Computed tomography of axial (A) and coronal (B) planes, showing reconstruction of a malar bone using calvarial split grafts, which are fixed with titanium plates and screws. Grafts are not covered by soft tissue on the inner side to the maxillary sinus.
Area of graft placement and number of transplants used in each region. Number of grafts are split into tumor, trauma, and other cases. Numbers in brackets are the number of grafts that were irradiated.
Three-dimensional computed tomographic scan of a zygoma reconstruction using 3 bone grafts: 1 to repair the anterior wall of the maxillary sinus and 1 in each location for latero-orbital rim and zygomatic arch restoration.
Complication rate of grafts totally covered with the surrounding soft tissue (n=17) compared with those partially covered (n=79).
Comparison between radiated and nonirradiated tumor patients according to graft survival.
Smolka W, Eggensperger N, Kollar A, Iizuka T. Midfacial Reconstruction Using Calvarial Split Bone Grafts. Arch Otolaryngol Head Neck Surg. 2005;131(2):131-136. doi:10.1001/archotol.131.2.131
To evaluate the success rate of free calvarial grafts for midfacial reconstruction, the relevance of soft tissue coverage, and the influence of radiotherapy.
University medical center.
Fifty-six patients (27 tumor cases, 24 trauma cases, and 5 others) underwent bony midface reconstruction using calvarial grafts in the past 11 years. Half of the patients with tumor were additionally treated with radiation.
A total of 95 bone transplants were used for reconstruction of the zygoma, orbit, and nasal bone. Graft survival and complications were evaluated. Grafts with total and partial soft tissue coverage were compared. The influence of radiotherapy in the tumor patient group was determined.
Graft survival was 95.8%. One nasal dorsum graft was totally resorbed. Infection occurred in 9 cases, leading to only 1 total and 2 partial graft losses. The incidence of dysfunction of the eye due to globe malposition after reconstruction of the orbital walls was low. A correlation between radiation and transplant loss as well as between soft tissue coverage and graft survival could not be found.
For midfacial reconstruction, it is not necessary to fully cover calvarial bone grafts by the surrounding soft tissue. Even in patients who will undergo postoperative irradiation, calvarial bone grafts are a reliable alternative in selected cases.
Bony reconstruction of midfacial defects caused by ablative tumor surgery or trauma is of utmost importance for functional and esthetic restoration. The facial contour is determined by the underlying skeleton, and its restoration is a goal for a pleasing esthetic result postoperatively.1 Reconstruction of the orbital walls avoids diplopia, enophthalmos, and—in case of orbital roof augmentation—pulsatile exophthalmos.2,3 Autogenous bone grafts of the iliac crest, rib, and calvarium for skeletal reconstruction of the midface have been stated to be superior to alloplastic material.4
Nonvascularized bone grafting is known to have some disadvantages, however.5,6 Transplant loss after graft infection can occur if the graft is not fully covered by surrounding soft tissue.6 Therefore, in cases of midfacial reconstruction it has been recommended to cover the bone grafts by a soft tissue flap on the inner side at the maxillary sinus and the nasal cavity.4,7 In addition, in case of ablative surgery of malignancies, radiotherapy often results in further local tissue injury and diminished vascularity. Nonvascularized bone grafts have been stated to be less successful because of the tendency toward resorption due to delayed revascularization of the grafts caused by radiotherapy.8 For this reason, the transfer of vascularized bone is preferred and various methods are available.
Since the early 1980s, calvarial bone grafts have been used for midfacial reconstruction of defects caused by trauma and for correction of craniofacial deformities in patients with Treacher Collins, Crouzon, and Apert syndromes.9 Because of low donor site morbidity and low complication rates, calvarial bone grafts have grown in popularity in the reconstruction of severe midfacial trauma.4,5,10,11 Based on the pleasing results in trauma cases, calvarial grafts were also increasingly used for bony reconstruction of the midface after ablative oncologic surgery.3,7,12- 14 Calvarial bone grafts have also been used for orbital wall and zygomatic bone reconstructions, in which it is not always possible to fully cover the bone grafts by soft tissue. The question of whether it is really necessary to cover the bone graft with soft tissues has not been definitively answered. In tumor cases, the question also remains as to whether calvarial bone grafts can be used in connection with radiation therapy. In previous studies, no attention was paid to soft tissue covering around the bone graft, nor has there been any detailed study of the use of free calvarial grafts in irradiated cases.
In our department, various reconstructions of the midface were performed, both in trauma and tumor cases, including patients who subsequently underwent radiotherapy. Cases of midfacial grafting using free calvarial bone transplants were retrospectively studied to evaluate the complication rate in this type of nonvascularized bone transfer. Data were analyzed to compare grafts that had been covered with soft tissue on the inner side at the paranasal sinuses and the nasal cavity with those that were not totally covered. Special attention was paid to the success or failure of grafts in cases of radiotherapy, and patients with tumor who underwent radiation were compared with those who did not.
The present retrospective study included 56 patients (36 male and 20 female) who underwent bony reconstructions of the midface between 1993 and 2003. The mean age of the patients at surgery was 46.6 years, ranging from 7 to 87 years. The follow-up time, defined as the interval between surgery and the latest follow-up examination, ranged from 6 months to 11 years (mean, 5 years). Six patients died during the follow-up period because of their tumor disease, and 1 died because of a myocardial infarction.
A total of 27 operations were performed in patients who had defects after ablative oncologic surgery. The types and numbers of tumors as well as the number treated with radiotherapy are listed in Table 1. Twenty-four patients underwent bony reconstructions following midfacial trauma. In 3 further cases the reconstructed defects were caused by a mucocele in the nasoethmoidal and maxillary region. One patient needed orbital floor reconstruction to correct facial asymmetry due to a craniofacial cleft. In addition, 1 bone transplantation was performed in a case of Romberg syndrome. Reconstruction was primary in 20 of the tumor patients and secondary in 7. For the trauma patients, 5 reconstructions were primary and 19 were secondary. Eleven tumor patients underwent postoperative radiation. The mean dose of the radiotherapy was 54.1 Gy (range, 50-70 Gy).
For reconstruction of the midfacial skeleton, calvarial bone split grafts were harvested from the parietal area. A small parietal incision was used to approach the harvesting area. In primary extensive reconstruction cases, a bicoronal flap was performed for both tumor and trauma surgery. Grafts were shaped to a suitable size and placed in the correct position of the midfacial defect. Then, they were fixed to the facial bone with titanium plates and screws and covered on the outer side by the surrounding soft tissue. The bone grafts were not covered by soft tissue on the inner side if they were placed in the region of the paranasal sinuses and the nasal cavity (Figure 1).
A total of 95 calvarial grafts were used for bony midfacial reconstruction (Figure 2). Bone transplants were placed for malar bone reconstruction in 28 cases (20 in tumor cases, with 6 that were irradiated), for orbital floor reconstruction in 23 cases (7 in tumor cases, with 4 that were irradiated), for medial orbital wall reconstruction in 15 cases (11 in tumor cases, with 6 that were irradiated), for lateral orbital wall reconstruction in 3 cases (2 in tumor cases, with 1 that was irradiated), for orbital roof reconstruction in 1 case, and for nasal bone reconstruction in 25 cases (11 in tumor cases, with 7 that were irradiated). In 15 patients, 17 grafts were fully covered by the surrounding soft tissue. All operations using totally covered grafts were performed in trauma patients and the patient with Romberg syndrome, except for 2 cases of lateral wall reconstruction after tumor resection. In 79 bone transplantations in 44 patients, the sinuses or nasal cavity was left uncovered. Twenty-five patients were augmented using more than 1 bone transplant (Figure 3), and 13 were augmented in more than 1 location.
Patient records and all available documents and radiographs were reviewed. Data were collected on graft survival and complication rates such as wound infection, dehiscence, and partial or total loss of bone transplant. In cases of orbital reconstruction, postoperative complications such as double vision, enophthalmos, exophthalmos, entropion, and ectropion were also evaluated. A comparison of the complication rates of tumor patients with and without radiation was performed to analyze the influence of radiotherapy on graft survival.
Most of the calvarial split bone grafts healed without any complications. Computed tomographic scans obtained a minimum of 12 months postoperatively were available for 28 patients (50 grafts). Of these, reconstruction of the malar bone was performed in 17 cases, of the orbital floor in 10 cases, of the medial orbital wall in 8 cases, of the lateral orbital wall in 2 cases, and of nasal bone in 13 cases. In the computed tomographic scans evaluated, resorption of bone transplants was insignificant for all except 1 case—a nasal dorsum reconstruction that had total resorption. In 3 further patients, infection at the grafts caused partial transplant loss in 2 cases of malar bone reconstruction and total transplant loss in 1 case of nasal dorsum reconstruction. Partial bone graft loss was minor. The partial loss had no influence on stability of the reconstruction or on facial contour. Overall graft survival was 95.8% in 95 bone transplants. All complications leading to partial or total graft loss occurred in tumor patients. No transplant loss occurred in trauma cases.
All postoperative complications are given in Table 2. In 9 cases (6 tumor patients and 3 trauma patients) infection occurred in the reconstructed area between 1 and 36 months postoperatively (mean, 14.6 months). Medial orbital wall and nasal dorsum reconstruction was affected in 3 cases each, malar bone reconstruction in 2 cases, and orbital floor reconstruction in 1 case. In 2 cases of nasal dorsum reconstruction, the grafts had been totally covered by the surrounding soft tissue. Of the 6 tumor patients, 2 received radiation therapy (1 medial orbital wall reconstruction and 1 nasal dorsum reconstruction). In 3 cases, infection led to partial (2 patients) or total (1 patient) graft loss. In the other 6 cases, infection healed after treatment with antibiotics and local therapy, without any partial or total graft loss.
Wound dehiscence immediately after the operation was found in 5 cases—4 tumor patients (1 that was irradiated) and 1 trauma patient after gunshot injury. It occurred in 3 cases of zygoma reconstruction, in 1 case of medial orbital wall reconstruction, as well as in 1 case of nasal dorsum repair. In 2 of the malar bone restorations, wound dehiscence led to the partial transplant loss already reported. Exposure of the graft was present in 2 further cases. Both patients had a reconstruction of the nasal dorsum. After resection of the exposed part of the bone transplant, the graft was again covered by the surrounding soft tissue. Donor site complication occurred in only 1 case where the skin along the operation wound in the parietal area showed a local infection. This complication was successfully resolved using antibiotic therapy and local treatment.
Dysfunction of the eye and globe malposition were evaluated in patients who underwent bony reconstruction of the orbital walls (40 patients). In 5 cases, temporary diplopia appeared. In 1 patient with severe orbital trauma, preoperative diplopia in all directions was found. After reconstruction of the orbital walls, it was reduced but still present on upward rotation of the globe. Slight enophthalmos was reported in 2 cases, which did not result in any diplopia or dysfunction.
Seventeen bone transplants had no connection to the nasal cavity or the paranasal sinuses postoperatively. Three of these grafts were used for lateral orbital wall reconstruction—2 in tumor cases and 1 in a trauma case. One zygoma transplantation was performed as an onlay graft in the patient with Romberg syndrome. Most of the bone grafts without any connection to the nasal cavity were nasal dorsum reconstructions after trauma (13 grafts). One wound dehiscence and 2 infections were found in these cases. However, no partial or total graft loss occurred (Figure 4).
Of the 27 patients who had undergone ablative oncologic surgery, 11 underwent postoperative radiation. Of these patients, 8 were reconstructed primarily and 3 secondarily. In the radiated patient group (11 cases), only 1 total transplant loss (9%) was observed in a patient with a juvenile nasopharyngeal angiofibroma involving the anterior skull base. The resection of the tumor was performed using a subcranial approach.15 For better visualization of the anterior skull base, the frontonasal segment was osteotomized and temporarily removed. To avoid saddle nose after replacement of the frontonasal segment, a nasal dorsum graft of calvarial bone was performed. The patient postoperatively received radiation therapy, with a total dose of 50 Gy in 2-Gy fractions, 25 times within 5 weeks, using a 6-MeV linear accelerator. After a postoperative follow-up of 3 years, a total resorption of the bone transplant was observed. In the tumor patient group without radiation therapy (16 cases), on the other hand, 1 total and 2 partial graft losses (19%) occurred as a result of the wound infection (Figure 5). Other surgery-related complications such as wound dehiscence, infection, or graft exposure were found equally in the radiated and nonirradiated groups. There was no difference in patients’ distribution to each complication between the groups.
Although calvarial bone grafts are used today for various midfacial reconstructions, the complication rates related to the bone graft are surprisingly low. Previous studies reported no resorption and no loss of calvarial transplants after midfacial repair on short-term follow-up ranging from 1 year to 3.7 years.5,12- 14 Minor complications such as graft exposure (4.8%), repositioning (3.2%), total transplant loss (3.2%), and infection (1.6%) were reported on a 2-year follow-up by Powell and Riley.11 Their graft survival rate was 96.8%. In a further study by the same authors reporting 4-year experience in an increased number of patients, the success rate increased to 99.3%.16 To our knowledge, there has been no longer follow-up than 4 years.
All complications observed in our long-term follow-up ranging up to 11 years appeared within the first 3 years after reconstructive treatment. The survival rate of 95.8% found in the present study is slightly lower than those previously reported. However, as all complications occurred within 3 years postoperatively, a longer follow-up seems to have no influence on the success rate of calvarial grafts.
Powell and Riley16 compared the different reconstructed areas of the midface. The nasal dorsum was found to be one of the regions of increased resorption and contour changes. In their study, all 7 cases of nasal dorsum reconstruction had partial bone resorption, ranging up to 30%. In our study, however, resorption of a nasal dorsum graft was present in only 1 (4%) of 25 reconstructions of the nasal dorsum. In the other 24 cases, no major resorption was noted. Cheney and Gliklich17 reported no resorption in 35 patients who underwent split calvarial grafting to the nasal dorsum, and concluded that split calvarial bone is a material of choice for nasal dorsum reconstruction. However, all their operations were performed to treat saddle noses, and grafts showed no connection to the nasal cavity. Our study included resorption of 1 nasal dorsum graft, in a reconstruction after ablative tumor surgery; in this case, the bone transplant was not covered with soft tissue on the inner side. Nevertheless, in all other cases using this grafting technique (10 patients), no resorption was observed, so an influence of soft tissue coverage on major graft resorption cannot be concluded.
Infection of nonvascularized iliac grafts leads to sequestration and total transplant loss.18 Studies on calvarial grafts, on the other hand, show that infection of the surrounding soft tissue rarely affects the bone transplant.11,16 Total healing under antibiotics and local treatment as well as only partial removal of the grafts is reported in a few cases. It has generally been stated that calvarial grafts have a relative resistance to infection in an infected soft tissue site. In our study, 9 infections occurred, resulting in only 1 total and 2 partial graft losses. Partial bone graft loss was minor, having no influence on stability of the reconstruction or on facial contour. Thus, the results of the present study are in accordance with those of the previous reports.
For midfacial reconstruction, it is generally recommended to cover free bone grafts with soft tissue to the sinuses to avoid infection and graft loss caused by sinus secretion and bacterial flora. With calvarial bone grafts as well, pedicled buccal fat pad graft and temporal muscle flap have been described for this purpose.7,19 In the present study, no major differences were observed in the complication rates of calvarial bone grafts totally or partially covered with soft tissue. Although partial and total transplant loss was only found in cases of partially covered grafts, it was low in this patient group (3%). Based on our results on using calvarial split grafts without soft tissue coverage to the sinuses and nasal cavity, bone transplants do not require full soft tissue coverage, even in irradiated patients.
One reason full soft tissue coverage is not needed seems to be the rapid vascularization of calvarial bone grafts. Animal studies document that membranous onlay bone grafts are more rapidly vascularized than enchondral grafts.20 Some authors suggest that the vascular ingrowth is affected by the gross morphology of the bone substitutes.21,22 Also, Hardesty and Marsh23 concluded that the different revascularization of enchondral and membranous grafts is related to the different 3-dimensional osseous architecture. The present study demonstrates that coverage of calvarial grafts on the outer side by the surrounding soft tissue is sufficient for revascularization and healing of the whole bone transplant.
The low resorption of calvarial grafts found in the present study is also reported from animal studies. Donovan et al24 compared histological findings of calvarial and iliac bone transplants in swine. Over a 12-month period, the mean average volume of iliac bone retention was 52.5% compared with 91.2% for the calvarial bone. Histological evaluation showed that calvarial grafts had less osteoclastic and increased osteoblastic activity than iliac bone transplants. The inflammatory response in iliac grafts is very high during the first 4 months postoperatively and correlates to the loss of graft volume. No inflammation is noticed in the calvarial grafts. A further reason for minor resorption of calvarial transplants is that compared with the corticocancellous iliac graft, the cortical calvarial graft has a greater than 2-fold higher density.
Treatment of malignancies of the midface leads to bony and soft tissue defects caused by ablative surgery, which need to be reconstructed and require radiotherapy. It is recommended that in these cases, transfer of vascularized bone is necessary because the decreased vascularity of the soft tissue in the grafting area increases the risk of transplant loss of free bone grafts.8,25 Grotting et al8 reported on 11 vascularized temporal osteomuscular flaps for midfacial and mandibular reconstruction after ablative tumor surgery. After a follow-up ranging from 2 to 34 months, neither infection, flap loss, nor any other complications were present. The authors concluded that facial defects associated with radiotherapy are poor recipient beds for nonvascularized bone grafts.8 Choung et al25 studied 13 patients who underwent midfacial and mandibular reconstruction, also using vascularized temporal osteomuscular flaps. In 2 irradiated patients, wound dehiscence occurred, which healed under local treatment. The authors concluded that vascularized bone is an essential component of the reconstruction in patients who have received radiotherapy.25 In the present study, there was no relationship between radiation and loss of transplanted free calvarial bone grafts. The incidence of transplant loss was even higher in the nonirradiated patients than in the patients who had subsequently received radiation therapy. To restore extensive bony defects of the midface in patients who subsequently received radiation, vascularized bone transplants are a safe reconstruction option. However, based on our results, the use of free calvarial transplants seems to be sufficient in many cases, even if the grafts are not covered at the inner side to the sinuses.
In conclusion, calvarial split grafts are a suitable material for bony midfacial reconstruction. Their thickness and shape are ideal for the restoration of facial bone. In addition, based on the low complication rates determined in this study, grafting without fully covering the bone transplant by the surrounding soft tissue is possible. This study also confirms that in irradiated patients, calvarial bone grafts for midfacial restoration are a reliable alternative, and in selected cases, vascularized bone transfer is not absolutely necessary.
Correspondence: Wenko Smolka, MD, DMD, Department of Cranio-Maxillofacial Surgery, University of Berne, Inselspital, CH-3010 Berne, Switzerland (firstname.lastname@example.org).
Submitted for Publication: August 2, 2004; accepted October 21, 2004.