Primary Immune-mediated Thrombocytopenia in 30 Dogs (1997–2003)
During a 6-year period, primary (idiopathic) immune-mediated thrombocytopenia was retrospectively evaluated in 30 dogs. Ages of the dogs ranged from 3 months to 10 years (median 4 years); female dogs were markedly overrepresented with 73%. Clinical examination revealed hemorrhages in 70% of the dogs. Platelet counts ranged from 0 to 111,000/μL (median 8000/μL); 77% of the dogs had platelet counts <30,000/μL. Seventeen dogs were anemic (hematocrit 9% to 36%; median 31%). Immunosuppressive therapy was performed in all but one dog. The recurrence rate of 19 dogs that were followed over an extended period (112 to 1684 days; median 340 days) was 26%. Twenty-nine (97%) dogs survived 14 days after initial presentation, and 27 (93%) dogs survived at least the following 15 to 1684 days (median 220 days).
Introduction
Immune-mediated thrombocytopenia (IMT) is a disease in which antibodies bound to the surface of platelets cause premature platelet destruction by macrophages. Platelet-bound antibodies can be antiplatelet autoantibodies, but they also can be immune complexes bound to the platelet membrane, antibodies bound to platelet antigens altered during the course of disease, or antibodies bound to foreign antigens adsorbed to the platelet surface.1,2 The glycoproteins IIb and IIIa have been identified as target antigens in canine idiopathic thrombocytopenic purpura.3
Immune-mediated thrombocytopenia can be differentiated as primary and secondary forms. Various infectious or neoplastic diseases (e.g., ehrlichiosis, babesiosis, lymphoma, hemangiosarcoma) as well as drugs (e.g., trimethoprim/sulfadiazine, phenylbutazone) may trigger the production of antibodies causing a secondary IMT (sIMT).2,4–6 Immune-mediated thrombocytopenia in the absence of other identifiable diseases is referred to as primary or idiopathic IMT (pIMT). The diagnosis is based on exclusion of known causes or underlying diseases.1,7
Thrombocytopenia is often the only abnormal laboratory result in pIMT; in the case of hemorrhage, an anemia may also occur.8 An increased bleeding tendency can be observed in dogs with platelet counts <30,000 to 50,000/μL.9,10 In thrombocytopenic dogs, mainly surface bleedings occur, such as petechiae, ecchymoses, melena, retinal hemorrhages, epistaxis, hematuria, hematemesis, and hemoptysis.9–13
Glucocorticoids are the initial therapy of choice for dogs with pIMT. Adjunctive therapies (including vincristine, azathioprine, cyclosporine, cyclophosphamide, danazol, human immunoglobulin G, and splenectomy) have been used in conjunction with glucocorticoids to treat dogs with IMT.1,10,12,14–16 In dogs with IMT, the transfusion of platelets is not recommended because of the short life span of transfused platelets. However, in cases of uncontrollable or life-threatening bleeding, platelet transfusions should be initiated.17,18
Only three older publications have reported the long-term follow-up in dogs with IMT, and only one of these studies dealt with the course of disease solely for dogs with primary IMT.10,12,14 Based on these studies, about 30% of dogs with IMT die, are euthanized during the acute phase of disease, or die because of relapses in the follow-up period. The recurrence rate is approximately 40%.1
The aim of this study is to analyze the medical records of dogs with pIMT that were presented within a period of 6 years (January 1997 to January 2003). Besides clinical findings, diagnostics, and therapy, particular emphasis is placed on the evaluation of the long-term course of disease. The results are compared with the literature.
Materials and Methods
Medical records of dogs seen at the Clinic of Small Animals, Free University of Berlin, between January 1997 and January 2003 were retrospectively reviewed in order to search for dogs with pIMT.
Inclusion criteria were complete medical records, platelet counts <150,000/μL, and no evidence of any underlying disease or any cause that might have triggered IMT, such as recent vaccination or drug exposure. To rule out underlying diseases, a complete diagnostic workup was performed, including a complete blood count (CBC), blood smear evaluation, clinical biochemical analysis, diagnostic imaging, and serological as well as immunological testing. In order to have evidence for an immune-mediated process, a platelet-bound antibody test had to be positive. Dogs were excluded from the study if identifiable underlying diseases were present, if recent exposure to drugs occurred that might trigger IMT, or a vaccination was administered in the 4 weeks before presentation. Dogs with an incomplete medical record were also excluded from the study. An incomplete medical record was counted as such when certain examinations were not performed in a dog (e.g., if no result of ehrlichia serology was available, and the dog had been in an endemic area; or if no abdominal ultrasonography had been performed).
Data collected from the medical records included signalment, history, clinical signs, and diagnostic imaging findings (e.g., thoracic and abdominal radiography, abdominal ultrasonography). A CBC and clinical biochemical analysis were performed in all dogs on the day of initial presentation.a,b In all cases, a blood smear evaluation was performed to confirm the low platelet counts. In cases with no sign of hemorrhage or with thrombocytopenia as an incidental finding, a second blood sample was taken.
All blood samples were macroscopically and microscopically evaluated for erythrocytic agglutination. In case of a spontaneous erythrocytic agglutination, the red blood cells (RBCs) were washed three times with physiological saline solution, and the macroscopic and microscopic evaluations for RBC agglutination were repeated. In 24 cases, manual differential cell counts were performed. For reticulocyte counting, brilliant cresyl blue smears were prepared. Coagulation testing (i.e., prothrombin time [PT] and activated partial thromboplastin time [aPTT]) was performed in 27 dogs.c The citrate plasma samples of dogs that were presented to the emergency service of the clinic were frozen at −4°F to be analyzed the next morning.
Further examinations included urinalysis (n=9), bone marrow aspirations (n=5), antinuclear antibody (ANA) titers (n=3), and infectious disease titers (e.g., Babesia canis, Ehrlichia canis, Leishmania, Dirofilaria immitis, Borrelia) (n=25). All of the aforementioned infectious disease titers were not carried out on all of the dogs. The choice of infectious disease titers varied from dog to dog and was based on clinician preference and whether the dog had been in an endemic area. The tests were performed in external laboratories. Dirofilaria immitis was tested by use of an enzyme-linked immunosorbent assay (ELISA), and the other infectious agents were analyzed by an immunofluorescence test.
The direct Coombs’ test (n=15) and the platelet-bound antibody test (n=30) were performed at the Immunology Unit, University of Veterinary Medicine, Hanover, Germany.4,19 Flow cytometric analysis was used to test for platelet-bound antibodies.d Separated and washed platelets were labeled by a platelet-specific monoclonal mouse antibody, followed by phycoerythrin-conjugated goat anti-mouse immunoglobulin G antibodies. Thus, particles having the same size as platelets (such as fragments of RBC) could be distinguished from platelets and excluded from the analysis. Fluorescein-conjugated polyclonal goat antidog immunoglobulin G antibodies (heavy and light chains) were used to detect those platelets that were carrying antibodies. Along with each blood sample (approximately 2 to 5 mL of ethylenediaminetetraacetic acid [EDTA] blood, taken on initial examination) of a dog, a control sample (EDTA blood of healthy dogs) was sent overnight without cooling and was analyzed simultaneously (that is, within 16 to 24 hours after the blood sample was taken, in order to avoid artefacts mimicking false-positive platelet reactions). Platelet-bound antibody testing was negative in 30 control dogs.
Initially, all dogs except one were treated with prednisolone (1 to 1.5 mg/kg body weight per os [PO] q 12 hours). If the platelet counts were in the reference range (150,000/μL to 500,000/μL), the dosage of prednisolone was reduced gradually, approximately every 2 weeks by one-fourth to one-fifth of the former dosage. The cytostatic agents (i.e., azathioprine, vincristine, or cyclosporine) were administered in combination with prednisolone either at the beginning of therapy, during treatment if the platelet count did not rise, or after a relapse. Azathioprine administration was 1.5 to 2 mg/kg body weight PO q 24 hours; vincristine was given at 0.02 mg/kg body weight intravenously (IV); and cyclosporine was given at 5 mg/kg body weight PO q 24 hours.
The dogs were divided into different treatment groups, depending on the type of immunosuppressive drugs they received. Furthermore, treatment consisted of the administration of H2-receptor antagonists (i.e., cimetidine, ranitidine), sucralfate, infusions of lactated Ringer’s solution, and antibiotics (i.e., doxycycline, amoxicillin with clavulanic acid, and/or enrofloxacin). In cases of bleeding and severe thrombocytopenia (platelet count <20,000/μL) and anemia (hematocrit [Hct] <15% to 20%), or if a rapid further Hct decrease was expected because of hemorrhage, the dogs were transfused with dog erythrocyte antigen 1.1-compatible fresh whole blood. Some dogs received packed RBCs in addition.
A response to therapy (complete remission) was defined as an increase of the platelet count to =150,000/μL. The platelet count was measured daily at the beginning of therapy, then every 2 to 3 days. If the number of platelets stabilized (>100,000/μL), then rechecks were carried out every 1 to 2 weeks and then every 4 weeks. A relapse was defined as a decrease of the platelet count to <150,000/μL after the value had already been within reference range. For the analysis of the recurrence rate, the only dogs considered were those that had been observed over a period of at least 90 days. Furthermore, the survival rates of dogs in the acute phase of disease (within the first 14 days) and in the following time were analyzed.
As descriptive statistics, the values of median as well as minimum and maximum were calculated; in some cases, average and standard deviation were calculated in addition. In order to compare two groups, cross tabulations were prepared and analyzed by use of the chi-square (χ2) test after Pearson’s correlation coefficient. A difference between groups was considered significant if the P value was <0.05. For comparison with other studies in regard to relapse and mortality rate, binomial confidence intervals (confidence coefficient 95%) were calculated. Statistical evaluations were performed with a standard statistical software program.e
Results
A total of 106 thrombocytopenic dogs tested positive for platelet-bound antibodies. Due to a lack of certain examinations, 17 dogs were not evaluated further. The other 89 dogs were divided into the following groups of disease: pIMT (n=30), Evans’ syndrome (n=14), sIMT (n=36), and sIMT with secondary immune-mediated hemolytic anemia (n=9).
The following diseases constituted the diagnosis of sIMT: neoplasia (n=8; [lymphoma n=3, hemangiosarcoma n=1, tumor of the liver n=1, tumor of the spleen n=1, leukemia n=1, mammary gland tumor n=1]); infectious diseases (n=23; [ehrlichiosis n=7, leptospirosis n=2, borreliosis n=1, prostatitis n=1, endometritis n=1, abscess n=1, bite wound infection n=1, pyometra n=1, wound infection n=1, babesiosis n=4, babesiosis/leishmaniosis n=2, babesiosis/ehrlichiosis n=1]); and other causes (n=5; [blood transfusion n=1, systemic lupus erythematosus n=1, drugs n=2, cause of illness unknown n=1]).
Thirty dogs with pIMT were included in the study. The ages of the dogs ranged from 3 months to 10 years (median 4 years). Twenty-two dogs were =6 years of age. The female dogs (73%; 22/30) were markedly overrepresented [Figure 1]. Eighteen different breeds were included, with two dogs each for the following breeds: rottweiler, cocker spaniel, golden retriever, and German shepherd dog. Furthermore, six mixed-breed dogs were included.
Twenty-two owners presented their dogs because of the following hemorrhages: gingival bleeding (n=13), skin and mucosal petechiae (n=10), hematochezia/melena (n=6), epistaxis (n=4), hematoma (n=2), hematuria (n=1), conjunctival hemorrhage (n=1), and hyphema (n=1). For the other eight dogs, CBCs were performed because of lameness (n=5), skin problems (n=2 with atopy), or a routine health check (n=1). Thrombocytopenia was a coincidental finding. Based on the history, the duration of disease varied between 1 and 30 days (median 2 days); nine dogs had travelled abroad. None of the dogs was vaccinated <4 weeks before first signs of disease. Twelve dogs had been pretreated with the following: glucocorticoids (n=2), vitamin K (n=8), and antibiotics (e.g., enrofloxacin, amoxicillin, penicillin, doxycycline; n=5).
In 70% (21/30) of the dogs, clinical examinations revealed the following hemorrhages: skin and/or mucosal petechiae (n=13), gingival bleeding (n=9), melena (n=6), ecchymoses (n=5), hyphema (n=4), epistaxis (n=3), hematochezia (n=2), hematoma (n=2), hematuria (n=2), and conjunctival hemorrhage (n=1). Four dogs exhibited a prolonged capillary refill time. Seven dogs had an increased rectal temperature (102.4°F to 104.5°F; median 103.1°F). Four of these dogs had a temperature above 103°F (103.1°F to 104.5°F; median 103.3°F). A slight general or local enlargement of lymph nodes was detected in two and three dogs, respectively. Three dogs had a tense but nonpainful abdomen. Conspicuous radiographic and ultrasonographic findings were splenomegaly (n=16), hepatomegaly (n=1), and hepatosplenomegaly (n=2).
On the day of admission, the platelet counts (reference range 150,000 to 500,000/μL) ranged from 0 to 111,000/μL (median 8000/μL) [Figure 2]. Twenty-three (77%) dogs had platelet counts <30,000/μL. In 19 of these 23 dogs, hemorrhages were detected during clinical examination. Two dogs with platelet counts >30,000/μL (33,000/μL and 73,000/μL) had hemorrhages. According to the χ2 test (P=0.014), platelet counts =30,000/μL were positively correlated with the occurrence of spontaneous bleeding.
The initial Hct (reference range 38% to 57%) was 9% to 61% (median 35%); 17 dogs were anemic (Hct 9% to 36%; median 31%). In 14 of the anemic dogs, hemorrhages were detected during clinical examination. Three dogs with mild anemia (Hct 34% [n=1] and 36% [n=2]) had no obvious hemorrhages. The RBC count (reference range 5.6 to 8.6 × 106/μL) varied between 1.5 and 9.6 × 106/μL (median 5.8 × 106/μL), and the hemoglobin concentration (reference range 13.5 to 20.0 g/dL) varied between 3.2 and 20.8 g/dL (median 12.1 g/dL). In nine dogs, the reticulocyte counts ranged between 19,800 and 210,080/μL (median 110,860/μL). Initially, five dogs had a regenerative anemia (reticulocyte counts >60,000/μL), and four had a nonregenerative anemia. The leukocytes (reference range 6000 to 15,000/μL) were in a range between 5000/μL and 33,800/μL (median 13,400/μL). A leukocytosis with leukocyte counts >15,000/μL (15,200 to 33,800/μL; median 19,000/μL) was found in 12 (40%) dogs. The differential cell count in 24 of the 30 dogs revealed monocytosis (n=15), eosinopenia (n=11), neutrophilia without left shift (n=10), lymphopenia (n=7), neutrophilia with left shift (n=6), and eosinophilia (n=4). In three dogs, a mild erythrocytic agglutination was present, which did not persist after washing the erythrocytes with physiological saline.
A coagulation panel was performed in 27 dogs. In all dogs, PT was in the reference range (15 to 20 seconds), and aPTT was in the reference range (14 to 18.5 seconds) in 25 dogs. Two dogs had a mildly prolonged aPTT (19.0 and 19.6 seconds). Clinical biochemical analysis revealed hypoproteinemia and hypoalbuminemia in five dogs, which was attributable to hemorrhages. One dog had an azotemia (a urinalysis was not available). In 15 dogs, slightly increased serum liver enzymes were detected; 10 of these dogs had isolated increased alkaline phosphatase activity. Two of these 15 dogs were 3 and 5 months old; three of the dogs had been pretreated with glucocorticoids. Urinalyses were performed in nine dogs. Erythrocytes were present in four samples: + (n=3) and ++ (n=1). A hemoglobinuria was found in two dogs, and a bilirubinuria was found in one dog.
The bone marrow was examined in five dogs before they received vincristine. Four of them had numerous megakaryocytes, and in one dog only a few megakaryocytes were found (suggesting a decreased megakaryocyte density). A core biopsy was not performed in this dog. For this dog, the time needed for a rise in the platelet counts to >150,000/μL was 13 days. The other dogs had platelet counts within the reference range 3 to 7 days (median 5 days) after vincristine was administered.
In 25 dogs, serological tests were performed for the following infectious diseases: Babesia canis (n=20), Ehrlichia canis (n=24), Leishmania (n=9), Dirofilaria immitis (n=3), and Borrelia (n=3). All results were negative. In three dogs, ANA titers were negative. A direct Coombs’ test was performed in 15 dogs; 11 of these dogs were anemic. The results were negative in all cases.
After 6 to 26 days (median 10 days), the platelet-bound antibody test was repeated in 13 dogs. Results were negative in 12 dogs. One dog, however, still had a positive test result after 10 days of immunosuppressive therapy (platelet count 134,000/μL).
Ninety-seven percent (29/30) of the dogs were treated with prednisolone. The cytostatic agents azathioprine, vincristine, or cyclosporine were given at the beginning of therapy (n=7) or 5 to 43 days (median 9 days) later (n=5) in the following combinations: prednisolone/azathioprine (n=3), prednisolone/vincristine (n=6), and prednisolone/cyclosporine (n=3). Seventeen dogs received only prednisolone, and one dog was not treated. Four dogs received cytostatic agents in addition to prednisolone after a relapse.
Due to severe thrombocytopenia, hemorrhages, and/or a poor general condition, 20 dogs were hospitalized. For the first 10 to 14 days, 27 dogs were treated with antibiotics (15 dogs received doxycycline, 10 received amoxicillin and clavulanic acid, and four received enrofloxacin). An H2 receptor-antagonist was administered to 24 dogs; in addition, five dogs received sucralfate. Nineteen dogs were infused with electrolyte solutions (lactated Ringer’s solution) because of hypovolemia or bleeding. Eleven dogs with acute blood loss and anemia were transfused one to three times over a 1- to 4-day period (median 2 days). At the time of blood transfusion, the Hct was =25% in all dogs, and the platelet count was =20,000/μL. All dogs received fresh whole blood (7.1 to 25.0 mL/kg body weight per transfusion; median 14.2 mL/kg; n=16 transfusions), and five dogs received packed RBCs (3.5 to 12.5 mL/kg body weight per transfusion; median 10.4 mL/kg; n=5 transfusions) in addition because of anemia.
In the present study, a total of 30 dogs were observed over a period of 5 to 1684 days (median 210 days). In 24 (96%) of 25 dogs with platelet counts <50,000/μL, an increase of the platelet counts to >50,000/μL was detected after 1 to 15 days (median 5 days). One dog did not reach this value during the first 5 days (the dog was euthanized).
After a period of 4 to 112 days (median 10 days), 90% (27/30) reached a platelet count of >150,000/μL. In one dog, the platelet count remained between 100,000/μL and 150,000/μL during a time period of 274 days. Another dog could only be followed for 21 days; on day 21, the platelet count was 90,000/μL. The third dog was euthanized at the owner’s request after 5 days of treatment (platelet count 15,800/μL). In the dog that was not treated, the platelet count increased from 47,000/μL to 112,000/μL within 1 day. Further measurements revealed platelet counts in the reference range (observation period 112 days).
In the prednisolone therapy group, 14 dogs (with platelet counts <50,000/μL; n=15) reached platelet counts =50,000/μL after 4 to 11 days (mean ± standard deviation [SD] 6±2.2 days; median 5 days). In the prednisolone/ vincristine therapy group, platelet counts in six dogs rose to =50,000/μL after a period of 2 to 7 days (mean ± SD 4±2 days; median 4 days). In the prednisolone/cyclosporine group, platelet counts increased to =50,000/μL in two dogs after 5 and 9 days. The observation period for these two dogs regarding the cyclosporine treatment was only over a period of 13 and 58 days. The third dog was treated with cyclosporine in addition to prednisolone on the 43rd day of treatment, as the platelet count did not increase. This dog never reached a platelet count >150,000/μL. The three dogs in the prednisolone/azathioprine therapy group reached a platelet count =50,000/μL after 4, 7, and 12 days.
Of 19 dogs that were observed over an extended period (112 to 1684 days; median 340 days), 26% (n=5; confidence interval 9% to 51%) had recurrence of thrombocytopenia. The first relapse was noted after a period of 19 to 286 days (median 66 days). These five dogs had one to four relapses (median two relapses) during their observation period. Most relapses (n=5) occurred after reduction of the prednisolone dosage. The prednisolone dosage was decreased by 0.3 to 0.5 mg/kg body weight per day (median 0.4 mg/kg) for 7 to 16 days (median 10.5 days) before recurrence of disease. At the time of recurrence, the dosage of prednisolone was 0.5 to 1.2 mg/kg body weight per day (median 1.0 mg/kg). Three relapses occurred after withdrawal of the medication by the owners. In three cases, the reason for relapse was unknown.
Ninety-seven percent (29/30) of the dogs survived within 14 days after first presentation. One dog developed neurological signs such as head tilt and ataxia during hospitalization; this dog was euthanized at the owner’s request after 5 days of treatment.
Ninety-three percent (27/29) of the dogs survived the following 15 to 1684 days (median 220 days). Six days after the beginning of azathioprine therapy, one dog developed pancreatitis with disseminated intravascular coagulation and was euthanized. Another dog died approximately 1 year after initial presentation, because acute gastrointestinal hemorrhage occurred after a relapse (platelet count 9200/μL). The mortality rate for the 30 dogs with pIMT was 10% (n=3; confidence interval 2% to 27%).
To be able to compare the different treatments in the individual dog, two groups were formed. One group consisted of dogs (n=17) that received prednisolone only, and the other group consisted of dogs (n=12) that received prednisolone and cytostatic agents. The only drugs considered were those administered at the beginning of treatment or a few days later if there was no increase in platelet counts. No significant difference regarding the relapse rate (P=0.676) and the mortality rate (P=0.367) was detected between the two groups.
After withdrawal of all drugs, eight (27%) dogs had stable platelet values for 49 to 1508 days (median 467 days). Sixteen dogs were administered prednisolone at the end of the observation period, and one dog received cyclosporine (observation period 10 to 686 days; median 72 days). One dog received no drugs, and the therapy of another could not be deduced from the files at the end of the observation period. The remaining three dogs either died or were euthanized.
Discussion
Several studies have dealt with canine IMT; however, only three older publications have evaluated the long-term follow-up in dogs with IMT.10,12,14 In one study, 54 dogs were followed for a period of 6 months to 8 years.10 Another study included 26 dogs over a period of 14 years.12 In these two studies, dogs with pIMT and dogs with sIMT were not classified separately. In another study, 15 dogs with pIMT were followed over a period of 2 years.14 In none of these three studies was a direct platelet-bound antibody test performed to show an immune-mediated basis for the disease. The present study observed the course of disease for a total of 30 dogs with pIMT over a period of 5 to 1684 days (median 210 days).
Immune-mediated thrombocytopenia is reported in dogs from 7 months to 14 years of age. The breeds most frequently mentioned are cocker spaniels, Old English sheep-dogs, and golden retrievers.9,10,12,14 Similar to other studies, a prevalence of IMT in rottweilers, cocker spaniels, golden retrievers, and German shepherd dogs was observed in the authors’ study. The female dogs were markedly over-represented (73%), which is comparable to results of former studies (69%; 65%).10,12
The most frequently described clinical signs in dogs with IMT are mucosal and cutaneous petechiae and ecchymoses, melena, hematuria, hyphema, epistaxis, retinal hemorrhage, hematemesis, pale mucous membranes, lethargy and weakness, and a systolic cardiac murmur.9–12 In one dog with pIMT, a hemothorax was diagnosed.20 In the authors’ study, hemorrhages were detected in 70% of the dogs during clinical examination; predominantly surface bleeding (87%) was observed. In other studies, surface bleedings were the most commonly described bleeding type in dogs with IMT (96%, 98%).9,10 Neurological signs may occur due to hemorrhages in the central nervous system (CNS),1,9 as the authors suspected in one of the dogs.
Fever, splenomegaly, and lymphadenopathy are uncommon in dogs with IMT.1 However, in several studies, fever was observed in dogs with IMT or pIMT (8/15 dogs, 2/13 dogs, 1/54 dogs).9–11 In the present study, seven dogs had an increased rectal temperature (>102.2°F). A slight enlargement of the lymph nodes (17%) and spleen and/or liver (63%) was observed in several dogs. Hepatosplenomegaly might be due to phagocytosis of platelets by macrophages of the spleen and liver.21,22
The following platelet counts in dogs with IMT were reported in other studies: <3000/μL to 60,000/μL, 500 to 66,000/μL, and <100,000/μL.9,11,12 An increased bleeding tendency is suspected if platelet counts are <30,000 to 50,000/μL.9,10 In the authors’ study, the platelets ranged between 0 and 111,000/μL. Seventy-seven percent of the dogs had platelet counts <30,000/μL. Eighty-three percent of the dogs with platelet counts <30,000/μL had hemorrhages. In two dogs with platelet counts >30,000/μL, bleedings were detected. In addition to thrombocytopenia, a platelet dysfunction might occur and might increase the bleeding tendency. In another study of 15 dogs with IMT, platelet function was examined. The platelet count ranged from 1000 to 68,000/μL. In 13 of the 15 dogs with IMT, maximal aggregation was significantly inhibited in response to adenosine diphosphate, thrombin, or collagen/epinephrine.23
Dogs with IMT might be anemic due to hemorrhages resulting from thrombocytopenia. In the present study, 17 (57%) of the dogs with pIMT were anemic. Fourteen of the anemic dogs were presented with hemorrhage. Other studies reported comparable results with 66% and 83%.9,10
Total and differential leukocyte counts are variable.24 In several studies, leukocytosis (18%; 53%) is reported in dogs with IMT. Leukopenia was a rare finding (9%; 7%) in dogs with IMT.9,11 In the manual differential cell count, neutrophilia with and without a left shift, monocytosis, lymphopenia, eosinophilia, and eosinopenia were found by the authors and in other studies.9–11 In the present study, leukocytosis was detected in 40% of the dogs.
In dogs with pIMT, PT and aPTT should be within the reference range as long as no disseminated intravascular coagulation complicates the disease.4,9,25 In the present study, two dogs had a slightly prolonged aPTT. As numerous dogs were presented at the emergency service of the clinic, some of the citrate plasma was frozen to be analyzed the next morning. Therefore, the slight prolongation of aPTT might be due to freezing and thawing of the samples. One study evaluated the effects upon PT and aPTT when canine and feline plasma samples were stored at −4°F and at room temperature. The study showed that aPTT of samples that were thawed and analyzed immediately was significantly prolonged compared to aPTT of fresh samples. On PT, however, the storage of samples at −4°F did not affect the results.26
Dogs with IMT mostly have normal or increased megakaryocyte counts in the bone marrow.10,27 Megakaryocytic hypoplasia or aplasia suggests a somewhat poorer prognosis and a more delayed response to appropriate therapy.28 In the present study, the bone marrow of five dogs was examined before vincristine was administered. In one of the dogs, only a few megakaryocytes were detected, and this dog needed 13 days to reach the reference range for platelets.
Serum antibody titers against infectious diseases were negative in all dogs of this study. In a recently published case report, a 10-year-old mixed-breed dog was infected with Anaplasma phagocytophilum and suffered from an immune-mediated thrombocytopenia and anemia.29 An infection with this agent cannot be completely ruled out in the dogs of the present study, as no serology or polymerase chain reaction studies for Anaplasma phagocytophilum were performed during the study period. However, granulocytic morulae were not found during blood smear evaluation.
In one study, the measurement of platelet surface-associated immunoglobulin G (PSAIgG) in a dog with pIMT was rechecked 1 month after therapy with prednisolone. The platelets returned to normal size, and PSAIgG levels decreased to within reference values.30 In the authors’ study, the platelet-bound antibody test was negative in 12 of 13 dogs after 6 to 26 days (median 10 days). In humans with IMT, a decrease in the amount of platelet-bound antibodies can be observed after treatment with glucocorticoids, cyclophosphamide, combination chemotherapy, and after splenectomy—suggesting that a major mechanism of action of these therapeutic maneuvers involves suppression of antibody production.31 If assays are performed after therapy is introduced, false-negative results may occur.32
Therapy for pIMT in 29 dogs consisted of treatment with prednisolone. Sixteen dogs were also treated with cytostatic agents. Treatment protocols were based on clinician preference. In other studies dealing with the course of disease, therapy for IMT also included glucocorticoids. Some dogs were additionally treated with cyclosporine, vincristine, azathioprine, or cyclophosphamide. In individual cases, a splenectomy was performed.10,12,14,16 In only one study is the administration of cyclosporine for dogs with IMT described. Four dogs with IMT received prednisone for 5 to 12 months prior to starting cyclosporine. Three dogs responded with normal platelet counts after 3 to 5 weeks, and one dog died of systemic aspergillosis.16
In the present study, three dogs were treated with cyclosporine in addition to prednisolone. Only two dogs responded to therapy. Since not enough dogs were treated with cyclosporine and since blood levels were not measured, the effectiveness of cyclosporine could not be determined.
In another study, the effect of prednisone alone, compared with a combination of prednisone and vincristine, on platelet counts in dogs with pIMT was evaluated. Dogs that received prednisone and vincristine had a significantly faster increase in platelet counts than did dogs that received prednisone alone.15 The efficacy of azathioprine in dogs with IMT is not documented.1 Also in the present study, no statement regarding azathioprine could be made, as the number of dogs (n=3) receiving it was too small.
In this study, all dogs with hemorrhages, with a significant thrombocytopenia (platelet count <20,000/μL), and with anemia (Hct <25%) were treated with fresh whole blood transfusions (n=11). Platelet transfusions are generally not recommended in dogs with IMT, unless uncontrolled or life-threatening bleeding occurs.17,18 In human medicine, platelet transfusions are recommended for IMT patients with severe mucosal bleeding or suspected CNS bleeding. The transfused platelets may be destroyed rapidly, but they protect the patient against catastrophic hemorrhage until specific therapy is beneficial.8,33
In one study, all dogs with pIMT reached a platelet count >50,000/μL within 2 to 9 days (median 4 days) after beginning immunosuppressive therapy.14 In a further study, the response time to reach a platelet count of >100,000/μL was 2 to 35 days (mean 8 days).12 In the authors’ study, similar results were achieved.
In one study, 54 dogs with IMT were monitored for 6 months to 8 years. A total of 43% (23/54) of the dogs had recurrence of disease. Eight of the 23 dogs had recurrences of one or more episodes of thrombocytopenia over periods of 2 to 8 months, but they had no further signs of IMT during follow-up periods of 1 to 5 years. Fifteen dogs experienced repeated recurrences over periods of up to 8 years.10 In another study, 47% (7/15) of dogs with pIMT had relapses. Five of the seven dogs were splenectomized, and withdrawal of medication was possible in four dogs after the surgery. One dog was euthanized. The two remaining dogs with relapsing IMT had been treated with medical therapy only, and they continued to require intermittent immuno-suppressive therapy over a period of 2 years.14 The recurrence rate in the authors’ study (26%; confidence interval 9% to 51%) did not differ from findings of other studies (43% to 47%).
During the first 14 days, 97% of the dogs survived. A possible reason for this high survival rate might be the intensive therapy with blood products. Ninety-three percent of the dogs survived the following 15 to 1684 days. In a retrospective study, a total of 23 (43%) of 54 dogs with IMT died or were euthanized during the course of disease. Eleven of the 23 dogs died during the first period of thrombocytopenia. Death was attributed to severe intestinal hemorrhage in seven dogs and to pneumonia in one dog. The cause of death was unknown in the three remaining dogs. Three dogs were euthanized during the first episode of IMT. Nine other dogs that repeatedly experienced relapses over periods of up to 8 years died or were euthanized.10
In another study, 29% of the dogs with IMT died or were euthanized because of IMT or related diseases. Examples of related diseases were hepatitis and systemic lupus erythematosus. Nineteen percent of the dogs died or were euthanized because of unrelated diseases.12
In an additional retrospective case series, 15 dogs with pIMT were described. Two dogs were euthanized after 9 and 12 months after diagnosis of IMT for problems unrelated to thrombocytopenia. In one dog, the severity of thrombocytopenia was not altered by splenectomy. The owner declined use of therapy with alkylating agents, and the dog was euthanized 3 months after surgery because of owner dissatisfaction with medical therapy (mortality rate 20%).14 In the authors’ study, the mortality rate amounted to 10% with a confidence interval of 2% to 27%. The mortality rate was comparable to the rate of one study (20%) and was distinctly lower than the rates of other studies (43% and 29%).10,12,14
The goal of treatment in dogs with IMT is to withdraw all therapy while maintaining remission.1 After withdrawal of all drugs, 27% of the dogs in the present study had stable platelet values. In other studies, 33% and 60% of the dogs remained in remission after withdrawal of drugs.12,14
Conclusion
Canine pIMT is a disease of good prognosis but with a high recurrence rate. The high survival rate in this study was attributed to the intensive medical care, including the use of blood products, in the acute phase of the disease. Especially important for a successful therapy is the cooperation of the owner, as the treatment is long term and needs to be controlled regularly.
Technicon H 1; Bayer, Leverkusen, Germany Cell Dyn 3500; Abbott, Wiesbaden-Delkenheim, Germany
Cobas Mira Plus; Roche Diagnostica, Grenzach-Wyhlen, Germany
Schnitger & Gross coagulometer; Amelung, Lemgo, Germany
FACScan; Becton-Dickinson, Heidelberg, Germany
SPSS, Version 12.0; SPSS GmbH Software, Munich, Germany
Acknowledgments
The authors thank Dr. R. Engelbrecht for the use of some patient records from his doctoral thesis.
Citation: Journal of the American Animal Hospital Association 44, 5; 10.5326/0440250
Citation: Journal of the American Animal Hospital Association 44, 5; 10.5326/0440250
Age and sex distributions of 30 dogs with primary immune-mediated thrombocytopenia.
Platelet counts (platelets/μL) of 30 dogs with primary immune-mediated thrombocytopenia on initial examination. (The box in the box plot includes values from the first quartile to the third quartile (25th to 75th percentile). The broad, black line in the box represents the median value. The highest and lowest values are depicted by thin, black lines. Outliers with a circle are values that lie between one and a half to three box lengths outside the box. Asterisks are values that lie more than three box lengths outside the box.
