Platelets play an important role in hemostasis, representing the first line of defense against bleeding at sites of vascular injury. Adhesion to damaged endothelium through specific cell surface receptors and interaction with coagulation factors lead to platelet activation, aggregation, and formation of the hemostatic plug at the site of vascular injury. When circulating platelets become activated, they secrete the contents of their alpha, dense, lysosomal granules through their canalicular systems to the exterior.1
Plasma MPC concentration is a measurement derived from the platelet refractive index (which is linearly related to platelet density) and can be determined only by use of a particular automated hematology analyzer system via 2-angle light scatter. This analyte can be used to assess platelet activation.2 When activated platelets degranulate, their density decreases, and the plasma MPC concentration then also decreases.3
Flow cytometry has been used in veterinary medicine for direct or indirect measurement of platelet activation through quantification of P-selectin expression by platelets, detection of platelet-bound fibrinogen, or measurement of platelet-leukocyte aggregates.1–6 P-selectin is a protein of the interior surface of alpha-granule membranes and is exposed on cell surfaces only after platelets degranulation.7–10 Degree of platelet P-selectin expression is correlated with plasma MPC concentration, suggesting that measurement of both variables may be useful in the detection of activated platelets in dogs.4 Quantification of P-selectin expression by platelets is clinically applicable but requires a flow cytometer and approximately 60 minutes of sample preparation time.11 On the other hand, plasma MPC measurement can be included in a routine CBC and therefore be performed inexpensively and quickly.4
Immune-mediated hemolytic anemia is a common hematologic disorder in dogs and can be primary or secondary in nature. Most dogs with IMHA have primary or idiopathic disease,12 and associated mortality rates can be high.13,14 Hemostatic disorders such as thrombosis and DIC can increase the risk of death associated with IMHA.5,13–24
In the past decade, platelet activation has been identified in dogs with septic and nonseptic inflammatory disease,4 including IMHA, on the basis of high platelet P-selectin expression.5,6 Activated platelets may therefore contribute to the development of micro- or macrothrombi, causing illness and death in dogs with IMHA.13,14 Treatment with combinations of corticosteroid drugs, azathioprine, and ultralow-dose aspirin reportedly results in significantly greater short-and long-term survival rates in dogs with pIMHA,23 and dogs with pIMHA receiving standard immunosuppressive treatment with clopidogrel, alone or in combination with ultralow-dose aspirin, have a short-term survival rate similar to that of dogs receiving ultralow-dose aspirin alone.25 These findings further suggest that platelet activation may be involved in the prothrombotic state of dogs with IMHA. Platelet hyperreactivity has also been identified via thromboelastography in dogs with IMHA, and for some of these dogs, this hyperreactivity may be the primary cause of hypercoagulability.26
The primary aim of the retrospective study reported here was to determine whether dogs with IMHA had detectable evidence of activated platelets as inferred by a low plasma MPC concentration at initial examination. Given that platelet activation may lead to a prothrombotic state and IMHA may be complicated by thrombotic conditions such as thromboembolism and DIC (which are both reported causes of death for dogs with IMHA),5,13–24 a secondary aim was to assess the prognostic value of plasma MPC concentration at initial examination for dogs with IMHA. In addition, we sought to identify whether signalment factors, clinicopathologic findings previously associated with death in dogs with IMHA (ie, serum urea and bilirubin concentrations, and Hct)14,17,23,25 or with risk of thrombosis (ie, plasma fibrinogen and D-dimer concentrations and platelet count),14,27,28 and plasma MPC concentration were associated with outcome in dogs with IMHA.
Materials and Methods
Animals
The electronic medical records databasea of the San Marco Veterinary Clinic, Padua, Italy, was searched to identify dogs evaluated at the clinic between August 1, 2005, and February 28, 2011, for inclusion in 3 groups. Dogs were included in the IMHA (case) group if they had anemia (Hct < 30%; reference interval, 38.6% to 54.5%) and anti-erythrocyte membrane antibody detected by flow cytometry. Dogs with IMHA and evidence of a concurrent condition such as neoplasia or infectious disease as well as dogs that received medications such as potentiated sulfonamides or cephalosporins29 or that were vaccinated within 30 days before anemia was detected were classified as having sIMHA. The remaining dogs were classified as having pIMHA. Dogs that had received prednisone treatment for < 14 days prior to initial examination were not excluded from the study; however, they were excluded if they had received prednisone for > 14 days. Dogs with IMHA were also excluded from the study if they had received in the 28 days prior to initial examination any immunosuppressant drugs other than prednisone (eg, azathioprine, cyclophosphamide, or cyclosporine), antiplatelet treatment (eg, aspirin or clopidogrel), or a blood transfusion. Dogs in the IMHA group were also required to have a minimum recorded follow-up period of 42 days after initial examination.
Two control groups were formed. Dogs were included in the healthy control group if the reason for initial examination was routine annual examination, elective surgery, blood donor health screening, or prebreeding examination. Dogs were included in the sick control group if they were examined for any health concern other than IMHA. All control dogs were individually matched to dogs with IMHA by age (within 6 months), reproductive status (ie, sex and neuter status), and breed. When no breed-matched dog of the same age and sex as a given case dog could be identified in the database, a dog of similar body weight (within 5 kg [11 lb]) was selected instead. Dogs were selected as closely as possible to the admission date of the corresponding case dog. When ≥ 2 dogs fulfilled these criteria, the records systema was used to randomly select 1 dog.
To be included in any group, dogs were required to have a complete medical record, including medical history and results of physical examination, CBC (including blood smear examination), serum biochemical analysis, coagulation profile analysis, and urinalysis. All dogs in the IMHA group and some dogs in the sick control group had also undergone thoracic radiography and abdominal ultrasonography. Dogs in these 2 groups also had other diagnostic tests performed when judged necessary by the attending clinician. To be included in the study, dogs in the IMHA group and sick control group were also required to have received a specific diagnosis relevant to the initial reason for examination. Dogs eligible for either control group were excluded if they had received during the 28 days prior to initial examination any antiplatelet treatment, such as aspirin or clopidogrel, or a blood transfusion.
Data collection
Data recorded at initial examination were retrieved from the electronic medical records of each dog regarding signalment (reproductive status, age, and breed) and plasma MPC concentration. For dogs with IMHA only, data were also retrieved from the electronic medical records regarding serum urea and bilirubin concentrations, Hct, plasma fibrinogen and D-dimer concentrations, platelet count, and prednisone treatment duration, when applicable. For dogs in the IMHA group that failed to survive for at least 42 days after initial examination, the time from initial examination to death was retrieved.
All clinicopathologic tests had been performed at the Laboratorio d'Analisi San Marco for all dogs at initial examination. A blood sample was collected from each dog via cephalic (for medium-sized or large dogs) or jugular (for small dogs) venipuncture into sterile 10-mL plastic syringes. Two milliliters of blood was immediately transferred to plastic tubes containing K3-EDTA for a CBC and anti-erythrocyte membrane antibody testing via flow cytometry. The plasma MPC concentration was measured as part of the routine CBC, which was performed by use of a hematology analyzerb ≤ 60 minutes after sample collection.
Percentages of erythrocytes with IgG or IgM membrane binding for dogs with IMHA were measured ≤ 60 minutes after sample collection by use of a flow cytometerc as described elsewhere.30 A sample was considered positive for antibody binding if a positive peak was observed in the fluorescence histogram in addition to > 1.47% of erythrocytes with membrane-binding IgG or 1.38% of erythrocytes with membrane-binding IgM.30
Statistical analysis
Continuous data were assessed for normality of distribution with the Shapiro-Wilk test. Normally distributed data were reported as mean ± SD, and nonnormally distributed data were reported as median (range). Differences among the 3 groups in mean age were evaluated by ANOVA. The plasma MPC concentration was compared among groups via 1-way ANOVA followed by the Bonferroni post hoc test.
Dogs with IMHA that survived to 42 days after initial examination were classified as survivors, and dogs with IMHA that did not were classified as nonsurvivors. For dogs with IMHA only, the prognostic importance of the following variables (as measured at initial examination) was determined through comparisons between survivors and nonsurvivors: plasma MPC concentration, reproductive status and age, serum urea and bilirubin concentrations, Hct, plasma fibrinogen concentration, platelet count, and plasma D-dimer concentration. The 2-sample t test, Mann-Whitney test, and χ2 test were used for these purposes. Also for dogs with IMHA, differences in plasma MPC concentration (2-sample t test) and mortality rate at 42 days after initial examination (χ2 test with Yates correction) were examined between dogs that received or did not receive prednisone treatment in the 2 weeks before initial examination.
In dogs with IMHA, ROC curve analysis was then used to identify the best cutoff value for plasma MPC concentration at initial examination for discriminating survivors from nonsurvivors. Differences in plasma MPC concentrations at initial examination, overall mortality rate at 42 days after initial examination, proportions of dogs with pIMHA and sIMHA with a plasma MPC value less than or equal to or greater than the optimal cutoff value identified through ROC curve analysis, and mortality rate of dogs with pIMHA and sIMHA with respect to the same cutoff value were also evaluated by use of the 2-sample t test (for concentrations) and χ2 test with Yates correction (all other variables). The optimal plasma MPC cutoff value derived from ROC curve analysis was used to calculate associations (ORs with 95% CIs, which were compared with the χ2 test with Yates correction) with survival status at 42 days after initial examination.
Kaplan-Meier curves were also constructed to investigate whether an association existed between plasma MPC concentration (according to the same cutoff value) and overall survival rate, and the log-rank test was used for statistical comparison of these curves. Cox proportional hazards regression was performed with survival status (at 42 days after initial examination) as the dependent variable, plasma MPC concentration as a predictor variable, and the other prognostic variables with a liberal P value (< 0.25) on univariate analysis as covariates, to estimate the mean hazard of death (reported as relative risk) for a dog with IMHA. The estimated risk of death per dog for each incremental unit of plasma MPC concentration at the time of initial examination was also calculated. For all statistical analyses, values of P < 0.05 were considered significant.
Results
Animals
Breed classifications and reproductive statuses of dogs in the IMHA group and the 2 control groups (healthy or sick without IMHA) were summarized (Table 1). During the study period, 191 dogs received a diagnosis of IMHA at the San Marco Veterinary Clinic, but only 95 (50%) fulfilled all inclusion criteria and were therefore included in the IMHA group. All of these dogs were anemic (mean ± SD Hct, 14.63 ± 5.89%) and had positive results of flow cytometric testing for erythrocyte membrane-binding IgG (n = 49), IgM (4), or both (42). Spherocytes were detected on blood smear examination for 66 (69%) dogs; 25 dogs had mild spherocytosis (ie, 1 to 10 spherocytes/hpf [1,000X]), 13 had moderate spherocytosis (ie, 11 to 50 spherocytes/hpf), 13 had marked spherocytosis (ie, 51 to 150 spherocytes/hpf), and 15 had severe spherocytosis (ie, > 150 spherocytes/hpf). Thirty-six (38%) dogs in the IMHA group received prednisone (mean ± SD dose, 1.90 ± 1.07 mg/kg/d [0.86 ± 0.49 mg/lb/d]; range, 0.18 to 5.0 mg/kg/d [0.08 to 2.27 mg/kg/d]) before initial examination, for a median duration of 3 days (range, 1 to 14 days). On the basis of whether dogs had concurrent disease in addition to IMHA, 64 (67%) dogs were classified as having pIMHA and 31 (33%) as having sIMHA. Underlying causes of sIMHA included bacterial infection (n = 11), neoplasia (8), ehrlichiosis (2), and other causes (10).
Number (%) of dogs with various signalment characteristics, grouped by whether dogs had IMHA (n = 95; case group), were healthy (95; healthy control group), or had health concerns other than IMHA (95; sick control group) at initial examination.
| Characteristic | IMHA | Healthy | Sick without IMHA |
|---|---|---|---|
| Breed or type | |||
| Mixed-breed dog | 31 (33) | 40 (42) | 32 (34) |
| English Cocker Spaniel | 5 (5) | 6 (6) | 5 (5) |
| Shih Tzu | 5 (5) | 4 (4) | 5 (5) |
| Boxer | 3 (3) | 3 (3) | 3 (3) |
| Toy Poodle | 3 (3) | 0 (0) | 3 (3) |
| Yorkshire Terrier | 3 (3) | 3 (3) | 3 (3) |
| Other | 45 (47) | 39 (41) | 44 (46) |
| Reproductive status | |||
| Sexually intact male | 30 (32) | 30 (32) | 30 (32) |
| Castrated male | 8 (8) | 8 (8) | 8 (8) |
| Sexually intact female | 29 (31) | 29 (31) | 29 (31) |
| Spayed female | 28 (29) | 28 (29) | 28 (29) |
These data are provided to show that matching of control dogs to case dogs on the basis of these variables was successfully achieved.
Dogs in the healthy control group (n = 95) were 100% matched to dogs in the IMHA group with respect to reproductive status and 77% with respect to breed (Table 1). Dogs in the sick control group (n = 95) were 100% matched to dogs in the IMHA group with respect to reproductive status and 98% with respect to breed. Causes of sickness for dogs in the sick control group included urogenital disease (n = 16), neoplasia (14), gastrointestinal disorder (11), brain and spinal cord disease (10), endocrinopathy (7), road traffic accident or other traumatic disease (7), respiratory problems (6), sepsis and infectious disease (5), inflammatory or immune-mediated disease (3), cardiac problems (3), liver disease (3), and other causes (10).
Comparisons among groups
No significant (P = 0.99) difference in mean age was identified among dogs with IMHA (mean ± SD, 7.32 ± 3.87 years; range, 8 weeks to 16.5 years), healthy dogs (7.29 ± 3.81 years; range, 16 weeks to 15.3 years), and sick dogs without IMHA (7.29 ± 3.80 years; range, 12 weeks to 15.4 years). Mean plasma MPC concentration (reference interval, 19.1 to 24.4 g/dL) at initial examination was significantly (P < 0.001) lower in dogs with IMHA (20.21 ± 2.38 g/dL) than in healthy dogs (22.01 ± 2.07 g/dL) or sick dogs without IMHA (21.31 ± 2.01 g/dL). Mean plasma MPC concentration in healthy dogs was significantly (P = 0.03) greater than in sick dogs without IMHA (Figure 1).
Box-and-whisker plots of plasma MPC concentrations at initial examination for dogs with IMHA (n = 95), healthy dogs (95), and sick dogs without IMHA (95). For each box, the central horizontal line represents the median, and the lower and upper boundaries represent the 25th and 75th percentiles, respectively. Whiskers represent the most extreme observations that were not outliers. Circles represent outliers (ie, values that were less than or greater than the 25th or 75th percentile values by > I.5 times the interquartile range). The shaded region represents the reference interval for plasma MPC concentration. a–cMean values of the represented data with different superscript letters differ significantly (P ≤ 0.03).
Citation: Journal of the American Veterinary Medical Association 252, 11; 10.2460/javma.252.11.1384
Among dogs with IMHA, no significant (P = 0.73) difference in mean plasma MPC concentration at initial examination was identified between those with pIMHA (20.27 ± 2.46 g/dL) and those with sIMHA (20.09 ± 2.23 g/dL). There was also no significant (P = 0.63) difference in mean plasma MPC concentrations between dogs that received (20.03 ± 2.36 g/dL) or did not receive (20.06 ± 2.36 g/dL) prednisone in the 2 weeks prior to initial examination.
By the end of the 42-day follow-up period, 39 of the 95 (41%) dogs with IMHA had died (n = 25) or were euthanized (14) because of worsening of their clinical condition. No difference in mortality rates was identified by IMHA type (pIMHA, 24/64 [38%]; sIMHA, 15/31 [48%]; P = 0.43) or whether prednisone had been received prior to diagnosis (yes, 16/36 [44%]; no, 23/59 [39%]; P = 0.76). Twenty-seven of the 39 (69%) nonsurvivors had died within 1 week after initial examination (overall mortality rate at this point, 27%), and 30 (77%) died within 2 weeks after initial examination (overall mortality rate at this point, 32%). At initial examination, mean plasma MPC concentration was significantly (P = 0.004) lower in nonsurvivors (19.37 ± 2.64 g/dL) than in survivors (20.79 ± 2.00 g/dL; Figure 2). Differences between survivors and nonsurvivors regarding sex and age, serum urea and bilirubin concentrations, Hct, plasma fibrinogen concentration, platelet count, and plasma D-dimer concentration at initial examination were summarized (Table 2).
Box-and-whisker plots of plasma MPC concentrations at initial examination for dogs with IMHA that survived (n = 56) or did not survive (39) to 42 days after initial examination. a,bMean values of the represented data with different superscript letters differ significantly (P = 0.004). See Figure I for remainder of key.
Citation: Journal of the American Veterinary Medical Association 252, 11; 10.2460/javma.252.11.1384
Comparison of various characteristics at initial examination for dogs with IMHA that survived (n = 56) or did not survive (39) to 42 days after that examination.
| Characteristic | Survivors | Nonsurvivors | P value |
|---|---|---|---|
| Age (y) | 7.0 ± 3.91 | 7.8 ± 3.83 | 0.33 |
| Male sex | 21 (38) | 17 (44) | 0.55 |
| Serum urea concentration (mg/dL) | 55 (12–366) | 61 (15–444) | 0.89 |
| Serum bilirubin concentration (mg/dL) | 0.45 (0.04–39.14) | 0.65 (0.06–33.94) | 0.22 |
| Hct (%) | 14.5 ± 6.5 | 14.8 ± 4.9 | 0.78 |
| Plasma fibrinogen concentration (mg/dL) | 647 ± 329 | 577 ± 291 | 0.29 |
| Platelet count (× 103/μL) | 286 (25–1,629) | 266 (28–880) | 0.19 |
| Plasma D-dimer concentration (μg/dL) | 0.21 (0.01–5.43) | 0.35 (0.01–7.53) | 0.05 |
Values for age, Hct, and plasma fibrinogen concentration are mean ± SD. Values for sex are number (%) of dogs in each group. Values for all other variables are median (range). Characteristics with P values < 0.25 on univariate analysis were included in a multivariate model.
ROC curve analysis and mortality rate
The optimal cutoff value identified through ROC curve analysis for plasma MPC concentration (ie, the value that maximized the sum of sensitivity and specificity [Youden index] in discriminating survivors from nonsurvivors) was 19.1 g/dL. Sensitivity at this cutoff value was 51%, and specificity was 84% (AUC, 0.67; 95% CI for the AUC, 0.56 to 0.79; P = 0.004; Figure 3). The frequency of dogs with pIMHA having a plasma MPC concentration ≤ 19.1 g/dL (19/64 [30%]) or > 19.1 g/dL (45/64 [70%]) was comparable to the frequency for dogs with sIMHA (10/31 [32%] or 21/31 [68%], respectively; P = 0.80).
Receiver operating characteristic curve for the use of plasma MPC concentration at initial examination for discriminating dogs with IMHA that survived (n = 56) from those that did not survive (39) to 42 days after that examination. The optimum cutoff value (19.1 g/dL; circle) was identified as the value that provided the greatest sum of sensitivity and specificity (Youden index). At that value, sensitivity was 51% and specificity was 84% (AUC, 0.67; 95% CI for the AUC, 0.56 to 0.79; P = 0.004).
Citation: Journal of the American Veterinary Medical Association 252, 11; 10.2460/javma.252.11.1384
The mortality rate of dogs with a plasma MPC concentration ≤ 19.1 g/dL (20/29 [69%]) was significantly (P < 0.001) higher than the mortality rate of dogs with values > 19.1 g/dL (19/66 [29%]). No difference in mortality rates was identified between dogs with pIMHA versus sIMHA regarding proportions with a plasma MPC concentration ≤ 19.1 g/dL (pIMHA, 13/19; sIMHA, 7/10; P = 0.73) or > 19.1 g/dL (pIMHA, 11/45 [24%]; sIMHA, 8/21 [38%]; P = 0.33).
A significant (P < 0.001) association was identified between plasma MPC concentration categorized by the cutoff value (≤ 19.1 g/dL or > 19.1 g/dL) and the probability of survival to 42 days after initial examination (OR, 5.50; 95% CI, 1.94 to 15.98). Kaplan-Meier analysis involving dogs with IMHA revealed a significant (P < 0.001) association between categorized plasma MPC concentration and the probability of survival to 42 days after initial examination (Figure 4). Dogs with values ≤ 19.1 g/dL were significantly less likely to survive for the 42-day period than dogs with values > 19.1 g/dL.
Kaplan-Meier survival curves for the 42-day period after initial examination for dogs with IMHA with an initial plasma MPC concentration ≤ 19.1 g/dL (the optimum cutoff value shown in Figure 3; solid line; n = 29) or > 19.1 g/dL (dotted line; 66).
Citation: Journal of the American Veterinary Medical Association 252, 11; 10.2460/javma.252.11.1384
Results of multivariate Cox proportional hazards regression revealed that plasma MPC concentration at initial examination was the only factor significantly associated with death after controlling for other factors (Table 3). The mean estimated risk of death for dogs with IMHA decreased by 16% (95% CI, 4% to 27%; P = 0.01) for every 1-unit increase in plasma MPC concentration. The estimated risk of death per incremental unit of plasma MPC concentration from 18 to 23 g/dL and associated survival rates were summarized (Table 4).
Results of multivariate Cox proportional hazards analysis of characteristics associated with failure to survive to 42 days after initial examination for 95 dogs with IMHA.
| Characteristic | Hazard ratio | 95% CI | P value |
|---|---|---|---|
| Plasma MPC concentration (g/dL) | 0.84 | 0.73–0.96 | 0.01 |
| Serum bilirubin concentration (mg/dL) | 1.02 | 0.99–1.06 | 0.18 |
| Platelet count (× 103/μL) | 0.999 | 0.997–1.001 | 0.22 |
| Plasma D-dimer concentration (μg/dL) | 1.0 | 0.81–1.24 | 0.99 |
Only variables with P values < 0.25 on univariate analysis were included in this model.
Discussion
In the case-control study reported here, the primary aim was to investigate whether dogs with IMHA had lower plasma MPC concentrations, consistent with high numbers of activated platelets, than both healthy dogs and sick dogs without IMHA. The use of healthy dogs as the sole control group would have introduced some bias because they can be generally expected to have different management characteristics than the general population (eg, routine vaccination and regular endo- and ectoparasite prophylaxis). For this reason, we also included a group of sick control dogs without IMHA that was unaffected by the healthy control bias. Findings indicated that the mean plasma MPC concentration at initial examination was significantly lower in dogs with IMHA than in healthy dogs or sick dogs without IMHA, and it was also significantly lower in sick dogs without IMHA than in healthy dogs. In accordance with findings of other studies,5,6 these data provided support for the hypothesis that platelets circulate in a more activated state in dogs with IMHA than in healthy dogs. Findings of the present study also suggested that in dogs with IMHA, platelets circulated with a low plasma MPC concentration, a finding consistent with high platelet activation, compared with the situation in sick dogs without IMHA.
In contrast, no significant decrease in plasma MPC concentration was identified in another study31 for dogs with IMHA versus healthy dogs or dogs with other types of inflammatory or noninflammatory diseases. A few factors may explain this discrepancy. First, each of the 3 groups in the present study contained 95 dogs, whereas dogs in the other study31 (n = 228) were separated into 8 groups containing between 9 and 47 dogs each (median, 30 dogs), limiting statistical power to detect small differences among the groups. Second, the previous study31 involved no strategy for matching dogs by signalment during enrollment; therefore, the influence of confounding factors such as age, sex (including neuter status), and breed was not controlled for. Third, some dogs with IMHA in the present study may have also had a concurrent immune-mediated thrombocytopenia owing to the occasional association of these 2 diseases with each other.32 Such concurrent disease would have resulted in a lower plasma MPC concentration in dogs with IMHA according to findings of the other study,31 in which dogs with immune-mediated thrombocytopenia had the lowest plasma MPC concentration.
In agreement with previously reported data,3 most dogs with IMHA in the present study had the idiopathic form or primary form of the disease. Interestingly, no difference in plasma MPC concentrations was identified between dogs with pIMHA and dogs with sIMHA. This finding suggested that, in dogs with IMHA, platelet activation may not be dependent on the underlying cause of IMHA but may instead stem from the IMHA itself. Similarly, no difference was identified in platelet activation, as measured by plasma MPC concentration, between dogs with IMHA that received prednisone before initial examination (mean dosage, 1.90 mg/kg/d for a median duration of 3 days) and dogs with IMHA that received no such treatment. This latter result is in agreement with reported findings that treatment of healthy dogs with prednisone at 2 mg/kg/d (0.9 mg/lb/d) for 15 days causes no alteration of platelet function, as measured by turbidimetric aggregometry.32 Likewise, despite the extensive use of corticosteroid drugs in the treatment of humans with cardiovascular and immune disorders, these drugs reportedly have minimal to no effect on platelet function in humans.33
Most reported mortality rates associated with IMHA in dogs have been high.13,14,17,20,23,34,35 Previously reported mortality rates for dogs with pIMHA range from 20% to 70%,35,36 and this wide range could be attributable, at least in part, to the duration of follow-up. In the present study, the overall mortality rate was 41%, with no significant difference identified between dogs with pIMHA (37%) and dogs with sIMHA (48%) or between dogs that briefly received (44%) or did not receive (39%) prednisone treatment before initial examination. In addition, most (77%) dogs that failed to survive in the present study died within the first 2 weeks after initial examination. By 2 weeks after initial examination, the overall mortality rate reached 32%. These results were similar to those in a previous study34 that included only dogs with pIMHA, for which the estimated mortality rate for a similar period was 22.5%, followed by a marked decrease in risk of death after this time. The period of 42 days after initial examination was chosen for survival-related calculations in the present study to allow inclusion of most of the surviving dogs, for which follow-up information had been obtained at the second reevaluation after initial examination.
Dogs with IMHA are often systemically ill and can have evidence of multiorgan dysfunction. Several factors previously identified as associated with a poor prognosis for dogs with IMHA include a high plasma urea concentration, high band neutrophil count, hyperbilirubinemia, thrombocytopenia, and the presence of petechiae at the time of diagnosis.14–16,23,34 Investigators who performed a systematic review37 of factors associated with mortality rate in dogs with IMHA excluded studies that included no adjustment for potential confounding factors via multivariate (vs univariate) analysis. Their aim in doing so was to avoid spurious associations between single biochemical or hematologic variables with survival times.37 For the same reason, a multivariate approach was used in the present study to control for variables that might confound the association between plasma MPC concentration and outcome. Moreover, dogs enrolled in each of the 3 groups were matched by potential confounders (age, reproductive status, and breed), minimizing the effect of these variables on the results involving plasma MPC concentration. Although plasma D-dimer and MPC concentrations appeared to have had prognostic value on univariate analysis, only plasma MPC concentration had a significant association with outcome on multivariate analysis. To the authors’ knowledge, the present study represents the first in which plasma MPC concentration was identified as a prognostic indicator for dogs with IMHA. Neither prednisone treatment prior to IMHA diagnosis (at the first plasma MPC evaluation) nor type of IMHA (pIMHA vs sIMHA) influenced the observed prognostic value of plasma MPC concentration.
Interestingly, the optimal cutoff value for plasma MPC concentration identified by ROC curve analysis to discriminate survivors from nonsurvivors coincided with the lower reference limit for this analyte at the study laboratory. The accuracy of plasma MPC concentration for predicting death over the 42-day period after initial examination for dogs with IMHA was 0.67. An AUC between 0.50 and 0.70 is generally considered to indicate low accuracy for a diagnostic test.38 The observed result may have reflected the fact that dogs with IMHA have multiorgan dysfunction and therefore multiple causes of death, which cannot be predicted by a single test.
To the authors’ knowledge, no other prognostic factors for dogs with IMHA have been investigated through ROC curve analysis, precluding comparisons of the findings for the plasma MPC assay used in our study with those of other diagnostic tests. The clinical importance of the ROC curve-derived cutoff value (19.1 g/dL) for plasma MPC concentration was supported by the significant degree of association between plasma MPC concentrations equal to or below this cutoff and mortality rate, which was an expected result given that both these analyses involved survival status at 42 days after initial examination for dogs with IMHA. However, the Kaplan-Meier curves of the probability of survival over the 42-day period further supported the clinical importance of the cutoff plasma MPC value. The significant mean estimated risk of death for dogs with IMHA with respect to plasma MPC concentration at initial examination allowed calculation of the estimated risk of death for each incremental unit of plasma MPC concentration (Table 4). In the authors’ opinion, these latter estimates would be more clinically informative and useful to attending clinicians and dog owners than a generic population mortality rate range of 20% to 70%.35,36
Survival rate (proportion [%]), cumulative survival rate (proportion [%]), and OR for death during the 42-day period after initial examination per incremental unit of plasma MPC concentration for 95 dogs with IMHA.
| MPC (g/dL) | Survival (%) | Cumulative survival (%) | OR for death (95% CI) | P value |
|---|---|---|---|---|
| < 18.0 | 4/14 (29) | 4/14 (29) | 4.48 (1.29–15.57) | 0.01 |
| 18.0–18.9 | 4/11 (36) | 8/25 (32) | 4.63 (1.74–12.36) | 0.001 |
| 19.0–19.9 | 6/11 (55) | 14/36 (39) | 3.88 (1.62–9.32) | 0.002 |
| 20.0–20.9 | 16/21 (76) | 30/57 (53) | 1.95 (0.83–4.60) | 0.12 |
| 21.0–21.9 | 11/18 (61) | 411/75 (55) | 2.49 (0.82–7.54) | 0.10 |
| 22.0–22.9 | 8/10 (80) | 49/85 (58) | 1.71 (0.41–7.09) | 0.45 |
| ≥ 23.0 | 7/10 (70) | 56/95 (59) | 0.58 (0.11–2.76) | 0.68 |
The overall mortality rate was significantly higher in dogs with a lower plasma MPC concentration than in dogs with a higher concentration in the study reported here. However, it was not possible to establish whether plasma MPC concentration was purely an indicator of platelet activation secondary to the prothrombotic state of dogs with IMHA or whether a low MPC concentration had direct harmful effects that contributed to hemostatic disorders such as thrombosis and DIC, which are potential causes of death in dogs with IMHA.5,13–24 Interestingly, 2 studies reported in the past decade showed that antiplatelet treatment with ultralow-dose aspirin could improve short- and long-term survival rates for dogs with IMHA23 and that clopidogrel was as effective as ultralow-dose aspirin for the short-term survival of dogs with pIMHA.25 These findings might indicate that a high degree of platelet activation, as suggested by a low plasma MPC concentration, has a direct effect on the likelihood of survival, although additional research would be needed to explore this possibility.
The hypercoagulable state in dogs with IMHA is an important cause of illness and death, resulting in an increased risk of thromboembolism and DIC.5,13–24,26 Results of the present study suggested that the hypercoagulable state characteristic of dogs with IMHA could be at least in part attributable to a higher degree of platelet activation, as indicated by a lower plasma MPC concentration at initial examination than in healthy dogs and sick dogs without IMHA. The identified prognostic value of plasma MPC concentration for dogs with IMHA regardless of type (primary or secondary) or brief prednisone treatment prior to plasma MPC measurement could theoretically justify the use of antiplatelet drugs such as clopidogrel or aspirin for all dogs with IMHA.
Acknowledgments
Presented in part in abstract form at the 21st European College of Veterinary Internal Medicine-Companion Animals Congress, Seville, Spain, September 2011.
ABBREVIATIONS
| AUC | Area under the curve |
| CI | Confidence interval |
| DIC | Disseminated intravascular coagulation |
| IMHA | Immune-mediated hemolytic anemia |
| MPC | Mean platelet component |
| pIMHA | Primary immune-mediated hemolytic anemia |
| ROC | Receiver operating characteristic |
| sIMHA | Secondary immune-mediated hemolytic anemia |
Footnotes
POA System-Plus, version 9.0, Ingegnere Luigi Coppola, Venezia Mestre, Italy.
ADVIA 120 and 2120 hematology system, Siemens Healthcare GmbH, Erlangen, Germany.
Epics XL-MCL, Beckman Coulter Inc, Brea, Calif.
References
1. Kamath S. Platelet activation: assessment and quantification. Eur Heart J 2001;22:1561–1571.
2. Chapman ES, Sorette M, Hetherington E, et al. A rapid, automated flow cytometric method to measure activated degranulated platelets by density determination. Thromb Haemost 2003;89:1004–1015.
3. Macey MG, Carty E, Webb L, et al. Use of mean platelet component to measure platelet activation on the ADVIA 120 haematology system. Cytometry 1999;38:250–255.
4. Moritz A, Walcheck BK, Weiss DJ. Evaluation of flow cytometric and automated methods for detection of activated platelets in dogs with inflammatory disease. Am J Vet Res 2005;66:325–329.
5. Weiss DJ, Brazzell JL. Detection of activated platelets in dogs with primary immune-mediated hemolytic anemia. J Vet Intern Med 2006;20:682–686.
6. Ridyard AE, Shaw DJ, Milne EM. Evaluation of platelet activation in canine immune-mediated haemolytic anaemia. J Small Anim Pract 2010;51:296–304.
7. Stenberg PE, McEver RP, Shuman MA, et al. A platelet α-granule membrane protein (GMP-140) is expressed on the plasma membrane after activation. J Cell Biol 1985;101:880–886.
8. Hsu-Lin S, Berman CL, Furie BC, et al. A platelet membrane protein expressed during platelet activation and secretion. Studies using a monoclonal antibody specific for thrombinactivated platelets. J Biol Chem 1984;259:9121–9126.
9. McEver RP. Properties of GMP-140, an inducible granule membrane protein of platelets and endothelium. Blood Cells 1990;16:73–80, discussion 80–83.
10. Michelson AD. Flow cytometry: a clinical test of platelet function. Blood 1996;87:4925–4936.
11. Moritz A, Walcheck BK, Weiss DJ. Flow cytometric detection of activated platelets in the dog. Vet Clin Pathol 2003;32:6–12.
12. McCullough S. Immune-mediated hemolytic anemia: understanding the nemesis. Vet Clin North Am Small Anim Pract 2003;33:1295–1315.
13. Scott-Moncrieff JC, Treadwell NG, McCullough SM, et al. Hemostatic abnormalities in dogs with primary immune-mediated hemolytic anemia. J Am Anim Hosp Assoc 2001;37:220–227.
14. Carr AP, Panciera DL, Kidd L. Prognostic factors for mortality and thromboembolism in canine immune-mediated hemolytic anemia: a retrospective study of 72 dogs. J Vet Intern Med 2002;16:504–509.
15. Klein MK, Dow SW, Rosychuk RA. Pulmonary thromboembolism associated with immune-mediated hemolytic anemia in dogs: ten cases (1982–1987). J Am Vet Med Assoc 1989;195:246–250.
16. LaRue MJ, Murtaugh RJ. Pulmonary thromboembolism in dogs: 47 cases (1986–1987). J Am Vet Med Assoc 1990;197:1368–1372.
17. Klag AR, Giger U, Shofer FS. Idiopathic immune-mediated hemolytic anemia in dogs: 42 cases (1986–1990). J Am Vet Med Assoc 1993;202:783–788.
18. Mischke R. Hemostatic disorders as a complication of autoimmune hemolytic anemia in dogs. Dtsch Tierarztl Wochenschr 1998;105:13–16.
19. Johnson LR, Lappin MR, Baker DC. Pulmonary thromboembolism in 29 dogs: 1985–1995. J Vet Intern Med 1999;13:338–345.
20. Burgess K, Moore A, Rand W, et al. Treatment of immune-mediated hemolytic anemia in dogs with cyclophosphamide. J Vet Intern Med 2000;14:456–462.
21. McManus PM, Craig LE. Correlation between leukocytosis and necropsy findings in dogs with immune-mediated hemolytic anemia: 34 cases (1994–1999). J Am Vet Med Assoc 2001;218:1308–1313.
22. Mason N, Duval D, Shofer FS, et al. Cyclophosphamide exerts no beneficial effect over prednisone alone in the initial treatment of acute immune-mediated hemolytic anemia in dogs: a randomized controlled clinical trial. J Vet Intern Med 2003;17:206–212.
23. Weinkle TK, Center SA, Randolph JF, et al. Evaluation of prognostic factors, survival rates, and treatment protocols for immune-mediated hemolytic anemia in dogs: 151 cases (1993–2002). J Am Vet Med Assoc 2005;226:1869–1880.
24. Breuhl EL, Moore G, Brooks MB. A prospective study of unfractionated heparin therapy in dogs with primary immune-mediated hemolytic anemia. J Am Anim Hosp Assoc 2009;45:125–133.
25. Mellett AM, Nakamura RK, Bianco D. A prospective study of clopidogrel therapy in dogs with primary immune-mediated hemolytic anemia. J Vet Intern Med 2011;25:71–75.
26. Hamzianpour N, Chan DL. Thromboelastographic assessment of the contribution of platelets and clotting proteases to the hypercoagulable state of dogs with immune-mediated hemolytic anemia. J Vet Emerg Crit Care (San Antonio) 2016;26:295–299.
27. Wagg CR, Boysen SR, Bedard C. Thrombelastography in dogs admitted to an intensive care unit. Vet Clin Pathol 2009;38:453–461.
28. Nelson OL, Andreasen C. The utility of plasma D-dimer to identify thromboembolic disease in dogs. J Vet Intern Med 2003;17:830–834.
29. Day M. Immune-mediated anemias in the dog. In: Weiss DJ, Wardrop JK, eds. Schalm's veterinary hematology. 6th ed. Ames, Iowa: Wiley-Blackwell, 2010;216–225.
30. Carli E, Tasca S, Trotta M, et al. Detection of erythrocyte binding IgM and IgG by flow cytometry in sick dogs with Babesia canis canis or Babesia canis vogeli infection. Vet Parasitol 2009;162:51–57.
31. Smith JR, Smith KF, Brainard BM. Platelet parameters from an automated hematology analyzer in dogs with inflammatory clinical diseases. Vet J 2014;201:406–411.
32. Romão F, Campos E. Hemostatic profile and thromboembolic risk in healthy dogs treated with prednisone: a randomized controlled trial. BMC Vet Res 2013;9:268.
33. Thong KL, Mant MJ, Grace MG. Lack of effect of prednisone administration on bleeding time and platelet function of normal subjects. Br J Haematol 1978;38:373–380.
34. Piek CJ, Junius G, Dekker A, et al. Idiopathic immune-mediated hemolytic anemia: treatment outcome and prognostic factors in 149 dogs. J Vet Intern Med 2008;22:366–373.
35. Reimer ME, Troy GC, Warnick LD. Immune-mediated hemolytic anemia: 70 cases (1988–1996). J Am Anim Hosp Assoc 1999;35:384–391.
36. Goggs R, Wiinberg B, Kjelgaard-Hansen M, et al. Serial assessment of the coagulation status of dogs with immune-mediated haemolytic anaemia using thromboelastography. Vet J 2012;191:347–353.
37. Swann JW, Skelly BJ. Systematic review of prognostic factors for mortality in dogs with immune-mediated hemolytic anemia. J Vet Intern Med 2015;29:7–13.
38. Swets JA. Measuring the accuracy of diagnostic systems. Science 1988;240:1285–1293.
