Giant platelet disorder in the Cavalier King Charles Spaniel

Experimental Hematology 32 (2004) 344–350

Giant platelet disorder in the Cavalier King Charles Spaniel Sara M. Cowana, Joseph W. Bartgesa, Rebecca E. Gompf a, Jimmy R. Hayesa, Tamberlyn D. Moyersa, Carolyn C. Sniderb, David A. Gerardc, Robert M. Craftb, Robert A. Muenchend, and Roger C. Carrollb a University of Tennessee College of Veterinary Medicine, Knoxville, Tenn., USA; bDepartment of Anesthesiology, University of Tennessee Medical Center, Knoxville, Tenn., USA; cDepartment of Oral Surgery, University of Tennessee Medical Center, Knoxville, Tenn., USA; dUniversity of Tennessee Statistical Consulting Center (Muenchen), Knoxville, Tenn., USA

(Received 9 September 2003; revised 19 November 2003; accepted 9 January 2004)

Objective. The aim of this study was to describe the clinical, functional, and morphologic characteristics of platelets in Cavalier King Charles Spaniel dogs (Cavaliers). Materials and Methods. Blood from 69 clinically normal Cavaliers was collected and anticoagulated with ethylenediamine-tetraacetic acid (EDTA) and citrate. Automated and manual platelet counts were obtained. Percent platelet aggregation in response to ADP (2, 4, 8, 16, and 32 mM) was determined. Electron microscopy was performed to examine platelet internal morphology and dense granule distribution. A cardiologist recorded the quality of murmurs. Results. Thrombocytopenia (⬍100,000/mL) was present in 51.43% (36/69) of Cavaliers. Macrothrombocytes (⬎3 mm) were present in 33.33% (22/69). Mean manual platelet count was 118,770/mL. Manual (EDTA blood) and automated (EDTA and citrated blood) methods of platelet counting were correlated. Prevalence of cardiac murmurs was 38% (26/69). There was no association between affected dogs and murmur, signalment, or coat color. Mean percent platelet aggregation was significantly higher in controls than in Cavaliers (79% vs 38%, p ⫽ 0.001). Response to ADP was unaffected by thrombocytopenia, macrothrombocytes, murmur, or any combination thereof. Platelet electron microscopy showed normal and giant sized platelets with normal internal morphology. Conclusions. A benign inherited giant platelet disorder affects approximately 50% of Cavalier King Charles Spaniels. It is characterized by thrombocytopenia, macrothrombocytes, or decreased platelet aggregation in response to ADP. Platelet ultrastructure is normal. Citrated or EDTA blood provides accurate platelet counts. Further studies are indicated to determine platelet glycoprotein structure and any association with mitral endocardiosis. Cavaliers may be useful models of inherited giant platelet disorders. 쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.

Inherited giant platelet disorders (IGPD) in humans are characterized by their inheritance pattern, prevalence, clinical manifestations, laboratory findings, concurrent diseases, morphologic features of platelets and megakaryocytes, platelet function, and genetic or molecular defects [1]. Some disorders, such as Montreal platelet syndrome, May-Hegglin anomaly, Fechtner syndrome, Sebastian syndrome, hereditary macrothrombocytopenia, and Epstein syndrome are inherited through an autosomal dominant mechanism [1,2].

Offprint requests to: Sara M. Cowan, D.V.M., University of Tennessee College of Veterinary Medicine, C247 Department of Small Animal Clinical Sciences, Knoxville, TN 37901; E-mail: [email protected]

0301-472X/04 $–see front matter. Copyright doi: 1 0 .1 01 6 /j.ex p h e m.2 0 0 4. 01 .0 0 8

In dogs, only a few platelet disorders have a proved mode of inheritance and molecular defect (Table 1) [3–7]. Although an IGPD in Cavalier King Charles Spaniels (Cavaliers) is suspected, laboratory findings, concurrent diseases, morphologic features, and platelet function have not been thoroughly studied. Benign thrombocytopenia in Cavaliers is prevalent in Sweden and Denmark [8,9]. Pedersen and others proved thrombocytopenia in Cavaliers to be inherited through an autosomal recessive mechanism [8], similar to Bernard-Soulier syndrome [10], and macrothrombocytopenia associated with mitral valve regurgitation [11]. Nevertheless, inherited thrombocytopenia in Cavaliers is still not as widely recognized by veterinarians as other disorders.

쑖 2004 International Society for Experimental Hematology. Published by Elsevier Inc.

S.M. Cowan et al. / Experimental Hematology 32 (2004) 344–350

345

Table 1. Inherited canine platelet disorders Disorder von Willebrand disease Basset Hound hereditary thrombopathia Type I Glanzmann thrombasthenia (otterhounds) Cyclic hematopoiesis of gray collies Type III Glanzmann thrombasthenia (Landseers)

Mode of inheritance Autosomal dominant with variable penetrance Autosomal recessive Autosomal Autosomal Recessive Autosomal Recessive

One study ascribed the condition to automated analyzers mistaking giant platelets for erythrocytes [12]. Cavaliers have many other conditions that may be related to defective endothelium, including mitral valve disease (MVD) due to endocardiosis, femoral artery occlusion [13], and pulmonary artery fibrosis [14]. Whether these conditions are related is unknown. MVD could be associated with IGPD of Cavaliers. For example, a common genetic glycoprotein defect on the valvular endothelium and the platelet could cause each condition; a diffuse inherited vascular defect could damage both the mitral valve and platelets; or primary MVD could cause shear stress damage to platelets during regurgitation. The IGPD of Cavaliers has not been fully characterized. The aim of this study was to determine the prevalence of thrombocytopenia and macrothrombocytes, report clinical findings in affected Cavaliers, compare the various methods of platelet counting, and study the aggregation responses and ultrastructural characteristics of platelets.

Methods Study population Power calculations performed prior to the onset of the study revealed that 40 subjects would be required for a power of 0.8. Seventy-one privately owned Cavaliers were recruited publicly and through breed-enthusiast newsletters between May 2001 and May 2002. Inclusion criteria were that each Cavalier was clinically healthy, weighed more than 5 kg, and was fasted for 12 hours prior to blood collection. Exclusion criteria were a history of thrombocytopathia, bleeding disorders, current illness, physical examination abnormality (with the exception of mitral endocardiosis), any dog taking medications within 12 weeks of the study (with the exception of enalapril or heartworm preventative), and platelet aggregates on examination of blood smears. In addition, four healthy control research dogs (Beagles, female spayed, aged 5–10 years) were included for aggregation studies. Upon arrival at the University of Tennessee College of Veterinary Medicine, the owner completed a questionnaire and complete history for each dog. Dogs originated from the United States, Canada, and western Europe. A complete physical examination of each dog was performed by a veterinary internist (S.M.C.), and a veterinary cardiologist (R.E.G.) ausculted each dog for the presence of a cardiac murmur consistent with mitral valve insufficiency. Informed consent was obtained from each owner. The protocol was approved by the Institute of Animal Care and Use at the University of Tennessee.

Molecular defect Decreased concentration of van Willebrand factor multimers Altered regulation of cAMP-phosphodiesterase Quantitative defect in GPIIb/IIIa receptor due to GPIIb gene mutation B3A subunit of AP3 defect Qualitative defect in GPIIb/IIIa receptor

Blood collection Using a modified two-syringe method, jugular venipuncture was performed with an 18-gauge butterfly catheter to establish 3 mL free-flowing blood. A 12-mL syringe was attached and filled while maintaining venipuncture. Two more 12-mL syringes were used in succession to obtain a total of 30-mL of blood. Appropriate blood volumes were transferred immediately from the first syringe into a 5-mL tube containing 15% potassium ethylenediaminetetraacetic acid (EDTA), and a 5-mL tube containing 105-mM (3.2%) trisodium citrate. The two remaining syringes contained 136 mM trisodium citrate and 110 mM glucose (1 vol) as the anticoagulant for platelet aggregation and electron microscopic studies. Samples were maintained at 37⬚C and analyzed within 3 hours of collection. Complete blood cell counts A complete blood count was performed using each sample (i.e., EDTA and citrate anticoagulants) on a Cell-Dyn 3500 automated hematology analyzer (Abbott Inc., Abbott Park, IL, USA). A 20µL aliquot of EDTA-anticoagulated blood was mixed with 380-µL stromalytic agent, and the manual platelet count was determined in duplicate by a single hematology technologist using hemacytometer chambers and phase contrast microscopy. Hematoxylin and eosin stained blood smears were examined by clinical pathology technicians at the University of Tennessee College of Veterinary Medicine for platelet aggregates, macrothrombocytes, and erythrocyte morphology on both blood samples from each patient. Thrombocytopenia was defined as a manual platelet count less than 100,000/µL [8,9]. Macrothrombocytes were defined as greater than 30% of platelets having a diameter of at least 3 µm on microscopic examination [12]. Platelet aggregation Platelet-rich plasma (PRP) and platelet-poor plasma (PPP) were prepared by differential centrifugation [6]. Platelet aggregation was performed by standard procedures with a Chronolog dual channel aggregometer (Chronolog Corporation, Havertown, PA, USA). Personnel performing platelet aggregation (C.C.S. and R.C.C.) were blinded with regard to the platelet count and the presence of giant platelets in each sample. A 495-µL aliquot of PRP without adjustment of platelet concentration (1–4.0 × 108 platelets/mL) was used in all experiments. Platelet activation by ADP was initiated by addition of a 5-µL aliquot of 100-fold concentrated stock solution made up with physiologic saline to a constantly stirred (1,000 rpm) cuvette thermostated at 37⬚C. Percent aggregation was calculated from the maximal change in light transmission, from 0% aggregation baseline of PRP, during the 5 minutes after adding ADP and taking PPP light transmission as equivalent to 100% aggregation. Established reference ranges

346

S.M. Cowan et al. / Experimental Hematology 32 (2004) 344–350

for healthy dogs [15] as well as four healthy control dogs were used for comparison. Electron microscopy Citrated whole blood was used for three different electron microscopy studies. Personnel performing electron microscopy (D.A.G.) was blinded with regard to the platelet count and the presence of giant platelets in each sample. PRP was prepared by differential centrifugation. This sample was fixed for 2 hours in 2.5% glutaraldehyde in 0.1 M phosphate buffer (pH 7.2) at room temperature with a volume of 1:2 PRP to fixative. Platelets were centrifuged at 200g for 15 minutes, washed in buffer, and divided into three tubes for further processing. One tube was prepared for standard transmission electron microscopy (TEM) observation by postfixation in 2% osmium tetroxide in 0.1 M phosphate buffer for 1 hour, followed by washing, dehydration, and embedding in Spurr’s plastic. Platelets from a second tube were air dried onto formvar/carbon-coated grids for observation with TEM of dense granules [16]. Platelets from the third tube were processed for scanning electron microscopy (SEM) by mounting platelets onto L-lysine coated coverslips, dehydration, and critical point drying. These coverslips then were coated with 100 to 150 A of goldpalladium. TEM observations of sectioned platelets and whole mounts were made with a Hitachi 800 immediate voltage TEM, and SEM observations were made with a LEO 435 analytical SEM. For comparison of platelet size, 12 Cavaliers were selected by one author (S.M.C.) and divided into groups based on manual platelet count and light microscopic platelet size, respectively, as follows: normal, normal (group 1); normal, giant (group 2); decreased, normal (group 3); or decreased, giant (group 4). Random fields of platelets from specimens were photographed with TEM at 5,000×. Approximately 8 to 10 platelets were included on each negative. To ensure that median sections of platelets were measured, each platelet had at least 2 granules, 2 mitochondrial profiles, and 2 vacuoles of the surface canalicular system. Each platelet was measured in two directions at 90⬚ to each other. Measures were done directly from the negatives using a modified IMAGE analysis system (National Institute of Health, Bethesda, MD, USA). Two hundred platelets from each sample were measured using this technique, again in a blinded fashion. Statistical analysis Variables were analyzed using the SPSS statistical software, version 11.0 (SPSS Inc., Chicago, IL, USA). “Affected” dogs were defined as dogs having thrombocytopenia, macrothrombocytes, or both. The platelet counts and distributions from each method were examined using histograms and scatterplot matrix. Nonparametric Spearman correlations then were calculated to determine the strength of their relationships. To examine the possibility that the automated method miscounted one blood cell for another, the Student’s t-test was used to compare mean hematocrit (HCT) and total white blood cell count (WBC) in normal and affected dogs. The Chi-square test was used to detect any association between affected dogs and murmur (present or absent), gender (male or female), neuter status (intact or neutered), and coat color (blenheim, tricolor, ruby, or black and tan). A three-way repeated measures analysis of variance was used with thrombocytopenia, macrothrombocytes, and murmur status as the between-subjects effects and ADP level as the withinsubjects factor. The average number of dense granules observed in 50 whole mount single platelets selected at random was calculated.

Results Sixty-nine Cavaliers were included in the study. Two dogs were excluded from analysis due to excessive platelet clumping and a diastolic murmur. Mean age was 4.9 years (SD 3.51; range 0.5–17). There were 28 males (41%) and 41 females (59%). Each of the four coat patterns of the breed was represented: 33 blenheim, 18 tricolor, 9 ruby, and 9 black and tan. Prevalence of thrombocytopenia was 51.43% (36/69), and prevalence of macrothrombocytes was 33.33% (22/ 69). Mean manual platelet count was 118,770/µL (SD 78,188; median 107,500; range 20,000–305,000/µL). Platelet count was positively skewed; therefore, a Mann-Whitney test was performed to compare platelet size with platelet count. The Cavaliers with giant platelets had a significantly lower platelet count (median 63,500/µL) than did the Cavaliers with normal sized platelets (median 118,000/µL; MannWhitney U ⫽ 290, p ⫽ 0.011). Prevalence of cardiac murmurs was 38% (26/69). All murmurs were consistent with mitral regurgitation on the basis of auscultation; one dog had a concurrent right-sided murmur consistent with tricuspid regurgitation. The Chisquare test showed that the prevalence of murmurs was the same for normal and affected dogs (χ2 ⫽ 0.155, p ⫽0.693). There was no association between affected dogs and gender (χ2 ⫽ 0.194, p ⫽ .693), reproductive status (intact vs neutered, χ2 ⫽ 0.656, p ⫽ 0.418), or coat color (χ2 ⫽ 0.378, p ⫽ 0.945). Mean HCT (mean EDTA: 43.6%, SE 0.413; mean citrate: 36.7%, SE 0.388) and WBC count (mean EDTA: 10,106/ µL, SE 390; mean citrate: 8,567/µL, SE 323) were normal in all dogs. None of these measures showed a significant relationship to platelet count or size. The platelet count distributions using each of the three methods (manual, automated citrate, automated EDTA) were examined using histograms and were found to vary from positively to negatively skewed. A scatterplot matrix showed that these measures were related linearly. Nonparametric Spearman correlations then were calculated to determine the strength of their relationships, and all methods of platelet counting were well correlated (Table 2). Macrothrombocytes occurred in both the EDTA and citrated samples of each dog having large platelets. A two-way repeated measures analysis of variance was used with Cavaliers (n ⫽ 69) vs beagle control dogs (n ⫽ 4) Table 2. Spearman correlation coefficients of platelet counting methods Method

Manual

Automated EDTA

Automated citrate

Manual Automated EDTA Automated citrate

1.000 0.796∗ 0.789∗

0.796∗ 1.000 0.984∗

0.789∗ 0.984∗ 1.000

∗Correlation is significant at the 0.01 level (two-tailed).

S.M. Cowan et al. / Experimental Hematology 32 (2004) 344–350

as the between-subjects effect and ADP level as the withinsubjects factor. Mauchley’s test of sphericity was significant (χ2 ⫽ 58.34, df ⫽ 9, p ⬍ 0.001), so the Greenhouse-Geisser version of the F-tests were used. Mean percent platelet aggregation was significantly higher in the beagles and historical controls (mean 78.90%) than in Cavaliers (mean 38.14%, F(1,54) ⫽ 11.45, p ⫽ 0.001). Both groups showed increased platelet aggregation in response to increasing concentrations of ADP (F(2.7,147.6) ⫽ 23.87, p ⬍ 0.001; Fig. 1). The difference in mean percent aggregation between groups did not depend upon the concentration of ADP (F(2.7,147.6) ⫽ 0.72, p ⫽ 0.53). A three-way repeated measures analysis of variance was calculated for the Cavaliers, with affected and nonaffected group and murmur status as the between-subjects effects and ADP level as the within-subjects factor. Mauchley’s test of sphericity was significant (χ2 ⫽ 44.26, df ⫽ 9, p ⬍ 0.001), so again the Greenhouse-Geisser version of the F-tests were used. Response to ADP was not different between affected dogs, those with a murmur, or any combination thereof. Post hoc comparisons using the Bonferroni method of correction found that each ADP level differed significantly from the others (p ⬍ 0.001). The assumption of homogeneity of variance was verified using Levene’s test. The residuals were found to be normally distributed using a histogram. Routine TEM of platelets showed normal morphology in all Cavaliers (Fig. 2). Platelets had α granules, mitochondria, dense granules, and evidence of the canicular system. These characteristics were seen in both normal and giant sized platelets. On whole mount preparations, there were 0 to 9 dense granules per platelet (average 5.1). This average did not vary significantly from dog to dog. On SEM, all platelets were smooth discs; there were no irregular shapes or filopodia to suggest improper handling of samples [17]. The average platelet size using TEM is listed in Table 3. There was no significant difference in the average, range, or mean size of platelets between groups. Subjectively,

Figure 1. Mean percent platelet aggregation (y-axis) in response to increasing ADP concentrations (x-axis) in control and Cavalier dogs.

347

some samples appeared to contain extremely giant platelets (⬎6 µm diameter): 14.6% of platelets from dogs in groups 2 and 4 were extremely giant, compared with 3% of platelets from dogs in groups 1 and 3. Discussion About half of the Cavaliers in this study had thrombocytopenia, which was similar to the reported prevalence in Sweden and Denmark (31%-56%) [8,9]. The high prevalence of IGPD in Cavaliers may be a result of the benign nature of the disorder, breeding of affected dogs due to the assumption that the disorder was artifactual, or that the breed was created by a small number of inbred dogs, increasing the prevalence of genetic disorders. Until now, all known inherited platelet disorders in the dog have been associated with prolonged bleeding times [4]. All Cavaliers in this study were clinically normal, had no other hematologic abnormalities, and showed no signs of disrupted primary hemostasis after blood collection. In this study, the disorder was present in Cavaliers regardless of gender, age, neuter status, or coat color, although a previous study found that males had a significantly lower platelet count than females [9]. This IGPD of Cavaliers is clinically benign with regard to bleeding, similar to the giant platelet syndrome associated with mitral valve regurgitation, benign Mediterranean macrothrombocytopenia, and, sometimes, Bernard Soulier syndrome [1]. The platelet count and size reported for IGPD in humans vary. In our study, there was variability among affected dogs as to whether they were affected by thrombocytopenia alone (20/43), macrothrombocytes alone (8/43), or both (15/43). These findings may reflect the heterogeneous expression of the disease, influence of unknown factors on the expression of the disease, or a progressive nature of the disease over time. These variable findings in affected dogs may have complicated the detection of any relationship with the heritable and clinical features examined, as there was no difference between affected and unaffected Cavaliers with respect to age, gender, neuter status, or murmur. Further studies are indicated to evaluate affected dogs over time and to continue to study possible factors that could affect expression of the disorder. The mechanism of developing macrothrombocytes in IGPD of Cavaliers is unknown. Bernard-Soulier syndrome is the only IGPD in which a genetic mutation has been proved (frameshift deletion of the gene encoding platelet glycoprotein Ibα) [18]. Other theories as to why macrothrombocytes might develop in IGPD of Cavaliers include 1) abnormally increased thrombopoietic signaling to megakaryocytes; 2) abnormal megakaryocyte glycoproteins responsible for platelet release, resulting in large fragmentation; 3) persistent low-grade damage to circulating platelets by diffusely abnormal endothelium, resulting in decreased platelet lifespan; and 4) shear stress through valvular lesions, resulting in decreased platelet lifespan. The absence of other markers of

348

S.M. Cowan et al. / Experimental Hematology 32 (2004) 344–350

Figure 2. Transmission electron micrography of a representative giant sized (A) and normal sized (B) platelet from a Cavalier, each showing normal morphology (magnification ×20,000; scale ⫽ 1 µm).

endothelial damage (bleeding tendencies, abnormal erythrocyte morphology) makes a primary endothelial disorder less likely but would not completely exclude the possibility. Normal bone marrow and megakaryocyte morphology using hematoxylin and eosin stained marrow cytology have been reported in Cavaliers with thrombocytopenia [19]. These Cavaliers were most likely affected by IGPD, because extensive diagnostic evaluation for acquired causes of thrombocytopenia was negative. Further studies are indicated to examine platelet and megakaryocyte glycoprotein structure and thrombopoietic responses and to determine platelet lifespan in Cavaliers. Canine platelets exposed to EDTA exhibited artifactual increases in light microscopic size in one study, and it was recommended that a citrated specimen maintained at 37⬚C be used to determine platelet size [17]. In our study, samples were maintained at 37⬚C and analyzed within 3 hours of collection. There was 100% correlation in recognizing macrothrombocytes between EDTA and citrate samples. Therefore, giant platelets were consistently detected in Cavaliers

using blood smears anticoagulated with either citrate or EDTA. Platelet count is affected by specimen collection, sample handling, splenic activity, methodology, human error, inherent 20% to 25% variation in platelet counts, and, with automated methods, microcytosis and macrocytosis of blood cells [20]. A study of 14 Cavaliers tested the hypothesis that thrombocytopenia of Cavaliers was artifactual; automated platelet counts were significantly less than manual platelet counts, but this also occurred in control dogs, raising concern over the accuracy of the automated method [12]. In this study, specimen collection, handling, and methodology were consistent for each dog. No other cell lines were abnormal, and a single hematology technologist determined the manual platelet count in duplicate. Therefore, the thrombocytopenia in the Cavaliers of this study was real, not artifactual. A limitation of this study was that common causes of acquired thrombocytopenia were not systematically ruled out. After excluding dogs with historical (tick exposure, vaccination within 3 months, drug history) or physical

S.M. Cowan et al. / Experimental Hematology 32 (2004) 344–350 Table 3. Average platelet size, including range and mean, for a subset of Cavaliers

Animal no. Group 1 1026 1028 1030 Group 2 1034 1096 2038 Group 3 2050 1016 1024 Group 4 1046 2040 1097 Groups 1 and 3 combined

Average measure (µM)

Range

2.2 2.3 2.3

1.2–9.5 1.3–9.7 1.2–8.7

2.6 2.5 2.6

6–3% 8–4% 5–2.5%

2.8 2.8 2.7

1.2–9.8 1.4–9.3 1.2–10.2

2.9 3.1 3.0

31–15.5% 28–14% 27–13.5%

2.4 2.3 2.3

1.1–9.6 1.4–10.1 1.1–8.9

2.4 2.7 2.4

5–2.5% 9–4.5% 4–2%

2.9 2.8 3.0

1.1–9.5 1.3–9.8 1.4–9.9

2.8 2.9 3.0

31–15.5% 34–17% 25–12.5%

2.4

1.1–10.1

2.5

37–3%

2.7

1.1–10.2

2.95

Mean

Platelets ⬎6 µM

Groups 2 and 4 combined 176–14.6%

Dogs are grouped according to manual platelet count and light microscopic size as follows: normal, normal (group 1); normal, giant (group 2); decreased, normal (group 3); decreased, giant (group 4). Transmission electron microscopy (× 5,000).

(petechiation, polyarthritis, pale mucous membranes, fever) markers of acquired thrombocytopenia, and on the basis of previous studies ruling out acquired causes in the breed [19], we believe it to be very unlikely that any Cavalier was affected by a pathologic acquired thrombocytopathy. MVD is common in Cavaliers. Studies on the prevalence of MVD suggest that 30% to 50% of Cavaliers are affected by age 5 [21]. MVD in Cavaliers is a polygenic threshold trait with possible multifactorial environmental influence [21]. There is no proven relationship between MVD and thrombocytopathy in dogs. In this study, Cavaliers did not undergo echocardiography to confirm that murmurs ausculted were due to MVD. However, most Cavaliers affected with MVD can be detected by auscultation [22]. Furthermore, the probability of misclassifying other cardiac disease as MVD by auscultation is very low in breeds with a high prevalence of MVD [23]. An association between MVD and an inherited giant platelet disorder has been identified in humans [11]. Whether these two conditions in Cavaliers are causally or genetically related remains to be determined. Testing the hypothesis that MVD is causally or genetically associated with the platelet disorder will require comparison of the histology of the mitral valve in Cavaliers with and without IGPD, as well as detailed genetic characterization of both disorders. Platelet responses to known aggregating agents is proposed to be a reliable measure of platelet function and of

349

severity of disease in inherited platelet disorders. Unfortunately, there is no standardized method of performing platelet aggregation studies in dogs, and there are few and conflicting results in canine platelet aggregation studies [24]. In this study, Cavaliers had decreased platelet aggregation regardless of platelet count, size, or murmur status. Dogs with MVD also have been shown to have decreased platelet aggregation [24], and Cavaliers with MVD showed a prolonged time to form a platelet plug [25]. In contrast, humans [26] and Cavaliers with mitral valve prolapse show increased aggregation, whereas Cavaliers with thrombocytopenia or macrothrombocytes had normal platelet aggregation (Table 4) [27]. One explanation for the variability among studies in canines platelet aggregation responses is that decreased platelet aggregation may be an independent characteristic of the IGPD, or that concurrent subclinical MVD is present without a murmur. Differences in methodologies also may account for the discrepant findings. It has been suggested that, for canine aggregation studies, platelet numbers be adjusted to a final concentration of 150,000/µL and mixed 1:1 with isotonic saline before adding the platelet agonist [28]. This methodology has not been studied for canine PRP as used in the present study and, in our opinion, may not reflect in vivo activity. Ultrastructurally, platelet internal morphology was normal. The internal features of the giant platelets were identical to those of the normal sized platelets, making an internal granule, mitochondrial, or cytoskeletal defect unlikely as the cause of the IGPD. On TEM, Cavaliers with giant platelets on light microscopy (⬎3 µm) had a greater percentage of platelets that were extremely giant (⬎6 µm). The power of the TEM study was insufficient to draw any conclusions on the significance of this finding. Whether these platelets represent abnormal megakaryocytic function, an increased platelet biomass to compensate for fragmentation in the circulation, or a unique quality of the inherited defect requires further investigation. A classification scheme recently proposed for inherited giant platelet disorders in humans suggested that each disorder be classified under one of four groups, although overlap may exist: 1) structural defects including glycoprotein abnormalities noted on platelet surface protein electrophoresis; 2) intracytoplasmic granule defects or abnormal neutrophil inclusions; 3) concurrent systemic disease such as deafness, cataracts, nephritis, or MVD; and 4) benign macrothrombocytopenia [1]. The IGPD of Cavaliers may be grouped under a benign macrothrombocytopenia, but further characterization is indicated to determine glycoprotein structure and to determine any association with systemic disease, especially MVD. From this study, the benign, autosomal recessive IGPD of Cavaliers was further characterized as a variably expressed disease. Thrombocytopenia affected approximately 50% of the breed, macrothrombocytes affected an overlapping 30% of the breed, and platelet aggregation in response to ADP was decreased compared with non-Cavalier control dogs

350

S.M. Cowan et al. / Experimental Hematology 32 (2004) 344–350

Table 4. Pairwise comparisons of platelet aggregation in response to increasing concentrations of ADP in control dogs vs Cavaliers

ADP (µM)

Group 1

Group 2

Mean Difference Groups 1 and 2 (%)

2 4 8 16 32

Control Control Control Control Control

Cavaliers Cavaliers Cavaliers Cavaliers Cavaliers

38.885∗ 46.192∗ 51.135∗ 33.808 33.788∗

95% confidence interval for difference† SE

Sig.†

Lower bound (%)

Upper bound (%)

7.700 13.948 17.714 16.936 14.200

0.000 0.002 0.006 0.051 0.021

23.446 18.228 15.620 ⫺0.146 5.320

54.323 74.156 86.649 67.762 62.257

Based on estimated marginal means. ∗Mean difference is significant at the 0.05 level. † Adjustment for multiple comparisons: least significant difference (equivalent to no adjustments).

regardless of platelet count, size, or presence of a murmur. Internal morphology was normal in all giant and normal sized platelets. Cavalier King Charles Spaniels may be used as a model of inherited giant platelet disorders in humans.

Acknowledgments This work was supported by grants from The American Kennel Club, The Cavalier Health Foundation, and The University of Tennessee College of Veterinary Medicine. Previously presented at the American College of Veterinary Internal Medicine Forum, Dallas, Texas, May 2002. The authors thank Marilyn Cottrell, B.S., for technical assistance, Beth Bales for recruiting participants, and all pets and pet owners for participating in the study.

References 1. Mhawech P, Saleem A. Inherited giant platelet disorders. Classification and literature review. Am J Clin Pathol. 2000;113:176–190. 2. Jantunen E. Inherited giant platelet disorders. Eur J Haematol. 1994; 53:191–196. 3. Riehl J, Okura M, Mignot E, Nishino S. Inheritance of von Willebrand’s disease in a colony of Doberman Pinschers. Am J Vet Res. 2000;61: 115–120. 4. de Gopegui RR, Feldman BF. Acquired and inherited platelet dysfunction in small animals. Comp Cont Ed Sm An Prac. 1998;20:1039–1052. 5. Boudreaux MK, Catalfamo JL. Molecular and genetic basis for thrombasthenic thrombopathia in otterhounds. Am J Vet Res. 2001;62: 1797–1804. 6. Boney CM, McDonald TP, Jones JB. Abnormal function and thromboxane release from platelets of dogs with cyclic hematopoiesis. Exp Hematol. 1985;13:586–590. 7. Katen LJ, Aprikyan AG, Dale DC, Osborne WR. Cloning and sequencing of the canine neutrophil elastase cDNA. DNA Seq. 2002;13: 221–223. 8. Pedersen HD, Haggstrom J, Olsen LH, et al. Idiopathic asymptomatic thrombocytopenia in Cavalier King Charles Spaniels is an autosomal recessive trait. J Vet Intern Med. 2002;16:169–173. 9. Eskell P, Haggstrom J, Kvart C, Karlsson A. Thrombocytopenia in the cavalier King Charles spaniel. J Small Anim Pract. 1994;35:153–155. 10. Lopez JA, Andrews RK, Afshar-Kharghan V, Berndt MC. BernardSoulier syndrome. Blood. 1998;91:4397–4418. 11. Becker PS, Clavell LA, Beardsley DS. Giant platelets with abnormal surface glycoproteins: a new familial disorder associated with mitral valve insufficiency. J Pediatr Hematol Oncol. 1998;20:69–73.

12. Brown SJ, Simpson KW, Baker S, Spagnoletti MA, Elwood CM. Macrothrombocytosis in cavalier King Charles spaniels. Vet Rec. 1994; 135:281–283. 13. Buchanan JW, Beardow AW, Sammarco CD. Femoral artery occlusion in Cavalier King Charles Spaniels. J Am Vet Med Assoc. 1997;211: 872–874. 14. Karlstam E, Haggstrom J, Kvarat C, Jonsson L, Michaelsson M. Pulmonary artery lesions in cavalier King Charles spaniels. Vet Rec. 2000; 147:166–167. 15. Nolte I, Mischke R. Investigations of platelet aggregation and platelet counts from stored canine whole blood. Res Vet Sci. 1995;58:190–192. 16. White JG. The dense bodies of human platelets: inherent electron opacity of the serotonin storage particles. Blood. 1969;33:598–606. 17. Handagama P, Feldman BF, Kono C, Farver T. Mean platelet volume artifacts: the effect of anticoagulants and temperature on canine platelets. Vet Clin Pathol. 1986;15:13–17. 18. Kunishima S, Kamiya T, Saito H. Genetic abnormalities of BernardSoulier syndrome. Int J Hematol. 2002;76:319–327. 19. Smedile LE, Houston DM, Taylor SM, Post K, Searcy GP. Idiopathic, asymptomatic thrombocytopenia in Cavalier King Charles spaniels: 11 cases (1983–1993). J Am Anim Hosp Assoc. 1997;33:411–415. 20. Prasse KW, Duncan JR. In: Latimer KS, Mahaffey EA, Prasse KW, eds. Duncan & Prasse’s Veterinary Laboratory Medicine and Clinical Pathology. 3rd ed. Ames, Iowa: Iowa State Press; 1994. pp. 3–36, 82–86. 21. Swenson L, Haggstrom J, Kvart C, Juneja RK. Relationship between parental cardiac status in Cavalier King Charles spaniels and prevalence and severity of chronic valvular disease in offspring. J Am Vet Med Assoc. 1996;208:2009–2012. 22. Pedersen HD, Haggstrom J, Falk T, et al. Auscultation in mild mitral regurgitation in dogs: observer variation, effects of physical maneuvers, and agreement with color Doppler echocardiography and phonocardiography. J Vet Intern Med. 1999;13:56–64. 23. Thrusfield MV, Aitken CG, Darke PG. Observations on breed and sex in relation to canine heart valve incompetence. J Small Anim Pract. 1985;26:709–717. 24. Tanaka R, Yamane Y. Platelet aggregation in dogs with mitral valve regurgitation. Am J Vet Res. 2000;61:1248–1251. 25. Tarnow I, Kristensen AT, Texel H, Pedersen HD. Decreased platelet function in dogs with clinically silent mitral valve regurgitation. Proc ESVIM. 2002;630. 26. Martini F, Zuppiroli A, Gori A, et al. Platelet and blood clotting activation in patients with mitral valve prolapse. Thromb Res. 1996; 83:299–306. 27. Olsen LH, Kristensen AT, Haggstrom J, et al. Increased platelet aggregation response in Cavalier King Charles Spaniels with mitral valve prolapse. J Vet Intern Med. 2001;15:209–216. 28. Juttner C, Rodriguez M, Fragio C. Optimal conditions for simultaneous measurement of platelet aggregation and ATP secretion in canine whole blood. Res Vet Sci. 2000;68:27–32.