J Oral Maxillofac Surg59:169-177, 2001

Tissue-Engineered Mucosa Graft forReconstruction of the Intraoral LiningAfter Freeing of the Tongue: A Clinical

and Immunohistologic StudyGunter Lauer, MD, DMD, PhD* and Ronald Schimming, MD, DMD†

Purpose: This article describes the use of tissue-engineered mucosal grafts instead of split-thicknessskin grafts after freeing of the tongue in patients who had previous resection of an oral squamous cellcarcinoma and initial primary wound closure.

Patients and Methods: Tissue-engineered mucosal grafts, up to 75 cm2 in size, were cultured frombiopsy specimens of the hard palate in 6 patients, starting 3 to 4 weeks before the operation. After freeingof the tongue, the engineered mucosa was implanted on the wound surface by using vaseline gauze ascarrier and fixed with an intraoral gauze wound dressing.

Results: A good glossoalveolar sulcus was formed in 5 patients, resulting in good mobility of the tongueand a satisfactory denture-bearing surface. In 1 patient, there was a disturbance of wound healing, leadingto severe shrinkage of the glossoalveolar sulcus and very limited improvement in tongue mobility.Preoperative bromodeoxyuridine (BrdU) labeling of the graft and postoperative immunohistochemicalstaining of biopsy specimens from the grafted areas with anti-BrdU showed that the cultured cells areintegrated into the newly formed mucosal epithelium. Postoperative histologic investigations showed adifferentiation process in the grafted mucosal epithelium, with a change in the expression of cytokera-tins. At 6 months postoperatively, the typical pattern of normal nongrafted mucosa was regained.

Conclusions: This investigation provides evidence that tissue-engineered mucosal cells can serve as agraft for large intraoral wounds. Complete intraoral lining is quickly reestablished, and normal epithelialdifferentiation is seen in the graft area within a 6-month postoperative period.© 2001 American Association of Oral and Maxillofacial Surgeons

After tumor resection in the floor of the mouth, themethod selected for reconstruction depends on thelocation and size of the defect. In the case of largetumors and, subsequently, large tissue defects, themethod of choice is primary reconstruction usingmicrosurgically anastomosed free grafts from the ra-dial forearm or the lateral upper arm.1 However, afterresection of small tumors, the tissue defects may becovered primarily by fixing the tongue to the adjacent

tissues, such as the mucosa of the alveolar crest or thevestibule (Fig 1). This usually leads to considerableimpairment in swallowing and speech and diminishesthe conditions for prosthetic rehabilitation. Further-more, phonation and food intake may be affectednegatively, overtly decreasing the postoperative qual-ity of life.

To rehabilitate these patients and to gain a suffi-cient base for dentures, freeing of the tongue is per-formed.2 During this procedure, an extensive woundsurface is created that is usually covered by a split-thickness skin graft from the upper thigh or by amucosal graft from the palate. Split-thickness skingrafting is a simple surgical technique, but the post-operative functional condition of the intraoral liningis poor. Skin used intraorally is an insufficient base fordentures and leads to hyperplasia around implants.3,4

Moreover, only a limited amount of oral mucosa isavailable, not enough for covering the large woundcreated after secondary freeing of the tongue.5

In the following pilot study, tissue-engineered mu-cosa grafts produced according to the technique de-

Received from the Department of Oral and Maxillofacial Surgery,

University of Freiburg, Germany.

*Associate Professor in Oral and Maxillofacial Surgery

†Resident and Research Assistant in Oral and Maxillofacial Sur-

gery.

Address correspondence and reprint requests to Dr Lauer:

Department of Oral and Maxillofacial Surgery, University Hospital

Carl-Gustav-Carus Dresden, Fetscher Straße 74, D-01307 Dresden,

Germany; email: guenterlauer@gmx.de

© 2001 American Association of Oral and Maxillofacial Surgeons

0278-2391/01/5902-0007$35.00/0

doi:10.1053/joms.2001.20489

169

scribed by Lauer in 19946 were used as an alternativemethod to cover the resulting wound created by free-ing of the tongue. The aim of this prospective inves-tigation was to assess the functional, histologic, andimmunhistologic results following this procedure.

FIGURE 1. The anterior floor of the mouth after resection of asquamous cell carcinoma. The tongue is fixed to the mucosa of thevestibule covering the tissue defect.

FIGURE 2. The tissue-engineered mucosal transplant attached tothe petroleum jelly gauze ready for transfer onto the intraoral woundsurface. The white clusters represent the explants in the primaryculture (arrows) from which the keratinocytes grew to form a con-fluent and coherent sheet of keratinocytes (arrowhead—border ofthe coherent keratinocyte sheet).

FIGURE 3. Patient 1. Intraoral wound site after gauze dressingremoval at day 11 postoperatively. The epithelialized wound surfaceis covered with some fibrin (arrow).

FIGURE 4. Patient 1. Intraoral wound after gauze dressing removalat day 18 postoperatively. A complete mucosal layer had beenformed.

FIGURE 7. Photomicrograph of gingival biopsy from the grafted site,taken close to the alveolar ridge at 6 months postoperatively. Strat-ification has formed in the epithelial layer (E), and rete ridges (R)are seen extending into the connective tissue (CT). All suprabasallayers are positive for cytokeratins 1, 2, 10, and 11. (Originalmagnification �250.)

170 TISSUE ENGINEERING OF MUCOSA GRAFTS

Patients and Methods

Tumor resection and local defect coverage with atongue flap was performed primarily in 6 patientswith squamous cell carcinoma of the anterior or lat-eral floor of the mouth (T stage: T1 and T2). After adisease-free interval between 6 and 18 months, free-ing of the tongue and grafting of the floor of themouth was performed to improve food intake, swal-lowing, and speech.

TISSUE ENGINEERING OF THE GRAFT

To tissue engineer the mucosa grafts, a biopsy 4mm in diameter was taken from the hard palate ap-proximately 4 weeks before the operation. In addi-tion, 30 mL autogenous serum was produced from avenous whole blood sample. The mucosal keratino-cyte grafts were cultured, using the explant techniqueas described previously.6 The primary cultures wereincubated in Dulbecco’s modified Eagle’s Medium,Nutrient Factor F 12 (ratio 3/1, Gibco, Eggenstein,Germany) containing the additives insulin 5 mg/L,hydrocortisone 0.2 mg/L, cholera toxin 0.0085 mg/L,epidermal growth factor 10 �g/L, 10% autogenousserum,7 penicillin 100,000 U/L, and streptomycin100,000 �g/L. Within a period of 3 to 5 weeks, mu-cosal grafts consisting of several layers of keratino-cytes were engineered. These were detached enzy-matically from the culture flask with 0.05% Dispase(Boehringer Co, Mannheim, Germany) and trans-planted to the wound surface, using vaseline gauze(Adaptic; Johnson & Johnson Co, Arlington, TX) as acarrier (Fig 2). Mucosal grafts up to of 75.0 cm2 in sizewere engineered.

To assess the outcome of the cultured and subse-quently grafted mucosal keratinocytes, the culturedcells were labeled with 20 �mol/L bromodeoxyuri-dine (Sigma, Munich, Germany) in culture mediumwithout serum in 2 cases 2 hours before grafting.8 Inan early biopsy specimen, the labeled cells were de-tected by using anti-Brdu immunohistochemistry.

SURGICAL TECHNIQUE

Freeing of the tongue was performed under generalanesthesia. After the mucosal incision was made, the

tissue was separated along the scar to mobilize thetongue. The created defects were between 4.0 � 6.0and 6.0 � 9.0 cm. Thorough hemostasis was ob-tained, and the cultured mucosa graft adherent to thevaseline gauze carrier (Fig 2) was transferred to thewound surface in the floor of the mouth and fixedwith single 4–0 Vicryl sutures (Ethicon, Norderstedt,Germany). Additionally, the cultured mucosa graftwas secured by an intraoral gauze dressing fixed tothe wound surface by single-suture loops for 8 to 10days. After removal of the dressing, no further specialwound care was performed.

POSTOPERATIVE ASSESSMENT

The healing of the cultured graft was assessed clin-ically, histologically, and immunohistologically.

CLINICAL ASSESSMENT

Careful clinical inspection of the intraoral site wasperformed up to days 8 to 10. Wound healing, espe-cially epithelialization, was observed every other dayup to day 30 after removal of the gauze dressing. In afurther postoperative observation period of up to 3years, the mobility of tongue was recorded. Becauseof the topographically difficult situation in the floor ofthe mouth, repeated measurements of the graftedarea could not be performed, but the gain of adherentmucosa was assessed, as was speech improvement.The assessment schedule is shown in Table 1.

Table 1. ASSESSMENT SCHEDULE FOR THE GRAFTED SITES

Assessment Score Tongue Mobility Extension of Attached Mucosa Speech Improvement

No improvement Poor mobility No adherent mucosa on thealveolar ridge

Speech not understandable

Fair Reaches the alveolar ridge Less than 5 mm of adherentmucosa

Speech fairly understandable

Good Reaches to the vermilion 5 to 10 mm of adherent mucosa Speech understandableExcellent Reaches to all parts of the

lower lipMore than 10 mm of adherent

mucosaNormal speech

Table 2. BIOPSY SCHEDULE FOR INTRAORAL SITESGRAFTED WITH TISSUE ENGINEERED MUCOSA

PatientNo. Sex

Age(yr)

Postoperative Interval toBiopsy

Days Weeks Months

1 F 55 16 14 —2 M 48 — 20 303 M 53 9 6 —4 F 65 20 7/23 85 M 38 28 — 76 M 45 8 21 19

LAUER AND SCHIMMING 171

HISTOLOGIC AND IMMUNOHISTOLOGICASSESSMENT

Biopsy specimens were taken from the grafted sitesbetween day 8 and 30 months postoperatively (Table2). The specimens were fixed in 70% alcohol, dehy-drated, and embedded in Technovit 8100 (Kulzer Co,Wehrheim, Germany). Semi-thin sections were cutand stained with hematoxylin and eosin. For immu-nohistology, the sections were predigested with a0.1% trypsin solution after rinsing with saline and block-ing of nonspecific binding sites with 10% horse serum.The immunohistochemical protocol was performed byusing the Vectastain method, incubating the primaryand secondary antibody for 30 minutes each.

As markers for wound healing, cytokeratin poly-peptides and the basement membrane proteins lami-nin and collagen IV were examined. There is a typicaldistribution cytokeratins in oral mucosa,9,10 cytokera-tins 5, 6 being specific for basal cells, and cytokeratins1,2,10,11, and 17 for cells in the suprabasal layers ofthe epithelium. Collagen IV and laminin are typical forthe basement membrane.11 The cytokeratin antibod-ies AE 2 (ICN, Munich, Germany) for cytokeratins 1,2, 10, 11, the antibodies for cytokeratin 5, 6, and 17(Boehringer Co), and those for laminin and collagenIV (Sigma) were used.

To provide some evidence about the outcome ofthe grafted keratinocytes, the incorporated BrdU was

demonstrated in slides of an early biopsy after DNAdenaturation with 0.07 mol/L NaOH for 10 seconds,using immunohistochemistry and Anti BrdU 1:100(Sigma) as described.8 The slides were assessed byusing a Zeiss Axiolab microscope (Zeiss Co, Munich,Germany) at 10� and 25� magnification.

Results

At the time of intraoral wound dressing removal,usually 8 to 11 days postoperatively, an epithelialzedwound surface covered with some fibrin was ob-served in 5 patients (Fig 3). In one patient,5 there wasstill wound debris and mainly a nonepithelizedwound surface was visible at 20 days (Table 3).

In 5 patients, a complete mucosal layer had formedby 15 to 18 days postoperatively (Fig 4). In contrast,in 1 patient5 there was still an incomplete, fibrindebris–covered, mucosal layer visible on day 20. Inthis case, complete wound healing (epithelialization)was finally observed on day 30. Subsequently, in thispatient, there was severe shrinkage of the glossoal-veolar sulcus, and only a very limited improvement intongue mobility was achieved. In the other 5 patients,there was also some shrinkage of the wound surface,and flattening of the glossoalveolar sulcus was ob-served, but the primary improvement of tongue mo-bility was maintained during the entire postoperativefollow-up period (Table 4).

In the postoperative interval, 4 patients receivedremovable overdentures and 1 received a fixedbridge. The prosthodontic appliances were implantborne, except in 1 patient, in whom it was mucosaborne. This patient did not wear the prosthesis. Inanother patient, a sufficient prosthodontic rehabilita-tion was not achieved because of less than 5 mm ofpoorly fixed mucosa on the alveolar ridge (Table 4).

The histologic evaluation of biopsies at dressingremoval on day 8 showed conglomerates of keratino-cytes forming a superficial layer on the wound over anedematous, extracellular, matrix-rich connective tis-sue (Fig 5A). At the basal border of these keratinocyteconglomerates, parts of the basal membrane were

Table 3. WOUND HEALING STATUS AFTERGRAFTING OF TISSUE-ENGINEERED CULTUREDMUCOSA

PatientNo.

Wound Healing Status Postoperative

Partial Healing (days)(Epithelialization and Fibrin Debris)

CompleteMucosal Lining

(days)

1 11 182 10 153 9 154 8 165 20 306 9 16

Table 4. LONG-TERM POSTOPERATIVE IMPROVEMENT IN FUNCTION (ASSESSMENT ACCORDING TOEWERS AND HOFFMEISTER2)

PatientNo.

TongueMobility

Extension ofAttached Mucosa

SpeechImprovement

ProsthodonticRehabilitation

1 Good Good Good Yes2 Excellent Good Excellent Yes3 Excellent Good Good Yes4 Good Good Fair Yes5 Fair Fair No impr. No6 Excellent Fair Good Disliked by patient

172 TISSUE ENGINEERING OF MUCOSA GRAFTS

FIGURE 5. A, Photomicrograph of gingival biopsy from the grafted site at day 8 postoperatively. There are conglomerates of keratinocytes on thewound surface, forming an epithelial layer (E). Underneath these keratinocyte conglomerates at the basal border with the connective tissue matrix(CT), there is a positive linear dotted staining for collagen IV of varying intensity (arrowheads). (Original magnification �150.) B, Photomicrographof gingival biopsy from the grafted site at day 8 postoperatively immunohistolchemically stained for cytokeratin 17. On the surface of the wound,there are conglomerates of keratinocytes (K) with a positive immunoreaction over an edematous extracellular matrix. Underneath the basal borderof those keratinocyte conglomerates (arrowheads), there is an extracellular, matrix-rich, connective tissue (CT). (Original magnification �475.)

FIGURE 6. A, Photomicrograph of gingival biopsy from the grafted site at day 16 postoperatively immunohistochemically stained for cytokeratins5/6. A multilayered epithelium (E) has formed, and the keratinocytes in all layers express these cytokeratins. (CT, connective tissue.) (Originalmagnification �250.) B, Photomicrograph of gingival biopsy from the grafted site at day 16 postoperatively immunohistochemically stained forBrdU, which was incorporated into the keratinocytes immediately before grafting. There is a dark brown reaction in some nuclei (arrows), givingevidence that the cultured and transplanted cells are present in the graft. Arrowheads indicate its border between the mucosal epithelium (E) and theconnective tissue (CT). (Original magnification �475.)

immunohistochemically visible in dotted patterns bythe reaction of anti-collagen IV and anti-laminin (Fig5A). The keratinocytes in these conglomerates re-acted strongly positively for cytokeratins (CK) 5, 6,and 17 (Fig 5B).

By day 16, a multilayered epithelium had formed.Within this multilayered epithelium, the nuclei ofsome of the keratinocytes bound anti-bromodeoxyuri-dine (BrdU) (Fig 6A). The keratinocytes of the supra-basal layers expressed CK 1, 2, 10, and 11, stainingwhereas CKs 5, 6, and 17 reacted in all layers (Fig 6B).The mucosal epithelium rested on a continuous basalmembrane that was positive for collagen IV and lami-nin in a linear pattern.

Four to 8 weeks postoperatively, a 20 to 40 celllayered epithelium, without differentiation into strata,was visible. The CKs 5, 6, and 17 were positive in alllayers; the anti-cytokeratins 1, 2, 10, and 11 stained inall layers except the basal stratum. In the furtherpostoperative observation period, the histologic con-figuration of the grafted mucosa changed to a normalappearance. From 6 months postoperatively onward,a differentiated mucosal epithelium was observed,and the different strata could be distinguished. Inbiopsies performed close to the alveolar ridge, a moreprominent formation of the rete ridges was observed(Fig 7), as well as a well-developed stratum corneum.The cell differentiation, as judged by the cytokeratinstaining, resembled that of the nongrafted mucosa.CK 5 and 6 were mainly expressed in the basal andadjacent suprabasal layers, whereas the CKs 1, 2, 10,and 11 reaction was restricted to the upper strata (Fig7). These findings were noted in biopsies performedat 30 months.

Discussion

The technique of producing a large graft from asmall biopsy and keeping it alive has opened a newway in the field of plastic and reconstructive surgery.Since Rheinwald and Green12 first produced culturedskin grafts about 20 years ago, the indications andlimitations of this method have gradually becomeclear. Different tissue engineering and reconstructionprocedures have been described.6,13–15 Ueda et al14

used oral mucosal cells for covering of donor siteepithelium defects after split-thickness skin grafts.They found that the site where the mucosal epithe-lium was transplanted keratinized normally within 4weeks and formed normal skin.

The current investigation provides evidence thattissue-engineered mucosa also can serve as a graftfor covering intraoral wounds resulting from sec-ondary freeing of the tongue in patients treated bytumor resection. In contrast to a split-thickness skin

graft, the extraoral harvesting wound and possibledonor side morbidity can be avoided. There is onlya small intraoral biopsy (maximum, 4 mm diameter)necessary, and this wound is closed primarily.When a split-thickness skin graft is used, greaterkeratinization occurs, which does not occur whengrafting cultured mucosa. Furthermore, a shorterinterval for wound healing (approximately 18 days)was observed. Although there was some degree ofwound shrinkage, good mobility of the tongue andan adequate denture bearing area were achievedand maintained so that satisfactory prosthodonticrehabilitation could be performed in 4 of 6 cases. Adifferentiation into a parakerathotic stratified epi-thelium takes place, as shown histologically andimmunohistologically.

During the postoperative healing period, there isa sequence and topographic order in the appear-ance of the extracellular and intracellular markers.With the cytokeratins used as intracellular markers,the appearance change ended with a distributionpattern similar to that described in normal mu-cosa9,10 and in split mucosal transplants 7 monthsafter grafting.16 When using the extracellular mark-ers laminin and collagen IV, there was evidence forde novo synthesis of a basement membrane at theepithelial-connective tissue junction, as describedduring the healing of cultured skin grafts.17 Similarclinical, cell-biologic, and histologic results havebeen described when grafting skin wounds withcultured keratinocytes17 that promote epithelialcoverage. However, the outcome of the graftedkeratinocytes remains unclear. The in vitro labelingof cultured and subsequently grafted keratinocytesprovides evidence that the grafted gingival kerat-inocytes are integrated in the newly formed muco-sal epithelium.

Although there were good clinical results with thetissue-engineered mucosa grafts in most of the cases,some wound shrinkage was seen. This might be be-cause these cultured mucosa grafts consist of only athin layer of keratinocytes without any submucosalconnective tissue. The additional transplantation ofsubmucosal connective tissue, as already establishedfor skin,18 may reduce this problem. However, woundshrinkage has also been described when using split-thickness skin grafts that consist of dermis and epi-dermis.19

In comparison with split-thickness skin grafts,which are performed simultaneously with secondaryfreeing of the tongue, tissue-engineered mucosa graftsrequire a considerable amount of preoperative workand laboratory preparation. This limits the applicationof the method to selected cases. Further develop-ments in tissue engineering may reduce this disadvan-tage, for example, when the mucosal layer can be

174 TISSUE ENGINEERING OF MUCOSA GRAFTS

engineered directly on a carrier.20 The use of a colla-gen matrix may allow the engineering of completemucosa–submucosa grafts, as already described forskin.21

References1. Gellrich N-C, Lauer G, Kwon T, et al: A comparative study of

lateral upper arm flap versus radial forearm flaps for intraoralreconstruction. Int J Oral Maxillofac Surg 28:64, 1999 (suppl1)

2. Ewers R, Hoffmeister B: Reconstruction of the mandibulardenature bearing area and freeing of the tongue after tumorsurgery. J Oral Maxillofac Surg 46:272, 1988

3. Ritter R: Zweckma�ige und unzweckmassige praprothetischeChirurgie. ZWR/Ref 68:351, 1967

4. Mitchell DL, Synnott SA, van Dercreek JA: Tissue reactioninvolving an intraoral skin graft and cp titanium abutments:A clinical report. Int J Oral Maxillofac Implants 5:79, 1990

5. Huybers TJM, Stoelinga JW, de Koomen HA, et al: Mandibularvestibuloplasty using a free mucosal graft. Int J Oral Surg 14:11,1985

6. Lauer G: Autografting of feeder-cell free cultured gingival epi-thelium: Method and clinical application. J CraniomaxillofacSurg 22:18, 1994

7. Lauer G: Autogenous serum for culturing keratinocyte au-tografts, in Phillips GO, von Versen R, Strong DM, et al (eds):Advances in Tissue Banking, vol 1. Singapore, World Scientific,1997, p 183

8. Sasaki K, Ogino T, Takahashi M: Immunological determina-tion of labeling index on human tumor tissue sections usingmonoclonal anti-BrdUrd Antibody. Stain Technology 61:155,1986

9. Ouhayoun JP, Gosselin F, Forest N, et al: Cytokeratin patternsof human oral epithelia: Differences in cytokeratin synthesis ingingival epithelium and adjacent alveolar mucosa. Differentia-tion 30:123, 1985

10. Juhl M, Reibel J, Stolze K: Immunhistochemical distribution ofkeratin proteins in clinically healthy human gingival epithelia.Scand J Dent Res 97:159, 1989

11. Salonen J, Pelliniemi LJ, Foidart JM, et al: Immunohistochemi-cal characterization of the basement membranes of the humanoral mucosa. Arch Oral Biol 29:363, 1984

12. Rheinwald J, Green H: Serial cultivation of strain of humanepidermal keratinocytes: The formation of keratinizing colo-nies from single cells. Cell 6:331, 1975

13. Ueda M, Ebata K, Kaneda T: In vitro fabrication of bioartificialmucosa for reconstruction of oral mucosa: Basic research andclinical application. Ann Plast Surg 27:540, 1991

14. Ueda M, Hata K-I, Horie K, et al: The potential of oral mucosalcells for cultured epithelium: A preliminary report. Ann PlastSurg 35:498, 1995

15. Izumi K, Takacs G, Terashi H, et al: Ex vivo development of acomposite human oral mucosal equivalent. J Oral MaxillofacSurg 57:571, 1999

16. Lauer G, Wiedmann-Al-Ahmad M, Otten JE, et al: Immuno-histochemical study during healing of free palatal mucosagrafts on plastic-embedded samples. J Oral Pathol Med (inpress)

17. Compton CC, Gill JM, Bradford DA, et al: Skin regeneratedfrom cultured epithelial autografts on full-thickness burnwounds from 6 days to 5 years after grafting. Lab Invest60:600, 1989

18. Sabolinski ML, Alvarez O, Auletta M, et al: Cultured skin as a“smart material” for healing wounds: Experience in venousulcers. Biomaterials 17:311, 1996

19. Gregory EW, Triplett RG, Connole PW: Comparison of freshautogenous and freeze-dried allogenic skin for mandibular ves-tibuloplasty. J Oral Maxillofac Surg 41:75, 1983

20. Gutwald R, Lauer G, Otten JE, et al: Epithelzellen und Fibroblastender Gingiva auf resorbierbaren Membranen: Gewebetransfer zurWundheilung? Deutsch Zahnarztl Z 49:1015, 1994

21. Hansbrough JF, Morgan J, Greenleaf G, et al: Evaluation ofgraftskin composite grafts on full-thickness wounds on athymicmice. J Burn Care Rehabil 15:346, 1994

175