolution that can be compared with histological imaging. Specifically, it offers the opportunity to study the sub-basal nerve plexus fibers responsible for corneal health and trophism [19].
Neurotrophic keratitis (NK) is a progressive, degenerative condition that is characterized by diminished corneal sensitivity and inadequate corneal repair. This condition makes the cornea more vulnerable to damage and reduces the tearing reflex. As

Discussion
Several studies reported the utility of high-tech instruments in the assessment of corneal and ocular surface pathologies. However, to our knowledge, there are no reports that demonstrate the benefit of these instruments in monitoring morphological changes occurring in corneal tissue in patients with NK.
Optical coherence tomography (OCT) is a non-contact, in vivo imaging technique that uses low coherence interferometry to obtain in vivo imaging of different ocular structures [28].
As applied to cornea with an anterior segment module, the OCT imaging confirmed its crucial role in the clinical examination of the cornea and anterior segment of the eye.
Several studies demonstrated the usefulness of the AS-OCT in the clinical evaluation of a large number of corneal pathologies and in assessing the differences between normal and abnormal epithelium in the ocular surface neoplasia [29][30][31]. The undisputed role of AS-OCT in corneal assessment was reported in cases of edema, corneal infiltrates in microbial keratitis, and in peripheral ulcers [9][10][11][12]22]. AS-OCT also became a fundamental tool in the study of stromal deposits and structural changes in corneal dystrophies and degenerations and in the evaluation of the progression of keratoconus [32][33][34][35][36][37][38][39]. However, to our knowledge, the present study represents an initial report of the meaningful role of AS-OCT in the monitoring of corneal structural changes in the eyes with severe NK as the response to the therapy [40].
In the present study, we used AS-OCT to analyze the corneal thickness changes in the response to the treatment with Cenegermin in patients with severe NK. The results showed a significant increase of the corneal thickness that permitted us to consider the importance of imaging in monitoring the healing process.
K5M is a video topographer that allows analysis of tear meniscus height, lipid layer interferometry, NIBUT, bulbar redness, and/or Meibomian glands through Meibo-Scan in an automated, non-invasive, and objective manner [13]. Several studies highlighted the importance of K5M in evaluation and follow-up of patients with Dry Eye disease, but no studies reported the utility of K5M in monitoring the healing of corneal defects in patients with NK [14][15][16][17].
IVCM is the main tool in the structural investigation of the cornea. It allows visualization of all corneal layers, with cells and nerves, with a high resolution that can be compared with histological imaging. Specifically, it offers the opportunity to study the sub-basal nerve plexus fibers responsible for corneal health and trophism [19].
Neurotrophic keratitis (NK) is a progressive, degenerative condition that is characterized by diminished corneal sensitivity and inadequate corneal repair. This condition makes the cornea more vulnerable to damage and reduces the tearing reflex. As a result of inadequate healing, epithelial breakdown may result in ulceration, infection, melting, and perforation [23]. NK may be caused by any permanent modification of the corneal sensory innervation that impairs the function of the post-ganglionic fibers [41].
Numerous ocular or systemic disorders, such as diabetes mellitus or neurosurgical surgeries, may damage the V nerve, but herpes virus infection is considered as the primary cause of NK, resulting in the loss of sensory ganglion fibers and/or cells [23,[41][42][43].
In our study, Herpes virus keratitis was found as the primary local cause of NK, followed by diabetes, using data consistent with those found in the literature [23].
Although the Mackie classification has been used for several years, Dua et al. proposed another adaption that is more therapeutically useful and includes severity and prognosis information. This classification divides the extent of pathology into: mild (characterized by epithelial irregularity without a true epithelial defect), moderate (characterized by epithelial defect without stromal defect), and severe, which is characterized by stromal involvement ranging from corneal ulceration to lysis and perforation, accompanied by corneal hypo-anaesthesia [23]. In 2019, Mastropasqua et al. proposed another classification of NK based on the sub-basal nerve fiber layer assessed with in vivo confocal microscopy (IVCM) and the stromal structure, evaluated with AS-OCT.
In their report, the authors highlighted the role of the high-tech instruments in the assessment of the corneal structure in NK, however, without follow up and quantification of data [44].
In the present study, we measured the extension and depth of the corneal defects, and we were able to provide a statistical assessment of the morphological changes occurring in the corneal tissue as a response to the therapy.
The treatment milestone was achieved when recombinant human nerve growth factor (rhNGF) was introduced and after proved efficacy in clinical trials approved by EMA and FDA.
The efficacy and safety of treatment with rh-NGF were reported in clinicals where the corneal defects extension was assessed using fluorescein staining [55,56,58]. In our study, we report the morphological corneal changes evidenced with new generation instruments such as AS-OCT and K5M during the Cenegermin administration in patients with NK ulcers. Both devices provided us with detailed information on the healing processes, improving our knowledge on corneal changes induced by therapy.
The IVCM allowed us to demonstrate the nerve's regrowth after the treatment, confirming the effectiveness of rhNGF in promoting corneal reinnervation in NK. Similarly, Mastropasqua et al. confirmed the ability of Cenegermin to heal persistent epithelial defects and corneal ulcers in a group of patients with NK, assessing the corneal nerves with confocal investigation [59].
The importance of IVCM in monitoring the effects of corneal healing and nerves restoration after neurotization in NK was also reported by Fung et al. [60].
The presented results proved the possibility of qualitative and quantitative analysis of corneal defects, allowing the specialist to follow the changes during the treatment.
Additionally, the use of IVCM, even if limited, allowed us to prove the effect of the treatment on nerve restoration with the presence of new fibers in the acquired pictures.
However, this device cannot be used in all patients due to the frequent poor cooperation and invasiveness related to the contact approach in significantly affected eyes.
The use of IVCM in NK patients is of high importance, as it proves the restoration of corneal health, together with morphological improvement as assessed with Keratograph and AC-OCT.
In conclusion, in this study we demonstrated that the use of new technologies offers inestimable opportunities to accurately monitor corneal response to therapies in several diseases where deep stromal defects or different structural alterations might occur. These results are based on the use of high-tech in assessing corneal healing in severe NK with deep defects. We believe that this instrumental approach could be extended to the other corneal pathologies, i.e., infectious, degenerative, and dystrophic, providing qualitative and quantitative information on the corneal morphology.