• Open Access

A Prospective, Randomized, Double-Blinded, Placebo-Controlled Study of Human Intravenous Immunoglobulin for the Acute Management of Presumptive Primary Immune-Mediated Thrombocytopenia in Dogs


  • Dr Bianco is presently affiliated with Red Bank Veterinary Hospital, Tinton Falls, NJ. Abstract was presented at the 2008 ACVIM Forum in San Antonio, TX.

Corresponding author: Dr Domenico Bianco, DVM, PhD, Dipl. ACVIM, Department of Veterinary Clinical Sciences, College of Veterinary Medicine, University of Minnesota, 1352 Boyd Avenue, St Paul, MN 55108; e-mail: bianc037@umn.edu.


Background: Immune-mediated thrombocytopenia (IMT) is a common hematologic disorder in dogs. Human intravenous immunoglobulin (hIVIG) may have a beneficial effect in canine IMT.

Hypothesis: A single hIVIG infusion (0.5 g/kg) in dogs with presumed primary IMT (pIMT) is a safe adjunctive emergency treatment to accelerate platelet count recovery and shorten hospitalization time without increasing the cost of patient care.

Animals: Eighteen client-owned dogs with a presumptive diagnosis of pIMT.

Methods: Prospective, randomized, double-blinded, placebo-controlled clinical trial.

Results: There were no identifiable immediate or delayed adverse reactions associated with hIVIG administration over a 6-month period. The median platelet count recovery time for the hIVIG group was 3.5 days (mean ± SD: 3.7 ± 1.3 days; range, 2–7 days) and 7.5 days (mean ± SD: 7.8 ± 3.9 days; range, 3–12 days) for the placebo group. The median duration of hospitalization for hIVIG group was 4 days (mean ± SD: 4.2 ± 0.4 days; range, 2–8 days) and 8 days (mean ± SD: 8.3 ± 0.6 days; range, 4–12 days) for the placebo group. There was no significant difference between groups with respect to expense of initial patient care, whereas significant reduction in platelet count recovery time (P= .018) and duration of hospitalization (P= .027) were detected in the hIVIG group.

Conclusions and Clinical Importance: Compared with corticosteroids alone, adjunctive emergency therapy of a single hIVIG infusion was safe and associated with a significant reduction in platelet count recovery time and duration of hospitalization without increasing the expense of medical care in a small group of dogs with presumed pIMT.


activated partial thromboplastin time


fibrin degradation products


fresh frozen plasma


fresh whole blood


human intravenous immunoglobulin


intensive care unit




immune-mediated thrombocytopenia


idiopathic thrombocytopenic purpura


neutered male


polymerase chain reaction


primary immune-mediated thrombocytopenia


packed red blood cell


prothrombin time


standard deviation


spayed female


Veterinary Medical Center of the University of Minnesota

Immune-mediated thrombocytopenia (IMT) in dogs is a common hematologic disorder in which antibodies bound to the surface of platelets cause premature platelet destruction and removal by the mononuclear phagocytic system.1 Platelet-bound antibodies can be antiplatelet autoantibodies, 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.1–4 It may be a primary condition or secondary to known antigenic stimuli, such as neoplasia, infectious diseases, and drugs.2,5–11 Methods for the detection of platelet-bound antibody, including ELISA,2 flow cytometry,3,11–14 and immunoradiometric methods,15 have been reported in dogs with IMT. Furthermore, successful immunosuppressive treatment of IMT with a decrease in platelet-bound antibodies also has been reported.a3,13,14

Immunosuppressive dosages of corticosteroids have been recommended as the standard therapy for canine IMT.1,3,10,13,16–22 Other immunosuppressive drugs and splenectomy also have been recommended for use as adjunctive therapy in canine IMT, although their efficacy has not yet been established in well-designed clinical trials.b,1,3,10,16–25 Based on clinical observations of dogs with primary IMT (pIMT), the risk of bleeding is minimized when platelet counts return to >40,000/μL.13 Using this criterion, most dogs with severe IMT respond to corticosteroid therapy with or without vincristine within 4.5–6.5 days of starting therapy, respectively.19 Platelet transfusions (fresh whole blood [FWB], platelet-rich plasma, or platelets concentrates) generally are not recommended in most dogs with IMT because the platelets are destroyed or consumed within minutes to hours of transfusion. In dogs with IMT experiencing uncontrolled or life-threatening bleeding (eg, into brain, myocardium, or lungs), however, platelet transfusions may provide short-term hemostasis despite the lack of a measurable increase in platelet count posttransfusion.26,27 The mortality rate of canine pIMT ranges from 10 to 30%,1,3 and in our clinical experience the most challenging cases are animals presented in the emergency setting for severe gastrointestinal hemorrhage that require massive blood product transfusions and experience a delayed response to conventional immunosuppressive therapy. For these reasons, we chose to investigate an alternative fast-acting treatment option for the acute management of canine pIMT.

Human intravenous immunoglobulin (hIVIG) is a sterile immunoglobulin (Ig) preparation that contains IgG and trace amounts of IgM, IgA, CD4, CD8, and human leukocyte antigen molecules from a healthy human donor population.28 Recent research in humans has demonstrated that the anti-inflammatory activity of IgG is completely dependent on sialylation on the N-linked glycan of the IgG Fc fragment.29 Blockade of Fc receptors on mononuclear phagocytic cells has been described as a mechanism for the observed response to hIVIG infusion in dogs.30 hIVIG has been administered to dogs to treat hematologic and dermatologic immune-mediated disorders,31–37 and retrospective studies have reported that hIVIG was well tolerated and associated with rapid platelet count recovery in most dogs with presumed pIMT.c38,39

The aims of this prospective clinical trial were to compare the effect of hIVIG plus corticosteroids versus corticosteroids alone on platelet count recovery, hospitalization time, transfusion requirements, cost of patient care, and mortality in dogs with presumed pIMT. We hypothesized that a single hIVIG infusion within 48 hours of diagnosis in dogs with presumed pIMT would be a safe adjunctive emergency treatment to accelerate platelet count recovery time, decrease transfusion requirements, and shorten hospitalization time without increasing the cost of initial care.

Materials and Methods

Study Population

Client-owned dogs were enrolled at the Veterinary Medical Center of the University of Minnesota (UM-VMC) between August 2006 and November 2007. The study was approved by the Institutional Animal Care and Use Committee at the University of Minnesota. All dogs were second-opinion, referral cases.

Inclusion Criteria

To be included in the study, dogs must have severe thrombocytopenia (platelet count <20,000/μL), weighed 5 kg or more, and had a presumptive diagnosis of pIMT based on extensive diagnostic investigation (outlined below) to exclude other causes of severe thrombocytopenia. Dogs were eligible for inclusion in the study only if their owners gave informed consent.

Exclusion Criteria

Dogs were excluded from the study if they had been treated with corticosteroids for >24 hours before presentation (to allow clinicians to start corticosteroid therapy on an emergency basis and complete the initial diagnostic investigation before possible study drug administration), had received any additional immunosuppressive therapy, were exposed to any drugs or toxins within 4 weeks before presentation, had relapsing IMT, suffered from concurrent immune-mediated hemolytic anemia, neoplasia, infectious disease, or had renal azotemia (serum creatinine concentration >1.4 mg/dL) or proteinuria (urine protein/creatinine ratio >0.5).

Diagnostic Investigation

Within 24 hours of admission, all dogs were evaluated with a complete history, (including recent vaccinations or any known exposure to drugs or toxins), physical examination, CBC, serum biochemistry, urinalysis on a voided midstream urine sample, urine protein/creatinine ratio, coagulation profile (including fibrinogen, prothrombin time [PT], activated partial thromboplastin time [aPTT], and fibrin degradation products [FDPs]), systolic blood pressure measurement, 3-view thoracic radiographs, abdominal radiography, abdominal ultrasonography, and vector-borne disease serology (Dirofilaria immitis,dEhrlichia canis, Rickettsia rickettsii, Anaplasma phagocytophilum, Babesia canis). Polymerase chain reaction (PCR) testing for Babesia spp. was performed at the time of presentation in all enrolled American Pitbulls and Greyhounds. Convalescent serologic titers were repeated 3–4 weeks after initial presentation in all patients with history of acute illness. Bone marrow cytological and histopathologic evaluation also was performed in all dogs. Bone marrow core biopsies and aspirates were taken from the iliac crest or proximal humerus. Bone marrow aspirates, measuring usually at least 0.5 mL in volume, were placed in 2% EDTA-treated tubes. For the preparation of bone marrow smears to assess cell counts and cell distribution, a particle-squash technique was routinely used.40 In all anemic patients (hematocrit <35%) a slide agglutination test and direct Coombs' test were performed before transfusion and immunosuppressive therapy were initiated. Canine antiplatelet antibody tests were not available at the UM-VMC during this clinical trial. Complete necropsy at the University of Minnesota Diagnostic Laboratory was performed for each enrolled dog that did not survive to the end of the study period. Imaging studies were reviewed by board-certified veterinary radiologists. Blood smears, bone marrow, and necropsy examinations were reviewed by board-certified veterinary pathologists.

Study Design

The study was a prospective single-center, randomized, double-blinded, placebo-controlled clinical trial. Randomization was based on blocks of 4 using a table of random numbers. Block randomization was used with a 1 : 1 allocation ratio to maintain similar sample sizes in both treatment groups.41 Allotment to the treatment groups was based on sequentially numbered sealed envelopes to be opened by an unblinded pharmacist. Investigators, owners, technicians, laboratory personnel, and veterinary students were blinded for the duration of the study. Unblinding occurred only after completion of the study and data entry. Study drug or placebo were reconstituted by an unblinded pharmacist who then gave an opaque infusion set and bag to an intensive care unit technician who was not otherwise directly involved in study patient management. Attending clinicians and pet owners were not permitted to visit the dogs during the time of infusion. To maintain pet owner blinding, charges for the hIVIG were assessed and paid by the clients 6 months after their pet's initial presentation (at the end of the study period).

Test Treatment

Within 24 hours of admission, enrolled dogs were randomized to receive either a single infusion of hIVIGe (0.5 g/kg at a 6% concentration over 6–12 hours) or placebo (0.9% NaCl) IV, using identical infusion volume and time via an opaque infusion set and bag. Before the infusion, all dogs were treated with diphenhydramine (0.5 mg/kg IV). Attitude, rectal temperature, heart rate, and respiratory rate were monitored immediately before and every 10 minutes for the first 40 minutes of infusion, and at the end of the infusion period. During the 6-month study period, any sign possibly attributed to an adverse drug reaction was recorded.

Study Endpoints

The primary endpoint of the study was the time needed for the dogs to reach a platelet count >40,000/μL. Secondary endpoints over a 6-month period included disease-related mortality, acute or delayed adverse reactions to study drug, duration of hospitalization, transfusion requirements, days to complete hematologic response to treatment (defined as the time required for the platelet count to return to within the laboratory reference range, 160,000–425,000/μL), rate of relapse (defined as a platelet count decrease of 50% compared with previous count or any count of <40,000 platelets/μL after initial response), complications due to immunosuppressive therapy, and costs to client from initial presentation to discharge (dogs were discharged from the hospital if they were clinically stable and had platelet counts >40,000/μL), including the cost of hIVIG in cases that were randomized to receive it or the cost of an hIVIG infusion after 1 week, if applicable. If a dog died spontaneously or was euthanized, the attending clinician specified whether they considered the cause of death to be IMT related or not.

Concomitant Treatments

All dogs were hospitalized in the intensive care unit (ICU) at the UM-VMC for 24-hour monitoring and care. From admission to day 7, all patients were treated with doxycyclinef (5 mg/kg PO or IV q12h), prednisoneg (1.5 mg/kg PO or IV q12h), famotidineh (0.5 mg/kg PO or IV q24h), and sucralfatei (50 mg/kg PO q8h). Cross-matched packed red blood cell (PRBC), fresh frozen plasma (FFP), FWB, and crystalloid fluids with electrolyte supplementation were administered at the attending clinician's discretion. The use of synthetic colloids was discouraged because of the potential of causing platelet dysfunction,42 but allowed at the attending clinician's discretion. On day 7, the attending clinician was allowed to use any additional immunosuppressive drugs, including hIVIG. In this case, placebo group dogs would receive hIVIG, whereas hIVIG group dogs would receive placebo in a blinded fashion as described above. Prednisone dosages were decreased by 25% at 2- or 4-week intervals based on clinical assessment and clinicopathologic findings at recheck appointments. The time to discontinuation of each drug was decided by attending clinicians.

Response Monitoring

Platelet counts were determined in EDTA-anticoagulated blood by use of automated cell counters.j Platelet counts <40,000/μL were confirmed by manual counting of platelets with a hemocytometer and blood smears were evaluated for additional confirmation of thrombocytopenia or evidence of platelet clumping. For each case, a platelet count was performed at least every 24 hours starting from admission until discharge. A platelet count also was repeated immediately before hIVIG or placebo administration. In all dogs, physical examination, CBC including platelet count, and serum biochemistry were performed on days 7, 21, 60, 120, and 180. Follow-up ended on day 180 after initial presentation.

Statistical Analysis

Power calculations were based on a previous prospective study of canine IMT.19 A study population of 20 dogs for each treatment group was estimated to provide a power of approximately 80% at 0.05 significance level to detect a 50% difference in median times to reach the primary endpoint between the treatment groups. An interim analysis was performed by a study monitor (PJA) after 50% enrollment was achieved. Data from dogs in hIVIG- and placebo group were compared by use of a Student t-test for data with normal distribution and a Mann-Whitney rank sum test was used for data with non-normal distribution. The Kaplan-Meier estimates of the distribution of times from diagnosis to death were computed, and the Mantel-Cox log-rank analysis was performed to compare the survival curves between the 2 groups of thrombocytopenic dogs. Results are presented as median and range unless indicated otherwise. Statistical analyses were performed using a standard statistical software package.k For all analyses, values P <.05 were considered statistically significant.


Study Population

Twenty-five client-owned dogs fulfilled the entry criteria, and 18 of them were enrolled in the study. Five dogs were eligible for inclusion in the clinical trial, but the owners declined (4 for financial reasons and 1 due to safety concerns). Two dogs were prematurely enrolled and randomized, but excluded before hIVIG or placebo administration because of laboratory error (normal platelet count in a repeated measurement) and a concomitant diagnosis of immune-mediated hemolytic anemia (positive direct Coombs' test), respectively. Because an interim analysis at 50% enrollment demonstrated a significant reduction in platelet count recovery time and duration of hospitalization with the use of hIVIG, the investigators considered it unethical to continue and elected to end the study before the intended enrollment goal of 40 dogs. All cases were referred on an emergency basis to the UM-VMC. Fourteen of 18 dogs were managed by one of the authors (DB). The enrolled dogs consisted of 16 purebreds and 2 mixed breed dogs with a mean age ± SD of 8.3 ± 1.7 years and a mean ± SD body weight of 20.9 ± 2.8 kg. Within the hIVIG group, there were 7 spayed females (SF) and 2 neutered males (NM). Within the placebo group, there were 7 SF, 1 NM, and 1 intact male dog. Overall, 8 pure breed (2 American Cocker Spaniels, 1 Rottweiler, 1 English Springer Spaniel, 1 American Staffordshire Terrier, 1 Border Collie, 1 Dachshund, and 1 Toy Poodle) and 1 mixed breed dog were represented in the hIVIG group, and 8 pure breeds (1 Golden Retriever, 1 Standard Poodle, 1 American Cocker Spaniel, 1 English Setter, 1 Italian Greyhound, 1 Pomeranian, 1 Shih Tzu, and 1 Maltese) and 1 mixed breed dog were represented in the placebo group.

Clinical signs noted by owners included lethargy (n = 18), decreased appetite (11), petechiae or ecchymoses (9), epistaxis (5), hematochezia (4), hematuria (4), and hematemesis (2). Based on the history, median duration of clinical signs before presentation was 2 days (range, 1–8 days). At presentation, the median rectal temperature was 101.2°F (range, 99.1–103.7°F), the median heart rate was 120 beats/min (range, 60–200 beats/min), and the median respiratory rate was 32 breaths/min (range, 16–60 breaths/min). On physical examination, all dogs had petechiae and ecchymoses. Other findings included melena or hematochezia (hIVIG group = 7; placebo group = 6), pale mucous membranes (hIVIG group = 6; placebo group = 6), abdominal organomegaly (hIVIG group = 4; placebo group = 3), epistaxis (hIVIG group = 2; placebo group = 3), systolic cardiac murmur (hIVIG group = 3; placebo group = 3), and scleral hemorrhage or hyphema (hIVIG group = 2; placebo group = 1). Neurologic examination was unremarkable in all dogs.

On the day of admission, the median platelet count was 2,000/μL (range, 1,000–18,000/μL; reference range, 160,000–425,000/μL), the median hematocrit was 26% (range, 11–49%; reference range, 39–57%), and the median serum albumin concentration was 2.5 g/dL (range, 1.5–3.5 g/dL; reference range, 2.8–3.9 g/dL). Results of measurements of coagulation were as follows: fibrinogen (median, 0.2 g/dL; range, 0.1–0.6 g/dL; reference range, 0.2–0.4 g/dL), PT (median, 6.6 seconds; range, 6.3–7.4 seconds; reference range, 6.2–7.7 seconds), aPTT (median, 11.4 seconds; range, 10.5–14.2 seconds; reference range, 9.8–14.6 seconds), and FDPs (median, <5 μg/mL; range, <5 to >20 μg/mL; reference range, <5 μg/mL). Direct Coombs' test results determined for 14 dogs were negative. On presentation, results of serologic testing for D. immitis, E. canis, B. canis, R. rickettsii, and A. phagocytophilum were negative in all dogs, and PCR for Babesia spp. was negative in the 2 tested dogs. Convalescent serologic titers for E. canis, B. canis, R. rickettsii, and A. phagocytophilum were negative in the 12 tested dogs.

Thoracic radiographs were normal in all dogs. Abdominal radiographic abnormalities included splenomegaly (n = 4), hepatosplenomegaly (2), and hepatomegaly (1). Abdominal ultrasonography revealed mild bilateral symmetrical adrenomegaly and diffusely coarse hepatic parenchyma in 1 dog that had concurrent pituitary-dependent hyperadrenocorticism; no clinically relevant findings were noted in any of the other dogs. Bone marrow aspirates and core biopsy specimens were considered of adequate quality for evaluation in all dogs. Bone marrow findings included megakaryocytic hyperplasia in 12 dogs, adequate megakaryocytes in 4 dogs, and megakaryocytic hypoplasia in 2 dogs (1 from each group). Increased numbers of plasma cells were reported in 7 dogs, and increased lymphocyte numbers were reported in 2 dogs. Dysplastic features were not observed in any cell lines, and erythrophagocytosis was not observed. On presentation, there was no statistical difference between the 2 groups with regard to age, body weight, duration of clinical signs, rectal temperature, heart rate, respiratory rate, platelet count, hematocrit, and serum albumin concentration (Table 1).

Table 1.   Comparison of baseline variables between treatment groups.
VariablehIVIG GroupPlacebo GroupP-Value
  1. Data are expressed as median and range.

  2. BPM, beats per minute; bpm, breaths per minute; hIVIG, human intravenous immunoglobulin.

Age (years)8.5 (2–12)7.5 (2–12.5).43
Body weight (kg)19.8 (7.2–36.9)21.4 (12.1–40.8).41
Duration of clinical signs (days)2 (1–8)3 (1–7).54
Rectal temperature (°F)101.4 (99.6–103.4)100.9 (99.1–103.7).61
Heart rate (BPM)140 (80–200)120 (60–200).38
Respiratory rate (bpm)32 (20–60)28 (20–48).35
Platelet count (platelets/μL)1,000 (0–16,000)2,000 (1,000–18,000).59
Hematocrit (%)23 (11–45)29 (23–49).26
Albumin (g/dL)2.6 (1.5–3.5)2.5 (1.6–3.5).75

Adverse Drug Reactions

There were no identifiable immediate or delayed adverse drug reactions detected over a 6-month period.

Primary Endpoint

There was a significantly shorter median lag time from the start of treatment until platelet count increased to >40,000/μL in the hIVIG group compared with the placebo group (Table 2). All dogs in the hIVIG group responded by day 7, whereas 4 dogs in the placebo group failed to respond to corticosteroids alone. On day 7, 2 of these dogs were treated with azathioprinel (2 mg/kg PO q24h) and vincristinem (0.02 mg/kg IV once), 1 with hIVIG and azathioprine, and 1 with hIVIG and cyclosporinen (5 mg/kg PO q12h). One of the placebo dogs subsequently treated with azathioprine and vincristine was humanely euthanized at the owners' request 2 days after discharge despite a platelet count of 58,000/μL because of severe muscle weakness. Three of the 9 dogs in the hIVIG group were treated with additional azathioprine after the 1st week of therapy to minimize the amount of corticosteroids used.

Table 2.   Comparison between treatment groups for the median time to reach study endpoints.
Study EndpointhIVIG GroupPlacebo GroupP-Value
  1. Data are expressed as median and range. Bold P-value numeral indicates statistical significance at P <.05.

  2. PRBC, packed red blood cell; hIVIG, human intravenous immunoglobulin.

Initial response (days to platelet count >40,000/μL)3.5 (2–7)7.5 (3–12).018
Duration of hospitalization (days)4 (2–8)8 (4–12).027
Complete response (days to platelet count >160,000/μL)8 (3–19)13 (5–32).093
PRBC transfusion requirement (mL/kg)14.7 (0–34.5)21.2 (0–68.7).62
Cost of initial hospitalization (US$)4,196 (1,868–6,143)4,954 (1,439–7,756).78

Secondary Endpoints

Two of 9 placebo-group dogs did not survive the study period due to IMT-related reasons. One dog died on day 3 because of failure to maintain blood volume with transfusions, and 1 dog was humanely euthanized on day 13. Postmortem examination of these 2 dogs did not reveal any underlying diseases associated with severe thrombocytopenia. All dogs in the hIVIG group survived to hospital discharge and over the 6-month follow-up. Survival analysis did not indicate any differences with regard to time of death from causes related to IMT, either comparing the 2 initial randomized groups (P= .79) or comparing all dogs that did not receive hIVIG versus those that did at some point during the 6-month study period (P= .53). Dogs in the hIVIG group had a significantly shorter duration of hospitalization compared with the placebo group (Table 2). Median time to complete platelet count recovery (platelet count >160,000/μL) for the hIVIG group was not different compared with the 7 surviving dogs in the placebo group (Table 2). Two dogs, 1 from each group, suffered a relapse with platelet counts <40,000/μL that responded to increased doses of corticosteroids. The neutrophil count was significantly increased on day 21 (P <.001) and on day 60 (P= .007) in both groups compared with the neutrophil count on day 1, but no other significant differences were identified in other white cell lines at each time point. All dogs developed signs of iatrogenic hyperadrenocorticism, including recurrent urinary tract infections in 2 dogs, 1 from each group. The median volume of PRBC transfusions received by hIVIG group dogs was not significantly different compared with the placebo group dogs (Table 2). Two dogs in the hIVIG group and 4 dogs in the placebo group received FWB transfusions, 2 dogs in the placebo group received FFP transfusions, and 2 dogs, 1 from each group, received synthetic colloids. All dogs received IV fluid therapy with crystalloids, and the daily amount of administered crystalloids was not different between the 2 groups. Blood products were not administered to 2 dogs of the hIVIG group and 1 dog of the placebo group. The median cost of initial hospitalization for the hIVIG group dogs was not significantly different compared with the cost of care for the placebo group dogs (Table 2). The median cost of the initial diagnostic investigation for all dogs was US$1,482 (mean US$1,451; range, US$1,290–US$1,846). The median cost of hIVIG treatment was US$995 (mean, US$948; range, US$412–US$1,957).


The results of this prospective clinical trial showed that a single infusion of hIVIG (0.5 g/kg), within 24 hours from starting concurrent corticosteroid therapy, is associated with a significant reduction in platelet count recovery time and duration of hospitalization without increasing the expense of medical care in dogs with presumed pIMT in a referral institution. Although 2 dogs were prematurely randomized and then excluded from the study, randomization did not appear to be disrupted at the interim analysis, with an equal number of dogs with very similar baseline variables in both treatment groups. No associated systemic or local adverse effects attributable to hIVIG administration were observed in this small group of dogs over a 6-month period. Immunoglobulin of canine origin is not commercially available, and antigenicity of hIVIG in dogs is unknown. Administration of other human proteins to healthy dogs may result in severe immediate and delayed adverse reactions. For example, serum antibodies against human albumin have been detected in 7% of healthy dogs with no history of previous exposure.43–45

Signalment, physical examination, clinicopathologic, imaging, and bone marrow findings of the dogs enrolled in our study were similar to those described in previous reports.1,3,11,15–21,46 Definitive diagnosis of IMT may be challenging, because no single test is available to confirm or refute the diagnosis.19 In this study, the presumptive diagnosis of pIMT was made on the basis of finding a very low platelet count (<20,000 platelets/μL) without evidence of any underlying disease, despite a comprehensive diagnostic investigation and a 6-month follow-up. Finally, a response to treatment with immunosuppressive agents is often used in support of a diagnosis of IMT.1,16,17,19 Canine antiplatelet antibody testsa1–4,10–15 can demonstrate an immune basis for thrombocytopenia but were not available at the UM-VMC during this clinical trial.

Canine IMT carries a fair to good prognosis,1 and a recent retrospective study with long-term follow-up data reported only 10% mortality rate for 30 dogs with pIMT.3 Similarly, the mortality rate in our study over a 6-month period was 11.1% with 2 fatalities, both in the placebo group, and the survival analysis did not reveal any differences between the 2 groups. A limitation of this study relates to the platelet count being only a surrogate measure for risk of life-threatening hemorrhage as the clinically relevant endpoint. It remains to be determined how effective platelet count recovery time is as a surrogate for survival in dogs with presumed pIMT. The impetus for treating acute idiopathic thrombocytopenic purpura (ITP) in children is the prevention of intracranial hemorrhage. In pediatric medicine, children treated with corticosteroids for acute ITP are 26% less likely to have a platelet count >20,000/μL after 48 hours of therapy when compared with children treated with IVIG.47 Although, neurological signs due to intracranial hemorrhage have been suspected in dogs with IMT,1,3,17 they were not noted in any dog in the present study. As previously observed,19 PRBC transfusion requirements varied substantially among individual dogs with presumed IMT in our study, despite similarly low platelet counts, and the most common site of severe hemorrhage was the gastrointestinal tract. Although the hIVIG group dogs received fewer PRBC transfusions compared with the placebo group dogs, this finding did not reach statistical significance.

Traditionally, corticosteroids were the first medication used to successfully increase platelet counts in dogs with IMT and became the standard therapy of this disorder.1,3,10,13,16–24 Consequently, controlled clinical trials assessing the efficacy of corticosteroids in dogs with IMT cannot be ethically performed.1,19 Scott-Moncrieff and colleagues treated 1 dog with IMT not responsive to prednisone, and observed a relapse 23 days after hIVIG treatment.38 Because the half-life of hIVIG in healthy dogs is 7–9 days,31 this therapy may only be effective only in the short-term management of canine IMT. Previous retrospective studies reported that a single infusion of hIVIG was effective in approximately 60–80% of dogs with presumed pIMT.c39 In the study presented here, a similar response rate was observed in most dogs with presumed pIMT shortly after completion of the hIVIG infusion.

The optimal dose for hIVIG in dogs remains to be fully identified, and other veterinary studies reported use of dosages ranging from 0.25 to 2.2 g/kg.c30–39 In the present clinical trial, the investigators evaluated the effects of a low dose of hIVIG (0.5 g/kg) because of the importance of cost containment in veterinary health care. In children with acute ITP, the use of high-dose immunoglobulin (2 g/kg) subdivided in a 2-consecutive day administration was superior by day 3 from start of therapy compared with a similar protocol using a low dose (0.6 g/kg) in terms of platelet count recovery and number of nonresponders.48 Recent studies in humans have demonstrated that the anti-inflammatory activity of IgG is completely dependent on sialylation on the N-linked glycan of the IgG Fc fragment, and a fully recombinant sialylated IgG1Fc with greatly enhanced potency was generated to help guide development of an IVIG replacement with improved activity and availability.29,49 Additional studies are needed to evaluate a possible dose-response relationship of hIVIG for the treatment of canine IMT.

In a previous study, dogs with severe IMT that were treated with a combination of vincristine and prednisone had a more rapid increase in platelet counts and shorter duration of hospitalization when compared with dogs treated with prednisone alone.19 Using the same study endpoints, the hIVIG group dogs in our study had a comparable platelet count recovery time and duration of hospitalization. The mortality rate also was similar between the 2 studies. Additional prospective clinical trials are needed to compare vincristine versus hIVIG as adjunctive emergency therapy in dogs with severe IMT. In human medicine, the cost-effectiveness of hIVIG versus corticosteroids for the treatment of ITP often has been debated.50 Although the cost of initial hospitalization for the hIVIG group dogs was not significantly different compared with the placebo group dogs, long-term cost-effectiveness was not evaluated and a referral population of IMT dogs was exclusively represented in this clinical trial. Therefore, because the difference in cost between hIVIG and corticosteroids with or without a single dose of vincristine is substantial (US$502/10 kg dog for prednisone/hIVIG versus US$11/10 kg dog for prednisone and vincristine, excluding drug administration costso), and the generally good response of acute canine IMT to corticosteroid therapy alone,1,3,13,16–19 the use of hIVIG is not necessarily recommended in all IMT cases. On the other hand, the judicious use of hIVIG should be considered in IMT dogs with active hemorrhage that are not responding to conventional therapy.


aKohn B, Engelbrecht R, Giger U, et al. Platelet-bound antibodies in dogs with thrombocytopenia and change with treatment. J Vet Intern Med 2000;14:134

bCook AK, Bertoy EH, Gregory CR. Effect of oral cyclosporine (CS) in dogs with refractory immune-mediated anemia (IMA) and thrombocytopenia (ITP). J Vet Intern Med 1994;8:170

cVilar P, Couto CG, Lara A, et al. Use of human intravenous immunoglobulin G in dogs: A retrospective study. J Vet Intern Med 2007;21:622-623

dSNAP 3Dx Test, IDEXX Laboratories, Westbrook, ME

eImmune Globulin Intravenous (Human), Carimune NF, ZLB Berhing, King of Prussia, PA

fDoxycycline tablets, IVAX Pharmaceuticals, Miami, FL

gPrednisone tablets, Watson Labs, Corona, CA

hFamotidine, Pepcid AC tablets, distributed by Goldline Labs, manufactured by IVAX Pharmaceuticals Inc

iSucralfate tablets, Major Pharmaceuticals, Livonia, MI

jCell-Dyne System 3500, Diagnostics Division, Abbott Laboratories, Santa Clara, CA

kSPSS 14.0 for Windows, Microsoft, Redmond, WA

lAzathioprine, Roxane Labs, Columbus, OH

mVincristine injectable, manufactured for Mayne Pharma Inc, Paramus, NJ

nCyclosporine capsules, manufactured for PLIVA Inc, East Hanover, NJ

oUniversity of Minnesota, Veterinary Medical Center Pharmacy, St Paul, MN


The investigators are grateful to all dog owners for their willingness to enroll their pets in the study and to Dr Alisa Craig, PharmD, for her help. We also thank the 2006 CVM Small Companion Animal Grants Committee for study funding.

Supported by a 2006 Small Companion Animal Grant from the College of Veterinary Medicine, University of Minnesota, St Paul, MN.