Treatment of Evans’ Syndrome With Human Intravenous Immunoglobulin and Leflunomide in a Diabetic Dog

Introduction

Evans’ syndrome (ES) is an uncommon, life-threatening hematological disease of dogs, characterized by the co-occurrence of immune-mediated hemolytic anemia (IMHA) and immune-mediated thrombocytopenia (IMT).¹ Diagnosis of IMHA is usually made based on the presence of anemia with a positive direct Coombs’ test or persistent autoagglutination and/or spherocytosis.² Because antiplatelet antibody assays have variable specificity and sensitivity and are not widely available, their use has limited clinical utility. The diagnosis of primary IMT is usually made on the basis of exclusion of other identifiable causes of thrombocytopenia and response to immunosuppressive therapy.^3–6

Primary canine ES carries a poor short-term prognosis, with a mortality rate of up to 80%. Immunosuppressive doses of glucocorticoids have been the mainstay of treatment.¹ Furthermore, long-term use of glucocorticoids often leads to adverse systemic effects, such as iatrogenic hyperadrenocorticism, gastrointestinal hemorrhage, predisposition to bacterial or fungal infections, and occasional insulin-resistant diabetes mellitus.^7–9 For these reasons, combination therapy consisting of glucocorticoids and other immunomodulatory drugs or splenectomy in dogs with IMHA and/or IMT have been reported with variable clinical response.², ^10–12a

Human intravenous immunoglobulin (hIVIg) is a sterile and highly purified preparation containing all the immunoglobulin G (IgG) and trace amounts of immunoglobulin M and immunoglobulin A obtained from the healthy human donor population. Human IVIg has been safely and successfully used in many canine immune-mediated diseases, such as IMHA, IMT, nonregenerative anemia, drug-induced Stevens-Johnson syndrome, other adverse cutaneous drug reactions, and pemphigus foliaceus. ^13–20 Blockade of Fc receptors on mononuclear phagocytic cells has been proposed as one of the most likely mechanisms of action for the observed therapeutic response to hIVIg infusion in dogs.²¹

Anecdotal evidence has been presented for the efficacy of leflunomide as a second-line and primarily adjunctive therapy in dogs with different immune-mediated disorders, including one dog with relapsing ES.²² The mechanism of action of leflunomide is not completely understood. Leflunomide is converted by hydrolysis in the plasma and intestinal tract mucosa to its primary metabolite, a malononitriloamide (A771726), which reversibly inhibits dihydro-orotate dehydrogenase, the rate-limiting enzyme in de novo pyrimidine synthesis. Unlike most other cells, lymphocytes need both the salvage and the de novo pathways to meet the pyrimidine demand for their activation and proliferation. At high concentration, A771726 also inhibits protein cytokine and growth factor-receptor- associated tyrosine kinase activity involved in T-cell activation. As a result, these are the leading hypotheses for the specific inhibitory effects of leflunomide on activated T- and B-cell proliferation and immunoglobulin production, while other cells maintain their basal cell division.^22,23

The purpose of this report is to describe the successful treatment of ES in a diabetic dog using a combination of hIVIg and leflunomide as first-line immunomodulatory therapy, without the use of glucocorticoids.

Case Report

An 11-year-old, spayed female miniature schnauzer weighing 7.6 kg was evaluated for an acute onset of weakness, lethargy, and decreased appetite of 2 days’ duration. No known exposure to toxin or drugs other than insulin was reported, and no travel had occurred outside of Minnesota. The dog was diagnosed with diabetes mellitus 2 years prior to presentation, and the diabetes was well controlled using 7 units of NPH^b insulin subcutaneously (SC) twice daily. The dog also had bilateral phacoemulsification to correct diabetes- induced cataracts 8 months before presentation. Vaccination for canine distemper virus, adenovirus (type II), parainfluenza, parvovirus, and rabies occurred 5 months before presentation. The dog was routinely treated with heartworm and flea and tick preventative medications during the six warmer months of the year, from May to October.

Pale mucous membranes, numerous petechiae on the oral mucosa, and large ecchymotic lesions on the ventral abdomen were detected on physical examination. A grade II/VI holosystolic murmur was ausculted at the left heart base. A complete blood count (CBC) detected severe thrombocytopenia (2000 platelets/μL; reference range 160,000 to 425,000/μL) and severe normocytic, normochromic, nonregenerative (0.003 reticulocytes/μL) anemia (hematocrit 12%; reference range 38% to 57%). Evaluation of red blood cell (RBC) morphology indicated the presence of spherocytosis. A slide agglutination test was negative, and a direct Coombs’ test was positive for IgG (antibody titer 1:256). A serum biochemical profile disclosed hyperbilirubinemia (0.5 mg/dL; reference range 0 to 0.3 mg/dL), hypercholesterolemia (420 mg/dL; reference range 143 to 373 mg/dL), hyperglycemia (128 mg/dL; reference range 75 to 117 mg/dL), and high activity of serum alkaline phosphatase (259 U/L; reference range 8 to 139 U/L).

Results of a coagulation profile (including measurements of fibrinogen level, prothrombin time, activated partial thromboplastin time, and fibrin degradation products) were within reference limits. Thoracic radiography was unremarkable, while abdominal radiography and ultrasonography detected splenomegaly. Serological testing for Ehrlichia canis, Babesia canis, Rickettsia rickettsii, and Anaplasma phagocytophilum was negative. Cytological evaluation of a bone marrow aspirate and histological evaluation of a core bone marrow biopsy specimen indicated moderate megakaryocytic and erythroid hyperplasia.

Based on these results, a presumptive diagnosis of ES was made. A cross-matched packed red blood cell (pRBC) transfusion (10 mL/kg intravenously [IV]) was administered. Because of the potential adverse effects of immuno-suppressive doses of glucocorticoids in a diabetic patient, hIVIg^c and leflunomide^d were used as first-line immunomodulatory therapy after informed owner consent was received.

Doxycycline^e was used at a dosage of 5 mg/kg by mouth (PO) q 12 hours (pending vector-borne disease serological testing). Leflunomide was initiated at a dosage of 2 mg/kg PO q 12 hours (based on the nontoxic dosage of 4 mg/kg per day used in studies of the canine renal transplantation model).²⁴ Additionally, a single infusion of hIVIg was administered at a dosage of 1.3 g/kg in a 5% solution over 8 hours. Twenty minutes prior to the hIVIg infusion, the dog was treated with diphenhydramine^f (1 mg/kg SC), and the rectal temperature, heart rate, and respiratory rate were monitored every 10 minutes for the first 40 minutes of the infusion.

No adverse reactions were noted during or after the infusion. A platelet count immediately after the infusion was 51,000/μL, and 24 hours later it was 116,000/μL. At this time, a cystocentesis and an ultrasound-guided fine-needle aspiration of liver and spleen were performed without complications. An ultra-low dosage of aspirin^g (0.5 mg/kg PO q 24 hours) was also initiated in an attempt to prevent thromboembolic disease. Cytological evaluation of hepatic and splenic aspirates indicated moderate extramedullary hematopoiesis with moderate cholestasis, and marked extramedullary hematopoiesis with moderately reactive lymphoid hyperplasia, respectively. Urinalysis disclosed a urine specific gravity of 1.023 (reference range 1.001 to 1.070), and no bacterial growth was seen on urine culture. Over the subsequent 8 days, three cross-matched pRBC transfusions were administered, and daily CBCs detected platelet counts within the reference range. On day 9, the hematocrit stabilized at 27%, the spherocytosis had completely resolved, and the dog was discharged from the hospital.

One week later, the owner reported the dog had progressive lethargy and decreased appetite since the discharge. A physical examination revealed resolution of petechiae and ecchymotic lesions, but mucous membranes were pale. A CBC detected mild thrombocytopenia (136,000 platelets/μL) and moderate, normocytic, normochromic anemia (hematocrit 19%). Evaluation of RBC morphology disclosed occasional spherocytes. A serum biochemical profile did not reveal any other significant changes compared to the initial clinicopathological findings. At that time, a cross-matched pRBC transfusion (10 mL/kg IV) was administered, the aspirin was discontinued, the leflunomide was continued as previously prescribed, and the dog was discharged at completion of transfusion.

Ten days later, the owner reported an improved appetite and physical activity level in the dog. A physical examination revealed diffuse hair thinning and mild symmetrical alopecia along the dorsum; the dog was not pruritic. Serum total thyroxine concentration, serum thyroid-stimulating hormone, and results of an adrenocorticotropic hormone stimulation test were within reference limits. At that time, a CBC was unremarkable, and the leflunomide dosage was decreased by 25%. Dermatological abnormalities completely resolved after 12 weeks.

A CBC and serum biochemical profile were also performed every 2 weeks for the following 2 months, and then every 4 weeks for another 6 months. The platelet count and the hematocrit were within reference ranges, and the serum biochemical profile did not reveal any other significant changes compared to the initial clinicopathological findings during that time. A leflunomide 12-hour serum trough level, measured at a commercial laboratory^h via high-performance liquid chromatography, was >20 μg/mL at months 1, 2, 3, and 4 (260 μg/mL, 170 μg/mL, 90 μg/mL, and 40 μg/mL, respectively). Leflunomide dosage was decreased by 25% every 4 weeks for the first 4 months in order to achieve a through level of approximately 20 μg/mL (leflunomide trough levels were based on studies of the canine renal transplantation model);²⁴ then the dosage was decreased every 8 weeks until discontinuation after 10 months of therapy. Aurine culture and a 12-hour blood glucose curve were also performed at 4, 10, and 19 months. Urine cultures were negative, and the dog has remained a well-regulated diabetic using 7 units of NPH insulin SC twice daily (as given prior to the presentation for ES).

Discussion

This dog was suspected to have had severe anemia and thrombocytopenia attributable to a primary immune-mediated mechanism, on the basis of the following criteria: presence of spherocytosis on blood smear and positive direct Coombs’ test; increased numbers of erythroid cells and megakaryocytes demonstrable in bone marrow aspirates and core biopsies; exclusion of other causes of anemia and severe thrombocytopenia, such as neoplasia, disseminated intravascular coagulation, and vector-borne diseases; and response to immunomodulatory therapy.

In the present case, this dog was an insulin-dependent diabetic for 2 years prior to presentation for ES. After informed owner consent was received, a single infusion of hIVIg and oral leflunomide were used to avoid severe adverse effects of immunosuppressive doses of glucocorticoids.

The serum half-life of hIVIg in healthy dogs is 7 to 9 days.¹⁴ In this dog, hIVIg was used to treat the acute phase of the disease; this seemed especially effective for the rapid platelet count recovery.

Immediate hypersensitivity reactions during hIVIg infusion in immunocompromised veterinary patients are uncommon; only one dog has been reported to vomit once during a hIVIg infusion.¹⁷ In previous reports, only a single infusion of hIVIg was recommended in dogs because of possible antibody-mediated anaphylactic shock and type III hypersensitivity reactions on subsequent administration.^15–17 Recently, a dog with severe pemphigus foliaceus was treated with prednisone, azathioprine, and multiple doses of hIVIg over several months without evidence of immediate and delayed hypersensitivity reactions.²⁰ One might hypothesize that the immunosuppression of that dog did not predispose to develop antihuman immunoglobulin antibodies. In the present case, no immediate or delayed adverse reactions attributable to hIVIg infusion were observed at any time, but antihuman immunoglobulin antibodies were not measured. In a previous study, evidence of thromboembolic disease was detected in 50% of dogs with IMHA after treatment with hIVIg.¹⁴ This is a common complication in dogs with IMHA, regardless of the type of immunomodulatory therapy used,²⁵ and the role of hIVIg administration in thromboembolic events is not completely understood.¹⁷ This complication was not encountered in the present case. Safety of treatment with hIVIg needs to be more clearly established before this therapy is routinely recommended.

The active metabolite of leflunomide, malononitriloamide A771726, has a long half-life (11 to 16 days) in humans, but canine T and B cells seem to be very sensitive to the antiproliferative effects of all the malononitriloamides. ²³ In contrast to hIVIg, leflunomide appears to have a very safe profile, and it was used as chronic therapy with long-term administration. A full clinical response was noted after 4 weeks of treatment.

In a previous study, side effects of leflunomide therapy in 26 dogs were rare and mild, including self-limited hematemesis and/or hematochezia in three dogs, decreased appetite in two dogs, and lethargy and mild anemia in one dog each. Most of the dogs in that study were concurrently receiving other immunosuppressive agents, which may have contributed to some of these adverse effects.²² Decreased appetite was also noted in the present case during the first weeks of therapy. Because of concurrent aspirin administration and persistent anemia, it is difficult to determine the contribution of leflunomide to this clinical sign. Furthermore, diffuse hair thinning and mild symmetrical alopecia along the dorsum were noted 4 weeks into therapy. Dermatological disorders in dogs with diabetes mellitus have been described.²⁶ In the present report, the dog was a well-regulated diabetic over the entire course of the ES, and concurrent endocrinopathy (such as hyperadrenocorticism or hypothyroidism) was ruled out. A thorough dermatological evaluation was not performed in this case, but dermatological abnormalities resolved completely without specific intervention after 16 weeks of therapy, when the leflunomide dosage was reduced to <2 mg/kg per day.

Availability and cost are important considerations when using hIVIg and leflunomide in veterinary patients. The costs of hIVIg ($1,000.00 for a single infusion at 1.3 g/kg in this 7.6-kg dogⁱ) and leflunomide ($350.00 per month for this 7.6-kg dog at a dosage of 4 mg/kg per dayⁱ) are significantly higher when compared to prednisone ($12 per month for this 7.6-kg dog at a dosage of 2.2 mg/kg per dayⁱ).

The case presented here suggests that hIVIg and leflunomide may be beneficial as alternative immunomodulatory therapy for dogs with ES when glucocorticoids are contraindicated.

Conclusion

To the authors’ knowledge, this is the first report of a diabetic dog with ES successfully treated with a combination of hIVIg and leflunomide as first-line immunomodulatory therapy, without the use of glucocorticoids. This treatment was well tolerated and did not negatively impact the short-and long-term management of diabetes mellitus in this animal. Prospective, randomized clinical trials comparing the use of hIVIg and leflunomide versus glucocorticoids in dogs with ES are necessary to assess safety and efficacy of this combination protocol before it is routinely recommended.