Correspondence Noel Clancey, Department of Pathology and Microbiology, Atlantic Veterinary College, University of Prince Edward Island, 550 University Avenue, Charlottetown, PEI, Canada C1A 4P3 E-mail: email@example.com
Background: Cardiac disease has the potential to alter platelet function in dogs. Evaluation of platelet function using the PFA-100 analyzer in dogs of multiple breeds and with a broad range of cardiac conditions would help clarify the effect of cardiac disease on platelets.
Objectives: The objective of this study was to assess differences in closure time (CT) in dogs with cardiac disease associated with murmurs, when compared with that of healthy dogs.
Methods: Thirty-nine dogs with cardiac murmurs and turbulent blood flow as determined echocardiographically were included in the study. The dogs represented 23 different breeds. Dogs with murmurs were further divided into those with atrioventricular valvular insufficiency (n=23) and subaortic stenosis (n=9). Fifty-eight clinically healthy dogs were used as controls. CTs were determined in duplicate on a PFA-100 analyzer using collagen/ADP cartridges.
Results: Compared with CTs in the control group (mean±SD, 57.6±5.9 seconds; median, 56.5 seconds; reference interval, 48.0–77.0 seconds), dogs with valvular insufficiency (mean±SD, 81.9±26.3 seconds; median, 78.0 seconds; range, 52.5–187 seconds), subaortic stenosis (71.4±16.5 seconds; median, 66.0 seconds; range, 51.5–95.0 seconds), and all dogs with murmurs combined (79.6±24.1 seconds; median, 74.0 seconds; range, 48.0–187 seconds) had significantly prolonged CTs (P<.01).
Conclusions: The PFA-100 analyzer is useful in detecting platelet function defects in dogs with cardiac murmurs, most notably those caused by mitral and/or tricuspid valvular insufficiency or subaortic stenosis. The form of turbulent blood flow does not appear to be an important factor in platelet hypofunction in these forms of cardiac disease.
Platelets play a vital role in hemostasis, particularly in the primary stage. Dogs can experience altered platelet function with a variety of conditions, including cardiac disease.1–5 Evaluation of platelet hypofunction and hyperfunction has traditionally involved flow cytometry or platelet aggregation. These techniques are not routinely available in diagnostic laboratories and are labor-intensive. Point-of-care assessment of platelet hypofunction in dogs has usually relied on the mucosal bleeding time, but this test has several problems, including poor standardization and insensitivity to minor platelet defects.6–8
The PFA-100 is a point-of-care platelet function analyzer developed to evaluate human platelet hyperfunction and hypofunction.9–15 The device assesses primary hemostasis by simulating high shear platelet function at a site of endothelial damage by causing rapid in vitro flow of a citrated whole blood sample through an aperture in a collagen-coated membrane containing either ADP (COL/ADP) or epinephrine (COL/EPI). The time until platelet plug formation occurs is measured in seconds and is called the closure time (CT). The PFA-100 is easy to use and has recently been investigated and validated for use in dogs.16–18 Use of COL/ADP is recommended over COL/EPI, as epinephrine does not consistently induce platelet aggregation in dogs.19
Cardiac disease has the potential to alter platelet function in humans and dogs, but there is conflicting information as to whether hypofunction or hyperfunction is present. In earlier studies in humans with cardiac disease, platelets were noted to be in an activated state.20,21 The few studies reported in dogs have instead supported platelet hypofunction in cardiac disease.1–5 Tanaka et al4,5 showed a decreased platelet lifespan and decreased platelet activity in dogs with mitral valve regurgitation; this latter study used platelet aggregometry. In the few studies of dogs with cardiac disease assessed using the PFA-100 analyzer, Cavalier King Charles Spaniels (CKCS) with mitral valve disease had increased CTs compared with healthy dogs.1,3
Further evaluation of platelet function using the PFA-100 analyzer in dogs with a broader range of cardiac conditions and representing a wider range of breeds would be of interest in order to better understand the role that heart disease may play in populations undergoing hematologic assessment for clinical disease. The objective of the present study was to determine whether significant differences in CT exist in dogs with cardiac disease, as defined broadly by the presence of a murmur, when compared with that of healthy dogs.
Materials and Methods
All experimental protocols were designed and performed in accordance with the standards of the Canadian Council on Animal Care and were approved by the University of Prince Edward Island Animal Care Committee. Written owner consent was obtained for all dogs used in this study.
Fifty-eight healthy hospital staff- and client-owned dogs were used as the control group to which the CTs from dogs with cardiac disease were compared. The control group was also used to develop a laboratory-specific reference interval for CT. The control group included 37 clinically healthy dogs used in a preliminary study and 21 additional healthy dogs. All dogs were deemed healthy for inclusion in the study if they had no history of a bleeding tendency noted by the owner; had not received any medications known or suspected to interfere with platelet function or number within the 7 days before blood collection; had no abnormalities on physical examination, in a CBC, or in a serum biochemistry profile; had a platelet count of >150 × 109/L; had no evidence of platelet clumps on blood smears; and had a plasma von Willebrand factor antigen (vWF:Ag) concentration >50%.
Dogs with cardiac disease evaluated by the Atlantic Veterinary College Cardiology Service between June 2006 and September 2007 and found to have a heart murmur on auscultation and corresponding turbulent blood flow demonstrated echocardiographically were included in the study. Additional inclusion criteria for dogs with cardiac disease were a HCT of >0.35 L/L, a platelet count of >150 × 109/L, no evidence of platelet clumping on blood smears, and no history of having received any medications known to interfere with platelet function or number within 7 days before blood collection. Plasma vWF:Ag concentration was measured in all dogs in which the CT was greater than the upper limit of the reference interval for the control group.
Dogs with cardiac disease were grouped into 2 subgroups: dogs with mitral and/or tricuspid valvular insufficiency and dogs with subaortic stenosis. Cardiac diagnoses were established echocardiographically according to standard definitions.22,23 Subaortic stenosis was considered to be present when transaortic valvular velocity was >2.2 m/s (>2.0 m/s if there was concurrent supraphysiologic aortic insufficiency) and/or when left ventricular myocardial hyperechogenicity was observed. Left atrial size was assessed echocardiographically using the aortic valve area:left atrial area ratio.24 For dogs with tricuspid valve regurgitation, the same approach was used for determining the aortic valve area:right atrial area ratio using the right parasternal short axis and left apical 4-chamber echocardiographic views during earliest diastole (initial aortic valve closure). To quantitatively assess the turbulence created by heart lesions, we measured the surface area of turbulent jets (ATJ) identified through color-flow Doppler echocardiography. The ATJ was indexed to the area of the receiving atrium (ATJ:AA) as described previously,25 and to the cross-sectional area of the patient's aortic valve (ATJ:Ao), as calculated above. All Doppler measurements were made during peak systole. For dogs with valvular insufficiencies, correlation analysis was performed between CT and ATJ, ATJ:AA, and ATJ:Ao. For dogs with subaortic and pulmonic stenosis, correlation analysis was performed between CT and the transvalvular velocity and pressure gradient. The investigator (E.C.) performing the Doppler measurements was blinded to the heart murmur information for all patients until all measurements were finalized. When >1 measurement was available, such as multiple echocardiographic views, the mean of the measurements was calculated. Dogs with cardiac disease were also subdivided by murmur grade based on a standard I–VI scale and by gender.
Blood collection and handling
All blood samples were obtained by careful jugular or lateral saphenous venipuncture with collection directly into evacuated glass tubes using a 21 or 22 G needle vacuum system (Vacutainer, Becton Dickinson, Franklin Lakes, NJ, USA) or a 21 G butterfly catheter system (Becton Dickinson). Hemolyzed blood samples or samples that resulted in venous collapse or stoppage of blood flow were discarded and a fresh venipuncture from a different vein with a new needle was performed. Blood was collected into tubes containing potassium EDTA (Becton Dickinson) for CBCs, manual platelet estimates, and blood smear examination for platelet aggregates, as well as into plain tubes (Becton Dickinson) for serum biochemical analysis. Tubes containing 105 nM (3.2%) trisodium citrate (Becton Dickinson) were filled to the appropriate ratio of 1 part anticoagulant to 9 parts blood for platelet function analysis, blood smear examination for platelet aggregates, and vWF:Ag analysis. Tubes were filled with blood in sequential order of plain tubes first, then trisodium citrate tubes followed by K-EDTA tubes, according to a standard operating protocol.
Plasma for vWF:Ag testing was obtained following centrifugation of the citrate tubes at approximately 1300g for 5 minutes at room temperature (∼23°C) within 30 minutes of collection. Plasma was immediately stored in cryovials at −20°C for no longer than 21 days, then transferred to −80°C freezers until shipment. Analysis of vWF:Ag concentration was performed by the Comparative Coagulation Section Diagnostic Laboratory at Cornell University (Ithaca, NY, USA). Autostained (Ames Hematek-1000, Bayer Corp, Elkhart, IN, USA) blood smears of K-EDTA or citrate anticoagulated blood were used for manual platelet estimates and examined for platelet clumping. Automated platelet counts were performed on K-EDTA anticoagulated blood samples using a CELL-DYN CD-3500CS (Abbott Diagnostics, Santa Clara, CA, USA). If a K-EDTA anticoagulated sample was not submitted for a given dog, CBC and blood smear evaluation were performed using the citrated sample, with appropriate volume corrections. In the rare event of an erroneous automated platelet count, platelet estimates were determined from smears as reported previously.26 Only results from samples with platelet counts >150 × 109/L were included in the study.
Platelet function analysis
The PFA-100 analyzer (PFA-100, Dade Behring, Deerfield, IL, USA) was used according to the manufacturer's instructions. The PFA-100 has been described in detail elsewhere.27 Briefly, the device is a microprocessor-controlled instrument utilizing disposable test cartridges that contain a biologically active membrane. Test cartridges consist of a sample reservoir from which fresh citrated whole blood is aspirated under constant vacuum through a small diameter capillary tube ending at a small aperture within a collagen-coated membrane containing ADP. During operation, platelets adhere to the membrane, become activated, and degranulate upon agonist contact. Platelets then aggregate on the membrane surface surrounding the aperture, forming a plug that gradually occludes the aperture and eventually arrests blood flow. The analyzer measures the time from initial blood flow until flow arrest in seconds and this result is recorded as the CT. Errors during testing may result from air leaks, insufficient sample volume, or flow obstructions due to microthrombi or particulates introduced into the sample or test cartridge. Maintenance of the PFA-100 was performed according to the manufacturer's recommendations.28
Citrated whole blood samples used for platelet function analysis were stored at room temperature (∼23°C) and analyzed between 30 minutes and 4 hours of collection. Test cartridges were stored at 4°C and prewarmed at room temperature for at least 10 minutes before analysis.28 Immediately before analysis, blood was mixed by consistently performing 10 gentle inversions of the tubes by hand. Once at room temperature, test cartridges were inserted into the housing carousel of the analyzer and loaded with 800 μL of mixed citrated whole blood. All blood samples from all dogs were run in duplicate using COL/ADP cartridges; the analyzer automatically calculates a coefficient of variation (CV, %), as the SD/mean × 100. In the event of a flow obstruction in the PFA-100, a repeat analysis using a single test cartridge was performed. If the second analysis also resulted in a flow obstruction, a new blood sample was obtained and duplicate testing was repeated. In case of an air leak error, repeat duplicate testing was performed with new cartridges. If the CV was >15% between the duplicate results, repeat duplicate analysis using new test cartridges was performed.28
Statistical analyses were performed using statistical software (Minitab 14, Minitab Inc., State College, PA, USA). Data were evaluated for normality using descriptive statistics and normal probability plots. The CT reference interval was calculated nonparametrically as the central 0.95 percentile.29 A Mann–Whitney U-test was used to determine if significant differences existed between categories of dogs with cardiac disease compared with controls. A Mann–Whitney U-test was also used for comparing the subgroups of dogs with cardiac disease to the control group. Gender and murmur grades for dogs within the cardiac group were analyzed by using 1-way ANOVA. When the overall F-test of the ANOVA was significant, least squared means were compared by use of t-tests with a Tukey adjustment for multiple comparisons. Results were considered significant when P<.05.
Control group and reference intervals
There was no significant difference between CTs of the 2 healthy dog populations (P=.87), so data from the 21 dogs were merged with data from the 37 healthy dogs used in a preliminary study to achieve the control group of 58 dogs. These dogs ranged in age from 6 months to 13 years, represented 17 breeds, and included 2 intact females, 23 spayed females, 4 intact males, and 27 neutered males. One female and 1 male dog had unknown neuter status.
CTs for the 58 control dogs ranged from 48 to 79 seconds (Figure 1) (mean±SD, 57.6±5.9 seconds; median, 56.5 seconds). CVs ranged from 0% to 10% (3.79±2.66%, median 3.00%). A CT reference interval of 48–77 seconds was calculated.
Dogs with cardiac disease
Fifty-five dogs with cardiac disease were initially sampled. Sixteen dogs were excluded due to low vWF:Ag results (n=6), medication histories with exclusionary drugs (n=6), low HCT (n=3), or a combination of low vWF:Ag and drugs (n=1). All dogs with cardiac murmurs that had plasma vWF:Ag analysis performed (33/39) had vWF:Ag concentrations >55%. Six dogs with cardiac murmurs that did not have vWF:Ag analysis performed had CTs within the reference interval. The 39 dogs included in the study were 8 intact females, 9 spayed females, 4 intact males, and 16 neutered males that ranged in age from 4 to 168 months and ranged in weight from 4.6 to 56.8 kg. No significant differences in CT were found between gender groups (P=.989). Twenty-three breeds were represented, including 6 CKCS, 4 Golden Retrievers, 3 Boxers, 3 mixed breed, 2 of each Border Collie, Bichon Frise, Bull Mastiff, and Boston Terrier, and 1 of each Cocker Spaniel, Dachshund, Miniature Poodle, Shih Tzu, West Highland White Terrier, Australian Shepherd, Bull Terrier, German Shepherd, Greyhound, Labrador Retriever, English Springer Spaniel, Pomeranian, Afghan, Beagle, and Newfoundland.
The CT for dogs with cardiac murmurs ranged from 48 to 187 seconds (79.6±24.1 seconds; median, 74.0 seconds). Of the 39 dogs with cardiac disease, 21 had CTs within the reference interval and 18 had CTs above the upper limit. A significantly longer median CT was found in the group of dogs with cardiac murmurs compared with controls (P<.001) (Figure 2). In subgroup analysis, a significant difference in median CT was found in both the valvular insufficiency (n=23; P<.001) and subaortic stenosis (n=9; P=.007) subgroups when compared with controls (n=58) (Figure 2). CT was not significantly different between the 2 cardiac subgroups (P=.227). For the 23 dogs with valvular insufficiency, correlation between CT and ATJ, ATJ:AA, and ATJ:Ao was poor (r2<.12; data not shown). For the 9 dogs with subaortic stenosis and the 1 dog with pulmonic stenosis, correlations between CT and velocity (r2=.46) and CT and pressure gradient (r2=.44) were also weak. While CTs tended to be longer with increasing murmur grade, no significant differences in CT were found between murmur grades.
The remaining 7 dogs with a murmur that were not included in either of the subgroups had various cardiac conditions that included physiological murmurs (n=3, CT 63, 72, and 94.5 seconds, murmur grades I/VI, II/VI, and IV/VI, respectively), ventricular septal defects (n=2, CT 48 and 118 seconds, murmur grades III/VI and V/V, respectively), and patent ductus arteriosus (PDA) (n=2, CT 74 and 111, murmur grades IV/VI and V/VI, respectively). One dog with valvular insufficiency had, in addition to congenital mitral valve dysplasia, a ventricular septal defect, and PDA. This dog had the longest CT observed (187 seconds) and a murmur grade of V/VI.
Three of the dogs with cardiac disease had CTs recorded both pre- and postsurgical intervention. These included 1 dog with PDA, 1 dog with PDA and mitral valve dysplasia, and 1 dog with valvular pulmonic stenosis and a ventricular septal defect. The dog with PDA had an initial CT of 110 seconds and CTs of 65 and 66 seconds, 2 and 3 days after complete surgical ligation of the patent ductus, respectively. No murmur was auscultable postoperatively and color-flow Doppler echocardiography results indicated an absence of turbulence. The dog with PDA and mitral valve dysplasia had an initial CT of 111 seconds and a CT of 187 seconds at 2 months postoperatively. Surgical correction had been attempted by thoracotomy but the lesions were not safely correctable and the patient was discharged without a change in cardiac status. Cardiac evaluation 2 months postoperatively showed worsening turbulence with color-flow Doppler echocardiography and the dog died of congestive heart failure 8 months postoperatively. The dog with valvular pulmonic stenosis and a ventricular septal defect had an initial CT of 93.5 seconds, with a CT of 118 seconds at 2 months postsurgical correction with thoracotomy and partial instrument dilation of the stenotic pulmonic valve and no treatment of the septal defect.
To ensure breed bias was not a confounding factor, statistical analyses were performed excluding the 6 CKCS dogs. Statistically significant differences were found between the remaining dogs with cardiac murmurs (n=23) and dogs with valvular insufficiency (n=17) compared with controls (for both groups, P<.001).
The goal of the present study was to determine if a significant difference in CT existed between dogs of several breeds with cardiac disease and healthy control dogs. The group of dogs with cardiac disease and turbulent blood flow, as defined by the presence of a cardiac murmur and confirmed with color flow Doppler echocardiography, had significantly prolonged CTs compared with control dogs. These results suggest that dogs with cardiac murmurs have altered platelet function. This is in contrast to some initial reports in the human medical literature in which patients with valvular cardiac disease had platelets in an activated state as defined by increased platelet diameter20 and increased numbers of spread-type platelets.21 However, studies to date in dogs have supported the finding of decreased platelet function in cardiac disease. Tanaka et al4 showed a decreased platelet lifespan, using in vitro biotinylation, and decreased platelet activity in dogs with mitral valve regurgitation using platelet aggregometry.5 Tarnow et al3 showed decreased platelet activity in CKCS with mild, moderate, and severe mitral valve regurgitation compared with healthy control dogs using the PFA-100. The results of the present study confirm and expand on the findings of these studies by showing decreased platelet function as assessed by the PFA-100 in a broader range of dog breeds with cardiac murmurs. Because CT was not significantly different in dogs with valvular insufficiency and subaortic stenosis, the form of turbulent blood flow does not appear to be an important factor in platelet hypofunction in these forms of cardiac disease.
The exact mechanism associated with decreased platelet function in these dogs is unknown. It has been postulated in humans30 and dogs5 that continuous high shear stress, evoked by turbulent blood flow, may be a leading factor in platelet degranulation and hypofunction. Another factor may be decreased plasma concentrations of vWF high-molecular-weight multimers, as recently shown in CKCS with mitral valve prolapse and in dogs with subaortic stenosis.1 In these patients, abnormally elevated shear stress may actually promote proteolysis of vWF, decreasing the largest and most functional vWF multimers, possibly causing reduced platelet function. Whether the consequences of such an acquired vWF condition in canine patients with valvular disease are clinically relevant is unknown.
The 6 CKCS in the current study comprised 15.8% and 26.1% of dogs in the overall and valvular insufficiency cardiac groups, respectively. Statistical analyses performed after exclusion of these dogs resulted in similar results, eliminating the potential of a breed-associated bias in the study population. Prolonged CTs have also been reported in healthy Cairn terriers31 and in Dachshunds with early stages of myxomatous mitral valve disease.2 In the current study, no Cairn terriers and only 1 Dachshund were included, the latter with a CT (71 seconds) within the reference interval, a III/VI physiological murmur, and no echocardiographic abnormalities.
The results of the current study support those previously reported in CKCS dogs.1 Platelet exhaustion and potentially decreased concentrations of vWF high-molecular-weight multimers may be reasons for the prolonged CTs in these dogs. If these factors are proportional to increasing shear force, finding a positive relationship between CT and murmur grade would be expected. Indeed, a general albeit not statistically significant trend of prolonged CTs relative to murmur grade was noted. Additionally, the extremely low correlation coefficients between CT and turbulent jet calculations, flow velocity, and pressure gradient suggested other factors were involved.
While only 3 dogs had repeated CT analysis performed postsurgically, the findings supported the connection between turbulent blood flow and platelet hypofunction. In humans, successful aortic valve replacement for Heyde syndrome, with confirmed hematologic recovery, has been documented.32 These findings also emphasize that a prolonged CT may be anticipated from a murmur alone, and should not be misinterpreted as indicative of platelet dysfunction caused by another disease process. Additionally, normalization of the CT following surgical correction of the cardiac condition suggests the PFA-100 analyzer may have potential as a postoperative measure of surgical success.
The CT reference interval (48–77 seconds) determined with 58 dogs using COL/ADP cartridges was comparable to that reported by others for this analyzer. In 1 study the canine CT reference interval was 47–81 seconds (n=45),16 in another it was 52–86 seconds (n=29),18 and in a third larger study it was 53–98 seconds (n=136).33 Reasons for difference in reference intervals include variations in collection technique, differences in citrate formulations, and normal patient variability. The previous 3 studies used 0.129 mM (3.8%) sodium citrate tubes while we used 0.105 mM (3.2%) sodium citrate tubes. In human34 and canine35 studies using COL/ADP cartridges, mean CTs were longer when 3.8% sodium citrate tubes were used compared with 3.2% sodium citrate tubes, likely due to stronger calcium chelating properties of higher concentrations of sodium citrate. At the time of the present study, the 3.8% sodium citrated tubes were no longer available. Additionally, use of 3.2% sodium citrate tubes for routine coagulation testing in humans is the recommended standard.36
In summary, the PFA-100 analyzer is useful in detecting platelet function defects in dogs with cardiac murmurs, most notably those caused by mitral and/or tricuspid valvular insufficiency or subaortic stenosis. Such alterations in platelet function should not be attributed automatically to extracardiac illness in patients with concurrent cardiac disorders. Additional investigations using this analyzer to detect platelet functional alterations in canine patients with various forms of cardiac disease or other illnesses are warranted.
This study was supported by a grant from the Sir James Dunn Animal Welfare Centre. The authors would like to thank Dr. Henrik Stryhn for his statistical expertise, Ms. Anne Dover, Ms. Linda Ruschkowski-Kerr, Ms. Karen Blackburn, and Mrs. Elaine Reveler for excellent technical assistance, and all clinicians, students, and pet owners for their participation and generosity during this study.