Considerations for Free-Flap Reconstruction of the Hard Palate FREE

DEFECTS OF the palate following ablative oncologic surgery can cause severe functional and cosmetic deformities. The creation of large oronasal and oromaxillary fistulae, as well as the loss of crucial tooth-bearing segments, can extensively impair phonation, oral alimentation, lip and cheek support, and midface projection.¹

Traditionally, these defects have been obturated with a bulky dental prosthesis.^{2 - 4} Although acceptable results can eventually be achieved in many cases, patients may become dissatisfied for several reasons. The prosthesis must be in place for adequate speech, swallowing, and cosmesis. The patient must maintain adequate hygiene of the surgical site and prosthesis. Poor retention due to denture bulkiness and poor residual dentition can create leakage and oronasal regurgitation.

Surgical options to seal the palate have included local palatal, pharyngeal, and nasal septal flaps.^{5 - 7} For larger defects, temporalis, forehead, and more distant tubed flaps have been described.^{7 - 11} These techniques are limited to the amount of tissue available and/or to pedicle length.

Developments in microvascular free-tissue transfers provide a unique alternative for reconstruction of the palate and midface. An adequate amount of bone and/or soft tissue can be transferred in a single stage. Restrictions of pedicle length and flap geometry are eliminated as well. Various free-tissue transfers have been advocated for palate and midface reconstruction, including scapular,^{12 - 15} fibular,^{14 ,16 - 22} radial forearm,^{14 ,23 - 25} rectus abdominus,^{26 - 27} iliac crest,^{14 - 15 ,28} and latissimus dorsi flaps.^{29 - 30}

A series of patients who underwent free-flap reconstruction of the hard palate is described. The success of this technique is evaluated, and factors that aid in flap selection for specific palatal defects are identified.

A retrospective series of 30 patients who underwent free-flap reconstruction of the hard palate at the University of Washington, Seattle, and the University of Utah, Salt Lake City, between September 1, 1993, and October 1, 1997, were examined. Defects were classified as hemipalate, subtotal or total palate, and anterior arch. All hemipalate defects had associated cheek and infraorbital defects. The plan for palatal and dental rehabilitation was made preoperatively with the maxillofacial prosthodontist. The type of free flap used and surgical complications were identified for each patient's defect.

Patient follow-up ranged from 3 months to 4 years. Type of diet and dental rehabilitation were identified in each patient. Twenty-three patients underwent formal speech evaluation by a speech pathologist at 6 months and 1 year. A subset of 20 patients completed the University of Washington's quality of life (QOL) questionnaire before the initiation of treatment and again at 6 and 12 months after treatment.³¹ The purpose of the questionnaire was explained. Informed consent was obtained in accordance with an approved human subjects protocol. Only overall QOL scores are reported in this study.

Table 1 lists the characteristics of patients undergoing free-tissue transfer reconstruction of the hard palate. Twenty-three patients underwent immediate reconstruction at the time of tumor ablation, and 7 patients underwent secondary reconstruction. The flaps used are listed as well.

Table 2 matches defect to flap choice. Determination of whether to use a bone-containing flap was largely based on the quantity and quality of the patient's residual dentition and/or denture-bearing alveolar arch. Myocutaneous free flaps were used in most hemipalate defects. Choice of myocutaneous or myofascial flap was dictated by defect volume considerations and surgeon preference. Two patients in this group received scapular osteocutaneous free flaps. The bony component was used to reconstruct the orbital floor in one patient, and the orbital floor and palate in the other patient. Anterior arch defects were generally small and well-suited to the radial forearm free flap without or with bone. The addition of radial bone was used to improve anterior arch and upper lip contour in the latter 4 patients, and provided a better aesthetic result than soft tissue alone.

All 10 patients with subtotal or total hard palate defects received bone-containing free-tissue transfers. In 8 of the 10 cases, vein grafts were required to allow the flap pedicle to reach recipient vessels in the neck. Vein grafts were not required with any flaps used to reconstruct hemipalate or anterior arch defects.

The complications of the free-tissue transfer reconstructions were as follow:

Two flaps required urgent return to the operating room for venous compromise, each on postoperative day 1. In one case, kinking of the venous system due to poor pedicle geometry was identified; the kinking was corrected by rerouting the vessels. The second case of venous compromise was due to torsion of the venous outflow system. The venous anastomosis was separated; proper vessel alignment was established, and venous outflow was restored. Both flaps were salvaged, and there were no flap failures in this series. One patient developed an abdominal hematoma after rectus abdominus muscle harvest, and another developed a neck hematoma. These hematomas were drained at the bedside, and the wounds completely healed by secondary intention.

Three patients developed partial dehiscence of the palatal wound within 10 days of the reconstructive surgery. Two patients were treated with local wound care and subsequently returned to the operating room on postoperative days 14 and 22 for delayed wound closure. Both wounds then healed completely. One wound healed by secondary intention.

All patients resumed oral diets (Table 3). At a minimum of 6 months' follow-up, 19 patients resumed a regular diet and 9, a soft diet. The 2 most recent patients examined at 3- and 4-month follow-ups have achieved a soft diet.

Twenty-three patients underwent formal speech evaluation at 6 and 12 months after reconstructive surgery. Seventy-eight percent of the patients have resumed normal phonation.

Dental rehabilitation is detailed in Table 3. Ten of 14 patients with hemipalate defects had ample dentition in their residual arch. Eight of these patients had a conventional partial denture created. In those 4 patients with hemipalate defects without dentition in their residual arch, 2 had osseointegrated implants placed in their native bone and wear implant-borne prostheses. All 6 patients with anterior arch defects had adequate posterior dentition on both sides of the maxilla. Four patients function with conventional partial dentures. Seven of 10 patients with subtotal or total palatal reconstructions underwent placement of osseointegrated implants, and 6 of these patients wear an implant-borne dental prosthesis.

Twenty patients completed the University of Washington's QOL questionnaire during a 1-year period (Figure 1). Overall, 12 patients rate their QOL as excellent; 6, good; and 2, fair.

Reconstruction of palatal defects is necessary to separate the oral cavity from the nasal and sinus cavities. Without this separation, speech is altered and swallowing is hindered, which creates a disability for these patients. The goals of reconstruction are to restore speech and allow the patient to maintain a normal diet. Re-establishment of oral competence, lower midface projection, and an aesthetically acceptable appearance are also essential. Functional dental restoration completes the reconstruction.^{1 - 2}

The traditional method of reconstructing these palatal and maxillary defects is by placing a maxillofacial prosthesis to obturate the cavity and seal the palate.^{3 - 4} These appliances can be cumbersome and difficult to clean. Patients must also maintain meticulous hygiene of the surgical cavity, which may be difficult in the elderly patient. There is often continued leakage between the nasal and oral cavities, caused by the difficulty in obtaining an adequate leak-tight seal without irritation that creates further difficulties with speech and oral hygiene. Dentition must also be adequate for the retention of the appliance, and, as the maxillary defect increases, problems with retention and motion of the implant escalate.

In an attempt to alleviate these problems, various pedicled autogenous tissues were initially used. As early as 1862, von Langenbeck⁵ described the use of local palatal flaps for small defects. In addition to other local tissue transfers, forehead flaps, temporalis flaps, and distant tubed flaps such as the deltopectoral and upper arm flaps were described for larger palatal defects.^{6 - 11} Vascularized cranial bone grafts have also been used in a small series of patients.^{32 - 33} All these techniques, however, are limited by the lack of tissue, the length of vascular pedicle, and/or the need for multiple stages to achieve the final result.

Microvascular free-flap techniques to reconstruct defects of the palate and maxilla allow immediate, 1-stage reconstruction. Abundant bone and soft tissues can be transferred without the limitations of vascular pedicle length and tissue orientation. Various donor sites have been recommended in the literature.^{12 - 30 ,34}

Small case series have described using the radial forearm free flap to close palatal defects due to failed cleft palate repair²³ and tumor extirpation.²⁴ Its thin, pliable tissue characteristics allowed the palate to be sealed and created a denture-bearing surface. Recently, Cordeiro et al²⁵ described the "sandwich" radial forearm osteocutaneous free flap for reconstruction of the subtotal maxillectomy defect. The bone was used to recreate the maxillary arch and the skin paddle was folded around it, so the palatal and nasal linings could be restored. Neither of the 2 patients received osseointegrated implants at the time of publication, but one patient has a tissue-borne prosthesis with "relatively normal speech and mastication."

LARGER SERIES of palatal and maxillary defects reconstructed with myocutaneous free flaps have been described. Shestak et al^{29 - 30} found that the latissimus dorsi free flap was not only able to seal the palate but provided excellent tissue bulk in the cheek and malar area when defects extended to these sites. In addition, the ample pedicle length allowed microvascular anastomoses in the neck without the need for vein grafts. Similar findings and flap attributes were demonstrated by the use of the rectus abdominus free flap.^{26 - 27} In the series described by Olsen et al,²⁷ 11 patients with massive sino-orbital defects underwent reconstruction with primarily rectus abdominus free flaps. Six of 11 patients underwent palate reconstruction with a microvascular free-tissue transfer, and all exhibited excellent speech and deglutition. Some of these patients also had dental prostheses placed. These patients also preferred living tissue reconstruction to the problems associated with cleaning and placing a prosthesis.

Most reports^{12 - 15} recommending free-tissue transfers for maxillary reconstruction describe the use of the scapular osteocutaneous flap. This flap has the advantage of having a soft tissue component that can be rotated around the adequate bone stock with greater freedom than other composite flaps. It is particularly useful in defects in which the orbital floor, zygoma bone, and palate are required to be reconstructed. Case reports using the fibular free flap^{14 ,16 - 22} and the iliac crest free flap²⁸ have emerged in the literature, as these tissue transfers have an adequate volume of bone to support osseointegrated implants.

Finally, some authors²² advocated the use of dual free flaps to reconstruct these complex defects. Although success has been reported in small series of patients, the advantages of 2 separate free flaps have not been demonstrated. The complexity of this effort has not been shown to be more beneficial than the use of a single free flap with excellent bone stock and pliable soft tissue.

Controversy exists in the literature concerning which flap is best to seal the palate. Schusterman et al¹⁴ recommended a bone-containing free flap for maxillary reconstruction when bony support was needed and previous irradiation precluded the use of nonvascularized grafts. Coleman¹⁵ suggested that the needs of the wound be matched with characteristics of the appropriate flap, including length of vascular pedicle, thickness or thinness of skin, muscle and fat, volume of soft tissue available, durability and thickness of bone, and donor site morbidity.

The criteria for palate closure with a soft tissue free flap in a series reported by Funk et al³⁴ included sufficient residual dentition to retain a dental prosthesis or, if the anatomy of the reconstruction would allow, retention of an upper denture despite the absence of teeth. If placement of osseointegrated implants was anticipated, the scapular osteocutaneous flap was preferred.

Recently, Davison et al² recommended free-tissue transfer closure of maxillectomy defects when substantial associated sino-orbital and/or soft tissue defects existed. They concluded their review, however, by stating "refinement of microsurgical techniques with free vascularized bone may provide an ideal answer in providing surgical reconstruction that could support an implant-bone prosthesis."² (p218)

In our series of patients, all of the previously mentioned flaps except the iliac crest flap have been used. Free-tissue transfer has been demonstrated to be reliable and to achieve separation of the oral and sinonasal cavities. In addition, deglutition, speech, and an acceptable QOL have been restored for most patients. Quality of life data for patients with obturators are incomplete at this time. However, a prospective study is under way examining these patients as well as a subset of patients who initially used an obturator and then underwent secondary free-flap reconstruction of the palate.

Flap selection has been determined by various factors. Whether to select a bone-containing free flap has been largely determined by the amount, location, and quality of residual dentition and/or denture-bearing alveolar arch. In most patients with hemipalatal defects, sufficient retention was present to support a conventional dental prosthesis so that myocutaneous free flaps were used in 12 of the 14 patients. Since many of these defects had associated upper midface deficits, soft tissue bulk restored these areas. In addition, 8 patients underwent cranial bone grafting to restore the zygomatic bone and orbital floor. This was achieved with the scapular flap in 2 patients.

In those patients in whom the anterior arch was missing, the radial forearm flap provided thin, pliable tissue with minimal bulk. The incorporation of radial bone may aid in arch contour but was not considered suitable for placement of osseointegrated implants.

In patients who underwent total or subtotal palatectomies, little retentive surface was available for a conventional dental prosthesis. Therefore, all of these patients received a bone-containing free flap (9 fibular and 1 scapular) to create the potential for implant-borne dental restoration. These were also the most complex palatal reconstructions, and 8 patients required vein grafts to achieve adequate pedicle length to the recipient vessels. Six patients have been fully restored with dental prostheses to date.

In conclusion, although maxillofacial prostheses still have a role in primary reconstruction of palatal defects, free-tissue transfers have been successful and well-accepted by various patients. Excellent contour and acceptable cosmesis can be achieved. In addition, the potential for dental rehabilitation with the restoration of masticatory function and normal phonation exists. The choice of free-tissue transfer should be tailored to the specific defect, denture-bearing potential of native residual tissues, and patient needs.

Accepted for publication July 15, 1998.

Presented at the joint meeting of the American Society of Head and Neck Surgery and Society of Head and Neck Surgeons, Palm Beach, Fla, May 16, 1998.

Corresponding author: Neal D. Futran, MD, DMD, Department of Otolaryngology–Head and Neck Surgery, University of Washington School of Medicine, 1959 NE Pacific St, Box 356515, Seattle, WA 98195-6515 (e-mail: futran@otomail.u.washington.edu).