Effects of whole blood storage in a polyolefin blood bag on platelets for acute normovolemic hemodilution

Acute normovolemic hemodilution (ANH) is a potential transfusion method for platelets, as well as for red blood cells. However, previous studies have shown that whole blood storage in ANH decreases platelet aggregability by 14.7–76.3% and that this decrease is not recovered by reinfusion. We investigated whether a new whole blood storage method for 6 h using a polyolefin bag, based on the platelet concentrates storage method, would maintain platelet function better than the conventional method using a polyvinyl chloride bag. We demonstrated that storage of whole blood in a polyolefin bag maintained ADP-induced aggregation rates at more than twofold higher than those in a polyvinyl chloride bag, and also significantly suppressed P-selectin expression, a platelet activation marker (ADP-induced aggregation rates: 24.6 ± 5.1% vs. 51.7 ± 11.5%, p = 0.002; P-selectin expression; 50.3 ± 8.4MFI vs. 31.6 ± 9.3MFI, p = 0.018). These results could be attributed to the high gas permeability of polyolefin, which lowered PCO2 and maintained a high pH with or without agitation. There were no significant changes in platelet count and red blood cell parameters due to the storage methods. Our results suggest that ANH using polyolefin bags is advantageous in improving hemostatic function compared to the conventional method.

The storage conditions and expiration date of each blood production are shown in Table 1 1,[15][16][17][18][19][20][21][22][23] . The major differences in storage for ANH and platelet concentrates are the bag material and use of agitation. Few studies mention the material of the storage bags used for ANH, but polyvinyl chloride (PVC) storage bags for red blood cells are generally used because of their excellent durability despite their low gas permeability 1,8,13,16,19,21 . In contrast, platelet concentrates are stored with agitation in bags made of highly oxygen-permeable materials such as polyolefin 16,22 , which provide sufficient oxygen for platelets to maintain aerobic respiration and ensure gas exchange between the storage medium and the atmosphere 24 .
We hypothesized that use of a PVC bag at rest is responsible for the significant reduction in platelet function in conventional ANH. Therefore, we examined if whole blood storage in a polyolefin bag with or without agitation could maintain platelet function more effectively than that achieved by the conventional method.

Results
The results are shown in Table 2 and Fig. 2 (see "Methods" for the storage groups). First, the pH of venous blood mixed with CPDA before 6 h storage was lower than the normal venous blood pH (7.31-7.41) 25 (Table 2 PRE). Compared to PRE, pH and ADP-induced platelet aggregation rates in PVN storage decreased significantly, and lactate and P-selectin increased significantly (Table 2 and Fig. 2 PRE-PVN). These results show that whole blood stored in the conventional ANH method severely impairs platelet function, as previous studies have shown [11][12][13][14] (Fig. 1). In contrast, whole blood storage in polyolefin bags maintained significantly higher pH and ADP-induced platelet aggregation rates and significantly lower PCO 2 and P-selectin, compared to those in the PVN group (Table 2 and Fig. 2 PON-PVN). PO 2 in the PON group was relatively high, but not significantly higher than in the PVN group. No parameters in the PVA group differed significantly from those in the PVN group, and the POA and PON group showed significant differences from PVN in the same parameters. There were no significant changes in platelet count, mean platelet volume, and red blood cell parameters among all groups. The conventional method of ANH cannot maintain platelet function. Cardiopulmonary bypass reduces platelet counts and function decreases to 57% and 10% of the respective preoperative values 10 . ANH also markedly reduces platelet function and these changes do not recover after transfusion [11][12][13][14] . This figure was drawn using Microsoft PowerPoint 2016 (Ver. 2103, https:// www. micro soft. com/ ja-jp/ micro soft-365/ power point).

Discussion
In this study, we showed that whole blood storage in a polyolefin bag maintained higher ADP-induced platelet aggregation rates and lower P-selectin expression, compared to the conventional method used in ANH (Table 2 and Fig. 2). The decrease in aggregation rate measured by light transmittance aggregometry correlates with the amount of bleeding 26,27 . P-selectin (CD62P) is a membrane protein present in α granules of platelets, and its expression level on the unstimulated platelet surface is a common measure of platelet activations 28 . Platelets stored in the container are affected by various factors including duration of storage, temperature, pH, solution, gas permeability of the container for O 2 and CO 2 and interruption of agitation 29 . The platelet injuries associated with extracorporeal storage are called platelet storage lesions (PSLs) 30 , which lead to platelet activation via the necrotic and apoptotic processes, resulting the reduction of survival in vivo and hemostatic activity after  29 . Thus, decline of agonist-induced aggregation rates, P-selectin expression, morphological change and increased lactate level are the major manifestations of PSLs. ANH using polyolefin bags is advantageous in reducing PSLs and improving hemostatic function over the conventional method. Some studies have shown that platelet function of whole blood stored in blood bags markedly decreases and does not recover after transfusion in ANH. Ramnarine et al. showed that collagen-induced formation of large stable hemostatic aggregates (platelet macroaggregation) were markedly decreased to 29.4% after collecting blood into a CPDA bag and to 14.7% after 80 min of storage, compared to that before the collection, and that this deterioration was largely irreversible by the reinfusion of ANH 12 . Scott et al. showed that blood in bag storage with CPDA for 300 min had significantly reduced platelet function, as measured by thrombin receptor activating peptide stimulation in multiple electrode aggregometry (Multiplate) analysis and maximum clot formation on ROTEM EXTEM 13 . Gallandat Huet et al. showed platelet aggregation response measured by Multiplate in ex-vivo stored blood with heparin decreased compared to the pre-CPB levels 11 . Therefore, a new preservation method is required to maintain the platelet function in ANH.
In our study, storage with highly gas permeable polyolefin maintained lower PCO 2 and higher pH compared to PVN group (Table 2 and Fig. 2). Of ATP production by platelets, 85% is derived from aerobic metabolism 24,29 . In the 1980s, platelet concentrates were stored in PVC bags, but could not maintain the platelet function because of the accumulation of CO 2 and lactate followed by a rapid decrease of pH 31 . Low pH generally causes platelet morphological change (below 6.7) and irreversible loss of viability (below 6.2) 30,32 . The whole blood stored in PVC bags in our study also showed a significant decrease in pH, but these declines were insufficient to cause the morphological changes. The polyolefin bags currently used for storage of platelet concentrates are 2.2 times more oxygen permeable and 3.6 times more CO 2 permeable than PVC bags 22,33 . Use of high gas-permeable containers with gentle agitation ensures O 2 and CO 2 exchange between the storage medium and the atmosphere, and prevents local hypoxia resulting in lactate production 22,24 . This storage condition prevents PSLs and increases the ability to stop bleeding after transfusion 22,29 . However, it is unclear whether the gas permeability of the bag affected the aerobic metabolism of platelets in our results. Krause et al. showed that just limiting gas exchange in storage bag of platelet concentrates causes accumulation of PCO 2 , decreased pH, increased lactate, and increased P-selectin expression 34 . Similarly, the high gas permeability of the storage bag in our study may have led to higher aggregation rates and low P-selectin expression, although the underlying biochemical pathway is unknown. Mean platelet volume of platelets generally increases inversely related to pH, indicating a poor quality of product 35 , which did not change in this study. In contrast, there were no significant differences caused by agitation (Table 2 and Fig. 2, PVN-PVA, PON-POA). Thomas 24 showed that interruption of agitation for several hours did not reduce platelet function. In addition to agitation, the platelet concentration and the surface area of the storage bag affect the oxygen partial pressure 33 . The effect of agitation on the platelet count and function may depend on the storage container, agitation speed and other conditions.
Preservative solution and temperature also influence PSLs. CPDA used in ANH is an anticoagulating solution suitable for long term storage of red blood cells in preoperative autologous blood donation 20 , but is not optimized for platelet storage. CPDA contains sodium citrate hydrate and dextrose, and the pH is 5.6-5.8. Several studies have reported that citric acid impairs platelet function even in short-time storage as practiced in ANH 12,36 . Hyperglycemia can induce hyperreactivity of platelets to high shear stress and increased P-selectin within 4 h 37,38 . For these reasons, CPDA is not a suitable solution for storage of platelets. The recommended temperature for platelet storage has changed over time. Platelet storage at 4 °C was performed in the 1970s, but 20-24 °C or room temperature is currently recommended based on reports that platelets stored below 20-24 °C rapidly change irreversibly and lose their viability after transfusion 24,32 . Room temperature storage is also recommended in ANH, but the temperature in the operating room during hypothermic CPB often falls below 20 °C, which may impair platelet function. However, some reports have indicated that cold-stored platelets contribute effectively to hemostasis 30,39 and the US Food and Drug Administration recently approved cold-stored platelets in resuscitation of patients with active bleeding 40 . Therefore, we are planning the next study to clarify the appropriate preservative solutions and temperature for ANH, which provide better platelet function and reduce perioperative blood loss.
Red blood cell storage lesions are measured with indicators of hemolysis and the ability to deliver oxygen. There were no significant changes in hemoglobin concentration in all groups in this study. In general, severalhours whole blood storage at room temperature does not cause unacceptable injury to red blood cell functions 41 ; 6-h whole blood storage at 20 °C with CPD decreases 2.3-DPG to 88%, which is acceptable 42 . Low pH suppresses the activity of glycolysis system, which in turn reduces 2,3-DPG required to supply oxygen to peripheral tissues and ATP production required to maintain erythrocyte morphology 43 , but 2,3-DPG and ATP levels recover in 7-72 h after transfusion 43 . High oxygen concentrations may confer oxidative stress, but the appropriate concentration range has not been determined 23,43 . Blood storage in polyolefin bags for several weeks are associated with greater red blood cells hemolysis compared to the storage in di-2-ethylhexil phthalate PVC bags 17 , but storage for several hours in our study had no effect. Mean corpuscular volume, a parameter of the erythrocyte morphology and storage lesions, tends to increase during storage 44,45 , but this also did not increase in our study. Red blood cells can be stored for a longer period than platelets, and the storage conditions such as higher PO 2 for several hours may not affect their function 44,46 .
There are several limitations of this study. First, in order to reduce the burden on the volunteers, the amount of sample blood per bag was set at 20 mL, which is much less than the recommended capacity. Therefore, the surface area, internal pressure and effect of agitation in our storage bags might not be the same as those in clinical use. Storage with small amounts will allow easier agitation and gas exchange. Our results should be supplemented by full volume experiments in the future. Second, in addition to the material and gas permeability, the different surface structures and plasticizers between the two bags used may have influenced the results. Other products may have different effects on blood cells, because the structures of blood bag varies from manufacturer to manufacturer 16 . Third, ANH also improves CPB-induced coagulopathy 47  www.nature.com/scientificreports/ conditions of ANH on coagulation factors were not examined. Coagulation factors, even the most affected factor VIII, are maintained at more than 70%, which is sufficient for hemostasis, in 24-h whole blood storage at room temperature 48 . Global hemostasis assay such as thromboelastography or rotational thromboelastometry should be considered as an additional functional assay. Fourth, as this study was performed in vitro, the hemostatic effect and survival time and hemolysis in vivo of stored platelets could not be evaluated. Within these limitations, our results show that whole blood stored in a polyolefin bag for 6 h maintains significantly higher platelet function compared to the conventional method of storage in a PVC bag, while agitation had no effect on the results.

Methods
The study was approved by the ethics committee of Kyoto University Hospital (R0978-1) and carried out according to the guidelines of the Declaration of Helsinki. All methods were performed in accordance with the institutional guidelines and regulations. Prior written informed consent was obtained from subjects. A study flow chart is shown in Fig. 3.  Blood collection and storage. A sample of 80 mL of venous blood was collected by venipuncture of forearm veins from 6 healthy volunteers who had not taken any medication for at least two weeks before blood sampling. The blood was mixed with 11.2 mL of CPDA (8.14 v/v%; the percentage specified for use of PVC bags currently used in ANH) gently. After removal of 11 mL of blood for testing (sample PRE), the rest of the blood was divided into 20 mL volumes for each of the four groups. The blood was injected into 4 bags and extra air was removed. PVN and PON bags were set at rest, while PVA and POA bags were placed on a horizontal rotatory agitator at 60 rpm, with all 4 bags stored at 22 °C in an incubator box for 6 h. After storage, the blood was agitated gently and tested (samples PVN, PVA, PON, POA).  www.nature.com/scientificreports/ Complete blood count and blood gas analysis. Each sample (PRE, PVN, PVA, PON, POA) was tested promptly after collection using the following methods. Complete blood counts and blood gas parameters were measured using an automated hematology analyzer (Celltac α Nihon Kohden, Tokyo, Japan) and an automated blood gas analyzer (RAPIDPoint 500 or RAPIDLab 1265 Siemens Healthineers, Munich, Germany).

Measurement of ADP-induced platelet aggregation rates.
Platelet-rich plasma (PRP) was prepared by centrifugation of a blood sample at 160g for 10 min at room temperature, followed by collection of the supernatant. The remaining lower portion was further centrifuged at 1600g for 15 min at room temperature and the clear supernatant was used as platelet-poor plasma (PPP). The platelet count was adjusted to 3 × 10 5 /μL by dilution with PPP (adjusted PRP). Aggregation induced by ADP was measured with a light transmission aggregometer (MCM Hema Tracer 212; MC Medical, Tokyo, Japan). Adjusted PRP (3 × 10 5 /μL, 200 μL) was pipetted into a cylindrical cuvette and incubated at 37 °C for 3 min, and then the adjusted PRP was stirred at 37 °C with a magnetic bar at 1000 rpm. A 10-μL volume of 200 μM ADP (final concentration: 9.5 μM) was added to the cylindrical cuvette and ADP-induced platelet aggregation rates was measured for 10 min as a change in light transmission from that of PPP, which was taken to be 100%.
Flow cytometry analysis of P-selectin. Flow cytometry was performed as we have described previously 49,50 . Adjusted PRP was diluted tenfold with phosphate-buffered saline (PBS) (pH 7.42) containing 139 mM NaCl, 8.1 mM NaHPO 4 , 1.5 mM KH 2 PO 4 , and 2.7 mM KCl. Samples were fixed with ice-cold 1% formaldehyde for at least 60 min in a refrigerator and washed twice with ice-cold PBS by centrifugation at 1600g for 15 min at 4 °C. The pellet was suspended in 100 μL PBS at 4 °C. 20 μL of the suspension was coincubated with PerCP-labeled anti-CD61 antibody and PE-labeled anti-CD62P (P-selectin) antibody in a final volume of 100 μL adjusted with PBS for 60 min at room temperature in the dark. PE-labeled IgG was used to estimate nonspecific binding. The reaction was stopped by adding ice-cold PBS. Samples were analyzed using a fluorescence-activated cell sorting instrument (FACSCalibur Becton Dickinson, San Jose, CA, USA). For each sample, data from 10,000 platelets were collected. Platelets were identified by forward and side scatter intensity and by CD61 expression. P-selectin levels on activated platelet surface membranes were recorded as the mean fluorescent intensity (MFI) of PE.
Statistical analysis. All data are expressed as a mean (standard deviation: SD) of 6 experiments. Group variances were tested by a Brown-Forsythe test and were statistically equal. All data were compared by oneway ANOVA, followed by a Dunnett test compared to PVN. All analyses were performed using JMP Pro 15.10 (https:// www. jmp. com/ ja_ jp/ softw are/ predi ctive-analy tics-softw are. html) (SAS Institute Inc., Cary, NC, USA) with P < 0.05 considered significant.