Abstract
Lingual biopsies are a common type of sample submission at the Athens Veterinary Diagnostic Laboratory (AVDL). Here we describe the pathology diagnoses of 793 canine and 406 feline lingual biopsies submitted to the AVDL in a 10-y period. Non-neoplastic lesions accounted for 450 diagnoses (57%) in dogs and 239 diagnoses (59%) in cats. Canine non-neoplastic lesions consisted of inflammatory lesions (286 cases; 64% of non-neoplastic lesions) and tumor-like proliferative lesions (164 cases; 36% of non-neoplastic lesions). Feline non-neoplastic lesions consisted of inflammatory lesions (228 cases; 95% of non-neoplastic lesions) and tumor-like proliferative lesions (11 cases; 5% of non-neoplastic lesions). The most common canine neoplasms were melanocytic neoplasms (103 cases; 30% of neoplasms) and epithelial neoplasms (102 cases; 30% of neoplasms), followed by mesenchymal neoplasms (90 cases; 26% of neoplasms) and round cell neoplasms (48 cases; 14% of neoplasms). Approximately 43% of melanocytic neoplasms affected Chow Chows and Labrador Retrievers, and 20% of epithelial neoplasms affected Labrador Retrievers. In cats, most tumors were epithelial (158 cases; 94% of neoplasms), followed by mesenchymal (8 cases; 5% of neoplasms) and round cell neoplasms (1 case; 1% of neoplasms). Over 50% of neoplasms of cats affected domestic shorthair cats. Although the percentage of lingual biopsies that had a neoplastic diagnosis was roughly the same between species, the diversity of neoplasms was much greater in dogs than in cats.
Oral lesions comprise a large proportion of surgical biopsy submissions of dogs and cats at the Athens Veterinary Diagnostic Laboratory (AVDL; Athens, GA, USA) and other institutions, but the veterinary literature on the topic is relatively scarce and less comprehensive than that which is available for other organ systems. 12 Depending on their biologic nature (nodular vs. non-nodular, neoplastic vs. non-neoplastic), degree of infiltration into surrounding tissues, and location in the oral cavity, oral lesions can lead to clinical signs associated with discomfort and pain or can be discovered incidentally during routine clinical examination. 11 In addition, non-neoplastic oral lesions may appear clinically as neoplasms, and their differentiation and diagnosis typically requires surgical biopsy and histology.9,11,15
Oral lesions in dogs and cats most commonly involve the gingiva.2,11 Canine lingual lesions are reported less commonly, with ~50% consisting of neoplasms, primarily melanomas, papillomas, and squamous cell carcinomas.2,4 Reports of lingual lesions in cats are much sparser than in dogs, and most focus primarily on squamous cell carcinoma and their classic sublingual location.7,10 Here we describe the surgical pathology diagnoses of lingual tissue submissions from 793 dogs and 406 cats at the AVDL over a period of 10 y.
We searched the AVDL web-based archive system for canine and feline surgical biopsy submissions of tongue tissue between January 2010 and December 2020 using the keywords “tongue”, “lingual”, and “glossal”. Submission forms and biopsy reports from retrieved cases were reviewed for patient signalment, clinical signs, anatomic location of the lesion (rostral, caudal, dorsal, ventral, lateral, or sublingual aspect of the tongue), and pathology diagnosis. Archived H&E–stained slides were reviewed, and tissue changes were categorized as non-neoplastic or neoplastic. Non-neoplastic lesions were further classified as inflammatory or tumor-like proliferative lesions, according to published recommended guidelines. 12 Neoplastic lesions were classified as melanocytic, epithelial, mesenchymal, or round cell neoplasms. Cases with no definitive diagnosis were not included in our investigation.
We selected and included 793 canine and 406 feline cases in our study (Tables 1, 2; Suppl. Tables 1, 2). Non-neoplastic lesions accounted for 450 diagnoses (57%) in the canine dataset and 239 diagnoses (59%) in the feline dataset. Neoplastic lesions were diagnosed in 343 (43%) dogs and 167 (41%) cats.
Diagnosis and anatomic location of 793 cases of lingual lesions in dogs between 2010 and 2020.
C = caudal; D = dorsal; L = lateral; MF = multifocal; NS = not specified; R = rostral; S = sublingual; V = ventral.
Diagnosis and anatomic location of 406 cases of lingual lesions in cats between 2010 and 2020.
C = caudal; D = dorsal; L = lateral; MF = multifocal; NS = not specified; R = rostral; S = sublingual; V = ventral.
Non-neoplastic lesions
Canine inflammatory lesions
Canine inflammatory lesions (286 cases; 64% of non-neoplastic lesions) consisted of ulcerative and/or suppurative glossitis (223 cases; 78% of inflammatory lesions), hyperplastic glossitis of unknown cause (49 cases; 17% of inflammatory lesions), eosinophilic granuloma (10 cases; 3% of inflammatory lesions), and ranula (4 cases; 2% of inflammatory lesions). Ulcerative and/or suppurative glossitis (Fig. 1) consisted of variable degrees of epithelial ulceration and accumulation of cell debris and neutrophils with submucosal proliferation of fibrovascular tissue. Hyperplastic glossitis (Fig. 2) consisted of extensive areas of mucosal epithelial hyperplasia with variable degrees of lymphoplasmacytic submucosal inflammation. Depending on the location of the lesion, ulcerative/suppurative or hyperplastic glossitis were often attributed to chronic friction with the adjacent gum or physical trauma caused by a foreign body or alterations in dentition. Eosinophilic granuloma (Fig. 3) occurred as dense clusters of epithelioid macrophages and eosinophils surrounding areas of collagenolysis that were in turn surrounded by eosinophils, lymphocytes, and plasma cells. Ranulas were distended salivary ducts lined by epithelial cells (Fig. 4) surrounded by mixed inflammatory infiltrates, fibrovascular tissue, and eosinophilic salivary secretion (Fig. 5).
Non-neoplastic lingual lesions in dogs and cats.
Canine tumor-like proliferative lesions
Tumor-like proliferative lesions (164 cases; 36% of non-neoplastic lesions) consisted of lingual polyps (134 cases; 82% of tumor-like proliferative lesions), calcinosis circumscripta (27 cases; 16% of tumor-like proliferative lesions), and histiocytic foam cell nodules (3 cases; 2% of tumor-like proliferative lesions). Lingual polyps were covered by hyperplastic mucosa supported by a stalk of submucosa (Fig. 6) containing dilated blood vessels and lymphoplasmacytic inflammation (Fig. 7). Calcinosis circumscripta describes well-demarcated areas of submucosal mineralization surrounded by epithelioid macrophages and multinucleate giant cells with fibrosis (Fig. 8). Histiocytic foam cell nodules consisted of nodular submucosal clusters of foamy macrophages (Fig. 9). Periodic acid–Schiff reaction with and without diastase did not highlight any positive material in the cytoplasm of these cells.
No overall breed predisposition was observed for non-neoplastic lesions of dogs, except for eosinophilic granulomas, which affected mainly Siberian Huskies (3 cases; 30% of eosinophilic granulomas) and Labrador Retrievers (2 cases; 20% of eosinophilic granulomas). No sex predisposition was evident for canine lesions. No lesions caused by a bacterial, fungal, or parasitic infection were diagnosed.
Feline inflammatory lesions
Feline inflammatory lesions (228 cases; 95% of non-neoplastic lesions) consisted of eosinophilic granuloma (136 cases; 60% of inflammatory lesions), plasmacytic glossitis (48 cases; 21% of inflammatory lesions), ulcerative and/or suppurative glossitis of unknown cause (39 cases; 17% of inflammatory lesions), and hyperplastic glossitis of unknown cause (5 cases; 2% of inflammatory lesions). The lesions in cases of feline eosinophilic granuloma were similar to those described in dogs (Fig. 10). Plasmacytic glossitis consisted of dense sheets of plasma cells with fewer lymphocytes and neutrophils within the submucosa (Figs. 11, 12). Lesions were attributed to feline chronic gingivostomatitis in 31 of 48 cases because of the widespread nature of the lesions in the examined biopsies. In the remaining cases, lesions were localized, and their cause was not determined. The lesions in cases of ulcerative and/or suppurative glossitis were morphologically similar to those described in dogs.
Feline tumor-like proliferative lesions
Tumor-like proliferative lesions (11 cases; 5% of non-neoplastic lesions) consisted of lingual polyps (11 cases; 100% of tumor-like proliferative lesions). The lingual polyps in cats were morphologically similar to those described in dogs.
More than 50% of non-neoplastic lesions of cats occurred in domestic shorthair cats (167 cases; 41% of affected cats), domestic longhair cats (31 cases; 7.5% of affected cats), and domestic medium hair cats (16 cases; 4% of affected cats). No sex predisposition was evident for feline lesions. No lesions caused by a viral, bacterial, fungal, or parasitic infection were diagnosed.
Neoplastic lesions
Canine neoplastic lesions
The most common canine neoplasms were melanocytic neoplasms (103 cases; 30% of neoplasms) and epithelial neoplasms (102 cases; 30% of neoplasms), followed by mesenchymal neoplasms (90 cases; 26% of neoplasms) and round cell neoplasms (48 cases; 14% of neoplasms).
Oral melanomas with high malignancy features (85 cases; 83% of melanocytic tumors) were poorly demarcated neoplasms with distinct degrees of invasiveness, pigmentation, and ulceration (Figs. 13, 14). Neoplastic cells were polygonal or elongate with increased nuclear atypia and more than 4 mitoses in 2.37 mm2 (10 FN22/40× fields). 1 Oral melanomas with low malignancy features (18 cases; 17% of melanocytic tumors) were typically heavily pigmented and had low cellular pleomorphism, low nuclear atypia, and none or <4 mitoses in 2.37 mm2 (Fig. 15). 1 Vascular invasion was observed in 24 of 85 (28%) melanomas with high malignancy potential. Because all melanocytic neoplasms had some degree of neoplastic cell pigmentation, immunohistochemistry for melan A and PNL2 was not necessary for the diagnosis and was performed in only one case (0.9% of melanoma cases) in our group.
Neoplastic lingual lesions in dogs and cats.
The most common epithelial tumors were papillomas (54 cases; 53% of epithelial tumors) and squamous cell carcinomas (47 cases; 46% of epithelial tumors). Papillomas consisted of exophytic proliferations of epithelial cells recapitulating the normal sequence of epidermal maturation (Fig. 16). Tumors were often ulcerated, covered with necrotic debris, and associated with a mixed inflammatory infiltrate (neutrophils, lymphocytes, and plasma cells) in the submucosa. Squamous cell carcinomas were invasive tumors; cords and trabeculae of neoplastic cells were often surrounded by a dense desmoplastic reaction (Figs. 17, 18). Neoplastic cells had moderate pleomorphism and abundant, eosinophilic cytoplasm with round-to-indented nuclei, finely stippled chromatin, and prominent nucleoli. Areas of dyskeratosis or keratin pearl formation were frequent. Neoplasms were often ulcerated and had extensive suppurative inflammation.
The most common canine mesenchymal tumors were granular cell tumors (32 cases; 35% of mesenchymal tumors) and spindle cell sarcomas (17 cases; 20% of mesenchymal tumors). Granular cell tumors were well-demarcated neoplasms composed of round-to-polygonal cells with eosinophilic, granular cytoplasm and round nuclei with dense chromatin (Figs. 19, 20). Cytoplasmic granules were positive for PAS and resistant to PAS-diastase reaction. Spindle cell sarcomas consisted of infiltrative tumors composed of bundles of elongate neoplastic cells with eosinophilic cytoplasm and round-to-elongate nuclei with dense to finely stippled chromatin (Fig. 21). Given the lack of morphologic features that would allow their classification as a specific neoplasm (e.g., fibrosarcoma, nerve sheath tumor), these neoplasms were routinely classified as spindle cell sarcomas.
Most of the canine round cell tumors were plasma cell tumors (38 cases; 79% of round cell tumors) and mast cell tumors (8 cases; 17% of round cell tumors). Plasma cell tumors consisted of sheets and packets of neoplastic cells supported by fine fibrovascular stroma (Fig. 22). Neoplastic cells had round, eosinophilic, glassy or vacuolated cytoplasm with the typical clear perinuclear area, and round or folded nuclei with dense chromatin. Binucleate and multinucleate neoplastic cells were common. In mast cell tumors, sheets of well-differentiated neoplastic mast cells contained numerous cytoplasmic granules. Nuclei were round and had finely stippled chromatin.
The most common canine breeds affected by melanocytic neoplasms included Chow Chows (23 cases; 22% of melanocytic neoplasms), Labrador Retrievers (22 cases; 21% of melanocytic neoplasms), mixed-breed dogs (19 cases; 18% of melanocytic neoplasms), and Golden Retrievers (12 cases; 12% of melanocytic neoplasms). Epithelial neoplasms occurred mainly in Labrador Retrievers (20 cases; 20% of epithelial neoplasms) and mixed-breed dogs (14 cases; 14% of epithelial neoplasms). Mesenchymal neoplasms most often affected Labrador Retrievers (11 cases; 12% of mesenchymal neoplasms) and Chihuahuas (9 cases; 10% of mesenchymal neoplasms). The breeds most affected by round cell neoplasms included Labrador Retrievers (8 cases; 17% of round cell neoplasms), Cocker Spaniels (5 cases; 10% of round cell neoplasms), and mixed-breed dogs (5 cases; 10% of round cell neoplasms). No sex predisposition was evident for canine neoplasms.
Feline neoplastic lesions
In cats, most tumors were epithelial (158 cases; 94% of neoplasms), followed by mesenchymal (8 cases; 5% of neoplasms) and round cell neoplasms (1 case; 1% of neoplasms). The overall morphologic features of feline lingual squamous cell carcinomas (148 cases; 94% of epithelial tumors) were like those described for their canine counterpart. Carcinomas (6 cases; 4% of epithelial tumors) and adenocarcinomas (4 cases; 2% of epithelial tumors) consisted of groups of neoplastic epithelial cells with polygonal cytoplasm that formed solid areas or tubules, respectively, and were separated by a fine fibrovascular stroma or desmoplasia. Nuclei were round and had finely stippled chromatin and prominent nucleoli. Mesenchymal tumors (8 cases; 5% of neoplasms) included 3 hemangiosarcomas, 3 fibrosarcomas (38% of mesenchymal tumors each), 1 fibroma, and 1 granular cell tumor (12% of mesenchymal tumors each). The only lingual round cell tumor in cats was a mast cell tumor.
Most neoplastic lesions of cats occurred in domestic shorthair cats (167 cases; 41% of affected cats), domestic longhair cats (31 cases; 7.5% of affected cats), and domestic medium hair cats (16 cases; 4% of affected cats). No sex predisposition was evident for feline neoplasms.
The 793 canine and 406 feline lingual biopsy submissions accounted for 1% and 4% of the total canine and feline biopsy submissions, respectively, to our laboratory during the studied period. Non-neoplastic lesions accounted for 57% of the canine diagnoses and 59% of the feline diagnoses; neoplastic lesions were diagnosed in 43% of dog and 41% of cat cases. Although the frequency of non-neoplastic and neoplastic lingual lesions in dogs and cats varies slightly in the veterinary literature, their overall distribution is similar among different investigations.2,4 None of the non-neoplastic lesions of dogs and cats in our series was attributed to an infectious cause. Although it is possible that some of the lingual papillomas could have been associated with papillomaviral infection, no ancillary testing was performed for diagnostic confirmation at the time of diagnosis.
The non-neoplastic lesion diagnoses in our cases comprised a heterogeneous group of oral lesions in which a definitive cause could not be determined based on histology. However, the most likely inciting factors for most of these lesions include chronic physical trauma to the mucosa with damage to the epithelium (hyperplastic and ulcerative and/or suppurative glossitis, lingual polyps), submucosa (calcinosis circumscripta), or to the salivary gland ducts (ranula); chronic antigenic stimulation (lymphoplasmacytic glossitis); and hypersensitivity (eosinophilic granuloma). 12 In our dogs, 52% of ulcerative and/or suppurative glossitis, 45% of hyperplastic glossitis, and 69% of lingual polyps affected the sublingual area; in cats, 61% of ulcerative and/or suppurative glossitis and 67% of lingual polyps occurred in the sublingual area. These findings may support the hypothesis that chronic friction of the sublingual mucosa with the adjacent gum or teeth could lead to sequential lesions that begin as reactive mucosal hyperplasia and evolve into polyps that eventually ulcerate.
Common tumor-like proliferative lesions diagnosed in dogs included calcinosis circumscripta and histiocytic foam cell nodules. Calcinosis circumscripta is one of the most common non-neoplastic lingual lesions of dogs.2,4 In our case series, it affected younger dogs (x̄ age = 3 y) compared to the other non-neoplastic lesions. This age predilection has been reported by other authors, 4 together with a higher risk for large canine breeds, which was not confirmed in our cases. Histiocytic foam cell nodules are reactive lesions commonly reported in Miniature Dachshund dogs but can also occur in other breeds. 8 The main differential diagnosis for these lesions includes granular cell tumors of the tongue, which, in contrast to histiocytic foam cell nodules, have PAS-positive cytoplasmic granules. 13
The most common non-neoplastic lesion of cats was eosinophilic granuloma, an inflammatory disorder of cats that is presumably associated with a hypersensitivity reaction to unknown antigens. 12 In our investigation, lesions occurred on the caudal or sublingual aspect of the tongue in 48% of cases, similar to what has been reported elsewhere. 6
Approximately 50–60% of reported lingual lesions in dogs are neoplastic, with 45–75% being malignant.2–4 The most frequently reported neoplasms include melanomas, squamous cell carcinomas, papillomas, and plasma cell tumors.3,4,14
As supported by our results, oral melanocytic neoplasms are commonly diagnosed in large dog breeds, particularly Chow Chows. 4 Although no sex or breed predilection has been confirmed for lingual SCC in dogs, 11 a study using a large number of cases found that these tumors were significantly more common in female dogs, Poodles, Labrador Retrievers, and Samoyeds. 4 There is no apparent sex or breed predisposition in cats.7,11 No such sex and breed predispositions were evident for lingual SCCs in our cases.
Melanomas reportedly have a predilection for the dorsal aspect of the body and root of the tongue. 14 In our cases, 41% of canine melanomas occurred on the dorsal or caudal (root) aspect of the tongue, confirming this anatomic predilection. Similarly, 28% of canine SCCs and 51% of feline SCCs affected the sublingual aspect of the tongue, also in concordance with previous findings. In addition, 31% of canine granular cell tumors occurred in the ventral aspect of the tongue, and 34% of canine plasmacytomas occurred in the caudal and dorsal aspects of the tongue. Canine granular cell tumors have a predilection for the tongue, but not enough cases have been reported to determine whether there may be a predisposition for a specific portion of the tongue.11,13,14 Similarly, no anatomic predisposition has been confirmed for lingual plasmacytomas.5,11,14
Because rostral lingual neoplasms are thought to have a better chance of early detection and complete surgical removal, and because the rostral portion of the tongue has a lower vascular and lymphatic vessel density, a rostral tumor location has been associated with a better prognosis. 3 However, this assertion has not been supported clinically. 14 Given that clinical follow-up data, including post-diagnostic survival and metastatic rate, were not available in our investigation, we cannot support or refute this potential link.
Supplemental Material
sj-pdf-1-vdi-10.1177_10406387241278888 – Supplemental material for A retrospective study of lingual lesions in 793 dogs and 406 cats at the Athens Veterinary Diagnostic Laboratory, 2010–2020
Supplemental material, sj-pdf-1-vdi-10.1177_10406387241278888 for A retrospective study of lingual lesions in 793 dogs and 406 cats at the Athens Veterinary Diagnostic Laboratory, 2010–2020 by Jesse Riker and Daniel R. Rissi in Journal of Veterinary Diagnostic Investigation
Footnotes
Declaration of conflicting interests
The authors declare no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors declared that they received no financial support for their research and/or authorship of this article.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
