The maxilla is the functional and esthetic keystone of the midface, forming part of each of the key midfacial elements; these are the orbits, the zygomatico-maxillary complex, the nasal unit, and the stomatognathic complex. Maxillary reconstruction is a challenging endeavor in functional and esthetic restoration. Given its central location in the midface and its contributions to the midface, maxillary defects are inherently complex because they generally involve more than one midfacial component. Maxillary defects are composite in nature, and they often require skin coverage, bony support, and mucosal lining for reconstruction. Reconstruction of maxillary defects secondary to warfare, trauma, ablative tumor surgery, or congenital deformities must meet the following goals namely: (1) obliteration of the defect; (2) restoration of essential functions such as mastication and speech, (3) provision for adequate structural support to each of the midfacial units and (4) esthetic restoration of facial features. This chapter will discuss the anatomic considerations, the historical approaches to maxillary reconstruction as well as state-of-the-art techniques in use today.
Understanding the complex three-dimensional anatomy of the maxilla and its relationship to contiguous structures is critical to approaching reconstruction of the midface. Conceptually, the maxilla can be described as a geometric structure with six walls (a hexahedron, Figure 1).
The roof of the box is the floor of the orbit; the floor forms the anterior hard palate and alveolar ridge; the lateral walls form the lateral walls of the maxillary sinuses and are a part of the lacrimal system. The maxillary sinus, the largest of the paranasal sinuses, is contained within the central portion of the maxilla. Anteriorly it comprises the midface supporting the nose and anterior teeth. Overlying the posterior pterygoid region of the maxilla is the cranial base.
The maxilla provides structural support between the skull base and the occlusal plane, supports the globe, separates the oral and nasal cavities and resists the forces of mastication. 
Finally, the overlying soft tissues, including the mimetic musculature of the midface, are supported by the maxilla and influence to a large extent one's unique facial appearance.
3. Historical procedures for maxillary reconstruction
Traditionally, reconstruction of large maxillary defects was accomplished by obturation of the defect with a prosthetic appliance. [2,3] Before the development of more sophisticated reconstructive techniques, prosthetic appliances were the only modality available to address the functional and esthetic requirements of such a complex defect. Both functional and esthetic results were far from optimal (Figure 4).
Edgerton and Zovickian  reviewed early attempts at autogenous reconstruction of the maxilla and reported a palatal reconstruction technique using cervical flaps. These early reconstructive endeavors progressed from local flaps, such as forehead, upper lip, cheek, pharyngeal, turbinate, and tongue flaps, to tube flaps from the upper extremity, thorax, and abdomen. [5,6] Numerous other local flaps have been described for maxillary and palatal reconstruction. Generally, these have been useful for small defects or to augment other tissue-transfer techniques used to reconstruct larger defects. [7-17]
One of the earliest descriptions of a staged maxillary reconstruction with both soft tissue and bone was by Campbell in 1948.  He combined a temporalis muscle flap with a rotational palatal mucosal flap for soft-tissue reconstruction. An iliac bone graft was then placed in a second procedure; this was followed by the placement of a vestibular skin graft. The resulting reconstructed maxilla was capable of supporting a conventional maxillary denture.During the 1960s and 1970s, pedicled myocutaneous flaps were developed and replaced the more cumbersome tube flaps previously used in reconstructive surgery. However, these flaps tended to be quite bulky and were limited in their capacity to replicate the complexities of the resected maxillary structures. During the 1980s, a revolution in reconstructive surgery was brought about by the introduction of free tissue transfer techniques. These techniques have been widely applied in maxillary reconstruction, [18-25] and they have made possible the use of less bulky fascial or fasciocutaneous and osseous flaps. [26-36] Alongside the development of these tissue transfer techniques was the development of osseointegration pioneered by Branemark. [37-39] This technology, in combination with free tissue transfer, has made autogenous reconstruction of the maxilla and dentofacial rehabilitation possible. [40-45]
4. Classifying midfacial defects
Because of the disparate shapes and sizes of defects affecting the maxilla, the complex three-dimensional anatomy and the contiguous relationship of the maxilla to the surrounding structures, the broad category of maxillectomy constitute a wide spectrum of diverse defects.  Thus, a classification system to group this wide array of possible composite tissue defects was needed to facilitate clinical decision-making by outlining preferred reconstructive options and their common functional and esthetic sequelae. Attesting to both the variety and complexity of midfacial defects, numerous different classification schemes have been proposed. Based on a combined experience with 45 maxillectomies, Brown et al. developed a classification scheme allowing a very detailed description of 10 possible defects involving the palate; defects of the midface not involving the palate were excluded from the classification.  Unfortunately, the status of the orbital floor and zygoma, which play an important role in both the function and cosmesis of the midface, were not specifically addressed and specific recommendations for the reconstruction of each type was not given.
Wells and Luce proposed a classification system based on the extent of maxillary resections.  The schema allows the distinct classification of defects; however, proposed treatment focuses on the use of prosthetic obturators and/or the use of regional flaps rather than the specific use of microsurgical tissue transfer. In contrast, Yamamoto advocated the use of complex microsurgical procedures, specifically, the combined latissimus dorsi myocutaneous free flap with scapular bone based on the angular branch of the thoracodorsal artery and the rectus abdominis myocutaneous flap combined with costal cartilage based on the vascular connection between the eighth intercostal and deep epigastric vascular system. [49-50] Based on their 10-year experience with 38 maxillary reconstructions, they designed a complex reconstructive algorithm that ultimately culminates in nine different clinical scenarios based predominantly on the aforementioned vascularized, composite-tissue flaps. Futran and Mendez presented an algorithm designed to depict options for midface reconstruction. Based on a thorough review of the literature, they classified defects as those involving the palate, the inferior maxilla, or the total maxilla with or without orbital exenteration.  Their algorithm was designed to delineate types of tissue required to reconstruct a particular defect, such as soft-tissue flaps or vascularized bone flaps rather than specific flap options. Spiro et al. proposed a relatively straightforward classification system that divides defects into three subtypes but does not specifically address the involvement of adjacent structures such as the orbit and zygoma.  based on a review of 108 patients, Davison et al. similarly divided patients into the two broad categories of “compete” or “partial” maxillectomy defects.  Although their group proposed a wide range of reconstructive techniques, the lack of a specific defect-oriented classification system outlining the remaining portion of the hard palate, dentition, orbit, and zygoma makes such an algorithm difficult to apply as a reconstructive guide. 
The same could be said for the classification proposed by Foster et al.  based on a single-surgeon series of 26 midfacial reconstructions; they classified defects into those involving soft-tissues and those involving bone. Bony midfacial defects were then subclassified into those involving more or less than half the palate. Triana et al. assessed 51 midfacial defects that had been treated with microvascular free-tissue transfer procedures.  The defects were classified as those seen after inferior partial maxillectomy, subtyped into the extent of palate lost and subdivided depending on the amount of malar bone and zygomatic arch lost. Okay et al. performed a retrospective review of 27 consecutive palatomaxillary reconstructions and designed a defect-oriented classification system designed to delineate the indication for prosthetic reconstruction, soft-tissue reconstruction, or vascularized bone-containing free flaps.  The authors concluded that the classification system does not address all factors required for decision-making. Although most of these classification systems allow for accurate descriptions of anatomical defects, many do not provide a clear algorithm for flap selection based on defect category. Others do not provide a comprehensive system for classifying defects of the midface that includes important structures such as the orbit or zygoma. One of the newer classifications has been proposed by McCarthy et al.56 They classify the maxillary defect of oncologic surgery origin into five distinct types; it is a rather straightforward classification but there are some deficiencies in this classification i.e. maxillary atrophy after radiation therapy (Figure 5,6).
An all inclusive classification is yet to be found; but as a rule of thumb maxillary reconstruction can be divided into three groups :
Upper maxilla which needs space filling or bulky flaps
Lower maxilla or alveolar ridge for which the prefabricated bone flaps are the best solution C: Combined or total maxillary defects in this group a single flap addressing both the problems is yet to be found.
The McCarthy classification is as follows:
4.1. Type l: Limited maxillectomy
Type l defects include resection of one or two walls of the maxilla, excluding the palate. In most cases, the anterior wall is partially removed with either the medial wall and/or, occasionally, the orbital rim. In addition, these resections commonly involve the overlying cheek and can extend onto the lips, nose, or eyelids. Thus, type l or limited maxillectomy defects usually require a significant amount of skin for resurfacing with minimal associated bone volume (Figure 7).
The radial forearm fasciocutaneous flap provides good external skin coverage and minimal bulk in this setting. Multiple skin islands can be designed and de-epithelialized when needed to wrap around bone grafts or supply nasal lining. If critical segments of bone are missing, such as the orbital rim or the anterior floor of the orbit, nonvascularized bone grafts can provide the needed support. Other flap options, depending on the amount of soft-tissue bulk required, include the lateral arm flap, anterolateral thigh flap,  and scapula flap. 
4.2. Type ll: Subtotal maxillectomy
Type ll defects include resection of the maxillary arch, hard palate, and anterior and lateral walls (five walls) with preservation of the orbital floor (Figure 8).
All type ll defects involving more than 50 percent of the transverse palate require flaps that provide a substantial surface area with which to reline the nasal floor and palatal roof, and bone for structural support. [59-61] Similarly, in patients who do not have sufficient retentive surfaces and/or teeth to support a conventional prosthesis, vascularized bone-containing free flap reconstruction is indicated.
The associated bulk provided by the skin and soft tissues is a significant disadvantage to using the fibula osteocutaneous flap, therefore we recommend the use of the prelaminated fibula free flap for the reconstruction of these defects (Figures 9-15).62
Various other donor sites have also been used to reconstruct these defects. Schliephake used a fasciocutaneous forearm flap followed by secondary bone grafting in two patients and reported that secondary nonvascularized bone grafting increases the risk of infection and is therefore not recommended. 
Use of the iliac crest free flap harvested with the internal oblique muscle has been reported by others. Iliac bone is plentiful and can provide a suitable bed for osseointegrated implants; however, its disadvantages include its short vascular pedicle and the potential for significant donor-site morbidity following its harvest. 
4.3. Type lll: Total maxillectomy
Type lll defects include resection of all six walls of the maxilla. These total maxillectomy defects are further subdivided into type lll a defects, where the orbital contents are preserved; and type lll b defects, where the orbital contents are exenterated.
4.3.1. Type llla
Reconstruction after total maxillectomy with preservation of the orbital contents is technically more challenging than maxillectomy with orbital exenteration. In this setting, reconstruction must: (1) provide support to the orbital contents, (2) obliterate any communication between the orbit and nasopharynx, and (3) reconstruct the palatal surface. When the orbital floor has been resected, support needs to be restored to the orbital contents; otherwise, the globe will prolapse downward, causing severe vertical dystopia with significant diplopia (Figure 16).
A variety of methods have been advocated to provide orbital support, including nonvascularized and vascularized bone grafts, alloplastic substitutes, and soft-tissue “slings.” [64-65] we strongly advocate the use of nonvascularized bone grafts to support the orbital contents. By contrast, the use of alloplastic substitutes in defects that potentially expose it to the oronasal cavity increase the opportunity for periprosthetic infection. The volume of a soft-tissue flap may change over time  secondary to muscle atrophy, scar contracture, or changes in nutritional status. In this setting, even minor changes in volume can translate into significant changes in the vertical position of the soft-tissue sling and consequently the volume of the orbital cavity.
By using the rectus abdominis free flap in combination with nonvascularized bone grafts, reconstruction of a three-dimensional defect is facilitated because the bone, skin, and soft-tissue components may be inset into their desired positions without compromising the microvascular aspect of the reconstruction. In addition, the rectus abdominis can be harvested easily during the resection and the pedicle can be extended up to 19 cm to reach the neck vessels.
Alternatively, the temporalis flap can be used to cover bone. Using this approach, however, requires the subsequent use of a palatal obturator; thus, the temporalis muscle flap is indicated primarily in older patients who are not candidates for free-tissue transfer. It is also useful for the patient who has an intact palate and preserved orbital contents (usually ethmoidal tumor resections), where access for free flap vessels is exceedingly difficult and muscle coverage is still needed to cover orbital bone grafts.  We however, support the use of vascularized bone flaps in this setting. The osteocutaneous free flap most frequently described for reconstruction of the maxillary region are the scapula, fibula, and radius. Each donor site has its own advantages and disadvantages.
The osteocutaneous radial forearm flap has been used for simultaneous reconstruction of the infraorbital margin and external skin in the midface.  Unfortunately, the volume of tissue transferred is rarely enough to obliterate the maxillary cavity completely, and palatal defects must be obturated with a prosthesis. 
Others have advocated the use of the subscapular flap to reconstruct defects caused by total maxillectomy with orbital preservation. Replacement of the alveolar arch inferiorly with the lateral scapular bone and the orbital floor and rim with the scapular tip has been described.  Schliephake reported difficulty however, in tailoring the scapular bone over the malar prominence, infraorbital rim, and maxillary wall at the same time that the lateral border of the scapula was to be positioned for placement of implants at the alveolar crest.  Yamamoto et al. have similarly reported using the scapular bone in conjunction with costal cartilage for reconstruction of all the maxillary buttresses in extended midfacial defects. 
Several authors have described the use of free fibula osteocutaneous flaps to reconstruct combined maxillary and mandible defects.  we think that the prefabricated fibula can address the alveolar ridge and the palate but cannot reconstruct both the mandible and maxilla in one setting and also the prefabricated fibula cannot act as a space filling flap for upper maxillary defects (Figure 17,18).
However, Futran et al. found that as the need for reconstruction of the zygomatic complex, infraorbital rim, and the floor increased, the fibula flap was limited in its ability to restore the entire maxillary form.  In addition, it was difficult to osteotomize and orient the bone to restore both the palate and the infraorbital area. Even with the harvest of additional soleus muscle bulk, it was difficult to rotate the skin paddle to resurface the palate and provide zygomatic and infraorbital contour. Based on this experience, their group concluded that when orbitozygomatic support is the primary objective, use of the fibular free flap is not advocated.
Brown presented three cases of reconstruction with the iliac crest myo-osseous flap with favorable functional results.  A “block” of iliac bone was used to restore alveolus, zygomatic prominence, and orbital rim with success. Genden et al. Described use of the iliac crest–internal oblique osteomusculocutaneous free flap in six patients, four of whom had type llla defects.  The iliac crest was fashioned to recreate the inferior orbital rim; the internal oblique muscle was used to reline the palate and resurface the ipsilateral lateral nasal wall. Based on their report, all four patients achieved facial symmetry and underwent placement of osseointegrated implants. Others have discouraged the use of this flap however, because of its potentially excessive bulk, limited soft-tissue mobility in relationship to the bone and short pedicle length. [70-73]
4.3.2. Type lllb
Patients with type lllb defects undergo resection of the entire maxilla in addition to exenteration of the orbit (also known as the extended maxillectomy). These defects are extensive and have both large-volume and large–surface area requirements. The palate needs to be closed; the medial wall of the maxilla often needs to be restored to maintain an adequate airway; and the often extensive external defect, which can involve the eyelids, cheek, and occasionally the lip, need to be reconstructed. In addition, the anterior cranial base in the area of the sphenoid is often exposed and coverage of the brain becomes essential ( Figure 19).
If the external skin of the cheek is intact, a rectus abdominis free flap with a skin island used to close the palate is a simple, straightforward solution. If the flap is not too bulky, a second skin island to restore the lateral nasal wall can be used. A third skin island can be used to provide closure of the external skin deficit if necessary.4 [73-75]
Shestak et al. successfully used the latissimus dorsi flap in three patients with type lllb defects to fill the orbital cavity, seal the palate, and recontour the soft tissue of the face and cheek.  The latissimus dorsi was used because of its bulk, reliable anatomy, and ample pedicle length.
Palatal closure has its advantages and disadvantages in these reconstructions. If the palate is not closed (and muscle alone is used to cover the brain), the resultant massive intraoral defect requires a very large obturator, which can be difficult to support if there are no teeth left in the remaining maxilla. Palatal closure, although not ideal, makes sense because these patients can usually speak well and eat soft solids without dentures. Denture fitting can be difficult if the skin bulges downward and there are no teeth to fit the prosthesis. However, because these patients would have similar difficulties with an open palate and function well even without a denture when closed, we feel that the palatal closure is generally advisable.
We do not attempt to reconstruct bony deficits in these patients because of the extensive nature of the defects. Bone-containing free flaps do not have the same versatility with regard to providing intraoral and extraoral lining and soft-tissue bulk and are therefore not generally indicated for the massive type lllb resections.
4.4. Type lV: Orbitomaxillectomy
Type lV or orbitomaxillectomy defects include five walls of the maxilla and the orbital contents, leaving the dura and brain exposed. The palate is usually left intact with these resections. Reconstructive objectives include the provision of adequate soft tissue and the resurfacing of external skin defects where necessary. Thus, a flap that provides a medium volume of soft tissue and has the potential to cover a medium/large surface area with one or more skin islands is required (Figure 20).
The rectus abdominis flap can meet these requirements. These are conceptually simple reconstructive procedures, but the principal challenge is technical; one needs to anastomose the flap to a donor vessel in the neck, as temporal and facial vessels are usually resected or are unreliable. Dissection of the rectus pedicle extends the length up to 20 cm. A superficial tunnel in the face-lift plane allows transfer of the vessels; or, if the maxillary tubercle is resected, access can be gained by a parapharyngeal approach medial to the mandible. Maintaining the nasal airway is often the most difficult problem in these patients; thus, a second skin island to address lateral nasal wall reconstruction is helpful. 
4.5. Reconstruction with vascularized autogenous tissue
Advances in tissue transfer techniques have made sophisticated reconstruction with autogenous tissues possible. In the past, it was thought that autogenous reconstruction after tumor surgery would interfere with examination for residual or recurrent disease. Advances in diagnostic techniques such as computerized tomography, magnetic resonance imaging, and endoscopy now enable the surgeon to evaluate the resection bed without direct inspection. [77-80] With the numerous free and pedicled flaps and the adjunctive modalities, such as enteral feeding tubes, tracheostomy, and osseointegrated dental prostheses now available to the reconstructive surgeon, many of the technical difficulties related to autogenous reconstruction can be circumvented, both in the perioperative period and over the long term.
The idea of “one wound one scar” has drastically altered our reconstructive approaches. Local flaps in extensive defects only make a defect a “larger” defect and a “larger scar” ensues and in extensive maxillary defects “new” tissue must be brought into the wound and enlarging the scar by local or adjacent flaps is not advisable. The free or prefabricated flaps are not the "last ditch measures" and they must be considered as the first line of treatment in these complex midfacial defects (Figure 21,22).
Figures 21 shows a war-wounded veteran after 25 operations by world famous surgeons; the midface defect has been treated by local flaps, the maxillary defect remains and maxillary nonvascularized bone grafts, have all resorbed, the face and forehead are scarred.
5. State of the art procedures: Flap prefabrication and prelamination
Flap prefabrication is a term that was first introduced and later clinically applied by Shen in the early 1980s.81-82 Flap prefabrication and prelamination are two closely related concepts. Clinical applications of flap prefabrication and prelamination are relatively new to the field of reconstructive plastic surgery. Although the two terms are often used interchangeably in the literature, they are two distinctly different techniques. Understanding their differences is helpful in planning the reconstructive strategy. They are primarily used in reconstructing complex defects where conventional techniques are not indicated.
5.1. Flap prefabrication
Flap prefabrication starts with introduction of a vascular pedicle to a desired donor tissue that on its own does not possess an axial blood supply. After a period of neovascularization of at least 8 weeks, this donor tissue can then be transferred to the recipient defect based on the newly acquired axial vasculature (Figure 23).
Cartilage and bone can be incorporated into these flaps but they are mostly suitable for ear and nose reconstruction and for maxillary or mandibular reconstructions the prelamination method is the better choice. Flap prelamination, begins with building a three-dimensional structure on a reliable vascular bed. This composite structure, once matured in approximately 6-8 weeks can then be transferred to the recipient defect.
5.2. Flap prelamination
Flap prelamination is a term first coined by Pribaz and Fine in 1994.  The definition of “lamination” means bonding of thin sheets together to give a multilayered construction. In reconstructive surgery, the term “flap prelamination” has been used to describe a process of two or more stages for constructing a complex three-dimensional structure. The first stage involves adding different layers to an existing axial vascular territory as composite grafts, allowing time for the tissues to mature before being transferred (Figure 24-26).
An intermediate stage may be needed to further modify the flap, such as thinning, delaying, or adding additional tissue.  At the next stage, when the remote composite flap is completed, it is transferred to the defect based on the original axial blood supply. As with any composite graft, these added layers have to be sufficiently thin or small for them to take. The rationale for prelaminating those layers at a different site before transfer results from the belief that this offers the best chance for the prelaminating layers to heal, stabilize, and assume their expected structures and positions if the construction is performed in a reliable vascular bed at a less conspicuous site instead of in situ, where local complicating factors can be numerous. This is particularly important for reconstruction of functional units that need to be transferred to complex local environments, where structural leaks may cause grave complications (e.g., neourethra in the perineum and neoesophagus in the mediastinum).
5.3. Flap maturation
Because the blood supply is not manipulated, the time for a prelaminated flap to mature is shorter than for a prefabricated flap,  usually between 4 and 6 weeks. Intuitively, this makes sense because it represents a similar amount of time for any composite graft to fully take, whereas in a prefabricated flap, neovascularization needs to take place over a much larger and sometimes thicker dimension of tissue. Intermediate manipulation may be required to obtain a thinner flap or to delay an extended portion of a flap or to add additional graft material (Figure 27).
5.4. Flap transfer
Because the layering of structures takes place in an established vascular territory, venous congestion is usually not a problem in a prelaminated flap as it is often in a prefabricated flap. However, all flaps, including prelaminated flaps, become edematous after transfer, and there is increased scarring at each tissue healing interface. In attempting to reconstruct complex three-dimensional structures, the multiple layers with scarring and contractile forces at each interface can result in distortion and loss of contour of the flap. Because of this, the initial result is often suboptimal, and generally several revisions are necessary. This occurs especially in the face, where prelamination is used for reconstruction of central facial features, such as the nose and surrounding tissues. Once the prelaminated flap is healed in place and a stable foundation has been obtained, the external part can be de-epithelialized and covered with local advancement flaps or, in the case of nostril reconstruction, with a forehead flap for final esthetic reconstruction (Figure 28,29).
6. Osseointegration techniques
The development of osseointegrated implants has revolutionized the approach to the dental rehabilitation of patients requiring maxillary reconstruction. The work of Branemark  and others has resulted in the development of the materials and techniques necessary to provide predictable and reliable implants that can be completely incorporated into grafted bone and support a fixed and stable dental prosthesis. [87-94] The use of osseointegrated implants in conjunction with free tissue transfer represents state-of-the-art reconstruction of large maxillary defects. The use of osseointegrated implants for dental rehabilitation has previously been much more extensively discussed in the context of mandibular reconstruction than that of maxillary reconstruction.  Some fundamental concepts of functional dental restoration with prosthetics should be understood. The reconstruction should provide for retention, support, and stabilization of the denture. Retention involves preventing the displacement of the prosthesis from the denture-bearing surface. Support implies that masticatory forces should not cause the prosthesis to impact vertically against the soft tissue of the load-bearing surface. Stabilization refers to the prevention of excessive lateral movement of the prosthesis. Dentures may be implant-borne, in which case the osseointegrated implants completely retain, support, and stabilize the prosthesis, or implant-retained, in which case the support and stabilization functions are shared by the denture-bearing surface and the retention of the prosthesis is completely dependent on the osseointegrated implants. Dentures that do not require osseointegrated implants are tissue-borne and tooth-supported, relying on the native tissues for retention and stabilization. 
Tissue-borne prostheses generally cannot be used in extensive maxillary defects because of insufficient residual palatal and alveolar tissues to provide support and retention. Funk et al.  defined such defects as those involving more than two-thirds of the maxillary arch. These defects typically require surgical reconstruction of the maxillary arch to provide neoalveolar bone of adequate thickness (approximately 10 mm) to accommodate osseointegrated implants, support a denture, and prevent its movement during mastication (Figure 30-34).
Bony reconstruction of the maxillary arch allows placement of the osseointegrated implants axial to the occlusal forces, a key factor for successful implant function.  Osseointegrated implants may be placed at the time of the reconstruction or secondarily, 6 to 8 weeks later.  Three to 8 months after placement, the osseointegrated implants are uncovered and prepared for final prosthetic reconstruction by a prosthodontist. 
The use of free tissue transfer techniques in combination with osseointegrated implants for maxillary reconstruction has been reported by various authors.  Holle et al.  described a two-stage procedure for the reconstruction of maxillectomy defects. Initially, an osseous flap was created from the lateral border of the scapula; it incorporated osseointegrated implants, was covered with skin grafts, and was protected with a PTFE membrane. Three months later, the flap was harvested and transferred to the face using a microsurgical technique. This procedure successfully restored facial contour and allowed full dental rehabilitation. Funk et al.  used free scapular osseocutaneous flaps with primary or secondary osseointegrated implants for large palatomaxillary defects in three patients. These patients all underwent successful dental rehabilitation, with 94 percent stability of the implants at an average of 18 months after the completion of rehabilitation. Nakayama et al.  reconstructed a bilateral maxillectomy defect with a free fibula osseocutaneous flap combined with osseointegrated implants. Igawa et al.  recently reported the use of a prefabricated iliac crest free flap, which was secondarily vascularized by a rectus abdominis muscle flap and covered by split-thickness skin graft, with the secondary placement of osseointegrated implants for functional alveolar ridge reconstruction after hemimaxillectomy.
Maxillary defects are one of the most challenging problems facing the reconstructive surgeon. Microsurgical tissue transfers evolved from the groin flap transfer to the complicated flap prefabrication and prelamination approaches to difficult reconstructive needs. These sophisticated techniques are distinctively different and yet can be perfectly complementary. Prelamination can add virtually anything to where there is a good axial blood supply, and prefabrication can bring an axial blood supply to almost anywhere in the body. The two techniques can even be combined when certain complex reconstructive needs are present. Prefabrication and prelamination can also serve as a conduit through which products of tissue engineering and embryonic stem cell technologies can be applied to the reconstruction of head and neck defects. Tissues synthesized in vitro with better structural, color, texture, and functional match can be prelaminated to a site that has already been prefabricated. Prefabrication of a bioabsorbable matrix system can create a well perfused scaffold to which more and larger subunits can be prelaminated.
As our understanding of the techniques evolves, the breadth of their usage will also expand. These techniques will continue to be useful to help solve many difficult problems that baffle even the very best reconstructive surgeons, and the potential for these techniques may be used to bring tissue engineering from the laboratory to clinical reality. Lastly, as progress is made in transplant pharmacology, the immunologic barrier to feasible composite tissue allograft transplantation may be overcome. This represents the beginning of a new era in reconstructive surgery.