Microplastic contamination of donated blood and red cell concentrates

Microplastics (MPs), defined here as 1 μm to 2 mm in size, have been identified in human tissues including blood.1-3 More than 20 MP polymer types have been found in blood, with a mean particle length of 127 ± 293 μm (range 7 to >2000 μm) and a mean particle width of 58 ± 89 μm (5–800 μm) with polyethylene the most abundant.1, 3 MPs can access the blood via different medical procedures such as intravenous infusions, percutaneous coronary interventions or bypass machines.4-8 The presence of MPs in blood has been associated with biological impacts such as vascular calcification, endothelial dysfunction and elevated thrombotic activity, potentially contributing to cardiovascular disease.9-11 Blood transfusions are given to patients, and it is crucial that they are as therapeutically active as possible and do not cause unintended harm. Whole blood (WB) is donated into plastic blood packs that undergo centrifugation, mechanical pressing and include leucodepletion filters to create the transfusable blood component. Such processes could cause spallation of MPs, elevating MP levels within blood components, potentially not only compromising the efficacy of the transfusion but also causing longer term harm. Therefore, it is critical to understand if blood packs, and the processes that they undergo, introduce MPs into blood components for transfusion. We sampled WB donations taken into top-and-top blood collection packs, where WB is leucodepleted by filtration prior to separation into plasma and red cell concentrate (RCC) only, with no buffy coat being produced. Samples were taken: (1) pre-leucodepletion; (2) after leucodepletion; and (3) from the final RCC to determine whether steps should be taken to reduce MP levels in donated WB or blood components. Ten WB donations, taken from healthy donors in accordance with UK Guidelines, were selected at random in the manufacturing laboratory.12, 13 Blood was collected into standard blood packs made from polyvinylchloride (PVC) plasticised with di(2-ethylhexyl) phthalate (DEHP) with citrate–phosphate–dextrose anticoagulant solution in the primary collection pack and in-line leucodepletion filters (FQE614B, Macopharma, Tourcoing, France) then processed according to standard procedures.12, 13 Samples (range 1.8–10.7 mL) were taken within 24 h of donation by docking PVC-DEHP sample pouches (Macopharma, as before) to the relevant line on the blood pack. Samples were taken from (1) donated WB pre-leucodepletion; (2) leucodepleted blood after filtration; and (3) the final RCC immediately following centrifugation, expression of the plasma and addition of Saline-Adenine-Glucose-Mannitol storage solution. Procedural controls were used to account for possible MP contaminants introduced from the airborne environment and sample pouch packaging. The procedural blank mimicked all sample processing steps. MPs found within procedural blanks represent contamination from indoor atmosphere at the point of sample analysis in the laboratory, contamination from laboratory reagents and equipment or fallout from the air during the transfer of samples from the incubation to filtration stage. All stages of processing and analysis were completed within an ISO class 7 clean room (Connect 2 Cleanrooms, Lancaster, UK) to minimise air fallout into the samples. Blood samples were digested with an enzymatic mix, placed in pre-filtered hydrogen peroxide (100 mL of 30% H2O2) alongside procedural blanks (n = 10) as previously described.3 The H2O2 used was triple filtered using an all-glass vacuum filtration kit and 47 mm glass fibre grade 6 filters (GE Healthcare Life Sciences, Marlborough MA, USA). All glassware underwent thorough manual cleaning, a dishwasher cycle using distilled water and three manual rinses with pre-filtered MilliQ water. All equipment and reagents were always covered with foil lids and a small opening made when pouring. When filtering digested samples, glassware and the sides of the filtration kit were rinsed three times with pre-filtered MilliQ water to avoid sample particle loss. Each sample was processed individually to prevent cross contamination. Plastic equipment was avoided, a cotton laboratory coat and a new set of nitrile gloves were used for each processing step. Flasks were placed in a shaking incubator at 60°C for 7 days, 80 rpm. The digest promotes removal of organic particles while maintaining MP integrity.14 Samples were filtered onto aluminium oxide filters (0.02 μm Anodisc, Watford, UK) using a precleaned glass vacuum filtration system. These were stored in petri dishes before chemical composition analysis alongside procedural blanks using μFTIR microscopy (as described previously).3 A subset of particles was analysed using scanning electron microscopy (SEM) to provide additional morphology and elemental composition as a validation step. GraphPad Prism 10 (Boston, USA) was used to analyse the data using a Friedman test for non-parametric one-way analysis of variance (ANOVA), followed by Dunn's multiple comparison tests. Consistent with previous observations,1-3 we found MP particles in the blood of healthy donors. MP particles were found in all unprocessed WB samples, 60% of leucodepleted samples, and 30% of RCC samples. Figure 1 shows selected images from μFTIR and SEM analysis of the selected MP particles. The identified MP particles matched (>70%) to the library spectra for a total of 21 different polymer types, mainly of fragment shapes, and of median length 64.5 (range 10–2150) μm and width 34.0 (range 5–1150) μm (Figures 1 and 2). The SEM approach validates the μFTIR findings: Images indicate the presence of hard edged, discrete shaped particles, as well as elemental analysis indicating fluorine in PTFE-identified particles. The size of particles is relevant as capillaries are typically 5–8 μm in diameter, presenting a theoretical barrier to any particle larger than this. MPs could potentially block or slow the flow of blood through the capillary bed, impairing oxygen delivery and increasing the opportunity for interaction with the vascular endothelium, with potentially inflammatory effects. Indeed, it has been shown that MPs can be phagocytosed by immune cells, with these cells then causing thrombotic responses that block brain blood vessels leading to neurological changes in mice.15 These data identify potential issues with the presence of MPs and why introducing these MPs into a patient could have additional serious consequences. Donated WB samples had the highest level of MP particles with a median of 4805 (range 1075–10 989) MP/L. Subsequent leucodepletion reduced MPs to 365 (0–10 526) MP/L, and after centrifugation, expression of plasma and the addition of storage solution, only three of the 10 final RCC contained any MP particles: 0 (0–4367) MP/L (Figure 2 and Figure S1). It had been anticipated that the processing of blood through collection into plastic blood packs, centrifugation and passage through leucodepletion filters might lead to an increase in the content of MPs in the final components. In contrast, these results indicate that the standard manufacturing process for leucodepleted RCC leads to a marked clearance of MPs from the final, transfusable component. This would suggest an additional benefit of leucodepletion, in addition to its primary aim of removing white cells and reducing the risk of infection and alloimmunisation as well as decreasing febrile non-haemolytic transfusion reactions. However, eight of the samples were found to contain phthalate-associated chemicals, which did not originate from the donors but could have been introduced from the anticoagulant or storage solutions. Equally, it is important to note that these results might underestimate the level of MPs within the blood. Previous studies have reported similar levels of MPs in blood samples from healthy donors, and the finding that this is significantly reduced in leucodepleted blood components suggests limited impact on patients, who will each have their own baseline level pre-transfusion.1, 3 Future studies should include samples taken directly from the blood donor and from blood components at the end of their shelf-life, to investigate more thoroughly the introduction or removal of MPs during processing and during storage. Method development is also required—removal of fat from the samples is technically difficult and fats can cause interference with the μFTIR analysis due to similar spectral characteristics. The present study examined RCC produced from top-and-top blood packs, where WB is filtered prior to centrifugation. However, blood components can be produced using other manufacturing processes which may result in different levels and types of MPs to be present. A significant amount of blood processing uses bottom-and-top packs which allow the manufacture of plasma, red cells and a buffy coat that is usually used for manufacture of platelet concentrates. These packs include a different leucodepletion filter for filtration of the red cells only, and platelet pooling packs contain yet another different filter, which may have a different impact on the MP levels. It will be important to study MP levels in blood components at the end of their shelf-life (in the UK 35 days for RCC, 7 days for platelets) to see if MPs are shed during storage. Furthermore, the production of blood components by apheresis not only uses different plastic consumables and separation technology; it also returns some blood components to the donor. It is therefore critical that all manufacturing processes for blood components are investigated for the presence of MPs to determine whether these processes may introduce MPs or, conversely, remove MPs from donor blood, to understand any potential risk to the patient. ST, GG, SC and JMR conceived and designed the study. NH, CA, HP, NR, PE and JMR collected the samples and performed the analysis. ST, GG, SC, JMR, NR, PE, NH, CA and HP reviewed the results. ST, GG, SC and JMR wrote the paper. We are grateful to all donors of blood for transfusion, including those whose anonymised donations were included in this study. This work was funded by a Research and Development grant from NHS Blood and Transplant. None of the authors have any conflicts of interest to declare. Data are available on request. Figure S1. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.