Introduction: Bacterial contamination of blood products presumably occurs mainly during blood collection, starting from low initial concentrations of 10–100 colony-forming units (CFUs) per bag. As little is known about bacterial growth behavior and distribution in stored whole blood (WB) and WB-derived blood products, this study aims to provide data on this subject. Methods: WB units were inoculated with transfusion-relevant bacterial species (Acinetobacter baumannii, Bacillus cereus, Escherichia coli, Klebsiella pneumoniae, Listeria monocytogenes, Pseudomonas fluorescens, Serratia marcescens, Staphylococcus aureus, Staphylococcus epidermidis, Streptococcus dysgalactiae, Streptococcus pyogenes, Yersinia enterocolitica; n = 12 for each species), stored for 22–24 h at room temperature, and then centrifuged for separation into plasma, red blood cells (RBCs), and buffy coats (BCs). The latter were pooled with 3 random donor BCs and one unit of PAS-E each to yield plasma-reduced platelet concentrates (PCs). Samples for bacterial colony counting were collected after WB storage and immediately after blood component production. Sterility testing in PCs (n = 12 for each species) was performed by bacterial culture after 7 days of storage. Results: Bacterial growth in WB varied remarkably between donations and species. Streptococcus species produced the highest titers in WB, whereas Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli, and Pseudomonas fluorescens did not multiply. Centrifugation resulted in preferential accumulation of bacteria in BCs, with titers of up to 3.5 × 103 CFU/mL in BCs and up to ≤0.9 × 103 CFU/mL in BC-derived PCs. Overall, 72/144 PCs (50%) tested positive for bacteria after storage. Sterility test results were species-dependent, ranging from 12 of 12 PCs tested positive for Streptococcus pyogenes to 1 of 12 PCs positive for Escherichia coli. Bacterial contamination of RBC and plasma units was much less common and was associated with higher initial bacterial counts in the parent WB units. Conclusions: Bacterial growth in WB is species-dependent and varies greatly between donations. Preferential accumulation of bacteria in BCs during manufacturing is a critical determinant of the contamination risk of BC-derived pooled PCs.

While rare, transfusion-transmitted bacterial infections (TTBIs) are regularly reported to national hemovigilance authorities [1, 2]. Although improvements in donor screening, testing and donation practices have significantly reduced the risk of TTBI, bacterial contamination of blood products still contributes to the most frequently reported serious adverse events. Bacterial contamination is presumed to occur mainly during the donation process. The initial number of bacterial contaminants, though too small to measure, is estimated to be around 10–100 colony-forming units (CFUs) per bag [3]. Due to the rarity of TTBIs and detection limit issues, little is known about the growth behavior and distribution of bacteria in stored whole-blood (WB) units and WB-derived blood products. Spiking experiments are needed to generate data on bacterial contamination of blood products. Most previous studies of bacterial growth in WB units are limited to single bacterial strains [4‒6] or high spiking concentrations of 102 CFU/mL or higher [7‒10], neither of which reflects the true strain diversity and low initial numbers of bacteria present at the time of blood donation.

Several factors influence the growth of bacteria in blood and blood components. Blood has antimicrobial properties because it contains different immune system components, such as WB cells (WBCs), platelets (PLTs), complement components, and antimicrobial peptides [10, 11]. Consequently, the proliferation of bacterial contaminants in WB may be not only strain-dependent but also donor-dependent [8]. Storage conditions such as storage time also influence bacterial growth. The risk for bacterial contamination is highest for platelet concentrates (PCs), the blood products with the most favorable storage conditions for bacterial growth.

Various mitigation strategies are used to minimize the risk of bacterial contamination, including diversion of the first millimeters of collected blood, bacterial testing, and pathogen inactivation (PI). Skin and donation site preparation/disinfection also plays a role in bacterial contamination of donated blood, as does technique of the phlebotomist.

PI systems are available for plasma and PLTs but not for red blood cells (RBCs) and WB. The latter systems are under development and will hopefully become available soon [12‒14]. Knowledge about the growth and distribution of bacteria in WB and WB-derived blood components is important for risk assessments, bacterial selection for validation experiments, and adequate inactivation study design. Two different panels of transfusion-relevant bacteria reference strains (TRBRS) have been tested in international studies, shown to reproducibly grow in PCs or RBC units, and recommended by the World Health Organization (WHO) for use in studies of the microbial safety of blood products [15‒17]. The aim of the present study was to determine the growth and distribution characteristics of twelve transfusion-relevant bacterial strains in WB units and WB-derived blood components with a special focus on PCs.

Blood Donors

Blood donors were selected based on our local donor selection standards. Only regular blood donors who met the requirements for blood donation and gave their informed consent approved by the Local Ethics Committee were included in the study.

WB Processing

WB (450–500 mL) was collected in bags containing 70 mL citrate phosphate dextrose anticoagulant solution (day 0) and stored overnight for 14–19 h at room temperature (22 ± 2°C) until inoculation and further processing. Using standard blood bank equipment under conditions equivalent to routine production, each WB unit was separated into components by centrifugation (4,000 g for 10 min at room temperature) using a Cryofuge 6,000i centrifuge (Heraeus). RBCs, buffy coat (BC), and plasma were prepared using a MacoSmart bag press (Macopharma). RBCs were manufactured with 110 mL of PAGGS-M additive solution and leukodepleted by filtration.

Plasma-reduced PCs were prepared from pools of four BCs according to the method of Eriksson and colleagues [18]. The pools were mixed with 280 mL of SSP+ PLT storage solution (MacoPharma), which is identical to PAS-E [19]. After soft spin centrifugation, PCs were isolated from suspension, leukoreduced by inline filtration, and transferred to standard storage bags (Haemonetics).

Spiking and Sampling

The study design is shown in Figure 1. We investigated the impact of WB processing on bacterial distribution and growth in blood components with a special focus on PCs. In PCs, we studied the contamination rate and bacterial load in fresh and stored products as well as the bacterial load at a time point when PI treatment may be performed. Each WB was tested for sterility before spiking using the BACT/ALERT system (bioMérieux, Nuertingen, Germany). In total, twelve different bacteria strains were used (Table 1). On the day after donation, each strain was inoculated in twelve WB units (inoculum size: approximately 10 CFU/bag). WB units were stored for additional 22–24 h at room temperature and then separated into blood components as described above. Study donor BCs were pooled with 3 random donor BCs and one unit of PAS-E to yield plasma-reduced PCs. Samples for bacterial testing were taken after 22–24 h of storage at room temperature (WB), immediately after preparation (RBCs, BCs, PCs, and plasma), and after 4–6 h of storage at 22°C (PCs), respectively, mimicking the time period after preparation that may be required to carry out PI treatment of PCs. After 7 days of storage, PC was sampled for sterility testing using the BACT/ALERT system (BioMérieux) and for determination of the bacterial concentration by plating.

Fig. 1.

Design of spiking experiments using 12 different bacterial species to spike 12 WB units per species.

Fig. 1.

Design of spiking experiments using 12 different bacterial species to spike 12 WB units per species.

Close modal
Table 1.

List and origin of bacterial species used in the study

BacteriaGramStrainReference
Acinetobacter baumannii Negative ATCC-17961 Isolated from blood 
Bacillus cereus Positive PEI-B-P-57 WHO PLT 2 
Escherichia coli Negative PEI-B-P-19 WHO PLT 1 
Klebsiella pneumoniae Negative PEI-B-P-08 WHO PLT 1 
Listeria monocytogenes Positive PEI-A-199 WHO RBC 
Pseudomonas fluorescens Negative PEI-B-P-77 WHO PLT 2, WHO RBC 
Serratia marcescens Negative PEI-B-P-56 WHO PLT 2 
Staphylococcus aureus Positive PEI-B-P-63 WHO PLT 2 
Staphylococcus epidermidis Positive PEI-B-P-06 WHO PLT 1 
Streptococcus dysgalactiae Positive PEI-B-P-71 WHO PLT 2 
Streptococcus pyogenes Positive PEI-B-P-20 WHO PLT 1 
Yersinia enterocolitica Negative PEI-A-105 WHO RBC 
BacteriaGramStrainReference
Acinetobacter baumannii Negative ATCC-17961 Isolated from blood 
Bacillus cereus Positive PEI-B-P-57 WHO PLT 2 
Escherichia coli Negative PEI-B-P-19 WHO PLT 1 
Klebsiella pneumoniae Negative PEI-B-P-08 WHO PLT 1 
Listeria monocytogenes Positive PEI-A-199 WHO RBC 
Pseudomonas fluorescens Negative PEI-B-P-77 WHO PLT 2, WHO RBC 
Serratia marcescens Negative PEI-B-P-56 WHO PLT 2 
Staphylococcus aureus Positive PEI-B-P-63 WHO PLT 2 
Staphylococcus epidermidis Positive PEI-B-P-06 WHO PLT 1 
Streptococcus dysgalactiae Positive PEI-B-P-71 WHO PLT 2 
Streptococcus pyogenes Positive PEI-B-P-20 WHO PLT 1 
Yersinia enterocolitica Negative PEI-A-105 WHO RBC 

Bacteria belong to the First WHO International Repository of Platelet Strains (WHO PLT 1) [15], to the enlargement panel (WHO PLT 2) [16] or to the RBC reference strains (WHO RBC) [17].

Bacterial Strains

The bacteria used in this study are gram-negative and gram-positive species known to contaminate blood products. Most belong to the WHO International Repository of Platelet Transfusion-Relevant Reference Strains for PLTs and RBCs and were obtained from the Paul-Ehrlich-Institute [PEI] in Langen, Germany. Acinetobacter baumannii was obtained from the ATCC (American Type Culture Collection) (LGC Standards, Wesel, Germany) (Table 1). Each WHO reference strain has been shown to grow reproducibly in PCs or RBCs after minimal inoculation (10–100 CFU/unit) in multiple international laboratories [15‒17].

Bacteria were cultured in tryptic soy broth and aliquots of 1 mL were frozen after the addition of 50% human serum albumin. The suspensions were stored at temperatures below −70°C, and the bacterial concentration of the stock was determined by plating assay.

Aliquots thawed at 37°C without further cultivation were used for spiking. The inoculum was prepared by serial dilutions of the stock in sodium chloride solution. Dilution factor and volume were calculated based on the known concentration of the stock. The dilution factor chosen resulted in an inoculum volume in the range of 100–1,000 μL. The spiking concentration was confirmed by plating the spiked volume of the inoculum.

Enumeration of Bacterial Cells and Automated Blood Culture

Bacterial concentrations were determined using standard plating methods and expressed as the number of viable CFUs per milliliter (CFU/mL). Streptococcus dysgalactiae was cultured on Columbia blood agar (Merck, Darmstadt, Germany). All other bacterial species were plated on CASO agar (Merck, Darmstadt, Germany). Samples with an expected count of >100 CFU/mL were inoculated using a spiral plater. Serial ten-fold dilutions of each sample were transferred to agar plates (2 plates/dilution, 50 μL/plate). Samples with an expected count of <100 CFU/mL were cultured using the pour plate method (2 plates/sample, 0.5–1 mL/plate). All agar plates were incubated at 37°C (except for Serratia marcescens, which was incubated at 30°C) for 1–2 days until colonies could be counted.

PC samples collected on day 7 of storage were tested for sterility with the BACT/ALERT 3D 60 Microbial Detection System (bioMérieux, Nuertingen, Germany). Briefly, after the collection of PC samples (20 mL each) under sterile conditions on day 7 of storage, half (10 mL) was injected into a BACT/ALERT SA (aerobic) and SN (anaerobic) culture bottle, respectively, and cultivated under automated blood culture conditions at 36°C in the BACT/ALERT incubator (bioMérieux) for 7 days.

WB was spiked with 3–15 CFU per bag (corresponding to approximately 0.006–0.03 CFU/mL) and stored for 22–24 h before sampling (Table 2). Bacterial growth in WB varied remarkably between donations and bacterial species. Streptococcus pyogenes and Serratia marcescens exhibited the most rapid growth in WB, whereas Staphylococcus aureus, Escherichia coli, and Pseudomonas fluorescens did not multiply but remained detectable in some spiked units. In WB units spiked with Staphylococcus epidermidis, Yersinia enterocolitica, or Acinetobacter baumannii no bacteria were detected after storage prior to component preparation.

Table 2.

Bacterial growth and distribution in blood products derived from spiked WB (n = 12 per species)

SpeciesSpike, CFU/bagTiter, CFU/mL, mean, min–maxPC BacT/Alert day 7a
WB 22–24 hRBCPlasmaBCPC after separationPC 4–6 hPC day 7*
Klebsiella pneumoniae 9–15 13.3 0.0 0.2 34.6 4.6 77.8 1.5 × 109 8/12 pos 
0–70 0–1.5 0–278 0–43 0–823 0.8–1.8 × 109 
Streptococcus pyogenes 7–9 252.9 48.1 162.4 1,085 225 3.9 × 103 1.3 × 106 12/12 pos 
25–539 0–202 0–695 163–3,506 18–906 2.5 × 102–2.6 × 104 0.2–3.6 × 106 
Staphylococcus aureus 6–7.5 3.5 0.1 0.0 0.9 0.0 0.0 8.6 × 107 6/12 pos 
0–20 0–1 0–4 3.6 × 106 – 1.6 × 108 
Streptococcus dysgalactiae 3–9 6.3 0.0 0.0 32.7 0.3 2.0 1.0 × 108 11/12 pos 
0–31 0–0.5 0–161 0–1 0–7 0.8–1.5 × 108 
Escherichia coli 8–14 2.2 0.0 0.2 0.6 0.0 0.0 3.8 × 108 n.a 1/12 pos 
0–21 0–2.5 0–6 
Staphylococcus epidermidis 7–10 0.0 0.0 0.0 0.0 0.0 0.0 3.7 × 107 n.a 1/12 pos 
Pseudomonas fluorescens 3–8 10.8 1.5 0.0 71.7 2.8 79.7 1.6 × 109 12/12 pos 
0–42 0–13 2–284 0–11 0–396 0.4–3.3 × 109 
Yersinia enterocolitica 8–12 0.1 0.1 0.1 0.0 0.0 0.0 n.a 0/12 pos 
0–2 0–1 
Serratia marcescens 15 40.1 0.2 3.2 95.0 2.3 63.1 0.9 × 1010 9/12 pos 
0–360 0–2 0–38 0–864 0–10 0–2.5 × 102 0.2–3.0 × 1010 
Acinetobacter baumannii 7–11 0.0 0.0 0.0 0.4 0.1 0.0 Titer could not be determinedb 2/12 pos 
0–2.5 0–1 
Listeria monocytogenes 9–15 0.0 0.0 0.0 0.0 0.0 0.0 1.3 × 107 4/12 pos 
0–0.5 0–1 0.1–2.8 × 107 
Bacillus cereus 2–8 23.2 0.0 0.3 53.4 0.4 23.2 4.1 × 106 6/12 pos 
0–124 0–1 0–392 0–3 0–44 2.7–7.1 × 106 
SpeciesSpike, CFU/bagTiter, CFU/mL, mean, min–maxPC BacT/Alert day 7a
WB 22–24 hRBCPlasmaBCPC after separationPC 4–6 hPC day 7*
Klebsiella pneumoniae 9–15 13.3 0.0 0.2 34.6 4.6 77.8 1.5 × 109 8/12 pos 
0–70 0–1.5 0–278 0–43 0–823 0.8–1.8 × 109 
Streptococcus pyogenes 7–9 252.9 48.1 162.4 1,085 225 3.9 × 103 1.3 × 106 12/12 pos 
25–539 0–202 0–695 163–3,506 18–906 2.5 × 102–2.6 × 104 0.2–3.6 × 106 
Staphylococcus aureus 6–7.5 3.5 0.1 0.0 0.9 0.0 0.0 8.6 × 107 6/12 pos 
0–20 0–1 0–4 3.6 × 106 – 1.6 × 108 
Streptococcus dysgalactiae 3–9 6.3 0.0 0.0 32.7 0.3 2.0 1.0 × 108 11/12 pos 
0–31 0–0.5 0–161 0–1 0–7 0.8–1.5 × 108 
Escherichia coli 8–14 2.2 0.0 0.2 0.6 0.0 0.0 3.8 × 108 n.a 1/12 pos 
0–21 0–2.5 0–6 
Staphylococcus epidermidis 7–10 0.0 0.0 0.0 0.0 0.0 0.0 3.7 × 107 n.a 1/12 pos 
Pseudomonas fluorescens 3–8 10.8 1.5 0.0 71.7 2.8 79.7 1.6 × 109 12/12 pos 
0–42 0–13 2–284 0–11 0–396 0.4–3.3 × 109 
Yersinia enterocolitica 8–12 0.1 0.1 0.1 0.0 0.0 0.0 n.a 0/12 pos 
0–2 0–1 
Serratia marcescens 15 40.1 0.2 3.2 95.0 2.3 63.1 0.9 × 1010 9/12 pos 
0–360 0–2 0–38 0–864 0–10 0–2.5 × 102 0.2–3.0 × 1010 
Acinetobacter baumannii 7–11 0.0 0.0 0.0 0.4 0.1 0.0 Titer could not be determinedb 2/12 pos 
0–2.5 0–1 
Listeria monocytogenes 9–15 0.0 0.0 0.0 0.0 0.0 0.0 1.3 × 107 4/12 pos 
0–0.5 0–1 0.1–2.8 × 107 
Bacillus cereus 2–8 23.2 0.0 0.3 53.4 0.4 23.2 4.1 × 106 6/12 pos 
0–124 0–1 0–392 0–3 0–44 2.7–7.1 × 106 

*Titers determined in PCs tested positive in BACT/ALERT.

apos = Samples of PCs tested positive in BACT/ALERT 3D 60 Microbial Detection System (Biomerieux).

bIn single PCs, the titer could not be determined accurately because the dilution factors were inappropriate and the number of colonies was too low or too high to be counted.

Bacteria preferentially accumulated in BCs during separation, reaching titers of up to 3.5 × 103 CFU/mL in BCs and up to 0.9 × 103 CFU/mL in BC-derived PCs immediately after preparation. Bacterial contamination of RBC and plasma units was much less frequently detected and was associated with higher initial bacterial counts in the parent WB units. Maximum bacterial counts were 202 CFU/mL for RBCs and 695 CFU/mL for plasma.

Bacterial counts in PCs directly after preparation were lower than those in the contaminated BCs. Count reduction is probably due to a combination of different effects like leukodepletion by filtration and the dilution effect obtained by BC pooling and addition of storage solution. However, bacteria started to proliferate during storage for 4–6 h at room temperature. The highest bacterial counts of up to 2.6 × 104 CFU/mL were obtained in PCs for Streptococcus pyogenes. Titers above 1 × 102 CFU/mL were detected for Klebsiella pneumoniae (up to 8.2 × 102 CFU/mL), Pseudomonas fluorescens (up to 4 × 102 CFU/mL), and Serratia marcescens (up to 2.5 × 102 CFU/mL), while the titers in PCs spiked with the other bacteria species were lower or zero at this point of time.

Overall, 72 out of 144 PCs tested positive for bacteria after 7 days of storage. Bacterial counts on day 7 in PCs that tested positive in BACT/ALERT were variable and ranged from 106 – 1010 CFU/mL (Table 2). PC contamination rates varied depending on the type of species used for WB spiking: 12/12 PCs tested positive for Streptococcus pyogenes and Pseudomonas fluorescens, while only 1/12 tested positive for Escherichia coli and Staphylococcus epidermidis, and none tested positive for Yersinia enterocolitica.

This systematic study of bacterial growth behavior and distribution in WB and WB-derived blood components demonstrated that very low-level inoculation with various transfusion-relevant bacterial species can cause bacterial contamination in all three blood products studied. Bacterial growth differed between bacterial species and blood donors. Contaminating bacteria preferentially accumulated in BCs during WB centrifugation and were transferred to the PC during further preparation, while accumulation of bacteria into RBC and plasma units was associated with higher initial bacterial counts in the parent WB units. Even if risk reduction measures such as skin disinfection and diversion of the first 10–50 mL of donated blood are taken, transient or commensal bacteria on the donor’s skin may enter the blood collection bag via venipuncture during the WB donation process. During storage, bacteria in blood components may grow to levels capable of inducing acute sepsis and life-threatening or fatal reactions in the recipient [20]. Bacterial contamination can be missed in early testing due to low bacterial concentrations. By interpolating from counts determined at a later time point of storage, Benjamin and Wagner estimated the number of contaminating bacteria to be less than 62 CFU/unit 12–36 h after donation [21]. In the current study, we mimicked routine transfusion conditions by contaminating WB units with low levels of transfusion-relevant bacteria thought to be capable of growing in blood products like PCs and RBCs [15‒17]. Our findings show that these bacteria were also able to survive and grow in the WB units, suggesting that they may also be suited as reference strains for WB spiking studies.

Several bacterial spiking studies in WB have been performed in the past to shed light on different aspects of transfusion safety. Some investigated the effects of leukocyte depletion [22] or different storage conditions [4, 7] on bacterial growth. Leukocyte depletion by filtration was shown to significantly reduce bacterial counts by partially removing bacterial contaminants [8, 22‒24]. Prolonged storage of WB in the presence of donor WBCs was shown to delay the growth of some bacteria [4]. Taha et al. [9] demonstrated that overnight holding of bacterially inoculated WB eliminated some bacterial strains, while others were able to survive and proliferate. Similar results were achieved by Siblini et al. [24] who also investigated the bactericidal properties of WB. Our results are consistent with data from other studies demonstrating differences in the growth of different bacterial strains or even isolates from the same species [22]. The donor-dependent differences in bacterial growth observed in our experiments have also been described previously and may be de due to differences in the presence and activity of immune factors between donors.

In our study, we observed no or limited growth of Yersinia enterocolitica and Staphylococcus epidermidis, respectively, in WB. This is in accordance with the findings of Siblini and Taha who both showed a reduced viability of Yersinia enterocolitica in WB during room temperature storage [9, 25]. We cannot conclude from our investigations whether bactericidal or other factors were responsible for the limited ability to grow. However, it was reported for Staphylococcus epidermidis that the clearance of bacteria from WB is dependent on the inoculum size [11]. While small inocula of Staphylococcus epidermidis were cleared rapidly from WB, it took up to 24 h of storage in WB to eliminate larger inocula. This might explain why we could not detect growth of Staphylococcus epidermidis in our study using low titer spiking concentrations, whereas others reported growth of this species. The fact that one of the PCs spiked with Staphylococcus epidermidis tested positive at the end of storage, although bacteria could not be detected in the underlying WB unit, is most probably based on a sampling error due to a low bacterial count in the WB. However, accidental contamination during inoculation of culture bottles cannot be excluded in this case, as identification methods were not applied.

Consistent with our findings, Taha et al. [9] and Mohr et al. [8] also described the segregation of bacteria into specific blood components. Because the density ranges of many bacteria overlap with those of RBCs and PLTs, centrifugation of WB results in the preferential accumulation of bacteria in BCs under routine blood banking conditions [25]. Taha et al. [9] investigated the distribution of bacteria through BC-based PC production and subsequent PC storage after spiking and overnight storage of WB. In this study, bacteria segregated preferably toward the cellular fractions (RBCs and BC) compared to plasma. Mohr et al. looked at the bacteria concentration in WB after storage for 8 h at 22°C and at different process steps during the production and storage of pooled PCs. While some of the investigated strains did not grow or were even killed during storage, the concentrations of bacteria in WB after 8 h and in the freshly prepared BC were in a similar order of magnitude. The main differences between these studies and our investigation are the higher spiking concentration (102–104 CFU/mL) used and the correspondingly higher titers achieved after WB storage. However, several other factors including the centrifugation time and speed, the anticoagulant chemistry, the bacterial load, and the geometry of the blood container also influence the separation of blood constituents and contaminants [26]. The less rigid centrifugation conditions usually used to recover PLTs from pooled BCs not only support the separation of the PLTs from RBCs but also seem to cause bacteria to accumulate with the PLTs in the supernatant used for BC-derived PC preparation. However, clinical data and our spiking experiments show that this separation effect is not perfect. Hemovigilance reports indicate that TTBI rates are highest for PCs and that TTBIs are less frequently caused by RBC and plasma transfusions [2, 27]. In the present study, bacterial contamination was also detected in RBCs and plasma and was associated with bacterial strains able to grow rapidly in WB. Clinically relevant concentrations of the fast-growing strain Streptococcus pyogenes were detected in all blood components examined here.

Bacterial growth in blood components is mainly determined by storage conditions. While RBCs are stored at 1–6°C and plasma is usually kept frozen during storage, PCs are stored at 20–24°C under continuous agitation for gas exchange to preserve function and survival. These storage conditions make PCs an excellent growth medium for many bacteria. Rapid-growing strains, such as S. aureus, E. coli, and S. pyogenes can reach clinically relevant counts in PCs within a few hours of storage [28]. Nevertheless, psychrophilic bacteria like Y. enterocolitica or S. marcescens [17, 29] can grow in RBCs at 4°C or in refrigerated plasma, stored for several days to ensure rapid availability for emergency purposes [30]. The high rates of PCs testing positive for bacteria after 7 days of storage in the present study confirm that PCs are particularly susceptible to bacterial contamination. With our experimental approach, we tried to define factors that influence the bacterial contamination of blood components during WB manufacturing. Our findings and those from other studies suggest that the distribution of bacteria in WB critically determines the contamination rate of blood components [4].

Different strategies to control bacterial risk exist, and most focus special attention on PCs. Bacterial testing poses significant logistical challenges, so many countries have decided to implement PI for PLTs. In the future, universal PI may be introduced as soon as PI technologies for RBCs or WB are approved and become commercially available. Production processes must be adapted for the implementation of PI systems, and time frames for PI procedures must be defined to ensure that processing does not exceed the inactivation capacity of the system. Whereas viruses or parasites do not replicate in blood products, bacteria are able to grow to high titers [31]. Inactivation of bacteria is a challenging task, and incomplete inactivation allows residual bacteria to recover and grow again. Several studies of different PI systems have revealed that fast-growing bacterial strains are not completely inactivated if the bacterial count at the time of inactivation exceeds the inactivation capacity of the PI system [3, 32‒34]. Our study provides information on relevant bacterial strains and conditions that can be used to challenge and validate PI systems. In addition, the results of this study on the occurrence of bacterial contamination in blood components in the early post-preparation period may be helpful for planning infection risk assessments and defining process parameters to optimize the efficacy of PI processes. Our results show that bacterial contamination may reach titers of up to 0.5 × 103 CFU/mL in WB after 24 h of storage, even though initial contamination was very low. Therefore, a bacteria inactivation capacity of more than three log10 reduction steps may be required for a PI system for WB to ensure sterility of the WB and the corresponding blood products.

There are some limitations to this study. First, WB units were spiked on the day after donation and not immediately for logistic reasons. Some immunologic factors like the complement system [35] and WBC activity may have been diminished during storage. Second, the design of our study does not cover the scenario of secondary contamination with environmental bacteria recently reported in the USA for PCs contaminated with Acinetobacter baumannii by storage container defects [36]. Third, production parameters like storage time and temperature, centrifugation conditions, or the timing of leukocyte depletion may have had an additional impact on bacterial distribution and growth [4, 8, 11, 22]. Thus, the bacterial concentrations observed in this study may differ slightly from those detected under routine conditions. Methods of manufacturing blood components can differ between production sites. Automated processes like apheresis were not investigated in this study. Therefore, the present study is only representative of bacterial growth and distribution under the manufacturing conditions used at our blood center. However, these production processes are similar to those used at blood services in many other European countries. Finally, differences in the growth of different bacterial strains or even isolates from the same species were shown in several studies [10, 11, 22, 37]. The results of our study therefore are limited to the strains and species described herein and cannot be used to extrapolate the growth of other isolates.

In conclusion, TTBIs still pose a serious risk for transfusion recipients. Although there are several risk mitigation strategies, like bacterial testing or PI, none of them completely eliminates the risk for bacterial contamination. Reasons for this include sampling errors in bacterial testing [38‒40] or secondary contamination during storage [41, 42]. However, PI technologies have been shown to significantly improve the bacterial safety of blood transfusion, particularly in countries using 100% pathogen-reduced PCs [43]. The present study adds to the knowledge about risk factors for bacterial contamination of blood components and provides information on critical factors to consider when validating and implementing bacterial risk reduction strategies.

We would like to thank Karin Oller, Jana Pravemann, and Birthe Nordmann for performing the bacterial growth experiments.

Samples used in this study were obtained as part of routine blood donation. Written informed consent was obtained from participants prior to the study. Use of these samples for research purposes did not require ethical approval in accordance with local/national guidelines.

The authors have no conflicts of interest to declare.

Ute Gravemann, Wiebke Handke, Torsten Schulze, and Axel Seltsam received grants from the Research Foundation of the German Red Cross Blood Services (Deutsche Forschungsgemeinschaft der Blutspendedienste des Deutschen Roten Kreuzes) and Maco Pharma for development of a UVC-based PI technology for PLTs.

Ute Gravemann, Wiebke Handke, and Axel Seltsam designed the study. Ute Gravemann, Torsten Schulze, and Axel Seltsam analyzed the data and wrote the manuscript. All authors read and approved the final manuscript.

All data generated or analyzed during this study are included in this article. Further inquiries can be directed to the corresponding author.

1.
Lafeuillade
B
,
Eb
F
,
Ounnoughene
N
,
Petermann
R
,
Daurat
G
,
Huyghe
G
, et al
.
Residual risk and retrospective analysis of transfusion-transmitted bacterial infection reported by the French National Hemovigilance Network from 2000 to 2008
.
Transfusion
.
2015
;
55
(
3
):
636
46
. .
2.
Orru
S
,
Oberle
D
,
Heiden
M
,
Muller
S
,
Krut
O
,
Funk
MB
.
Analysis of transfusion-transmitted bacterial infections according to German hemovigilance data (2011-2020)
.
Transfus Med Hemother
.
2023
;
50
(
2
):
144
53
. .
3.
Gravemann
U
,
Handke
W
,
Muller
TH
,
Seltsam
A
.
Bacterial inactivation of platelet concentrates with the THERAFLEX UV-Platelets pathogen inactivation system
.
Transfusion
.
2019
;
59
(
4
):
1324
32
. .
4.
Sanz
C
,
Pereira
A
,
Vila
J
,
Faundez
AI
,
Gomez
J
,
Ordinas
A
.
Growth of bacteria in platelet concentrates obtained from whole blood stored for 16 hours at 22 degrees C before component preparation
.
Transfusion
.
1997
;
37
(
3
):
251
4
. .
5.
Braathen
H
,
Sivertsen
J
,
Lunde
THF
,
Strandenes
G
,
Lindemann
PC
,
Assmus
J
, et al
.
Effect of leukoreduction and temperature on risk of bacterial growth in CPDA-1 whole blood: a study of Escherichia coli
.
Transfusion
.
2021
;
61
(
Suppl 1
):
S80
9
. .
6.
Ramirez-Arcos
S
,
Kou
Y
,
Kumaran
D
,
Culibrk
B
,
Stewart
T
,
Schubert
P
, et al
.
Assessment of bacterial growth in leukoreduced cold-stored whole blood supports overnight hold at room temperature prior to filtration: a pilot study
.
Vox Sang
.
2022
;
117
(
5
):
678
84
. .
7.
Wagner
SJ
,
Robinette
D
,
Nazario
M
,
Moroff
G
.
Bacteria levels in components prepared from deliberately inoculated whole blood held for 8 or 24 hours at 20 to 24 degrees C
.
Transfusion
.
1995
;
35
(
11
):
911
6
. .
8.
Mohr
H
,
Bayer
A
,
Gravemann
U
,
Muller
TH
.
Elimination and multiplication of bacteria during preparation and storage of buffy coat-derived platelet concentrates
.
Transfusion
.
2006
;
46
(
6
):
949
55
. .
9.
Taha
M
,
Kalab
M
,
Yi
QL
,
Maurer
E
,
Jenkins
C
,
Schubert
P
, et al
.
Bacterial survival and distribution during buffy coat platelet production
.
Vox Sang
.
2016
;
111
(
4
):
333
40
. .
10.
Taha
M
,
Kyluik-Price
D
,
Kumaran
D
,
Scott
MD
,
Toyofuku
W
,
Ramirez-Arcos
S
.
Bacterial survival in whole blood depends on plasma sensitivity and resistance to neutrophil killing
.
Transfusion
.
2019
;
59
(
12
):
3674
82
. .
11.
Hogman
CF
,
Gong
J
,
Eriksson
L
,
Hambraeus
A
,
Johansson
CS
.
White cells protect donor blood against bacterial contamination
.
Transfusion
.
1991
;
31
(
7
):
620
6
. .
12.
Mufti
NA
,
Erickson
AC
,
North
AK
,
Hanson
D
,
Sawyer
L
,
Corash
LM
, et al
.
Treatment of whole blood (WB) and red blood cells (RBC) with S-303 inactivates pathogens and retains in vitro quality of stored RBC
.
Biologicals
.
2010
;
38
(
1
):
14
9
. .
13.
Reddy
HL
,
Doane
SK
,
Keil
SD
,
Marschner
S
,
Goodrich
RP
.
Development of a riboflavin and ultraviolet light-based device to treat whole blood
.
Transfusion
.
2013
;
53
(
Suppl 1
):
131S
36S
. .
14.
Handke
W
,
Gravemann
U
,
Muller
TH
,
Wagner
FF
,
Schulze
TJ
,
Seltsam
A
.
New ultraviolet C light-based method for pathogen inactivation of red blood cell units
.
Transfusion
.
2022
;
62
(
11
):
2314
23
. .
15.
Stormer
M
,
Arroyo
A
,
Brachert
J
,
Carrero
H
,
Devine
D
,
Epstein
JS
, et al
.
Establishment of the first international repository for transfusion-relevant bacteria reference strains: ISBT working party transfusion-transmitted infectious diseases (WP-TTID), subgroup on bacteria
.
Vox Sang
.
2012
;
102
(
1
):
22
31
. .
16.
Spindler-Raffel
E
,
Benjamin
RJ
,
McDonald
CP
,
Ramirez-Arcos
S
,
Aplin
K
,
Bekeredjian-Ding
I
, et al
.
Enlargement of the WHO international repository for platelet transfusion-relevant bacteria reference strains
.
Vox Sang
.
2017
;
112
(
8
):
713
22
. .
17.
Prax
M
,
Spindler-Raffel
E
,
McDonald
CP
,
Bearne
J
,
Satake
M
,
Kozakai
M
, et al
.
Establishment of transfusion-relevant bacteria reference strains for red blood cells
.
Vox Sang
.
2021
;
116
(
6
):
692
701
. .
18.
Eriksson
L
,
Shanwell
A
,
Gulliksson
H
,
Hogman
CF
,
Svensson
LA
,
Kristensen
J
, et al
.
Platelet concentrates in an additive solution prepared from pooled buffy coats. In vivo studies
.
Vox Sang
.
1993
;
64
(
3
):
133
8
. .
19.
Gulliksson
H
,
Larsson
S
,
Kumlien
G
,
Shanwell
A
.
Storage of platelets in additive solutions: effects of phosphate
.
Vox Sang
.
2000
;
78
(
3
):
176
84
. .
20.
Stormer
M
,
Vollmer
T
,
Kleesiek
K
,
Dreier
J
.
Spore-forming organisms in platelet concentrates: a challenge in transfusion bacterial safety
.
Transfus Med
.
2008
;
18
(
6
):
371
6
. .
21.
Benjamin
RJ
,
Wagner
SJ
.
Bacterial pathogen reduction requires validation under conditions of intended use
.
Transfusion
.
2015
;
55
(
9
):
2060
3
. .
22.
Holden
F
,
Foley
M
,
Devin
G
,
Kinsella
A
,
Murphy
WG
.
Coagulase-negative staphylococcal contamination of whole blood and its components: the effects of WBC reduction
.
Transfusion
.
2000
;
40
(
12
):
1508
13
. .
23.
Buchholz
DH
,
AuBuchon
JP
,
Snyder
EL
,
Kandler
R
,
Piscitelli
V
,
Pickard
C
, et al
.
Effects of white cell reduction on the resistance of blood components to bacterial multiplication
.
Transfusion
.
1994
;
34
(
10
):
852
7
. .
24.
Siblini
L
,
Lafeuillade
B
,
Ros
A
,
Garraud
O
,
Pozzetto
B
.
Influence of blood prestorage conditions and white blood cell filtration on the bacterial load of blood deliberately inoculated with Gram-positive and Gram-negative pathogens
.
Vox Sang
.
2004
;
87
(
4
):
241
9
. .
25.
Alizadeh
M
,
Wood
RL
,
Buchanan
CM
,
Bledsoe
CG
,
Wood
ME
,
McClellan
DS
, et al
.
Rapid separation of bacteria from blood: chemical aspects
.
Colloids Surf B Biointerfaces
.
2017
;
154
:
365
72
. .
26.
Pitt
WG
,
Alizadeh
M
,
Husseini
GA
,
McClellan
DS
,
Buchanan
CM
,
Bledsoe
CG
, et al
.
Rapid separation of bacteria from blood-review and outlook
.
Biotechnol Prog
.
2016
;
32
(
4
):
823
39
. .
27.
Funk
MB
,
Lohmann
A
,
Guenay
S
,
Henseler
O
,
Heiden
M
,
Hanschmann
KM
, et al
.
Transfusion-transmitted bacterial infections: haemovigilance data of German blood establishments (1997-2010)
.
Transfus Med Hemother
.
2011
;
38
(
4
):
266
71
. .
28.
Brecher
ME
,
Holland
PV
,
Pineda
AA
,
Tegtmeier
GE
,
Yomtovian
R
.
Growth of bacteria in inoculated platelets: implications for bacteria detection and the extension of platelet storage
.
Transfusion
.
2000
;
40
(
11
):
1308
12
. .
29.
Ramirez-Arcos
S
,
Perkins
H
,
Kou
Y
,
Mastronardi
C
,
Kumaran
D
,
Taha
M
, et al
.
Bacterial growth in red blood cell units exposed to uncontrolled temperatures: challenging the 30-minute rule
.
Vox Sang
.
2013
;
105
(
2
):
100
7
. .
30.
Selleng
K
,
Greinacher
A
.
10 Years of experience with the first thawed plasma bank in Germany
.
Transfus Med Hemother
.
2021
;
48
(
6
):
350
7
. .
31.
Muller
TH
,
Montag
T
,
Seltsam
AW
.
Laboratory evaluation of the effectiveness of pathogen reduction procedures for bacteria
.
Transfus Med Hemother
.
2011
;
38
(
4
):
242
50
. .
32.
Goodrich
RP
,
Gilmour
D
,
Hovenga
N
,
Keil
SD
.
A laboratory comparison of pathogen reduction technology treatment and culture of platelet products for addressing bacterial contamination concerns
.
Transfusion
.
2009
;
49
(
6
):
1205
16
. .
33.
Schmidt
M
,
Hourfar
MK
,
Sireis
W
,
Pfeiffer
U
,
Gottig
S
,
Kempf
VA
, et al
.
Evaluation of the effectiveness of a pathogen inactivation technology against clinically relevant transfusion-transmitted bacterial strains
.
Transfusion
.
2015
;
55
(
9
):
2104
12
. .
34.
McDonald
CP
,
Bearne
J
,
Aplin
K
,
Sawicka
D
.
Assessing the inactivation capabilities of two commercially available platelet component pathogen inactivation systems: effectiveness at end of shelf life
.
Vox Sang
.
2021
;
116
(
4
):
416
24
. .
35.
Walport
MJ
.
Complement. First of two parts
.
N Engl J Med
.
2001
;
344
(
14
):
1058
66
. .
36.
Gammon
RR
,
Reik
RA
,
Stern
M
,
Vassallo
RR
,
Waxman
DA
,
Young
PP
, et al
.
Acquired platelet storage container leaks and contamination with environmental bacteria: a preventable cause of bacterial sepsis
.
Transfusion
.
2022
;
62
(
3
):
641
50
. .
37.
Alabdullatif
M
,
Osman
IE
,
Alrasheed
M
,
Ramirez-Arcos
S
,
Alyousef
M
,
Althawadi
S
, et al
.
Evaluation of riboflavin and ultraviolet light treatment against Klebsiella pneumoniae in whole blood-derived platelets: a pilot study
.
Transfusion
.
2021
;
61
(
5
):
1562
9
. .
38.
Fuller
AK
,
Uglik
KM
,
Savage
WJ
,
Ness
PM
,
King
KE
.
Bacterial culture reduces but does not eliminate the risk of septic transfusion reactions to single-donor platelets
.
Transfusion
.
2009
;
49
(
12
):
2588
93
. .
39.
McDonald
C
,
Allen
J
,
Brailsford
S
,
Roy
A
,
Ball
J
,
Moule
R
, et al
.
Bacterial screening of platelet components by National Health Service Blood and Transplant, an effective risk reduction measure
.
Transfusion
.
2017
;
57
(
5
):
1122
31
. .
40.
Prax
M
,
Bekeredjian-Ding
I
,
Krut
O
.
Microbiological screening of platelet concentrates in europe
.
Transfus Med Hemother
.
2019
;
46
(
2
):
76
86
. .
41.
Fridey
JL
,
Stramer
SL
,
Nambiar
A
,
Moayeri
M
,
Bakkour
S
,
Langelier
C
, et al
.
Sepsis from an apheresis platelet contaminated with Acinetobacter calcoaceticus/baumannii complex bacteria and Staphylococcus saprophyticus after pathogen reduction
.
Transfusion
.
2020
;
60
(
9
):
1960
9
. .
42.
Fadeyi
EA
,
Wagner
SJ
,
Goldberg
C
,
Lu
T
,
Young
P
,
Bringmann
PW
, et al
.
Fatal sepsis associated with a storage container leak permitting platelet contamination with environmental bacteria after pathogen reduction
.
Transfusion
.
2021
;
61
(
2
):
641
8
. .
43.
Benjamin
RJ
,
Braschler
T
,
Weingand
T
,
Corash
LM
.
Hemovigilance monitoring of platelet septic reactions with effective bacterial protection systems
.
Transfusion
.
2017
;
57
(
12
):
2946
57
. .