9-year-old spayed female DSH feline, Felis cats. Three-month history of progressively worsening left hind limb lameness. On external examination, a large, firm mass was noted on the left proximal tibia. Radiographs of the limb revealed a focally extensive area of osteolysis with periosteal elevation along the proximal tibia. The left rear leg was amputated mid-femur and the entire limb was submitted for histopathological analysis.

Gross Description:  

Submitted for histopathology was the entire left hind limb that had been amputated at the level of the middle femur. Dissection revealed a pronounced thickening of the proximal tibia with irregular and lytic areas of periosteum and cortical bone. 

Histopathologic Description:

Examined is a section of bone and surrounding soft tissue (tendon, muscle) where the bone is markedly expanded and focally replaced by a poorly demarcated, non-encapsulated, densely cellular mass. Neoplastic cells fill greater than 50% of the marrow spaces, surrounding and replacing trabeculae, multifocally replacing the cortex and extending into the periosteum. Neoplastic cells are arranged in sheets and streams, supported by a fine fibrovascular stroma. Cells are pleomorphic; most cells are polygonal to stellate with poorly defined cell borders, moderate eosinophilic fibrillar cytoplasm, eccentric ovoid nuclei, finely stippled chromatin, and a single amphophilic prominent nucleolus. There is moderate anisocytosis and anisokaryosis There are numerous scattered binucleate and multinucleate giant cells, sometimes containing >20 nuclei, often adjacent to osteoid matrix. Mitotic figures are rare with 1 or fewer per 10 hpf. Neoplastic cells appear to produce a dense fibrillar to homogenous eosinophilic matrix (osteoid). There are extensive multifocal to coalescing regions of cartilaginous differentiation. Scattered throughout and adjacent to the neoplasm there is bone lysis, necrotic bone, and mild multifocal to coalescing areas of hemorrhage. The cortical bone is discontinuous, interrupted by clusters of neoplastic cells surrounded by abundant fibrous connective tissue (scirrhous reaction). Multifocally, there is periosteal proliferation of reactive bone (exostosis) and few perivascular infiltrates of lymphocytes, plasma cells, and fewer macrophages that extend into the adjacent soft tissue. 

Morphologic Diagnosis:  

Bone, proximal tibia: Osteosarcoma



Contributor Comment:  

Primary bone tumors are an uncommon finding in the feline patient with reported incidence of 3.1 to 3.9 per 100,000 cases.(5) Of the primary bone tumors, osteosarcomas (OS) account for 70 to 80 percent of findings.(3) Differing from canine OS, which exhibits a biphasic impact distribution, feline OS tends to impact middle age to older cats (average age 8-10 years).(3) Tumor locations can be divided into axial, appendicular, and extraskeletal. Extraskeletal sites have been noted to occur sporadically in multiple tissues and anatomical locations with a propensity for occurring in locations commonly associated with vaccine administration. The most common skeletal locations for feline OS include distal femur and proximal humerus and tibia and overall hind limbs are more commonly impacted. Further differing from canine OS, feline OS exhibits a low rate of (5-10%) of pulmonary metastasis.(5) Radiographic features of osteosarcoma in cats are variable with the aggressive periosteal proliferation often noted in canine OS being less prevalent.(4)

Cats with appendicular or extra skeletal forms of OS tend to survive longer when compared to those with axial forms. In a study published by Dimopoulou, survival prognosis for cats with osteosarcoma was related to histologic grade and mitotic index. 

Spugnini, reported similar findings with median survival of 49 months for cats following amputation alone with appendicular OS, compared with a median survival time of only 5.5 months for cats with axial skeletal OS. If the tumor permits removal, surgery alone may be curative with extended survival time for those undergoing advanced adjunctive therapies. The histologic grade in this case was relatively low (rare mitotic figures, low to moderate tumor cell density, abundant tumor matrix, minimal to mild necrosis, moderate pleomorphism) and is suggestive of a fair to guarded prognosis. Although amputation is often curative, neoplastic cells were noted rarely in blood vessels within the section in this case. Interestingly, tumor invasion into vessels was not found to be a significant prognosticator in one retrospective study of feline osteosarcoma.(3) No long-term follow up is available for this patient. 

This tumor was made up of predominantly osteoblast-like neoplastic cells with numerous clusters of multinucleate giant cells scattered amongst a prominent matrix of osteoid, mature bone, and cartilage. Scattered multinucleate giant cells are not uncommon in feline osteosarcoma, though a giant cell variant osteosarcoma such as this one, with numerous giant cells, is unusual. The origin of multinucleated giant cells in OS is poorly understood and both osteoclast and osteoblast origins have been postulated. In a report by Negrin(8), both osteoclast and osteoblast immune staining features were noted in a feline OS of the calvarium. Osteoclast like features included TRAP, vimentin, and S-100 positive staining with cytokeratin-negative staining. An osteoblast like feature includes MHC IInegative reaction.(8) No prognostic significance has been related to the giant cell variant osteosarcoma seen in cats.

JPC Diagnosis:  

Bone, tibia: Osteosarcoma.

Conference Comment:  

Osteosarcomas (OSA) are commonly described as malignant long bone tumors of large breed dogs, with the distal radius, distal tibia, and proximal humerus being the most common sites of occurrence(1) and are much more common than their feline counterpart. The differences between OSA of dogs and cats have been highlighted by the contributor. Conference participants discussed the particulars of obtaining a definitive diagnosis of OSA. In many cases, diagnosis is often complicated by a small sample submission and the fact that these neoplasms are often heterogenous and admixed with reactive bone which may result from a proliferative response due to nonneoplastic mechanisms. Bone reacts to local and systemic stimuli through systematic and regimented processes, regardless of etiologic stimulus (e.g. fracture, benign or malignant neoplasm or infectious disease). This case nicely illustrates this point, as in some areas, woven bone is overtly neoplastic and characterized by large, atypical, irregularly clustered osteoblasts haphazardly oriented to disorganized bone trabeculae, or embedded within lacy deposits of osteoid. The disorganized areas of bone are intermixed with reactive or reparative woven bone characterized by plump osteoblasts coalescing around, and oriented perpendicular to the longitudinal axis of woven bone trabeculae, thus demonstrating a structural uniformity distinguisible from neoplastic bone islands. However, in small samples without clinical or radiographic context, distinguishing osteosarcoma from reactive bone can prove problematic. As stated in the WHO tumor fascicles "practice and experience cannot be supplanted," in the case of interpreting proliferative bony and reactive periosteal lesions.(9)

The variety of histologic presentations of OSA has led to the description of nine separate subclassifications in a recent paper.(7) Subclassifying these tumors is also hindered by their heterogeneic nature as often several histologic subtypes are evident within a single neoplasm; however, their subclassification, with the exception of the aggressive telangiectatic variant, does not appear prognostically significant.(7) Histologic grade, on the other hand, is often cited as offering considerable prognostic value, though only mitotic index was identified specifically in cats.(3) Evidence has accumulated in the literature supporting the hypothesis that cyclooxgenase-2 (COX-2) is involved in the pathogenesis of osteosarcoma, and recently prostaglandin E2 was pinpointed as the downstream culprit of interest, lending credence to the use of COX-2 inhibitors in chemotherapy regimens and suggesting selective inhibition of PGE2 may be equally effective.(6)


1. Carlson CS, Weisbrode SE. Bones, joints, tendons and ligaments. Zachary JF, McGavin MD, eds. Pathologic Basis of Veterinary Disease. 5th ed. St. Louis, MO: Elsevier Mosby; 2012:959.

2. Culp WT, Olea-Popelka F, Sefton J, et. al. Evaluation of outcome and prognostic factors for dogs living greater than one year after diagnosis of osteosarcoma: 90 cases (1997-2008). J Am Vet Med Assoc. 2014;245(10):1141-1146.

3. Dimopoulou M, Kirpensteijn J, Moens H, Kik M. Histologic prognosticators in feline osteosarcoma: a comparison with phenotypically similar canine osteosarcoma. Vet Surg. 2008 Jul;37(5):466-71.

4. Frankel CS, Young TL, Alvarez-Berger F, Spencer CP. What Is Your Diagnosis? Journal of the American Veterinary Medical Association. August 1, 2013, Vol. 243, No. 3, Pages 329-331

5. Heldmann E, Anderson MA, Wagner-Mann C. Feline Osteosarcoma: 145 Cases (19901995). J Am Anim Hosp Assoc 2000;36:51821. 

6. Millanta F, Asproni P, Cancedda S, Vignoli M, Bacci B, Poli A. Immunohistochemical expression of COX-2, mPGES and EP2 receptor in normal and reactive canine bone and in canine osteosarcoma. J Comp Pathol. 2012;147(2-3):153-160.

7. Nagamine E, Hirayama K, Matsuda K, et. al. Diversity of histologic patterns and expression of cytoskeletal proteins in canine skeletal osteosarcoma. Vet Pathol. 2015 Mar 13. pii: 0300985815574006. [Epub ahead of print]

8. Negrin A, Bernardini M, Diana A, and Castagnaro M. Giant Cell Osteosarcoma in the Calvarium of a Cat. Vet Pathol 2006 43: 179.

9. Slayter MV, Boosinger TR, Inskeep W, Pool RR, Dammrich K, Larsen S. Histologic Classification of Bone and Joint Tumors of Domestic Animals. 2nd series. Vol. I. Washington, D.C.: Armed Forces Institute of Pathology; 1994:9-11. 

Click the slide to view.

1-1. Tibia

1-2. Tibia

1-3. Tibia

Back | VP Home | Contact Us |