Suspected immune-mediated neutropenia and corticosteroid responsive pancytopenia in a Portuguese water dog

Abstract

An 8-year-old spayed Portuguese water dog was presented with dysuria, lethargy, and anorexia. A profound neutropenia and pancytopenia were identified. Bone marrow aspirates revealed neutrophilic hyperplasia, a significant left shift, and toxic changes, suggesting immune-mediated destruction as a likely underlying mechanism. Immunosuppressive therapy was instituted and clinical signs improved.

In May 2016, an 8.7-year-old spayed female Portuguese water dog was presented to the Central Island Veterinary Emergency Hospital on Vancouver Island with clinical signs of lethargy, dysuria, stranguria, and possible hematuria. A urine dipstick identified proteinuria, hematuria, and a urine specific gravity (USG) of 1.033. The patient was treated with amoxicillin/clavulanic acid (Clavaseptin; Vétoquinol, Lavaltrie, Quebec), 12.5 mg/kg body weight (BW), PO, q12h for 10 d, resulting in marked improvement. In early June 2016, the dog ate a piece of clothing, vomited 2 to 3 times, and remained anorexic for several days. Her owners reported that she spontaneously improved and therefore they did not seek veterinary attention. On June 20, 2016, the patient was presented to the Benson View Veterinary Hospital for recurrence of dysuria, stranguria, lethargy, and anorexia. On presentation she was mildly pyrexic (39.8°C), the remainder of her vital parameters were within normal limits, and no abnormalities were noted on physical examination. The patient was up-to-date on rabies, canine distemper virus, canine adenovirus type-2, canine parvovirus, canine parainfluenza virus (DA2PP), and Leptospira spp. vaccinations and had no history of travel or known allergies. A complete blood (cell) count (CBC) revealed pancytopenia, characterized by a mild normocytic, normochromic, non-regenerative anemia [hematocrit 0.34 L/L; reference interval (RI): 0.38 to 0.57 L/L], thrombocytopenia with shift platelets (9.0 × 10⁹/L; RI: 143 to 448 × 10⁹/L), mild lymphopenia (1.0 × 10⁹/L; RI: 1.1 to 5.0 × 10⁹/L), and profound mature neutropenia (0.3 × 10⁹/L; RI: 2.9 to 12.7 × 10⁹/L) with mild toxic changes. Biochemistry revealed an increased symmetric dimethylarginine (SDMA) (0.09 μmol/L; RI: 0 to 0.07 μmol/L), total protein (77 g/L; RI: 55 to 75 g/L), aspartate transaminase (AST) (58 IU/L; RI: 16 to 55 IU/L), creatine kinase (295 IU/L; RI: 10 to 200 IU/L), hypoalbuminemia (26 g/L; RI: 27 to 39 g/L), hyperglobulinemia (51 g/L; RI: 24 to 40 g/L), and hyperbilirubinemia (6.5 μmol/L; RI: 0 to 5.13 μmol/L). Total T4 was within normal limits (16.0 nmol/L; RI: 13.0 to 53.0 nmol/L). The dog urinated normally while in the clinic and urinalysis revealed a USG of 1.044, pH of 8.0, bilirubinuria, and proteinuria, while all other values were unremarkable. Urine sediment did not reveal red blood cells, white blood cells, crystals, or bacteria. An IDEXX SNAP 4Dx test (IDEXX laboratories, Westbrook, Maine, USA) was negative for Borrelia burgdorferi, Ehrlichia canis, Ehrlichia ewingii, Anaplasma phagocytophilum, Anaplasma platys antibodies, and Dirofilaria immitis antigen. Treatment with Cephalexin (500 mg, PO, q12h) was initiated and the dog was referred to the internal medicine service at Central Victoria Veterinary Hospital (CVVH).

On presentation at CVVH, the patient was moderately pyrexic (41.1°C) and lethargic. Physical examination identified prominent mandibular lymph nodes and mild dehydration. Biochemistry and CBC and biochemistry panels were repeated revealing a further decrease in hematocrit (0.25 L/L; RI: 0.37 to 0.61 L/L), lymphocytes (0.75 × 10⁹/L; RI: 1.05 to 5.10 × 10⁹/L), and neutrophils (0.25 × 10⁹/L; RI: 2.95 to 11.64 × 10⁹/L). Albumin, bilirubin, and total protein were normal. Prothrombin time and partial thromboplastin time were within normal limits. Samples for a blood culture were collected and treatment with ampicillin (Ampicillin Sodium; Novopharm, Toronto, Ontario), 440 mg, IV, q8h, and enrofloxacin (Baytril; Bayer, Toronto, Ontario) 200 mg, IV, q24h, was initiated.

Abdominal ultrasound showed hyperechoic material with linear striations in the mid small intestine, dilating the lumen of the intestine for a distance greater than 3 cm. The wall of the intestine at this location was thickened (0.5 cm) and the mesentery surrounding this location was voluminous and hyperechoic. These findings were consistent with an intestinal foreign body — presumed to be the clothing eaten recently, causing mechanical obstruction. Intestinal wall inflammation and possible peritonitis were also identified. The splenic parenchyma contained at least 3 hypoechoic foci up to 0.8 cm in diameter. The differentials for this finding include extra-medullary hematopoiesis, benign hyperplasia, or round cell neoplasia. All other abdominal organs were within normal limits. An exploratory laparotomy was recommended for removal of the suspected intestinal foreign body.

Prior to surgery, the patient’s blood type was determined to be Dog Erythrocyte Antigen (DEA) 1.1+ and a fresh whole blood transfusion (500 mL) was started at a rate of 3.33 mL/min. The patient was premedicated with fentanyl (Fentanyl Citrate; Summit Veterinary Pharmacy, Richmond, British Columbia), 3 μg/kg BW, IV, and lidocaine (Xylocaine 2%; Vétoquinol), 1 mg/kg BW, IV, boluses, induced with ketamine (Ketalean; Bimeda-MTC Animal Health, Cambridge, Ontario), 1 mg/kg BW, IV, and propofol (Propofol; Fresenius Kabi, Richmond Hill, Ontario), 10 mg/kg BW, IV, and maintained with isoflurane inhalant and oxygen on a circle system (fresh gas flow rate of 100 mL/kg BW per minute). Analgesia was provided throughout the surgery with a fentanyl (8 μg/kg BW per hour) and lidocaine (33 μg/kg BW per minute) continuous rate infusion (CRI). An incision was made from the xyphoid to pubis, upon which a small amount of serous peritoneal effusion was noted and collected for culture. All organs were grossly normal, although a small amount of hemosiderosis was noted on the ventral aspect of the tail of the spleen. The entire length of the small bowel was palpated and appeared grossly normal, with no foreign body identified. However, the distal colon contained firm compressible material and a foreign body could not be differentiated from fecal material. Finally, a 1 cm × 1 cm firm, purple nodule in the omentum was discovered and removed using blunt dissection for histopathology. This was later determined to be ectopic splenic tissue. A cystocentesis was performed during surgery and urine was collected for culture. The patient’s abdomen was closed using a routine three-layer closure and ampicillin, enrofloxacin, fentanyl/lidocaine CRI, and Normosol-R (2.2 mL/kg BW per hour) IV fluids were continued after surgery.

The following morning, the patient passed a small piece of rubbery, eraser-like material in her feces, which was likely the object palpated in her colon during the exploratory laparotomy. The patient’s vital parameters normalized (temperature 38.4°C); however, her lethargy and anorexia persisted. Blood, urine, and serous abdominal fluid cultures yielded no aerobic, anaerobic, or fungal growth. Metronidazole (Metronidazole; Hospira, Saint-Laurent, Quebec), 10 mg/kg BW, IV, q12h, maropitant (Cerenia; Zoetis, Kirkland, Quebec), 1 mg/kg BW, IV, q12h, sucralfate (Sulcrate Suspension Plus; Aptalis Pharma, Mont-Saint-Hilaire, Quebec), 0.25 mL/kg BW, PO, q8h, pantoprazole (Pantoprazole Sodium; Fresenius Kabi), 1 mg/kg BW, IV, q24h, and buprenorphine (Vetergesic; Alstoe Animal Health, Whitby, Ontario), 0.01 mg/kg BW, IV, q8h, were added to her regimen and the fentanyl/lidocaine CRI was discontinued. The dog’s packed cell volume (PCV) was maintained at approximately 31.6% and her lymphopenia (0.31 × 10⁹/L; RI: 1.05 to 5.10 × 10⁹/L), thrombocytopenia (15 × 10³/μL; RI: 148 to 484 × 10³/μL), and severe neutropenia (0.17 × 10⁹/L; RI: 2.95 to 11.64 × 10⁹/L) persisted.

Further diagnostics were pursued at this point, including ventral-dorsal, left and right lateral radiographs of the thorax, which were within normal limits. The tick polymerase chain reaction (PCR) performed through IDEXX (Rickettsia rickettsii, Mycoplasma haemocanis, Neorickettsia risticii, Leishmania spp., Hepatozoon spp., Ehrlichia spp., Babesia spp., Anaplasma spp., Bartonella spp., Candidatus Mycoplasma haematoparvum) was negative. The patient was premedicated for a bone marrow aspirate with buprenorphine (0.01 mg/kg BW, IV) and midazolam (Midazolam; Sandoz, Boucherville, Quebec), 0.2 mg/kg BW, IV, induced with propofol (5 mg/kg BW, IV) and maintained on isoflurane inhalant and oxygen on a circle circuit (fresh gas flow rate of 100 mL/kg BW per minute). The bone marrow aspirates were reviewed by True North Veterinary Diagnostics (Langley, British Columbia) and revealed a myeloid:erythroid (M:E) ratio of 8:1, with the neutrophilic series predominating. There was a significant left shift within the neutrophilic series as well as dysplastic features, including abnormal nuclear:cytoplasmic maturation, retention of primary granules, giant band cells, and toxic changes. Megakaryocytes were subjectively decreased with some containing irregular segmented nuclei. Macrophages were noted occasionally, containing hemosiderin and some phagocytosed cellular debris. Both stem cell injury (as a result of transient insult/toxicity or immune-mediated destruction) and myelodysplastic syndrome were noted as possible causes for these findings, with stem cell injury considered more likely as evidenced by the toxic changes. Blast cells did not exceed 20% of the nucleated cell count, thereby ruling out leukemia as a possible cause for the dog’s illness.

In light of these findings, 1 dose of dexamethasone (Dexamethasone 5; Vétoquinol), 0.22 mg/kg BW, IV, was given, after which the patient was placed on prednisone (NOVO-Prednisone; Novopharm, Toronto, Ontario), 1.5 mg/kg BW, PO, q24h. The dog remained anorexic in hospital and resistant to having a liquid diet fed by syringe. Her owners elected to try managing her at home in the hopes that she would improve clinically and feel more comfortable eating in a familiar environment. She was discharged and prescribed oral prednisone (NOVO-Prednisone; Novopharm), 1 mg/kg BW, q12h for 7 d; tapering to 1 mg/kg BW, q24h for 7 d then 1 mg/kg BW, q48h for 14 d, omeprazole (Omeprazole; Sandoz), 0.5 mg/kg BW, PO, q24h for 7 d, enrofloxacin (Baytril; Bayer), 7.5 mg/kg BW, PO, q24h for 7 d, and amoxicillin/clavulanic acid (Clavaseptin; Vétoquinol), 12.5 mg/kg BW, PO, q12h for 7 d.

After returning home, the dog’s energy level greatly improved, her appetite returned to normal, and she continued to urinate normally. A CBC and a biochemistry panel were repeated at the Island Veterinary Hospital after the patient had completed her 4-week long prednisone trial. The dog’s anemia had resolved (hematocrit 0.40 L/L; RI: 0.38 to 0.57 L/L) and a regenerative response was present (reticulocytes 188.0 × 10³/μL; RI: 10 to 110 × 10³/μL). Her thrombocytopenia had also resolved (186 × 10⁹/L; RI: 143 to 448 × 10⁹/L); however, her lymphopenia (0.2 × 10⁹/L; RI: 1.1 to 5.0 × 10⁹/L) and neutropenia (0.3 × 10⁹/L; RI: 2.9 to 12.7 × 10⁹/L) remained. Alanine transaminase (ALT; 137 IU/L; RI: 18 to 21 IU/L), alkaline phosphatase (ALP; 266 IU/L; RI: 5 to 160 IU/L), and gamma-glutamyltransferase (GGT; 17 IU/L; RI: 0 to 13 IU/L) showed mild elevations, likely as a result of corticosteroid therapy. The dog was placed on doxycycline (Teva Doxycycline; Teva Pharmaceuticals, Scarborough, Ontario), 5 mg/kg BW, PO, q12h for 14 d to combat any underlying infectious causes prior to repeating a CBC and biochemistry panel in August. Results were unavailable at the time of writing.

Discussion

Neutropenia in dogs can be due to increased use (during severe suppurative inflammation/infection), decreased production (as a result of primary or secondary insults to bone marrow neutrophilic precursor cells), or accelerated destruction of neutrophils through immune-mediated mechanisms (1). Primary immune-mediated neutropenia (IMN) is an uncommon disorder in veterinary medicine and is diagnosed clinically based on exclusion of other mechanisms that cause neutropenia, as well as response to corticosteroid immunosuppressive therapy (1–9). The disease has been documented in dogs (1–4,7–9), cats (6), and horses (10) and involves the direct targeting of neutrophils by autoantibodies, resulting in subsequent opsonization and phagocytosis of neutrophils (5,7,9–11).

In humans, several methods have been described for detecting anti-neutrophil antibodies in circulation which may provide a more definitive diagnosis for IMN when other causes have been ruled out (3–6,8–9,11). Most of these tests have proven to be unreliable for use in veterinary patients, although 1 study showed promising results using flow cytometry to detect anti-neutrophil antibodies in dogs affected with IMN (3). One of the key limitations in detecting anti-neutrophil antibodies lies within the neutrophils themselves as neutrophils are fragile cells that do not store well without activation, consequently leading to autolysis (4,7,11). Also, false positives are frequently encountered due to non-specific binding of immunoglobulin G (IgG) to their neutrophilic target cells (6,11) and false negatives may occur in severely neutropenic patients which are too neutrophil depleted to permit proper performance of the test (3,6). Therefore, there is a need for a safe, reliable, and efficient method for diagnosing immune-mediated neutropenia in veterinary patients, as arriving at a diagnosis continues to be an expensive and difficult challenge.

In veterinary patients, the process of diagnosing IMN is a tedious and often expensive process of elimination. Individuals affected by IMN typically present with non-specific signs that do not reflect the severity of their underlying disease. This commonly includes lethargy, anorexia, and vomiting (9), all of which were reported in this particular case. When a CBC is performed and reveals a profound neutrophil depletion, often accompanied by deficiencies in other cell lineages, the extent of the patient’s disease is discovered. Infectious causes of neutropenia (viral, bacterial, and fungal) can be ruled out through vaccination history, blood and urine cultures, 4Dx SNAP, and tick PCR tests, all of which were performed in the case presented here. In addition, septic peritonitis and gastric or enteric perforation due to a foreign body were definitively ruled out in this patient through gross examination of the abdominal organs and culture of the abdominal effusion.

Primary and secondary insults to bone marrow precursor cells can be excluded through obtaining a thorough medical history, including recent drug administration and eliminating gross neoplasia through thoracic and abdominal imaging as well as bone marrow analysis. An extensive list of drugs that are potentially myelotoxic was reviewed by this patient’s owners and it was determined that there was no history of any exposure prior to her clinical signs appearing in May. Her abdominal ultrasound, thoracic radiographs, and bone marrow aspirates did not support the presence of a neoplastic process. Bone marrow aspirates from IMN patients typically reveal normal erythroid and megakaryocytic cells, an abnormally high M:E ratio, and neutrophilic hyperplasia with dysplastic features (5), which were evident in this patient.

One of the consistent hallmarks of immune-mediated neutropenia in dogs and cats is a rapid response to immunosuppressive therapy (1–4,6–9), with some patients’ neutrophil count achieving normal levels as early as 3 d after initiating treatment (2). However, relapse after cessation of immunosuppressive therapy is common. Concurrent use of prednisone and azathioprine is recommended for refractory cases or those whose response is suboptimal (7). Despite resolution of clinical signs and her owners reporting that she was doing well, the patient’s neutropenia did not resolve after 4 wk of prednisone therapy.

It is worth noting that predisposition to secondary infections is most often mild to moderate in human patients with primary autoimmune neutropenia (2). Treatment in these patients aims to control secondary infections when they occur through use of antibiotics, with large doses of immunosuppressive steroids used in instances that require more aggressive treatment (2). In support of these findings, 2 dogs affected by IMN showed marked improvement in their clinical signs following administration of antibiotics alone without the use of corticosteroids, although their neutropenia remained unchanged (2). Similar results were observed in the case presented here, as the patient’s lethargy, stranguria, dysuria, and anorexia quickly resolved following antibiotic treatment although her neutropenia persisted. Interestingly, the dog’s anemia and thrombocytopenia responded completely to corticosteroid therapy. Therefore, it is possible that the dose, duration of therapy, the immunosuppressive drug chosen, or a combination of these factors were not sufficient in treating her neutropenia.

Immune-mediated neutropenia remains a challenging and complex disease. This case serves to highlight both the difficulty in establishing a diagnosis of IMN and the importance of finding a treatment regimen that is tailored to the individual animal. Further efforts are needed to validate diagnostic tests for this disease in veterinary medicine in order to establish a timely diagnosis and initiate treatment promptly. Finally, the use of various immunosuppressive drugs and treatment protocols in controlling this disease should be explored to provide veterinarians different options to use under their clinical judgment.