Epidemiologic analysis of salivary gland tumors over a 10-years period diagnosed in a northeast Brazilian population

Background Salivary gland tumors (SGT) correspond to a heterogeneous group of lesions with variable biological behavior. The present study aimed to determine the distribution and demographic findings of salivary gland neoplasms in a northeast Brazilian population. Material and Methods A retrospective descriptive cross-sectional study was performed. A total of 588 cases of SGT were diagnosed between 2006 and 2016 of 4 pathology services in the state of Sergipe, Brazil. All cases were reviewed, and data such as sex, age, anatomical location, and histopathological diagnosis were collected. Results A total of 470 (79.9%) tumors were benign and 118 (20.1%) were malignant. The majority of the patients were females (n=328, 55.8%) with an overall female:male ratio of 1.2:1. The major salivary glands were affected more than the minor glands (69.5% vs. 30.5%). Pleomorphic adenoma (n=419, 71.3%) and mucoepidermoid carcinoma (n=29, 4.9%) were the most frequent benign and malignant tumors, respectively. In addition, both benign and malignant tumors occurred more frequently in the parotid gland (n=300, 51%, p<0.05). Conclusions The epidemiologic profile and clinical characteristics of SGT were similar to those described in other countries and other regions of Brazil. Epidemiological studies of SGT help to understand their clinical and pathological features and are essential to establish the proper management and prognosis. Key words:Salivary gland, tumors, epidemiology, head and neck pathology.


Introduction
Salivary gland tumors (SGT) are uncommon lesions that present a wide variation in relation to the clinical, histological, and biological aspects (1,2). In addition, these lesions often represent a diagnostic challenge for the pathologist due to the overlapping of morphological findings (2,3). SGT account for about 3 to 6% of all tumors in the head and neck region, with an annual estimated global incidence ranging from 0.4 to 13.5 cases per 100,000 individuals (2,4). Although several studies evaluate the frequency and incidence of these tumors in Brazil (1,2,(5)(6)(7)(8)(9)(10) and other countries of the world (11)(12)(13)(14)(15)(16), the epidemiological data of these lesions is not well established because there is a wide variation in the incidence and prevalence of these tumors across countries, indicating a geographic variation in the frequency of these neoplasms (2,9). Thus, the objective of the present study was to describe the demographic and clinical aspects of salivary gland neoplasms diagnosed in 4 reference pathology centers in the state of Sergipe (Aracaju, Brazil), and to compare the findings with epidemiological data from different geographic locations.

Material and Methods
-Study design In this study, the files of 4 surgical pathology centers in Aracaju, Sergipe State, Brazil were retrospectively reviewed: Laboratory of Surgical Pathology of the University Hospital of the Federal University of Sergipe (HU-UFS), Oral Pathology Service of the Tiradentes University (UNIT), and two private general pathology services. During a 10-year period, between January 2006 and December 2016, 588 cases of salivary gland neoplasms were retrieved from these archives.
-Sample All cases of salivary gland tumors were retrieved, and data such as gender, age, anatomical location, and histopathological diagnosis were obtained from clinical records and analyzed. The lesions were classified into benign and malignant tumors in accordance with the current WHO classification of the head and neck tumors (17). Microscopical slides of all cases were examined by two independent pathologists with more than 25 years of experience. Immunohistochemical and histochemical analyses were performed when routine staining (hematoxylin-eosin) was not sufficient to establish the final diagnosis.
-Analysis Descriptive and quantitative data analysis was performed using the Statistical Package for the Social Sciences for Windows 20.0 (SPSS, Inc., Chicago, IL, USA). Continuous variables were expressed as mean, median and standard deviation values. Categorical variables were expressed as absolute number of cases and percentage values. Person's chi-square test and Fisher's exact test were used to evaluate association between biological behavior (malignant vs benign tumors) and clinical and demographic characteristics, adopting a pvalue of ≤0.05 and a 95% confidence interval.

Discussion
In the last two decades  several studies performed worldwide have been published on the epidemiology of salivary gland neoplasms, as shown in Fig. 3.
According to the WHO [2017], overall, female patients are slightly more affected by salivary gland neoplasms than male patients (17). However, some variations can be found when analyzing specific tumor subtypes (2,9,17,18). In the present study, the female-to-male ratio was 1.2:1, which is in agreement with most studies (19,20), including Brazilian reports (2,6,8,9).   In addition, benign neoplasms presented a male-to-female ratio of 1:1.4, while malignant neoplasms demonstrated a male-to-female ratio of 1:0.8, indicating that benign tumors were more common in female patients, whereas malignancies were slightly more common in males. These data are in accordance with several previous studies (2,9). However, a previous study performed in Mexico showed that female patients were more affected by malignant neoplasms than male patients (21).
In the present study, most tumors were benign (79.9%), data similar to other studies where these tumors correspond to about 51.5 to 86.4% of all salivary gland neoplasms (2,(5)(6)(7)(8)(9)13). However, some studies conducted on the African (22,23) and Asian continents (24,25) have shown a higher incidence of malignancies, and suggest geographic variation in the frequency of these tumors.
On the other hand, Silva et al.
(2) performed a retrospective multicentric study in Brazil and observed that basal cell adenoma and cystadenoma were the second and third most common neoplasms, respectively, different from most studies published in Brazil that shows the Whartin's Tumor as the second most common benign neoplasm (1,5,6,(8)(9)(10). These results suggest that multicenter studies can better characterize the heterogeneity of tumors in large territories, such as Brazil, for example, and contribute to the comprehension of epidemiological differences in the population (2). Also, the fact that this study was performed in oral pathology services may explain this apparent difference. The Warthin's Tumor is a neoplasm that affects almost exclusively the parotid gland, and some studies show that most cases from surgical pathology centers affect the major salivary glands, particularly the parotid gland, whereas tumors of the minor salivary glands represent the majority of cases diagnosed in oral pathology services (9). In addition, some other benign tumors, such as oncocytomas, were diagnosed in our study (n = 2, 0.3%). However, these tumors are very rare and are usually observed only in large sample studies (2). The most common malignant tumor was the mucoepidermoid carcinoma, accounting for 24.6% of the cases, followed by adenocarcinomas not otherwise specified (AcNOS), which represented 20.3%, and cystic adenoid carcinoma (17.8%), that corroborate with previous studies (12)(13)(14)(15)19,(24)(25)(26)29,30). On the other hand, other studies indicate cystic adenoid carcinoma as the most frequent malignant tumor (1,4,7,8,11,(21)(22)(23)28). In general, the four most frequent malignant tumors are mucoepidermoid carcinoma, cystic adenoid carcinoma, acinar cell carcinoma, and AcNOS. The morphological diagnosis of salivary gland tumors is challenging due to a large number of histological subtypes, overlapping of morphological findings, and different classifications (2,3). The diagnosis of polymorphous adenocarcinoma, in particular, can be difficult, especially in pathology centers without an experienced pathologist in oral and maxillofacial lesions, since this tumor shares some morphological characteristics with several other tumors (2). In our study, 13 cases of polymorphous adenocarcinomas were diagnosed, of which 8 affected minor intraoral salivary glands, 3 affected parotid glands, and 3 cases with unspecified anatomical sites. This strong predilection for polymorphic adenocarcinoma by minor salivary glands, especially in the palate region, is well established in the literature (17). Furthermore, until the last WHO classification, polymorphous adenocarcinoma was called "low-grade polymorphous adenocarcinoma", because in most cases, it exhibits indolent behavior. However, approximately 10% to 33% of patients develop local recurrences, 9% to 15% have nodal metastases, and some cases are extremely aggressive, with imprecise clinical behavior (2). Considering the variation in the biological behavior of these lesions, the new classification proposed by the WHO for salivary gland neoplasms abandoned the term "low grade" and renamed these tumors only as polymorphous adenocarcinoma (2,17). The purpose of this modification is to avoid possible terminological confusion and to facilitate the choice of treatment, especially for the most unusual cases (2,17). In addition, in our study, some other malignant tumors were very rare, such as salivary duct carcinomas (n=3, 0.5%) and myoepithelial carcinomas (n=1, 0.2%). Although these entities are well recognized, they are also rarely reported in studies with small samples (2). Regarding the anatomical location, most of the SGTs of this study were diagnosed in the parotid gland, followed by the minor salivary glands of the palate and submandibular gland. In general, this result was also reported by other studies (1,4,9,11,14,15,19,20,23,26). However, some studies have shown that malignant neoplasms preferentially affect the minor intraoral salivary glands (1,2,9). In summary, although several studies evaluate the frequency and incidence of salivary gland neoplasms, continuous studies that report the incidence and characteristics of these lesions are essential to keep physicians and surgeons up to date, especially when the classification of these tumors undergoes some change (2).