• Open Access

Sonoclot® Evaluation of Single- and Multiple-Dose Subcutaneous Unfractionated Heparin Therapy in Healthy Adult Dogs


  • This work was performed at the College of Veterinary Medicine, Athens, GA.
  • Portions of this work have been presented in abstract form at the International Veterinary Emergency and Critical Care Symposium 2010 (San Antonio, TX).

Corresponding author: B.M. Brainard, VMD, Department of Small Animal Medicine and Surgery, College of Veterinary Medicine, University of Georgia, 501 D.W. Brooks Dr, Athens, GA 30602; e-mail: brainard@uga.edu.



Heparin therapy is difficult to monitor due to variation in animal response. While laboratory measurements of activated partial thromboplasin time (aPTT) and Anti-Xa activity (AXA) accurately describe heparin effect, their availability is limited.


Sonoclot analysis would be as sensitive as AXA and aPTT to monitor effects of unfractionated heparin (UFH) in healthy adult dogs.


Six adult mixed-breed dogs.


A prospective study design was employed. On day 1, baseline samples were collected (CBC, PT, aPTT, and Sonoclot), and UFH (300 U/kg SC) was administered to 6 dogs following an IV loading dose of 50 U/kg. Sonoclot and aPTT were performed hourly for 12 hours. AXA was assayed at hours 3, 6, 9, and 12. UFH (300 U/kg q8 h SC) was administered at 12 hours, and subsequently (q8 h) for 2 additional days. On day 4, a final dose of UFH was administered, and a sampling protocol identical to day 1 was performed.


Sonoclot activated clotting time (ACT) and clot rate (CR) correlated with AXA (R = 0.69, R = 0.65, respectively, P < .001), although to a lesser degree than aPTT (R = 0.75, P < .001). Linear regression using ACT and CR as covariates indicated a stronger correlation with AXA (R = 0.73, P < .001). ACT values strongly correlated with aPTT (R = 0.87, P < .001).

Conclusions and Clinical Importance

Administration of UFH to healthy dogs results in progressive changes in Sonoclot values. AXA was correlated with a combination of ACT and CR and with aPTT. Sonoclot may play a role in monitoring UFH therapy; however, prospective studies evaluating its utility in clinical cases are warranted.


Sonoclot-derived activated clotting time


activated partial thromboplastin time




area under the curve


anti-Xa activity


complete blood count


Sonoclot-derived clot rate


Sonoclot-derived platelet function


prothrombin time


receiver-operator characteristic




unfractionated heparin

Critically ill dogs and cats might be susceptible to thrombosis and thromboembolic disease.[1] Increased platelet aggregation, excessive activation of coagulation factors, deficiencies of natural anticoagulants, and decreased fibrinolysis can result in a hypercoagulable state.[2-4] In companion animals, thromboembolic complications attributed to a hypercoagulable state have been identified in animals with immune-mediated hemolytic anemia, pancreatitis, sepsis, glomerular disease, neoplasia, endocrine disease, and heart disease.[4-7] Treatment for thromboembolic disease includes therapy for the underlying cause of hypercoagulability and administration of anticoagulant or antiplatelet drugs such as unfractionated heparin (UFH), clopidogrel, or aspirin.[8, 9]

Heparin therapy is difficult to monitor in animals due to wide variation in response to a standard dose.[9, 10] Unfractionated heparin therapy is traditionally monitored by evaluating activated partial thromboplastin time (aPTT).[11] Target prolongation of aPTT to 1.5–2.0 times the baseline has been accepted in clinical practice, but is still considered controversial and unreliable because of laboratory variation in the measurement of aPTT.[12] Determination of anti-Xa activity is a more accurate measure of circulating plasma heparin levels.[13, 14] The chromogenic anti-Xa activity assay, however, is not always available for same-day therapeutic adjustments, and a cage-side, clinically applicable, accurate, and precise test to monitor anticoagulation with UFH currently does not exist for dogs. Thrombelastography (TEG) has been reported to be useful in guiding heparin therapy in humans.[15] However, TEG using citrated whole blood activated with calcium chloride was unable to reliably distinguish plasma UFH concentrations greater than 0.075 U/mL, limiting its utility as a test for managing anticoagulation.[16] The therapeutic range for unfractionated heparin in dogs is believed to be between heparin concentrations of 0.35 and 0.7 U/mL.[13]

Point-of-care viscoelastic coagulation monitors measure the entire clotting process by tracking changes in viscosity caused by the process of coagulation. The Sonoclot measures changes in blood viscosity using a vertically oscillating probe immersed in a cuvette of whole blood.[17] Glass beads in the cuvette activate coagulation, and might make the Sonoclot a more accurate method to evaluate dogs receiving heparin therapy. The Sonoclot analyzer provides information on the entire hemostasis process in both a qualitative graph (Sonoclot Signature) and as quantitative results: the activated clotting time (ACT), the clot rate (CR), and the platelet function (PF).[17, 18] Reference intervals for the sonoclot parameters in healthy adult dogs are available.[19]

The objectives for this study were to evaluate the effects of in vivo UFH therapy on the Sonoclot signature of healthy adult dogs, and to correlate our findings with plasma aPTT and anti-Xa activity. We hypothesized that heparin would cause progressive, detectable changes in the Sonoclot signature. Specifically, we expected the ACT to become prolonged and the CR to decrease. We also hypothesized that Sonoclot analysis would correlate well with current tests used to monitor heparin therapy.

Materials and Methods

Baseline Measurements

Six random-source neutered dogs determined to be healthy based on a normal physical examination and normal complete blood count (CBC),1 serum chemistry,2 plasma antithrombin (AT)3 concentration and coagulation parameters (prothrombin time [PT]4 and activated partial thromboplastin time [aPTT])5 were used for this study. All study protocols were reviewed and approved by the Institutional Animal Care and Use Committee of the University of Georgia.

On day 0, all dogs were equipped with a sampling catheter6 placed in the right jugular vein using a modified Seldinger technique. An initial blood sample was collected to establish a baseline including a CBC, PT, aPTT, and Sonoclot (SCT). To collect samples, a 2 mL discard sample was drawn directly from the jugular catheter port into a 3 mL syringe.7 Following the discard sample, a 3.6 mL sample of blood was obtained directly from the catheter port via a 5 mL syringe7 and transferred by vacuum using a 20 SWG needle into tubes containing 3.2% sodium citrate,8 for a final citrate to blood ratio of 1 : 9. One tube reserved for SCT analysis was kept at room temperature (20–22°C, 68–72°F) for 30 minutes before evaluation. The other citrate tube was centrifuged within 10 minutes of collection at 1,500 × g for 10 minutes. The plasma portion was separated and aliquoted into plastic vials and stored at −80°C. During periods of hourly sampling, the catheter was flushed with 3 mL of 0.9% NaCl9 after samples were obtained. When not actively used for sampling (days 2 and 3), the catheter was flushed with 1 mL of a 50% dextrose solution10 every 8 hours to maintain catheter patency.[16]

Blood samples for Sonoclot analysis were rested at room temperature for 30 minutes prior to analysis. All Sonoclot samples were performed on a single chamber of a dual chamber Sonoclot analyzer.11 To minimize potential variation, the same chamber was used for every dog throughout the duration of the study. For each sample, 340 μL of citrated blood was added to a glass bead-activated (gbACT+) cuvette12 containing 20 μL of 0.2 M calcium chloride.13 All samples were allowed to run until an ACT, CR, and PF were recorded, generally less than 25 minutes. Prior to each study day, the machine calibration was verified as per manufacturer's instructions using an oil of known viscosity.14

Summary Study Protocol

On study day 1, dogs received an IV loading dose of UFH,15 followed by a single subcutaneous dose (discussed below as “single heparin dose protocol”). Coagulation parameters in these dogs were studied for 12 hours, after which a second phase of the study to evaluate the clinically relevant scenario of multiple subcutaneous heparin doses was begun. Phase 2 (described below in the “multiple heparin dose protocol” section) consisted of intermittent subcutaneous dosing of UFH for 2 days, terminating with a final subcutaneous dose on study day 4, after which the same coagulation parameters were studied for at least 10 hours.

Single Heparin Dose Protocol

A single loading dose of UFH (50 U/kg) was administered intravenously to each dog via direct venipuncture of the cephalic vein, immediately followed by a subcutaneous injection of UFH (300 U/kg) over the region of the left or right thorax. At 1-hour intervals, samples were obtained from the catheter for SCT and aPTT analysis as previously described. SCT analysis was performed on one of these samples and the other was centrifuged as previously described to obtain plasma for aPTT analysis. At the 3, 6, 9, and final (hour 11 or 12) time points, an additional (1.8 mL) sample was collected for evaluation of anti-Xa activity. Plasma from this tube was separated and stored as described above for batch analysis of anti-Xa activity. The Factor Xa inhibitory activity (anti-Xa activity) of UFH was measured in a commercial chromogenic assay,16 previously validated for use in dogs.[13] The assay was performed using the manufacturer's calibrators and quality control materials.17 In addition, aliquots of a canine plasma control sample that was spiked with 0.5 IU/mL UFH were assayed simultaneously with each test run. Results were reported as units of anti-Xa activity/mL (U/mL) relative to the human heparin standard. The final sample time was determined by a return to baseline Sonoclot values.

Multiple Heparin Dose Protocol

At 12 hours following the initial UFH dosing, the same dogs were administered UFH at a dose of 300 U/kg SC over alternating hemithoraces, every 8 hours for 2 days. Patency of the sampling catheter was maintained for the duration of the study by flushing with 1 mL of a 50% dextrose solution11 every 8 hours. Dogs were monitored 3 times daily for signs of hemorrhage or discomfort. Sampling procedures were not performed on day 2 or day 3.

On day 4, baseline samples were collected for evaluation of CBC, PT, aPTT, AXA, and SCT. Following this collection, a final dose of UFH (300 U/kg) was administered subcutaneously. As on day 1, blood was sampled from the jugular catheter at 1-hour intervals for aPTT and SCT analysis for 11–16 hours after UFH administration. Each sample was handled as described above. Sampling endpoints were determined by the return of the tracing to baseline Sonoclot values. After the final samples for aPTT, SCT, and AXA were collected, an additional 3.8 mL sample was collected for an additional CBC and PT. Jugular catheters were removed after the final sample collection.

Statistical Analysis

Statistical analysis was performed using commercial software programs.1819 Data were evaluated for normality using the Kolmogorov–Smirnov test, and are described as mean ± standard deviation (SD) for parametric data, and median and range for nonparametric data. Values for individual days were compared with the baseline values for that day using an ANOVA with repeated measures, followed by posthoc correction for multiple comparisons using the Holm-Sidak method, where indicated. Spearman correlation analysis was used to compare the Sonoclot parameters and AXA. Statistical significance was set at a P value < .05 for all initial comparisons.

Receiver operator characteristic (ROC) curves were generated to describe Sonoclot parameters associated with clinically relevant AXA activity (0.3–0.7 U/mL). Sensitivity and specificity were calculated and a receiver operator characteristic curve was plotted to determine optimal values to predict therapeutic plasma heparin concentrations.


Six adult, middle-aged, random-source dogs weighing an average of 18.0 kg (range 12.6–21.8 kg) were used. Two dogs were neutered males and 4 were spayed females. Baseline PT and aPTT and Sonoclot tracings were normal in all dogs.

Three dogs had baseline hematocrit and platelet count within institutional reference intervals on day 1. Three dogs had relatively decreased platelet counts (206, 226 and 231 × 103/μL [institutional reference interval 235–694 × 103/μL]), although review of a blood smear showed the presence of platelet clumping in two. A significant decrease in hematocrit was observed over the duration of the project comparing baseline values with both day 4 and the final sample (P < .001), but all dogs remained within our institutional reference range for hematocrit (36.6–59.6; Table 1). Thrombocytopenia was noted in 1 dog on day 4 (150 × 103/μL), and another dog had a decreased platelet count with platelet clumping (186 × 103/μL) on day 4. Platelet count and PT did not differ significantly between day 1 and day 4 (P = .071 and 0.534, respectively, Table 1). Baseline AT concentrations in the dogs were within the reference range (118 ± 6%, reference range = 65–145%).

Table 1. Hematocrit, platelet count, and prothrombin time (PT) in 6 dogs administered unfractionated heparin over 4 days. Time points are baseline on day 1, day 4, and at the end of day 4
 Hematocrit (%) (36.6–59.6)Platelet Count (×109/L) (226–500 × 109/L)Prothrombin Time (sec) (5.8–9.8)
  1. a

    Indicates a significant difference between baseline hematocrit and that on day 4 and at the end of study (P < .001).

Day 148.2a (±6)276 (±99)8.0 (±0.7)
Day 445.4a (±7)234 (±64)7.78 (±0.4)
End study39.2a (±7)211 (±52)N/A

aPTT and AXA Analysis

On day 1, no dog reached a target aPTT of 1.5–2.0 times baseline; however, at hour 3 and hour 6, 5/6 dogs reached recommended therapeutic heparin levels (0.35–0.7 U/mL) as determined by AXA. On day 4, 4/6 dogs reached therapeutic heparin effect based on aPTT prolongation of 1.5–2.0 times baseline and five of 6 dogs reached therapeutic heparin effect based on AXA. One dog never reached therapeutic heparin effect based on either test on day 4. This dog did, however, reach suggested AXA levels on day 1. One dog developed a large subcutaneous hematoma (on the left hemithorax) on day 4, and this dog is the only dog that had an aPTT value prolonged greater than 2 times baseline. AXA greater than 0.7 U/mL was not measured in any dog during the study (Tables 2 and 3). The aPTT for all dogs on day 1 was significantly different from baseline (T0) starting at hour 1, and continuing until hour 11 (P < .001, Table 2). For day 4, aPTT measures were significantly different from T0 starting at hour 1 and lasting until hour 6 (P < .001, Table 3). Compared with the baseline measure on day 1, the day 4 aPTT values from hour 1 through hour 10 were significantly prolonged (P < .001, Tables 2 and 3). On day 1, the AXA values at T3 were significantly greater than those at T6, T9, and T11 (P < .001, Table 2). On day 4, the AXA values at T3 and T6 were significantly prolonged compared with T12 (P = .002, Table 3). The AXA values for any measured time point (T3, T6, T9) were not different between day 1 and day 4.

Table 2. Coagulation data for 6 dogs treated with unfractionated heparin (50 U/kg IV, followed immediately by 300 U/kg SQ)
  1. a

    Significant difference from baseline (T0) values, P < .001.

  2. b

    Significant difference from T3 measurement, P < .001. In this table, Sonoclot parameters ACT, CR, and PF are described, in addition to activated partial thromboplastin time (aPTT) and plasma Anti-Xa activity (AXA).

0113 ± 1335.6 ± 10.03.48 ± 1.0110.5 ± 0.6Not measured
1239 ± 67a11.8 ± 5.2a2.53 ± 0.5814.3 ± 1.1a 
2260 ± 84a13.6 ± 7.1a2.80 ± 0.6214.0 ± 0.9a 
3227 ± 95a11.1 ± 4.3a2.75 ± 1.4413.5 ± 0.8a0.433 ± 0.08
4219 ± 59a14.3 ± 6.6a2.77 ± 1.4612.8 ± 1.1a 
5192 ± 6123.3 ± 19.2a2.52 ± 0.2612.6 ± 1.3a 
6184 ± 4418.4 ± 9.6a3.00 ± 0.4312.7 ± 1.0a0.267 ± 0.12b
7170 ± 2017.3 ± 8.3a3.00 ± 0.9112.5 ± 1.2a 
8174 ± 2919.1 ± 11.0a3.00 ± 0.4612.3 ± 1.0a 
9149 ± 3520.0 ± 10.7a3.07 ± 1.0712.1 ± 1.4a0.170 ± 0.13b
10149 ± 2422.7 ± 7.3a2.68 ± 1.1411.8 ± 0.8a 
11132 ± 2727.5 ± 7.73.33 ± 1.1611.8 ± 1.20.057 ± 0.08b
Table 3. Coagulation data for 6 dogs treated with 2 days of unfractionated heparin therapy (300 U/kg SQ q. 8 h) with an additional dose administered at time 0
  1. a

    Significant difference from baseline (T0) values, P < .001.

  2. b

    Significant difference from baseline measurement (T0) from day 1, P < .001.

  3. c

    Significantly different from AXA values at T3 and T6 (P = .002). For hours 12–16, analysis was performed on 5 dogs. Please see the legend for Table 2 for definitions of abbreviations.

0209 ± 12716.2 ± 11.2b3.48 ± 0.9012.7 ± 1.5 
1385 ± 148a9.2 ± 7.62.52 ± 1.11a15.8 ± 2.6a,b0.237 ± 0.19
2334 ± 1565.0 ± 3.2a0.94 ± 0.4816.5 ± 2.7a,b 
3334 ± 1276.5 ± 3.41.62 ± 1.3316.1 ± 2.5a,b0.418 ± 0.21
4320 ± 1217.9 ± 3.52.02 ± 0.5215.8 ± 2.4a,b 
5345 ± 18310.4 ± 9.51.30 ± 0.9515.5 ± 2.7a,b 
6337 ± 1559.4 ± 7.22.10 ± 0.4415.1 ± 2.7a,b0.367 ± 0.21
7292 ± 14413.2 ± 11.61.62 ± 1.2114.5 ± 2.7b 
8284 ± 16013.0 ± 14.62.30 ± 1.5814.0 ± 2.4b 
9230 ± 13911.6 ± 6.82.42 ± 0.6013.3 ± 2.5b0.238 ± 0.28
10240 ± 12617.5 ± 13.31.92 ± 0.8013.0 ± 2.0b 
11201 ± 10020.6 ± 15.12.37 ± 1.0912.6 ± 1.1 
12192 ± 3915.1 ± 6.52.80 ± 0.6412.6 ± 1.00.105 ± 0.12c
13164 ± 4020.7 ± 7.91.50 ± 0.0711.9 ± 1.1 
14129 ± 1427.8 ± 11.4a3.48 ± 0.4311.7 ± 0.6 
15127 ± 3434.4 ± 11.9a3.50 ± 0.2211.7 ± 0.8 
16128 ± 1630.4 ± 7.8a3.72 ± 0.4211.3 ± 0.5 

Sonoclot Analysis

Progressive, measureable changes were noted in the Sonoclot Signature as plasma heparin concentration increased; the ACT became increasingly prolonged, and the CR decreased (see Tables 2 and 3). ACT and CR values were obtained at every time point for all dogs, although timing necessitated the termination of some runs before the PF could be calculated. Baseline ACT and CR fell within the institutional reference interval (ACT 56.0–154.0 seconds, CR 14.9–46.0[19]) at hour 0 on day 1 except for 1 dog with a slightly steeper clot rate (52). The final sample for all dogs on day 4 also fell within the institutional reference interval.

In all dogs, the ACT was progressively prolonged after heparin administration, and was significantly different from baseline values at times T1–T4 on day 1 (P < .001), but only at T1 on day 4 (P < .001 compared to day 4 T0, Tables 1 and 2). Although comparatively prolonged, the day 4 T0 ACT was not significantly different from the day 1 T0 ACT (P = .160). Peak effect on ACT occurred at hour 2 on day 1 and hour 1 on day 4. ACT returned to reference intervals by hour 11 in all dogs on day 1, and in all dogs by hour 14 on day 4 (Table 2).

Clot rate decreased after heparin administration in all dogs. Peak effect on CR was seen at hour 3 on day 1, and at hour 2 on day 4. Differences from baseline CR were seen in dogs on day 1 at all time points except for T11 (P < .001, Table 2). Baseline CR on day 4 was significantly decreased from the baseline CR on day 1 (P = .001). On day 4, significant differences were seen from the T0 CR at T2, where it was prolonged, and at T14–T16, where it was increased (P < .001, Table 3). On day 1, CR in all dogs returned to the reference interval by hour 10. On day 4, CR in all dogs returned to the reference interval by hour 14.

There were no significant differences in the value for PF on day 1 (P = .795). The PF value for T0 on day 4 was not significantly different from the baseline PF (P = .121), and only the T1 value on day 4 was significantly different from T0 (P = .013).


Sonoclot ACT and CR each moderately correlated with AXA (R = 0.69, R = −0.65, respectively, P < .001), although to a lesser degree than the correlation between AXA and aPTT (R = 0.75, P < .001). Sonoclot platelet function (PF) was not significantly correlated (P = .68) with AXA. A multiple linear regression using ACT and CR as covariates [equation: AXA = 0.373 + (0.000450 × ACT) − (0.0127 × CR)] resulted in a stronger correlation with AXA than either ACT or CR alone (R = 0.81, P < .001). ACT values strongly correlated with aPTT (R = 0.87, P < .001). As the ratio of baseline aPTT to treatment aPTT is thought to estimate AXA by predicting therapeutic levels of heparin, this aPTT ratio (aPTT treatment/aPTT baseline) was evaluated for correlation with AXA. The ratio did not have a strong correlation with AXA (R = 0.57, P < .001, Fig 1).

Figure 1.

Correlation of anti-Xa activity (AXA) with the activated partial thromboplastin (aPTT) ratio (aPTT treatment/aPTT baseline) in dogs treated with unfractionated heparin. The solid line represents the linear regression, with an R value of 0.567 (P < .001).

The ROC curve for ACT had an area under the curve (AUC) of 0.81 (P < .0001), and indicated that an ACT of > 185.5 seconds provided a sensitivity of 82% and a specificity of 82% for AXA > 0.3 U/mL (Fig 2). The ROC curve for CR had an AUC of 0.88 (P < .0001), and determined that a CR of < 13.8 is associated with a sensitivity of 82% and specificity of 82% for anti-Xa activity > 0.3 U/mL (Fig 3). Using the multivariate equation combining ACT and CR, the AUC was 0.87 (P < .0001), and a value >0.27 provided a sensitivity of 79% and a specificity of 86% for anti-Xa activity > 0.3 U/mL. By comparison, an aPTT of > 12.45 seconds gave a sensitivity of 90% and a specificity of 71% for anti-Xa activity > 0.3 U/mL, with an area under the curve of 0.86 (P < .0001, Fig 4). The ROC curve for the aPTT ratio had an AUC of 0.86 (P < .0001) and indicated that an aPTT > 1.173 × animal baseline gave a sensitivity of 89.6% with a specificity of 77.78 for being >0.3 AXA (Fig 5).

Figure 2.

Receiver operator characteristic curve describing the ability of Sonoclot parameter ACT to predict a therapeutic anti-Xa activity (AXA) (> 0.3 U/mL) after unfractionated heparin dosing in 6 healthy dogs. The area under the curve (AUC) was 0.8068 (P < .0001), and indicated that an ACT of > 185.5 seconds provided a sensitivity of 82% and a specificity of 82% for AXA > 0.3 U/mL.

Figure 3.

Receiver operator characteristic curve describing the ability of Sonoclot parameter CR to predict a therapeutic Anti-Xa activity (AXA) (> 0.3 U/mL) after unfractionated heparin dosing in 6 healthy dogs. The area under the curve (AUC) was 0.8801 (P < .0001), and showed that a CR of < 13.8 is associated with a sensitivity of 82% and specificity of 82% for AXA > 0.3 U/mL.

Figure 4.

Receiver operator characteristic curve describing the ability of activated partial thromboplastin time (aPTT) to predict a therapeutic anti-Xa activity (AXA) (> 0.3 U/mL) after unfractionated heparin dosing in 6 healthy dogs. aPTT of > 12.45 seconds gave a sensitivity of 90% and a specificity of 71% for AXA > 0.3 U/mL, with an area under the curve of 0.8553 (P < .0001).

Figure 5.

Receiver operator characteristic curve (ROC) for activated partial thromboplastin time (aPTT) ratio (aPTT treatment/aPTT baseline) ability to predict a therapeutic anti-Xa activity (AXA) (> 0.3 U/mL) after unfractionated heparin dosing in 6 healthy dogs. The area under the curve was 0.8646 (P < .0001) and indicates that an aPTT > 1.173 × patient baseline has a sensitivity of 89.6% with a specificity of 77.8% for an AXA > 0.3 U/mL.


Plasma heparin inhibitory activity (as measured by AXA) up to 0.7 U/mL in healthy adult dogs resulted in progressive and measureable changes in the Sonoclot signature. The ACT became prolonged with increasing plasma heparin inhibitory activity, and the slope of the CR decreased. Sonoclot ACT and CR individually, as well as in combination, correlated well with AXA. The aPTT also correlated well with AXA; however, the aPTT ratio did not correlate as strongly with AXA. The Sonoclot is a sensitive and specific test for monitoring plasma heparin inhibitory activity, although changes in aPTT had equivalent correlations and sensitivity.

Significant changes in the Sonoclot signature of human patients receiving heparin during hemodialysis have been described.[18] A significant linear correlation was shown between heparin dose and ACT and CR; however, changes in clot rate proved most useful for monitoring heparin therapy in this population.[18] In the current study, neither ACT nor CR appears to have an advantage over the other in identifying dogs with AXA > 0.3 U/mL. Overall, Sonoclot parameters did not correlate more strongly than aPTT with AXA.

The Sonoclot showed quantifiable changes even at high AXA, which was not demonstrated in a prior study using recalcified TEG.[16] The use of an activator for TEG analysis, analogous to the glass beads used in the Sonoclot gbACT+ cuvettes, might result in increased sensitivity to therapeutic levels of heparin, although this hypothesis remains to be tested.

Although correlation of AXA activity with the aPTT ratio was moderate, the ROC curve analysis indicated that in this population, a ratio closer to 1.2 times the baseline value would be a better value to indicate plasma heparin inhibitory activity > 0.3 U/mL. Failure of the recommended aPTT ratio of 1.5–2.0 times baseline to strongly correlate with therapeutic heparin levels based on AXA might be a result of our institution's reagent and testing methodology. The aPTT ratio of a canine plasma sample with a defined heparin activity depends on the reagent used, and these differences reflect variations in the chemical composition of the activator (phospholipid type, contact activator), activation time, and the method used (mechanical versus photooptical).[12] Consequently, it is recommended that each laboratory should establish an individual therapeutic aPTT ratio range for its own reagent and testing methodology.[12] There is also marked individual variation with regard to heparin dosing; and variables such as plasma antithrombin (AT) concentration or concurrent inflammatory states can also influence how a specific dose of UFH affects coagulation measures.[9] As the concentration of AT decreases with prolonged heparin therapy,[10] the possibility also exists that the drug might become less potent as treatment is continued, exacerbating the difference between a functional assay such as aPTT and an assay that is directed specifically at quantifying the factor Xa inhibition.

Administration of 300 U/kg of UFH SC to healthy adult dogs produced recommended therapeutic levels of heparin (based on AXA) in multiple dogs during this study period. More dogs reached therapeutic levels on day 4 versus day1 based on both prolongation of aPTT and AXA. The 1 dog in this study that developed a subcutaneous hematoma was the only dog to have a prolongation of her aPTT beyond the recommended upper limit of 2.0 times baseline (maximum ratio of 2.12 × baseline). Her AXA at the time that this increased aPTT ratio was measured was 0.6 U/mL, still within the presumed therapeutic range of 0.35–0.7 U/mL. This may represent a delayed heparin effect, where the anticoagulant performance of UFH is related more to the degree of thrombin inhibition, which persists even as AXA values are decreasing, although further study is necessary to investigate this hypothesis.

Evaluation of the aPTT ratio might be more predictive of the possibility for bleeding during heparin therapy, although the timing of the measurement after heparin dosing should be standardized. Only 4 dogs achieved an aPTT ratio >1.8, yet they remained between 0.6 and 0.7 U/mL of plasma heparin concentration derived from AXA. On the basis of results of this study and the possibility for an increased risk of clinical bleeding, we have revised our institutional guidelines to target an aPTT prolongation of 1.2–1.8 times the dog's baseline. The variability of the aPTT ratio vis-a-vis plasma AXA or heparin concentrations has been previously reported by Mischke[20] who noted wide variability in the aPTT ratio for a given UFH concentration depending on the type of reagent and protocol used to measure aPTT. This variability has also been reported in a number of human studies.[21-23] In another study of UFH in dogs,[12] the correlation between the aPTT ratio and the amidolytic AXA showed an R value of 0.877 or 0.890, depending on the aPTT reagent. The correlation coefficients are comparable to those in the current study for the aPTT:AXA correlation alone, but much better than the aPTT ratio:AXA correlation described herein. Testing methodology for both the coagulation aspects, as well as the AXA measurement, might have contributed to these differences, and it is recommended that each institution generate an individual reference interval if the aPTT is to be used for monitoring of heparin therapy.[24]

Sonoclot analysis of citrated whole blood has some benefits over the laboratory measurement of aPTT, especially if coagulation assays must be sent out to a reference laboratory. The standardization of the Sonoclot assay across machines (all cuvettes are packaged with identical glass bead activators) might eliminate the problems inherent in comparing aPTT measurements across different coagulometers or reagents. The Sonoclot is a cage-side machine, requires a small sample volume, and provides rapid analysis (generally under 25 minutes). In addition, the Sonoclot generates a global picture of hemostasis that is easily interpreted, both visually and using quantitative parameters. Sonoclot disposables are inexpensive (approximately $5/test at the time of writing), and maintenance, operation, quality control, and set-up of the instrument are uncomplicated. For these reasons, the Sonoclot analyzer may be a useful point of care coagulation monitor for monitoring of UFH therapy that can be reasonably managed in veterinary laboratories or ICUs.

There are several limitations in this study. First, the study was performed in a small number of dogs, which becomes a limitation because of known individual variation in response to dosing of UFH.[10, 25] Ideally, AXA would have been measured at every time point to confirm plasma heparin inhibitory activity. As noted above, the effects of a certain plasma level of UFH on aPTT measurement will depend on the machine utilized for the coagulation analysis. For this reason, the results of this study cannot necessarily be generalized to all monitors and dogs, and individual institutional reference intervals should be established for monitoring of UFH therapy in dogs.

Although Hct decreased significantly over the study period, we do not believe that this resulted in altered evaluation of the effects of heparin on whole blood coagulation. Although it is known that Hct can affect the results of viscoelastic coagulation monitoring, a recent study evaluating pre- and postcardiopulmonary bypass showed no significant difference in ACT values, despite a significant decrease in hematocrit,[26] and it is possible that the Sonoclot (with glass bead activator) might not be as sensitive to alterations in hematocrit as other viscoelastic coagulation monitors. The Hct for most dogs also remained within the institutional reference range, and large variations were not seen. We suspect that our frequent sampling of blood over the course of the study contributed to the drop in hematocrit. The hematocrit of the dog that had subcutaneous bleeding secondary to the heparin treatment dropped from 42% at time 0 on day 1 to a hematocrit of 31% by the end of the study, and this contributed to the overall decrease in Hct.

UFH produces measurable effects on the Sonoclot Signature. ACT was gradually prolonged, and the clot rate gradually decreased at increasing plasma heparin concentrations. As the heparin effect waned, these parameters also returned toward normal. Sonoclot ACT and CR, both separately and together, significantly correlated with anti-Xa levels. The aPTT also correlated well with both ACT and AXA; however, the aPTT ratio correlated only fairly with AXA. The Sonoclot analyzer is a sensitive and specific test that is useful for monitoring UFH therapy in healthy adult dogs. Further prospective studies in critically ill patients receiving heparin therapy are warranted.


This work was supported by a grant from the American College of Veterinary Emergency and Critical Care (ACVECC) Foundation.


  1. 1

    Advia, Bayer Corp, Shawnee Mission, MO

  2. 2

    P module, Hitachi, Tarrytown, NY

  3. 3

    TriniCHROM Antithrombin IIa, T Coag, Inc, Parsippany, NJ

  4. 4

    Triniclot PT Excel, T Coag, Inc

  5. 5

    Alexin, T Coag Inc

  6. 6

    18 g × 12 cm single lumen catheter, Arrow Teleflex Inc, Reading PA

  7. 7

    Syringe, Monoject, Covidien, Mansfield MA

  8. 8

    Vacutainer, B–D, Franklin Lakes, NJ

  9. 9

    0.9% Sodium Chloride, Hospira, N. Chicago, IL

  10. 10

    Dextrose 50%, Hospira, Inc

  11. 11

    Sonoclot Coagulation and Platelet Function Analyzer, model SCP-2, Sienco, Arvada, CO

  12. 12

    Sienco Inc

  13. 13

    0.2 M CaCl2, Haemoscope/Haemonetics, Boston, MA

  14. 14

    SonOil, Sienco Inc

  15. 15

    Heparin sodium, 1,000 U/mL, Baxter Healthcare Corporation, Deerfield, IL

  16. 16

    Rotachrom Heparin, Diagnostica Stago, Parsippany, NJ

  17. 17

    Hepanorm and Heparin Control, Diagnostica Stago

  18. 18

    SigmaStat v 3.5, Systat Inc, Chicago IL

  19. 19

    GraphPad Prism, v.5.0, GraphPad Software, La Jolla CA