Assessment of a dimethyl sulfoxide–stabilized frozen canine platelet concentrate (2024)

Materials and Methods

Control dogs and platelet preparation—Blood was collected from 6 dogs owned by staff veterinarians or veterinary students from the Veterinary Medical Teaching Hospital at the University of California, Davis. Consent was obtained from owners prior to study enrollment; dogs and blood were handled according to institutionally approved protocols. Dogs (5 castrated males, 1 spayed female; median age, 5.5 years [range, 3 to 13 years]) were considered healthy on the basis of history, physical examination findings, and results of a CBC.^d None of the dogs had received drugs during the 10 days preceding blood collection.

Fresh whole blood was obtained by venipuncture and immediately transferred into a commercial specimen tube^e containing acid-citrate-dextrose to achieve a 1:7 acid-citrate-dextrose–to–blood ratio. Platelet-rich plasma as well as platelet-rich plasma with 6% DMSO was prepared for subsequent studies.

Platelet-rich plasma was prepared by centrifugation^f,g of whole blood at 2,000 × g for 3 minutes at 23°C within 30 minutes of blood collection. Platelet-poor plasma (ie, < 10,000 platelets/μL) was prepared by centrifugation of the platelet-rich plasma at 20,000 × g for 10 minutes. Dimethyl sulfoxide was also added to the previously prepared platelet-rich plasma to achieve a final 6% concentration, gently mixed, and allowed to rest for ≥ 10 minutes at 37°C before analysis.

Platelet concentrate—Eleven units of frozen canine PC were purchased,^a and each PC was placed in a −20°C freezer immediately upon receipt according to the instructions of the manufacturer.¹⁸ All analyses were performed within 10 days of receipt of the product. The PC was thawed at room temperature, visually inspected, and gently swirled every 5 minutes during the thaw process. The time at which the PC was entirely fluid was defined as the thaw time. The PC was allowed to rest 15 minutes before starting analysis (T0), and analysis was completed within 4 hours of thawing, without washing the product.

Hematologic and metabolic analysis—At T0, pH, PCO₂, and PO₂ and HCO₃⁻, glucose, and lactate content were measured with an intensive care unit analyzer.^h Thaw time was also recorded.

For hematologic analysis, platelet count and MPV were measured at T0 and T0 + 2 hours by use of an automated cell counter.^d The cell counter used a pulsatile change in the electric field for cell count and sorted cells on the basis of size. Pulses that represent cells from 2 to 20 fL were classifieds as platelets. Platelet pulses were sorted by size into 64 channels to produce a platelet histogram, and the MPV was derived from the platelet histogram. Accuracy of results was maximized by triplicate counting and internal voting criteria.

Freeze-thaw recovery was calculated at T0 and T0 + 2 hours; it was defined as the concentration of platelets after thaw divided by the concentration of platelets stated by the manufacturer to be present in the bag prior to freezing (ie, 1 × 10⁶/μL)¹⁸ multiplied by 100.

Annexin V labeling of surface phosphatidylserine and flow cytometry—Translocation of phosphatidylserine from the inner membrane leaflet of platelets to the outer membrane leaflet is a sensitive indicator of platelet activation that can be detected by use of FITClabeled annexin V, which specifically binds to phosphatidylserine. Five frozen PCs were used for the flow cytometry analysis. Fresh platelet-rich plasma and fresh platelet-rich plasma with 6% DMSO were used as a source of control platelets. Platelets in each product were diluted in Tyrode-HEPES buffer (12mM NaHCO₃, 138mM NaCl, 2.9mM KCl, 10mM HEPES, 5mM dextrose; pH, 7.2). Calcium chloride from a 10mM stock in Tyrode-HEPES was added in 5 equal aliquots at 5-minute intervals to achieve a final concentration of 1mM CaCl₂ and 5 × 10⁷ cells/mL. The FITC-labeled annexin Vⁱ was used to detect phosphatidylserine translocation to the outer membranes of platelets from control dogs and PC.

Fresh platelet-rich plasma, fresh platelet-rich plasma with 6% DMSO, and frozen PCs were analyzed at rest and after activation. Washed platelets were activated with thrombin^j (1 U/mL) for 15 minutes at 37°C. Platelets (5 × 10⁵) were removed from the reaction, diluted to 100 μL with Tyrode-HEPES plus 1mM CaCl₂, and labeled with FITC-labeled annexin V according to the recommendation of the manufacturer. Flow cytometric^k analysis was performed after a 10-minute incubation at 37°C.

Forward and side scatter voltages were set to detect machine noise, which was removed during subsequent analyses. In experiments with FITC-labeled annexin V, the fluorescence channel 1 detector was set such that 95% of the autofluorescence of the control platelets was binned in the first logarithm of the detector response.

Statistical analysis—Nonparametric measures of comparison were used because platelet count, MPV, and FTR were not normally distributed. Differences between groups were assessed by use of a Mann-Whitney U test.^l Data were reported as median and range values (n = number of samples). Flow cytometry data were normally distributed; therefore, a Student t test was used for data analysis. A value of P < 0.05 was considered to be significant.

Results

Hematologic and metabolic analyses—Analyses were completed on 11 PCs (Table 1). Visual and microscopic examination of the PC at thaw revealed swirling platelets, platelet aggregates, and air bubbles. The aggregates were of different sizes, ranging from 2 mm to 2 cm in diameter. Over time (up to 6 hours), the swirling decreased, and aggregates decreased in number but increased in size (up to 5 to 6 cm in diameter). Median thaw time was 77.5 minutes (range, 70 to 100 minutes; n = 6).

Platelet count at T0 ranged from 243,000 to 742,000 platelets/μL (median, 590,500 platelets/μL; n = 10) with an FTR of 59%. Using the 5 PCs for which we had paired analyses, median platelet counts decreased from 680,000 platelets/μL at T0 (range, 577,000 to 742,000 platelets/μL; n = 5) to 509,000 platelets/μL at T0 + 2 hours (range, 496,000 to 568, 000 platelets/μL; 5; P = 0.009). The FTR was 68% at the time recommended to begin a transfusion and decreased to 51% (P = 0.009) after 2 hours, the time recommended by the manufacturer to have concluded the transfusion.

Median MPV of the fresh platelets was 6.75 fL (range, 5.6 to 8.5 fL; n = 6). Median MPV of the PC was 11.15 fL at T0 (range, 9.6 to 13.8 fL; n = 10), significantly (P = 0.001) greater than that of fresh platelets. For the paired sample, median MPV was similar at T0 and T0 + 2 hours, 11.1 fL (range, 9.6 to 13.8 fL; n = 5) and 11.1 fL (range, 9.7 to 17.4 fL; 5), respectively (P = 0.83; Tables 2 and 3).

Flow cytometry—In the absence of cryopreservative (ie, DMSO), the percentage of fresh platelets with detectable phosphatidylserine increased from 19 ± 2.5% in the resting state to 67 ± 7.7% after stimulation with thrombin (Figure 1). This increase in the percentage of fresh platelets with detectable phosphatidylserine was accompanied by an increase in the amount of phosphatidylserine on their surface, which was reflected by an increase in the MFI from 280 ± 52 RFU to 1,100 ± 220 RFU.

In contrast, 64 ± 14% of fresh platelets in 6% DMSO in the resting state had phosphatidylserine detected on their surface with a mean MFI of 1,200 ± 460 RFU. These results were not significantly different from the treatment of resting platelets with thrombin (P = 0.761 and P = 0.751 for percentage and RFU values, respectively). The fresh platelets in 6% DMSO responded to thrombin, although this response was not significant. After stimulation with thrombin, the percentage of fresh platelets in 6% DMSO that were positive for annexin V was 84 ± 3.2% and had a mean MFI of 1,900 ± 310 RFU (P = 0.174 and P = 0.119 for percentage and RFU values, respectively, compared with fresh platelets in 6% DMSO in the resting state).

Platelets in the resting state obtained from frozenthawed PCs were 99 ± 0.92% positive for phosphatidylserine on their surfaces and had a mean MFI of 1,300 ± 560 RFU (Figure 1). After thrombin activation of these platelets, 98 ± 0.82% were positive for phosphatidylserine and had a mean MFI of 1,400 ± 280 RFU. Therefore, detection of phosphatidylserine on platelets in the resting state obtained from frozen-thawed PC did not change significantly after stimulation with thrombin (P = 0.107 and P = 0.730 for percentage and RFU values, respectively).

The progressive activation of platelets in response to the DMSO treatment and cryopreservation was also apparent in the physical characteristics measured by use of flow cytometry, forward scatter and side scatter. After stimulation with thrombin, a typical population of untreated (ie, no treatment with DMSO) fresh platelets in the resting state shifted to lower forward and side scatter values (Figure 2), which corresponded to the change in shape that platelets undergo after activation. Activation of non-DMSO incubated fresh platelets with thrombin was also accompanied by a significant (P < 0.001) increase in the mean number of microparticles from 910 ± 140 to 6,570 ± 1,040. Incubation of fresh, resting platelets with 6% DMSO resulted in a split platelet distribution in which 32 ± 11% of the platelets had lower forward and side scatter values, characteristic of platelets activated with thrombin. The number of microparticles, 1,530 ± 90, in these preparations also increased, compared with that of untreated fresh platelets in the resting state. Thrombin activation of fresh platelets in 6% DMSO induced a shift in the population that was indistinguishable from thrombin-activated fresh platelets that were not treated with DMSO. Although the difference was not significant (P = 0.082), slightly fewer microparticles were produced (4,450 ± 1,210).

Unstimulated platelets from frozen PCs were shifted to lower forward and side scatter values in a manner characteristic of activated platelets. These platelets had a compact distribution, consistent with a uniform population of cells, which did not change further upon stimulation with thrombin. Interestingly, no increase in the number of microparticles was observed in platelets from PC either before or after stimulation with thrombin (760 ± 200 and 840 ± 200 [P = 0.545], respectively).

Discussion

Variability was found in platelet numbers between the units of the commercial DMSO-stabilized frozen PCs as well as loss of platelets during the freeze-thaw process and after thaw. The platelet population in the frozen PC had maximal amounts of phosphatidylserine at rest and after addition of the platelet agonist thrombin. This is consistent with the presence of an activated platelet population in the frozen PC at thaw.

Prior studies^16,19 evaluating frozen canine PC reported platelet concentration corresponding to 1.6 to 5 × 10¹¹ platelets/100-mL bag after thaw. In veterinary medicine, there is currently no universally accepted definition of platelet numbers in a PC. The manufacturer of the commercial frozen PC indicated that the initial PC was 1.0 × 10¹¹ platelets/100-mL bag, similar to the definition used in human medicine. In our study, the median platelet count in PCs was 5.90 × 10¹⁰ platelets/100-mL bag, corresponding to a 59% recovery rate, consistent with the previously published literature. The veterinary and human literature report a recovery of 60% to 70%, depending on storage temperature, duration of storage, and blood collection procedure.^{13,16,17,19,20} The FTR may be further decreased by washing the product before transfusion or by the use of a traditional 18-μm transfusion filter.

The term platelet storage lesion is used in human medicine to describe structural, biochemical, and functional alteration of platelets caused by platelet product collection, preparation, and storage.^21,22 Storage lesions are influenced by several factors such as mechanical manipulation during platelet apheresis or centrifugation,^21,23 storage conditions (eg, cooling at 4°C^24,25 or freezing methods^13,17,19), storage containers,²⁶ and the use of DMSO²⁷ (the most commonly used cryopreservative for human and canine platelets^17,28,29). Those changes are considered deleterious and are the hallmark of platelet activation.²⁶

Increases in platelet volume, as we found in our study, have been used as an indicator of platelet storage lesions. Because of the technology used for size measurement, increased MPV can be caused by platelet microaggregates, cell swelling, shape change (from discoid to spherical), or extension of pseudopods or can represent platelets with a longer lifespan.^21,26,30 Increased MPV and irreversible loss of discoid shape are associated with increased reactivity³¹ and can be induced by prolonged cold (4°C) storage.²⁵ Microvesiculation, or the generation of microparticles, is associated with storage lesions and is considered a marker of activation.²⁶ Furthermore, these platelet microparticles are reported to have greatly increased procoagulant activity.³² Flow cytometric analysis of the frozen PC revealed an increase in the number of microparticles, as well as smaller platelets. As noted, platelet microaggregates can lead to a general increase of MPV in the product.

Platelet storage lesions also are characterized by accumulation of lactic acid and acidosis, with a pH below 6.2 associated with decreased in vivo survival.^8,22,33 Aging of the PC is associated with decreased PCO₂ and glucose content in both human³⁴ and veterinary^d studies. The acid-base status of the studied PCs is consistent with the use of citrate-phosphate-dextrose-adenine as the anticoagulant and storage agent.^16,34,35,d In our study, the PO₂ and PCO₂ were higher in the platelet storage container than in room air (176 mm Hg, compared with 160 mm Hg in room air, and 37.9 mm Hg, compared with 0.25 mm Hg in room air, respectively) and may be a sign of decreased gas exchange between the PC and the environment, which is highly dependent on the plasticizer used.³⁶ Plasticizers are chemical components added to the hard plastic bag to improve its flexibility and durability. Each one has innate chemical and physical characteristics that modify gas exchange through the collection system.³⁶

Loss of cellular function associated with room temperature platelet storage lesions has been reported in the human literature.³³ Increased amounts of phosphatidylserine or P-selectin on the platelet surface, decreased aggregometry response to agonist, or decreased recovery after hypotonic stress reaction are all evidence of a loss of platelet function.³³ In our experiment, the platelets in the frozen concentrate already had maximal amounts of phosphatidylserine at rest, compared with fresh platelets. Moreover, platelets in the frozen concentrate are not able to translocate additional phosphatidylserine to the outer leaflet of the plasma membrane when stimulated with thrombin, whereas the fresh platelets and the fresh platelets in 6% DMSO could. In fresh platelets, with or without 6% DMSO, a portion of the population (33% and 16%, respectively) remained unactivated, even in the presence of thrombin, because the concentration used in these experiments (1 U/mL) is insufficient to induce translocation of phosphatidylserine in 100% of the fresh platelet population.³⁷ However, the concentration of thrombin used in this experiment is closer to physiologic amounts than that which would be required for 100% translocation of phosphatidylserine.

In conclusion, findings in our study revealed that the commercial frozen PC had a decreased platelet number after thaw. The flow cytometric analysis revealed that the platelets in the frozen concentrate already had maximal amounts of phosphatidylserine and, when stimulated with thrombin, were unable to translocate additional phosphatidylserine to the outer leaflet of the platelet plasma membrane as required for platelet activation.³²