To investigate the possibility of replacing the metaplastic ocular surface with nasal mucosa, and to evaluate the results of autologous nasal and oral mucosal transplantation in cicatricial ocular surface diseases.
Retrospective interventional case series.
We studied 6 eyes in 6 patients with chemical burns, which were characterized by a cicatricial ocular surface. After removal of cicatricial tissues and symblepharolysis, autologous nasal mucosa was transplanted in all patients. In 3 patients with extensive damage, oral mucosal autografting was performed concurrently. The nasal and oral mucosa was evaluated using immunohistochemical analysis for p63, K3, MUC5AC, and CD34. Clinical outcomes were assessed based on visual acuity, ocular manifestations, and liquid-based cytology.
Immunohistochemical analysis revealed a plentitude of p63 and K3 in nasal mucosal epithelium. Goblet cells and MUC5AC expression were only observed in nasal mucosal epithelium, not in oral mucosal epithelium. Well-developed parallel vasculature was demonstrated in the nasal mucosa. In contrast, perpendicular vasculature was demonstrated in the oral mucosa. This vascular feature remained after transplantations. In all patients, ocular surface stability recovered with no major complications and increased goblet cells were observed on ocular surface. However, delayed epithelialization and ischemic thinning were seen at oral mucosal graft sites.
Nasal mucosa, which has the advantage of well-developed parallel vasculature, enriched goblet cells, and plenty of stem cells, may be an ideal substitute for a cicatricial ocular surface. Transplantation of autologous nasal mucosa is a very effective method for achieving ocular surface reconstruction in cicatricial ocular surface diseases.
Cicatricial ocular surface diseases such as Stevens-Johnson syndrome, ocular cicatricial pemphigoid, and chemical/thermal burns are very difficult and challenging disorders for ophthalmologists. These chronic and excessive inflammatory diseases induce destruction of corneal epithelial stem cells located in the limbus, loss of conjunctival goblet cells producing mucins, and ultimately squamous metaplastic epithelium-like skin. These changes cause severe visual loss, ocular pain, limitations in eyeball movement, and cosmetic problems. Regarding the treatment of these diseases, ocular surface reconstruction may be required to replace the cicatricial tissues with stem cells or a stem cell niche and conjunctival goblet cells.
Normal, healthy ocular surface tissues of the fellow eye, such as conjunctiva or limbus, may be the best source of ocular surface reconstruction materials. However, in bilateral disease, this approach is not an option. Furthermore, harvesting of the fellow eye is restrictive and often provides an insufficient quantity of tissue. Therefore, amniotic membrane (AM) and various autologous mucosal tissues, such as oral mucosa, hard palate mucosa, maxillary sinus mucosa, or nasal mucosa have been used for ocular surface reconstruction. With the recent development of tissue engineering techniques, methods using autologous cultivation of corneal epithelial cells, limbal cells, or oral mucosal epithelial cells have come into the spotlight. However, tissue engineering techniques are possible in only some restricted hospitals, because these techniques are fastidious and require much effort and expense. Hence, nasal mucosa and oral mucosa, which can be easily accessed and obtained, may be practical tissue sources for clinicians. Especially, nasal mucosa has enriched goblet cells in the epithelium and submucosa, which secrete the mucin required for tear film stabilization. In 1990, Naumann and associates introduced autologous nasal mucosa transplantation in severe conjunctival mucus deficiency syndrome. After that, there were a few reports in which the nasal mucosa was used in eyelid reconstructive surgeries. However, nasal mucosal transplantation has not been used broadly in ocular surface reconstruction as of yet because most ophthalmologists hesitate to harvest the nasal mucosa. There have been no studies assessing the immunohistochemical characteristics of the nasal mucosa and comparing them to those of the ocular surface.
Therefore, we investigated the possibility of replacing the metaplastic ocular surface with autologous nasal mucosa by analyzing the histologic and immunohistochemical characteristics of the nasal mucosa. We also evaluated and compared the results of ocular surface reconstruction with autologous nasal mucosa and oral mucosa in cicatricial ocular surface diseases, especially ocular chemical burns.
We obtained informed consent from 6 patients (6 eyes) with chemical burns. They had all undergone prior ocular surgery with AMs more than twice. Nevertheless, they all had devastating cicatricial squamous metaplasia, recurrent corneal ulcers, total limbal deficiency, conjunctivalization, fibrovascular proliferation, and symblepharon. Hence, we performed ocular surface reconstruction using AM transplantation, nasal mucosal autografting, oral mucosal autografting, and conjunctival limbal autografting on these 6 damaged eyes.
Preparation of Autologous Oral Mucosa and Nasal Mucosa
All patients were examined carefully before the operation by an otorhinolaryngologist in order to determine the state of the nasal and oral mucosa. After preparation of a 5% povidone-iodine solution for antiseptic purposes, nasal mucosal tissues, each measuring 2 cm 2 , were stripped in situ from the one-third portion to the two-thirds portion of the inferior nasal turbinate using a blade and scissors in all patients. After harvest, bleeding was controlled by bipolar cauterization under an endoscope. In 3 patients who exhibited entire cicatricial ocular surface change, and for whom ocular surface reconstruction using only AM and nasal mucosa was expected to be insufficient, oral mucosal tissues, each measuring 1 cm 2 , were taken from the lower inner lip mucosa. Nasal and oral submucosal connective tissues were removed and trimmed into thin, flat mucosal tissues 2 mm in thickness with scissors and a blade.
All surgical procedures were performed by one surgeon (J.C.K.) on a case-by-case basis, with each patient under retrobulbar anesthesia. All patients underwent symblepharolysis and excision of cicatricial and fibrovascular tissue. In 3 patients who had no vision, but who had cosmetic problems, corneal tattooing was performed using India ink (True Black Eye Liner; Biotouch Products, Walnut, California, USA). Thereafter, permanent AM grafting was performed on bare sclera and whole cornea in all patients using cryopreserved AM (AmniSite-Cornea; Bioland Ltd, Cheonan, Korea) with interrupted 10-0 nylon sutures. Nasal mucosal autografting (in 6 eyes) and oral mucosal autografting (in 3 eyes; Cases 1 to 3) were performed on sclera over the permanent AM grafting with interrupted 10-0 nylon sutures. Conjunctival limbal autografting was performed in the perilimbal area over the permanent AM grafting in 2 patients (Cases 2 and 5) using the method described by Kenyon and Tseng. In the 4 remaining patients, it was impossible to harvest limbal tissue from the fellow eye, because of previous harvesting of conjunctivolimbal tissue or because of chemical burn–induced damage to the ocular surface. Lastly, temporary AM patches were placed in all patients. After surgery, 20% autologous serum eye drops were instilled every 2 hours, 0.5% levofloxacin (Levaquin; Ortho-McNeil Pharmaceutical, Titusville, New Jersey, USA) was used topically 4 times a day, and topical steroid ointment (dexamethasone [Decadron; Merck & Co Inc, Whitehouse Station, New Jersey, USA] polymyxin B sulfate, and neomycin sulfate) was administered 3 times daily until epithelialization was complete. During the first week after surgery, prednisolone (30 mg per day) and third generation cephalosporins (cefditoren pivoxil; [Spectracef; Purdue Pharmaceutical Products, Stamford, Connecticut, USA] 300 mg per day) were administered orally. Five days after surgery, temporary AM patches were removed and preservative-free artificial tears were used.
Preoperative and postoperative best-corrected visual acuity (BCVA) and intraocular pressure were measured. Ocular surface manifestations, the presence of corneal erosion, corneal neovascularization, the recurrence of symblepharon, epithelial integrity, and survival of the mucosal graft were evaluated at the first day, third day, first week, second week, fourth week, and every month postoperatively with a slit-lamp microscope (SL 130; Carlzeiss Meditec Inc., Dublin, California, USA) and fluorescein staining. Clinical success was defined as intact ocular surface epithelium without epithelial erosion, maintained for greater than 4 weeks.
Histologic and Immunohistochemical Analysis
Hematoxylin-eosin staining, periodic acid–Schiff (PAS) staining, and immunohistochemical staining for p63, keratin 3 (K3), CD34, and MUC5AC were performed to evaluate the epithelial characteristics in the remnant nasal mucosa and oral mucosa of Case 1, after the mucosa was trimmed. Monoclonal antibodies against K3, p63, CD34, and MUC5AC were purchased from Santa Cruz Biotechnology (Santa Cruz, California, USA). The mucosal tissues were fixed with 2% paraformaldehyde in phosphate-buffered saline (PBS) at 4 C for 10 minutes and then permeabilized with 0.2% Triton X-100 in PBS at room temperature for 10 minutes. The endogenous peroxidase was quenched with 0.3% H 2 O 2 in 0.5% horse serum in PBS and incubated with 5% horse serum to block the nonspecific sites. Monoclonal antibodies against p63, K3, CD34, or MUC5AC were applied and incubated for 1 hour at room temperature, followed by incubation with biotinylated second antibodies, anti-mouse, or anti-rabbit IgG, using a Vectastain Elite ABC Kit (PK-6101; Vector Laboratories, Burlingame, California, USA), according to the manufacturer’s protocol. Samples were finally incubated with diaminobenzidine (DAB) peroxidase substrate to give a brown stain and were then counterstained with hematoxylin. After washing with PBS, the samples were mounted and analyzed.
Liquid-based cytology was performed to evaluate for the presence of goblet cells, before surgery and 1 year after surgery, in Cases 4 and 5. Each patient turned his head to the operative eye in the supine position. After that, 1 ml of balanced salt solution (BSS; Alcon Laboratories, Fort Worth, Texas, USA) was dropped from the medial bulbar conjunctiva to the lateral canthus and collected in an Eppendorf tube. A commercially manufactured fixative (CytoRich Red; TriPath Imaging Inc, Burlington, North Carolina, USA) was added to the collected specimen and vortex-mixed for 15 minutes. After the specimen sat for 30 minutes, it was centrifuged at 600 g for 10 minutes; the supernatant was then decanted. As for washing, 10 ml of water was added to the specimen, which was then vortexed. The specimen was then centrifuged at 600 g for 5 minutes, and the supernatant was decanted again. After the specimen was loaded onto a PrepStain system (TriPath Imaging Inc) for processing, a uniform thin layer of cells was produced in a 13-mm-diameter circle. Interpretations were made by applying fundamental cytologic guidelines practiced by pathologists.
Histologic and Immunohistochemical Analysis
PAS staining and MUC5AC immunostaining demonstrated goblet cells containing mucin only in the nasal mucosal epithelium, but not in the oral mucosal epithelium ( Figure 1 ). K3 was strongly expressed in the nasal mucosal epithelium, similar to cornea and conjunctiva. Furthermore, P63, a stem cell marker, was expressed throughout the whole basal layer of the nasal mucosal epithelium ( Figure 2 ). Flat-mount immunostaining for CD34, which was expressed in vascular endothelial cells, showed a longish pattern representing the parallel vascular network to the epithelium in the nasal mucosa. However, the oral mucosa stained with CD34 showed a mainly round-dot vascular pattern (though a partially longish pattern was found) ( Figure 3 ). This implies that the oral mucosa has a perpendicular vascular network to the epithelium, as contrasted with the parallel vasculature of the nasal mucosa. These vascular features remained at 1-year after transplantation of nasal or oral mucosa in all patients ( Figure 3 ).
The mean follow-up period in our study was 14.7 months (range, 12 to 18 months). Nasal and oral mucosal tissues were safely excised without any complications, and their mucosal harvest sites were completely healed in 2 weeks. In all patients, the ocular surface was stabilized, and there were no recurrences of ulceration, conjunctivalization, fibrovascular tissue, or symblepharon ( Figure 4 ). Clinical success was achieved without any major postoperative complications. Eyeball movement was improved in all patients. All patients noted improvement in ocular comfort, and 3 patients with corneal tattooing were very satisfied with their cosmetic results. We noted rapid epithelial healing, well-developed horizontal vascularization, and abundant mucin in all nasal mucosal graft sites. The existence of mucin and goblet cells was confirmed by liquid-based cytology. Before surgery, only squamous cells and inflammatory cells without any goblet cells were observed on liquid-based cytology. However, 1- year after surgery, several goblet cells were easily found ( Figure 5 ). All transplanted conjunctival limbal tissues around the nasal mucosal graft were well healed and living. However, poor, delayed epithelialization and pale ischemic thinning were found at the oral mucosal graft sites ( Figure 6 ). Preoperative BCVAs were no light perception (3 eyes), light perception (1 eye), and 20/1000 (2 eyes). Postoperative BCVAs had improved in 3 eyes (excluding the 3 eyes that underwent corneal tattooing) at the last follow-up examination ( Table ).