Regeneration of Vocal Fold Mucosa Using Tissue-Engineered Structures with Oral Mucosal Cells

Abstract

Objectives: Scarred vocal folds result in irregular vibrations during phonation due to stiffness of the vocal fold mucosa. To date, a completely satisfactory corrective procedure has yet to be achieved. We hypothesize that a potential treatment option for this disease is to replace scarred vocal folds with organotypic mucosa. The purpose of this study is to regenerate vocal fold mucosa using a tissue-engineered structure with autologous oral mucosal cells.

Study design: Animal experiment using eight beagles (including three controls).

Methods: A 3 mm by 3 mm specimen of canine oral mucosa was surgically excised and divided into epithelial and subepithelial tissues. Epithelial cells and fibroblasts were isolated and cultured separately. The proliferated epithelial cells were co-cultured on oriented collagen gels containing the proliferated fibroblasts for an additional two weeks. The organotypic cultured tissues were transplanted to the mucosa-deficient vocal folds. Two months after transplantation, vocal fold vibrations and morphological characteristics were observed.

Results: A tissue-engineered vocal fold mucosa, consisting of stratified epithelium and lamina propria, was successfully fabricated to closely resemble the normal layered vocal fold mucosa. Laryngeal stroboscopy revealed regular but slightly small mucosal waves at the transplanted site. Immunohistochemically, stratified epithelium expressed cytokeratin, and the distributed cells in the lamina propria expressed vimentin. Elastic Van Gieson staining revealed a decreased number of elastic fibers in the lamina propria of the transplanted site.

Conclusion: The fabricated mucosa with autologous oral mucosal cells successfully restored the vocal fold mucosa. This reconstruction technique could offer substantial clinical advantages for treating intractable diseases such as scarring of the vocal folds.

Fig 1. Fabrication protocol of tissue-engineered vocal fold mucosa.

A 3 mm by 3 mm specimen of buccal mucosal tissue was excised. The mucosal section was divided into the epithelium and subepithelium. Isolated epithelial cells from the epithelium and isolated fibroblasts from the subepithelium were separately cultured for a two-week culture period. Following this, using cell-culture inserts, the proliferated fibroblasts were mixed into concentrated type I collagen gel on an oriented collagen sheet and cultured for one week. The proliferated epithelial cells were seeded and co-cultured on the collagen gel containing fibroblasts. After an additional two-week co-culture period, fabrication of the organotypic cultured tissue with stratified epithelial cells was complete.

Fig 2. Schematic view of the extent of vocal fold resection.

Vocal fold mucosa was resected to the full depth of the lamina propria in the coronal view (A). The vocal muscle was preserved to the extent possible. The superior to inferior dimensions of injury were about one-third the thickness of the vocal fold. The resection extended from the vocal process to the anterior commissure in the horizontal view (B). Line a: median approach by laryngofissure.

Fig 3. Diagram of stroboscopic observation of removed larynx.

Glottal closure was achieved at the posterior portion by suturing the bilateral cartilaginous portion of the vocal folds of the excised larynges. A contact microphone was attached firmly to the excised larynges and connected to the laryngostroboscope. Experimental phonation was artificially induced by blowing air through the trachea. The upper surfaces of the vocal folds were illuminated by the stroboscopic light from the laryngostroboscope via a flexible endoscope. Vocal fold vibrations were recorded by a video processor.

Fig 4. Tissue-engineered vocal fold mucosa in vitro.

A tissue-engineered vocal fold mucosa was successfully fabricated (20 mm diameter).

Fig 5. Scanning electron microscopy of tissue-engineered vocal fold mucosa.

Microvilli developed on the epithelium of the tissue.

Fig 6. Evaluation of an organotypic cultured vocal fold.

Hematoxylin and Eosin staining revealed that the tissue-engineered vocal fold mucosa consisted of the epithelium and lamina propria. The epithelium comprised two to four cell layers (A). Cells in the upper layer of the tissue-engineered vocal fold mucosa immunohistochemically expressed cytokeratin (B). Cells in the lower layer of the tissue-engineered vocal fold mucosa immunohistochemically expressed vimentin (C).

Fig 7. Transplantation of an organotypic cultured vocal fold to the mucosa-deficient vocal fold.

After laryngofissure was performed, the unilateral membranous portion of the vocal fold was resected (A) to the extent shown in Fig 2. The organotypic cultured vocal fold was transplanted to the mucosa-deficient vocal fold (B). TC: median cut surface of thyroid cartilage, CC: cricoid cartilage, Tr: trachea, ET: endotracheal tube.

Fig 8. Stroboscopic findings.

Mucosal waves on the transplanted side (right vocal fold) were regular but slightly smaller than those on the normal side (left vocal fold) (A). Mucosal waves on the control side (right vocal fold) in the controls were few (B).

Fig 9. Whole organ section of a removed larynx (H&E stain).

The regenerated vocal fold mucosa on the transplanted portion (B) was morphologically similar to that of the normal side (A). The vocal fold mucosa on the control side (C) was morphologically atrophic (B).

Fig 10. Histological comparison between normal (A, D, E), transplanted (B, E, H), and control (C, F, I) portions of the removed larynges.

Epithelial cells were not immunohistochemically stained with anti-cytokeratin 8 antibodies in normal (A) or control (C) portions, but stained positively on transplanted portions (B). The stratified epithelium on both normal and transplanted portions consisted of five to seven uniform cell layers, but differed from the control portion, which consisted of approximately ten uniform cell layers. Cells stained with vimentin, including fibroblasts, were identified in equal density in the lamina propria on every portion (D, E, F). Vascular endothelial cells stained with vimentin were identified in greater number in the lamina propria on normal (D) and regenerated (E) portions than on control (F) portions. EVG staining revealed both collagenous and elastic fibers on transplanted portions (H), with denser collagenous fibers than those on normal portions (G). Scar tissue formation and proliferation of collagenous fibers were observed on control portions (I).