Biclonal Gammopathy Associated with Immunoglobulin A in a Dog with Multiple Myeloma

Case Presentation

A 10-year-old neutered male Airedale Terrier was presented to the Veterinary Medical Teaching Hospital (VMTH) with a 2–month history of generalized stiffness and lameness and a 5–month history of lethargy, inappetance, and mild weight loss. At the initial presentation to the referring veterinarian for inappetance 4 months prior to referral, results of a CBC and serum chemistry profile revealed mild anemia (PCV 35.7%, reference interval 37%-54%), hyperproteinemia (9.7 g/dL, reference interval 5.2–8.2 g/dL), and hyperglobulinemia (6.9 g/dL, reference interval 2.5–4.5 g/dL). The referring veterinarian prescribed a 10–day course of amoxicillin (15 mg/kg PO every 12 hours), and mild improvement was reported by the owner. The lameness was treated pallia-tively with aspirin (325 mg PO every 12 hours). Previous medical history was unremarkable with the exception of recurrent superficial pyoderma, which had been treated with multiple 3–week courses of cephalexin (30 mg/kg PO every 12 hours).

On admission to the VMTH, the dog was thin (body condition score of 3/9), with severe muscle wasting of the rear limbs. An elevated body temperature of 103.9°F was noted. Diffuse pain was noted on palpation and manipulation of all limbs. No joint effusion was palpable. Because of severe aggression and a tense abdomen, abdominal palpation was unrewarding. Neurologic examination was unremarkable with the exception of questionable conscious proprioceptive deficits in all 4 limbs. Fundic exam revealed tortuous retinal vessels and retinal hemorrhages, consistent with hyperviscosity syndrome due to hyperglobulinemia. The lameness was difficult to localize; differential diagnoses included joint pain (osteoarthritis, erosive or nonerosive immune-mediated or infectious polyarthritis, or neoplasia), bone pain (neoplastic or infectious lytic lesions or trauma), soft-tissue injury, and neuromuscular disease. Differential diagnoses for hyperglobulinemia in the absence of hyperalbuminemia included polyclonal gammopathy caused by chronic inflammation or infection, and monoclonal gammopathy caused by lymphoid tumors (multiple myeloma, extramedullary plasmacytoma, lymphoma, or lymphocytic leukemia) or certain chronic infections (ehrlichiosis, leishmaniasis, or dirofilariasis).

Results of a CBC revealed normocytic, normochromic anemia (PCV 27.9%) with mild thrombocytopenia (123×10³ platelets/mUL, reference interval 160–430×10³ platelets/μL). Anemia was nonregenerative (reticulocytes <3310/μL or <0.1%), and marked rouleaux was noted. Results of serum chemistry analysis revealed hyperproteinemia (10.5 g/dL, reference interval 5.5–7.2 g/dL) with hyperglobulinemia (8.4 g/dL, reference interval 2.1–3.6 g/dL) and hypoalbuminemia (2.1 g/dL, reference interval 3.1–4.2 g/dL). The serum calcium concentration was within the reference interval (11.2 mg/dL, reference interval 9.3–11.3 mg/dL). However, when corrected for the hypoalbuminemia, serum calcium concentration was increased (12.6 mg/dL). Urinalysis results were unremarkable with the exception of 2+ proteinuria by both dipstick and sulfosalicylic acid tests. Urine specific gravity was 1.030. Urine protein:creatinine ratio was increased (1.28, reference interval <1.00), suggestive of protein-losing nephropathy such as that caused by Bence-Jones proteinuria or glomerulonephritis. The urine tested negative for Bence-Jones proteins by qualitative thermal test (Antech Diagnostics, Farmingdale, NY, USA). The serum was negative for antibodies against Ehrlichia canis and Dirofilaria. Prothrombin time (PT), activated partial thromboplastin time (PTT) and fibrinogen concentration were within reference intervals. Fibrin degradation products (FDPs) were increased (5–20μg/mL, reference interval <5 μg/mL).

Thoracic radiographs revealed no evidence of metastatic lung lesions. Abdominal radiographs revealed an abdominal mass in the region of the spleen. In addition, several small lucent foci were identified in the mid- and proximal diaphysis of the femur and in the region of the distal physis and epiphysis. Scanning radiographs of the long bones and thoracolumbar spine revealed multiple focal lucencies in all long bones and in multiple spinous processes. Abdominal ultrasound confirmed the splenic mass accompanied by diffusely abnormal splenic parenchyma.

Cytologic evaluation of fine needle aspirates of the splenic mass revealed high numbers of atypical plasma cells. Moderate to marked anisocytosis, anisokaryosis, and frequent binucleation were noted in the plasma cell population. Because of the multiple lytic bone lesions in this patient, multiple myeloma was considered the likely diagnosis. Splenic plasmacytosis was the likely reason for the mass. Extramedullary plasmacytomas may cause hyperglobulinemia but are not associated with bony lysis. Serum protein electrophoresis (SPE) and bone marrow aspiration samples were assessed to confirm the suspicion of multiple myeloma. Protein electrophoresis showed a monoclonal peak in the (β₂-γ region accounting for 75.6% (7.78 g/dL) of the total serum protein (10.3 g/dL) (Figure 1A). Bone marrow aspirates were highly cellular with mild megakaryocytic hyperplasia. Also visible were confluent masses of immature round cells and rare blast forms with eccentric nuclei, abundant cytoplasm, and distinct Golgi zones suggestive of plasma cell neoplasia. Plasma cells accounted for 18% of cells in the bone marrow. A diagnosis of multiple myeloma with splenic involvement was made.

The dog was sent home with a 7.5–mg fentanyl patch for pain relief and begun on a 10–day induction protocol of melphalan (0.1 mg/kg PO every 24 hours) and prednisone (0.5 mg/kg PO every 24 hours). On day 17, the dog returned for reevaluation.The owner reported some improvement in lameness and general attitude. Results of a CBC revealed mild nonregenerative anemia (PCV 31.2%; reticulocytes 22,600/μL), moderate rouleaux, and resolution of the thrombocytopenia (398×10³ platelets/μL). Results of coagulation testing revealed significantly prolonged PTT (>25% increase over the upper reference limit), normal PT, normal FDPs (<5 μg/mL), and decreased fibrinogen concentration (φ100 mg/dL; reference interval 200–400 mg/dL). The coagulation abnormalities were consistent with inhibition of major coagulation factors by paraprotein rather than with disseminated intravascular coagulation. The SPE on day 17 (Figure 1B) showed β₂-γ bridging and a slight decrease in the size of the monoclonal peak, which accounted for 65.9% or 6.03 g/dL of the total serum protein concentration (9.0 g/dL). A maintenance protocol of melphalan (0.05 mg/kg PO every 24 hours) and prednisone (0.5 mg/kg PO every 48 hours) was initiated.

After 1 month of maintenance chemotherapy, the dog was reevaluated. The lameness and inappetance had resolved, and the dog had gained 1.7 kg. Results of a CBC revealed persistent normocytic, normochromic anemia (PCV 29.5%), mature neutrophilia (23×10³ cells/μL, reference interval 3–11×10³ cells/μL), and monocytosis (1400 cells/μL, reference interval 150–1350 cells/μL). These changes, in addition to a mild increase in serum alkaline phosphatase activity (142 IU/L, reference interval 20–126 IU/L), were consistent with the effects of corticosteroid therapy. The total serum protein concentration had decreased to 7.9 g/dL, with a serum globulin concentration of 4.7 g/dL. On SPE, a split β–pat-tern was observed (Figure 1C), rather than the monoclonal gammopathy noted previously. Abdominal ultrasound evaluation revealed significant enlargement of the splenic mass. Proteinuria (2+, specific gravity 1.011) was again seen on urinalysis; however, a urine pro-tein:creatinine ratio was not done because of an inflammatory sediment (3–6 WBC/high-power field and many bacilli). The urinary tract infection was treated using amoxicillin with clavulinic acid (15 mg/kg PO every 12 hours) for 3 weeks.

On day 115 after initial presentation, the dog returned with recurrent lameness, inappetance, and lethargy. The CBC results revealed persistent mild anemia (PCV 29.3%). Serum biochemical abnormalities included increased alkaline phosphatase activity (133 IU/L), hyperproteinemia (9.6 g/dL) with hyperglobulinemia (6.5 g/dL), and mild azotemia (BUN 28 mg/dL, reference interval 6–25 mg/dL) with normal creatinine concentration (0.7 mg/dL, reference interval 0.3–1.6 mg/dL). Urinalysis with bacterial culture and sensitivity demonstrated reinfection of the urinary tract, such that the proteinuria (2+ on dipstick, 4+ by sulfosalicylic acid test) was difficult to interpret. The SPE showed a biclonal gammopathy (Figure 1D), which represented 50.3% (4.83 g/dL) of the total serum protein concentration (9.6 g/dL). The chemotherapy protocol was changed to cyclophosphamide (50 mg/m² PO every 24 hours, 4 days/week) and prednisone (0.5 mg/kg PO every 48 hours). Cephalexin (30 mg/kg PO every 12 hours) for 3 weeks was prescribed again for the urinary tract infection. The dog improved clinically until 2 months later, when clinical signs returned. The dog was euthanized on day 175 postdiagnosis. Necropsy examination confirmed the diagnosis of multiple myeloma involving spleen and bone marrow, with diffuse myelophthisis and focal myelofibrosis, and glomerulonephritis.

Paraprotein Characterization

Prior to euthanasia, serum was collected and submitted for immunoquantitation, using immunoelectrophoresis (IEP) and immunofixation electrophoresis (IFE) techniques to determine the immunoglobulin (Ig) class involved. All antisera were prepared, absorbed and affinity purified, and both assays were performed at Bethyl Laboratories, Inc., Montgomery, Tex, USA. Briefly, goat anti-dog IgA and IgM and sheep anti-dog IgG antisera were made by injecting goats with purified canine IgA and IgM or sheep with purified canine IgG. The resulting whole serum antisera reacted with both Ig heavy and light chains. Anti-light chain antibodies specific for light chains (k and λ) were then prepared by affinity purification. Both anti-IgA heavy chain and anti-light chain antisera were used in this procedure. The goat anti-dog IgA was specific for epitopes on the heavy chains of IgA; the goat anti-dog IgA did not crossreact with canine IgG, IgM, IgE, or IgG (Fab fragments). The results of IEP revealed an IgA gammopathy in serum from the dog in this case by IEP using goat anti-dog IgA (Figure 2). A decreased amount of normal IgG was noted compared with normal dog serum controls. IgM was not detected using anti-IgG and anti-IgM antibodies.

Because the IgA migrated to the β-region of the gel, IFE was performed to further evaluate its migration pattern. IFE is a 2–stage process combining agarose gel electrophoresis with immunoprecipitation, and was carried out on commercially-obtained plates (Helena Labs, Beaumont, Tex, USA). Diluted (1:10 and 1:20) patient serum was applied to each track of the gel using an application mask (Helena Labs). The gel was then elec-trophoresed (Helena IFE system, Helena Labs) for 25 min at 125v in Helena's IFE buffer, pH 8.4–8.8, separating proteins according to their charge. Protein fixative and monospecific antisera to IgG, IgA, and IgM heavy chains and to light chains (k andλ) were applied to the gel. During incubation, insoluble immune complexes were formed. Insoluble immune complexes (antigen-antibody complexes) fix the antigen of interest in the gel prior to staining. The gel was then washed, pressed, dried, stained, and examined for paraprotein bands. Anti-IgA antiserum diluted 1:10 and 1:20 was used for detection. Results of IFE showed a major protein in the β₁ region and a second IgA band in the β₂ region, suggesting a biclonal gammopathy (Figure 3).The IgA clonal components appeared to be intact, and consisted of both heavy and light chains. The IgA concentration (934 mg/dL, reference interval 65–90 mg/dL) was determined by multiplying the fa percentage from the densitometric tracing by the total protein concentration (7.0 g/dL). IgG (420 mg/dL, reference interval 670–1650 mg/dL) and IgM (<50 mg/dL, reference interval 100–400 mg/dL) levels were below the reference intervals.

To rule out dimerization of IgA as a cause of the 2 IFE bands, SPE was performed after an experiment using addition of the reducing agent 2–mercaptoethanol (2–ME) to the sample. When 2–ME was mixed with the sample, however, a semigel formed. Consequently, the IgA paraprotein did not enter the agarose when applied to SPE plates. Other serum proteins, however, did enter the gel and were electrophoresed. Because SPE was unsuccessful, an additional Western blotting experiment was performed with 2–ME to determine the approximate molecular weights of the proteins using polyacry-lamide gel electrophoresis. Canine IgA, IgG, and IgM antisera specific for heavy chains were used. Western blots showed 3 bands greater than approximately 200 kd that were specific for IgA heavy chain (data not shown). These results indicated that dimerization was not a cause of the biclonality.The origin of the third band was unknown.

Discussion

Multiple myeloma is a neoplasm of plasma cells or plasma cell precursors. Diagnosis of multiple myeloma requires the demonstration of at least 2 of the following criteria¹,²: 1) monoclonal gammopathy, 2) lytic bone lesions, 3) atypical plasma cell proliferation within the bone marrow, and 4) Bence-Jones proteinuria. The diagnosis of multiple myeloma in the dog of this report was based on the presence of a clonal gammopathy on SPE, multiple lytic bone lesions, and atypical plasma cells within the marrow and spleen. Bence-Jones proteinuria was not identified in the dog in this case. Although proteinuria was observed, glomerulonephritis was confirmed at necropsy and was the most likely cause of the proteinuria. The qualitative thermal test for detection of Bence-Jones proteinuria has been described as unreliable in human beings^1,3 because of a high percentage of false-positive and false-negative results, which may explain the negative finding in this dog. Because of the unreliability of the thermal test, more sensitive methods are required for detection of Bence-Jones proteins. IEP or IFE techniques can be applied to concentrated urine samples using antisera against Ig heavy and light chains.³ These tests allow classification of the light chain and/or heavy chain component of the urine paraprotein. Although both IEP and IFE are useful in detecting Bence-Jones proteinuria, IFE is the preferred method because of faster turnaround time, increased sensitivity, and easier interpretation of results.³ These additional tests were not performed on urine from this dog.

Probable biclonal gammopathy was identified by SPE in the dog of this report. Biclonal gammopathy with different heavy chain classes has been reported previously in a dog¹ and a cat⁴ with multiple myeloma and hepatic plasmacytoma, respectively. In the dog with multiple myeloma, biclonal gammopathy consisted of 2 heavy chain classes (IgG and IgA).¹ In the cat with hepatic plasmacytoma, biclonality was noted on SPE, but the results of peroxidase-antiperoxidase tests for IgG, IgA, and IgM were inconclusive.⁴ Biclonal gammopathy of a single heavy chain class (IgG) has been reported in a dog with multiple myeloma and cutaneous lymphoma.⁵

The IgA gammopathy in the dog in this report was identified by IEP. A presumptive biclonal gammopathy involving intact Igs with a single heavy chain of the α class (IgA) was identified by IFE. Another possibility for the apparent biclonality was dimerization of IgA.⁶ Recently, a biclonal peak involving IgA dimers was identified in a cat with multiple myeloma.⁶ Gammopathy with 2 M-components, involving a dimer and trimer or tetramer of IgA, also has been reported in a dog with multiple myeloma.⁷ To reduce Ig dimers, 2–ME is frequently used, although there is no report in the literature describing reduction of IgA. When we added 2–ME to serum from this dog, a semigel formed that precluded SPE. The Western blot, however, showed 3 bands greater than approximately 200 kd that were specific for the IgA heavy chain. These results confirmed that the biclonal peaks were distinct monoclonal proteins rather than dimers of a single monoclonal protein. The origin of the third band was uncertain. It may have been a J-chain or an artifact.

Biclonal gammopathies may arise from a single clone of plasma cells or may occur as independent monoclonal gammopathies arising from 2 different plasma cell clones.⁸ In this dog, splenic involvement may have been the source of 1 of 2 independently-arising monoclonal gammopathies. IFE confirmed that the biclonal spike was of the IgA class. The anti-IgA we used specifically recognizes epitopes on the heavy chains of IgA and identified 2 bands that were still present when anti-dog light chain antiserum (k- andλ-nonspecific) was used. These results suggested that the biclonal spike observed on SPE was the result of a single heavy chain class (IgA) with intact heavy and light chains. Light chain classes probably differed between the 2 IgA clones. However, we did not identify the light chains as K orλ. We are unaware of antisera available from commercial or research labs that can distinguish between K and λ light chains.

A distinct biclonal gammopathy was not observed in early SPEs (Figure 1A through C), although probable biclonal gammopathy was noted later (Figure 1D). In an earlier retrospective study of human multiple myelomas, more than half of cases with single monoclonal peaks by SPE had biclonal peaks on IEP.⁹ Another explanation for the lack of distinct biclonality in early samples from this dog may be that globulin concentrations were higher in initial samples, resulting in decreased sensitivity of SPE. As chemotherapy continued, the clonal component may have decreased in concentration, increasing the sensitivity of SPE and unmasking the biclonal profile. Although there was a significant increase in IgA levels, there was a reduction in IgG and IgM types, possibly as a result of chemotherapy. In a typical monoclonal gammopathy, the clonal component is abnormally increased in concentration, resulting in a profile in which 1 heavy chain class is increased but the others are decreased.⁸

The median survival time in dogs with multiple myeloma treated with melphalan, cyclophosphamide, and prednisone is 540 days.¹⁰ Negative prognostic factors for multiple myeloma in the dog include extensive bony lesions, hypercalcemia, and light chain proteinuria.¹⁰ Renal insufficiency and poor initial response to therapy also may be associated with decreased survival times.¹⁰ Glomerulonephritis in the dog in this case may have resulted from impaired tubular catabolism of excess light chains, glomerular Bence-Jones protein deposits, amyloid deposition, or a combination of these factors. Although amyloid stains and electron microscopy can be used to further characterize glomerulonephritis, they were not done in this case. The dog was hypercalcemic and had multiple extensive lytic bone lesions, suggesting a poor prognosis. Initial clinical response to therapy was short-lived and was not accompanied by a decrease in size of the splenic tumor.

In a recent study, a possible correlation of prognosis with Ig isotype was suggested in cats with multiple myeloma.⁶ That correlation should be interpreted with caution, however, because of the small number of cases (n=5) in the study. Although IgA isotypes are less common in cats than IgG isotypes, myelomas with an IgA paraprotein were associated with visceral involvement and decreased survival time (ranging from a few days to 6 months) as compared with overall mean survival times for feline multiple myeloma of 2–9 months.^6,10

Hyperviscosity associated with hyperglobulinemia can result in cardiac insufficiency, retinal hemorrhage, renal disease, hemorrhagic diatheses, and neurologic signs, which can contribute to decreased survival time in animals with multiple myeloma.^11–13 In this dog, retinal vessel tortuosity and hemorrhage were consistent with hyperviscosity. The degree of hyperviscosity is related to the size of the paraprotein and the severity of hyperglobulinemia. Although hyperviscosity can occur in myelomas of any isotype, it is more frequently associated with IgA or IgM isotypes than with IgG gam mopathies,^12–14 consistent with the larger dimeric or multimeric size of these molecules.

To our knowledge, this is the first report of an undimerized biclonal gammopathy in a dog caused by a single heavy chain class involving IgA. Based on this report and other recent findings, biclonal gammopathy may be more common in veterinary patients than previously thought. Identification of the isotype in biclonal multiple myeloma is possible using IEP and IFE. Future studies are needed to determine the relationship, if any, between the results of these assays and prognosis.