A 10-year retrospective study of the clinical presentation, pathology, treatment and outcomes of ameloblastoma and ameloblastic carcinoma in a teaching hospital in Lagos, Nigeria

INTRODUCTION

Ameloblastoma is the most common benign, odontogenic tumour of the head and neck region, comprising about 1% of all cysts and tumours of the jaws and up to 10% of all odontogenic tumours.¹ It arises from epithelial remnants of the developing root sheath.^2,3 It can affect all age groups but is most common in the third to seventh decades, with peak incidence in the third and fourth decades.^4,5 It has a racial preponderance with higher prevalence in Africans and Asians than in Caucasians.⁶ The lesion can occur anywhere in the mandible and maxilla but is most commonly found in the posterior mandible.⁷ A slight male preponderance has been reported with a male-to-female ratio of 1.17:1.⁸

The tumour is usually asymptomatic, presenting as a painless slow-growing swelling, which may explain the reason for its neglect among patients with low socioeconomic status.⁹ However, neglected lesions can grow into large disfiguring sizes (Fig. 1).^9,10 Despite being benign, it is a locally aggressive tumour with a high recurrence rate if not completely resected. Radiological investigations such as orthopantomograms are useful first-line investigations that aid its diagnosis.¹¹ Advanced imaging such as computed tomography delineates the extent of the tumour and aids treatment planning; however, orthopantomograms can be used for treatment planning if advanced imaging is not available,¹¹ especially in cases affecting the mandible (Fig. 2). The main goals of surgical treatment of ameloblastoma are to completely remove the tumour and to restore orofacial function and aesthetics.¹² Complete removal of the tumour prevents tumour recurrence and/or transformation to ameloblastic carcinoma.¹² The definitive treatment is surgical resection with wide margins (1– 2 cm)¹³ with some authors recommending immediate bone reconstruction.¹⁴ Conservative management in the form of enucleation and curettage is no longer recommended, even for early lesions, due to the high recurrence rate.¹⁵ In view of the high prevalence of ameloblastoma in Nigeria, and the public health burden it confers on the Nigerian population, it is important to determine its current epidemiological and clinicopathological presentation as well as current treatment modalities and their postoperative outcomes.

Close

FIG 1.

Large expansile lesion of the anterior mandible extending posteriorly

Export to PPT

Close

FIG 2.

Orthopantomograms showing unicystic ameloblastoma of the anterior mandible (top left), multicystic ameloblastoma of the left posterior mandible (top right), recurrent ameloblastoma of the right posterior mandible (bottom left), and reconstruction of mandibular continuity defect using reconstruction plate, alloplastic condylar prosthesis and iliac crest bone graft secondary to resection of ameloblastoma of the right posterior mandible (bottom right)

Export to PPT

We aimed to study the current presentation, treatment and treatment outcomes of ameloblastoma in Lagos, Nigeria, and compare these with what have been previously reported in the population. The specific objectives were to describe the current epidemiology and clinicopathology of ameloblastoma and to describe the current pattern of management and incidence of recurrence of ameloblastoma surgically treated during a 10-year period in a teaching hospital in Lagos, Nigeria.

METHODS

RESULTS

Sixty-three patients were diagnosed with and managed for ameloblastoma within the study period. The number of patients treated in a year ranged from 4 (4/63, 6.3%) in 2013 to 10 (10/63, 15.9%) in 2016 with mean (SD) number of patients of 6.3 ( 1.6) per year (Fig. 3). All 63 patients were followed up for at least 1 year postoperatively with a follow-up ranging from 1 to 10 years and a mean (SD) of 2.7 (2.1) years.

Close

FIG 3.

Yearly number of patients treated surgically for ameloblastoma

Export to PPT

There were 32 (50.8%) males and 31 (49.2%) females with a male-to-female ratio of 1.03 to 1. Their age was between 8 and 67 years with a mean (SD) of 34.2 (14.8) years. Participants’ labour categories were unskilled (5/63, 7.9%), semi-skilled (35/63, 55.6%) and skilled (2/63, 3.2%). Twenty-one participants (21/63, 33.3%) were students, retirees or unemployed. Only one patient had a positive family history of tumour (mother had breast cancer), and 14 (14/63, 22.2%) patients had habits of cigarette smoking, alcohol consumption or chewing kola and tobacco.

The DOC ranged from 2 months to 18 years with mean (SD) of 4.6 (3.8) years. There were 9 (9/63, 14.3%) patients with DOC <12 months, 8 (8/63, 12.7%) with DOC ≥12 months but <2 years, 17 (17/63, 27.0%) with DOC ≥2 years but <5 years and 29 (29/63, 46.0%) with DOC ≥5 years. No significant relationship was noted between age, gender, occupation and DOC (p>0.05 for all associations, Table 1).

The size of the lesion was recorded for 23 patients, and the largest diameter ranged from 3 cm to 17 cm with a mean (SD) of 8.1 (4.4) cm. The tumour originated more often from the mandible than from the maxilla and, in a decreasing order of frequency, from the right posterior mandible in 27 (27/63, 42.9%) patients, from the left posterior mandible in 22 (22/63, 34.9%) patients, from the anterior mandible in 9 (9/63, 14.3%) patients, from the right posterior maxilla in 3 (3/63, 4.8%) patients and from the left posterior maxilla in 2 (2/63, 3.2%) patients. None of the included tumours originated from the anterior maxilla. About half (50.8%) the lesions at presentation involved 1 (19.1%) or 2 (31.7%) adjacent regions.

The most common radiological features on orthopantomo-gram or CT scan were multilocular radiolucencies (57/63, 90.5%) and truncation of roots (58/63, 92.1%). Other less common radiological features included unilocular radiolucencies (6/63, 9.5%), cystic necrosis of the jawbone (4/63, 6.3%) and marked destruction of submandibular gland (1/63, 1.6%).

Fifty-three (53/63, 84.1%) patients were diagnosed with conventional ameloblastoma, 4 (6.3%) with unicystic ameloblastoma, 2 (3.2%) with peripheral ameloblastoma and 4 (46.3%) with ameloblastic carcinoma. The most common histological type was follicular ameloblastoma (32/63, 50.8%; Table 2).

The surgical intervention in 56 (88.9%) patients was mandibulectomy, 5 (7.9%) had a maxillectomy and 2 (3.2%) had surgical excision with peripheral ostectomy (Table 3). Eight (12.7%) patients received additional surgery: 6 (9.5%) had additional surgical excision of soft-tissue lesion, 1 (1.6%) had additional neck dissection and 1 (1.6%) had additional skin graft placement secondary to excision of involved skin.

Complete data on treatment characteristics and complications were available for 52 of 63 patients. Over half the patients had nasotracheal intubation (35/52, 67.3%) and extraoral approach (39/52, 75%) as the anaesthetic technique and surgical approach, respectively. Other less common anaesthetic techniques included elective tracheostomy (11/52, 21.2%), endotracheal intubation (4/52, 7.7%), local anaesthesia plus conscious sedation (1/52, 1.9%) and emergency tracheostomy (1/52, 1.9%; Table 3).

Reconstruction of the jaw defect after surgical resection was either immediate or delayed. Of the 47 patients who had mandibulectomy and complete treatment data, only 4 (48.5%) patients had immediate reconstruction. For maxillectomy patients, only 1 of 3 (33.3%) had immediate reconstruction. For patients who had surgical excision with peripheral ostectomy secondary to peripheral ameloblastoma, reconstruction was not needed.

At the time of data collection, only 19 of 50 (38%) patients with complete data had received reconstruction for jaw defect. In contrast, 31 (62%) patients were either awaiting treatment, refused treatment or were not ready for treatment (financially or psychologically). For mandibulectomy defects, reconstruction was done using non-vascularised iliac crest bone graft for 12 of 16 (75%) patients, vascularised fibular flaps for 3 of 16 (18.8%) patients and clasp retained denture for 1 (6.2%) patient who received marginal resection under local anaesthesia plus conscious sedation; 3 of the iliac crest graft reconstructions were immediate reconstructions and only 1 of the vascularised fibular flap reconstruction was an immediate reconstruction. For maxillectomy defects, immediate reconstruction was done with feeding plate for 1 (33.3%) patient and delayed reconstruction was done with a definitive obturator for 2 (66.7%) patients.

The minimum LOH stay was 1 day (day case) for the patient who received marginal resection under local anaesthesia along with conscious sedation and the maximum LOH stay was 16 days with a mean (SD) of 9.4 (2.4) days. There was a significant difference between surgical approach and LOH stay; transoral approach was associated with shortened LOH stay (Table 4). No preoperative complications were recorded. Immediate postoperative complications were recorded in only 3 of 52 (5.8%) patients: 1 (1.9%) had surgical wound breakdown at postoperative day 14 and 2 (3.8%) developed trismus at postoperative days 7 and 9, respectively. The majority (9/10, 90%) of patients who had extraoral surgical approach developed postoperative complications compared to transoral approach; however, this relationship was not statistically significant (Table 4). No significant relationships were noted between surgical intervention, type of reconstruction, immediate postoperative complications and LOH stay (p>0.05 for all associations, Table 5). Only 7 patients had delayed postoperative complications (Table 6).

Recurrence was recorded in only 3 (4.8%) patients, 2 with conventional ameloblastoma and 1 with ameloblastic carcinoma. Although recurrence appeared to occur in relationship with larger tumour size (>8 cm), certain tumour diagnosis and histological type (conventional ameloblastoma [follicular and granula] and ameloblastic carcinoma) and surgical approach (extraoral), none of these were statistically significant (Table 7).

Time-to-event analysis was done for incidence of recurrence and potential prognostic factors. Kaplan–Meier plot was used to describe probability of recurrence within 10 years postoperatively for each of the prognostic factors (Figs 4–7). The probability of recurrence was >10%; >20% for tumours >7– 11 cm, 20% or more, >20% for ameloblastic carcinoma, about 10% for follicular ameloblastoma, about 50% for granular ameloblastoma and almost 20% for extraoral surgical approach. However, no significant associations were noted between recurrence and potential prognostic factors (p>0.05 for all associations, Table 7).

Close

FIG 4.

Probability of recurrence within 10-years post-treatment by largest diameter of tumour

Export to PPT

Close

FIG 5.

Probability of recurrence within 10-year post-treatment by tumour diagnosis

Export to PPT

Close

FIG 6.

Probability of recurrence within 10-year post-treatment by histological type of ameloblastoma

Export to PPT

Close

FIG 7.

Probability of recurrence within 10-years post-treatment by surgical approach

Export to PPT

DISCUSSION

Ameloblastoma is the most common odontogenic tumour of the head and neck region with a high preponderance in the African population.^1,6 Nigeria alone bears over 90% of the total ameloblastoma burden in sub-Saharan Africa; despite an unclear reason, it may be partly due to the large amount of research literature from the region.^19,20 Previous studies^7,20–23 have reported the epidemiological distribution and clinicopathological presentation of ameloblastoma in Nigeria; however, regular review of this information is important for early identification of secular changes due to its high public health impact. In addition, we explored the surgical management of ameloblastoma and outcomes of management after a minimum follow-up of 1 year.

Similar to existing reports,^7,20–23 our study showed a peak incidence of ameloblastoma in the third and fourth decades of life and confirmed previous reports of the rarity of ameloblastoma in the first decade of life.^7,22,23 Conflicting reports exist on the sex distribution of ameloblastoma. Most studies^7,20,23 similar to results of our study, suggest equal sex distribution or a slight male preponderance, while Chukwuneke et al.²² reported a female predisposition to ameloblastoma.

Although ameloblastoma is not known to be caused by any occupation, it has been reported to be more common among the low socio-economic class.^24,25 This may be explained by the fact that ameloblastoma sometimes develops from untreated odontogenic cysts, and people of low socio-economic class are less likely to seek healthcare for financial reasons, resulting in neoplastic transformation to ameloblastoma. In our study only 2 patients were in the skilled labour category while majority were either semi-skilled or unskilled, which supports the current assumptions. The DOC in our study raises public health concerns in that a large proportion of patients lived with their condition despite its grotesque appearance for over 5 years. These are supported by previous reports from within^21,23 and outside Nigeria,²⁶ pointing to an unaddressed issue of healthcare inequality and inaccessibility.

Ameloblastoma is largely known as a painless, slow-growing, benign lesion. Small lesions are often seen among Europeans and Americans compared to large lesions among Africans and Asians.⁶ However, the clinical presentation of the lesion is dependent on the DOC, and large grotesque swellings are seen only with longer DOC.^21,23 Our study, similar to previous Nigerian studies^21,23 reported large lesions with mean tumour diameter of 8.1 cm, and maximum up to 17 cm, with about half involving multiple regions of the mandible or maxilla. The mandible is considered the most common site for ameloblastoma, with some studies^7,19 reporting it predominantly in the posterior mandible, while others^22,23 reported it in the anterior mandible. Similar to previous reports, our study showed that the majority (92.1%) of the lesions occurred in the mandible, and 84.4% of cases involved the left or right posterior mandible. No consensus has been reached on the reason for the predominance in the mandible, but an assumption is that ameloblastoma originates from reduced enamel epithelium and rest cells of Malassez which are predominantly found in the mandible due to its size compared to the maxilla.²¹

In agreement with previous studies from in^21–23 and outside^26,27 Nigeria, the most common radiological features reported were multilocular radiolucency and root truncation representing a multicystic locally invasive odontogenic lesion characteristic of ameloblastoma. Similar to previous reports from Nigeria and/or Africa,^21–23 the most common morphological presentation was conventional ameloblastoma,^21–23 and the most common histological type was follicular ameloblastoma.^7,22 However, Soyele et al.²¹ reported plexiform type, as the most common type.

A similar study done in the same hospital decades ago by Ladeinde et al.⁷ reported a prevalence of ameloblastic carcinoma of 4.35%. In our study, the prevalence of ameloblastic carcinoma was 6.3%. Although both reports indicate prevalence <10%, the slight increase in prevalence may suggest an increase in predisposing factors among the Nigerian population and the need for further research into the aetiopathogenesis of ameloblastoma in Nigeria.

Jaw resection and surgical excision with wide margins (1– 2 cm) are the recommended treatment for both intraosseous and extraosseous ameloblastoma, respectively.^13,14 Oral and maxillofacial surgeries performed under general anaesthesia require special consideration as routine endotracheal intubation may not be ideal. After non-total mandibular jaw resection, occlusion must be established between the remaining jawbone and maxillary arch for proper placement and fixation of reconstruction plate and for ideal rehabilitation and functional restoration of the maxillomandibular system after surgery. Endotracheal intubation will prevent this and, hence, the need for alternatives such as nasotracheal intubation, submental intubation and surgical airway or tracheostomy.²⁸ We used nasotracheal intubation as the most common anaesthetic technique and was only used for mandibulectomies. Endotracheal intubation was used for all maxillectomies as establishment of occlusion was not required. Similar to previous studies²⁹ this study showed that routine tracheostomy was unnecessary for maxillectomies. No patients in this study had submental intubation; however, it has been shown to be less invasive with minimal intraoperative and postoperative complications compared to tracheostomy and should be considered when endotracheal and nasotracheal approaches are contraindicated.³⁰

The more commonly practiced surgical approach for mandibulectomy is extraoral.³¹ However, proponents of the transoral approach have suggested its use for resection and immediate reconstruction of benign pathologies when sufficient access can be achieved, with the benefits of limiting postoperative morbidities associated with the extraoral approach.³¹ In this study, the transoral approach was used in 12 of 47 mandibulectomy patients and in 1 of 2 patients with surgical excision and peripheral ostectomy. Only 1 of 13 patients in the transoral approach group developed a postoperative complication, although this association was not statistically significant. In addition, statistically significant reduction in LOH was noted with the transoral approach.³¹ These issues do not exist for maxillectomy as the universally accepted surgical approach is extraoral using the Weber–Fergusson incision with or without a Trotter or Lynch extension, which is corroborated by the results of our study.

Reconstruction of mandibular jaw resection defects can either be performed immediately at the time of resection or delayed till a later date. Improved infection control and high degree of success following radiotherapy and/or chemotherapy are advantages of the delayed approach, while increased healthcare cost and negative psychological effect from multiple surgeries are disadvantages.^32,33 On the other hand, the immediate approach offers economic advantages and a single surgery experience but at the cost of the increased risk of infection.^32,33 In the immediate approach, non-tension, water-tight seal must be achieved to prevent ingress of saliva into the surgical site, subsequent wound dehiscence, infection and graft failure. In this study, four mandibulectomy patients received immediate reconstruction and one of them developed osteomyelitis 1-month postoperative requiring sequestrectomy of bone graft, curettage of surgical site and delayed secondary reconstruction. Despite the merits of an immediate reconstruction, we recommend it only when non-tension, water-tight seal with adequate infection control can be guaranteed. Reconstruction of maxillary jaw resection defects can be done either by surgical procedure or by prosthetic rehabilitation.³⁴ Similar to mandibular defects, maxillary defects may be reconstructed using free flaps or grafts from the iliac crest, fibula, radial forearm or scapula.³⁵ However, the complex anatomy of the maxilla poses a unique challenge in their retention. Like the results of this study, maxillary defects of ameloblastoma are often reconstructed through prosthetic rehabilitation with feeding plates and obturators.

The high rate of recurrence of ameloblastoma following conservative surgical management led to the recommendation of a radical approach for all cases including early stage lesions.¹⁵ In this study, radical resection and excision was done for all patients. Despite this approach, recurrence of ameloblastoma is not uncommon due to its local invasiveness, tissue components and histologic types.³⁶ Previous studies in Nwoga³⁷ and outside Nigeria³⁸ have reported a recurrence rate of 33.3% and 35%, respectively. We report a relatively low recurrence rate of 4.8%, possibly due to the short average follow-up period.

Previous studies by Reichart et al. and Ghai^39,40 have reported an increase in recurrence of ameloblastoma when the lesion is multicystic, located in the maxilla and follicular or granular histologic types. The multicystic nature of the lesion points to a more aggressive form of the tumour compared to its unicystic variant. In addition, the lesion spreads more aggressively and recurs more frequently in the maxilla due to the thin maxillary cortical bone that provides a weak barrier for the spread of tumour. Our study corroborates these previous findings. Two-thirds of recurrence reported were seen in multicystic ameloblastoma, and their histologic variants were follicular and granular ameloblastomas. The remaining one-third was seen in ameloblastic carcinoma. Ameloblastic carcinomas are more aggressive and prone to recurrence than conventional ameloblastoma due to their cellular atypia features such as nuclear pleomorphism, hyperchromatism, high mitotic activity, tissue necrosis and vascular invasion.⁴¹ This is supported by the results of this study and found that ameloblastic carcinoma showed a 25% recurrence rate and time-to-recurrence of 3 months, compared to 3.8% and 3 years, respectively, for conventional ameloblastoma. Nonetheless, this and other studies reported that conventional ameloblastoma may undergo malignant transformation during recurrence.^37,38

We did not find any statistically significant association between recurrence and potential prognostic factors, which may be due to the low recurrence incidence that, in turn, may be due to the short average follow-up period. Ameloblastoma recurrence is usually recorded at 1–10 years postoperative.^36,42 The mean follow-up duration of our study was 2.7 years which may not be adequate for observing recurrence; this risks underestimating the recurrence incidence.

One of the major limitations of this study was the retrospective nature of the data which made it difficult to correlate histopathology and recurrence of the tumours as information such as presence of positive margins after surgery were not recorded. In addition, data on study variables were not evenly available for all included patients. Second, the histopathologic diagnosis reported was based on the acceptable classification of odontogenic tumour at the time of surgical intervention. However, some of these histopathologic diagnoses are no longer recognized by the WHO, due to the ever changing definitions of odontogenic tumours.⁴³ Third, the study reported a relatively small sample size which was due to the strict inclusion and exclusion criteria. Previous Nigerian studies have reported over 1000 Nigerians living with ameloblastoma.²⁰ This study, however, did not include all patients with ameloblastoma, but only those who had surgical intervention. The low number of patients who received treatment may be due to the economic challenges and peculiarities of the study location. Finally, the mean follow-up time was not adequate to assess for recurrence following treatment of ameloblastoma. Although, our study included patients treated from 2012 to 2021, and patients treated during the late study period were not likely to have more than a couple of years of follow-up at the time of data collection, patients treated during the early study period also reported low follow-up time which may be due to patient burnout and economic challenges. We, therefore, recommend that these patients are followed up for a longer period of at least 10 years.