Transport Osteogenesis in the Maxillofacial Skeleton: Outcomes of a Versatile Reconstruction Method Following Tumor Ablation

Objectives  To report our clinical experience using bifocal distraction osteogenesis (BDO) with internal devices to treat patients having bony defects of the maxillofacial skeleton following tumor ablation and to focus on outcomes of dental implant placement in patients having maxillomandibular segmental defects.

Design  Retrospective case series.

Setting  Academic research.

Patients  Patients were selected according to the following inclusion criteria: a bony defect in the maxillofacial skeleton, moderate soft-tissue defect, local or general conditions that preclude more aggressive surgery, and adequate patient compliance. Types of BDO included horizontal mandibular or maxillar alveolar, bilateral alveolar, vertical mandibular or maxillar, ramus and body, mandibular angle, symphysis, the 2-step procedure, temporalis muscle flap reconstruction, vascularized free–fibular flap reconstruction, radial forearm free-flap reconstruction, and pectoralis muscle flap reconstruction.

Main Outcome Measures  The latency period was 10 days, after which distraction was initiated at a rate of 0.5 mm/d. The distraction period continued until the transport disk reached the distal stump. The consolidation period ranged from 8 to 48 weeks. Seven patients required additional bone grafting to complete union with the residual bone.

Results  Twenty-eight patients having bony defects of the maxillofacial skeleton underwent BDO. The mean (SD) bony defect length was 47.0 (20.1) mm. The mean (SD) distracted bone lengthening was 36.5 (20.0) mm, with a mean (SD) consolidation period of 16.4 (8.0) weeks. The bony defect involved the hemimandibular body in 12 patients, with greater involvement of the body and symphysis in 4 patients and of the bilateral mandibular body in 2 patients. Complications after BDO included the following: discomfort in 8 patients, complete intraoral exposure and infection in 3 patients, partial cutaneous exposure in 1 patient, premature consolidation in 1 patient, and temporomandibular joint ankylosis in 1 patient. Overall, BDO for reconstruction of bony defects was successful in 22 patients and failed in 6 patients. At the end of the follow-up period, 11 patients had undergone endosseous dental implant placement.

Conclusions  Bifocal distraction osteogenesis potentially benefits patients with bony defects following tumor ablation at various locations in the maxillofacial skeleton. Sufficient bone is gained to allow dental implant placement, an important functional outcome.

Since first described by McCarthy et al¹ in 1992, distraction osteogenesis (DO) of craniofacial bones has increasingly become a mainstay in bone regeneration. In patients having bony defects of the maxillofacial skeleton after tumor ablation, bifocal DO (BDO) allows regeneration of loosened bone following osteotomy and gradual separation of bony fragments. Distraction osteogenesis for reconstruction of segmental mandibular defects has been previously described.^2,3

The use of microvascular free flaps has been successful for reconstruction of large segmental mandibular defects.⁴ A vascularized free-fibular flap (VFFF) aids in reconstructing bone and soft tissue in most adult patients undergoing mandible ablation in a single setting. However, the following 2 conditions preclude the use of mandibular BDO, despite the availability of a composite microvascular free flap: (1) patients with increased surgical risk who are unsuitable for prolonged surgical time or for more aggressive surgical procedures and (2) patients who require improved alveolar soft-tissue quality before endosseous dental implant placement. Bifocal DO allows reconstruction of segmental bony defects with an alveolar height that closely mimics the native mandible. It also provides attached gingiva that resemble the original. This is of utmost importance, as overlying soft tissue is often too thick following microvascular free-flap reconstruction, leading to unfavorable dental implant placement and maintenance because of the absence of attached gingiva.

Most mandibular defects or other defects involving the maxillofacial skeleton require reconstruction of bone and surrounding soft tissue, for which BDO and VFFF reconstruction are effective. A previous study³ focused on patients with adequate compliance and limited soft-tissue defects for inclusion in a distraction osteogenesis protocol. However, few investigators have reported patient inclusion criteria for various types of reconstruction; therefore, controversy exists about reconstruction of the mandible and other craniofacial bone defects.

In the present study, we report our clinical experience using BDO with internal devices to treat patients having bony defects of the maxillofacial skeleton following tumor ablation. We also evaluated the outcomes of endosseous dental implant placement as ideal restorative treatment.

Patients treated from January 1, 2006, to December 31, 2008, were retrospectively evaluated and were selected according to the following inclusion criteria: (1) bony defect in the maxillofacial skeleton, (2) moderate soft-tissue defect, (3) local or general conditions that preclude more aggressive surgery, and (4) adequate patient compliance. Bone transport was used to reconstruct defects in all patients. The following clinical and treatment data were examined: sex, age, primary and secondary diagnoses, defect location and length, transport disk length, distracted bone lengthening, consolidation period, device location, number of devices, macroscopic quality of newly generated bone, radiation therapy before BDO, complications, success rate, follow-up period, bone grafting, and dental implant placement.

Primary diagnoses included squamous cell carcinoma, giant cell granuloma, odontogenic myxoma, ameloblastoma, sarcoma, leiomyosarcoma, mucoepidermoid carcinoma, cleft palate, periimplantitis, and meningioma (Table 1). Secondary diagnoses included VFFF reconstruction failure, oroantral fistula, soft-tissue and hard-tissue defects, radionecrosis, and recurrence of second malignant neoplasm. In all patients, computed tomography was performed before surgery to determine the extent of the defect and the involvement of surrounding tissue. Panoramic radiographs were obtained monthly during the distraction and consolidation periods to evaluate the progress of BDO and to determine the vector of the distracted transport disk. Types of BDO included horizontal mandibular or maxillar alveolar, bilateral alveolar, vertical mandibular or maxillar, ramus and body, mandibular angle, symphysis, the 2-step procedure, temporalis muscle flap reconstruction, VFFF reconstruction, radial forearm free-flap reconstruction, and pectoralis muscle flap reconstruction.

During the procedure, periosteum over the designed transport disk was preserved to ensure its vascularity. In 1 patient, periosteum or dura matter was preserved at the opposite side of the transported bone. An osteotomy was performed at a variable distance from the defect to create the transport disk. To create new bone in the gap, we used 2 types of unidirectional distraction devices (Cranimaxillofacial Distraction System; Synthes Inc, West Chester, Pennsylvania; and Modus MDO 1.5; Medartis AG, Basel, Switzerland). The movable part of the distraction device was fixed to the transport disk by 2 bicortical 10-mm to 14-mm screws. Two additional bicortical screws fixed the device to the residual bone. After verifying the mobility of the distractor, the device was returned to its original position with the distractor activator placed intraorally or percutaneously. Some challenging cases (eg, patient 4 in Table 1 with sarcoma of the maxilla and failure of VFFF reconstruction) required the use of a stereolithographic model before surgery to design adequate positioning of the BDO device (Figure 1).

The distraction device allowed maximal distracted bone lengthening to 40 mm. For this reason, the following patients required 2 consecutive distraction devices at the same site: (1) patient 11 with an 80-mm calvarial defect, (2) patient 16 with an 80-mm mandibular angle defect, and (3) patient 19 with a 50-mm parasymphyseal defect. The following additional patients underwent reconstruction with 2 devices at 2 different sites: (1) patient 2 with a 70-mm left mandibular body and symphyseal defect; (2) patient 7 with an 80-mm mandibular body and symphyseal defect; (3) patient 9 with a 45-mm lateromaxillary defect; (4) patient 12 with a 20-mm premaxillary defect; (5) patient 18 with a 30-mm right mandibular body and a 40-mm left mandibular angle defect; and (6) patient 21 with right 25-mm and left 26-mm mandibular angle defects. The following patients required the use of 2 simultaneous distraction devices at the same site to achieve adequate bone outcome, followed by the use of 2 other distraction devices because of the presence of defects larger than 40 mm: (1) patient 5 with a 60-mm premaxillary defect, and (2) patient 23 with a 45-mm mandibular body defect.

The latency period was 10 days, after which distraction was initiated at a rate of 0.5 mm/d. The distraction period continued until the transport disk reached the distal stump. The consolidation period ranged from 8 to 48 weeks. Seven patients required additional bone grafting to complete union with the residual bone as follows: (1) anteroiliac crest in patients 4, 5, and 18, (2) posteroiliac crest in patient 6 and patient 20, (3) mandibular ramus in patient 2, and (4) rib in patient 23. A VFFF was used in patient 13 because of complete failure of BDO. The rest of the patients underwent corticotomy in the distal stump of the transported bone and corticotomy in the residual bone. Both poles were fixed with titanium miniplates or bridging plates. Two patients (patient 14 and patient 16) underwent greenstick fracture at the proximal stump to achieve better curvature of the reconstructed mandible.

Commercially available statistical software (SPSS 13.0; SPSS Inc, Chicago, Illinois) was used to analyze descriptive statistics for continuous and categorical variables. Institutional review board approval was obtained for the study.

A 36-year-old woman (patient 2) with giant cell granuloma of the left mandibular body and symphysis underwent bone and soft-tissue resection and subsequent reconstruction using 2 horizontal alveolar BDO devices (Figure 2). The defect length was 70 mm. Two 10-mm transport disks were designed, and 2 opposing BDO devices were placed for 60-mm distracted bone lengthening. The consolidation period was 12 weeks, and excellent new bone was observed. Bone graft from the mandibular ramus was used to achieve proper union between the distracted bone segments. Six endosseous dental implants were placed over the distracted bone.

A 66-year-old man (patient 16) with squamous cell carcinoma of the oropharynx underwent bone and soft-tissue resection and further reconstruction using a pectoralis major pedicled flap (Figure 3). Following radiation therapy, the patient experienced radionecrosis of the mandible, which required segmental mandibulectomy and further reconstruction by BDO of the mandibular body and symphysis for a defect length of 40 mm. After a 12-week consolidation period, greenstick fracture was performed at the distal stump, and union with the remaining bone was fixed using a 2.0-mm titanium miniplate. No complications were observed, and endosseous dental implants were placed in distracted bone.

A 42-year-old woman (patient 9) with ameloblastoma of the right mandible underwent bone and soft-tissue reconstruction and subsequent reconstruction using horizontal alveolar BDO (Figure 3). A 30-mm defect was present. An 8-mm transport disk was designed to create a 25-mm bone segment. The consolidation period was 12 weeks, and excellent new bone was observed. No additional bone grafting was necessary. Subsequent restoration was achieved using 3 endosseous dental implants.

Twenty-eight patients having bony defects of the maxillofacial skeleton underwent BDO. Fifteen patients were male, and 13 patients were female, with a mean (SD) age of 51.79 (17.46) years (age range, 14-80 years). Primary diagnoses included the following: (1) squamous cell carcinoma in 15 patients; (2) giant cell granuloma in 3 patients; (3) ameloblastoma and cleft palate in 2 patients each; and (4) odontogenic myxoma, leiomyosarcoma, meningioma, mucoepidermoid carcinoma, sarcoma, and periimplantitis in 1 patient each. The primary defect location was the mandibular gingiva in 10 patients, floor of the mouth in 6 patients, maxilla in 5 patients, oropharynx in 2 patients, and skull, palate, retromolar trigone, base of tongue, and hypopharynx in 1 patient each. Local treatment of the primary defect consisted of the following: (1) bone and soft-tissue resection plus direct closure in 7 patients; (2) bone and soft-tissue resection plus VFFF reconstruction in 7 patients; (3) soft-tissue resection plus radial forearm free-flap reconstruction in 2 patients; (4) chemoradiotherapy in 2 patients; and (5) combinations of bone or soft-tissue resection plus pectoralis muscle flap, temporalis muscle flap, and bridging plates in the rest of the patients. Fifteen patients underwent modified type III radical neck dissection (Table 1). As adjuvant treatment following the first intervention, 12 patients underwent postoperative radiation therapy before BDO (mean [SD] radiation therapy dosage, 62.66 [3.93] J; range, 60-70 J) (to convert joules to grays, multiply by 1.0). Secondary diagnoses included the following: (1) radionecrosis in 7 patients, (2) oroantral fistula and residual soft-tissue or hard-tissue defects in 4 patients each, (3) relapse or second malignant neoplasm in 3 patients, and (4) loss of VFFF in 2 patients. Eight patients underwent primary BDO without a secondary diagnosis. The mean (SD) follow-up period was 41.09 (19.86) months (range, 4-60 months). At the end of the follow-up period, 5 patients had died of their disease, 21 patients had no evidence of disease, and 2 patients were alive with disease.

The mean (SD) bony defect length was 47.0 (20.1) mm (range, 20-80 mm). The mean (SD) transport disk length was 13.7 (8.0) mm (range, 4-29 mm). The mean (SD) distracted bone lengthening was 36.5 (20.0) mm (range, 15-80 mm), with a mean (SD) consolidation period of 16.4 (8.0) weeks (range, 8-48 weeks). The bony defect involved the hemimandibular body in 12 patients, with greater involvement of the body and symphysis in 4 patients and of the bilateral mandibular body in 2 patients. Two patients had mandibular symphyseal region involvement, and 1 patient had mandibular angle involvement. Three patients had premaxilla involvement, and 1 patient had premaxilla and lateromaxilla involvement. One patient had a parietofrontal calvarial defect (Table 2). Various BDO device locations were chosen relative to the bony defect. The most frequent transcutaneous locations for the device were the gonial region (6 patients) and the symphyseal-parasymphyseal region (8 patients). Other intraoral locations allowed more favorable manipulation of the device, including the lateromaxilla (5 patients) and the mandibular body (5 patients). Six patients required 2 BDO devices.

The overall success rate of BDO in the series was 79% (22 of 28 patients). Bone formation failed to occur in 6 patients with complete intraoral exposure and infection, screw loosening, or poor general health. According to surgical findings after the distraction period, newly generated bone was macroscopically good or excellent in 20 patients and poor or moderate in 6 patients. Seven patients required additional bone grafting to complete union between the distal stump and the distracted bone. The rest of the patients underwent corticotomy in the distal stump of the distracted bone and corticotomy in the residual bone. Both poles were fixed with miniplates or bridging plates (Table 3).

Complications after BDO included the following: (1) discomfort in 8 patients, (2) complete intraoral exposure and infection in 3 patients, (3) partial cutaneous exposure in 1 patient, (4) premature consolidation in 1 patient, and (5) temporomandibular joint ankylosis in 1 patient. Overall, BDO for reconstruction of bony defects was successful in 22 patients and failed in 6 patients. At the end of the follow-up period, 46 endosseous dental implants (MG Osseous; Mozo-Grau SL, Valladolid, Spain) had been placed over distracted mandible or maxilla in 11 patients (Table 3).

Several studies^1-4 have focused on the possibility of reliable segmental mandibular reconstruction following tumor ablation. Because some patients undergo BDO as a secondary procedure, reconstructive goals in terms of function and aesthetics are of utmost importance. Adequate masticatory function is desirable. As previously noted by Costantino et al,⁴ the diameter of newly generated bone is similar to that of the residual mandible and transport disk. Moreover, the inferior alveolar nerve and artery are recannulated in most cases.⁵ These conditions favor placement of an ulterior endosseous dental implant, which may lead to better results in terms of function.

In the present series, BDO was usually performed as a secondary procedure following severe complications or failure of the first treatment. Reconstruction using the VFFF was performed in defects involving the mandible (6 patients) and the maxilla (1 patient). Two patients with segmental mandibular defects experienced failure of the VFFF, while the patient with a defect of the maxilla developed an oroantral fistula. Mandibular osteoradionecrosis and development of a second malignant neoplasm occurred in the rest of the patients. Therefore, BDO was performed as secondary surgical treatment for these patients with recalcitrant disease. Consideration of BDO as a secondary choice in healthy patients avoids controversy about the use of BDO vs a vascularized free-osseous flap for maxillomandibular reconstruction. In patients who are not candidates for more aggressive surgery or prolonged surgical time because of poor general health, BDO is an appropriate primary surgical treatment. Poor peripheral vascularization of the lower part of the leg was bilaterally present in 2 of our patients with VFFF failure. In such patients, mandibular BDO is our preferred treatment option primarily and secondarily.

Seven patients required additional bone grafting to fill the gap between the distracted bone and the distal stump. This is controversial, as the main criticism would be that the minimal benefit of BDO in ulterior bone grafting was necessary. In fact, the distraction device should apply compression forces at the docking sites to achieve fusion with the residual distal skeleton stump, although this is not always possible. We advocate BDO because the needed additional small bone graft is often much smaller than the one required if BDO is not used. Moreover, soft tissue is accordingly expanded. In our series, the mean defect length was 10 mm longer than the mean distracted bone lengthening. This was remedied by the use of additional bone grafting in selected patients. In most patients, bone grafting material was harvested from the anteroiliac or posteroiliac crest, which is a safe and easy donor site. In more than two-thirds of patients, bone grafting was unnecessary, and direct union between bone poles was achieved following corticotomy and plate fixation.

Few studies have evaluated the role of adjuvant radiation therapy before BDO relative to distraction outcomes. A previous study³ among 7 patients found 6 patients (85.7%) having successful DO following postoperative radiation therapy, in accord with the findings of other authors among smaller samples.^6,7 In the present series, 12 of 28 patients received adjuvant postoperative radiation therapy before BDO. Ten of 12 patients (83.3%) had successful BDO. Because more patients in our study received radiation therapy, this value is more accurate than previously published BDO success rates. However, outcomes among patients who received radiation therapy were poorer than those among patients who did not receive radiation therapy. For patients undergoing radiation therapy, the transport disk was designed to be as far as possible from the maximum dosage point in the radiation field, and the consolidation period was doubled. By using BDO in patients receiving radiation therapy, necrotic bone was eliminated, and soft tissue was expanded. We also eliminated associated fistulae and generated sufficient healthy bone to assure a bridging plate to improve facial symmetry and patient quality of life. All patients who underwent radiation therapy received high dosages of irradiation up to 70 J in accord with standard adjuvant treatment for head and neck cancer. Even with high irradiation dosages, at least 3 of 4 patients receiving radiation therapy will undergo successful BDO following maxillomandibular segmental resection.

Thirty-six internal devices were used in the entire series, 8 of them consecutively. In 14 of 28 patients (50.0%), the activator of the device was located transcutaneously. The external activator of these semiburied devices allowed easier access for activation. Previously, decreased infection and dehiscence with this type of access was observed because an intraoral approach is avoided.³ However, 10 of 28 patients (35.7%) with defects of the lateromaxilla or mandibular body required intraoral devices. Because the mean defect length was 47.0 mm for the entire series, 8 of 28 patients (28.6%) required 2 BDO devices placed simultaneously or in consecutive steps. The distractor used in this study achieves maximal distracted bone lengthening of 40 mm. Although other distraction devices allow longer distractions such as 50 mm,⁸ the shorter BDO devices used in the present series allowed better control of the distraction vector. In fact, the direction traversed by the transport disk after initial osteotomy may cause inferior results, as it tends to move parallel to the proximal segment of the residual bone, and greenstick fracture osteotomy following distractor removal may be necessary to reproduce the curvature of the mandible or maxilla in some patients. Slight changes in the distraction vector following the first distraction step may be useful to replicate adequate maxillomandibular curvature and remedy this complication. Consequently, we prefer shorter devices over more complicated systems such as the use of a traction wire plus a bridging reconstruction plate and bracket with miniplates for fixing the transport disk.⁹

Overall, BDO success with adequate bone formation was observed in 22 of 28 patients in our series. Good or excellent bone quality was macroscopically observed in 20 of 28 patients (71.4%). These results seem acceptable for a procedure that is a secondary reconstruction option in many patients in whom VFFF reconstruction has failed. To achieve better functional outcome, 11 of 28 patients (39.3%) underwent endosseous dental implant placement. A total of 46 dental implants were placed, with an overall dental implant success rate of 89.1% (41 of 46 dental implants). All failed dental implants were observed in 1 patient with previous VFFF loss, soft-tissue sequelae, and intraoral exposure of distracted bone. Bifocal DO allows reconstruction of the defect with an alveolar height that closely mimics the native mandible, with the advantage of a near-original attached gingiva. In contrast, the overlying soft tissue of microvascular free flaps is often thick, and dental implant placement is less favorable because of the absence of attached gingiva.

Complications were common. The most frequent complaint was discomfort by 8 patients. This was not secondary to pain during distraction, which was present in only 2 patients at the end of the distraction period. Discomfort was reported by 5 patients with an intraoral activator placed in the maxilla and by 3 patients with the activator placed in the mandibular body. This represents interference with the normal surrounding soft tissue such as the cheeks or lips. Discomfort was absent when the activator was placed externally, but this was socially more disturbing. The primary cause of poor final results was complete intraoral exposure and infection, which were observed in 3 patients with complete failure of the distraction procedure. The overall success rate of BDO was 79% (22 of 28 patients), and surgeons must expect a potential 20% failure rate despite rigorous patient selection and adequate BDO protocol.

In summary, BDO is of potential benefit for patients with bony defects following tumor ablation at various locations in the maxillofacial skeleton. It is useful for treating patients who are unsuitable for more aggressive surgery or prolonged surgical time because of poor general health or for patients in whom primary treatment using a vascularized free-osseous flap has failed. Bifocal DO is a reliable tool to gain sufficient bone to allow dental implant placement, an important functional outcome.

Correspondence: Raúl González-García, MD, Calle Los Yébenes 35, 8C, 28047, Madrid, Spain (raulmaxilo@gmail.com).

Submitted for Publication: March 23, 2009; final revision received September 14, 2009; accepted September 29, 2009.

Author Contributions: Dr González-García had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: González-García. Acquisition of data: Naval-Gías. Analysis and interpretation of data: González-García. Drafting of the manuscript: González-García. Critical revision of the manuscript for important intellectual content: Naval-Gías. Statistical analysis: González-García. Study supervision: Naval-Gías.

Financial Disclosure: None reported.