Polycythemia and Inappropriate Erythropoietin Concentrations in Two Dogs with Renal T-cell Lymphoma
Two dogs presented with suspected renal disease and polycythemia. Abdominal ultrasound examinations performed on both dogs revealed coalescing masses causing bilateral renomegaly. Serum erythropoietin (EPO) concentrations were physiologically inappropriate. Postmortem examinations revealed renal T-cell lymphoma in both dogs. One of the two dogs also had involvement of the liver and mesentery. EPO-immunohistochemistry on tissue samples demonstrated positive staining in tumor cells and occasional normal renal cells. This report illustrates that paraneoplastic EPO production may induce polycythemia. The pattern of EPO-immunohistochemistry staining suggested that the mechanism of production was due to tumor production of EPO and local hypoxia-induced EPO production from compression of normal renal cells and vasculature.
Introduction
Primary renal lymphoma is an uncommon form of lymphoma with a poor long-term prognosis.1 Anemia is often noted in patients with lymphoma; however, polycythemia has also been reported.2,3 Polycythemia is an uncommon disorder that is characterized as either primary or secondary. Primary polycythemia is due to the neoplastic proliferation of red blood cells (RBC) independent of erythropoietin (EPO) stimulation, whereas secondary polycythemia results from increased EPO concentrations. Secondary polycythemia due to inappropriately elevated EPO is a rare manifestation usually due to infiltrative renal neoplasia. Two pathophysiologic mechanisms have been proposed to cause this inappropriate increase in EPO production.2 These include: ectopic production of EPO or an EPO-like substance from neoplastic cells; and local hypoxia in the kidney due to compression of normal tissue or vasculature by an infiltrative neoplasm resulting in physiologically-induced EPO production.
This case report describes two dogs that presented with polycythemia and renal T-cell lymphoma. Physiologically inappropriate serum EPO concentrations were detected in both cases. EPO-immunohistochemical staining of renal tissue supported both the production of EPO from tumor cells and hypoxia-driven EPO production from compressed renal tissue.
Case Report
Case 1
A 3 yr old castrated male Cocker spaniel dog that weighed 9.3 kg was presented to the primary care veterinarian following a 5 day history of intermittent hematuria. The referring veterinarian prescribed a 14 day course of amoxicillin trihydrate/clavulanate potassiuma (13.9 mg/kg per os q 12 hr). Improvement was noted; however, hematuria recurred 1 day after completion of the antimicrobial and the dog returned to the referring veterinarian. Urinalysis revealed hematuria and dysplastic epithelial cells. A positive contrast cystogram was normal. The dog was referred to the Internal Medicine Department at the Mississauga-Oakville Veterinary Emergency Hospital and Referral Group (MOVEH) for further evaluation.
Physical examination on presentation was normal, including neurologic assessment, digital rectal examination, and fundic examination. Initial diagnostic tests included a complete blood cell count (CBC), serum biochemical profile, prothrombin time (PT), activated partial thromboplastin time (aPTT), and pre- and postprandial bile acid assays. Urinalysis was performed on a sample collected via cystocentesis. Blood gas analysis was not performed. The blood work revealed polycythemia. The HCT was 0.61 L/L (reference range, 0.39–0.56 L/L), RBC count was 9.1×1012/L (reference range, 5.8–8.5×1012/L), and hemoglobin (Hb) was 214 g/L (reference range, 133–197 g/L). A normal concurrent total plasma protein was measured (63 g/L; reference range, 55–75 g/L) as well as prolongation of both the PT (>100 sec; reference range, 5.5–9.8 sec) and aPTT (26.8 sec; reference range, 9.8–19.6 sec). The remainder of the CBC was within normal limits including RBC indices (i.e., mean cell volume was 67 fL [reference range, 62–72 fL], mean corpuscular hemoglobin (MCH) was 24 pg [reference range, 21–25 pg], and mean corpuscular hemoglobin concentration (MCHC) was 352 g/L [reference range, 330–360 g/L]). The red cell distribution width (RDW) was mildly increased (14.5%; reference range, 11–14%). Abnormalities in the biochemical profile included elevations in creatine kinase and amylase activities, which were 702 U/L (reference range, 40–255 U/L) and 981 U/L (reference range, 299–947 U/L), respectively. Serum phosphorus, urea, creatinine, and electrolytes were evaluated and were normal. The urinalysis confirmed hematuria (blood 3+ and 300–500 RBCs per 400× field) and showed isothenuric urine (specific gravity was 1.009). An abdominal ultrasound was performed, which revealed bilaterally enlarged kidneys with complete loss of normal architecture. The renal parenchyma had been entirely replaced with multiple, ill-defined, coalescing masses of mixed echogenicity. All other visualized organs, including the urinary bladder and lymph nodes, were within normal limits. Differential diagnoses included neoplastic, infectious, or congenital renal disease. Systolic (120 mm Hg), diastolic (85 mm Hg), and mean (108 mm Hg) blood pressures were normal. Serum EPO concentration was measuredb to further investigate the cause of the polycythemia. The EPO level was markedly increased (40.4 mU/mL; reference range, 8.4–28.0 mU/mL). The polycythemia was therefore suspected to be due to inappropriate EPO production from abnormal or hypoxic renal tissue. As there was no evidence of lower urinary tract disease, the cause of the hematuria was likely renal in origin, presumably from the renal masses. Renal biopsy, fecal floatation, repeat PT and aPTT, and urine culture were recommended. The owner declined, and the dog was discharged.
The dog presented to MOVEH 20 days following initial presentation for recurrent hematuria. Physical examination, including fundic examination, was within normal limits. Repeat abdominal ultrasound revealed progressive enlargement of the kidneys with masses of similar appearance. Further diagnostic testing was again declined. The owners were advised to assess the HCT q 2–4 wk, to determine serum urea and creatinine concentrations and blood pressure q 4 wk, to determine the urine protein:creatinine ratio on a sample without gross hematuria (if possible) in 1 mo, and to perform an abdominal ultrasound in 1 mo to assess progression of renal lesions. The owners were instructed to contact a veterinarian immediately if the dog deteriorated.
On day 55, the dog suddenly developed neurologic abnormalities. The dog was in lateral recumbency with paddling motion of all limbs and ptyalism; however, urination and defecation were not noted and the dog appeared conscious. The owners presented the dog immediately to the referring veterinarian. The PCV was 65% and the systolic blood pressure was elevated, ranging from 160–200 mm Hg on serial blood pressure measurements. Total protein was not assessed at that time. The dog was treated by the referring veterinarian with nitroglycerin ointmentc (a 1 inch strip placed on the inner right pinna) to reduce blood pressure. In addition, a phlebotomy was performed and 100 mL of blood was removed from the jugular vein. In addition, 200 mL of 0.9% NaCl crystalloid solution was infused via a peripheral intravenous (IV) catheter. For this procedure, the dog was sedated using diazepamd (0.5 mg/kg IV) and intramuscular hydromorphone hydrochloridee (0.1 mg/kg). The dog was then transferred to the MOVEH.
On presentation to MOVEH, the physical examination was within normal limits, including vital parameters, mucous membrane color and capillary refill time, neurologic and fundic examinations, and digital rectal examination. Serum biochemistry and CBC results revealed numerous abnormalities. Significant changes in the CBC included polycythemia (HCT was 0.61 L/L, RBC count was 8.9×1012/L, Hb was 190 g/L, and MCHC was 309 g/L) with a normal concurrent total protein (63 g/L) as well as a mature neutrophilia and lymphopenia (neutrophil count was 13.9×109/L; reference range, 2.9–10.6×109/L and lymphocyte count was 0.63×109/L; reference range, 0.8–5.1×109/L). The RDW was increased, and rubricytosis was present (RDW was 15.5% and the rubricyte count was 0.16×109/L). The remainder of the RBC indices was normal (i.e., MCV was 69 fL and MCH was 21 pg). Biochemical profile abnormalities consisted of mild increases in phosphorus and total bilirubin concentrations and alanine aminotranspherase, alkaline phosphatase, gamma glutamyl transferase, amylase, lipase, and creatine kinase activities as well as moderate panhypoproteinemia. Serum urea and creatinine concentrations were normal. Abdominal ultrasound revealed enlargement of the kidneys bilaterally due to progression of renal masses, bilateral pyelectasia, evidence of severe pancreatic edema based on patchy hypoechoic regions within the pancreas, and nodules in the liver, pancreas, and mesentery. In addition, there was a moderate amount of ascites. Interestingly, azotemia was not present even with severe ultrasonographic renal changes, suggesting that there remained a proportion of functional renal tissue. Based on the poor prognosis, acute onset of possible seizure activity, and abnormalities noted on abdominal ultrasonography, euthanasia was elected.
A postmortem examination was performed, which grossly revealed firm, pale, red to white colored masses throughout both renal cortices and medullae, as well as the liver, pancreas, and mesentery. Pancreatic edema was also noted. Bone marrow biopsy and removal of the brain were not performed. All other assessed organs, including lungs, heart, adrenal glands, urinary bladder, stomach, duodenum, jejunum, ileum, colon, and lymph nodes were grossly and histologically normal, providing no evidence of systemic hypoxia. Histopathology revealed an extensively infiltrative multifocal round cell tumor displacing normal renal tissue (Figure 1). Intravascular tumor cells were also noted. The tumor cells were round but had moderate anisocytosis and were interpreted as lymphoma. There were tumor cells with similar morphology and associated hemorrhage and fibrosis in the liver and mesentery. Immunohistochemistry for round cell markers revealed strong and moderate detection of CD3f and CD18g, respectively, and lack of detection of CD79αh, confirming a diagnosis of T-cell lymphoma.
Citation: Journal of the American Animal Hospital Association 47, 2; 10.5326/JAAHA-MS-5614
Case 2
A 9 yr old spayed female Border collie dog that weighed 19.4 kg was referred to MOVEH for evaluation of inappetence, weight loss, and renal disease. The dog had a chronic history of infrequent vomiting and diarrhea when stressed as well as urinary incontinence since ovariohysterectomy as an immature dog. The CBC, biochemical profile, and urinalysis performed during a wellness examination 3 mo prior to presentation to MOVEH revealed moderate polycythemia and dilute urine. The HCT was 0.61 L/L, RBC count was 8.9×1012/L, Hb was 224 g/L, and urine specific gravity was 1.015. A normal concurrent serum protein concentration (72 g/L) was also measured. The remainder of the RBC indices were normal (MCV was 70 fL, MCH was 24 pg, and MCHC was 359 g/L), as were the total and differential white blood cell counts and platelet count. Four weeks before presenting the dog to MOVEH, the owners had noted inappetence, intermittent diarrhea, weight loss, polyuria, and polydipsia. Diagnostic testing was performed by the referring veterinarian. The CBC and biochemical profile revealed severe polycythemia (HCT was 0.72 L/L, RBC count was 11.7×1012/L, and Hb was 259 g/L), a mildly increased serum protein concentration (78 g/L), azotemia (urea was 11.4 mmol/L; reference range, 3.0–10.0 mmol/L, creatinine was 231 μmol/L; reference range, 30–140 μmol/L), and an increased amylase (1,560 U/L; reference range, 150–1,350 U/L). The remainder of the CBC was normal, including total and differential white blood cell counts, platelet count, and RBC indices (MCV was 62 fL, MCH was 22 pg, and MCHC was 358 g/L). Urinalysis collected via cystocentesis demonstrated proteinuria (1+), a pH of 6.0, and isosthenuria (urine specific gravity was 1.011). Medications included diethylstilbestroli (0.05 mg/kg PO q 4 days long-term) and metronidazolej (12.5 mg/kg PO q 12 hr, prescribed for the previous 10 days based on a telephone consultation with the referring veterinarian). The inappetence, polyuria, and polydipsia progressed and the dog was referred to MOVEH for further testing.
Physical examination on presentation to MOVEH revealed injected mucous membranes, bilateral conjunctivitis, and mild abdominal pain. All other parameters were within normal limits, including neurologic, fundic, and digital rectal examinations. Initial diagnostic procedures included a CBC and an abdominal ultrasound. Biochemical profile and blood gas analysis were declined by the owner. The CBC revealed severe polycythemia (HCT was 0.72 L/L, RBC count was 11.5×1012/L, and Hb was 260 g/L) with a normal concurrent plasma protein concentration (72 g/L). The RDW was mildly increased (14.5%), and rubricytosis was present (rubricyte count was 0.09×109/L). The remainder of the CBC was within normal limits. The abdominal ultrasound revealed bilateral renomegaly caused by multiple ill-defined, coalescing masses of mixed echogenicity that obliterated all normal renal ultrasonographic architecture. Several splenic masses of mixed echogenicity measuring approximately 1.0 cm in diameter were also noted. The most likely differential diagnosis was renal neoplasia. Thoracic radiographs followed by a renal biopsy were recommended; however, the owner elected euthanasia. Measurement of serum EPO concentration was performedb due to the severe polycythemia and a value that was physiologically inappropriate in light of the polycythemia was obtained (22.2 mU/mL).
On postmortem examination, bilateral renomegaly with multiple pale nodules throughout the renal cortices and medullae were identified. In addition, hemorrhagic masses were noted diffusely throughout the spleen. Histopathologic examination revealed a bilateral renal pleomorphic, hyperchromatic round cell tumor interpreted as lymphoma (Figure 2). The splenic nodules were hemorrhagic and necrotic, consistent with splenic infarction. Neoplastic cells were not noted in the spleen. All other assessed organs, including liver, lungs, heart, pancreas, adrenal glands, urinary bladder, stomach, duodenum, jejunum, ileum, colon, and lymph nodes were grossly and histologically normal, providing no evidence for systemic hypoxia. The brain was not removed. Immunohistochemical assessment of the renal lymphoma indicated strong and moderate expression of CD3f and CD18g, respectively, and lack of expression of CD79αh, confirming a diagnosis of T-cell lymphoma. Bone marrow biopsy performed immediately after euthanasia revealed erythroid hyperplasia with complete maturation as well as a normal proportion and morphology of WBC and platelet precursors. These findings were consistent with excessive bone marrow stimulation by EPO in secondary polycythemia.4
Citation: Journal of the American Animal Hospital Association 47, 2; 10.5326/JAAHA-MS-5614
Immunohistochemical EPO Staining
Immunohistochemical staining for EPO was performed on renal tumor tissue for both dogs. Tissue sections were deparaffinized, heated to 95°C for 10 min, and then incubated with a 1:100 dilution of a polyclonal rabbit antiserum to human EPOk for 1 hr. Sections were washed and bound antibody was detected with a visualization systeml and 3,3′-diaminobenzidine chromogenm. Although the antibody was not specifically validated for canine tissue, it had been previously used to detect canine EPO and showed clear staining of villous trophoblast cells in canine placenta used as positive control.5,6 Negative control sections were incubated with normal rabbit serum in place of antiserum. In sections of kidney from the dog described in case 1, occasional individual tumor cells expressed EPO. In the dog described in case 2, prominent (although focal) expression of EPO in tumor cells was apparent (Figure 3). Occasional nonneoplastic interstitial cells were also observed to express EPO in the dog described in case 2, and in the first case, rare tubular cells showed EPO immunolabeling.
Citation: Journal of the American Animal Hospital Association 47, 2; 10.5326/JAAHA-MS-5614
Discussion
Primary renal neoplasia in dogs is rare, often malignant, and most commonly caused by either carcinoma or adenocarcinoma.1,7 Dogs can present with any combination of nonspecific signs such as inappetence, weight loss, lethargy, vomiting, abdominal pain, or hematuria.7 The best diagnostic test for confirming neoplasia is a renal biopsy;1 however, ancillary testing such as a CBC, biochemistry profile, urinalysis, and abdominal ultrasound can provide important information regarding renal anatomy and function and indicate the presence of paraneoplastic syndromes such as polycythemia.8
Polycythemia consists of an increased concentration of erythrocytes relative to nonerythroid blood constituents. Polycythemia can be categorized into either relative polycythemia (which refers to hemoconcentration typically resulting from a decrease in plasma volume or transient increase in circulating RBC such as with splenic contraction) or absolute polycythemia, which is caused by an increase in total RBC mass.9 Absolute polycythemia can then be further categorized into primary and secondary polycythemia. Primary polycythemia, or polycythemia vera, is a rare myeloproliferative disorder where neoplastic precursor cells proliferate independent of stimulation by the glycoprotein hormone EPO.10 Secondary polycythemia is defined as being physiologically appropriate or inappropriate based on the stimulus for the increase in EPO.10,11 Physiologically appropriate increases in EPO result from perceived tissue hypoxia, such as with severe pulmonary disease, hemoglobinopathies, or malformation such as venoarterial shunting.9,10 The most common cause of an inappropriate increase in EPO resulting in excess production of erythrocytes in dogs is renal neoplasia.8–11 The development of secondary polycythemia as a consequence of an infiltrative renal disease is very rare with few reported cases.2,3,7,8,12–15 Other causes of physiologically inappropriate secondary polycythemia include nonrenal neoplasia. The ectopic production of EPO was demonstrated in two hepatic tumors and a carcinoma in horses,16–18 nonrenal lymphoma and mesenchymal tumors in dogs,2,19–21 and two cases of canine pyelonephritis.10,15
The pathophysiologic mechanism underlying the physiologically inappropriate increase in EPO associated with neoplasia, both renal and extrarenal, has been proposed to include either autogenous production of EPO or EPO-like factors by the tumor cells or EPO production secondary to local tissue hypoxia resulting from compression of normal cells and vascular supply by neoplastic masses.9 In human medicine, autonomous production of EPO from renal cell carcinoma cells has been convincingly shown.22 Polycythemia associated with increased EPO in a man with small lymphocytic lymphoma has been shown. Resolution of the polycythemia and normalization of EPO levels was noted after successful chemotherapy in that case.23 Direct expression of EPO in lymphoma cells in humans has not been evaluated.23,24 Evaluation of EPO production by tumor cells has been recommended as a prognostic marker and potentially to guide novel therapeutic strategies.25 In veterinary medicine, EPO production by tumor cells has been documented in only a small number of cases.13,20,26
In the two cases described here, both dogs had physiologically inappropriate concentrations of EPO associated with polycythemia. Histologic examination of tumors in both cases confirmed renal T-cell lymphoma. One case also had involvement of the liver and mesentery. Bone marrow biopsy was performed in one of the two cases, which revealed erythroid hyperplasia consistent with polycythemia due to stimulation from excessive EPO.4 The IHC suggested that the cause of inappropriate increases in EPO included both production by neoplastic cells and by benign renal tissue. In both dogs, tumor cells stained positive for EPO, suggestive of ectopic EPO production; however, nonneoplastic renal cells in both dogs also had cytoplasmic staining, suggesting a possible physiologic response to local hypoxia as a result of tumor compression within the kidney. It might also have been informative to assess EPO gene expression in tumor and other tissues; however, suitably fresh tissues were not available.
Chronically elevated EPO concentrations result in polycythemia with numerous physiologic consequences. Increased blood viscosity results in impaired blood flow, which can lead to local hypoxia or thrombosis.2,8,9 Diminished blood flow can manifest neurologically in the form of seizures and behavioral changes or physically as blindness, ataxia, or tremors.8 In the dog described in case 1, clinical signs consistent with abnormal brain activity were noted, and the HCT, measured within hours of this abnormal brain activity, was 65%. This high HCT could have resulted in reduced oxygenation of the brain. Other causes of the abnormal brain activity are also possible including primary central nervous system lymphoma or a cerebrovascular accident. Removal of the brain was not performed during the postmortem examination; therefore, the cause of the possible seizure activity is not known. The increases in PT and aPTT noted in the dog described in case 1 could be due to chronic disseminated intravascular coagulation, liver disease, sampling error, or laboratory error. No evidence of disseminated intravascular coagulation was present on either physical examination or postmortem examination, and chronic liver disease was ruled out with histopathology. Exposure to rodenticide was ruled out based on lack of exposure. Sampling error or laboratory error remained the most likely differential diagnoses. Repeat evaluation of these parameters was not performed. The cause of the hematuria in both cases was likely due to intermittent hemorrhage from the renal masses. There was no evidence of lower urinary tract disease on the ultrasound or postmortem examinations. Hematuria due to a coagulopathy could not be ruled out in the dog described in case 1; however, there was no visible hemorrhage from any other sites in this dog either clinically or during the postmortem examination.
Conclusion
This report describes two dogs that presented with polycythemia associated with inappropriate serum EPO concentrations, which were subsequently diagnosed with renal T-cell lymphoma. Polycythemia secondary to lymphoma is rarely reported in dogs.2,3 In addition, rarely in the veterinary literature has polycythemia secondary to renal neoplasia included measurement of EPO concentration and immunohistochemical evaluation of both neoplastic and nonneoplastic tissue for EPO.26 EPO expression in the analyzed tissues suggested that production was due to both production from tumor cells and from hypoxic renal cells. Monitoring EPO in treated cases may aid in assessing response to therapy and prediction of relapse.25
A section of kidney from the dog described in case 1 diagnosed with renal lymphoma and polycythemia. There is complete effacement of renal cortical architecture by a round cell tumor separating glomeruli and tubules. Hematoxylin and eosin stain, original magnification ×100. Bar=50 μm.
A section of kidney from the dog described in case 2 diagnosed with renal lymphoma and polycythemia. A round cell tumor infiltrate surrounds a glomerulus and renal tubules. Hematoxylin and eosin stain, original magnification ×400. Bar=10 μm.
A section of kidney from the dog described in case 2 diagnosed with renal lymphoma and polycythemia. Immunohistochemistry for erythropoietin (EPO) revealed that individual tumor cells express cytoplasmic EPO (brown stain). Original magnification ×1,000. Bar=10 μm.
Contributor Notes
