Purpose
To study the corneal nerves in patients with chronic bullous keratopathy.
Design
Prospective observational case series with histologic evaluation.
Methods
We studied 25 eyes of 25 bullous keratopathy patients of different etiologies (17 female, 8 male; mean age, 76.3 years) as well as 6 eyes of 6 normal control subjects (5 male, 1 female; mean age, 38 years). All subjects were scanned by laser scanning confocal microscope. Five corneal buttons obtained following penetrating keratoplasty from 5 of the above patients and 6 normal control corneal buttons were stained as whole mounts with acetylcholinesterase (AChE) method for corneal nerve demonstration and scanned in multiple layers with digital pathology scanning microscope.
Results
The density, branching pattern, and diameter of sub-basal nerves were significantly lower in corneas with bullous keratopathy compared with normal corneas (density: 4.42 ± 1.91 mm/mm 2 vs 20.05 ± 4.24 mm/mm 2 ; branching pattern: 36.02% ± 26.57% vs 70.79% ± 10.53%; diameter: 3.07 ± 0.64 μm vs 4.57 ± 1.12 μm). Aberrations such as localized thickenings or excrescences, abnormal twisting, coiling, and looping of the (mid) stromal nerves were observed in the study group both by in vivo confocal microscopy and on histology.
Conclusions
Striking alterations in corneal innervation are present in corneas with bullous keratopathy that are unrelated to any specific etiology of bullous keratopathy. This study provides histologic confirmation of novel in vivo confocal microscopy findings related to corneal nerves in bullous keratopathy.
Bullous keratopathy is a common clinical endpoint of corneal endothelial dysfunction or damage resulting from a variety of causes. Common conditions associated with bullous keratopathy include cataract extraction with or without intraocular lens implantation, primary or secondary (immune rejection) corneal graft failure, absolute glaucoma, and endothelial dystrophies such as Fuchs endothelial corneal dystrophy (Fuchs dystrophy). Unfortunately, bullous keratopathy of iatrogenic origin accounts for about 60% of cases. The 3 most common causes of bullous keratopathy are pseudophakic bullous keratopathy, aphakic bullous keratopathy, and Fuchs dystrophy. Currently, bullous keratopathy is the leading indication of penetrating keratoplasty/endothelial transplant and regraft. Initially, the corneal stroma becomes edematous and thickened and eventually intraepithelial and subepithelial fluid-filled vesicles, “bullae,” appear.
Clinical features can vary from asymptomatic early disease, glare, and painless decrease in vision to profound loss of vision attributable to subepithelial scarring in advanced cases. Painful episodes associated with photophobia and tearing are attributed to nerve stretching and irritation by epithelial and subepithelial bullae and to rupture of surface bullae with exposure of nerve endings.
The histologic features of bullous keratopathy include intracellular epithelial edema and bullous separation, stromal thickening, and endothelial loss. In addition, bullous keratopathy secondary to Fuchs dystrophy also shows Descemet membrane thickening with posterior guttata. With the advent of in vivo confocal microscopy our understanding of the pathologic findings in bullous keratopathy has been enhanced by our ability to examine the tissue in vivo, thereby avoiding artifacts resulting from tissue handling and processing.
Although pain, attributed to nerve stretching and exposure as stated above, is a common manifestation and is often intractable, there is very little information on the state of the corneal nerves in bullous keratopathy. Routine histologic techniques and transverse sections of corneal tissue are not conducive to a proper evaluation of the nerves. Laser scanning confocal microscopy has provided new insights into the orientation and distribution of human corneal nerves in health and disease. However, it is not always clear what structures equate to the in vivo confocal findings. Studies on direct correlation of confocal microscopy findings with histology of the examined tissue are few. We used this mode of examination to assess the corneal nerves in corneas with bullous keratopathy, including some that were scheduled for penetrating keratoplasty (PKP). We were able to examine whole mounts of corneal buttons removed at PKP with a special nerve stain to delineate corneal nerves with a view to correlate the in vivo confocal microscopy findings with histologic observations. We also intended to elucidate any nerve-related anatomic basis for pain, which is a dominant feature of advanced bullous keratopathy.
Materials and Methods
A prospective consecutive case series of 25 eyes of 25 patients with bullous keratopathy were examined by in vivo confocal microscopy. Patient demographics and clinical data are given in Table 1 . Fuchs dystrophy and pseudophakic bullous keratopathy accounted for 88% of the cases. There were 17 female and 8 male patients with a mean age of 76.3 ± 7.3 (58-88) years. All patients were white. All the patients were treated with corneal transplant surgery. Twelve of 25 patients (48%) had penetrating keratoplasty alone, 9 of 25 (36%) had triple procedure (penetrating keratoplasty and lens extraction with implant) and 4 of 25 (16%) underwent Descemet stripping endothelial keratoplasty (DSEK). Five host corneal buttons were available for histology. Whole mounts were stained for corneal nerve demonstration using acetylcholinesterase technique and examined by light microscopy.
| Patient | Sex | Age | Primary Diagnosis | Surgery | Method of Examination | Duration of Disease (Months) | CCT (μm) | Detection of the Sub-basal Nerve by IVCM | Abnormal Stromal Nerves by IVCM |
|---|---|---|---|---|---|---|---|---|---|
| 1 | F | 78 | ABK | Triple | IVCM+histology | 9 | 700 | − | + |
| 2 | F | 81 | PBK | PKP | IVCM+histology | 34 | 628 | − | + |
| 3 | F | 88 | PBK | Triple | IVCM+histology | 9 | 691 | − | − |
| 4 | M | 74 | PBK | Triple | IVCM+histology | 14 | 650 | + | − |
| 5 | M | 74 | FED | Triple | IVCM | 31 | 598 | + | − |
| 6 | F | 73 | FED | Triple | IVCM | 48 | 760 | + | − |
| 7 | F | 78 | FED | PKP | IVCM | 7 | 750 | + | − |
| 8 | F | 72 | FED | Triple | IVCM | 24 | 642 | + | + |
| 9 | F | 76 | FED | DSEK | IVCM | 11 | 776 | + | − |
| 10 | F | 78 | FED | PKP | IVCM | 4 | 760 | − | + |
| 11 | M | 79 | FED | PKP | IVCM | 8 | 588 | − | − |
| 12 | F | 58 | FED | PKP | IVCM | 14 | 730 | + | − |
| 13 | F | 82 | FED | DSEK | IVCM | 8 | 760 | + | − |
| 14 | F | 71 | FED | PKP | IVCM | 19 | 613 | + | + |
| 15 | F | 82 | FED | PKP | IVCM | 48 | 570 | − | − |
| 16 | F | 84 | Idiopathic | Triple | IVCM | 5 | 737 | + | + |
| 17 | M | 83 | FED | Triple | IVCM | 7 | 624 | + | + |
| 18 | F | 70 | PBK | PKP | IVCM | 10 | 688 | − | − |
| 19 | F | 74 | PBK | PKP | IVCM | 12 | 670 | + | − |
| 20 | F | 85 | Glaucoma | Triple | IVCM | 11 | 749 | + | − |
| 21 | M | 71 | PBK | DSEK | IVCM | 6 | 766 | + | − |
| 22 | M | 86 | PBK | PKP | IVCM | 28 | 633 | − | + |
| 23 | M | 60 | PBK | PKP | IVCM | 18 | 663 | − | + |
| 24 | M | 72 | FED | PKP | IVCM+histology | 39 | 695 | − | + |
| 25 | F | 78 | PBK | DSEK | IVCM | 11 | 625 | − | − |
The diagnosis of bullous keratopathy was based on history, slit-lamp examination, and central corneal thickness (CCT) (mean ± SD, 682.6 ± 63.7 μm; range 570-776 μm). An ultrasonic pachymetry device (Tomey SP-3000, Pachymeter; Tomey Corporation, Nagoya, Japan) was used for the measurements of CCT. The average time between diagnosis of bullous keratopathy and in vivo confocal microscopy scan was 17.4 months (range 4-48 months). None of the patients had corneal vascularization.
Confocal Microscopy
All 25 eyes were examined preoperatively by laser scanning confocal microscope (Heidelberg Retina Tomograph II Rostock Corneal Module [RCM]; Heidelberg Engineering GmbH, Heidelberg, Germany). The device uses a class I diode laser (670-nm wavelength) with a 63× water-immersion lens (Olympus, Tokyo, Japan). The images obtained using this lens are 400 × 400 μm, and have 2- and 4-μm lateral resolution and optical depth resolution, respectively (provided by the manufacturer at http://www.accessdata.fda.gov/cdrh_docs/pdf4/K042742.pdf ). Image magnification on screen was 300×. In vivo confocal microscopy was performed under topical anesthesia with MINIMS oxybuprocaine hydrochloride 0.4% (Bausch & Lomb Ltd, Surrey, United Kingdom). A digital camera mounted on a side arm furnished a lateral view of the eye and objective lens to monitor the position of the objective lens on the surface of the eye. A drop of 0.2% polyacrylic gel (Viscotears liquid gel; Novartis Pharmaceuticals Ltd., Surrey, United Kingdom) was used as coupling medium between the contact cap and objective lens of the microscope.
Central and paracentral regions (approximately 7 × 7 mm) of the cornea were scanned through all the layers. Frames from sub-basal (beneath basal cells of corneal epithelium) and stromal layers containing nerves were selected for analysis. Standard quantitative descriptors for nerve studies were examined. These were nerve density, which is in mm/mm 2 ; branching pattern which is expressed as the percentage of nerve branches per total number of sub-basal nerve fibers within a single frame; and diameter of sub-basal nerves, in microns. The thickest region of each main sub-basal nerve within a single frame was selected for the thickness analysis and the average diameter of 3 measurements for each nerve was calculated. Qualitative morphologic evaluation of sub-basal and stromal nerves was also carried out.
Acetylcholinesterase Technique for the Demonstration of Corneal Nerves
In 5 patients where in vivo confocal microscopy examination of their corneas was performed preoperatively, their corneal buttons obtained after penetrating keratoplasty for bullous keratopathy were processed and stained as whole mounts for cholinesterase enzyme using the acetylcholinesterase (AChE) technique. The protocol of the staining method has been described previously. Briefly, corneal buttons were fixed in cold 4% formaldehyde (pH 7) for 4 hours and then rinsed overnight in phosphate-buffered saline (PBS). Specimens were incubated in the stock solution containing acetylthiocholine iodide as a substrate for 24 hours at 37°C. The acetylcholinesterase enzyme in the nerves reacted with acetylthiocholine iodide in the substrate to produce a brown coloration of the nerves. The color was intensified with a dilute solution of ammonium sulfide. Specimens were dehydrated by immersion in alcohol and cleared in xylene, as is standard for histologic preparation. The specimens were finally mounted between a slide and coverslip and scanned en face using a Hamamatsu NanoZoomer digital pathology (NDP) microscope system (Hamamatsu, Hamamatsu City, Japan). The corneal buttons were examined at 40× magnification in multiple layers from epithelium to endothelium at 10-μm intervals. The images were then stacked and merged to give a single, holistic, detailed anatomic view of the stained corneal nerves. Image analysis was carried out using the software provided by the manufacturer and the areas of interest were then selected, automatically scaled, and exported to JPEG format.
Control
Six normal eyes from 6 healthy subjects with no previous ocular problems or surgeries were selected as controls for in vivo confocal microscopy. There were 5 male and 1 female subjects with a mean age of 38 ± 10.4 (range 31-49). Although mean age of the controls was less than that of the study group, it has been shown in the literature that no correlation exists between age and sub-basal nerve parameters. Six fresh postmortem corneas donated with family consent from 3 deceased patients (2 male, 1 female; mean age 57.3) with no previous ocular pathology or surgery were also stained with the AChE technique and used as controls. Causes of death were metastatic prostate carcinoma, adenocarcinoma of the lung, and post-renal transplant sepsis.
Statistical Testing
Data were analyzed using an analysis tool pack for Microsoft Excel 2007 (Microsoft Corp., Redmond, Washington, USA) and SPSS 16.0 (SPSS Inc., Chicago, Illinois, USA). A P value of <.05 was taken as the threshold of statistical significance. The Student t test for independent samples was used to establish whether any differences in the sub-basal nerve density, branching pattern, and sub-basal and stromal nerve diameters between the normal subjects and patients with bullous keratopathy were significant. It was also used to test the difference in central corneal thickness in patients with and without demonstrable in vivo confocal microscopy findings. Mann-Whitney U test was used to test the difference in the duration of bullous keratopathy in patients with and without demonstrable in vivo confocal microscopy findings.
Results
Confocal Microscopy Findings
Sub-basal nerves
Sub-basal nerves were found in 14 of 25 cases (56%) and were absent in 11 cases (44%). In cases where sub-basal nerves were detected, the mean sub-basal nerve density was 4.42 ± 1.91 mm/mm 2 (range 1.13-7.81 mm/mm 2 ). This was significantly lower than the density in controls (20.05 ± 4.24 mm/mm 2 ) ( Figure 1 , Top left) ( P = .001). In addition, these nerves showed a decrease in number and in percentage of branching (branching pattern). Branching pattern of sub-basal nerve plexus in patients with bullous keratopathy (36.02% ± 26.57%) was significantly lower than that of the controls (70.79% ± 10.53%) ( P = .001).
Furthermore, these nerves appeared thinner and attenuated compared to the normal sub-basal nerves ( Figure 1 , Top right). The diameter of the main sub-basal nerves in patients with bullous keratopathy (3.07 ± 0.64 μm) was significantly lower than that of the controls (4.57 ± 1.12 μm) ( P = .001). Some nerves were abnormally tortuous and showed a bizarre orientation ( Figure 1 , Middle left).
Stromal nerves
On in vivo confocal microscopy examination, stromal nerves were seen in all patients studied but changes were observed in 40% (10 out of 25) of bullous keratopathy cases. These consisted of relatively thin, tortuous, and convoluted nerves present mainly at the mid stroma (305.34 ± 71.39 μm depth). Their mean diameter was 5.35 ± 1.3 μm (range 3.12-8.86) ( Figure 1 , Middle right, Bottom left, and Bottom right). Some larger stromal nerves showed localized thickenings or excrescences suggestive of early sprouting ( Figure 2 , Top left). At the site of nerve bifurcations, hyper-reflective expansions with ill-defined blurry edges were noted in the central cornea ( Figure 2 , Middle left). At several places the stromal nerves formed distinct coils or loops appearing as hyper-reflective lines surrounding dark areas within the corneal stroma ( Figure 1 , Middle right, Bottom left, and Bottom right; Figure 2 , Bottom left).
