Intraoral soft tissue lipomas: clinicopathological features from 91 cases diagnosed in a single Oral Pathology service

Background intraoral soft tissue lipomas are relatively uncommon mesenchymal neoplasms. Few papers have been published comparing the clinicopathological features of these tumors in different populations. The aim of the present study was to analyze the clinicopathological features from intraoral soft tissue lipomas diagnosed in a Brazilian population. Material and Methods all cases diagnosed as intraoral soft tissue lipomas in an Oral Pathology laboratory from 2005 to 2019 were retrieved and descriptively analyzed; statistical analysis was performed for comparison of the clinical and demographic parameters. Results 91 intraoral lipomas were retrieved, including 56 lipomas, 30 fibrolipomas, 2 spindle cell lipomas, 2 angiolipomas, and 1 chondrolipoma. Mean age of the patients was 62.2 years and females represented 57.1% of the sample. Mean time of complaint was 45.4 months and mean size of the lesions was 16.2 millimeters. Buccal mucosa (38.8%), lower lip (18.8%) and tongue (16.5%) were the most commonly affected locations. Fibrolipomas were more common in females (p=0,037) and presented as smaller lesions (p=0,011) in comparison to lipomas. Conclusions report of clinicopathological data from intraoral lipomas aid in establishing their differential diagnostic criteria and clinical profile in this specific location. Key words:Lipoma, fibrolipoma, oral, mouth, spindle cell lipoma, angiolipoma, chondrolipoma.


Material and Methods
The files of the Oral Pathology laboratory, School of Dentistry, State University of Rio de Janeiro, Rio de Janeiro/RJ, Brazil, were reviewed from 2005 and 2019 and all cases diagnosed as lipomas and their subtypes were retrieved. Cases diagnosed through fine-needle aspiration biopsy, cases not located in the oral cavity (eg. submandibular area) and cases presenting no histological slides and/or paraffin blocks for review were excluded from the sample. Demographic and clinicopathological information were retrieved from the patient´s registries and included: gender; age; location and clinical features of the lesions; time of onset; and final diagnosis. All hematoxylin and eosin (HE)-stained histological slides were carefully reviewed and lipoma subtypes were classified according to reference literature on the field (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14). When applicable, immunoreactions were performed with antibodies directed to mast cell tryptase (clone AA1, Dako Cytomation, Glostrup, Denmark, dilution 1:10.000) and CD34 (Biocare Medical, Concord, US, dilution 1:500) for spindle cell lipomas. Data were descriptively analyzed for all cases and each of the vari-ants and statistical analysis (Pearson chi-square and t test) was performed for comparison of the clinical characteristics from lipomas and fibrolipomas, with statistical significance considered when p<0,05.
The results of the present study reinforce these findings. Although these general characteristics can be applied to most intraoral lipomas, several lipoma subtypes have been reported in the oral cavity with variable frequencies and, due to their rarity, few information is presently available ( Table 2) (1-14). Apart from conventional lipomas (the most common subtype, representing 36 to 83% of the cases), other histological variants including fibrolipomas (1 to 40%), intramuscular lipomas (2 to 17%), spindle cell lipoma (1 to 44%), sialolipoma (2 to 10%), myxoid lipoma (up to 5%), angiolipoma (1 to 4%), osteolipoma (1 to 2%), chondrolipoma (1 to 2%) and atypical lipomatous tumor (up to 9%) have been reported in the literature (1)(2)(3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14). Cytogenetic analysis has shown several different chromosomal alterations in some lipomatous tumors (10), but, in clinical practice, their diagnosis remain based on conventional HE features, eventually associated to specific immunoreactions. For this reason and, as shown by the variable frequency of the histological subtypes reported in the published series, it is essential to follow very strict diagnostic criteria when classifying these tumors, allowing reliable comparisons on their frequencies, morphological patterns, clinical features and biological behavior.  Conventional lipomas are characterized by a proliferation of mature adipocytes interspersed by variable amounts of connective tissue containing bundles of collagen fibers and blood vessels (4). Despite these tumors rarely present a fibrous capsule, the proliferation of adipocytes shows a well-defined limit to the adjacent normal tissue. The term fibrolipoma is used when, apart from the typical adipocyte proliferation seen in a conventional lipoma, dense fibrocollagenous connective tissue bands are interlacing the mature adipocytes (2).
There are no strict criteria to define whether the amount and distribution of fibrous tissue is sufficient to define the lesion as a fibrolipoma, and this seems to explain the variable lipoma:fibrolipoma ratios reported in the literature (1,2,4). Most studies have shown that conventional lipomas are more common than fibrolipomas, except for one study where fibrolipomas (38%) outnumbered conventional lipomas (36%) (9), probably due to the histological criteria applied more than by a true geographic and/or populational variation. Apart from that, it is essential to rule out other similar lesions, such as herniated adipose tissue and fibrous hyperplasias with fat entrapment, when diagnosing these two histological subtypes (3). One interesting feature is that fibrolipomas seem to be more common in females (5,11,12,14,16), in accordance with the present results. Myxoid lipoma has been considered a subtype of lipoma characterized by the presence of a myxoid background associated to and partially replacing the proliferative adipose tissue (2,8). As conventional lipomas and spindle cell lipomas can also present myxoid areas, it is essential to rule out these two histological subtypes before rendering a diagnosis of myxoid lipoma. None of the intraoral lipomas reported in the present study presented an exclusive myxoid background and, for this reason, none was classified as myxoid lipoma. More comparative studies are necessary to define if myxoid lipoma is a true histological subtype of lipomas or if these lesions are characterized by myxoid degenerative connective tissue changes. Supporting this controversial issue, Furlong et al. (4) have reported that 11 of their cases previously diagnosed as myxolipoma were classified as spindle cell lipoma with prominent myxoid change after revision. Intramuscular (or infiltrative) lipomas are slow-growing lesions usually found in the extremities, thigh and trunk of adults (4). This variant is characterized by the presence of mature adipocytes showing no atypia infiltrating adjacent muscle layers and, especially in the extremities, these lesions can show infiltrative growth and higher recurrence rates. Intraoral intramuscular lipomas have been reported, but in this specific anatomical location, they usually show a biological behavior and a proliferative activity similar to conventional lipomas (1,3,7). Other subtypes of intraoral lipomas can show this close relationship to muscle fibers, suggesting that this feature should be interpreted mostly as a consequence of anatomical proximity, without changing the typical benign and indolent behavior of intraoral lipomas (5,7). To reinforce this anatomical relationship, most intraoral intramuscular lipomas have been reported in the tongue, where muscle fibers are in close superficial contact with the adipose tissue (1,3,4,10). For this reason no intramuscular lipomas were diagnosed in the present study. However, even with these evidences, it is worthwhile to consider that any infiltrating oral adipocytic tumor should be carefully analyzed to rule out the possibility of oral liposarcoma. In the latter, the presence of pleomorphic cells with nuclear hyperchromatism and vacuolization, as well as lipoblasts and higher mitotic activity can aid diagnosis, although they can be only focally present (1,10). Chondroid lipomas are characterized by the presence of mature adipose tissue, lipoblast-like cells and a chondroid matrix, and should be differentiated from liposarcomas and chondrosarcomas (10). This entity should be differentiated from chondrolipoma, as in the latter there is mature cartilaginous metaplastic tissue, probably secreted by pluripotential mesenchymal cells (10). This feature is similar to what occurs in osteolipomas, where metaplastic bone is formed (10). One chondrolipoma was found in the present sample, affecting the tongue of an 82-year-old male. Sialolipomas are characterized by the presence of adipocytic mature cells associated to atrophic salivary acini and dilated salivary ducts (3), but no case was found in the present series. Angiolipoma usually affects the subcutaneous tissue of the upper limbs and trunk of young adult males, being frequently multiple and tender to painful at palpation (10). They show a proliferation of mature adipose tissue interspersed by small thin-walled blood vessels containing fibrin thrombi in their lumen. Few cases have been reported in the oral cavity and most affected the buccal mucosa and mucobuccal fold as painless solitary submucosal masses (8,10,15). It is essential to call attention that the presence of a higher amount of blood vessels associated to mature adipose tissue is not sufficient for diagnosis of angiolipoma, as the presence of fibrin thrombi is essential for diagnosis (15). The present series revealed two angiolipomas affecting the pterigomandibular raphe and mucobuccal fold of adults. Both cases were reported in detail elsewhere, with a discussion on their main clinicopathological features and differential diagnosis (15). Spindle cell lipoma commonly arises as a painless slowgrowing mass in the subcutaneous tissue from the neck and trunk of adult males. The first description of spindle cell lipoma was rendered by Enzinger and Harvey in 1975 (17), but McDaniel et al. in 1984 (18) were the first to report an intraoral spindle cell lipoma. This entity usually affects adults without gender predilection, being characterized by a submerse painless mass covered by normal mucosa that affects mostly the tongue, buccal mucosa and floor of the mouth (2,(19)(20)(21). Although it has been considered that lipomas and fibrolipomas are the most common intraoral lipoma subtypes, Furlong et al. (4) reported that spindle cell lipoma represented 44% of their intraoral lipomas, affecting mostly the lips, buccal mucosa and tongue. This frequency probably do not reflect a true geographical or populational predilection but, as a reference center, a bias due to the referral of more complex cases to second opinion/revision. Curiously, none of the cases reported in the present study affected the buccal mucosa, also supporting that spindle cell lipomas present a slightly different site distribution in comparison to conventional intraoral lipomas and fibrolipomas.
Oral spindle cell lipoma is composed by variable amounts of mature adipocytes, CD34-positive spindle cells and mast cells in a background of myxoid and collagenous connective tissue (20,21). Spindle cells are characterized by a single elongated nucleus, bipolar cytoplasmic process and occasional cytoplasmic vacuoles (2,19). As previously discussed, it is possible that some spindle cell lipomas with prominent myxoid stroma were classified as myxoid lipomas in the past, and, also, more collagenous lesions could have been misinterpreted as fibrolipomas (4,7). These features call attention to the importance of CD34 immunostaining in the spindle cells (21). Although some spindle cell lipomas can show focal atrophic changes in adipocytes and focal atypia (including hyperchromatic nuclei and floret-like multinucleated cells), lipoblasts and widespread atypical features are not a feature and this is useful in differentiating these entities from oral atypical lipomatous tumors and well-differentiated liposarcomas (9). The reported features of focal atypia found in spindle cell lipomas were mostly associated to another lipoma subtype, called pleomorphic lipoma, which is presently considered as closely associated to spindle cell lipoma, being part of the same spectrum. The spindle cells are, as a rule, positive to CD34 and vimentin, and negative for most other markers, such as S100 protein, alpha smooth muscle actin, desmin and factor VIII (2,7,(19)(20)(21). It is also important to highlight that focal chondroid changes can be found in spindle cell lipomas and should be taken in account when diagnosing these entities (21). Fregnani et al. (3) have shown that intraoral spindle cell lipoma shows a higher proliferative rate in comparison to conventional intraoral lipomas, but the biological behavior seems to be similar and no recurrences are expected after conservative surgical approach (2,3,19,20). Although oral liposarcoma is an uncommon neoplasm, it is important to rule out this possibility when diagnosing oral adipocytic tumors (9,21,22,23). The term atypi-cal lipomatous tumor has been used instead of welldifferentiated liposarcoma for oral tumors presenting few lipoblasts and discrete infiltrative pattern, as in this location most lesions have an excellent prognosis, in contrast to well-differentiated liposarcomas elsewhere. Some atypical lipomatous tumors can show focal areas of stromal myxoid degeneration and, due to the discrete presence of lipoblasts and cellular atypia in most cases, they can be misinterpreted as myxoid lipomas. However, as this specific subtype has been a matter of controversy, it is important to consider other microscopic differential diagnosis, such as conventional lipomas with myxoid background or spindle cell lipoma, when dealing with these specific cases. Additionally, lipoblasts can be sparse in many atypical lipomatous tumors, reinforcing their resemblance to most conventional lipoma subtypes (22). For further confusion, oral liposarcomas have a very similar clinical profile than oral lipomas, mostly affecting adults and being characterized as long-lasting submerse nodules located especially in the tongue, buccal mucosa and floor of mouth (9,(21)(22)(23).
There were no cases diagnosed as atypical lipomatous tumor or liposarcoma in the present sample. It seems that, considering the previously exposed data, there are several clinical and microscopic features that differ intraoral lipoma subtypes from their counterparts in other anatomical locations. For this reason, clinicopathological series including intraoral lipomas will offer the possibility of establishing these differential criteria and, consequently, determine the most appropriate diagnostic and management steps for these tumors when affecting this specific anatomical location.