• Open Access

Lymphoma-Associated Polymyositis in Dogs


  • Dr Gruenenfelder is presently affiliated with Applied Neurobiology Group, Department of Cell Science, Institute for Comparative Medicine, Faculty of Veterinary Medicine, University of Glasgow, Glasgow, UK

Corresponding author: Dharshan Neravanda, BS, DVM, diplomate ACVIM (Neurology), Department of Small Animal Medicine and Surgery, College of Veterinary Medicine, University of Georgia, 501 DW Brooks Drive, Athens, GA 30602; e-mail: dneravanda@gmail.com


creatine kinase


generalized inflammatory myopathy




primary skeletal muscle lymphoma

A 3-year-old castrated male Pit Bull Terrier was referred with a 2-month history of progressive tetraparesis that began with an acute onset of exercise intolerance, weight loss, and muscle atrophy. The dog's gait was short strided and accompanied by marked exercise intolerance. Generalized muscle atrophy was present, with the temporalis and masseter muscles most severely affected. The patellar and withdrawal reflexes were decreased. Neurologic signs were consistent with a diffuse neuromuscular disorder. Enlarged sacral lymph nodes were palpated on rectal examination. Clinicopathologic abnormalities included thrombocytopenia (84,000 platelets/μL; reference range, 211,000–621,000 platelets/μL) and increased serum creatine kinase (CK) activity (1,878 U/mL; reference range, 63–350 U/mL). An enlarged sternal lymph node was identified on thoracic radiographs (Fig 1). Abdominal ultrasonography disclosed enlarged mesenteric and iliac lymph nodes with ill-defined regions of hypoechogenicity. Cytologic evaluation of an iliac lymph node aspirate indicated reactive lymphoid hyperplasia with reactive histiocytes and evidence of chronic hemorrhage. Positive antibody titers were present for Toxoplasma (IgM, 1 : 256; IgG, 1 : 32), Borrelia (IgG, 1 : 4,096), and Leptospira (icterohemorrhagica, 1 : 400; grippotyphosa, 1 : 200). Borrelia titers were interpreted to be vaccine induced, although confirmatory Western blot analysis was not performed. Toxoplasma gondii and Leptospira spp. titers were interpreted as evidence of exposure but active infection was not excluded. Electromyography (temporalis, masseter, epaxial muscle along the vertebral column, and appendicular musculature) and nerve conduction studies (tibial-sciatic nerve and ulnar nerve) performed under general anesthesia identified no abnormalities. Histopathologic examination of frozen sections of temporalis and triceps muscle biopsies utilizing H&E, modified Gomori trichrome, PAS, ATPases at pH 9.8 and 4.3, esterase, NADH-TR, acid phosphatase, alkaline phosphatase, oil red O, and staphylococcal protein A-horseradish peroxidase were performed.1 Excessive variability in myofiber size was observed with numerous atrophic fibers of both fiber types having a round shape. Multifocal areas of mixed mononuclear cell infiltrations were present with an endomysial, perimysial, and perivascular distribution (Fig 2). Neither mitotic figures nor atypical lymphocytes were identified. Immunophenotyping showed predominantly CD8+ T lymphocytes and γδ T-cell receptor (TCR) utilization (Fig 3).2,a B lymphocytes were not identified. The peroneal nerve was biopsied and evaluated in resin sections. With the exception of resolving subperineural edema, the nerve was normal. The differential diagnosis included infectious myositis, immune-mediated polymyositis (PM), and paraneoplastic inflammatory myopathy.

Figure 1.

 (A) Lateral thoracic radiograph of Case 1 at initial presentation demonstrating an enlarged sternal lymph node. (B) Lateral thoracic radiograph of the same dog 29 days later. There is resolution of the sternal lymph node enlargement and development of pulmonary nodules.

Figure 2.

 Frozen biopsy sections of the triceps (A) and temporalis (B) muscles of Case 1, taken at initial presentation. Excessive variability in myofiber size was observed with numerous atrophic fibers having a round shape and of both fiber types. There were multifocal areas of mixed mononuclear cell infiltration having an endomysial, perimysial, and perivascular distribution consistent with a generalized inflammatory myopathy of infectious, autoimmune, or paraneoplastic etiology. Neoplastic changes were not seen initially or on re-evaluation. H&E staining.

Figure 3.

 Immunophenotypic analysis of T-cell infiltrates in frozen biopsy sections from the triceps muscle of Case 1. CD4+ (A) and CD8+ (B) T cells were localized with canine specific monoclonal antibodies. CD8+ T cells predominated with only small numbers of CD4+ cells. Canine-specific monoclonal antibodies against αβT-cell receptor (C), and γδT-cell receptor (D), documented the predominance of γδT-cell receptor utilization.d An avidin-biotin immunoperoxidase method was performed as described previously.2

Antibiotic treatment with clindamycin (12 mg/kg PO q8h) and doxycycline (6 mg/kg PO q12h) was initiated. After 5 days without a response, prednisone therapy (0.5 mg/kg PO q12h) was initiated. After the addition of prednisone, the dog's strength, stamina, and muscle mass improved for 10 days at which time it developed intermittent episodes suggestive of vestibular dysfunction. The episodes consisted of an involuntary head turn to the left side followed by stumbling and falling to the right; no abnormal nystagmus was noted. Laboratory evaluation indicated a resolution of thrombocytopenia and decrease in CK activity (664 U/mL). Repeated thoracic radiographs revealed development of several (2–3 cm) ill-defined pulmonary nodules and resolution of the sternal lymphadenopathy (Fig 1). Repeated abdominal ultrasound examination indicated less severe abdominal lymphadenopathy. Analysis of bronchoalveolar lavage (BAL) fluid indicated severe mucopurulent inflammation with macrophages, neutrophils and a few eosinophils. Magnetic resonance imagingb of the brain demonstrated multifocal to coalescing, noncontrast-enhancing lesions in the brainstem. The lesions were hyperintense on T2-weighted images and hypointense on T1-weighted images, relative to gray matter. No changes were noted in the head musculature. Analysis of cerebrospinal fluid (CSF) obtained from the cisterna magna indicated a pleocytosis (6 cells/mL; reference range, 0–5 cells/mL) and normal protein concentration (11.4 mg/dL; reference range, <24 mg/dL). Cytologically, there was a mixed pleocytosis with an increased percentage of eosinophils (45% monocytes, 28% lymphocytes, 21% eosinophils, and 6% neutrophils), suggestive of fungal, protozoal, or neoplastic disease. The prednisone dosage was decreased (0.25 mg/kg PO q24h) and fluconazole treatment (5 mg/kg PO q12h) was initiated while awaiting results of protozoal and fungal testing (serum titers for Cryptococcus neoformans, Blastomyces dermatitidis, Histoplasma capsulatum, T. gondii, and Neospora caninum; pan-fungal PCR on CSF; PCR for C. neoformans on BAL fluid and CSF; fungal culture and Gram stain of BAL fluid; all results were negative). Three days later, the dog became anorexic and profoundly lethargic and was euthanized.

At postmortem examination, multicentric lymphoma was diagnosed involving the lung, brain, skeletal muscle, a cervical lymph node, and spleen. In multiple paraffin sections of skeletal muscle there was widespread separation and replacement of degenerating myofibers by a lymphocytic infiltrate. Lymphoid cells ranged from small to large with the larger cells appearing atypical and containing large hyperchromatic nuclei. The brain lesions were predominantly in gray matter of the pons and medulla. Perivascular cuffs contained moderate numbers of small and large atypical lymphoid cells admixed with eosinophils. Mitotic figures were present among the larger lymphoid cells. Similar abnormal lymphocytes were found in the neuropil of gray matter. The neoplastic lymphocytic infiltrate in lung, muscle, and brain specimens was positive for CD3c and negative for CD79α antigens,d characterizing the neoplastic infiltrate as T cell in origin (Fig 4). Re-evaluation of the cytology from the initial lymph node aspirate disclosed abnormal lymphocytes suggestive of lymphoma. Review of frozen muscle histopathology from the original biopsy did not identify mitotic figures or atypical lymphoid cells.

Figure 4.

 Lymphoma diagnosed in formalin-fixed paraffin embedded sections of muscle (A, B, C) and lung (D, E, F) obtained at necropsy from Case 1. (A) H&E staining of muscle demonstrating a monomorphic infiltrate of atypical lymphocytes consistent with lymphoma. (B) Atypical lymphocytes in muscle stain positive for CD3 demonstrating T-cell immunophenotype. (C) Negative CD79α staining for B cells in muscle. (E) H&E staining of lung tissue demonstrating atypical lymphocytes consistent with lymphoma. (F) Positive CD3 staining demonstrates atypical cells in lung are T lymphocytes. (C) Negative CD79α staining for B cells in lung.

A 7-year-old castrated male Labrador Retriever was referred with an 8-month history of progressive exercise intolerance, episodes of restricted neck movement, lethargy, weight loss, weakness, and abnormal gait. Before referral, mild improvement was noted after antibiotic therapy (details unknown) for 1 month. On referral examination, the dog walked with a stiff gait and guarded neck posture. With manipulation of the neck, the dog had full range of motion without discomfort. The remainder of the neurologic examination was within normal limits. The dog's gait and posture were consistent with a diffuse neuromuscular disorder. Physical examination including lymph node palpation identified no abnormalities. Laboratory evaluation was normal. Serum CK activity was not evaluated. Thoracic radiographs were suggestive of pulmonary osteomas but otherwise were normal. Abdominal radiographs and ultrasound examination disclosed hepatomegaly and splenomegaly and enlarged iliac lymph nodes. Histopathologic examination (evaluated using the same staining techniques described in Case 1) of biceps femoris and triceps muscle biopsies disclosed multifocal mixed mononuclear cell infiltration having an endomysial distribution with invasion of nonnecrotic fibers. The cellular infiltrates were composed of clusters of lymphocytes and scattered macrophages. Occasional necrotic fibers were present undergoing phagocytosis with several fibers showing central necrosis. Endomysial fibrosis and myofiber atrophy were present in areas of marked cellular infiltration. The findings were consistent with an inflammatory myopathy of infectious, immune-mediated, or paraneoplastic causes. In this case, immunophenotyping, infectious disease testing, electrodiagnostics, and liver and spleen aspirates were not performed. The dog was treated successfully with a tapering course of prednisone (beginning at 2 mg/kg PO q24h) over an 11-month period. Thirteen months after initiation of prednisone therapy, the dog was presented for exercise intolerance, lethargy, stiffness, weight loss, inappetence, polyuria, polydypsia, abdominal distention, and left pelvic limb swelling. Clinical signs had begun 2 months earlier, coinciding with the discontinuation of prednisone. Hematologic evaluation indicated a hematocrit of 29.4%, increased nucleated red blood cells (6,700 cells/mL; reference range 0 cells/mL) and a mature neutrophilic leukocytosis (49,500 cells/mL; reference range, 6,000–18,000 cells/mL). Microscopic evaluation of a peripheral blood smear identified vacuolated monocytes and a few atypical lymphocytes. Serum CK activity was increased (177 IU/L; reference range, 21–56 IU/L). The dog was euthanized because of clinical deterioration.

Complete necropsy was performed. Histologic examination of routine paraffin-embedded sections disclosed thick interstitial bands and dense aggregates of lymphocytes, histiocytes, and plasma cells in the appendicular, epaxial, and abdominal wall musculature and diaphragm. Small numbers of neutrophils were present in areas of active muscle degeneration. Some extremely dense areas of the infiltrate obliterated muscle fibers and were composed primarily of large atypical lympho-histiocytic cells and scattered mitotic figures. These cells had round to ovoid nuclei showing fine chromatin stippling, small to medium-sized nucleoli, thickened nuclear membranes, and minimal cytoplasm. Immunohistochemistry of the dense atypical cell infiltrate identified 60% of the population as T cells (CD3+), including a substantial number of the larger cells sometimes showing mitotic figures. Only small numbers of B cells (CD79α+) were identified in the infiltrate. Pathologic findings were interpreted as PM and lymphoma based on the presence of atypical cell morphology as well as possible malignant transformation of an existing inflammatory infiltrate. A necrotic Gram-positive bacterial infection in the paw of the left pelvic limb was likely responsible for the limb swelling.

Typical clinical signs associated with PM include weakness, exercise intolerance, and a characteristic short-strided, stilted gait in all 4 limbs. Less common signs include generalized muscle atrophy, regurgitation, pyrexia, myalgia, and neck ventroflexion.3–5 Increased serum CK and aspartate aminotransferase activities reflect muscle damage but do not correlate with the severity of clinical signs or with the degree of cellular infiltration.5 Histopathologic findings in PM are relatively consistent.4 Microscopic hallmarks include variability in myofiber size and endomysial, perimysial, and perivascular infiltration of predominantly CD8+ (cytotoxic) T lymphocytes into nonnecrotic muscle fibers with lesser numbers of macrophages and dendritic cells.2,6 T lymphocytes with TCR αβ predominate.2 Treatment for PM typically requires long-term corticosteroid-based immunosuppression and prognosis can be favorable except in cases with concurrent aspiration pneumonia or chronic extensive muscle fiber loss and fibrosis.3–5 Some dogs require long-term alternate day corticosteroid treatment or other forms of immunosuppression to prevent relapses.5

Infectious agents are associated with approximately one-third of canine generalized inflammatory myopathies (gIMS), although definitive etiologies are seldom identified.4 Canine gIMS have been associated with histopathologically confirmed N. caninum, Hepatozoon americanum, Hepatozoon canis, and Leishmania infantum.4,7–9 Although a causal relationship has not been confirmed, several infectious agents have been associated with inflammatory myopathies. An inflammatory myopathy also has been reported in 2 dogs with Ehrlichia canis morulae in circulating monocytes, 1 of which had a positive E. canis blood culture.10 In addition, positive serum antibody titers for T. gondii, N. caninum, Borrelia burgdorferi, Rickettsia rickettsii, E. canis, H. americanum, and Leptospira australis have been reported in canine inflammatory myopathies.4,11,12 It is unclear whether the latter organisms are direct causes of canine gIM or provide an antigenic stimulus for a secondary autoimmune myositis.10,11 Cases lacking microscopic, serologic, immunohistochemical, or PCR evidence of infection or neoplasia typically are considered to represent PM, implying an underlying immune-mediated pathogenesis.

There are 3 potential pathogeneses for the association between PM and lymphoma in these 2 cases. First, infiltrating T cells in PM may have undergone malignant transformation to lymphoma. Second, PM may be a paraneoplastic syndrome secondary to extramuscular, multicentric lymphoma. Third, generalized primary skeletal muscle lymphoma (PSML) may have been misdiagnosed as PM.

The malignant transformation of inflammatory cells may follow a chronic antigen-driven inflammatory process. In cats, a transition from inflammatory bowel disease to gastrointestinal lymphoma has been suspected.13,14 In people, some patients with Helicobacter pylori-associated chronic gastritis have developed gastric lymphoma.12 In Case 1, neither mitotic figures nor atypical cells were identified in the original muscle biopsies in frozen or paraffin-embedded sections. However, the predominance of T cells utilizing TCR γδ is unusual for canine PM and could be indicative of clonal expansion of a specific T-cell population (as in human PM, the majority of T cells in canine PM have the most common TCR, the αβ).2 Molecular studies would be necessary to confirm this hypothesis. In both cases, initial muscle biopsies were consistent with PM, whereas at the time of diagnosis of lymphoma muscle histopathology indicated a neoplastic lymphocytic infiltrate, implying malignant transformation as a likely cause. Alternatively, the temporal relationship between PM and lymphoma may be associated with a paraneoplastic process. Paraneoplastic disease occurs in tissues of the body distinct from the primary tumor or metastases and may have an autoimmune basis.15 The PM in these 2 dogs cannot be classified as a paraneoplastic process associated with lymphoma because paraneoplastic syndromes must occur in tissues distinct from the primary tumor. In these dogs, both the PM and the lymphoma affected skeletal muscles. A 3rd possible relationship between PM and lymphoma is the misdiagnosis of PSML as PM on muscle biopsy. Neoplastic lymphocytes may be overlooked on muscle biopsies.5 People with PSML and 1 previously reported dog with PSML manifested nonpainful muscle enlargements caused by infiltrating neoplastic lymphocytes.16,17 The 2 dogs reported here had no signs of focal muscle enlargement and therefore were unlikely to have had PSML. Furthermore, diffuse neuromuscular dysfunction is not a recognized manifestation of PSML in humans or dogs.

Detailed reports of 4 dogs with similar clinical courses are reported in the literature.18–21 Duration of clinical signs before the diagnosis of PM ranged from 1 to 2 months, which is similar to the duration of 2 and 8 months reported here.18–21 The diagnosis of lymphoma was made 2–7 months after the diagnosis and treatment of PM, which is similar to the 1 and 13 months reported here.18–21 Retrospective review of samples from the initial presentation of 1 previously reported dog was consistent with lymphoma that was overlooked originally.5 In 2 previously reported dogs, thoracic radiographs demonstrated a nodular pattern similar to that observed in Case 1.18,21 This radiographic pattern typically is not associated with lymphoma in dogs.22 Two of the previously reported dogs had intramuscular B-cell lymphoma (1 with concurrent cutaneous T-cell lymphoma) unlike the T-cell immunophenotypes reported here.19,20 Immunophenotyping was not performed in the 2 other reported cases.18,21 Additionally, in a study of 200 dogs with gIM, 12 dogs developed neoplasia within 12 months of the diagnosis of gIM.4 Eight of these dogs were Boxers, of which 6 were diagnosed with lymphoma, 1 with an anaplastic round cell tumor, and 1 with a plasmacytoma.4

A few recommendations can be made to assist in the recognition of lymphoma-associated PM in dogs. Persistently increased serum CK activity should increase suspicion for ongoing muscle cell damage, even if CK is only mildly increased. Ultimately, muscle biopsies are critical for diagnosis of PM. Infiltration of CD8+ T lymphocytes with few to no B lymphocytes is characteristic of canine PM.2,23 As observed here, this pattern also can be seen in intramuscular lymphoma. Based on the results of this study and review of previous cases, immunophenotyping of cellular infiltrates in muscle biopsies from dogs with PM should be performed. Both frozen and paraffin-embedded sections should be evaluated for mitotic figures and atypical cells.24 After the diagnosis of PM is confirmed, periodic physical examination and thoracic radiographs should be performed. The observation of lymphadenopathy or pulmonary nodules in a dog with a diagnosis of PM should increase the level of suspicion for the development or coexistence of a neoplastic disorder such as lymphoma.

It is critical to collect all biopsies before the initiation of glucocorticoid therapy. Prior glucocorticoid therapy may delay or mask the definitive diagnosis of lymphoma.25 Although not performed in these cases, PCR tests of appropriate samples for clonally rearranged immunoglobulin or TCR genes may help distinguish a neoplastic population of lymphocytes from a reactive population, and should be pursued along with immunophenotyping if an increased suspicion for lymphoma exists.13


aCD4, CD8, αβTCR, and γδTCR antibodies were from Dr Peter Moore, University of California, Davis, CA

bGE 1.0T Signa, GE Health, Milwaukee, WI

cT-cell marker, diluted 1 : 800, Dako, Carpinteria, CA

dB-cell marker, clone HM57, diluted 1 : 400, Dako


The authors thank Dr Melinda Camus for providing photomicrographs of histologic sections, and Dr Frederic Almy for critical review of the manuscript.