Ameloblastoma resection with immediate rib reconstruction: addressing the problem of mandibular angle and central bone bulk

Introduction

The treatment of ameloblastoma has been problematic because there have been high recurrence rates with conservative debridement but also fears of high complication risks with aggressive surgical treatment and the perceived need for complicated reconstructive techniques. Therefore, several approaches have been attempted including resection alone and resection with immediate reconstruction. Previous studies have documented the feasibility of different grafts and the difference of bone durability as evidenced by radiologic criteria.

Clinical experience with treatment by tracheostomy and immediate reconstruction with free rib graft forms the planned programme for this study. The first patient was followed for three years. Resorption of a significant amount of bone was evident on radiography after 18 months but there was a very good functional result in several categories: cosmesis; speech; deglutition; satisfactory dietary intake; and airway control. There were clear pathologic margins and, after three years, no evidence of recurrence. Because clinical functional parameters continued to score well despite radiological changes, this method was then investigated in a series of patients.

Methods

Our study was initiated to demonstrate the safety in limited resource settings of such a regularly planned treatment plan. Tracheostomy is performed at the outset of the intervention and is followed by resection and immediate reconstruction. A nasogastric tube is left in situ afterwards to facilitate early feeding until postoperative oedema resolves and the patient regains his ability to swallow without much difficulty. Patients do not normally require postoperative intensive care; they are safely observed in the high dependency section of a general surgical ward. Usually the tracheostomy tube and nasogastric tube are removed by the tenth postoperative day. Both short-term complications and long-term functional results are carefully followed.

Our series of patients included those who presented to our dental clinic with enlarged mandibular swelling and where dental infection or trauma had been excluded. Half of these patients had previous debridement and had pathology results confirming ameloblastoma. Others had cystic lesions of the mandible consistent with ameloblastoma and subsequent pathologic confirmation.

This work was carried out in Queen Elizabeth Central Hospital (Bantyre, Malawi), Kamuzu Central Hospital (Lilongwe, Malawi), and Mzuzu Central Hospital (Mzuzu, Malawi). Records were kept as surgeon’s case files.

Results

A total of 25 patients were followed according to this operative programme. The demographics showed a nearly equal frequency of men and women with a median age of 29 years. The tumours were large, in the range of 5–20 cm, with a median length of 12 cm, many of them involving the symphysis and the angle of the mandible (Table 1).

OPEN IN VIEWER

Table 1.

Demographics.

Sex	9 women, 16 men
Age (years)	18–55 (median = 29)
Tumour size (cm)	6–20 (median = 12) Half involve angle of mandible Two-thirds involve midline
Follow-up (months)	1–15 (median = 9)

We had no deaths and no crises related to airway control. Local hospital staff resources decided whether most patients were transferred simply to high dependency units of surgical wards for postoperative monitoring (QECH and KCH) or to intensive care (MCH).

There was one major complication with significant intraoperative haemorrhage controlled by packing and planned re-exploration 72 h later. This patient had a free rib graft inserted at the second exploration with good results.

Blood was transfused in 50% of patients because of significant blood loss during the operation.

Minor complications included five patients who had a small extrusion of bone graft in the first six months after operation; these extruded pieces were < 1 cm² in size. Five other patients required removal of fixating wires. One patient had significant absorption of the bone graft and shortening of the reconstructed arch without interference of speech or diet. Five patients had purulent local infection over the bone graft which resolved readily with local wound care and oral antibiotics (Table 2).

OPEN IN VIEWER

Table 2.

Complications.

Minor	5 patients required removal of fixation wire after 2 months postoperatively; no loss of soft tissue contour
	1 patient has significant absorption and shortening of restored mandible; little interference speech and diet
	8 patients had minor extrusion of bone, less than 1 cm2
Major	1 patient had severe intraoperative haemorrhage controlled by pack which was removed at 48 h; free rib graft was placed without complication except for minor bone extrusion after 2 months

There were good functional results in terms of cosmesis, swallowing and speech in follow-up. There were no long-term airway problems related to the tracheostomy or to the reconstruction (Table 3).

OPEN IN VIEWER

Table 3.

Long-term function.

Cosmesis	Mouth opens over 3 cm in all patients 20/25 symmetrical mouth opening
Speech	7/25 clear speech 16/25 understandable 2/25 difficult
Swallowing	All patients manage soft diet well; stable weight 15/25 able to chew with opposite side 10/25 report drooling with slight difficulty of control
Airway	25 patients sleep without difficulty 25 patients report no difficulty from tracheostomy wound

Results were kept in the surgical log and outpatient records.

Discussion

The first clinical consideration must be clarifying the diagnosis. Extensive description of radiologic features is found in reviews by Scholl. ¹ An extensive description of pathologic features of dental tumours is found in a review by Mullapudi. ² The rationale offered for recurrence after debridement-curettage is that the capsule is incomplete and residual tumour finds room for recurrence within the bone marrow. Some recurrences can be massive extending beyond 10 cm. ³ Hence, many authors advocate radical resection with margins of 0.5–1.0 cm. ⁴

When a decision is made that curettage is unsatisfactory and resection must be performed, additional considerations come to the fore. The chief concern is to protect the airway and then to restore speech, swallowing and cosmesis. The reconstructive problems increase progressively with larger resections involving the midline, the angle of mandible and perhaps the entire mandible. Wilson described the different pulls of muscle groups distorting the residual fragments because of attachments of the pterygoid and the hyoid muscles. ⁵ Any defect > 5 cm warrants reconstruction. Free grafts can be constructed from rib or iliac crest. Composite grafts can be constructed from rib or fibula with musculovascular attachments. When placed within the soft tissue envelope, the residual muscles are stabilised but strength for mastication may remain weak. The details of different techniques are described elsewhere.

An important aspect of reconstruction is the management of the inferior alveolar nerve. Although others have been able to preserve this nerve and even graft it for reconstruction, so far we have not been able to duplicate their good results. ⁶ Because of the large size of tumours in our series, the nerve has regularly been sacrificed on the involved side.

Our distribution of demographic considerations (gender, age) and location of tumour (premolar, molar, angle of mandible) is consistent with other reports. We have followed a system of standardising treatment similar to that reported in Ghana. ⁷

This series demonstrates that free rib grafts can be used well beyond the generally accepted limits of a span of 6–9 cm. ⁸ Several modifications were found necessary and these are detailed in the following description (length, bulk, angle and fixation).

The first modification is the length of the span and the remodelling required to shape the substituted bone graft. The eighth rib is usually chosen for harvesting (usually from the chest of the non-dominant arm) in a subperiosteal manner; most often about 25 cm of rib is obtained. On occasion, a second rib can be taken without difficulty, even from the same side. The rib is scored on the convex side to increase the angulation to approximate the shape of the mandible (Figure 1). OPEN IN VIEWER

A second modification is reconstructing the mandibular height and thickness when segmental resection has been carried out. El-Sheikh described a method of bundling a free rib graft to make a broader replacement. ⁹ This technique splits the entire rib and overlaps the splits. Because of the large span required in our patients, we have devised a simple technique of partially thickening the graft by splitting a second length and tying the additional graft over the primary rib graft. This provides additional bulk especially in patients where the primary rib graft appears too thin. As El-Sheikh mentioned, the split rib incorporates readily into the mandibular bed. We have not noted any extra bone extrusion on follow-up (Figure 2). OPEN IN VIEWER

A third modification is fashioning the rib graft to reconstruct the angle of the mandible. This is done by cutting a short segment to serve as the vertical pillar and splicing the rib so as to overlap the split edges (secured with wire). This preserves the ability to open the mouth with adequate distance between the lips (Figure 3). OPEN IN VIEWER

The fourth modification relates to the fixation of the graft. Routinely we fashion a well in the remaining healthy mandible into which the graft can be wedged and then held securely with wire. Where the condyle has had to be resected, the residual capsule of the temporomandibular joint is sutured with long-lasting absorbable suture to the reconstructed rib graft. The graft provides soft tissue support and there is enough function to enable opening the mouth (Figure 1).

Rana has described two problems with the free rib graft – absorption and graft failure. ¹⁰ Of several possible sources of bone, a free rib graft was associated with 64% resorption as determined from X-ray films at one year. A small number of grafts failed because of dehiscence of the wound, the need for debridement or the recurrence of tumour. No major graft failures occurred. Often a small piece of bone (< 1 cm²) may extrude. In a rib graft of > 20 cm, this is usually of no consequence. It should be emphasised that firm apposition of soft tissue around the graft is important for graft survival. Simple closure of the oral mucosa and platysma muscle can leave space for seroma accumulation and subsequent graft loss. We place several mattress sutures in the remaining myelohyoid and geniohyoid muscles and the soft tissue around the submandibular gland. In our follow-up, we did not routinely measure radiolucency of residual graft but have concentrated on the clinical assessment of function, which remains satisfactory (Table 3).

Conclusion

This series documents the evaluation of several important clinical endpoints beyond mere radiological assessment of the durability of the graft. Early consequences of an ameloblastoma are progressive disability with deglutition, denutrition, speech and control of secretions as tumours progress. In a limited resource setting, tracheostomy, mandibular resection and free rib graft substitution is a safe and reliable method. This treatment programme addresses the clinical problems well and scores well both in patient satisfaction and clinician assessment. The functional outcome is more important than subsequent radiological appearances of the graft.