• Open Access

Cryopreservation of Canine Platelets


  • This study was performed at the School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA.

  • Abstract was presented at the 2008 ACVIM Forum in San Antonio, TX.

Corresponding author: Mary Beth Callan, School of Veterinary Medicine, University of Pennsylvania, 3900 Delancey Street, Philadelphia, PA 19104-6010; e-mail: callan@vet.upenn.edu.


Background: Platelet cryopreservation allows long-term storage and immediate availability of transfusion products.

Hypothesis: The addition of a preparation inhibiting platelet activation (Thrombosol, in 2% dimethyl sulfoxide [DMSO]) will enhance in vitro function and prolong in vivo survival of cryopreserved platelets compared with those preserved in 6% DMSO.

Animals: Thirty-three research dogs.

Methods: Prospective study. Eleven fresh canine apheresis platelet concentrates (PCs) were each split into 3 units: fresh and cryopreserved in 6% DMSO or Thrombosol. Platelet analysis, performed 1–10 weeks postfreezing, included in vitro functional testing and in vivo survival assessed by administration of biotinylated platelets.

Results: Platelet aggregation was diminished in cryopreserved PC. Cryopreserved platelets could be activated, as based on mean thrombin-stimulated P-selectin expression (6% DMSO, 23.0%; Thrombosol, 18.4%), although to a lesser extent than fresh PC (49.1%) (P < .0001). The mean maximum in vivo platelet recovery for fresh PC was 80.3%, significantly greater than recovery for 6% DMSO (49.2%) and Thrombosol PC (43.7%) (P≤ .001). The half-life (days) of fresh PC (3.8 ± 0.4) was significantly (P < .002) greater than that of 6% DMSO (1.9 ± 1.0) and Thrombosol (2.4 ± 1.1) PC, with no difference (P= .3) between cryopreserved PC.

Conclusions and Clinical Importance: Cryopreservation of canine platelets using Thrombosol did not provide any advantage over preservation using 6% DMSO. Cryopreserved platelets can be activated in vitro and provide therapeutic benefit when fresh platelets are unavailable. Further studies are needed to assess their in vivo hemostatic function.

Platelet transfusions are indicated in the management of patients with uncontrolled or life-threatening hemorrhage due to severe thrombocytopenia or thrombopathia. In most veterinary facilities, canine platelet products are not readily available and must be prepared on an as-needed basis. Platelet transfusion options include fresh whole blood (FWB), platelet-rich plasma (PRP), and platelet concentrate (PC), with the latter 2 harvested from FWB following centrifugation at 1,000 ×g for 4 minutes for generation of PRP and then subsequent centrifugation of PRP at 2,000 ×g for 10 minutes for production of PC. An alternative approach to preparation of a high-quality canine PC is plateletpheresis.1,2 Regardless of collection method, short-term storage restrictions are placed on all fresh platelet products, which must be stored at room temperature (22 °C) due to the deleterious effect of refrigeration on platelet viability posttransfusion.3 Concerns over bacterial proliferation at room temperature limit storage of human PRP and PC to 5 days.3 In vitro studies evaluating biochemical and functional changes in canine PC during room temperature storage have also supported a time limit of up to 5 days.4–6

During the past few decades, platelet cryopreservation has been investigated as a means to provide long-term storage and immediate availability of platelet products for transfusion. The American Association of Blood Banks (AABB) has approved cryopreservation of human platelets in 6% dimethyl sulfoxide (DMSO) as an acceptable storage procedure, although cryopreserved platelets inevitably demonstrate impaired in vitro function and shortened posttransfusion survival in comparison with fresh platelets.7 Recent studies have focused on a novel method for cryopreservation of human platelets utilizing 2% DMSO plus Thrombosol (note: 2% DMSO plus Thrombosol will be referred to as Thrombosol throughout the article), a mixture of second messenger effectors that inhibit premature platelet activation, which may allow significant improvement in platelet function and survival compared with the standard 6% DMSO method.7,8 However, at this time, cryopreservation of human platelets is largely considered a research technique, and human platelet products for transfusion are stored solely at room temperature.3

Despite the limitations of cryopreservation, ready availability of a stored platelet product with acceptable function and viability would be invaluable for veterinary clinics, particularly those without access to fresh platelet products. Two previous reports evaluating cryopreservation of canine platelets offer somewhat differing conclusions regarding efficacy.4,9 Therefore, the aim of this study was to evaluate 2 methods for cryopreservation of canine platelets, the freezing of platelets in 6% DMSO and Thrombosol in a −80 °C freezer, on in vitro function and in vivo survival in comparison with fresh canine platelets to determine if either cryopreservation method results in an acceptable product for transfusion. It is hypothesized that canine platelets cryopreserved in Thrombosol will have enhanced in vitro function and prolonged in vivo survival compared with platelets cryopreserved in 6% DMSO.

Materials and Methods

Platelet Cryopreservation

Fresh PC (each unit containing approximately 3 × 1011 platelets in 210 mL volume) was collected via apheresis from 11 hematologically normal research dogs as previously described.2a All handling of the PC units took place under a laminar flow hood. Within 1 hour of collection, a 5-mL aliquot was removed from the platelet collection bag for aerobic bacterial culture (0.5 mL PC placed into thioglycollate broth) and in vitro evaluation of fresh platelets. The PC was then divided into 3 separate units: fresh, cryopreserved in 6% DMSO, and cryopreserved in Thrombosol. The fresh PC was stored in the original platelet collection bag (citrated-polyvinylchloride) for no longer than 2 hours before transfusion. For each of the cryopreserved units, 70 mL of PC was transferred to a 500 mL cryocyte freezing container.b For 6% DMSO cryopreservation, 23 mL of 24% DMSO (5.5 mL 100% medical grade DMSOc diluted with 17.5 mL autologous fresh plasma collected during plateletpheresis) was added gradually over 20 minutes with gentle, continuous mechanical rotation at room temperature. A 50-fold concentrated (50 × ) solution of Thrombosol (12.5 mM amiloride, 5 mM adenosine, and 2.5 mM sodium nitroprusside, SNP) was prepared in 100% medical grade DMSO and sterile filtered using a 0.2 μm DMSO-compatible filter. For Thrombosol cryopreservation, 1.4 mL of 50 × solution was added rapidly to 70 mL PC (1 : 50 ratio), yielding a final DMSO concentration of 2%, and the PC unit was mixed with gentle, continuous mechanical rotation for 20 minutes. A 0.5 mL aliquot was obtained from Thrombosol-treated platelets for aerobic culture before freezing. Immediately following mixing, the cryocyte containers were placed horizontally in an aluminum freezing cassetted in a −80 °C freezer, which is usually associated with a cooling rate of 2–3 °C/min, and stored for 1–10 weeks before analysis.

In Vitro Platelet Evaluation

Metabolic Evaluation of PC and Platelet Count and Recovery. Cryopreserved platelets were thawed in a 37 °C waterbath for 5 minutes and manually mixed gently. Neither the 6% DMSO- nor the Thrombosol-cryopreserved platelets were washed after thawing for in vitro evaluation or before transfusion. A 5 mL aliquot was removed from each thawed unit for in vitro evaluation. Within 1 hour of fresh PC collection or postthawing, acid-base/electrolyte analysise was performed and platelet counts were measured using an automated hematology analyzer.f PC samples were diluted (1 : 2 to 1 : 5) with phosphate-buffered saline, as needed, to obtain a platelet count within the measurement range of the analyzer. In vitro platelet recovery for the cryopreserved platelets was calculated as follows: in vitro platelet recovery (%) = total platelet number of thawed PC ÷ total platelet number of PC prefreezing. Total platelet number = platelet concentration of PC (per μL) × volume of PC (mL) × 1,000.

Platelet Morphology. Platelet morphology was assessed by a single board-certified clinical pathologist (RP) who was blinded to the PC group. Fresh and cryopreserved PC (50 μL) was pipetted onto a glass slide and coverslipped. Platelets were then examined using a differential interference contrast microscopeg with a 40 × phase objective. Two hundred platelets were counted and categorized as spherical, discoid, or unclassified (ie, platelets that were not clearly discoid or spherical).

Flow Cytometric Analysis of Platelet Activation. Platelet suspensions were prepared and analyzed in a flow cytometerh as previously described.10 Baseline P-selectin expression was evaluated to determine if cryopreservation resulted in increased platelet activation compared with fresh platelets. In addition, P-selectin expression was evaluated postthrombin stimulation (0.5 U thrombin added per milliliter of platelet suspension, with gentle mixing for 25 minutes at room temperature) to determine if cryopreserved platelets could be activated to the same extent as fresh platelets. Binding of 0.1 μg anti-human CD61-fluorescein isothiocyanate (FITC)i to the platelet membrane glycoprotein (GP) IIIa was used to identify platelets. Binding of 0.045 μg CD62-phycoerythrinj (PE) was used to detect P-selectin exposure. Isotype control antibodies, mouse IgG1-FITC,k and mouse IgG1-PE,l were included in the analyses.

Platelet Aggregation. Fresh and cryopreserved (unwashed) platelets were diluted in autologous platelet-poor plasma to achieve a final platelet count of 300,000/μL. The PRP was allowed to rest at room temperature for 30 minutes before use. Platelet aggregation was studied in a standard aggregometer,m using convulxin (final concentration 20 nM) and γ-thrombin (final concentration 100 nM) as platelet agonists.

Platelet Transfusion and In Vivo Platelet Survival

Fresh and cryopreserved PCs were labeled with N-hydroxysuccinimide ester (NHS) biotinn for determination of in vivo platelet survival. Briefly, NHS-biotin, suspended in 100% medical-grade DMSO to a concentration of 20 mg/mL, was added to fresh and cryopreserved PC for a final concentration of 0.3 mg/mL. The biotinylated PC was mixed gently with continuous mechanical rotation for 10–20 minutes at room temperature before transfusion.

Allogeneic biotinylated PCs were administered via a standard blood filter over 20–30 minutes to 33 healthy mixed breed dogs randomly assigned to the PC groups. The platelet donor and recipient dogs were descendents from outbred dogs within a research colony comprising 12 different lines of dogs, resulting in low level of inbreeding. Platelet donors were not closely related to platelet recipients, although siblings were used as platelet recipients. Three of 33 platelet transfusion recipients were excluded from analysis of in vivo platelet survival; 1 dog (fresh PC group) was eliminated after being inadvertently vaccinated 1 day following transfusion, and 2 dogs (1 each in 6% DMSO and Thrombosol PC groups) were excluded due to technical difficulties with the flow cytometer. To avoid potential problems with platelet alloimmunization and, therefore, shortened platelet survival, the recipients had not been previously transfused or pregnant. The weight of platelet recipients ranged from 7 to 20 kg. All dogs were monitored clinically (heart rate, pulse quality, respiratory rate, and temperature) during and 2 hours following the transfusion for signs of any adverse reaction, and a brief clinical assessment was performed daily for approximately 1 week. A CBC was performed immediately pre- and 2 hours posttransfusion. In addition, EDTA-anticoagulated blood was obtained from each recipient 2 hours following completion of the platelet transfusion and then once daily until biotinylated platelets could no longer be detected.

Binding of 5 μg PE-labeled streptavidino was used to detect biotinylated platelets, as previously described.11 To optimize binding, PE-labeled streptavidin and 0.1 μg CD61-FITC antibodies were incubated separately. For each analysis, nonspecific platelet binding by streptavidin was assessed by mixing 5 μg streptavidin-PE and 0.1 μg CD61-FITC antibodies with nonbiotinylated EDTA blood samples from control dogs.11,12

In vivo platelet recovery, defined as the percentage of transfused biotinylated platelets that survived in the recipient, was determined at 2 hours or 1-day posttransfusion. In vivo recovery was calculated as follows: % platelet recovery = (measured % circulating biotinylated platelets) ÷ (calculated % transfused platelets) × 100. The calculated % transfused platelets = total number of platelets transfused (platelet count of PC [/μL] × volume of PC [mL] × 1,000) ÷ calculated total number of platelets in recipient posttransfusion (total number of platelets transfused + total number of platelets in recipient pretransfusion). The recipient's total number of platelets pretransfusion was calculated (based on an estimated total blood volume [TBV] of 90 mL/kg) as TBV (mL) × recipient's pretransfusion platelet count [/μL] × 1,000. This calculation assumes that 100% of platelets in the PC were biotinylated before transfusion. For determination of platelet half-life, the measured maximum % circulating biotinylated platelets (at 2 hours or 1-day posttransfusion) was assigned a value of 100% and plotted on the y-axis versus time (days) on the x-axis. Platelet half-life was determined by the intersection on the x-axis where the y-axis coordinate =50%. Platelet lifespan was defined as the length of time (days) that the percentage of remaining biotinylated platelets was recorded at >1%.

The experimental protocols described above were approved by the Institutional Animal Care and Use Committee of the University of Pennsylvania.

Statistical Analysis

Standard software was used for analysis.p Data were divided into 3 groups: fresh, 6% DMSO, and Thrombosol PC. Normally distributed continuous variables were summarized as mean ± standard deviation, whereas median and range was used to describe nonnormally distributed continuous variables. Pairwise comparisons using t-tests were used to compare continuous variables that were normally distributed between groups, and the Mann-Whitney test or Wilcoxon's signed rank test was used to compare continuous variables that were not normally distributed. Values of P≤ .05 were considered significant.


In Vitro Platelet Evaluation

Platelet Number, Yield, and Recovery and Metabolic Evaluation of PC. The mean in vitro platelet recoveries, total platelet numbers, and platelet counts (per μL) of fresh and cryopreserved PC are summarized in Table 1. Mean in vitro platelet recovery was significantly higher for Thrombosol versus 6% DMSO units (P= .02). The pH of the 6% DMSO PC was significantly higher than the pH of fresh PC and Thrombosol PC, although all were within a physiologic range (Table 1). There were no significant differences in mean ionized calcium concentration or median lactate concentration between any PC groups (Table 1). Aerobic bacterial cultures from all tested platelet units (11 fresh apheresis PCs and 11 Thrombosol PC units) yielded no growth.

Table 1.   In vitro evaluation of fresh and cryopreserved (6% DMSO and 2% DMSO plus Thrombosol) canine apheresis platelet concentrates.
 Fresh6% DMSOThrombosol
  • a

    Values provided as median and range. All other values reported as mean ± SD.

  • b

    Thrombosol versus 6% DMSO, P= .02.

  • c

    6% DMSO versus fresh, P= .03; 6% DMSO versus Thrombosol, P= .005.

  • d

    6% DMSO versus fresh; Thrombosol versus fresh, P≤ .0001.

  • e

    6% DMSO versus Thrombosol, P= .02.

  • DMSO, dimethyl sulfoxide.

Platelet count (per μL)
 Prefreezinga1,409,273 (934,000–2,605,000)1,060,561 (785,000–1,960,750)1,381,640 (915,680–2,553,920)
 PostthawingaN/A1,029,500 (696,000–1,494,000)1,317,800 (956,000–2,200,000)
Total platelet number
 Prefreezinga1.2 × 1011 (8.2 × 1010–2.2 × 1011)9.9 × 1010 (6.5 × 1010–1.8 × 1011)9.9 × 1010 (6.5 × 1010–1.8 × 1011)
 PostthawingaN/A7.0 × 1010 (5.5 × 1010–1.1 × 1010)9.0 × 1010 (6.3 × 1010–1.6 × 1011)
In vitro platelet recovery (%)N/A72.8 ± 15.2b95.2 ± 23.3
Metabolic parameters
 pH7.34 ± 0.17.42 ± 0.07c7.33 ± 0.07
 Lactate (mmol/L)a2.0 (1.3–7.3)2.0 (1.1–5.4)2.4 (1.6–6.3)
 Ionized calcium (mmol/L)0.23 ± 0.020.21 ± 0.020.23 ± 0.02
Platelet morphology (%)
 Discoid48.8 ± 17.541.6 ± 11.834.5 ± 12.5
 Spherical48.9 ± 17.052.7 ± 11.061.7 ± 12.2
 Unclassified2.3 ± 2.05.7 ± 3.83.8 ± 1.9
P-selectin expression (%)
 Unactivateda0.3 (0.2–2.6)0.3 (0.3–1.3)0.4 (0.1–0.9)
 Thrombin-stimulated49.1 ± 9.923.0 ± 7.6d18.4 ± 6.3d
Platelet aggregation (%)
 γ-thrombin (100 nM)a64 (43–78)10 (0–37)e,d0 (0–16)d
 Convulxin (20 nM)79 ± 4.113 ± 8.8d17 ± 9.6d

Platelet Morphology. Phase microscopic evaluation of PCs did not reveal any significant differences in mean percentage of discoid and spherical platelets between fresh and cryopreserved PC, or between 6% DMSO and Thrombosol PC (Table 1).

Flow Cytometric Analysis of Platelet Activation. P-selectin expression was found in 0.3–0.4% of unactivated platelets, secondary to the plateletpheresis procedure or cryopreservation, with no significant difference between 6% DMSO and Thrombosol PC, or fresh and cryopreserved PC (Table 1). Following thrombin stimulation, platelet P-selectin expression increased in all PC groups, indicating that cryopreserved platelets can be activated. There was no difference in mean thrombin-stimulated P-selectin expression between the cryopreserved platelets, but both exhibited significantly less activation than fresh platelets (Table 1).

Platelet Aggregation. Cryopreserved platelets demonstrated minimal ability to aggregate in vitro. Both γ-thrombin–induced and convulxin-induced platelet aggregation were significantly decreased in 6% DMSO and Thrombosol PCs in comparison with fresh PC (Table 1). The median amplitude of γ-thrombin–induced platelet aggregation was significantly higher in platelets cryopreserved with 6% DMSO compared with Thrombosol; no difference in convulxin-induced platelet aggregation was noted between cryopreservative groups (Table 1). An initial trial of washing both the 6% DMSO and Thrombosol cryopreserved platelets and resuspending in autologous plasma did not improve aggregation responses (data not shown) and was, therefore, discontinued.

Platelet Transfusion and In Vivo Platelet Survival

There was no difference in median pretransfusion platelet count among the dogs in the 3 transfusion groups. There was a significant increase in platelet count posttransfusion only in the fresh PC group (Table 2), with a median increase of 27,000/μL and individual increases ranging from 10,000 to 134,000/μL. The 1 dog achieving the highest increase in platelet count posttransfusion was the smallest of the 33 recipient dogs (7 kg), and the fresh PC had the highest total platelet number (2.2 × 1011). The median change in platelet count after transfusion was significantly higher in recipients of fresh PC in comparison with cryopreserved PC (fresh versus Thrombosol, P= .0013; fresh versus 6% DMSO, P= .0047) with no significant difference among cryopreserved PCs (P= .6692). When the total number of platelets administered per transfusion was normalized for the weight of each recipient dog, dogs in the fresh PC group received significantly (fresh versus 6% DMSO, P= .02; fresh versus Thrombosol, P= .003) more platelets than dogs in either cryopreserved PC group (Table 2).

Table 2.   Characteristics and in vivo recovery and survival of fresh and cryopreserved (6% DMSO and 2% DMSO plus Thrombosol) allogeneic canine apheresis platelet concentrates administered to healthy dogs.
 Fresh (n = 11)6% DMSO (n = 11)Thrombosol (n = 11)
  • a

    6% DMSO versus fresh, P= .02; Thrombosol versus fresh, P= .0003.

  • b

    Values provided as median and range. All other values reported as mean ± SD.

  • c

    Platelet count 2 hour posttransfusion versus pretransfusion, P= .0035.

  • d

    6% DMSO versus fresh; Thrombosol versus fresh, P < .001.

  • e

    6% DMSO versus fresh, P < .0001; 6% DMSO versus Thrombosol P= .3.

  • f

    Thrombosol versus fresh, P= .0016.

  • DMSO, dimethyl sulfoxide.

Total number of platelets administered per kg body weight of recipientb9.9 × 1096.0 × 109a6.8 × 109a
(7.5–31.4 × 109)(3.8–13.3 × 109)(4.5–9.5 × 109)
Recipient platelet count (per μL)
 Pretransfusionb372,000 (227,000–516,000)335,000 (191,000–471,000)329,000 (281,000–534,000)
 2 hour posttransfusionb411,000c (246,000–540,000)337,000 (237,000–439,000)326,000 (252,000–564,000)
 Fresh (n = 10)6% DMSO (n = 10)Thrombosol (n = 10)
Maximum in vivo platelet recovery (%)80.3 ± 14.849.2 ± 19.1d43.7 ± 8.9d
Platelet half-life (days)3.8 ± 0.51.9 ± 1.0e2.4 ± 1.1f

Administration of fresh and cryopreserved PCs was well tolerated by recipient dogs. The amount of DMSO infused with the cryopreserved PCs ranged from 0.07 to 0.15 mL/kg (mean 0.11 mL/kg) in the Thrombosol group and 0.34 to 0.61 mL/kg (mean 0.45 mL/kg) in the 6% DMSO group. During transfusion of the cryopreserved platelets, several dogs exhibited lip-licking, and an odor consistent with DMSO was noted on the breath of all dogs. Pallor of the oral mucous membranes was noted in 1 dog (11 kg) during administration of 6% DMSO-cryopreserved platelets; however, heart rate, respiratory rate, pulse quality, and temperature remained normal. After temporary discontinuation of the transfusion for 10 minutes, the dog's coloring quickly improved and remained so for the duration of the transfusion. This same dog also developed facial swelling 1-hour posttransfusion that resolved within 2 hours following treatment with diphenhydramine (2 mg/kg IM). No adverse reactions were noted in any other dog during the platelet transfusion or the 1-week follow-up period.

The mean maximum in vivo platelet recovery for fresh PC was significantly greater than maximum in vivo recoveries for either cryopreserved PC group (Table 2). Six of 10 dogs in the fresh PC group and 2 of 10 dogs in the Thrombosol PC group had a higher percentage of biotinylated platelets present 1-day posttransfusion in comparison with 2 hours posttransfusion; the maximum percentage of biotinylated platelets in the remainder of dogs, including all 10 dogs in the 6% DMSO group, occurred at 2 hours posttransfusion. Fresh biotinylated platelets survived 7–9 days in all 10 recipients, whereas only 4 of 10 dogs in each cryopreserved group had >1% circulating biotinylated platelets on day 7 (Fig 1). Fresh platelets had a significantly longer mean half-life than both 6% DMSO and Thrombosol cryopreserved platelets (Table 2).

Figure 1.

 Percentages of mean (with 95% confidence intervals) in vivo platelet recovery of fresh and cryopreserved canine apheresis platelet concentrate (PC) assessed by allogeneic transfusion of biotinylated platelets. Mean in vivo recovery of fresh PC was significantly higher than both cryopreserved PCs at all time points except time ≥8 days. No difference was detected in mean in vivo recovery between Thrombosol and 6% dimethyl sulfoxide (DMSO) PC.


Results of this study demonstrate that Thrombosol did not provide any appreciable advantage for improving in vitro function or prolonging in vivo survival of cryopreserved platelets in comparison with 6% DMSO. Canine platelets cryopreserved in either 6% DMSO or Thrombosol can be activated in vitro, as demonstrated by thrombin-stimulated P-selectin expression, and survive in the circulation long enough to potentially be of benefit in the management of life-threatening hemorrhage in severely thrombocytopenic or thrombopathic patients. While cryopreserved PCs fall far short of fresh PC, particularly in terms of in vivo platelet recovery, their administration should be considered for patients with acute, life-threatening bleeding when fresh platelet products are unavailable. Either method of platelet cryopreservation may be a reasonable and more rapid alternative to fresh platelet products in veterinary clinics without ready access to canine blood donors.

Limited information is available regarding the efficacy and feasibility of cryopreservation of canine platelets. Human blood banks typically rely on fresh platelet transfusions prepared from collection of whole blood or apheresis due to the technical challenges of platelet cryopreservation, as well as the considerable loss of platelets and impaired function associated with the freeze/thaw process.7 As apheresis is not commonly available in veterinary medicine, fresh canine platelets are predominantly collected in whole blood on an as-needed basis. In this study, PC collected via apheresis provided a high-quality product, as evaluated through in vitro functional studies and assessment of in vivo lifespan. In comparison with PC prepared from centrifugation of a unit of FWB, the apheresis PCs reported here demonstrated greatly increased platelet numbers with negligible RBC and WBC contamination.2

Regardless of the collection method, fresh platelet products remain a limited resource in veterinary blood banks. Consequently, cryopreservation of canine platelets may offer a practical and much-needed alternative to use of fresh preparations. In this study, the cooling rate during the freezing process was not measured, but standardization of the PC volume (total volume of 93 mL for 6% DMSO and 71.4 mL for Thrombosol PC) in each cryocyte bag (all 500 mL) would be expected to result in a uniform rate of freezing within each PC group. It is well established in human medicine that all cryopreservation techniques cause platelet damage known as the “storage lesion,” which encompasses a wide array of biochemical, structural, and functional alterations.13 The AABB requires a postthaw washing step to remove 6% DMSO from cryopreserved platelets before transfusion in humans.7 However, platelet washing requires technical proficiency and equipment that may not be available in many veterinary clinics. Because commercially available canine platelets cryopreserved in 6% DMSO are currently administered without washing, neither 6% DMSO nor Thrombosol cryopreserved PCs were washed postthawing in order to mimic as closely as possible the situation in clinical practice.

Thrombosol is a mixture of second messenger effectors including amiloride, adenosine, and SNP that are proposed to reversibly inhibit premature platelet activation.7 Comparative studies of human platelets cryopreserved in 6% DMSO versus Thrombosol demonstrated enhanced in vitro and in vivo characteristics for the Thrombosol group, including a higher percentage of platelets retaining discoid morphology, lower expression of baseline platelet P-selectin, and improved posttransfusion platelet recovery and survival time.7,8 Both 6% DMSO and Thrombosol-cryopreserved human PCs were washed in these studies. However, a proposed benefit of the use of Thrombosol is the ability to reduce the final percentage of DMSO in the platelet cryopreservative solution, thereby allowing elimination of the washing step.8

There is a considerable loss of platelets during the freeze/thaw/wash process with all methods of cryopreservation, as quantified by the in vitro platelet recovery.14 Studies of 6% DMSO cryopreserved human platelets reported in vitro recoveries ranging from 50 to 80%.15–17 Similar in vitro recoveries of 70–80% have been reported in studies of canine platelets cryopreserved in 6% DMSO.4,9 In this study, 6% DMSO platelets had a mean in vitro recovery of 72.8%, significantly lower than that of Thrombosol platelets (95.2%, P= .02). However, as in vitro recoveries of over 100% were documented for 3 of the 11 Thrombosol cryopreserved PC units, platelet count measurements and, thus, calculation of in vitro recoveries may have been inaccurate. There are several potential sources of error in measurement of platelet counts for this study, including platelet clumping or insufficient mixing of PC units before counting. Also, platelet counts were obtained by an impedance hematology analyzer validated to count platelets in whole blood, not in PRP or PC, as the counting mechanism is directly dependent on the presence of red blood cells.18 Dilution of platelet samples has also been reported to influence the accuracy of the platelet count.18 These factors likely affected the platelet counts reported in this study, and subsequently the percentages of in vivo and in vitro recoveries, as those percentages are calculated based on platelet count measurements.

A variety of in vitro tests have been utilized to assess the quality of human platelet components for transfusion, although no single assay accurately predicts posttransfusion platelet viability and function.14,19 Expression of P-selectin, a platelet α granule GP exteriorized to the platelet surface following granule release, is considered a marker of platelet activation.7,20,21 In studies of cryopreserved human platelets, baseline (or unstimulated) P-selectin expression was increased (35%) compared with fresh platelets (9%), suggesting premature platelet activation during the freeze-thaw-wash process.17 Cryopreserved human platelets showed reduced P-selectin expression after ADP stimulation, reflecting their decreased ability for normal activation.22 Unexpectedly, minimal P-selectin expression was noted in the unstimulated fresh and cryopreserved canine platelets in this study, potentially due to technical differences in flow-cytometric gating that have been previously noted to cause difficulty in directly comparing interlaboratory values of P-selectin expression.23 Both 6% DMSO and Thrombosol cryopreserved canine platelets can be activated, as demonstrated by increased thrombin-induced P-selectin expression, although to a lesser extent than fresh platelets.

Platelet aggregometry is often considered the “gold standard” of in vitro platelet functional testing. The 2 platelet agonists, convulxin and γ-thrombin, utilized in this study were chosen for their reproducibility in promoting canine platelet aggregation in fresh PRP. Given that previous studies of cryopreserved canine (6% DMSO)4 and human (6% DMSO and Thrombosol)8,22 platelets have demonstrated markedly diminished aggregation responses, it was not surprising that both γ-thrombin- and convulxin-induced aggregation were significantly impaired in cryopreserved canine platelets in this study. Impaired in vitro platelet aggregation does not necessarily correlate with reduced in vivo survival and hemostatic function, as it has been recognized that stored human platelets have the ability to rapidly reverse some aspects of the storage lesion posttransfusion.22

Based on the limitations of various in vitro assays to predict platelet viability posttransfusion, assessment of in vivo platelet survival is essential in the evaluation of various platelet storage options. Radiolabeling of platelets with chromium-51 (51Cr) or indium-111 (111In) is typically utilized for platelet kinetic studies in humans. Platelet biotinylation represents an alternative method for determining platelet lifespan while avoiding regulatory issues associated with use of radio-isotopes. Previous studies have demonstrated that the lifespan of biotinylated and radiolabeled platelets is equivalent in dogs.11 In this study of canine PCs, the mean maximum in vivo platelet recovery for the fresh PC was 80.3%, in comparison with 49.2 and 43.7% for 6% DMSO and Thrombosol PC, respectively. These recovery values are similar to those reported in a previous study of canine platelets cryopreserved in 6% DMSO, with autologous fresh apheresis platelets and washed cryopreserved apheresis platelets having mean in vivo recoveries of 83 and 40%, respectively.9 As documented in this study, in vivo platelet recovery has been reported to be higher 1-day posttransfusion in comparison with 1–2 hours posttransfusion in both dogs9 and humans24 receiving fresh platelets, suggesting that some transfused platelets may be sequestered, perhaps in the spleen or liver, shortly after transfusion.

As expected, in vivo platelet lifespan was significantly longer for fresh versus cryopreserved PCs in this study. In Valeri's study of autologous platelet transfusions in dogs, fresh apheresis platelets had a lifespan of 8 days and a half-life of 3.5 days, almost identical to the results of this study, whereas the 6% DMSO cryopreserved apheresis platelets (stored for <6 months) had a half-life of 2 days.9 One notable difference between studies is the administration of autologous platelets by Valeri and allogeneic platelets in the present study. Although naturally occurring platelet alloantibodies in recipient dogs could affect survival of transfused platelets, there is no information in the literature describing the incidence of such antibodies in dogs. It is clear from experimental studies using a canine model for development of platelet transfusion refractoriness that dogs will develop platelet alloantibodies after repeated transfusions.25 Donor selection played a key role in the frequency of development of platelet alloantibodies, with 95% of recipients developing platelet alloimmunization after an average of 3 transfusions from unrelated donors, while the administration of platelets from littermates (DLA identical or nonidentical) resulted in 31% of recipients developing platelet alloimmunization.25 Given that the in vivo platelet recovery and half-life of both fresh and 6% DMSO cryopreserved platelets in the present study of allogeneic platelet transfusions were identical to those of Valeri's study of autologous platelet transfusions, the presence of naturally occurring platelet alloantibodies resulting in shortened survival of cryopreserved platelets seems unlikely.

No significant improvement in canine platelet survival was detected with Thrombosol in comparison with 6% DMSO cryopreservation. A significant increase in mean platelet count after transfusion was documented only for recipients of fresh PC. This may be partially attributed to the significantly higher number of fresh platelets transfused per kilogram of body weight as compared with cryopreserved platelets. In addition, due to the lower in vivo recovery of cryopreserved platelets, it has been estimated that approximately 2.5 U of cryopreserved PCs must be given to achieve a comparable increase in platelet count as for a single unit of fresh PC in thrombocytopenic human patients.7 Based on the calculated maximum in vivo recoveries of 49.2 and 43.7% for DMSO and Thrombosol PC, respectively (assuming minimal change in recipient TBV following transfusion), the expected increase in recipient platelet count following transfusion of cryopreserved PC ranged from 20,000 to 70,000 platelets/μL. Given the inherent error in automated whole blood platelet count measurements, ie, measured platelet count is generally assumed to be ±20,000/μL, relatively small expected increases in the recipient dogs' platelet counts may not have been detected. Hemostatic efficacy of the cryopreserved PCs was not evaluated in this study and warrants further investigation. Valeri reported a reduction in clinical bleeding (ie, reduction in occult blood in stool, ecchymosis, and bleeding from venipuncture sites) following administration of 6% DMSO cryopreserved platelets to lethally irradiated thrombocytopenic dogs.9

Transfusion of cryopreserved platelets was associated with a transient adverse event in only one of 22 recipient dogs in this study. Reported adverse effects of administration of DMSO cryopreserved cells (most commonly hematopoietic progenitor cells) to human patients range from mild, such as nausea, vomiting, flushing, fever, and abdominal pain, to severe complications, including dyspnea, hypo- or hypertension, and abnormalities of the renal, cardiovascular, and neurologic systems.26 Given that administration of 6% DMSO cryopreserved platelets was well tolerated by 10 of 11 dogs in this study receiving ≤10 mL/kg of PC, washing of platelets to remove DMSO does not appear necessary when moderate volumes of PC are transfused. However, administration of a large volume of 6% DMSO cryopreserved platelets to critically ill dogs may warrant additional monitoring for potential adverse effects of DMSO.

A major limitation of this study is the inherent difficulty in obtaining accurate platelet counts in PC samples, a problem not unique to this study, as discussed above. Calculation of in vitro and in vivo platelet recoveries is dependent on accurate determination of platelet counts in the PCs. While the in vitro and in vivo platelet recoveries reported in this study are comparable to those previously reported for fresh and cryopreserved platelet transfusions in dogs and humans, error may exist due to inaccurate platelet counts. In addition, the small number of dogs in this study may have precluded the ability to determine differences in platelet function and survival between the 6% DMSO and Thrombosol PCs. Further studies are needed to assess in vivo hemostatic function of cryopreserved canine platelets.


This study was funded in part by a grant from the ACVIM Foundation, a departmental research grant from the University of Pennsylvania, and NIH RR02512.

The authors acknowledge Dr Mark Haskins for generously providing dogs for this study and the staff of the Clinical Laboratory and Emergency Service for technical assistance.


aCOBE Spectra, Gambro BCT, Lakewood, CO

bCryocyte freezing container—500 mL PL 269 plastic, Baxter, Deerfield, IL

cCryoserv cryopreservative solution, Edwards Lifesciences, Bedford, OH

dFreezing cassettes, Custom Biogenic Systems, Shelby Township, MI

eStat Profile Critical Care Xpress, Nova Biomedical Corporation, Waltham, MA

fScil Vet abc, Scil Animal Care Company, Gurnee, IL

gLaborlux S phase contrast microscope, Leitz, Wetzlar, Germany

hFACSCalibur, Becton Dickinson, San Jose, CA

iAnti-human CD61, clone Y2/51, Dako, Carpinteria, CA

jAnti-CD62-PE, clone AC1.2, Becton Dickinson

kMouse IgG1-FITC, clone DAK-GO1, Dako

lMouse IgG1 PE control, Becton Dickinson

mModel 560Ca, Chrono-log Corporation, Havertown, PA

nBiotin-NHS, EMD Chemicals Inc, Gibbstown, NJ

oStreptavidin R-PE, Invitrogen Corporation, Carlsbad, CA

pStata 8.0 for Windows, Stata Corporation, College Station, TX