Canine monocytic ehrlichiosis (CME) is a tick-borne (Rhipicephalus sanguineus) disease of canids, caused by the rickettsia Ehrlichia canis, and is now recognized worldwide (1). CME consists of acute, subclinical and chronic phases. Dogs may recover from the acute phase of the disease without treatment, however some remain subclinically infected, and several months to years later may develop the severe pancytopaenic chronic stage. This pancytopaenic chronic phase of CME has a grave prognosis despite therapy (1,2,3,4,5). Hematopoietic growth factors have been proposed as a rational treatment for the chronic pancytopaenic phase (5). However, no clinical trials using this mode of therapy have been published. This report describes the successful treatment of a severely pancytopaenic dog in the chronic stage of CME, with long-term corticosteroid therapy in combination with two synthetic recombinant haematopoietic factors: human granulocyte colony stimulating factor (rhG-CSF) and human erythropoietin (rh-EPO).  

A three-year old male weimaraner dog was referred to the Hebrew University Veterinary Teaching Hospital with chief complaints of exercise intolerance for the last 3 months, fever, and a history of tick infestation. Two weeks prior to presentation, the referring veterinarian performed a complete blood count (CBC) which revealed a severe pancytopenia {white blood cells (WBC) - 0.29x103/mm3, platelets (Plt) - 2x103/mm3, red blood cells (RBC) - 3.94x106/mm3}. Serologic testing for E. canis with a commercial dot-blot enzyme-linked immunosorbent assay kit (Biogal Laboratories, Galed, Israel) was strongly positive. A presumptive diagnosis of CME was made, and the dog was treated with doxycycline (5mg/kg q12h PO for 14 days) and imidocarb-dipropionate (5mg/kg IM, single injection).

On the day of presentation (day 1) the dog’s general condition had deteriorated, it was lethargic, anorexic, emaciated and had a fever (40.60C), tachycardia (160/min), tachypnea (60/min), severe dyspnea, increased respiratory sounds and pale mucous membranes. CBC revealed severe pancytopenia (Table 1), severe anemia (RBC - 2.48x106/mm3), and no evidence of RBC regeneration in peripheral blood smears (reticulocyte count was 0%). The dog had hyperglobulinemia (globulin - 6.2g/dl), mild hypoalbuminemia (albumin — 2.4 g/dl), with an A/G ratio of 0.39. Thoracic radiographs showed signs consistent with severe bronchopneumonia. The E. canis IgG indirect immunofluorescence antibody test (IFA) titer was 1:5120. Severe chronic CME, with secondary bronchopneumonia, were diagnosed.

Table 1: Hematological values of a dog with severe chronic pancytopenia due to CME, successfully treated with hematopoietic growth factors and prednisone, during a 476 day period

Nd - not determined, WBC - White blood cells, RBC - Red blood cells, HB - Haemoglobin, Ht - Haematocrit, MCV - mean corpuscular volume, MCH - mean corpuscular haemoglobin, MCHC - mean corpuscular haemoglobin concentration, Plt - Platelets, TS - total solids.

The hematological results over a 476 days period are summarized in table 1. The main treatments administered, comments of diagnoses, major clinical signs and additional data are summarized in Table 2. Initial treatment included supportive care, cefazolin  (20mg/kg IV q8h for 10 days; Cefamezin, Teva , Israel), sulfadoxin/trimethoprim (15mg/kg IV q12h for 10 days; Trimoxin, Teva, Israel), gentamicin (2.2mg/kg IV q8h for 10 days; generic, Abic, Israel), aminophylline (10mg/kg IV q8h for 10 days; generic, Teva, Israel), and a single injection of imidocarb-dipropionate (5mg/kg IM; Imizol, Pitman-Moore, England).

As the dog was stabilized it was discharged on day 10 with oral cephalexine (20mg/kg q8h; Ceforal, Teva, Israel), sulfamethoxazole/trimethoprim (20mg/kg q12h; Resprim, Teva, Israel) and theophylline (10mg/kg q8h; Theo-Dur, Key Pharmaceuticals). On this day, recombinant-human granulocyte colony stimulating factor (rhG-CSF) (Neupogen, Roche) therapy was initiated (Table 2) for an overall period of 31 days (Tables 1 & 2). Resolution of bronchopneumonia was noticed clinically, and confirmed by thoracic radiography on day 26 (table 2). Four days later, the antibiotic treatment was changed to doxycycline (5mg/kg q12h; Doxylin, Dexxon, Israel), and prednisone (40mg q24h; generic, Rekah, Israel), and recombinant-human erythropoietin (rhEPO; Eprex, Cilag-AG, Germany) therapy was initiated. Erythropoietin was injected SQ q3d for a period of 36 days, followed by 5 consecutive treatments q14d (table 2). During EPO treatment, iron (20 mg/kg IM q7d), Abidex, Abic, Israel) and vitamins (3ml SQ q7d, Shefa-vit, Abic, Israel) were supplied continuously.  The E. canis IFA titer was 1:1280 on day 95.

Vincristine-sulfate (Vincristine, Abic, Israel) was given at 0.5mg/m2 IV on day 191, when the dog showed severe thrombocytopaenia and petechiae, and repeated 5 days later. The platelet numbers (Plt) increased to 3.6x103/mm3, 6 days later. 

Table 2: Hematological findings, related clinical signs, and treatments during a 476-day follow-up period of a dog with severe chronic CME and successfully treated with haematopoietic growth factors and prednisone.  

Treatment with immunosuppressive doses of prednisone (2mg/kg q24h PO) was initiated on day 26, when the bronchopneumonia was resolved, and was continued at that dose for 165 days, then reduced to 0.5mg/kg q24h PO. The dose was reduced further (to 0.5mg/kg EOD) on day 309. At that time the dog was still mildly anemic (Ht - 31.7%) and moderately thrombocytopenic (Plt - 9.8x103/mm3). On day 427, when all haematological values were within reference intervals, the dose was reduced further. On day 476, the dog was normal, the E. canis IFA titer was 1:20, and medication was discontinued.

The diagnosis of chronic CME in the present case was based on typical clinical, haematological and biochemical findings, a very high ehrlichial IgG IFA titer and clinical deterioration with no response to anti-rickettsial therapy. Marked pancytopenia has been reported as the hallmark of the severe chronic phase of CME (5), and occurs as the result of a hypocellular bone marrow (1,3,5,6). Although bone marrow analysis was not performed in the present case due to the owner’s objection, hypocellularity was presumed as the severe anaemia was non-regenerative and peripheral cytopaenia was evident in all three myeloid cell lines. In our experience, severe pancytopenia due to chronic E. canis infection with the Israeli strain carries a grave prognosis and to date conventional therapy has not been successful (3). 

The main objective of the treatment in the first period of therapy was to resolve the bronchopneumonia by providing a broad-spectrum bacteriocidal antibiotic. However, the very low numbers of circulating neutrophilic granulocytes might have led to failure. Human and canine recombinant granulocyte colony stimulating factors were reported previously to increase in a dose-dependent manner, the number and function of neutrophils (7). Phagocytosis, superoxide production and antibody-dependent cytotoxicity are increased by G-CSF. In a study performed on normal healthy dogs with recombinant canine G-CSF (rcG-CSF), the elevation in neutrophil counts in response to treatment occurred within 24 hours, with a peak on the 19th day of treatment (7). It is unknown if the response of a hypocellular bone marrow due to severe chronic CME, as in the present case, should be compared with normal healthy marrow. However the mega-dose pharmacological effect could preserve the low numbers of myeloid stem cells in the bone marrow and promote their proliferation. It has been shown that the long-term use of rhG-CSF in dogs led to a subsequent decrease in neutrophil numbers, presumably as a result of antibody formation to this drug (7). Thus, it was recommended that its use in dogs should be limited to short-term therapy (7). As rcG-CSF was unavailable commercially, we decided to use rhG-CSF. The treatment with rhG-CSF did not lead to an elevation in neutrophil numbers throughout the treatment period (Table 1), although it might have aided in the resolution of bronchopneumonia by increasing neutrophil function. It was decided to continue rhG-CSF treatment, at a lower dose, with concurrent prednisone therapy given in immunosuppressive doses. The high glucocorticoid levels may have partially prevented the destruction of rhG-CSF by antibodies. Immunosuppression induced by high prednisone doses may also have been partially prevented by concurrent use of rhG-CSF, by improving neutrophil function. 

The use of rhEPO in veterinary medicine is increasing, mainly in erythropoietin deficiency conditions (such as end stage renal disease) (8), but has not been reported in severe chronic CME. In the present case, the total rhEPO treatment period was 116 days. The first response was observed 3 days after initiation of treatment, evidenced by mild polychromasia in the peripheral blood smear and macrocytosis. This reflects a prompt bone marrow response to EPO. The mean corpuscular volume (MCV) reached a peak of 89fl on day 64 (Table 1). During the first phase of treatment severe hypochromia developed {mean corpuscular hemoglobin concentration (MCHC) - 24.1 g/dl}, despite a high dose oral iron therapy. This complication of EPO treatment is well-documented (8). At this point the platelet count increased markedly from 13x103/mm3 to 241x103/mm3. This could have been due to severe depletion of body iron stores, as the number of circulating platelets was shown to be inversely related to iron stores (9). As the MCHC improved, platelet numbers decreased markedly to thrombocytopaenic levels. On day 95 despite a rather long (14 days) rhEPO treatment interval, the haematocrit had improved markedly to 32.2%, the MCV was still high (77fl) and peripheral blood smears revealed evidence of regeneration (table 1 & 2). We can assume that even with this relatively long dosing interval rhEPO still had a beneficial effect on erythropoiesis. When the MCV returned to normal, despite rhEPO therapy, the treatment was no longer considered effective, and was discontinued. It is possible that immunosuppressive doses of glucocorticoids may have delayed the production of anti-rhEPO antibodies and its destruction, and thus allowed a long beneficial effect of EPO.

Thrombocytopaenia was the last haematological abnormality to resolve in this case. It is possible that correction of the anemia and a gradual decrease in circulating anti-platelet antibodies and complement concentrations, as a result of the long high-dose prednisone treatment, allowed increased platelet production with a decline in consumption. Thrombopoietic growth factors are commercially unavailable at present, and may prove to be important in the treatment of thrombocytopaenia secondary to bone marrow hypocelullarity. Induction of platelet release with vincristine-sulphate treatment has been documented previously (10). Its effect depends on the presence of satisfactory numbers of megakariocytes in the bone marrow (11). In the present case, this mode of treatment probably had a mild effect on platelet numbers. The platelet numbers increased from 3x103/mm3 to 21x103/mm3 after one treatment, and to 36x103/mm3 following the second treatment. Thereafter the platelet numbers continued to increase gradually with no additional vincristine therapy over a long period.

The pathogenesis of CME has been proposed to involve immunologic mechanisms, mediated mainly by the monocyte-macrophage system and the production of excess antibodies by plasma cells (1,3,5,6). The use of glucocorticoids in immunosuppressive doses seems rational to overcome the excessive aberrant immune response, and has been suggested previously in severely sick animals in the acute phase of CME (3,5,12). By the same rationale, this treatment would appear to be appropriate in the chronic phase of CME, providing that bacterial infections are eliminated, a broad- spectrum antibiotic treatment is administered, and neutrophil function is maintained. The beneficial effects of glucocorticoids in CME may include the decrease in splenic destruction of blood cells and also a decrease in immunoglobulin, complement and cytokine production, thus allowing survival and proliferation of bone marrow stem cells and an increased life span of mature blood cells. A decrease in immunoglobulin production was probably the main mechanism responsible for the decrease in globulin concentrations and the improvement of the A/G ratio during treatment.

The initial E. canis antibody IFA titer was very high (1:5120, Table 2), however on day 476, when all treatments were discontinued, the E. canis IFA antibody titer was found to be negative (less then 1:20). Judging from the complete disappearance of E. canis antibodies and the resolution of all clinical and hematological abnormalities, we assumed that the typical prolonged humoral reaction has ceased, and that the parasite has been eliminated. It has been previously suggested that negative serological results and resolution of thrombocytopenia represent elimination of the rickettsia (13).

In conclusion, this paper describes the first successful treatment of severe pancytopenia secondary to chronic CME, caused by the Israeli strain of E. canis. The treatment included the use of the haematopoietic growth factors rhG-CSF and rhEPO, combined with immunosuppressive doses of glucocorticoids. The treatment described in the present case, although prolonged and expensive, may provide a different option for therapy of severely pancytopenic CME patients. Further investigations are needed to assess the efficacy of such therapy in a larger number of patients.

Acknowledgement

The authors thank Dr. Trevor Waner of the Israel Institute for Biological Research, Ness-Ziona for his professional assistance.

References