Clinical data, clinicopathologic findings and outcome in dogs with amegakaryocytic thrombocytopenia and primary immune-mediated thrombocytopenia

ttp://www.bsava.com/ PAPER Clinical data, clinicopathologic findings and outcome in dogs with amegakaryocytic thrombocytopenia and primary immune-mediated thrombocytopenia S. A. Cooper, * A. A. Huang, * R. E. Raskin, H.-Y. Weng, and J. C. Scott-Moncrieff *,1 * Department of Veterinary Clinical Sciences, Purdue University College of Veterinary Medicine, West Lafayette, IN 47907, USA Department of Comparative Pathobiology, Purdue University College of Veterinary Medicine, West Lafayette, IN 47907, USA 1 Corresponding author email: scottmon@purdue.edu O BJECTIVES : The aim of this study was to identify distinguishing characteristics between dogs diagnosed with amegakaryocytic thrombocytopenia and those diagnosed with presumed primary peripheral immune-mediated thrombocytopenia. Presenting clinical and clinicopathologic data and outcomes were compared between the two groups. M ETHODS : Retrospective study performed on seven client-owned dogs diagnosed with amegakaryocytic thrombocytopenia and 34 client-owned dogs with primary peripheral immune-mediated thrombocytopenia. R ESULTS : All dogs in the amegakaryocytic thrombocytopenia group were anaemic on presentation with a median haematocrit of 23% (range 9 4 to 36), while the primary peripheral immune-mediated thrombocytopoenia group had a median presenting haematocrit of 35% (range 10 to 53). Dogs with amegakaryocytic thrombocytopenia had a median of five (range 4 to 7) clinical signs of bleeding compared to a median of three (range 0 to 6) in the primary peripheral immune-mediated thrombocytopenia group with 86% (6 of 7) of amegakaryocytic thrombocytopenia dogs requiring a blood transfusion compared to 41% (14 of 34) of primary peripheral immune-mediated thrombocytopenia dogs. Six of the seven amegakaryocytic thrombocytopenia dogs did not survive to discharge, while only five of the 34 primary peripheral immune-mediated thrombocytopenia dogs did not survive to discharge. C LINICAL S IGNIFICANCE : The clinical presentation of dogs with amegakaryocytic thrombocytopenia and primary peripheral immune-mediated thrombocytopenia is similar, but dogs with amegakaryocytic thrombocytopenia had a more severe clinical course compared to primary peripheral immune-mediated thrombocytopenia dogs. The prognosis for dogs with amegakaryocytic thrombocytopenia is poor. Journal of Small Animal Practice (2016) 57, 142 147 DOI: 10.1111/jsap.12441 Accepted: 10 December 2015; Published online: 22 January 2016 A.A. Huang s current address is Seattle Veterinary Specialists, Kirkland, WA 98034, USA S.A. Cooper s current address is United States Army Public Health Command, Fort Bragg, NC 28307, USA The views expressed in the article are those of the author and do not reflect the official policy or position of the Department of the Army, Department of Defense or the US Government. Veterinary Medical Database, Purdue University Veterinary Teaching Hospital, West Lafayette, IN (http://www.vmdb.org); VMDB does not make any implicit or implied opinion on the subject of the paper or study. 142 Journal of Small Animal Practice Vol 57 March 2016 2016 British Small Animal Veterinary Association

Amegakaryocytic thrombocytopenia INTRODUCTION Immune-mediated thrombocytopenia (IMT) can be categorised as either primary or secondary. Primary IMT is caused either by the peripheral destruction of platelets or by the decreased production of platelets because of immune-mediated destruction of megakaryocytes in the bone marrow (Lewis & Meyers 1996, Rozanski et al. 2002, Miller & Lunn 2007, Putsche & Kohn 2008, O Marra et al. 2011). Most dogs with IMT have adequate to increased numbers of megakaryocytes in the bone marrow, representing an appropriate bone marrow response to peripheral platelet destruction (Lachowicz et al. 2004, Miller & Lunn 2007). Conversely, amegakaryocytic thrombocytopenia (AMTP) is defined as thrombocytopenia with concurrent low to absent numbers of megakaryocytes within the bone marrow (Lachowicz et al. 2004, Miller & Lunn 2007). AMTP is believed to be due to primary or secondary immune-mediated destruction of megakaryocytes (Feldman et al. 1988, Lachowicz et al. 2004). In humans, AMTP is an uncommon cause of thrombocytopenia, with most reports in the literature consisting of small case series or case reports (Agarwal et al. 2006). In the veterinary literature, only one case report and one small case series of AMTP have been reported (Murtaugh & Jacobs 1985, Lachowicz et al. 2004), although some studies have described cases as part of larger IMT studies (Williams & Maggio-Price 1984, Rozanski et al. 2002, Miller & Lunn 2007, Putsche & Kohn 2008). The prognosis for IMT in dogs is reported to be good with survival rates ranging from 70 to 93% (Lewis & Meyers 1996, Putsche & Kohn 2008, O Marra et al. 2011). The prognosis for dogs with AMTP is unclear because of the small number of documented cases, although an unfavourable outcome has been reported in some studies (Williams & Maggio-Price 1984, Lachowicz et al. 2004). Given the difference in prognosis for primary peripheral immune-mediated thrombocytopenia (pimt) compared to AMTP, early identification of AMTP would provide important prognostic information. Currently, the most definitive way to determine if a patient has pimt or AMTP is to collect either a bone marrow aspirate and/or core sample. Miller & Lunn (2007) concluded that bone marrow examination was unlikely to provide specific diagnostic or prognostic information in dogs with severe thrombocytopenia. Because of this report, and the invasiveness of bone marrow collection, a bone marrow examination is not routinely performed in dogs with severe thrombocytopenia that are suspected to have IMT. Thus, a non-invasive means of differentiating AMTP from pimt would be clinically beneficial. The objective of this study was to compare the presenting clinical and clinicopathologic data from dogs with pimt and AMTP to identify any distinguishing characteristics between these two disorders. A secondary objective was to determine the severity of disease as evidenced by transfusion dependence and survival of dogs with AMTP compared to those with pimt. Our hypothesis was that patients with AMTP would have a more severe clinical presentation when compared to dogs with pimt, demonstrated by increased evidence of clinical bleeding, lower presenting haematocrit (HCT) and lower presenting platelet counts. MATERIALS AND METHODS Patient selection AMTP A search of the computerised Veterinary Medical Database and Hospital Information System from January 1, 1991 to December 31, 2011 was performed to identify patients with a final diagnosis of AMTP. Search terms included megakaryocytic hypoplasia and megakaryocytic aplasia. Inclusion criteria for patients with AMTP included a presenting platelet count less than 40,000/µL, a diagnosis of megakaryocytic aplasia or hypoplasia on a bone marrow aspirate or core sample and no underlying cause of thrombocytopenia identified through a combination of history, physical examination findings, diagnostic imaging and laboratory testing, including vectorborne disease testing. Exclusion criteria included previous treatment with cytotoxic drugs or lack of a complete blood count at the time of bone marrow sample collection. Bone marrow slides FIG 1. Bone marrow histological evaluation from dogs presenting with thrombocytopenia. (A) Case of presumed primary peripheral immune-mediated thrombocytopenia demonstrating megakaryocytic hyperplasia. Six mature megakaryocytes are shown in one high-power field. 40 objective. HE stain. (B) Low magnification of bone marrow demonstrating predominately erythroid precursors and the lack of mature megakaryocytes in a dog with amegakaryocytic thrombocytopoenia. 20 objective. HE stain. (C) Antibody against factor VIII-related antigen demonstrates immature cells consistent with megakaryoblasts as indicated by the dark brown cytoplasm in a dog with amegakaryocytic thrombocytopenia (same case as B). 60 objective. Benzidine chromogen/haematoxylin. Journal of Small Animal Practice Vol 57 March 2016 2016 British Small Animal Veterinary Association 143

S. A. Cooper et al. were retrieved and reviewed by one pathologist to confirm the diagnosis of megakaryocytic aplasia or hypoplasia. Aplasia was defined as an apparent lack of megakaryocytes. Hypoplasia was defined as less than adequate numbers of megakaryocytes in the bone marrow with adequate defined as an average of 2 to 5 megakaryocytes per 40 objective by histology or per particle by cytology using a 10 objective (Silva et al. 2012 ) (Fig 1 ). Patient selection pimt The pimt dogs included in this study were part of a group of dogs included in a previous study (Huang et al. 2012 ) that were determined to have primary IMT. Inclusion criteria for pimt included a presenting platelet count of less than 40,000/ µ L, a bone marrow aspirate or core sample documenting normal or increased numbers of megakaryocytes and no underlying cause of thrombocytopoenia identified through a combination of history, physical examination findings, diagnostic imaging and laboratory testing, including vector-borne disease testing. Exclusion criteria included previous treatment with cytotoxic drugs or lack of a complete blood count at the time of bone marrow sample collection. Data collection Variables that were recorded for both groups included age, sex, breed, weight, historical and physical examination findings, complete blood count (CBC) results, bone marrow aspirate and core results; treatments administered including blood transfusion products, infectious disease titres, subsequent CBC data, duration of hospitalisation until death or discharge and cause of death. To assess the severity of clinical signs of bleeding, the number of sites demonstrating bleeding was recorded (bleeding from the oral cavity, haemoptysis, epistaxis, hyphaema, haematemesis, melaena, haematochezia, presence of petechiae, presence of ecchymosis and gross haematuria). Only the deaths that related to clinical signs of bleeding or poor prognosis were considered events of interest. For survival statistics, patients that expired of other causes or that survived the initial hospitalisation were censored. neoplasia (n=1), previous treatment with cytotoxic drugs (n=2), a platelet count greater than 40,000/ µ L at the time of presentation (n=4) or no CBC available at the time of bone marrow sample collection (n=1). Seven patients met all the inclusion criteria and were included in the study. Dogs in the AMTP group included Maltese (n=2), Rottweiler (n=1), Pembroke Welsh corgi (n=1), Newfoundland (n=1), miniature poodle (n=1) and mixed breed (n=1). There were five spayed females, one neutered male and one intact male. The median age at time of presentation was 6 04 years (range 1 to 9) and median weight was 17 kg (range 4 06 to 54 05). The median presenting clinicopathologic data indicated marked thrombocytopenia, mildly elevated white blood cell count and moderate anaemia (Table 1 ; Fig 2 ). All of the dogs had clinical evidence of bleeding at the time of presentation and the median number Table 1. Clinicopathological and transfusion data for dogs with amegakaryocytic thrombocytopenia ( AMTP ) (n=7) and presumed primary peripheral immune-mediated thrombocytopenia ( pimt ) (n=34). Median (range) is reported unless otherwise stated AMTP pimt Platelet count 230/ µ L (0 to 4300) 800/ µ L (0 to 28,000) Haematocrit 23% (9.4 to 36) 35% (10 to 53) White blood cell 23 0 K/ µ L (15 8 to 45 7) 18 1 K/ µ L (6 5 to 100 0) Clinical signs of bleeding 5 (4 to 7) 3 (0 to 6) Whole blood transfusions 0 (0 to 5) 0 (0 to 7) Packed red blood cell 1 (0 to 6) 0 (0 to 1) transfusions Number of dogs receiving 6 of 7 (86%) 14 of 34 (41%) a blood transfusion Days of hospitalisation 6 (1 to 9) 6 (0 to 19) until death or discharge Number of dogs that survived to discharge 1 of 7 (14%) 29 of 34 (85%) Laboratory methods Values for the white blood cell count, red blood cell count and platelet count were determined by standard automated methods. A manual leucocyte differential is standard protocol on all CBCs. Serum samples for determination of infectious disease titres were submitted to various reference laboratories. Data analysis Descriptive statistics of the study variables are reported. The Kaplan-Meier method was used to estimate median survival time until discharge. Statistical analyses was performed using a commercially available software programme. RESULTS AMTP population Fifteen records of patients with AMTP were identified and reviewed. Eight cases were excluded because of haematopoietic FIG 2. Presenting haematocrit ( HCT ) of dogs with amegakaryocytic thrombocytopenia ( AMTP ) (n=7) and presumed primary peripheral immune-mediated thrombocytopenia ( pimt ) (n=34). Dots are HCT of individual dogs and horizontal lines are the reference interval (37 to 55%) for HCT. 144 Journal of Small Animal Practice Vol 57 March 2016 2016 British Small Animal Veterinary Association

Amegakaryocytic thrombocytopenia of clinical signs of bleeding was higher than that of the pimt group. A higher percentage of dogs required blood transfusions in the AMTP group than in the pimt group (Table 1 ). Only one dog achieved a platelet count greater than 40,000/ µ L 5 days after initial presentation. This dog had a platelet count of greater than 200,000/ µ L on day 12 and was the only dog in this group that survived to discharge. pimt population Thirty-four dogs with pimt met the criteria for inclusion in the study. There were 17 breeds represented in the pimt group including mixed breed (n=15), miniature schnauzer (n=2), shih tzu (n=2), Doberman pinscher (n=2) and one each of the following: Yorkshire terrier, pug, German shepherd dog, bull mastiff, Siberian husky, rottweiler, Chinese crested, bichon frise, American cocker spaniel, fox terrier, Australian shepherd, Italian greyhound and golden retriever. There were 15 spayed females, 15 neutered males, 1 female of unknown neuter status and 3 males of unknown neuter status. The median age at the time of presentation was 7 years (range 2 to 15), and the median weight was 25 08 kg (range 4 02 to 58). The presenting clinicopathologic data indicated marked thrombocytopenia, mild elevation in the white blood cell count and mild anaemia (Table 1 ). Thirty-one of the dogs (91%) had clinical evidence of bleeding at the time of presentation, with the median number of clinical signs of bleeding being lower than the AMTP group. The median duration of hospitalisation was similar for the two groups (Table 1 ). For dogs that had platelet recovery (n=27), it took a median of 5 days (range 2 to 14) to achieve a platelet count greater than 40,000/ µ L and a median of 10 days (range 2 to 42) to achieve a platelet count within the reference interval. Five dogs did not achieve a platelet count greater than 40,000/ µ L, and none of these dogs survived to discharge. Information on time to platelet recovery was not available for two dogs, but both of these dogs survived to discharge. Infectious disease testing Five of the seven dogs in the AMTP group were tested for vectorborne infectious diseases including Ehrlichia canis (n=5), Babesia canis (n=5), Rickettsia rickettsii (n=5), Borrelia burgdorferi (n=4), Anaplasma phagocytophilum (n=3), Bartonella henselae and B. vinsonii (n=1), Anaplasma platys (n=1 ) and Neoricketssia risticii (n=1). Among those tested, all had negative antibody titres except one dog that had a positive titre at 1:40 for Rickettsia rickettsii. All of the dogs with pimt were tested for vector-borne infectious diseases, including E. canis (n=34), R. rickettsii (n=33), B. canis (n=29), B. burgdorferi (n=28), A. platys (n=25), A. phagocytophilum (n=13) and N. risticii (n=8). All tested dogs had negative antibody titres except for three dogs that were tested for B. burgdorferi. One dog was positive at 1:64, and a Western blot suggested this was consistent with vaccination. The second dog was positive at 1:40, and convalescent titres performed 46 days later were unchanged at 1:40. The third dog was positive at greater than 1:64, but again, convalescent titres 2 weeks later showed no change in serum titre. Survival data Six of the seven AMTP dogs did not survive to discharge. The median survival time was 6 days. All of the AMTP dogs that did not survive were euthanased because of failure to respond to treatment and concern for a poor prognosis. The one surviving dog responded to treatment with intravenous (IV) 0 2 mg/ kg dexamethasone daily followed by 3 4 mg/kg oral prednisone daily and 11 4 mg/kg doxycycline daily for 14 days. One infusion of 0 57 gm/kg iv immunoglobulin was also administered. This dog was euthanased 868 days after presentation for unknown reasons. Five of the 34 pimt dogs (15%) did not survive to discharge. The median survival time for these dogs was 19 days. Four dogs were euthanased because of perceived poor prognosis, and one dog expired. Of the surviving dogs, 14 dogs (41%) survived greater than 1 year. Eleven dogs (32%) survived between 31 days and 364 days. Two dogs (6%) survived less than 1 month, and survival data was not available for two dogs (6%). DISCUSSION The clinical presentation for dogs with AMTP and pimt was similar in most regards, but the AMTP group had more severe anaemia, more clinical signs of bleeding and a greater proportion of dogs required blood transfusions. Prognosis for dogs with AMTP was grave, while dogs with pimt had a much better prognosis. Both groups of dogs were severely thrombocytopenic at presentation, with a median platelet count of less than 1000/ µ L, but dogs with AMTP were more anaemic at presentation than dogs with pimt. This finding could represent either more prolonged duration or increased severity of clinical bleeding in dogs with AMTP. As both groups had low platelet counts, they should both have been at high risk for bleeding. The reason for this disparity is not clear. An increased tendency for bleeding in dogs with pimt has been thought to be due to concurrent platelet dysfunction (Stuart et al. 1981, Kristensen et al. 1994, Lewis & Meyers 1996 ), but this has not been investigated in dogs with AMTP. The dogs in the AMTP group also had more clinical signs of bleeding, again suggesting that bleeding was more likely in the AMTP group. The lower HCT, increased clinical signs of bleeding and higher percentage of dogs requiring blood transfusions in the AMTP group all suggest that this group of dogs had a more severe clinical course than the pimt group. The lack of platelet recovery in all but one of the dogs with AMTP reflected the decreased survival times compared to the pimt group. Six of the seven AMTP dogs did not survive to discharge, indicating a grave prognosis. In the one dog that survived, testing did not support a diagnosis of a vector-borne infectious disease, but convalescent titres were not available. The patient was treated with doxycycline, so a vector-borne disease could not be ruled out. A bone marrow aspirate was performed in this dog but not a bone marrow core. The bone marrow aspirate was of excellent diagnostic quality with very low numbers of immature megakaryocytes. Although a core sample was not obtained, the Journal of Small Animal Practice Vol 57 March 2016 2016 British Small Animal Veterinary Association 145

S. A. Cooper et al. quality of the aspirate was sufficient to indicate overall decreased megakaryocyte numbers in this case. AMTP is rare in dogs (Murtaugh & Jacobs 1985, Lachowicz et al. 2004 ). This was also true in our study, with only seven cases being identified over 20 years. Of the five previous cases reported, three dogs survived. Infectious disease ( E. canis and B. burgdorferi ) was suspected in two of these three dogs and, in one, an infectious cause could not be excluded because only acute Ehrlichia serology was performed (Murtaugh & Jacobs 1985 ). All three dogs with suspected infectious disease as a cause of AMTP were treated with prednisone and tetracycline and survived. The two dogs with AMTP that had negative infectious disease testing did not survive. Williams & Maggio-Price ( 1984 ) described four dogs with severe megakaryocytic hypoplasia as part of a larger study of IMT, and all four died. Putsche & Kohn ( 2008 ) described 30 dogs with IMT with one of the dogs having decreased megakaryocytes on a bone marrow aspirate. This dog did survive, taking 13 days to achieve a platelet count greater than 150,000/ µ L. Infectious disease testing results were not available in this dog. Another study investigating treatment and predictors of outcome in 73 dogs with primary and secondary IMT described 11 dogs with bone marrow evaluation (Rozanski et al. 2002 ). These dogs had longer platelet recovery times, and treatment failure prompted the bone marrow examination in most of the 11 patients. Four of the dogs had absolute or relative megakaryocytic hypoplasia, and three of the dogs survived to discharge. In that study, the number of megakaryocytes observed on bone marrow evaluation did not correlate with survival, but infectious disease testing for these 11 dogs was not discussed, making direct comparison with our study difficult. Miller & Lunn ( 2007 ) described 7 of 58 dogs with severe thrombocytopenia as having megakaryocytic hypoplasia. These dogs did not have a worse prognosis in terms of survival, time to discharge or normalisation of platelet count compared to the other dogs in the study; however, the study included dogs with all causes of thrombocytopenia, again making direct comparison difficult. In the present study, AMTP was believed to be primary in all seven dogs, although vector-borne infectious disease could not be completely excluded. This suggests that the prognosis is very poor for dogs with primary AMTP. There were several limitations to this study including the small sample size of the AMTP group. Due to the rarity of this disease, larger study populations would be very difficult to obtain, but a multi-centre study could increase recruitment of more patients. Also, given the paucity of AMTP cases in the literature, the true prevalence of the disease is not known. Due to the retrospective design of the study, infectious disease testing varied greatly in the AMTP group, so it was not possible to completely eliminate vector-borne disease as a cause of secondary platelet destruction. More information on platelet indices such as mean platelet volume, reticulated platelets or giant platelets would have been interesting to compare between the two groups, but mean platelet volume is not reported at our hospital, and archived blood smears were not available for review. Additionally, comments on platelet morphology were not consistent between patients. Although a few patients had giant or enlarged platelets noted on their CBC, a lack of comments could not be interpreted as a lack of giant or enlarged platelets, so comparisons could not be made. Anti-platelet antibody testing was not performed in any of the patients. Treatment methods also varied between patients of both groups, as with any observational study. All patients were treated with immunosuppression (dexamethasone and/or prednisone) and doxycycline. Other treatments generally included gastroprotectants with some patients receiving analgesics, vincristine, azathioprine, human IV immunoglobulin and antibiotics. Again, due to the lack of standardisation of treatment, comparison between groups or survivors and non-survivors could not be made. The results of this study suggest that the clinical presentation of AMTP is more severe than that of pimt. The presenting HCT and clinical signs of bleeding were more severe in the AMTP group, but there was overlap between the groups; therefore, evaluation of a bone marrow sample currently appears to be the only way to definitively distinguish between AMTP and pimt. Although the exact mechanism of primary AMTP has not been established in dogs, immunosuppressive therapy appears warranted because some dogs do appear to respond to this treatment. Given the favourable response to treatment in cases of AMTP suspected to be secondary to infectious causes, infectious disease testing should be pursued in all dogs with suspected IMT. Based on the findings of Miller & Lunn ( 2007 ) and the low prevalence of AMTP, bone marrow sampling may not be indicated in the initial work-up for severe thrombocytopenia; however, bone marrow evaluation should be considered earlier in thrombocytopenic dogs with more severe anaemia, those with severe clinical signs of haemorrhage and those that do not respond to initial treatment within 5 days. Conflict of interest None of the authors of this article has a financial or personal relationship with other people or organisations that could inappropriately influence or bias the content of the paper. References Agarwal, N., Spahr, J. E., Werner, T. L. et al. ( 2006 ) Acquired amegakaryocytic thrombocytopenic purpura. American Journal of Hematology 81, 132-135 Feldman, B. F., Thomason, K. J., Jain, N. C. ( 1988 ) Quantitative platelet disorders. Veterinary Clinics of North America: Small Animal Practice 18, 35-49 Huang, A. A., Moore, G. E., Scott-Moncrieff, J. C. ( 2012 ) Idiopathic immune-mediated thrombocytopenia and recent vaccination in dogs. Journal of Veterinary Internal Medicine 26, 142-148 Kristensen, A. T., Weiss, D. J., Klausner, J. S. ( 1994 ) Platelet dysfunction associated with immune-mediated thrombocytopenia in dogs. Journal of Veterinary Internal Medicine 8, 323-327 Lachowicz, J. L., Post, G. S., Moroff, S. D. et al. ( 2004 ) Acquired amegakaryocytic thrombocytopenia four cases and a literature review. The Journal of Small Animal Practice 45, 507-514 Lewis, D. C. & Meyers, K. M. ( 1996 ) Canine idiopathic thrombocytopenic purpura. Journal of Veterinary Internal Medicine 10, 207-218 Miller, M. D. & Lunn, K. F. ( 2007 ) Diagnostic use of cytologic examination of bone marrow from dogs with thrombocytopenia: 58 cases (1994-2004). Journal of the American Veterinary Medical Association 231, 1540-1544 Murtaugh, R. J. & Jacobs, R. M. ( 1985 ) Suspected immune-mediated megakaryocytic hypoplasia or aplasia in a dog. Journal of the American Veterinary Medical Association 186, 1313-1315 O Marra, S. K., Delaforcade, A. M., Shaw, S. P. ( 2011 ) Treatment and predictors of outcome in dogs with immune-mediated thrombocytopenia. Journal of the American Veterinary Medical Association 238, 346-352 Putsche, J. C. & Kohn, B. ( 2008 ) Primary immune-mediated thrombocytopenia in 30 dogs (1997-2003). Journal of the American Animal Hospital Association 44, 250-257 146 Journal of Small Animal Practice Vol 57 March 2016 2016 British Small Animal Veterinary Association

Amegakaryocytic thrombocytopenia Rozanski, E. A., Callan, M. B., Hughes, D. et al. ( 2002 ) Comparison of platelet count recovery with use of vincristine and prednisone or prednisone alone for treatment for severe immune-mediated thrombocytopenia in dogs. Journal of the American Veterinary Medical Association 220, 477-481 Silva, L. F. N., Golim, M. A., Takahira, R. K. ( 2012 ) Measurement of thrombopoietic activity through the quantification of megakaryocytes in bone marrow cytology and reticulated platelets. Research in Veterinary Science 93, 313-317 Stuart, M. J., Kelton, J. G., Allen, J. B. ( 1981 ) Abnormal platelet function and arachidonate metabolism in chronic idiopathic thrombocytopenic purpura. Blood 58, 326-329 Williams, D. A. & Maggio-Price, L. ( 1984 ) Canine idiopathic thrombocytopenia: clinical observations and long-term follow-up in 54 cases. Journal of the American Veterinary Medical Association 185, 660-663 Journal of Small Animal Practice Vol 57 March 2016 2016 British Small Animal Veterinary Association 147