Clinical and Pathological Data of 17 Non-Epithelial Pancreatic Tumors in Cats

Tumors of mesenchymal origin are rarely reported in the pancreas. Therefore, this study characterized 17 feline non-epithelial pancreatic tumors, including clinical data, histopathology, and immunohistochemistry. Seventeen feline pancreatic tissue samples were investigated histopathologically and immunohistochemically. Selected pancreatic and inflammatory serum parameters, e.g., feline pancreatic lipase immunoreactivity (fPLI), 1,2-o-dilauryl-rac-glycero-3-glutaric acid-(6′-methylresorufin) ester (DGGR) lipase and serum amyloid A (SAA), were recorded, when available. The neoplasms were characterized as round (n = 13) or spindle (n = 4) cell tumors. Round cell tumors included 12 lymphomas and one mast cell tumor in ectopic splenic tissue within the pancreas. Lymphomas were of T-cell (n = 9) or B-cell (n = 3) origin. These cats showed leukocytosis (3/3) and increased fPLI (5/5), DGGR lipase (3/5) and SAA (4/5) values. Spindle cell tumors included two hemangiosarcomas, one pleomorphic sarcoma and one fibrosarcoma. The cat with pleomorphic sarcoma showed increased SAA value. Overall survival time was two weeks to seven months. These are the first descriptions of a pancreatic pleomorphic sarcoma and a mast cell tumor in accessory spleens within feline pancreas. Although rare, pancreatic tumors should be considered in cats presenting with clinical signs and clinical pathology changes of pancreatitis. Only histopathology can certainly distinguish solitary pancreatitis from a neoplasm with inflammation.

Lymphomas are common neoplasms in most domestic species. Gastrointestinal lymphoma is the most common form of lymphoma in cats and is associated with clinical presentation such as weight loss, vomiting, diarrhea anorexia, and lethargy [9]. However, pancreatic involvement has been reported in only some cases and pancreas-specific blood parameters were not included in the studies [10][11][12][13]. While retroviral diseases (feline leukemia virus (FeLV) and feline immunodeficiency virus) are risk factors, a high number of cats with alimentary lymphomas have been reported to be virus negative [14].
As a mutant of FeLV, the feline sarcomavirus can cause sarcomas in cats [15], but literature about feline pancreatic sarcomas is scant. There is one case report of a feline primary pancreatic carcinosarcoma with metastases in uterus, omentum, and diaphragm [16]. Mesenchymal pancreatic tumors in humans are reported to have a poor prognosis with a median overall survival time of µg/L, indicative for pancreatitis: >5.4 µg/L) and feline trypsin-like immunoreactivity (fTLI, reference interval: 12-82 µg/L, indicative for pancreatitis >100 µg/L) values were determined by an ELISA assay at LABOKLIN GmbH & Co. KG.

Results
In pancreatic tissue from 17 cats, round (n = 13) and spindle (n = 4) cell tumors were diagnosed histopathologically ( Table 2). Round cell tumors included 12 lymphomas and one mast cell tumor. Spindle cell tumors were classified as two hemangiosarcomas, one pleomorphic sarcoma, and one fibrosarcoma. Clinical pathology was available for five cats with lymphoma and one cat with pleomorphic sarcoma (Table 3).    3.1. Round Cell Tumors, n = 13 Cats with lymphomas were one to 16 years old (median 11) and predominantly male (75.0%). The majority of cats were Domestic Shorthair (66.7%) ( Table 2). Clinical reports were available for 9 out of 12 cats: The cats showed vomiting (n = 3), lethargy (n = 3), inappetence (n = 3), diarrhea (n = 2), painful abdomen (n = 2), ascites (n = 2), palpable abdominal mass (n = 1), and seizures (n = 1). Clinical pathology results were available for five cats ( Table 3). The fPLI values were increased in all cats (5/5) and SAA values were high in four out of five cats. DGGR lipase values were increased in three out of five cases. Alpha-amylase and fTLI were within the reference range in all cats (5/5). A complete blood count (CBC) was available for three cats: leukocytosis was present in all animals, but lymphocytosis with severe leukocytosis and neutrophilia only in one cat (No. 7). One 10-year-old British Shorthair (No. 3) showed lymphopenia.
The macroscopic appearance of pancreatic tissues varied from normal to diffusely enlarged with fine granulated texture ( Figure 1A). Masses varied in size from 0.
Lymphoma was diagnosed in most of the cases (11/12) with the standard staining (HE). However, case No. 12 showed undifferentiated malignant neoplasia and immunohistochemistry was necessary for the diagnosis. Immunohistochemically, lymphomas were classified as T-cell (n = 9, Figure 1B) or B-cell (n = 3) lymphoma. In two cats with T-cell lymphoma (Nos. 5, 11), FeLV was detected.
The remaining ten cats were tested negative.
Seven cats were euthanized during surgery. A one-year-old Domestic Shorthair (No. 1) had a survival time of two months and a 13-year-old Siamese (No. 4) survived seven months. Treatment data for these cases were not available.
Vet. Sci. 2020, 7, x 6 of 13 Figure 1b) or B-cell (n = 3) lymphoma. In two cats with T-cell lymphoma (Nos. 5, 11), FeLV was detected. The remaining ten cats were tested negative. Seven cats were euthanized during surgery. A one-year-old Domestic Shorthair (No. 1) had a survival time of two months and a 13-year-old Siamese (No. 4) survived seven months. Treatment data for these cases were not available.

Mast Cell Tumor, n = 1
A 16-year-old female British Longhair cat (No. 13) was presented to the clinician for routine vaccination, and during palpation a mass in the abdomen was identified. Ultrasonography showed marked splenomegaly. During laparotomy, a diffusely enlarged (22.0 × 7.5 × 4.0 cm) spleen was removed. Two dark brown masses (0.3 × 0.3 × 0.3 and 1.0 × 0.6 × 0.5 cm) were found within the pancreas. Microscopically they were identified as accessory spleens. Both, spleen and accessory spleens showed marked diffuse infiltration by well-differentiated large mast cells (Figure 2a). Nuclear atypia and anisokaryosis were mild. The mitotic count was 1/HPF. No vessel infiltration could be observed, but some mast cells infiltrated the mildly hyperplastic pancreatic tissue. Giemsa staining showed intense metachromatic granularity of mast cells (Figure 2b). However, there was no c-kit expression detectable. FeLV testing was negative. No further treatment was performed. The cat was euthanized 2 months after diagnosis due to progressive renal and cardiac insufficiency.  (Figure 2A). Nuclear atypia and anisokaryosis were mild. The mitotic count was 1/HPF. No vessel infiltration could be observed, but some mast cells infiltrated the mildly hyperplastic pancreatic tissue. Giemsa staining showed intense metachromatic granularity of mast cells ( Figure 2B). However, there was no c-kit expression detectable. FeLV testing was negative. No further treatment was performed. The cat was euthanized 2 months after diagnosis due to progressive renal and cardiac insufficiency.

Spindle Cell Tumors, n = 4
In two cats (Nos. 14, 15), pancreatic masses were hemangiosarcomas ( Table 2). One cat was a 13year-old Domestic Shorthair (No. 14) with a palpable abdominal mass. Clinical findings in Cat No. 15 were not reported. Blood samples were not investigated in these cases. In both cats, histopathology

Spindle Cell Tumors, n = 4
In two cats (Nos. 14, 15), pancreatic masses were hemangiosarcomas (Table 2). One cat was a 13-year-old Domestic Shorthair (No. 14) with a palpable abdominal mass. Clinical findings in Cat No. 15 were not reported. Blood samples were not investigated in these cases. In both cats, histopathology showed infiltrative vasoformative growth of neoplastic endothelial cells, marked hemorrhage, thrombosis, mixed inflammation, and necrosis in the pancreas ( Figure 3A). Mitotic count was 1/HPF and 3/HPF, respectively. The neoplastic cells co-expressed vimentin, vWF and CD31 ( Figure 3B). They were tested negative for FeLV. Both cats were euthanized during surgery.
Vet. Sci. 2020, 7, x 8 of 13  One 17-year-old Domestic Shorthair (No. 16) was lethargic and had a palpable abdominal mass. Pancreas-specific results were within reference intervals, but the SAA value was increased and CBC revealed mild leukocytosis with neutrophilia and lymphopenia (Table 3). Multiple non-encapsulated firm masses (0.1 × 0.1 × 0.1 to 2.0 × 2.0 × 2.0 cm) within the pancreas were noted. Further neoplastic masses were not identified by clinical investigation of the skin or during laparotomy. Microscopically, the multinodular pleomorphic sarcoma with myofibroblastic differentiation was characterized by poorly differentiated spindle cells as well as pleomorphic and multinucleated giant cells within the pancreatic interstitium ( Figure 3C). Mitotic count was 9/HPF. Vascular infiltration was present. There were multifocal mild mixed inflammation and small areas of necrosis. The pleomorphic sarcoma showed a diffuse expression of vimentin and inhomogeneous co-expression of actin ( Figure 3D). Other sarcomas such as rhabdomyosarcoma, gastrointestinal stromal tumor, or mast cell tumor were excluded by negative immunohistochemical reaction with anti-desmin and c-kit antibodies and Giemsa stain negative. Furthermore, FeLV immunohistochemistry was negative. The cat was euthanized during surgery.
One 13-year-old Domestic Shorthair (No. 17) was presented with poor appetite and ascites. No other intraabdominal masses or cutaneous sarcomas were noted by the clinician. Macroscopically, the submitted tissue samples were brown and firm. Histopathologically, a fibrosarcoma with interstitial infiltrative growth of moderately differentiated neoplastic spindle cells with heterogeneous morphology ( Figure 3E) was diagnosed. Mitotic count was 2 mitoses/HPF. Neoplastic cells exclusively expressed vimentin ( Figure 3F). FeLV was negative. The cat was euthanized two weeks after surgery.
In general, lymphoma is the most common hematopoietic feline tumor [31]. A clinical diagnosis of gastrointestinal lymphoma can be challenging because of its unspecific clinical presentation, especially if neither masses nor lymphocytosis are present. During the clinical work up of such uncertain cases, clinical pathology is commonly performed. In our study, cats with lymphoma for which blood values were available (n = 5) showed high (n = 3) or questionable (n = 2) fPLI values, as well as increased DGGR lipase (n = 3) and SAA values (n = 4). This may be explained by an accompanying purulent pancreatitis in four cases. Further studies about pancreas-specific laboratory parameters in cats with lymphomas are necessary to investigate the prevalence of pancreatic involvement. Furthermore, inflammation was also reported in numerous feline epithelial pancreatic tumors [29] and as far as serological data were available, fPLI value was increased in several cases [32]. One may conclude that increased serological pancreatic enzymes can be indicative of an acute pancreatitis. However, varying underlying pancreatic neoplasms should be taken into consideration especially if empiric treatment is unsuccessful or if there is imaging evidence of neoplasia. Without histopathological investigation, pancreatic neoplasms may be misdiagnosed as an unsuccessfully treated pancreatitis. However, one cat with T-cell lymphoma (No. 7) showed increased fPLI and DGGR lipase values and normal SAA value without histological signs of pancreatitis. This is consistent with the literature, where false negative (sensitivity) and false positive (specificity) results have been described previously for serological pancreatitis diagnostic in some cases [33].
Morphological differentiation between inflammatory and neoplastic small lymphocytes in the pancreas may require further investigation. Immunohistochemical assessment is an important diagnostic tool [34]. In contrast, polymerase chain reaction for antigen receptor gene rearrangement analysis (PARR) is well-known to have some limitations in cats [35]. In our study, T-cell lymphomas were more frequent (75%) than B-cell lymphomas. This correlates with the adjoining anatomy of small intestine and pancreas as interstitial T-cell lymphomas are more frequent than B-cell lymphomas in cats [36]. Thus, the pancreatic sites of lymphomas were interpreted as part of a multicentric systemic disease. Immunohistochemistry for FeLV was positive for two cats with T-cell lymphoma (No. 5,11). This is in accordance with the literature, where retroviral diseases (FeLV or feline immunodeficiency virus) were listed as a risk factor in lymphomas [14]. In conclusion, without cytological/histopathological examination and/or imaging techniques, feline pancreatic lymphoma is easily missed if solitary pancreatitis is suspected because of clinical pathology and non-specific clinical findings. This has a strong potential impact on therapy and prognosis, especially considering the treatability of lymphomas by chemotherapy. Survival time of lymphoma subtypes may vary. Sato et al. [37] showed a shorter survival time in alimentary (48 days) lymphomas than in cats with nasal (135 days) or mediastinal (143 days) lymphomas. Wolfesberger et al. [38] reported the highest survival time in intestinal T-cell lymphomas (1.7 years) compared to large B-cell lymphoma (4.5 months) or peripheral T-cell lymphoma (6.1 months). In our study, seven out of 12 cats with lymphomas were either euthanized or died during or shortly after surgery. Only two cats with lymphomas lived on for 2 months (No. 1, B-cell lymphoma) and 7 months (No. 4, T-cell lymphoma), respectively.
Feline mast cell tumors are predominantly cutaneous neoplasms but can also appear in the spleen [24]. In our study, a mast cell tumor in accessory splenic tissue was described for the first time. Accessory splenic tissue is usually an incidental finding during laparotomy and appears as small dark brown masses within the pancreas [39]. Ectopic splenic tissue was seen in 24 out of 540 cases (4%) of routinely submitted feline pancreas samples in our laboratory (unpublished data). As in our case (cat No. 13), a high percentage of mast cell tumors in cats are negative for different immunohistochemical markers, including c-kit [40], which can impede a diagnosis in poorly differentiated cases. Prognostic factors are not entirely clear in cats, but splenectomy has been recommended for feline splenic mast cell tumors [41,42]. Careful exploration of the pancreas is recommended because of the risk of incomplete resection of splenic tissue in the case of affected accessory spleens, as in our case.
Sarcomas of the feline pancreas are very rare. Because of the retrospective nature of our study, there were incomplete data on signalment, therapy, and clinical outcome. Previous clinical and surgical reports were often unavailable. Therefore, a distinction between those sarcomas that were primary pancreatic neoplasms and metastases from elsewhere cannot be made with certainty in all cases. Nevertheless, because of the tumors being exclusively reported within the pancreas, both macroscopically and histologically, there is a high likelihood of them being primaries.
Feline visceral hemangiosarcomas were mostly seen in the liver, intestine, and lymph nodes but rarely (2/26 cats) in the pancreas [20]. Both cats with hemangiosarcomas included in our study died or were euthanized shortly after surgery. According to the literature, prognosis for feline visceral hemangiosarcoma with a multifocal presentation is poor, with a median survival time of 77 days [20]. Correlation between the degree of differentiation, mitotic rate or size of feline hemangiosarcoma and the clinical outcome could not be found [43].
The present study included a fibrosarcoma and a pleomorphic sarcoma in the feline pancreas. Nomenclature of pleomorphic sarcomas is still controversial. The terms "undifferentiated pleomorphic sarcoma," "anaplastic sarcoma with giant cells," or "malignant fibrous histiocytoma" are used synonymously in veterinary pathology [15]. In cat No. 16, immunohistochemistry revealed small areas of myofibroblastic differentiation within the neoplasm. In literature, this combination of histological type of sarcoma and immunohistochemical expression pattern has been described for feline cutaneous post-vaccinal sarcomas [44,45]. Cutaneous sarcomas may undergo malignant transformation of fibrocytes due to an inflammation [46]. Furthermore, there is one case report of a fibrosarcoma arising at the site of a retained surgical sponge in a cat [47]. In human medicine, inflammatory pseudotumors have been reported as a benign change in various organs, predominantly in the lungs [48]. The term "inflammatory fibrosarcoma" was suggested for some of these cases with metastases [49]. The cat with the pleomorphic sarcoma in our study had increased SAA values of an unknown cause and the pathogenesis of the tumor remained unclear. In human medicine, recurrence of metastatic sarcomas in the pancreas is common and a median overall survival time of 21 months [17] as well as moderate complication rates [18] have been described. The cats in our study were euthanized during surgery because of the poor prognosis.

Conclusions
In summary, our study described the characteristics of five different pancreatic tumor types of round or spindle cell origin in 17 cats. It includes the first descriptions of a pancreatic pleomorphic sarcoma and a mast cell tumor in accessory spleens within the pancreas. Epithelial pancreatic neoplasms have been reported previously [29] and metastases of other carcinomas may affect the pancreas also (unpublished data). In general, pancreatic neoplasms may be accompanied by inflammation, and histopathological investigation is required to differentiate pancreatitis from an inflamed pancreatic neoplasm. This may have a strong impact on therapy and prognosis.