Schematic drawing of tongue resection model.
Hematoxylin-eosin–stained sagittal sections of the control (A) and operated-on (B-D) sides of tongues (original magnification ×40). The isotonic sodium chloride solution (saline) group (B) shows gross atrophy and loss of tongue volume. The hydrogel group (C) exhibits preservation of tongue volume. The hydrogel-myoblast composite group (D) also has preserved volume with islands and cords of eosinophilic cells, which is more obvious at higher magnification in Figure 3. The arrow marks a persistent island of gel. Groups are described in the "Methods" section.
Hematoxylin-eosin–stained sections from the hydrogel-myoblast composite group at original magnification of ×100 (A), ×200 (B and C), and ×400 (D). The islands and cords of lighter pink cells (arrows in C) and stroma interdigitating with the darker red tongue muscle represent the new skeletal muscle. Groups are described in the "Methods" section.
Hematoxylin-eosin–stained sections of the isotonic sodium chloride solution group at original magnification of ×100 (B), ×200 (A and C), and ×400 (D). Disorganization of myofibril architecture with atropy and scarring (arrows in A and C) and multinucleate giant cells (arrows in C and D) are seen. Groups are described in the "Methods" section.
Hematoxylin-eosin–stained sections of the hydrogel group at original magnification of ×100 (A), ×200 (B and C), and ×400 (D). Adjacent to native tongue muscle (m) are persistent islands of gel (g) in (B-D). Preservation of tongue volume with no evidence of new tissue growth is seen. Groups are described in the "Methods" section.
Kim J, Hadlock T, Cheney M, Varvares M, Marler J. Muscle Tissue Engineering for Partial Glossectomy Defects. Arch Facial Plast Surg. 2003;5(5):403-407. doi:10.1001/archfaci.5.5.403
From the Divisions of Facial Plastic and Reconstructive Surgery (Drs Kim, Hadlock, and Cheney) and Head and Neck Surgical Oncology (Dr Varvares), Massachusetts Eye and Ear Infirmary, and the Department of Surgery, Children's Hospital Boston (Dr Marler), Boston, Mass. Dr Kim is now with the Department of Otolaryngology, University of Michigan, Ann Arbor.
Copyright 2003 American Medical Association. All Rights Reserved.
Applicable FARS/DFARS Restrictions Apply to Government Use.2003
Background Tongue reconstruction represents a difficult reconstructive problem, based on the tongue's complex multimodality function. Existing methods of tongue reconstruction often result in significant deficits in speech and deglutition. A functional neotongue requires adequate soft tissue bulk and restoration of coordinated muscle function. Tissue engineering, a scientific approach that allows introduction of isolated cell populations of interest within 3-dimensional polymer scaffolds to create new tissue, may allow the generation of more highly functional tissue in tongue reconstruction.
Objectives To apply muscle-tissue engineering techniques in the reconstruction of partial glossectomy defects in rats, and to compare the gross and histological nature of tissue found after reconstruction of the hemiglossectomy defect with acellular vs tissue-engineered composite material.
Materials and Methods Thirty mature Lewis rats underwent a left-sided mucosa-sparing partial glossectomy. The defects were then filled with 1 of the following 3 substances: isotonic sodium chloride solution, a collagen-rich hydrogel, or hydrogel containing a suspension of neonatal myoblasts from syngeneic rats. The animals were killed after 6 weeks and the tongues were harvested. The control and operated-on tongue halves were evaluated for weight differences and histological features.
Results The group receiving the hydrogel-myoblast composite injections demonstrated a statistically significant increase in tongue weight of the operated-on side compared with the control side. In contrast, the isotonic sodium chloride solution and hydrogel groups demonstrated loss of tongue weight. These findings correlated with the results of the histological evaluation. Hemitongues from the composite group demonstrated formation of new tissue with areas of musclelike tissue extending from islands of residual hydrogel, and we found evidence of neovascularization and possible neurotization. In contrast, the isotonic sodium chloride solution group exhibited dense fibrous scar with loss of muscle architecture and dramatic loss of tongue volume. The hydrogel group demonstrated preservation of tongue volume with persistent islands of gel, but no clear evidence of new tissue formation.
Conclusions The introduction of a hydrogel into the rat hemiglossectomy pocket appears to promote volume preservation and/or muscle regeneration. The addition of myoblasts suspended in collagen gel supports the development of new tissue that preserves weight and volume after hemiglossectomy and may possess muscle properties similar to the tissue desired. This tissue-engineering approach represents a promising new strategy in tongue reconstruction and merits further investigation into the possible functional advantages it offers compared with current techniques.
THE TONGUE is a multifunctional organ with complex anatomy. It plays a critical role in articulation, airway protection, and deglutition. The complexity of tongue anatomy makes duplication of its form and function a challenge. Because it is the most common subsite of cancer of the oral cavity, the treatment of which often requires significant tissue resection, a method of preserving or restoring function after resection would be extremely useful. Because to date it has not been possible to replace the complex musculature dynamics of the tongue, treatment is often aimed at preserving the remaining mobility in the residual tongue after partial glossectomy.1
Of great significance is the volume and location of tongue resection. Many surgeons categorize defects according to mobile tongue vs base of the tongue because of the different functions that the 2 regions serve and the extent of resection, to guide them in treatment options.1 Urken et al1 presented a classification scheme for tongue defects with optimal reconstructive methods for each type. The use of thin, pliable sensate cutaneous flaps while maintaining the mobility of the remaining tongue has been the goal for the reconstruction of oral tongue defects, whereas the goal of reconstruction of base of tongue defects is the achievement of tissue bulk and sensation.1
The prevailing reconstructive option for significant glossectomy defects involves free tissue transfer. Radial forearm,2 gracilis,3 rectus,4 and latissmus5 flaps have all been used. These tissues provide bulk and the potential for sensory or motor innervation. Nevertheless, innervated free flaps do not ideally restore the form or the function of the native tongue. Most flaps lose bulk owing to atrophy or lose placement owing to gravity and at best provide weak unidirectional movement.6 As a result, patients live with suboptimal speech, airway protection, and swallowing.
Tongue transplantation has been explored as another reconstructive option. Haughey et al7 performed microneurovascular allotransplantation of the canine tongue in 10 animals. Although only half of the animals achieved long-term survival, clinical recovery of tongue function was observed, and results of histological studies demonstrated reinnervation of the hypoglossal and lingual nerves. Results of light microscopy also showed preservation of muscle, mucosal, and stromal ultrastructure of the transplanted tongue. This strategy, however, has not been applied in humans because of the obvious immunologic issues it poses.
Tissue engineering involves combining polymer scaffolds with isolated cell populations to create new tissue for repair or replacement, with the ultimate goal of restoring form and function.8 It remains an attractive alternative to transplantation in the reconstruction of tongue defects, because syngeneic cells can be used in animal studies, and ultimately, autologous cells expanded in culture would provide the cell source for human application. The goals of this study were (1) to apply tissue-engineering techniques toward reconstruction of partial glossectomy defects in the rat and (2) to compare the gross and histological nature of tissue found after reconstruction of the hemiglossectomy defect with acellular vs tissue-engineered composite material.
We followed institutional guidelines regarding animal experimentation.
The hydrogel vehicle was prepared by adding 1.7-mg/mL collagen (Cellagen; ICN, Mississauga, Ont) in Hams F10 medium (Sigma-Aldrich Corp, St Louis, Mo) and Matrigel (1:6 vol/vol; Collaborative Biomedical Products, Bedford, Mass) to 3-ng/mL basic fibroblast growth factor neutralized to a pH of 7.1 with sterile 7.5% sodium bicarbonate. The gel was warmed in a water bath to 37°C before injection to obtain the appropriate viscosity.
Paraspinal and extremity muscles were dissected from neonatal Lewis rats, placed in chilled calcification- and magnesium-free phosphate-buffered saline (Sigma-Aldrich Corp), and minced. Tissue pieces were digested by incubation in a collagenase-dispase solution (1% type II collagenase [Worthington Biochemical Corp, Lakewood, NJ], 2.4-U/mL dispase [Boehringer Mannheim, Indianapolis, Ind], and 2.5-mM calcium chloride]) for 30 minutes at 37°C, triturated 10 times each with 25-, 10-, and 5-mL pipettes, and strained through a 70-µm filter.
Myoblast medium (consisting of Hams F-10 nutrient mixture [Sigma-Aldrich Corp], 20% fetal bovine serum [Sigma-Aldrich Corp], 3-ng/mL basic fibroblast growth factor [Promega Corp, Madison, Wis], and a combination of 1% glutamine, penicillin, and streptomycin [Sigma-Aldrich Corp]) was added to the resulting cell suspension. Cells were spun at 1200 rpm for 5 minutes and transferred to plates coated with laminin (Sigma-Aldrich Corp), 5 µg/mL, in phosphate-buffered saline for 24 hours at 4°C. After 24 hours, cells were enzymatically dissociated with 0.05% (weight-volume ratio) calcification- and magnesium-free trypsin (Sigma-Aldrich Corp), rinsed with phosphate-buffered saline, and added to chilled collagen gel at a concentration of 3 to 5 million cells/mL.
Thirty male 200- to 300-g Lewis rats were anesthetized with an intramuscular injection of ketamine hydrochloride and metomadine. We then injected 0.1 mL of 1% lidocaine hydochloride with 1:100 000 epinephrine bitartrate into the left hemitongue for local vasoconstriction. All rats underwent a left-sided subtotal oral hemiglossectomy via an incision along the left lateral tongue, under microsurgical magnification. The tongue muscle was sharply dissected while maintaining the mucosa intact. Resection of the tongue musculature extended medially to the midline raphe, and posteriorly to the anterior margin of the circumvallate papillae.
The incision was closed with a running locking 7-0 polyglactin 910 (Vicryl; Ethicon, Inc, Somerville, NJ) stitch. The enucleated cavity was then filled with 1 of the following 3 substances using a 20-gauge needle: 0.3 mL of isotonic sodium chloride solution, hydrogel, or hydrogel-myoblast composite material. Postoperatively, the animals were fed food paste and monitored for weight gain.
Animals were killed at 6 weeks and perfused with 4% paraformaldehyde. The tongues were harvested by means of transection at the circumvallate papillae (Figure 1). Each tongue was blotted dry, weighed wet, and then bisected along the midline raphe. The 2 halves were identified as the control or the operated-on side. Each half was then weighed separately and recorded. We performed statistical analysis using a 2-tailed t test to compare weight data.
Histological sections were obtained centrally and along the longitudinal axis of the hemitongue. The sections were stained with hematoxylin-eosin and Masson trichrome and processed for routine microscopy. Desmin staining was also performed.
Control and experimental tongue weights for the 3 groups are presented in Table 1. The group receiving the cell-containing hydrogel demonstrated a statistically significant increase in tongue weight of the operated-on side compared with the control side (P = .03). In contrast, the isotonic sodium chloride solution (P = .006) and hydrogel groups (P = .006) demonstrated loss of tongue weight.
These findings correlated with results of histological evaluation (Figure 2). Tongues reconstructed with the tissue-engineered composite demonstrated formation of new tissue with islands and cords of eosinophilic cells interdigitating with existing muscle architecture (Figure 3). Immunohistochemical evidence of these areas being skeletal muscle cells is supported by positive results of desmin staining. In contrast, the isotonic sodium chloride solution group exhibited dense fibrous scar with disorganization of myofibril architecture and dramatic loss of tongue volume (Figure 4). The hydrogel group demonstrated preservation of tongue volume with persistent islands of gel, but no clear evidence of new tissue growth as seen in the hydrogel-myoblast composite group (Figure 5).
Previous investigators were able to cultivate skeletal muscle using isolated myoblasts attached to synthetic biodegradable polymer scaffolding for tissue replacement in the enhancement of muscle regeneration.9 In another tissue engineering study, Marler et al10 looked at soft tissue augmentation with an injectable alginate gel polymer with and without suspended syngeneic fibroblasts in rats. Our study used both concepts to cultivate skeletal muscle by suspending myoblasts in an injectable hydrogel-polymer composite.
We have successfully developed an animal model for reconstructing a partial glossectomy defect in the rat. Consistent anatomic landmarks in the circumvallate papillae and midline raphe allow the creation of reproducible defects for study. We have shown that rats are able to survive after the surgical insult and tolerate the defect while maintaining appropriate body weight.
The statistically significant increase in tongue weight in the rats receiving tissue-engineered composite material provides solid evidence of new tissue growth. Results of the histological examination revealed formation of segments of new muscle tissue with adjacent neovascularization. Desmin staining for skeletal muscle was positive in the area of new tissue formation, establishing the myogenic origin of the new tissue. However, to confirm that the new tissue is from transferred myoblasts, future labeling studies need to be performed. In contrast, we found dense scarring and loss of tongue volume and weight in the isotonic sodium chloride solution group. The hydrogel group exhibited preservation of tongue volume on results of the histological examination, with intact islands of gel surrounded by blood vessels and nerves. However, we found a decrease in tongue weight with no solid evidence of new tissue formation.
Of parallel importance is the ultimate function of the neotongue formed. This study does not examine function of this new tissue; however, it demonstrates the formation of muscle fibers and sets the groundwork for future studies to examine function of the newly formed tissue. It will be necessary to quantify the contractility of the muscle tissue, to promote tissue organization in such a way as to permit the generation of directed force and organized neural control. Assessment of electrophysiologic and behavioral function of the neotongue is also an immediate future goal.
The introduction of a hydrogel into the rat hemiglossectomy pocket promotes volume preservation and/or muscle regeneration. The combination of myoblasts and collagen gel in a tissue-engineered composite supports the development of new tissue, which preserves weight and volume after hemiglossectomy.
Myoblast transplantation via injectable gel suspension to the rat tongue is a promising experimental approach in the reconstruction of glossectomy defects and merits investigation into neotissue functionality.
Corresponding author and reprints: Jennifer Kim, MD, Department of Otolaryngology, 1904 Taubman Center, 1500 E Medical Center Dr, Ann Arbor, MI 48109-0312 (e-mail: email@example.com).
Accepted for publication October 2, 2002.
This study was supported by the Laurence Murphy Cancer Research Fund at the Massachusetts Eye and Ear Infirmary, Boston, Mass.
This study was presented at the Fall Meeting of the American Academy of Facial Plastic and Reconstructive Surgery; September 7, 2002; Denver, Colo.