Esthetic and Functional Reconstruction of the Posttumoral Interrupted Mandible With Double-Barrel Fibular Free Flap Graft: Rationale for a Microsurgical and Prosthodontic Approach

The revascularized fibular transplant is considered the gold standard over other reconstructive techniques for segmental mandibulectomy when the defect is more than 6 cm long.1–4 Its double medullar and periosteal vascularization allows this long bone to be shaped into a good morphologic and occlusal fit at the receiving site. This adaptive donor site may provide muscle for filling and tissue for coverage (skin, aponeurosis). Whereas the esthetic imperatives for these reconstructive techniques are well defined,1–4 few teams have addressed the functional—and therefore prosthetic—considerations. The size of a simple fibular bone graft makes it difficult to envisage an implant-supported fixed partial denture under optimal conditions; for this reason, some authors have proposed a secondary distraction osteogenesis of the fibular graft,5–7 or the direct use of the double-barrel technique on the initial graft.8,9 In this technique, 2 fragments of fibula (the 2 bone barrels) are shaped and adapted to each other and transfixed side by side with osteosynthesis screws in order to mimic the contour and volume of the mandible; this double-barrel fragment can then replace the resected mandible portion and reconstitute a more natural bone height than a fragment of fibula alone. Unfortunately, final results often diverge, sometimes significantly, from the original anatomy. The challenge presented by these surgeries, which may involve young patients, is to enable an implant-supported fixed rehabilitation under the best conditions—as similar as possible to those encountered in routine implantology.

The novelty of this concept, as described below, lies in bringing together the latest advances in microsurgical reconstruction (double barrel technique) and the timing of prosthetic stages (immediate loading of implants). In routine implant practice, it is the type of prosthetic rehabilitation that determines the surgical treatment plan (the number and site of implants, the volume and spatial distribution of bone grafts, if used). Whenever the prognosis for causal pathology allows, this rule should mandate all mandibular microsurgical reconstruction modalities.

A man, originally from the French Caribbean, nonsmoker and with an unremarkable medical history, was first operated on in 1978 for enucleation of a bone cyst that had developed in the left mandibular corpus. The initial histologic examination was in favor of an ameloblastoma with a follicular architecture and a multicystic aspect. Clinical history revealed 5 recurrences that resulted in iterative enucleations and the avulsion10 of teeth from the mandibular left second molar to the left lateral incisor.

When the patient came for consultation in 2003, he was 49 years old. He presented with a voluminous mandibular tumefaction, 4 cm in diameter, between the central incisor and the left wisdom tooth (Figure 1). An interruptive mandibulectomy was performed; the canine on the opposite side and the third homolateral molar were left in place. The bone resection included the symphysis and all of the left mandibular body, leaving a mucosal defect of 4 × 8 cm, which corresponds to type La/Aa in Cariou's classification.3 

Reconstruction was performed during the same surgery using an osteo-aponeurotic left fibular transplant. This was adjusted in situ, with a resin stereolithographic template, made preoperatively, based on computerized tomography data in DICOM format and using an appropriate software (Facilitate software, Astra Tech AB, Mölndal, Sweden). Three osteotomies were necessary, followed by bone modeling to obtain 4 fragments respectively of 4, 3, 3, and 4 cm long (from proximal to distal). The 120° axial angulation between the symphysis and the horizontal branch was reproduced; osteosynthesis was achieved using titanium miniplates. Both distal fragments were finally synthesized in a double barrel,11 the fibular crests being adapted to each other through osteoplasty followed by osteosynthesis.

The fibular flap was removed with an aponeurotic platform that replaced the resected gingival and soft tissues. Magnifying optics were used to perform lateroterminal vascular microanastomosis from the graft to the external carotid artery and internal jugular vein.

After allowing 3 months for mucosal scarring and bone consolidation, the vitality of the transplant was clinically determined by the characteristic appearance of the aponeurotic paddle. Vitality was also confirmed with dynamic magnetic resonance imaging, which showed that the vascularization of the reconstructed bone was similar to the vascularization of the native mandibular stump.12 Implant planning was undertaken with the Facilitate software (Figure 2). Dental implants were thus inserted 3 months after reconstruction. The crestal incision and periosteal flap were limited; it confirmed that the tissue obtained through healing of the aponeurotic paddle was similar to the initial oral mucosa.

Six implants (OsseoSpeed, Astra Tech AB) were placed: 2 fixtures of 4 mm in diameter and 13 mm in length in the symphysis, and 4 fixtures of 4 mm in diameter and 15 mm in length in the premolar and molar sites (Figure 3). It is important to choose implants that are sufficiently long so their osseointegration and stability in the fibula are optimized.13 Their excellent primary stability (torque ≥45 Ncm for each of the 6 implants) allowed—as anticipated—immediate implant loading with a temporary resin prosthesis reinforced with a metal framework (Figure 4). The implant-supported fixed prosthesis was screwed in a totally passive way. Eating was resumed that same evening and food remained soft for 4 weeks. Patient follow-up and dismounting of the temporary prosthesis was uneventful. Eight months after the implant surgery, impression was performed and the final prosthesis was made, using the temporary bridge as a valid model. The permanent ceramometallic bridge was finally placed (Figures 5 and 6). Patient follow-up is currently 5.5 years after implant loading (Figure 7). While it is still relatively early, all the Albrektsson criteria for success14 are met (particularly with stable peri-implant bone levels assessed on retroalveolar X rays), and the patient has regained all chewing function that had degraded progressively over 30 years.

Ameloblastoma is a benign tumor, but according to Gardner,15 55% to 90% of these lesions relapse after a simple enucleation. This patient history illustrates the recidivist potential and aggressive pattern of these tumors. To minimize this risk, Feinberg and Steinberg16 recommend an interruptive bone resection with safety margins of 1 to 2 cm, which were implemented on this patient.

The extension (symphysis and mandibular horizontal branch) and the composite (bone and mucosa) nature of the loss of substance in this still young and otherwise healthy patient justified a free micro-anastomosed flap reconstruction. The donor sites—iliac, scapular, and fibular—provide sufficient bone volume. However, only the fibula allows a mucosal reconstruction compatible with an implant-supported fixed rehabilitation of good quality due to the thinness of its aponeurotic component, one of several key factors that justified its use here.17 As our goal was also to achieve a prosthesis that most closely resembles what is achieved in routine implant practice, we settled on a fixed bridge over implants, with no fake gum. This requires bone reconstruction to be perfect, both at the basal and crestal level as would be the case for all pre-implant bone augmentation technique surgery. The wax-up studies and software simulation confirmed the need for a particularly fine-tuned work-up for this patient. The fibula was adapted to the recipient site using a double barrel technique that was first described by Horiuchi et al18 and later codified by Bähr et al.11 Though delicate to perform, this technique is increasingly used for these defects.9,19–21 It allows the available bone height to be doubled (up to 3 cm) and to rebuild the entire mandibular height: basal bone and alveolar bone. Available bone mass was quantitatively sufficient and topographically suited to receive several implants, and the 6 implants were inserted ideally following the planned prosthesis.

Loss of mucosa was rebuilt with an aponeurotic paddle, which has the advantage of being finer than the fascio-cutaneous paddle described by other teams.8,9 This paddle distributes more naturally over the double barrel fibular bone graft and reconstitutes a real crest with a vestibule and a pelvilingual sulcus. The quality of this mucosa, even if it is not an attached mucosa, is far superior to the cutaneous paddle; it allows the subsequent placement of implants, often without further mucosal adjustment.

Despite the constraints linked to the cutaneous, bone, and mucosal defects, together with the patient and the prosthodontist, we ruled out the implant-stabilized removable prosthesis and even the Branemark-type fixed denture. Instead, we chose at the start of treatment an implant-supported fixed bridge without fake gum, in order to achieve high-quality routine implantology results for this young, partially edentulous patient.

Insertion of implants in the transplant was done 3 months after surgery. Some teams, however, place implants on the same day as reconstruction.22 Requirements, in terms of prosthetic results, led us to wait and allow for bone consolidation and mucosal healing. Indeed, after 3 months, there was no residual macro- or micro-movement of bony pieces, and the mucosa had healed even though it had not keratinized. It seemed to be the ideal time to place the implants, directed by the surgical guide and the definitive occlusal relation.

It is not unusual for the mucosa to heal in an aberrant mode; after reconstruction, and even after implant placement, peri-implant granulomas have been described.23 Keratinized mucosa grafts can be useful to create a biologic space that is more favorable for durable fixtures. In this patient's case, the quality of the aponeurotic paddle epithelization allowed to avoid a palatal fibromucosal graft.

Implantation of free transplants is a simple evolution from the development of dental implants in the edentulous patient.24 In 1992, free fibular transplants were implanted in order to stabilize removable dentures.25 Since then, the experiences of microsurgical teams26–30 have established that implantation on transplanted bone is clearly reliable (with success rates close to the results on native jaw). Implant survival rates in fibular bone, using Albrektsson's success criteria,14 range between 96.1% and 98.6% for the first 2 years. It should be noted that these encouraging success rates were achieved without any surgical mucosal adjustment.

In the present case, this patient from the French Caribbean was being treated 8000 km away from home; this situation reinforced the decision for an immediate implant loading. The patient stayed several weeks in Paris after implantation for a thorough early follow-up, but the immediate loading allowed him to avoid several transatlantic travels (for the connection of the transgingival screws and the preparation of the temporary implant-supported fixed prosthesis). The use of bicortical anchorage of implants may seem unusual, but fibular bone has quite different characteristics from the native mandible and presents a cortical thickness of 4 mm and no medullar bone. With these particulars in mind, we made the empirical decision to go with bicortical and even tricortical anchorage of the implants, as has been confirmed by the first animal studies on this type of bone.13 The primary stability thus obtained was very satisfactory. The good bone quality and the absence of infection and parafunction were further indications for immediate loading.31–33 

In accordance with current consensus,34 an important number of implants were inserted (6 implants for a partial bridge of 9 teeth). Their excellent primary stability allowed immediate implant loading with a temporary resin prosthesis reinforced with a metal framework that held the implants in place, limited their micro-movements, and allowed good bone healing.35,36 In accordance with the results of Chiapasco and Gatti,37 whose experience with 2 maxillary full bridges was a success, the immediate loading of a partial mandibular bridge on a free fibula transplant appeared in this case as a reliable option.

Mandibular reconstruction using the fibular free flap is a technique that is well mastered; this surgery is ready to enter the era of optimization. Patients no longer need “make do” with reconstructions that are solely cosmetic: surgeons have the means to reconstruct, almost entirely, wide defects so that the dentoalveolar elements can be restored very realistically. The challenge in the years ahead lies in making this meticulous reconstruction more widely available, bringing together the esthetics of facial restoration and the rehabilitation of mandibular function.