To choose an appropriate reconstruction method, the following factors should be considered in reconstructing the mandible: (1) the length and location of the mandibular defect, (2) associated soft tissue loss, (3) the overall health and well-being of the patient, (4) the patient's potential prognosis, (5) potential donor sites, (6) primary versus delayed repair, and (7) the patient's dental health and potential for dental rehabilitation.
The easiest form of mandibular reconstruction is no reconstruction. That is, when faced with a segmental defect, simply close the surrounding soft tissues over the defect, leaving one or two “free-swinging” mandibular segments. This leaves the patient with a significant cosmetic and functional deficit, although for a small lateral defect in an edentulous patient, the cosmetic and functional deficit may be smaller than expected. Certainly, however, for a person missing a large segment of mandible or the anterior segment of the mandible, this leaves a significant deformity where the lower lip and the mentum are extremely retrusive, a situation known as the “Andy Gump deformity” (Figure 24–1).
An Andy Gump cigar label. Note how the character's face appears to end at the upper lip (absence of the mandible). Ironically, the character is smoking a cigar.
The earliest soft tissue closures used local tissues, cheek or tongue flaps, or pedicled flaps from the neck, scalp, forehead, or deltopectoral region, many of which required a staged reconstruction. However, without any rigid structural support, neither the form nor function of the mandible was reliably reconstructed with these methods. Sometimes, bone fragments pedicled on local flaps were tried, although they had less than desired reliability, particularly in the face of radiation before or after surgery.
The advent of alloplastic implants helped correct some of the problems that soft tissue closure alone did not address. These include steel, titanium, and other alloy (eg, Vitallium) implants that are fashioned into either a bar or a tray, which can then be conformed to the shape of the missing mandibular segment. Over time, titanium has become the most common metal with which to fashion implants as it retains strength, biocompatibility, and rigidity, but can still be contoured using handheld instruments. All alloplastic implants can eventually suffer metal fatigue and fracture owing to the repetitive stress put on the material through mastication. Unfortunately, materials strong enough to withstand the forces without the risk of fracture are too strong to be contoured in the operating room by the surgeon.
Mandibular replacement with alloplastic implants can provide a rapid, effective mandibular reconstruction without a secondary donor site defect. However, in addition to the risk of plate fracture, there can be a significant risk for plate extrusion and exposure with subsequent infection (Figure 24–2). Experience has demonstrated that a mandibular reconstruction plate, particularly if “wrapped” or otherwise insulated with a muscle pedicle flap (eg, a pectoralis major flap), is an adequate reconstructive option for lateral mandibular defects. Although microvascular free-tissue transfer provides some improvements and benefits, for a lateral defect, a mandibular bar is an acceptable contemporary reconstruction. However, for defects involving the anterior mandible, as well as the symphysis and the parasymphysial regions, mandibular reconstruction with a metal bar has a significantly higher risk of complications than reconstruction with revascularized bone. This may be due to the increased arc of rotation that the bar passes through at the anterior mandible, which causes excessive force on the overlying soft tissue, eventually leading to bar exposure. Also, unfortunately in many cases, metal bars without underlying bone grafting seem to have an increased chance of becoming exposed following radiation treatments. This can create a very complicated wound-healing challenge, typically requiring removal of the implant.
Patient after partial mandibulectomy with alloplastic (titanium) implant extruding a year after surgery. Note granulation tissue and purulent drainage.
Alloplastic trays filled with bone chips have been used and in some cases have been successful, although some physicians have noted that 50% or more of their patients end up with an unsatisfactory result of using this method. Other times, bone grafts can be used in the form of cancellous bone chips without a tray. In addition, irradiated bone grafts are used. However, in all of these cases, the grafted bone serves as a scaffold for osteoblasts to create new bone, and in the inevitably infected field encountered at the time of ablation, new bone growth is unpredictable and unreliable. Furthermore, irradiated fields create an additional impediment to good bone healing.
If bone substitutes (eg, metal bars), and free-bone grafts (in the form of chips or irradiated grafts) are unreliable, the next thing to try would be vascularized bone. Indeed, a variety of pedicled and free-bone flaps have been utilized.
Initially, pedicled bone flaps were used. Physicians have tried to rotate the clavicle on the sternocleidomastoid muscle, the trapezius muscle, or even on the deltopectoral flap. Regrettably, only mixed success was obtained because the blood supply to the bone in each of these situations was unreliable and random. Somewhat better results were obtained with the pectoralis major muscle with the fifth rib, but again, only unreliable results were achieved. Rib grafts were also pedicled off of the latissimus dorsi muscle, but are not a great choice because, in general, this flap provides an unnecessarily large amount of muscle and soft tissue, with a usually inadequate amount of bone to provide a good reconstruction. Better results have been obtained by transferring the spine of the scapula onto the trapezius muscle. This flap provides approximately 10 cm of bone, and as long as the transverse cervical vessels are not injured during any part of the ablative procedure, the flap has fair reliability.
At various times, a series of sliding osteotomies has been designed for use in the remaining mandible to allow the bone to be advanced to fill in gaps. Although interesting, the nature of the mandibular defect and the subsequent radiation may make these osteotomies unreliable.
To date, the best results have been achieved by free-tissue transfer. This technique provides both vascularized bone and soft tissue and has no restriction on pedicle range or length.
Free-tissue transfer techniques were not widely known even a few years ago, but at this time most academic medical center departments of otolaryngology have at least one surgeon who is trained in microvascular methods. Unfortunately, microvascular transfer remains a relatively long and complex procedure, and although certainly valuable for large defects, it is more difficult to decide what to do for small defects. Often the extensive surgery required for a “free flap” seems to be too much when faced with a small anterior defect, however; still, no better alternative is available. Several flaps have been tried, and the four most commonly used osseous free flaps are (1) the radial forearm, (2) the scapula, (3) the iliac crest, and (4) the fibula. Each differs in the amount and nature of the soft tissue and bony components. All of these flaps, except the scapula flap, are sufficiently distant from the head and neck to allow for a second team of surgeons to (conveniently) simultaneously harvest the flap while the ablation is being performed.
The radial forearm flap allows the transfer of a large amount of pliable thin fascia and skin from the ventral surface of the forearm. Arterial supply is through the radial artery; therefore, an Allen test must be carefully performed before harvesting this flap to be certain that the hand has adequate vascular supply from the ulnar artery alone. Venous drainage is through the vena comitans of the radial artery or through the cephalic vein. Approximately 10 cm of bone can be taken. Although the bone is strong cortical bone, it is not thick because only one-third of the cross-sectional area of the radial bone can be taken without greatly increasing the risk of stress fractures of the forearm. Tapering the edges of the graft in a “boat tail” fashion further reduces the risk of postoperative fractures, as does a prolonged immobilization of the arm in a splint (3 weeks or longer). Overall, since only a small amount of bone is obtained, this flap is useful for only certain mandibular defects and is probably best suited for reconstruction in which a large amount of soft tissue is required with only a small segmental mandibular defect.
The scapular flap is among the most versatile of free flaps, since a very large amount of soft tissue is available with the bone. Unfortunately, the usual need to change the position of the patient during surgery from supine to lateral to harvest the flap makes this flap less desirable to harvest than its usefulness might suggest.
The lateral scapula provides 12 cm of bone that can support an osseointegrated implant for dental rehabilitation (unlike the bone of the radial forearm flap). The circumflex scapular system provides the blood supply to the flap and, with dissection, the bone and large skin islands can be harvested off of the subscapular artery. Two venae comitantes (veins that travel closely in approximation to the artery) accompany this artery for venous drainage. In general, the scapular system of flaps may be the most versatile of all reconstructive options, providing a good amount of bone and the most independently mobile soft tissue components of any of the osseous composite flaps.
Based on the deep circumflex iliac artery and vein, the iliac crest flap has proved to be quite useful for mandibular reconstruction. Because of the great variety in which the bone can be harvested and contoured, three-quarters or more of the mandible can be reconstructed with this flap (Figure 24–3). Furthermore, the natural curvature of the iliac crest bone can be used to help approximate the natural shape of the mandible. The bone is thick and can more than make up for the thickness of the mandible (Figure 24–4). A relatively thick and nonpliable skin flap can be harvested with the iliac crest, although it is often helpful to use a Doppler probe to initially identify perforating vessels to the skin.
Inset iliac crest bone flap at the left mandibular angle. Note how the bone can be contoured without osteotomies due to the large bone stock available at the iliac crest.
Comparison of cross sections of a mandible (left) and an iliac crest (right). Note that the iliac crest bone is more than thick enough to recreate the mandible and accept implants for dental reconstruction.
The versatility of this flap was greatly enhanced when it was noted that the internal oblique muscle is reliably vascularized by an ascending branch off the deep circumflex iliac artery. This provides a thin, pliable muscle flap that can be used to reconstruct a soft tissue defect. For example, with a “through and through” defect of the lateral mandible and cheek, three effects are achieved: (1) the iliac crest bone can replace the mandibular bone, (2) the skin paddle can replace the skin of the external cheek, and (3) the internal oblique muscle can be used to reconstruct the mucosal surface defect and then left either to have mucosa grow over it or to be covered with a skin graft. Removing the internal oblique muscle necessitates great care in closure to prevent an abdominal hernia.
Another excellent use for the iliac crest flap is for the reconstruction of a near-total glossectomy with mandibulectomy. In this situation, the iliac crest bone can be positioned transversely so that the bone forms the floor of the mouth. This then elevates the soft tissue skin flap of the iliac crest flap into a good position to assist with swallowing once it is fashioned into a “neo-tongue.”
Probably the most commonly used free-tissue flap for mandibular reconstruction, the fibular flap has several advantages. A very long segment of bone is available (approximately 25 cm), since the entire fibula can be harvested except for 8 cm, which should be preserved at the proximal and distal ends for joint stability (Figure 24–5). In addition, a reliable skin paddle is obtained and additional vascular soft tissue is available as the flexor hallucis longus muscle can be harvested with the flap (Figure 24–6).
Harvesting of the right fibula. The bone and skin have been isolated on the right peroneal vessels.
Postoperative photograph of viable skin from the fibula flap reconstructing the mucosa of the left “alveolar ridge” over the fibula bone.
The fibular flap is based on the peroneal artery and veins. Preoperative vascular imaging is helpful to protect vascularity to the foot because vascular disease and anatomic irregularities can eliminate the normal three vessels that supply blood to the leg. Angiograms were once routinely ordered, although magnetic resonance imaging can be modified in protocol to provide adequate imaging of vascular anatomy. In addition, if desired, this flap can have sensory reinnervation through the lateral cutaneous branch of the peroneal nerve. Blood supply to the bone is through the periosteum; therefore, as long as most of the periosteum is not disturbed, numerous osteotomies can be made into the harvested fibula bone to allow for good custom contouring of the bone in recreating the mandible (Figure 24–7).
Fibular flap contoured to reconstruct the missing segments of the mandible. This tissue will then be inset and microvascular anastomoses will be performed to recreate a vascular supply.
Distraction osteogenesis is a new technique that has some value in mandibular reconstruction. In this technique, an appliance is attached to the mandible, and a thin piece of the end of the mandibular segment is cut free from the rest of the mandible. This thin segment is slowly advanced through the use of a “key” attached to the appliance. As it is advanced, the space between the advancing segment and the bulk of the mandible is filled in with new bone. Once enough new bone has been made, the free ends are “roughened” and then the remaining segments of mandible are plated, as for a mandibular fracture. Although exciting in concept, this technique does not allow for primary reconstruction because the distraction process takes time. Plus, at least one additional procedure is required. Thus far, the technique has been used primarily in cases of congenital mandibular insufficiency.
Of current interest is the possibility of creating new bone through the use of growth factors and hydroxyapatite mixtures. The nature of the expression of the various bone morphogenic proteins is becoming increasingly understood, and soon it may be likely to bridge a bone gap with a bone-like powder that is mixed with bone growth factors, resulting in new, strong bone within a predictable period. Certainly, the ability to generate bone is far closer than the ability to generate good soft tissue coverage; therefore, the use of the osseous microvascular flap may be limited.