INVITED MEDICAL REVIEW
Ameloblastoma: current etiopathological concepts and management
OA Effiom1, OM Ogundana1, AO Akinshipo1, SO Akintoye2
1Department of Oral and Maxillofacial Pathology/Biology, Faculty of Dental Sciences, University of Lagos, Lagos, Nigeria;
2Department of Oral Medicine, School of Dental Medicine, University of Pennsylvania, Philadelphia, PA, USA
Ameloblastoma is a benign odontogenic tumor of epithelial origin. It is locally aggressive with unlimited growth capacity and has a high potential for malignant transformation as well as metastasis. Ameloblastoma has no established preventive measures although major- ity of patients are between ages 30 and 60 years. Molec- ular and genetic factors that promote oncogenic transformation of odontogenic epithelium to ameloblas- toma are strongly linked to dysregulation of multiple genes associated with mitogen-activated protein kinase, sonic hedgehog, and WNT/b-catenin signaling pathways.
Treatment of ameloblastoma is focused on surgical resection with a wide margin of normal tissue because of its high propensity for locoregional invasion; but this is often associated with significant patient morbidity.
The relatively high recurrence rate of ameloblastoma is influenced by the type of molecular etiological factors, the management approach, and how early the patient presents for treatment. It is expected that further eluci- dation of molecular factors that orchestrate pathogene- sis and recurrence of ameloblastoma will lead to new diagnostic markers and targeted drug therapies for ameloblastoma.
Oral Diseases (2018) 24, 307–316
Keywords: ameloblastoma; jaw; etiopathogenesis; histopatho- genesis; targeted therapies
Introduction
Ameloblastoma is an aggressive odontogenic tumor that forms from odontogenic epithelium within a mature fibrous stroma devoid of odontogenic ectomesenchyme (Sciubba et al, 2005). Although classified as a benign
tumor, ameloblastoma is also the most common odonto- genic tumor of epithelial origin with severe clinical impli- cations (Bassey et al, 2014). Ameloblastoma has a locally aggressive growth pattern; about 70% of cases undergo malignant transformation, and up to 2% metastasize to other sites (Odukoya and Effiom, 2008; DeVilliers et al, 2011).
Ameloblastoma constitutes about 14% of all jaw tumors and cysts, and it is the most prevalent odontogenic tumors in developing countries (Lasisi et al, 2013; Oginni et al, 2015). The global incidence of ameloblastoma is 0.5 cases per million persons per year (Brown and Betz, 2015), and it is a highly encountered odontogenic tumor in Africa and China (Bassey et al, 2014). In the Western Hemi- sphere, ameloblastoma is second to odontoma as the most common odontogenic tumor, but the African American population isfive times more likely to develop ameloblas- toma compared to the Caucasian population (McClary et al, 2016). Most patients with ameloblastoma are between ages 30 and 60 years, but average age at time of diagnosis varies from continent to continent estimated to be approximately 42.3 and 30.4 years in Europe and Africa, respectively (Olusanya et al, 2013; Oomens and van der Waal, 2014). Only 10–15% of ameloblastoma cases occur in the pediatric population, but this can be as high as 25% in Africa and Asia (Bansal et al, 2015).
Ameloblastoma histologically resembles the enamel organ of a developing tooth that has no intention of form- ing dental hard tissues because the stroma lacks the prop- erties of dental mesenchyme. Despite the similarities, it is intriguing that ameloblastoma still displays a distinctive clinically invasive and aggressive growth pattern. Due to naivety and limited healthcare facilities, ameloblastoma patients in developing countries often present with mas- sively grown lesions before seeking care (Anyanechi and Saheeb, 2014; Bassey et al, 2014; Figure 1).
Advances in etiopathogenesis of ameloblastoma
Etiological factors associated with ameloblastoma have evolved over the years and are yet to be conclusively established. Earlier etiological theories were related to
Correspondence: Sunday O. Akintoye, BDS, DDS, MS, Department of Oral Medicine, School of Dental Medicine, University of Pennsylvania, The Robert Schattner Center Room 211, 240 South 40th Street, Philadelphia, PA 19104, USA. Tel: 215 898 9932, Fax: 215 573 7853, E-mail: akintoye@
upenn.edu
Received 19 January 2017; accepted 20 January 2017
All rights reserved www.wiley.com
trauma, inflammation, nutritional deficiencies, non-specific irritation from extractions, and dental caries (Brown and Betz, 2015). As development of odontogenic tumors was associated with remnants of the migrating epithelium at the cervical loop of the enamel organ, it was not surpris- ing that development of ameloblastoma was also linked to the enamel organ, remnants of odontogenic epithelium, and lining of odontogenic cyst (Sciubba et al, 2005). This odontogenic etiological origin was further supported by the similarities in the expression profiles of cytokeratin and vimentin between the developing tooth germ and ameloblastoma (Brown and Betz, 2015). Another earlier theory was associated with morphodifferentiation of pre- ameloblasts to ameloblasts during the bell stage of tooth development. It was believed that affected pre-ameloblasts propagate in the bell stage during tooth development instead of functionally inducing enamel protein synthesis and matrix deposition (Fan et al, 2012). Other studies have proposed that the absence of stratum intermedium hinders the differentiation of pre-ameloblasts to amelo- blasts because the stratum intermedium produces alkaline phosphatase needed to breakdown nutritional elements that will be passed on to ameloblasts during the bell stage.
This theory has been strengthened by the impaired amelo- blast function and enamel deposition observed in Msx-2 null mice that lacked functional stratum intermedium cells (Jussila and Thesleff, 2012). It should also be taken into account that the stellate reticulum within the tumor nests of columnar epithelium can degenerate to form micro- scopic cysts. The coalescing of these microcysts to form larger cystic spaces gives the multicystic features of ameloblastoma (Gupta et al, 2011).
At the molecular level, the genetic factors involved in tooth development, morphogenesis, cytodifferentiation, and tooth patterning have been associated with develop- ment of ameloblastoma because some of these are altered
significantly in ameloblastic tissues. An analysis of 34 dif- ferent genes demonstrated 11 overexpressed and 23 under- expressed genes relative to normal (Heikinheimo et al, 2002). Some of the overexpressed genes include c-fos proto-oncogene (FOS), tumor necrosis factor receptor 1A (TNFRSF1A), collagen type VIII alpha 1 (COL8A1), cyclin-dependent kinase inhibitor 1A (CDKN1A), matrix metalloproteinase 12 (macrophage elastase) (MMP-12), and matrix metalloproteinase 13 (collagenase 3) (MMP- 13). The genes highly underexpressed included sonic hedgehog (SHH), TNF receptor-associated factor 3 (TRAF3), deleted in colorectal carcinoma, Rho GTPase- activating protein 4 (ARHGAP4), cadherin 12 (CDH12), cadherin 13 (CDH13), teratocarcinoma-derived growth factor 1 (TDGF1), transforming growth factor beta 1 (TGFB1), patch (PTCH), and bone morphogenetic protein 2 (BMP2; Heikinheimo et al, 2002). More recent studies have also identified overexpression of WNT5A (wingless- type MMTV integration site family, member 5A) and WNT-1 that suggests they might be associated with the development of ameloblastoma (Sukarawan et al, 2010;
Siar et al, 2012b). These earlier genetic studies were per- formed using microarrays, but more recently, some of these candidate genes have been sequenced to identify the mutations and variants associated with ameloblastoma (Heikinheimo et al, 2002). The molecular pathogenesis of ameloblastoma is now attributed to dysregulation of the mitogen-activated protein kinase (MAPK) pathway based on studies using ameloblastoma tissues, cell lines, and transgenic mice (Brown and Betz, 2015). BRAF, a serine/
threonine protein kinase activating the MAPK/ERK signal- ing pathway strongly associated with melanoma, has also been implicated in over 63% of ameloblastoma (Brown et al, 2014; Kurppa et al, 2014; Sweeney et al, 2014).
Interestingly, over 90% of BRAF mutations involve a sub- stitution of valine for glutamate at codon 600 (V600E;
Kurppa et al, 2014; Brown et al, 2015; Fregnani et al, 2017). The mutation causes constitutive activation of BRAF protein downstream of MEK/ERK that ultimately results in neoplastic transformation (Niault and Baccarini, 2010). To further strengthen the association of MAPK sig- naling with ameloblastoma, the mutations in the RAS gene that acts upstream of BRAF and fibroblast growth factor receptor 2 (FGFR2), a membrane-bound activator of MAPK signaling, have also been identified in ameloblas- tomas (Brown et al, 2014; Sweeney et al, 2014). Addi- tionally, mutations in non-MAPK signaling genes especially smoothened (SMO), a G protein-coupled recep- tor and signaling effector component of the SHH signaling pathway, have also been described in ameloblastoma (Mishra et al, 2015). Taken together, these more recent molecular data strongly indicate the existence of unique genetic abnormalities that eventually lead to development of ameloblastoma (Sweeney et al, 2014; Brown and Betz, 2015; Brown et al, 2015).
Advances in histopathogenesis of ameloblastoma
As our understanding of odontogenic tumors including ameloblastoma increases, the classification of
Figure 1Maxillary Ameloblastoma. Extensive maxillary ameloblastoma in an African male patient. Note the extensive ameloblastoma-induced facial disfigurement and teeth displacement due to delayed access to treat- ment (Picture courtesy of Professor G. T Arotiba, Faculty of Dental Sciences, University of Lagos, Lagos Nigeria)
308
ameloblastoma will continue to evolve (Wright et al, 2014). According to an established World Health Organi- zation (WHO) report on odontogenic tumors, ameloblas- toma was classified into benign and malignant types based on its biologic behavior (Sciubba et al, 2005). Based on this report, each type was further subdivided into four sub- types based on anatomic location and histopathology. In order of frequency, the benign ameloblastomas include (i) solid/multicystic ameloblastoma, (ii) unicystic amelo- blastoma, (iii) peripheral (or extraosseous) ameloblastoma, and (iv) desmoplastic ameloblastoma. Similarly, malignant ameloblastomas based on order of frequency include (i) metastasizing ameloblastoma, (ii) primary ameloblastic carcinoma, (iii) secondary intraosseous ameloblastic carci- noma, and (iv) secondary peripheral ameloblastic carci- noma (Table 1; Reichart et al, 1995; Sciubba et al, 2005).
While the WHO classification provided a good guidance for management, it lacks precise terminology which limits its diagnostic application and international acceptance (Wright et al, 2014). A more contemporary classification of odontogenic tumors including ameloblastoma was pro- posed recently (Wright et al, 2014). The recommendations indicate that there is an authentic and existing concept of
unicystic ameloblastoma, and most conventional ameloblastomas microscopically and macroscopically pre- sent with cystic degeneration. Therefore, the term ‘multi- cystic’ adds no worth but misperception to the designation
‘solid multicystic ameloblastoma’. Hence, conventional lesions should simply be called ‘conventional ameloblas- toma’ and the term ‘solid/multicystic’ be discarded. Addi- tionally, the predominance of plexiform ameloblastoma in the posterior maxilla is not a reason for separating it as a different entity because it is still the distinctive site of occurrence for conventional ameloblastoma, so it should be considered as a histological type of conventional ameloblastoma. It was also proposed that the unique radio- graphic features of desmoplastic ameloblastoma, distinct from other histological types/subtypes, are a function of the desmoplasia. So, desmoplastic ameloblastoma should be simply noted for its distinctive radiographic and clinical presentation and considered a histological type like plexi- form ameloblastoma rather than being considered as a separate entity (Wright et al, 2014). The proposed recom- mendations also stated that unicystic ameloblastoma is comprised of luminal and mural histopathological vari- ants. Luminal unicystic ameloblastoma presents as a cystic
Table 1Clinicohistological types of ameloblastoma. Clinical, histologic, and radiographic features of the histological subtypes of ameloblastoma Clinicohistologic types of ameloblastoma
Types of
ameloblastoma Synonyms Salient features
Conventional
radiographic features Histopathological variants
Benign
Solid/multicystic Conventional/Classical ameloblastoma
Mean age: 36 years Male> female
Slightly higher in mandible
Unilocular radiolucency Multilocular radiolucency Unerupted tooth Root resorption
Cystic, acanthomatous, granular, basaloid, spindle, clear cell, hemangiomatous
Unicystic Cystogenic
ameloblastoma
Dentigerous type:
Mean age: 16.5 years Male> female
Unilocular radiolucency Multilocular radiolucency Unerupted tooth
Luminal (plexiform unicystic, intraluminal), mural
Non-dentigerous type:
Mean age: 35.2 years Female> male Slightly higher in mandible (posteriorly)
Unilocular radiolucency Multilocular radiolucency
Peripheral Extraosseous/soft tissue ameloblastoma
Mean age: 51 years Male> female
Slightly higher in mandible Exophytic
Mean size 1.3 cm
Saucerization Not applicable
Desmoplastic Ameloblastoma with
pronounced desmoplasia
Mean age: 41.6 years Female= male Maxilla= mandible
Mixed radiolucent/
radiopaque Root resorption
Hybrid
Desmoplastic osteoplasia
Malignant
Metastasizing Malignant ameloblastoma Mean age: 34.4 years Male> female
Slightly higher in mandible Distant sites: lungs and other areas
Same as solid/multicystic Same as solid/multicystic
Primary ameloblastic carcinoma
Not applicable Mean age: 53 years Male> female Higher in mandible (posteriorly)
Ill-defined multilocular radiolucency Foci of calcification
Not applicable
Secondary ameloblastic carcinoma (intraosseous)
Carcinoma ex intraosseous ameloblastoma
Rapid growth, 7th decade Male> female
Slightly higher in mandible
Ill-defined multilocular radiolucency Foci of calcification
Not applicable
Secondary ameloblastic carcinoma (peripheral)
Carcinoma ex peripheral ameloblastoma
Male= female
Alveolar bone resorption
Interradicular radiolucency Not applicable
309
lesion lined by ameloblastomatous epithelium which may protrude into the lumen in plexiform proliferations (re- ferred to as intraluminal subtype), while mural unicystic ameloblastoma presents with ameloblastomatous epithelial cells within the cyst wall. The recommendations suggested that mural invasion should be viewed as conventional ameloblastoma due to its relatively higher recurrence rate when compared with the luminal type (Wright et al, 2014).
The most common ameloblastoma is the solid/multicys- tic/conventional type, making up about 91% of all cases of ameloblastoma. It is slow growing and runs a benign course. Histologically, the solid/multicystic/conventional ameloblastoma displays two distinct histological patterns:
the follicular and plexiform types. The follicular type dis- plays proliferating odontogenic epithelial cells arranged in islands, while plexiform type displays epithelial cells arranged in continuous anastomosing strands (Figure 2). It is not uncommon for an ameloblastoma to display both histological patterns. In addition to these two histological types, cystic, granular, acanthomatous, spindle cell, basal cell, clear cell, and other microscopic subtypes have been reported (Table 1).
Unicystic ameloblastoma is the second most common ameloblastoma and accounts for about 5–15% of all cases (Dhanuthai et al, 2012). It is most often seen in younger patients with average age of 26.1 years, and its main loca- tion is the posterior mandible where it often presents as an asymptomatic swelling (Bansal et al, 2015). The majority of unicystic ameloblastomas resemble dentigerous cyst because of their association with an unerupted tooth. The WHO classification (Sciubba et al, 2005) and the more recent recommendations (Wright et al, 2014) consider two main histopathological variants of unicystic ameloblas- toma, the luminal and mural (Figure 3). The luminal vari- ant displays a cystic pattern lined by ameloblastomatous epithelium that protrudes into the lumen as plexiform pro- liferations that look like an intraluminal subtype. The mural variant displays either follicular or plexiform arrangement of ameloblastomatous epithelial cells within the cystic wall. It is not uncommon for both variants to be observed in the same ameloblastoma lesion (Figure 3).
Peripheral ameloblastoma is the least common variant of ameloblastoma, accounting for just 1% of ameloblas- toma cases (Odukoya and Effiom, 2008; Siar et al, 2012a). Mostly affected are middle-aged patients with an average age of 52 years. These lesions are more common in the mandible than the maxilla and are found on poste- rior gingiva or alveolar sulcus. Histologically, peripheral ameloblastoma consists of islands of ameloblastic epithe- lium with similar histological pattern as solid/multicystic/
conventional ameloblastoma (Figure 4a).
Desmoplastic ameloblastoma presents as a slow-grow- ing painless swelling, but radiologically, it displays a mixed radiolucent/radiopaque pattern and irregular bor- ders. The histological feature of extensive stromal dys- plasia is pathognomonic (Figure 4b). It consists of islands of odontogenic epithelium with variable shapes and sizes proliferating within a highly collagenous connective tissue.
The thick collagen fibers tend to compress the odonto- genic epithelial islands from the periphery, giving rise to the bizarre shapes and sizes. It is not uncommon for desmoplastic ameloblastoma to contain metaplastic bone formations.
Metastasizing or malignant ameloblastoma is a previ- ously benign ameloblastoma that has metastasized to a distant site usually the lungs. It is diagnosed based on clinical features and the rationalization that both primary and metastatic lesions display similar histological features of solid/multicystic/conventional ameloblastoma. Amelo- blastic carcinoma can develop de novo: This is the pri- mary type. Alternatively, it can develop secondarily from an initially benign ameloblastoma that loses differentiation to become a carcinoma. Ameloblastic carcinomas grow more rapidly and aggressively and can present as painful swellings that perforate the cortical bone. Histologically, ameloblastic carcinoma combines the overall histological patterns of an ameloblastoma with cytological atypia con- sisting of abnormal mitotic activities, cellular and nuclear hyperchromatism, and focal necrosis (Sciubba et al, 2005) (Figure 5).
Ameloblastoma presents clinically as a slow-growing relatively painless tumor. Due to its locally aggressive growth characteristics, ameloblastoma can rapidly become
(a) (b)
Figure 2Solid/Multicystic/Conventional ameloblastoma. (a) Follicular ameloblastoma connective tissue (Ct) contains islands of odontogenic epithelium (black arrows) delineated by peripheral columnar cells that display reverse polarization. There are islands of cystic degeneration containing central loose stellate reticulum-like cells (*). (b) Plexiform ameloblastoma connective tissue contains long anastomosing cords of odontogenic epithelium (black arrowheads) delineated by peripheral columnar cells that display reverse polarization and central loose stellate reticulum-like cells 310
a massive and expansile tumor causing tooth mobility, tooth displacement, and a grotesque facial appearance if the patient delays getting treatment (Figures 1 and 6). The constitutive activation of BRAFV600E associated with development of ameloblastoma has also been attributed to its progression (Kurppa et al, 2014). This mutation corre- lates with several clinicopathological features of
ameloblastoma such as location within the jaw, age of patient at diagnosis, histology, and possibly prognosis.
For example, BRAF mutations have been shown to occur more in the mandible and in younger patients (Brown and Betz, 2015), while BRAF wild-type ameloblasto- mas occurred more in the maxilla and displayed shorter recurrence-free survival (Brown et al, 2014).
(a) (b)
Figure 3Unicystic ameloblastoma. (a) Luminal unicystic ameloblastoma demonstrates afibrous cystic wall (black arrowhead) lined by ameloblastic epithelium (black arrow) and loose stellate reticulum-like cells (*). (b) Mural unicystic ameloblastoma demonstratesfibrous cystic wall (black arrow), infiltration of ameloblastic epithelium into the cystic wall (black arrowhead), and loose stellate reticulum- like cells (*)
(a) (b)
Figure 4Peripheral and desmoplastic ameloblastomas. (a) Peripheral ameloblastoma demonstrates odontogenic epithelial islands within the connective tissue overlaid by stratified squamous epithelium of the oral mucosa (black arrow). (b) Desmoplastic ameloblastoma demonstrates hyalinized densely collagenized connective tissue stroma (Ct) embedded with islands of cords of odontogenic epithelium (black arrowheads)
(a) (b)
Figure 5Ameloblastic carcinoma. (a) At low magnification, ameloblastic carcinoma contains odontogenic epithelial islands (black arrows) lined by peripheral ameloblast-like cells within the connective tissue stroma (Ct). The epithelial islands contain central stellate reticulum-like cells (*) that display acanthomatous changes (black arrowhead). (b) At higher magnification, the central stellate reticulum-like cells (*) surrounded by peripheral ameloblast-like cells display nuclear hyperchromatism,
pleomorphism, and mitoticfigures
311
Normal bone remodeling is regulated by the interactions of receptor activator of nuclear factor kappa B (RANK) on osteoclast precursors with its osteoblast membrane- bound ligand (RANKL). Osteoblasts also secrete osteopro- tegerin (OPG), a soluble receptor that interacts with RANKL to control RANK–RANKL interactions (Stefanik et al, 2008). As RANK, RANKL, and OPG are expressed in different variants of ameloblastoma, dysregulated RANK-RANKL signaling and altered levels of OPG have been associated with lesional bone loss in ameloblastoma (de Matos et al, 2013). The caspase-mediated apoptotic system is also disparately regulated in different variants of ameloblastoma leading to aberrant survival of ameloblastic tissue. The strong immunoreactivity of pAKT and PI3K in some variants of ameloblastoma, especially the plexiform histological pattern, suggests that AKT/PI3K pathway may be promoting proliferation of ameloblastic cells (Jhamb and Kramer, 2014).
Matrix metalloproteinases (MMPs) are capable of degrading extracellular matrixes, and ameloblastomas express high levels of MMP-1, MMP-2, and MMP-9 (Ribeiro et al, 2009). The local aggressiveness of ameloblastomas correlates with activity of MMP-2 that can degrade type IV collagen abundant in the basement membrane. As MMP-2 level is regulated by the activating and inhibitory actions of MMP-14 and tissue inhibitor of
matrix mettaloproteinases-2 (TIMP-2), respectively, altered levels of the MMP-14/MMP-2/TIMP-2 complex especially in solid multicystic and recurrent ameloblastoma have been attributed to the invasive growth properties of ameloblastoma (Zhang et al, 2010; Florescu et al, 2012).
Also, as MMP-2, MMP-9, and MMP-13 interact with WNT signaling pathway, the local invasiveness of ameloblastoma is further enhanced by WNT-mediated pro- liferative signals (Yamagata et al, 2012). In the same vein, podoplanin, a mucin-type transmembrane glycoprotein, has been shown to be much more strongly expressed in the peripherally located columnar cells than the central stellate reticulum-like cells of ameloblastomas. Hence, a differential expression pattern of podoplanin possibly con- tributes to the migration, aggregation, and recurrence of ameloblastoma cells (Siar et al, 2015).
Other molecular pathways that have been associated with the pathogenesis, invasiveness, and recurrence of ameloblastoma include but are not limited to p53-MDM2 (Kitkumthorn et al, 2010) and Notch signaling pathways (Siar et al, 2010) as well molecular markers such as syn- decan-1 (CD138) (Al-Otaibi et al, 2013; Safadi et al, 2016) and CD10 (Abdel-Aziz and Amin, 2012). Some of these have also been considered as either diagnostic or prognostic markers of ameloblastoma (Jhamb and Kramer, 2014).
(a) (b)
(c) (d)
Figure 6Postsurgical recurrence of mandibular ameloblastoma. Extensive mandibular ameloblastoma (a) in a young female African patient was treated by radical mandibulectomy (b) with acceptable postsurgical appearance (c).
Another ameloblastoma developed in the maxilla 14 years later (d) (Picture courtesy of Professor G. T Arotiba, Faculty of Dental Sciences, University of Lagos, Lagos Nigeria) 312
Current management approaches in ameloblastoma
Diagnostic imaging
The outcomes of a thorough clinical evaluation combined with different imaging modalities and histopathology are paramount to successful management of ameloblastoma irrespective of the histological subtype. Depending on how early the patient presents for evaluation, the clinical appearance of ameloblastoma may range from an innocu- ous intraoral swelling that the patient is unaware of to a grotesque orofacial swelling. Due to delayed access to health care in some developing countries in Africa and Asia, patients often present with a dramatically large ameloblastoma lesion (Figures 1 and 6). Different imaging modalities may have to be combined for evaluation, diag- nosis, and treatment planning of ameloblastomas. These include plain film radiography, cone-beam computed tomography (CT), conventional CT, magnetic resonance imaging (MRI), and functional imaging that combines positron emission tomography (PET) with conventional CT (PET/CT; Fujita et al, 2013). The use of plain film radiography is a good starting point. While it displays to some extent the multilocular pattern of ameloblastoma, it cannot demonstrate the three-dimensional structural expanse of ameloblastoma. Conventional CT with or with- out contrast is the gold standard for evaluation of both pri- mary and recurrent ameloblastomas. It accurately defines the radiodensity as well as multilocular and marginal details of ameloblastoma, which are vital for treatment planning (Figure 7). The use of MRI provides valuable details of the bone marrow and soft tissue components within and beyond the lesional margins of
ameloblastomas. This is especially useful in delineating the extensions of maxillary ameloblastomas within the maxillary sinuses, orbits, and skull (Fujita et al, 2013).
Functional imaging combining PET/CT is particularly use- ful for diagnosing malignant ameloblastoma as well as its extensive soft tissue infiltration and distant metastasis. As imaging studies do not provide definitive diagnosis of an ameloblastoma, it is imperative to biopsy the lesion for histopathological analysis and subtyping.
Surgical approaches
Both primary and recurrent ameloblastomas are treated by either surgical or non-surgical approach. The surgical approach could be conservative (type I) or radical (type II) surgery (Figure 6). The conservative surgical treatment could be in the form of enucleation and cauterization, curettage, cryotherapy, or marsupialization. Conservative surgery preserves the patient’s normal tissues, minimizes facial disfiguration, and supports adequate quality of life postsurgery; but it is prone to higher recurrence especially if the ameloblastoma is the aggressive subtype (Dandriyal et al, 2011). Radical surgical treatment is customarily the treatment of choice for biologically aggressive subtype of primary and recurrent ameloblastomas. It involves en bloc tumor resection with wide bone margin followed by immediate or delayed bone reconstruction of the surgical defect with tissue grafts and prosthetic rehabilitation (Shen et al, 2015). The interrelationship between clinical and histological properties of the ameloblastoma determines its aggressiveness which in turn dictates the treatment approach and recurrence. However, treatment is also affected by the patient’s physical and medical conditions, the patient’s wishes regarding potential facial deformity,
(a) (b)
(c) (d)
Figure 7Tomographic imaging of ameloblastoma. Reconstructed conventional computer tomography (a) and volume rendering (b, anterior; c, right; d, left views) demonstrate outlines of anterior mandibular ameloblastoma (red arrows). (Image courtesy of Drs. Mel Muparappu and David M. Graham, School of Dental Medicine, University of Pennsylvania)
313
compliance, and the psychological effect on quality of life postsurgery. The impact of surgery on facial growth and development in pediatric patients should also be consid- ered during treatment planning. Surgical treatment in pedi- atric ameloblastoma patients is still controversial.
Advocates for conservative surgery favor maintaining a good quality of life post-treatment over recurrence for a pediatric ameloblastoma patient (Huang et al, 2007).
Others, however, support radical surgery for cases with multiple recurrences, in poorly compliant pediatric patients and in environments where follow-up procedures are lim- ited (Odukoya and Effiom, 2008; Bassey et al, 2014).
However, if recurrence is the major consideration, sur- geons are encouraged to select aggressive radical surgery irrespective of the patient’s age except in patients with poor health due to other underlining medical conditions.
Conservative surgical excision with peripheral ostectomy is the usual treatment approach for primary or recurrent peripheral ameloblastomas because it presents as a periph- eral soft tissue lesion. Interestingly, rate of recurrence of peripheral ameloblastomas after conservative surgery is very low (Hertog et al, 2011).
Non-surgical approaches
Different forms of radiation therapy have been used suc- cessfully for non-surgical management of ameloblastomas especially in patients medically unstable for surgery (Kennedy et al, 2016). These include helical tomotherapy, image guided radiation therapy, intensity-modulated radiation therapy, and proton beam therapy. Some of these treat- ment modalities have been combined with surgery and/or chemotherapy. The therapeutic use of adjuvant radiother- apy with or without chemotherapy for positive margins of recurrent and unresectable ameloblastomas has resulted in mixed outcomes. However, their use is still strongly advo- cated to treat ameloblastic carcinoma and recurrent ameloblastoma after multiple postsurgical recurrences (Huang et al, 2014). During treatment planning, it is imperative to balance the efficacy of radiotherapy with risks of developing future life-threatening malignant trans- formations.
The efficacy of chemotherapy in the management of primary and recurrent ameloblastomas is still being explored as chemotherapy can improve clinical outcomes in non-surgical patients. Several drug regimens may be used in combination with surgical resection and/or or radiotherapy. These include the combinations of vinbl- astine + cisplatin + bleomycin; adriamycin + cisplatin + cyclophosphamide; doxorubicin + cisplatin; and gemc- itabine + carboplatin (Van Dam et al, 2010; Amzerin et al, 2011). However, there is still a need for more multi- center randomized controlled clinical studies to validate the use of radiation and chemotherapy as treatment options for ameloblastoma. Interestingly, bone loss in ameloblastoma associated with dysregulated RANK/
RANKL/OPG interactions (de Matos et al, 2013) has led to the suggestion that antiresorptives such as denosumab may be effective in controlling the local aggressiveness of ameloblastoma (Jhamb and Kramer, 2014). Unfortunately, a major side effect of antiresorptives is osteonecrosis of the jaw (Akintoye and Hersh, 2016; Omolehinwa and
Akintoye, 2016), so the potential clinical benefits must be balanced with their established side effects to justify their use in ameloblastoma therapy.
Targeted directed therapies
Recent advances in the molecular signaling pathways associated with pathogenesis of ameloblastoma have led to development of targeted therapies for management of ameloblastoma (Sauk et al, 2010). Several MAPK-specific drugs selectively inhibit the functions of mutated BRAF and MEK to stop the dysregulated proliferation and differ- entiation of ameloblastic cells. These include vemurafenib and dabrafenib, which inhibit mutated BRAF gene; trame- tinib, an inhibitor of mutated MEK gene; and ponatinib and regorafenib that inhibit mutated FGFR2 genes.
Regrettably, resistance mechanisms such as compensatory activation of the MAPK kinase pathway by epidermal growth factor receptor have been associated with vemu- rafenib treatment for ameloblastoma (Kurppa et al, 2014;
Heikinheimo et al, 2015). This had led to the suggestion that mutated MEK inhibitors may be desirable than mutant BRAF inhibitors for treating ameloblastoma (Heikinheimo et al, 2015).
Similarly, targeted therapies have been developed to control the effect of SMO mutation associated with patho- genesis of ameloblastoma (Mishra et al, 2015). These include vismodegib and itraconazole, which unfortunately have been less successful in controlling ameloblastoma associated with SMO mutations W535L and L412F due to resistance mechanisms that block binding of SMO-targeted drugs (Sweeney et al, 2014). On the contrary, arsenic tri- oxide and KAAD-cyclopamine are known to be highly effective against these same mutations and may be useful in the treatment of ameloblastoma associated with SHH signaling pathway (Sweeney et al, 2014). As SHH expres- sion is high in ameloblastomas, several drugs already developed to antagonize SHH signaling offer other non- surgical targeted therapeutic options for ameloblastoma patients (Mishra et al, 2015). Among these, cyclopamine is the most widely used, but its main drawback is the inhi- bition of osteoblastic proliferation and differentiation that are important for bone healing (Stanton and Peng, 2010;
Schaefer et al, 2013).
Recurrence of ameloblastoma
The relatively high post-treatment recurrence of ameloblastoma is a major challenge (Figure 6). This can be attributed to its local invasiveness, different histological variants with peculiar tissue components, the treatment approach, and how early the patient presents for treatment (Ribeiro et al, 2009; Zhang et al, 2010). The potential for tumor seeding at the surgical site is also attributed to high recurrence of ameloblastoma. Solid/multicystic/conven- tional ameloblastoma is associated with the highest rate of recurrence (Antonoglou and Sandor, 2015) especially if treated by conservative surgery. Similarly, the tendency to treat luminal unicystic ameloblastoma by conservative approach also leads to recurrence, but recurrence is much lower for unicystic than solid multicystic ameloblastoma and malignant ameloblastoma (Antonoglou and Sandor,
314
2015). Interestingly, SMO gene mutation appears to be associated with higher recurrence of ameloblastoma than other genetic mutations identified in ameloblastoma (Sweeney et al, 2014). This has led to the assertion that BRAF and other non-SMO genetic mutations might confer better prognosis (Brown et al, 2014). A conservative sur- gical approach to treat ameloblastic carcinoma also leads to high recurrence, so radical surgery is mostly advocated for treatment of ameloblastic carcinoma irrespective of its histological features. The recommended treatment approach for recurrent ameloblastoma is radical surgery which confers disease-free survival and the absence of secondary recurrence for at least 10 years (Hertog et al, 2011). There are promising results that higher expression levels of cancer stem cells as indicated by CD44 in ameloblastoma may be associated with tumor recurrence, but the sensitivity and specificity of CD44 as a test for ameloblastoma recurrence are yet to be fully clarified.
(Sathi et al, 2012).
Taken together, successful management of primary and recurrent ameloblastomas involves balancing radical sur- gery that has a margin wide enough to prevent recurrences with another less tissue destructive therapeutic option (Kennedy et al, 2016). The goals are to minimize morbid- ity and improve survivorship and the patient’s quality of life. Adequate periodic clinical and radiographic follow-up for at least 10 years is highly essential. There is optim- ism that ongoing research will eventually lead to newer therapies targeting both MAPK and non-MAPK signaling pathways.
Acknowledgements
This work was supported by United States Institute of Interna- tional Education/the Council for International Exchange of Schol- ars (IIE/CIES) Fulbright scholarship (awarded to Sunday O.
Akintoye). The authors thank Professor G. T Arotiba (Faculty of Dental Sciences, University of Lagos, Lagos Nigeria) for the clinical pictures and Drs. Mel Muparappu and David M. Graham (School of Dental Medicine, University of Pennsylvania) for the radiology images.
Conflict of interest
The authors declare that they have no conflict of interest.
Author contributions
O.A. Effiom, O.M. Ogundana and A.O. Akinshipo participated in literature review and writing of the manuscript. S.O. Akintoye conceptualized the review project and participated in the litera- ture review, writing and editing of the manuscript.
References
Abdel-Aziz A, Amin MM (2012). EGFR, CD10 and proliferation marker Ki67 expression in ameloblastoma: possible role in local recurrence. Diagn Pathol 7: 14.
Akintoye SO, Hersh EV (2016). Impact of communication between physicians and dentists on the incidence of jaw osteonecrosis caused by bone anti-resorptives. Curr Med Res Opin 32: 1455–1456.
Al-Otaibi O, Khounganian R, Anil S, Rajendran R (2013). Syn- decan-1 (CD 138) surface expression marks cell type and dif- ferentiation in ameloblastoma, keratocystic odontogenic tumor, and dentigerous cyst. J Oral Pathol Med 42: 186–193.
Amzerin M, Fadoukhair Z, Belbaraka R et al (2011). Metastatic ameloblastoma responding to combination chemotherapy: case report and review of the literature. J Med Case Rep 5: 491.
Antonoglou GN, Sandor GK (2015). Recurrence rates of intraosseous ameloblastomas of the jaws: a systematic review of conservative versus aggressive treatment approaches and meta-analysis of non-randomized studies. J Craniomaxillofac Surg 43: 149–157.
Anyanechi CE, Saheeb BD (2014). A review of 156 odontogenic tumours in Calabar, Nigeria. Ghana Med J 48: 163–167.
Bansal S, Desai RS, Shirsat P, Prasad P, Karjodkar F, Andrade N (2015). The occurrence and pattern of ameloblastoma in children and adolescents: an Indian institutional study of 41 years and review of the literature. Int J Oral Maxillofac Surg 44: 725–731.
Bassey GO, Osunde OD, Anyanechi CE (2014). Maxillofacial tumors and tumor-like lesions in a Nigerian teaching hospi- tal: an eleven year retrospective analysis. Afr Health Sci 14:
56–63.
Brown NA, Betz BL (2015). Ameloblastoma: a review of recent molecular pathogenetic discoveries. Biomark Cancer 7: 19–24.
Brown NA, Rolland D, McHugh JB et al (2014). Activating FGFR2-RAS-BRAF mutations in ameloblastoma. Clin Cancer Res 20: 5517–5526.
Brown NA, Betz BL, Weigelin HC, Elenitoba-Johnson KS, Lim MS, Bailey NG (2015). Evaluation of allele-specific PCR and immunohistochemistry for the detection of BRAF V600E mutations in hairy cell leukemia. Am J Clin Pathol 143: 89–99.
Dandriyal R, Gupta A, Pant S, Baweja HH (2011). Surgical management of ameloblastoma: conservative or radical approach. Natl J Maxillofac Surg 2: 22–27.
DeVilliers P, Suggs C, Simmons D, Murrah V, Wright JT (2011). Microgenomics of ameloblastoma. J Dent Res 90:
463–469.
Dhanuthai K, Chantarangsu S, Rojanawatsirivej S et al (2012).
Ameloblastoma: a multicentric study. Oral Surg Oral Med Oral Pathol Oral Radiol 113: 782–788.
Fan K, Andrews BT, Tabit CJ, Bradley JP (2012). The Sarnat studies in craniofacial biology. J Craniofac Surg 23: 37–43.
Florescu A, Margaritescu C, Simionescu CE, Stepan A (2012).
Immunohistochemical expression of MMP-9, TIMP-2, E-cad- herin and vimentin in ameloblastomas and their implication in the local aggressive behavior of these tumors. Rom J Morphol Embryol 53: 975–984.
Fregnani ER, Perez DE, Paes de Almeida O et al (2017). BRAF- V600E expression correlates with ameloblastoma aggressive- ness. Histopathology 70: 473–484.
Fujita M, Matsuzaki H, Yanagi Y et al (2013). Diagnostic value of MRI for odontogenic tumours. Dentomaxillofac Radiol 42:
20120265.
Gupta N, Saxena S, Rathod VC, Aggarwal P (2011). Unicystic ameloblastoma of the mandible. J Oral Maxillofac Pathol 15:
228–231.
Heikinheimo K, Jee KJ, Niini T et al (2002). Gene expression profiling of ameloblastoma and human tooth germ by means of a cDNA microarray. J Dent Res 81: 525–530.
Heikinheimo K, Kurppa KJ, Elenius K (2015). Novel targets for the treatment of ameloblastoma. J Dent Res 94: 237–240.
Hertog D, Schulten EA, Leemans CR, Winters HA, Van der Waal I (2011). Management of recurrent ameloblastoma of the jaws; a 40-year single institution experience. Oral Oncol 47:
145–146.
315
Huang IY, Lai ST, Chen CH, Chen CM, Wu CW, Shen YH (2007).
Surgical management of ameloblastoma in children. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 104: 478–485.
Huang CM, Chen JY, Chen CH, Huang CJ (2014). Radiotherapy for a repeatedly recurrent ameloblastoma with malignant trans- formation. Head Neck 36: E1–E3.
Jhamb T, Kramer JM (2014). Molecular concepts in the patho- genesis of ameloblastoma: implications for therapeutics. Exp Mol Pathol 97: 345–353.
Jussila M, Thesleff I (2012). Signaling networks regulating tooth organogenesis and regeneration, and the specification of dental mesenchymal and epithelial cell lineages. Cold Spring Harb Perspect Biol 4: a008425.
Kennedy WR, Werning JW, Kaye FJ, Mendenhall WM (2016).
Treatment of ameloblastoma and ameloblastic carcinoma with radiotherapy. Eur Arch Otorhinolaryngol 273: 3293–3297.
Kitkumthorn N, Yanatatsaneejit P, Rabalert J, Dhammawipark C, Mutirangura A (2010). Association of P53 codon 72 polymor- phism and ameloblastoma. Oral Dis 16: 631–635.
Kurppa KJ, Caton J, Morgan PR et al (2014). High frequency of BRAF V600E mutations in ameloblastoma. J Pathol 232: 492–
498.
Lasisi TJ, Adisa AO, Olusanya AA (2013). Appraisal of jaw swellings in a Nigerian tertiary healthcare facility. J Clin Exp Dent 5: e42–e47.
de Matos FR, de Moraes M, das Neves Silva EB, Galvao HC, de Almeida Freitas R (2013). Immunohistochemical detection of receptor activator nuclear kappaB ligand and osteoprote- gerin in odontogenic cysts and tumors. J Oral Maxillofac Surg 71: 1886–1892.
McClary AC, West RB, McClary AC et al (2016). Ameloblas- toma: a clinical review and trends in management. Eur Arch Otorhinolaryngol 273: 1649–1661.
Mishra P, Panda A, Bandyopadhyay A, Kumar H, Mohiddin G (2015). Sonic hedgehog signalling pathway and ameloblastoma – a review. J Clin Diagn Res 9: ZE10–ZE13.
Niault TS, Baccarini M (2010). Targets of Raf in tumorigenesis.
Carcinogenesis 31: 1165–1174.
Odukoya O, Effiom OA (2008). Clinicopathological study of 100 Nigerian cases of ameloblastoma. Niger Postgrad Med J 15: 1–5.
Oginni FO, Stoelinga PJ, Ajike SA et al (2015). A prospective epidemiological study on odontogenic tumours in a black Afri- can population, with emphasis on the relative frequency of ameloblastoma. Int J Oral Maxillofac Surg 44: 1099–1105.
Olusanya AA, Adisa AO, Lawal AO, Arotiba JT (2013). Gross surgical features and treatment outcome of ameloblastoma at a Nigerian tertiary hospital. Afr J Med Med Sci 42: 59–64.
Omolehinwa TT, Akintoye SO (2016). Chemical and radiation- associated jaw lesions. Dent Clin North Am 60: 265–277.
Oomens MA, van der Waal I (2014). Epidemiology of ameloblastomas of the jaws; a report from the Netherlands.
Med Oral Patol Oral Cir Bucal 19: e581–e583.
Reichart PA, Philipsen HP, Sonner S (1995). Ameloblastoma:
biological profile of 3677 cases. Eur J Cancer B Oral Oncol 31B: 86–99.
Ribeiro BF, Iglesias DP, Nascimento GJ, Galvao HC, Medeiros AM, Freitas RA (2009). Immunoexpression of MMPs-1, -2, and -9 in ameloblastoma and odontogenic adenomatoid tumor.
Oral Dis 15: 472–477.
Safadi RA, Quda BF, Hammad HM (2016). Immunohistochemi- cal expression of K6, K8, K16, K17, K19, maspin, syndecan-1
(CD138), alpha-SMA, and Ki-67 in ameloblastoma and ameloblastic carcinoma: diagnostic and prognostic correlations.
Oral Surg Oral Med Oral Pathol Oral Radiol 121: 402–411.
Sathi GA, Tamamura R, Tsujigiwa H et al (2012). Analysis of immunoexpression of common cancer stem cell markers in ameloblastoma. Exp Ther Med 3: 397–402.
Sauk JJ, Nikitakis NG, Scheper MA (2010). Are we on the brink of nonsurgical treatment for ameloblastoma? Oral Surg Oral Med Oral Pathol Oral Radiol Endod 110: 68–78.
Schaefer GI, Perez JR, Duvall JR, Stanton BZ, Shamji AF, Schreiber SL (2013). Discovery of small-molecule modulators of the Sonic Hedgehog pathway. J Am Chem Soc 135: 9675–
9680.
Sciubba JJ, Eversole LR, Slootweg PJ (2005). Odontogenic tumours. In: Barnes L, Eveson JW, Reichart P, Sidransky D, eds. World Health Organization classification head and neck tumours. IARC Press: Lyon, pp. 283–328.
Shen YF, Rodriguez ED, Wei FC, Tsai CY, Chang YM (2015).
Aesthetic and functional mandibular reconstruction with imme- diate dental implants in a free fibular flap and a low-profile reconstruction plate: five-year follow-up. Ann Plast Surg 74:
442–446.
Siar CH, Nagatsuka H, Chuah KS et al (2010). Notch4 overex- pression in ameloblastoma correlates with the solid/multicystic phenotype. Oral Surg Oral Med Oral Pathol Oral Radiol Endod 110: 224–233.
Siar CH, Lau SH, Ng KH (2012a). Ameloblastoma of the jaws:
a retrospective analysis of 340 cases in a Malaysian popula- tion. J Oral Maxillofac Surg 70: 608–615.
Siar CH, Nagatsuka H, Han PP et al (2012b). Differential expression of canonical and non-canonical Wnt ligands in ameloblastoma. J Oral Pathol Med 41: 332–339.
Siar CH, Ishak I, Ng KH (2015). Podoplanin, E-cadherin, beta-catenin, and CD44v6 in recurrent ameloblastoma: their distribution patterns and relevance. J Oral Pathol Med 44:
51–58.
Stanton BZ, Peng LF (2010). Small-molecule modulators of the Sonic Hedgehog signaling pathway. Mol BioSyst 6: 44–54.
Stefanik D, Sarin J, Lam T, Levin L, Leboy PS, Akintoye SO (2008). Disparate osteogenic response of mandible and iliac crest bone marrow stromal cells to pamidronate. Oral Dis 14:
465–471.
Sukarawan W, Simmons D, Suggs C, Long K, Wright JT (2010). WNT5A expression in ameloblastoma and its roles in regulating enamel epithelium tumorigenic behaviors. Am J Pathol 176: 461–471.
Sweeney RT, McClary AC, Myers BR et al (2014). Identifica- tion of recurrent SMO and BRAF mutations in ameloblas- tomas. Nat Genet 46: 722–725.
Van Dam SD, Unni KK, Keller EE (2010). Metastasizing (malig- nant) ameloblastoma: review of a unique histopathologic entity and report of Mayo Clinic experience. J Oral Maxillofac Surg 68: 2962–2974.
Wright JM, Odell EW, Speight PM, Takata T (2014). Odonto- genic tumors, WHO 2005: where do we go from here? Head Neck Pathol 8: 373–382.
Yamagata K, Li X, Ikegaki S et al (2012). Dissection of Wnt5a- Ror2 signaling leading to matrix metalloproteinase (MMP-13) expression. J Biol Chem 287: 1588–1599.
Zhang B, Zhang J, Huang HZ, Xu ZY, Xie HL (2010). Expres- sion and role of metalloproteinase-2 and endogenous tissue regulator in ameloblastoma. J Oral Pathol Med 39: 219–222.
316
