|Home | About | Journals | Submit | Contact Us | Français|
A 3-year-old Standardbred gelding with a history of pyrexia, persistent hemorrhage from the oral cavity, and a large, soft swelling at the junction of the caudal aspect of the mandibular rami and proximal neck was evaluated. The horse had neutropenia and anemia, with atypical granulated cells in a blood smear. Additional tests confirmed acute myeloid leukemia with basophilic differentiation, which has been reported in humans, cats, dogs, and cattle but not horses.
Leucémie myéloïde aiguë avec différenciation basophile chez un hongre Standardbred âgé de 3 ans. Nous avons évalué un hongre Standardbred âgé de 3 ans avec une anamnèse de pyrexie, d’hémorragie persistante de la cavité orale et d’une grosse enflure molle à la jonction de l’aspect caudal des rameaux mandibulaires et du cou proximal. Le cheval souffrait de neutropénie et d’anémie avec des cellules granulées au frottis sanguin. Des tests additionnels ont confirmé une leucémie myéloïde avec différenciation basophile, qui avait déjà été signalée chez les humains, les chats, les chiens et le bétail, mais pas chez les chevaux.
(Traduit par Isabelle Vallières)
A 3-year-old Standardbred gelding (trotter) was presented to the Ontario Veterinary College — Health Sciences Centre for evaluation of a large swelling under the mandibular rami and proximal neck, pyrexia, anemia, and anorexia. The horse had been castrated approximately 30 d earlier with no complications reported. He was lightly exercised for 7 d following the procedure; however, this was discontinued when the horse became lethargic. Twelve days prior to presentation, mild bleeding from the oral cavity was noted and was attributed to retained deciduous teeth. The teeth were removed but the hemorrhage continued. Two days prior to presentation, the horse became pyrexic and developed a large fluctuant swelling at the junction of the caudal aspect of the mandibular rami and proximal neck. On the day of presentation, the owners noted that the gelding’s abdomen appeared distended.
On physical examination, the horse was quiet but responsive to his environment. There was active but mild hemorrhage from the oral cavity as well as venipuncture sites from previous medication administered by the referring veterinarian. Those treatments included sodium penicillin (Novapharm, Toronto, Ontario), gentamicin (Merck Animal Health, Kirkland, Quebec), and flunixin meglumine (Pfizer Animal Health, Kirkland, Quebec). The horse was tachycardic [68 to 80 beats/min (bpm)] with a regular rhythm and a strong pulse quality. The respiratory rate and effort were within normal limits and although lung sounds were detected on both sides, no crackles or wheezes, or nasal discharge was noted. The rectal temperature was mildly elevated at 38.6°C [reference interval (RI): 37.0°C to 38.0°C]. Gastrointestinal sounds were present and there was moderate to marked abdominal distension. There were no petechial or ecchymotic hemorrhages on the oral mucous membranes and no obvious trauma to the upper or lower incisors, mucous membranes, or tongue. There was mild swelling of the left hind tendon sheath and distension of both radio-carpal joints. No heat was detected in any joint and the horse ambulated around the stall with ease. A plaque of ventral edema was noted, extending from the pectoral region down through the elbows distally and caudal towards the prepuce. Blood was collected from the left cephalic vein using a 22-gauge needle and placed into vacutainer tubes containing EDTA and citrate. The samples were submitted for complete blood (cell) count (CBC) and serum biochemistry profile, as well as coagulation assays including pro-thrombin time (PT), partial thromboplastin time (PTT), thromboelastography (TEG), plasma fibrinogen concentration, and D-dimer concentration. Urine was collected for complete urinalysis.
The CBC revealed a moderate leucopenia (2.7 × 109 cells/L, RI: 5.1 to 11.0 × 109 cells/L), decreased hemoglobin concentration (107 g/L, RI: 112 to 169 g/L), decreased hematocrit (0.31 L/L, RI: 0.38 to 0.55 L/L), and severe neutropenia (0.86 × 109 cells/L, RI: 2.8 to 7.7 × 109 cells/L). The red cell indices were within reported reference intervals. The PT and PTT were both greater than 100 s (RI: 10.1 to 13.2 s and 35.8 to 49.2 s, respectively). The platelet count was within the reference interval, while the plasma fibrinogen concentration was markedly decreased (0.4 g/L, RI: 1.2 to 2.3 g/L). The thromboelastography failed to produce a clot within the test period, indicating hypocoagulability. The D-dimer concentration was markedly elevated at 6.8 μg/mL (control value: 0.33 μg/mL). The urinalysis was unremarkable. The serum biochemistry profile revealed moderate hypoalbuminemia (18 g/L, RI: 30 to 37 g/L), hyperglobulinemia (42 g/L, RI: 26 to 41 g/L), an altered album in to globulin (A:G) ratio (0.43, RI: 0.8 to 1.3), and elevated serum amyloid A (270 mg/L, RI: 0 to 19 mg/L).
Based upon the clinical examination and laboratory data, differential diagnoses for multiple-site hemorrhage included vitamin K antagonism via rodenticide toxicity or ingestion of dicoumarol (moldy sweet clover), immune-mediated destruction of clotting factors, idiosyncratic drug reaction, and disseminated intravascular coagulation related to sepsis or neoplasia.
A thoracic ultrasound examination (ALT ultrasound and Curve Array C52 probe Model HDI 3000; Bothell, Washington, USA) revealed only a very mild visceral pleural roughening on both the right and left thorax. An abdominal ultrasound was unremarkable except for a moderate to marked amount of free fluid. An ultrasound examination of the mass at the caudal aspect of the mandibular rami revealed flocculent material with a honeycomb appearance.
An abdominocentesis was performed and the mass at the mandibular rami was aspirated. Cytological analysis confirmed the presence of hemoabdomen and hematoma, respectively. The aspiration sites bled at a rate of 1 to 2 drops/s. This rate was unchanged even after prolonged compression.
An indwelling intravenous catheter was placed in the left lateral thoracic vein for administration of medication and fluid therapy. Three liters of hyper-immune plasma (Equiplas; Plasvacc USA, Templeton, California, USA) were administered with the rationale of providing clotting factors to re-establish hemostasis. Vitamin K therapy (Phytonadione Injection; Vétoquinol, Lavaltrie, Quebec) was also initiated at 500 mg, SQ, q12h. The antibiotic regimen which had been initiated by the referring veterinarian was continued. This regimen consisted of gentamicin (Merck Animal Health), 6.6 mg/kg body weight (BW), IV, q24h, and sodium penicillin (Novapharm), 22 000 IU/kg BW, IV, q6h. The gelding was maintained in a quiet, low traffic room, and offered free choice hay and water.
Bleeding from the oral cavity, jugular venipuncture sites, abdominocentesis, and mass aspiration sites continued at a rate of 1 to 2 drops/s. Progressive distension of multiple joints, presumed to be due to hemarthrosis, also continued which increasingly made the horse more uncomfortable when moving around the stall. The packed cell volume had significantly decreased since admission (0.16 L/L, RI: 0.31 to 0.55 L/L) while the total solids concentration remained unchanged. An additional 3 L of plasma were transfused (Equiplas; Plasvacc USA) and dexamethasone (Dominion Veterinary Laboratories, Winnipeg, Manitoba), 0.2 mg/kg BW, IV, q24h, was added to the treatment. The dose and route of vitamin K administration were modified to 1.0 mg/kg BW, IV, diluted in 1 L of saline q12h in order to avoid further bleeding from injection sites.
The horse continued eating and drinking well but remained mildly tachycardic (48 to 54 bpm) with strong peripheral pulses and mildly elevated temperature (38.5°C).
The rate of hemorrhage had significantly slowed from the abdominocentesis site and had subsided from the mass caudal to the mandibular rami, oral cavity, and venipuncture sites by day 3 in the hospital. However, the packed cell volume had decreased to 0.09 L/L, necessitating a 5 L whole blood transfusion. The vitamin K1 administration was modified to 0.5 mg/kg BW, IM, q12h. The corticosteroid and antibiotic therapy was discontinued.
Repeat CBC performed on day 4 of hospitalization revealed an improvement in the white blood cell count (4.3 × 109 cells/L, RI: 5.1 to 11.0 × 109 cells/L) but persistent non-toxic neutropenia (0.77 × 109 cells/L, RI: 2.8 to 7.7 × 109 cells/L), moderate anemia (RBC 4.1 × 109 cells/L, RI: 6.9 to 10.7 × 109 cells/L; hemoglobin 64 g/L, RI: 112 to 169 g/L; hematocrit 0.19 L/L, RI: 0.38 to 0.55 L/L), an increase in red cell distribution width (21.2%, RI: 16.3% to 20.4%) suggestive of erythrocyte regeneration, and hypoproteinemia (56 g/L, RI: 57 to 75 g/L). Other red cell indices and the platelet count remained within reference limits. The manual differential blood cell examination revealed occasional non-segmented leukocytes with basophilic cytoplasmic granules. The serum biochemistry profile variables had also improved except for mild hyponatremia (133 mmol/L, RI: 136 to 144 mmol/L), increased glutamate dehydrogenase (GLDH) activity (16 U/L, RI: 1 to 7 U/L), and a consistently elevated serum amyloid A (250.6 mg/L, RI: 0 to 19 mg/L).
An anticoagulant screen for brodifacoum, bromadiolone, chlorophacinone, diphacinone, pindone and warfarin using high performance liquid chromatography was negative.
A bone marrow aspirate and core biopsy were collected from the right tuber coxae using a bone marrow biopsy needle (Kendall Monoject Bone marrow Biopsy needle, 11G×4; Tycohealthcare Group, Mansfield, Massachusetts, USA) due to the presence of abnormal leukocytes and the persistent cytopenias. The bone marrow aspirate was poorly cellular and the predominant cell was a population of atypical round cells that contained few to many basophilic granules. These cells were large with pale basophilic cytoplasm, moderate nuclear:cytoplasmic ratio, and round nuclei occasionally with a single prominent nucleolus (Figure 1). Limited differentiation was present, as occasional cells with indented or segmented nuclei contained similar granules. Other cell lines were minimally represented, although a few erythroid precursors were noted. Toluidine blue staining of additional aspirate slides confirmed metachromatic granules typical of either mast cells or basophils. Cells of myeloid lineage with basophilic differentiation were suspected based upon nuclear morphology and paucity of metachromatic granules (Figure 1). These findings were interpreted as acute myeloid leukemia with basophilic differentiation. Due to the guarded prognosis for survival, the gelding was euthanized with an intravenous overdose of pentobarbital (Merck Animal Health).
A postmortem examination revealed ecchymotic and petechial hemorrhages affecting the heart, lungs, and abdominal cavity. Subcutaneous edema was present along the ventral abdomen and forelimbs as well as within the mesenteric fat and wall of the urinary bladder. There was 5 L of serous fluid within the peritoneal cavity and approximately 250 mL of serous fluid within the pericardial sac.
The bone marrow grossly appeared approximately 80% active with 20% fat. Histologically, cellularity of the marrow was > 95%. Approximately 70% of the cells were large, undifferentiated blast cells or myeloid precursors with eosinophilic cytoplasm, large, central round nuclei with occasional clefting and finely stippled chromatin (Figure 2). Approximately 10% to 15% of these cells had distinct intracytoplasmic granules when stained with toluidine blue. The myeloid to erythroid ratio was 30:1. Similar cells were found in the liver and lymph nodes. Immunohistochemical staining showed that both CD3+ T lymphocytes and CD20+ B lymphocytes were scattered throughout the sections, but the neoplastic cells were negative for both stains, ruling out lymphoid origin of the cells. Specific immunohistochemical identification of the cells as myeloid in origin could not be performed due to the lack of equine antibody markers. The final diagnosis was acute myeloid leukemia and diffuse coagulopathy.
A rare acute myeloid leukemia with an unusual manifestation of disseminated intravascular coagulation (DIC) in a 3-year-old Standardbred gelding is described. Myeloproliferative disorders have been infrequently reported in the horse (1) and have included chronic granulocytic leukemia (2), monocytic leukemia (3), myelomonocytic leukemia (4), and eosinophilic myeloproliferative disorders (5).
The unexplained bleeding from the oral cavity and hematoma of the throat latch region prompted referral of this case. In horses, hemorrhage can occur as a result of trauma, thromobocytopenia due to immune-mediated destruction (6), chronic liver failure (7), deficient or defective coagulation factors (8,9), vitamin K consumptive coagulopathies including rodenticide intoxication (10), and moldy sweet clover poisoning (11), and DIC secondary to sepsis and neoplasia (6,12). In the present case, the uncontrollable hemorrhage was likely associated with uncompensated DIC, as indicated by the prolonged PT and aPTT, decreased fibrinogen concentration, increased D-dimers, and the failure of the blood sample to clot using thromboelastography (TEG).
In humans, DIC is a well-documented complication in patients with hematopoietic malignancies (13) and has been associated with altered platelet function (14,15), increased cell membrane expression of tissue factor (16), neoplasm associated procoagulant and prothrombotic agents (16), and decreased circulation of anti-coagulant factors (16–18). In some patients with hematopoietic malignancies, initial hypercoagulability causes platelet and coagulation factor consumption (13,16). Enhanced fibrin deposition results in up-regulation of the fibrinolytic system (13). Since there is no single diagnostic test to confirm DIC, the clinical findings and results of various coagulation assays (including PT, APTT, fibrinogen concentration, platelet count, fibrin degradation products and/or D-dimers) are used to support the diagnosis (19). A scoring system developed for use in human medicine takes into account the findings of specific coagulation assays for the diagnosis of overt and non-overt DIC, and has been utilized in patients with hematologic malignancies (19,20). This score incorporates the results of platelet count, the concentration of fibrin degradation products, prothrombin time, and activity of antithrombin and protein C and elevated thrombin-anti-thrombin (TAT) complexes as indicators of the presence of DIC (20). A similar validated scoring system does not exist for horses.
Interestingly, the platelet count in this horse was within our laboratory’s reference interval and remained so throughout the duration of hospitalization. No assessment of platelet function was attempted; there was no evidence of excessive removal or consumption based upon consecutive platelet counts. However, the coagulation assays confirmed a profound deficiency of activity of the coagulation factors of both the intrinsic and extrinsic pathways. The TEG, which measures the coagulation process from initial clot formation to activation of the coagulation factors (21) failed to develop any clot within the allotted time. TEG analysis for coagulation has been evaluated in healthy Thoroughbred horses (22), in foals (23), and critically ill adult horses (24). However, TEG analysis has not been reported and validated as a means to support the diagnosis of DIC in horses. Both the PT and APTT were significantly prolonged, indicative of ineffective or insufficient function and activation of the extrinsic and intrinsic coagulation pathways, consistent with the TEG result. Activation of the fibrinolytic system was documented by the markedly elevated D-dimer concentration, which has been associated with poor prognosis and non-survival in horses with colic (25).
In humans, coagulopathy associated with acute promyelocytic leukemia is a defining characteristic of the disease, and patients will often present with signs of DIC prior to diagnosis of the primary condition (26). The coagulopathy is a syndrome of hypercoagulability, DIC, fibrinolysis, and proteolysis causing hemorrhage and thrombosis (26). Thrombocytopenia is often noted and is partially attributed to myelophthisis (26). The leukemic cells express factors which contribute to the development of bleeding, including increased expression of annexin II, enhanced cytokine production, increased proteolysis, and the creation and activation of microparticles (26). Microparticles are cell-derived membrane fragments measuring 0.1 to 1.0 μm, originating from normal cells, such as platelets, blood cells, and endothelial cells, or malignant cells. Microparticles are involved directly and indirectly in activating coagulation and has been associated with reduced survival and higher rates of venous thromboembolism in patients with metastatic breast and pancreatic cancer (26). The presence of microparticles was not evident in this horse, but wasn’t investigated further.
Similar to the presentation frequently observed in humans, the patient was presented with evidence of a bleeding disorder and a diagnosis of DIC was supported by the analysis of various coagulation assays. Thrombocytopenia was not a feature of the syndrome in this horse; however, platelet function was not evaluated and therefore the functionality of the platelets may have been less than optimal.
Successful resolution of DIC is dependent on the identification and aggressive treatment of the initiating disease process (6). In this particular case, the inciting cause was not initially apparent, therefore supportive therapy with vitamin K administration to support the function of the vitamin K-dependent clotting factors as well as 2 plasma transfusions to provide coagulation factor proteins was administered. This treatment appeared to result in minor clinical improvement in hemostatic function and general demeanor. However, the effects were short-lived and ultimately overwhelmed by the primary condition.
The clinical presentation of myeloproliferative disorders in horses is generally non-specific with intermittent episodes of pyrexia, anorexia, and lethargy (1). Lethargy and exercise intolerance were the first clinical signs observed in this case and were followed by undulating pyrexia. Evidence of involvement of the hematopoietic system was first observed following the development of the fluctuant swelling at the caudal aspect of the mandibular rami, likely following self-trauma, and marked anemia. The first possible indication of a leukemic process was observed in the peripheral blood 4 d following admission when occasional non-segmented leukocytes with basophilic cytoplasmic granules were noted, suggestive of basophilic myelocytes. The bone marrow aspirate revealed megakaryocyte and erythroid hypoplasia as well as myeloid hyperplasia, although the immaturity of the cells made it difficult to differentiate them from cells of lymphoid origin. The definitive diagnosis of acute myeloid leukemia was made upon postmortem examination in which ~70% of the cells in the bone marrow were large undifferentiated blast cells which had toluidine blue positive granules and were negative for lymphocyte markers.
Basophilic leukemias have been reported infrequently in the veterinary literature, and may occur as a myeloproliferative neoplasm or as an acute process (27). Both develop from the neoplastic transformation of multipotent bone marrow stem cells and are further subdivided into granulocytic and monocytic leukemias (27). Basophilic leukemias have been reported in dogs (28,29), cats (30,31), and a calf (32), but we are unaware of reports of any in horses.
Acute basophilic leukemia is a rare form of acute myeloid leukemia (33,34). It is diagnosed based on cellular morphology, metachromasia with toluidine blue staining, electron microscopy (34,35), and flow cytometric analysis (35), although it is difficult to differentiate from other similar myeloid leukemias (34). Peripheral basophilia may or may not be present, similar to the initial presentation of the case reported here (34); however, blast cells with variable metachromatic granules compose a significant proportion of cells in bone marrow aspirates (34) or core biopsy (33). Diagnosis in this case was based upon cellular morphology and by the presence of toluidine blue positive granules in tumor cells. Toluidine blue positive granules are not present in cells of lymphocytic or monocytic origin. Other specific diagnostic modalities including flow cytometry and cytogenetic studies were not performed due to lack of availability of equine specific reagents.
Treatment with chemotherapeutic agents such as hydroxyurea has been reported in the human and canine literature; however, chemotherapy was not attempted in this case (36) due to overt signs of DIC, poor prognosis for survival, and costs associated with hospitalization and critical care.
Use of this article is limited to a single copy for personal study. Anyone interested in obtaining reprints should contact the CVMA office (gro.vmca-amvc@nothguorbh) for additional copies or permission to use this material elsewhere.