Radically reduced ex vivo cell activation by using "in-line" filtered whole blood as a source of platelet concentrate.

Abstract

Dear Sir, Factors affecting the quality of platelet concentrates, obtained from whole blood (WB) are: (i) the processing method used (originating from platelet-rich plasma or buffy coat-derived platelet concentrates [BC-PC]); (ii) the ex vivo manipulation applied (such as cryopreservation, pathogen reduction); (iii) cell storage conditions/duration; and (iv) quantity of residual white blood cells, particularly mononuclear cells1,2. As described previously3, higher mononuclear counts in platelet concentrates are associated with ex vivo cytokine release and cell/cytokine-mediated adverse events. Furthermore, mononuclear cells result in increased platelet surface antigen/marker expression and platelet activation ex vivo, followed by morphological and ultrastructural changes. Commonly, GP140/CD62p expression ≤40% correlates with acceptable clinical efficacy of platelet concentrates. However, discharge of alpha-granules and elevated GP140/CD62p expression (≥50–60%) can lead to decreased platelet recovery and impaired function. A high level of GP53/CD63 expression (due to the release of the contents of lysosomes) is typically detectable only following major platelet activation1,2,4. The aim of this research was to compare the quality of non-filtered platelet concentrates (NF-PC or control group) with “in-line-leucodepleted” platelet concentrates (LD-PC or study group) determined on the basis of residual mononuclear cell count and expression of activation markers (GP/140/CD62p and GP53/CD63), directly after collection and processing of whole blood (day 0), as well as on the 3rd and 5th days of storage of the platelet concentrates (day 3 and day 5). We assumed that the use of “in-line” filtered whole blood (as an optimised source of platelet concentrate) could radically prevent ex vivo platelet activation in LD-PC. For the preparation of NF-PC, whole blood (450±45 mL) from donors, non-reactive for infectious disease markers (hepatitis B and C, AIDS and syphilis) in serological screening tests (Biokit S.A, Llissa de Munt, Spain and Ortho Clinical Diagnostics, Raritan, JS, USA) and with normal clinical data, was collected in quadruple bags (Macopharma, Tourcoing, France). This bag system contains 63 mL citrate phosphate dextrose (CPD) and 100 mL saline-adenine-glucose-mannitol (SAGM) solutions. For the LD-PC, the same volume of whole blood from donors with normal clinical and laboratory findings was collected into Imuflex WB-SP bags (Terumo, Tokio, Japan) with CPD, SAGM, and an “in-line” filter that “saves” (not retains platelets)5. NF-PC were obtained from whole blood using a “standard buffy-coat” method. Non-filtered whole blood (total=180 units) was initially centrifuged with a “hard-spin” technique (3,890×g for 10 minutes, at 20±2 °C) in a Cryofuge 8500 (Heraeus, Langenselbold, Germany) and then processed using a T-ACE device (Terumo, Japan). NF-BC units were obtained from the separated buffy coat units were obtained following “light-spin” centrifugation (310×g for 7 minutes, at 20±2 °C) and second (manual) processing, as described previously2,3. Six NF-PC were pooled using a sterile connection system (TSCD, Terumo, Tokio, Japan) and stored on a shaker (Teknolabo Instruments Srl, Milan, Italy) at 22±2 °C for 5 days. In the same manner “in-line” filtered whole blood (total=180 units) underwent “hard-spin” centrifugation. The leucocyte-poor buffy-coat (platelet rich, but leucoreduced cell layer) was separated on a T-ACE device. The second (“light-spin”) centrifugation of these “leucocyte-poor buffy coats” was performed as described above. LD-PC were obtained by processing; the pooling and storage procedures for LD-PC were completed in the same manner as for the NF-PC. Standard haematological parameters were analysed on a Beckman-ACT device (Beckman-Coulter, Fullerton, CA, USA). The residual white blood cell count in LD-PC units was quantified manually in a Negeotte haemocytometer chamber, as described previously3. The platelet activation markers (GP140/CD62p and GP53/ CD63) were investigated using monoclonal antibodies (Becton Dickinson, San Jose, CA, USA) and flow cytometry. Briefly, samples of platelet concentrates (1×106 cells per tube) were incubated with fluorescein isothiocyanate-labelled monoclonal antibodies, in accordance with the manufacturer’s recommendations and instructions. Non-specific binding was detected using control cells, which were incubated with phosphate-buffered saline alone. After incubation, cells were washed and analysed on an Epics XL (Beckman-Coulter). The results were calculated and expressed as percentage of total cell number4. Differences between groups were analysed by Student’s t-test and were considered as statistically significant at p≤0.05. Our preclinical study confirms that the mononuclear cell count was lower in LD-PC (0.05±0.02×106/unit) than in NF-PC (5.32±1.7×106/unit) (t=6.693; p<0.001) (Table I). Table I The expression of platelet activation markers in NF-PC and LD-PC during 5 days of storage. The levels of GP140/CD62p and GP53/CD63 expression in both NF-PC and LD-PC over the storage period are shown in Table I. GP140/CD62p expression (calculated as a percentage) on day 0 was significantly higher in NF-PC than in LD-PC (t=10.642; p<0.001). Significant differences in GP140/CD62p expression between NF-PC and LD-PC (“intergroup” disparity) were also observed on day 3 (t=12.754; p<0.001) and day 5 (t=10.724; p<0.001). A marked linear increase of GP140/CD62p expression was confirmed in both investigated groups during the 5 days of storage (p<0.001). However, GP53/CD63 expression was significantly higher at each measurement (day 0, day 3, and day 5) in NF-PC than in LD-PC (p<0.001). Finally, GP53/CD63 expression increased significantly during the 5 days of storage in both NF-PC and LD-PC (p<0.001). As stated, the aim of this study was to determine the effect of leucodepletion on the expression of platelet activation markers in platelet concentrates. The results obtained showed that mononuclear cell count was significantly lower in LD-PC than in NF-PC (prepared by the “standard buffy-coat” method). This is consistent with research conducted by Paunovic et al., who showed that the “in-line” WB-SP filter (which “saves” platelets) achieves optimised, practical leucodepletion of blood products-packed (resuspended) red blood cells and platelet concentrates5. Our earlier research showed better quantitative and qualitative platelet recovery -that is, superior hypotonic shock response, aggregation ability, morphological score of platelets, ultrastructural properties (intact microtubules, minimal membrane changes, etc.)- in platelet concentrates treated by multiple, but optimised ex vivo manipulations2–4. The quality of platelet concentrates for transfusion is optimal when no activation markers are expressed, since activated platelets have inferior quantitative and qualitative recovery, and could promote prothrombotic and procoagulant events. Generally, platelet activation is accompanied by increased expression of GP140/CD62p and GP53/CD63 antigens on the platelet membrane. The expression of these antigens rises progressively during storage, and correlates with the degree of cell activation in platelet concentrates. The structural and adhesive glycoprotein GP140/CD62p (or P-selectin) is an integral constituent of the membrane of alpha-granules of “resting” platelets and Weibel-Palade bodies of endothelial cells. Its expression on the platelet surface upon alpha-granule exocytosis has been used as an ex vivo marker of platelet secretion and activation. In fact, GP140/CD62p expression is a marker of the platelet storage-lesion. Thus, GP140/CD62p expression is extensively used to evaluate platelet concentrate quality following whole blood processing, filtering, liquid-state storage or cryopreservation. Nevertheless, despite numerous investigations, the clinical relevance of GP140/CD62p expression levels or their value as an ex vivo measure of in vivo platelet activity or viability remains uncertain1,2,4. GP53/CD63 was the first identified marker of platelet activation that is elevated on the cell surface following releasing from granules2,4. Our recent study established that GP53/CD63 expression was really lower in LD-PC over the entire storage period. The expression of GP53/CD63 was also four times higher in NF-PC on day 5 than on day 0, which is in expected correlation. In summary, the use of optimised (with regards to timing) and effective (with regards to degree) “in-line” leucodepletion (Imuflex WB-SP system) undoubtedly reduces white blood cell and mononuclear cell counts. In the LD-PC investigated, GP140/CD62p and GP53/CD63 expression was less during the 5 days of storage, a finding strongly correlated with lower platelet activation. In accordance with our earlier research, we speculate that the use of LD-PC, with improved ex vivo cell quality, could have a better clinical effect.