A multi-centre evaluation of malignant odontogenic tumours in Nigeria

Introduction odontogenic tumors originate from neoplastic transformation of the remnants of tooth forming apparatus. There are varying degrees of inductive interactions between odontogenic ectomesenchyme and epithelium during odontogenesis, leading to lesions that vary from benign to malignant. Malignant odontogenic tumours (MOTs) are very rare and are classified according to embryonic tissue of origin. Recently, there has been a few changes to the classification of MOTs according to the World Health Organization's (WHO) classification in 2017. This study aims to evaluate and reclassify MOTs, using a multi-centre approach in some major tertiary dental hospitals in Nigeria. Methods this study reviewed the clinicopathological data on 63 cases of MOT diagnosed over 25 years in five major tertiary dental hospitals in Nigeria. All MOT cases were reclassified according to the recent revision to the 2017 WHO classification of odontogenic tumours. Results from a total of 10,446 biopsies of oral and jaw lesions seen at the 5 study centres over the 25-year study period, 2199 (21.05%) cases were found to be odontogenic tumours (OTs), of which 63 were MOT. MOTs constituted 0.60% of the total biopsy cases and 2.86% of OTs. Odontogenic carcinomas presented with a mean age higher than odontogenic sarcomas. According to our 2017 WHO reclassification of MOTs, odontogenic carcinomas, ameloblastic carcinomas and primary intraosseous carcinomas were found to be the top three lesions, respectively. Carcinosarcomas were found to be extremely rare. Conclusion using a multi-centre approach is a robust way to reduce diagnostic challenges associated with rare maxillofacial lesions such as MOTs.


Introduction
Odontogenic tumours (OTs) constitute a wide range of lesions that are derivatives of tooth forming apparatus via neoplastic transformation of remnants of odontogenesis and odontogenic cyst [1]. Tooth forming apparatus and their embryogenic rest cells such as dental lamina (and its cell rests of Serres); enamel organ (reduced enamel epithelium); Hertwig epithelial root sheath (HERS) and its residue; epithelial cell rests of Malassez; dental papilla; and dental follicle, have all been described as possible sources of OTs.
There is a varying degrees of inductive interaction between these embryogenic components of the developing tooth germ [2,3]. These odontogenic remnants are capable of developing into epithelial and mesenchymal tissues. This differentiation potential forms the basis for the World Health Organization (WHO)'s classification of OTs into benign and malignant tumors [4].
Based on published literature, odontogenic tumors are rare lesions with varying frequency. While some authors have estimated its frequency of occurrence to be 1% [2,3], others have reported higher values around 32% [3,5]. Based on other studies, it constitutes about 4% of oral and maxillofacial biopsy specimens in oral pathology services [6,7]. Similar to the classification of benign OTs, malignant odontogenic tumours (MOTs) are classified based on their histogenesis. They can emerge from epithelial components of odontogenesis; ectomesenchymal/mesenchmal remnants; or from mixed origin, consisting of both the epithelial and mesenchymal aspect [8]. MOTs are believed to constitute between 0-6.1% of OTs [9], hence they are extremely rare lesions that are exclusively located in the jaws. They arise within the jaws either as a primary lesion (de novo); from epithelial cystic linings; or via malignant transformation of a benign OTs, even though reports of malignant transformation of odontogenic cystic lining is rare [10]. This category of lesions present with diagnostic and therapeutic dilemma, consequent to their rarity coupled with cumbersome and complex histopathological features [9,10]. MOTs are often locally aggressive with radical surgery being the mainstay of treatment.
Histopathologically, MOTs can be carcinomas, sarcomas or carcinosarcomas; however the most common types are the carcinomas [8,9].
The etiology of MOTs is idiopathic, even though research is ongoing to elucidate the underlying molecular pathogenetic mechanisms [10].
There is dearth of knowledge in the scientific literature regarding MOTs and most of the current information as to their origin, clinicopathological features, biological behavior, and therapeutic options are derived from case reports and a few series of published cases. This results in failure to develop standardized guidelines for management, diagnostic criteria and treatment protocols for MOTs [9]. Recently, World Health Organization in 2017 updated their classification of odontogenic tumors with addition of new entity in the carcinomatous and sarcomatous groups of MOTs, viz: sclerosing odontogenic carcinoma and odontogenic carcinosarcoma, and reclassification of metastasizing ameloblastoma from malignant epithelial carcinoma to benign neoplasm [11][12][13][14]. Despite the evolution of modern diagnostic techniques, arriving at a precise diagnosis of MOTs is still a difficult task [15]. Hence, this study represents one of the largest descriptive epidemiology of MOTs, compiled using a multi-centre approach among academic referral institutions in Nigeria. Using this approach, we determined the incidence, demographics and clinicopathological features of MOTs in sub-Saharan Africa.

Study design, participating centers and data sources
This was a 25 years retrospective review from five academic medical centers located in the south-western and eastern geopolitical region Page number not for citation purposes 3

Case selection and exclusion criteria
Cases without adequate clinical and histopathological information were excluded from the study. A total of 63 qualified cases of MOTs were selected for this study from the five participating centres. Cases were classified according to the 2017 WHO classification of odontogenic tumours. Selected cases were categorized by histological types, site of primary tumour, age and gender distribution.

Data analysis
Data from the selected 63 cases were collated and processed using the SPSS data analysis software (version 20.0; SPSS Inc. Chicago, IL).
Categorical variables were analyzed as frequencies and percentages, while quantitative variables were summarized as means and standard deviation. cases of OTs were found, out of which 63 were cases of MOT. MOTs constituted (0.60%) of the total biopsy cases and 2.86% of OTs.

Age distribution
The mean age of MOT in this study was 39.9 years ± 16.4, ranging from 4 and 76 years, with a peak age incidence observed in the 4 th decade of life; although majority of the cases were seen among the 3 rd , 4 th and 5 th decades. Odontogenic carcinoma (OC) presented with a mean age of 40.7 ± 16.6, which is higher than that of odontogenic sarcoma (OS) and carcinosarcoma although it was not statistically significant. Mean age duration for males was 40.8 ± 13.5 and 38.7 ± 20.4 for females while the overall mean duration of the lesion prior to presentation was about 3.7 years with a range of 4 months to 16 years. The age distribution data obtained are summarised in Table 1.

Swelling and pain
Swelling was observed consistently in all cases (100%) however about 91.1% of the lesions showed buccolingual cortical bone expansions. Pain was present in 67.7% of cases while ulceration was observed in 46%. The mean size of lesions was 11.3cm ± 5.9 at the widest diameter.

Gender and site distribution
Malignant odontogenic tumour showed male preponderance of 39 cases (61.9%) as compared to females which were 24 cases (38.1%); with a male: female ratio of 1.6:1. We observed a marked mandibular predilection of 96.82% as compared to only 1 (1.59%) maxillary case.
Left mandibular side was the most affected (n=48, 77.4%) site in our study ( Table 2).

Histopathological distribution
Distribution of tumor was reported according to the WHO 2017 classification ( Figure 1). OC constituted 57 (90.5%) of the cases; 5 (7.9%) cases were OS; while only 1 (1.6%) case of carcinosarcoma was seen. Of the OC in this study, ameloblastic carcinoma (AC) constituted the highest frequency of 50 (87.7%) cases and 79.4% of the total MOT; while primary intraosseous carcinomas (PIOC) made up 12.3% cases of OC and 11.1% of MOT.

Discussion
MOT is a very rare subgroup of OTs. Previous studies have shown that the incidence of MOTs varies with OTs distribution (in terms of ethnics, racial and geographic variation) in the range of 0-6.1% [4,5,8]. In the present study, incidence of about 2.9% (n=63) of MOTs out of 2199 OTs was recorded and this is in consonance with the previous reports on rarity of this lesion. Previous studies that documented higher incidence are mostly from Africa and Asia with a range from 2.7-6.1% while those from America and Europe documented a lower range of about 1% [4 ,9, 10, 15, 17]. Contrary to the observation of a frequency of 2.9% for MOTs in our study, previous studies among Chinese population documented an incidence of 6.1% [18] and 4.7% [19]. Frequencies observed from Latin and South American studies vary between 1.2% and 2.2% [9,20].
However in Europe, Rubini et al. (2017), in a 25-years retrospective study of odontogenic neoplasm among Italian population recorded an incidence of 1.1% [7]. A frequency of about 0.36% was observed in a study from the United State of America [15]. Among African studies, the incidence varies from 1.26-5.6% [10,[21][22][23][24][25] Similar to other studies [10,20], the overall mean age of MOTs in our study was 39.9±16.4 years (ranging from 4 to 76, and a peak age in the 4 th decade of life), and most cases (n=43, 68.3%) were seen within the 3 rd , 4 th and 5 th decade of life. However, our findings were contrary (and lower) to other reports [7,9,26]. In the study of Males are frequently more affected in MOTs [25], which conforms with our observation of a male to female ratio of 1.6:1 in this study.
Our study showed a marked mandibular involvement with a frequency of 98.4%; while the frequency for maxillary MOT lesion was 1.6%.
This mandibular predominance tallies with the findings from other studies [19,20,[26][27][28]; even though a maxillary predilection of 66.7% had been previously observed [7]. The reason for the high mandibular predilection may be attributed to similar observation in benign OTs such as ameloblastoma [6,18]; and also the retention of the epithelial odontogenic rest with potential to develop into cysts or tumours in the mandible [6]. MOT can be classified into odontogenic Carcinomas (OC), odontogenic sarcoma (OS) and carcinosarcoma (CS); although OC is observed much more frequently than others [3,4,10,22]. On the other hand, studies have shown that sarcomas are extremely rare malignancy accounting for about 1% of all malignancy [29]. We identified 5 (7.9%) cases of OS in our study.
We recorded just one case (1.6%) of CS, showing the extent of its rarity. The reason for the low percentage of CS in our study may be due to the fact that it has not been a well recognised entity until the recent WHO 2017 classification.
OC presented with the highest value of mean age 40.7±16.6 years, compared to OS and CS which accounted for 34.6±15.9 years and 25 years, respectively. This is probably an indication that carcinomas are likely to be seen more in late adolescent/elderly. Similar difference in mean age between OC and OS was also observed in another study [20], while it is contrary to the findings of Jing et al. (2007) [19].
AC was the most common OC, with 50 (87.7%) cases in this study and 79.4% of all MOTs. This has equally been observed in other studies [4,[8][9][10]19], except for a study in Thailand where clear cell odontogenic carcinoma (CCOC) was found to be the most frequent OC and MOT, with a frequency of 70% and 46.7%, respectively. This was followed by OS which was 33.3% of MOTs [30]. CCOC and ghost cell odontogenic carcinoma (GCOC) are very rare malignant epithelial odontogenic neoplasm. CCOC was previously considered as a benign OT with biologic aggressive nature and later reclassified as a malignant lesion in WHO 2005 edition [16], based on its destructive and metastatic behaviour. GCOC is an OC with features of calcifying cystic odontogenic tumor (CCOT) and/or dentinogenic ghost cell tumor (DGCT), presenting variable histopathologic features [16,31].
Although neither of the lesions were found in our study, Martinez et al. (2014) [20], however observed a CCOC frequency of 12% of the MOTs and 15.8% of OC. In addition, Jing et al. (2007) [19], reported the frequency of CCOC in their study as 4% of MOTs and 4.2% of OC; and GCOC as 10% and 10.4% of MOTs and OC, respectively. AC in our study, presented with a mean age of 39.4±15.9 years, with a peak age incidence in 4 th decade. This result was similar to other studies [10,19]. It is presumed that the frequency of AC would be much higher in prospective MOTs studies, and this could be due to the reclassification of malignant ameloblastoma (MA) as a benign OTs in 2017 WHO classification. Similar to AC, PIOC was also narrowed from the previous three sub-classifications in the 2005 WHO classification, to a single entity in the 2017 WHO classification [16].
It develops probably from residues of odontogenic epithelium or its rest cell, and is located within bone, without demonstrable evidence of a primary or metastatic carcinoma in other sites or from the oral or sinonasal mucosa [3]. PIOC was the second most common MOTs in our series, constituting 12.3% and 11.1% of OCs and MOTs, respectively. This is in agreement with other studies [10,19,20]. It presented with a mean age of 50.3±19.6 years, which concur with other studies [8,19], but is however lower than a mean age of 56. Sclerosing odontogenic carcinoma is a new entity that was first described by Koutlas et al. (2008) [32]. About 10 cases of this lesion have been reported so far [16,32]. It is a distinct entity with histopathology features of a densely sclerotic stroma, bland cytology characterized by small single-file cords and strands of epithelium, and aggressive infiltrative growth into the muscles and nerves [16]. Being a recently discovered entity, pathologists should be vigilant when encountered with a lesion of such characteristic features. under the umbrella of odontogenic sarcomas [3,16]. These lesions are less frequent than carcinomas as observed in this study, and supported by others [25,26] need for proper clinicopathological review of these lesions in order to avoid wrong histopatological impression is essential.
Generally, treatment advocated for the proper management of these lesions is wide surgical excision followed by post-operative radiotherapy. Chemotherapy is not an effective option in treating the MOTs, although studies have suggested that the prognosis is dependent on the clinical type and location of tumour, age of the patient at diagnosis, tumour size and treatment modality [3,35].
Advanced age at diagnosis, large tumour size, post-operative radiotherapy and OS have been found to have a negative impact on survival [3,35].

Conclusion
We

Competing interests
The authors declare no competing interests.     Fisher's exact p=0.011 (significant)