TRANSFUSION COMPLICATIONS Evaluation of the effectiveness of a pathogen inactivation technology against clinically relevant transfusion-transmitted bacterial strains €ttig,2 Michael Schmidt,1 Michael K. Hourfar,1 Walid Sireis,1 Ulrich Pfeiffer,1 Stephan Go Volkhard A.J. Kempf,2 Carl P. McDonald,3 and Erhard Seifried1
BACKGROUND: To increase blood safety, various procedures are currently implemented, including donor selection, optimized donor arm disinfection, and diversion. In addition, pathogen inactivation (PI) techniques can be used for platelets (PLTs) and plasma concentrates. STUDY DESIGN AND METHODS: This study investigated the clinical efficacy of an inactivation technique for different blood components at two time points (12 and 35.5 hr). Eight transfusion-relevant bacterial strains were spiked at two different concentrations (100 and 1000 colony-forming units [CFUs]/bag) into whole blood (WB), apheresis PLTs (APs), and buffy coat (BC)-derived minipool PLTs. RESULTS: The bacterial concentrations were higher than 106 CFUs/mL within 24 hours after spiking depending on the particular bacterial strain. PI was absolute for all of the APs performed 12 hours after inoculation, but the bacterial strains of Klebsiella pneumoniae and Bacillus cereus were not completely inactivated in WB or BC PLTs, performed 35.5 and 12 hours after inoculation, respectively. CONCLUSION: The INTERCEPT PI system was not 100% effective for high concentrations of certain K. pneumoniae strains or spore-forming B. cereus. A critical observation was that the period between blood donation and inactivation needs to be minimal to enable efficient PI. In the case where PI cannot be performed immediately after preparation, a combination of a PI technology after the production of blood components with a rapid bacterial screen test on Day 4 or 5 after donation may offer a solution to further prevent the risk of bacterial transmission by transfusion.
T
he prevention of transfusion-related infectious diseases is one of the highest goals in transfusion medicine. The Hippocratic principle (primum non nocere) is important for transfusion medicine as well as for other disciplines.1 Blood safety with respect to bacterial contamination has improved significantly over the past decade, although the residual bacterial infection risk is still one to two logarithmic steps higher than the residual risk of a viral infection.2-6 Blood services have implemented the interventions of improved donor arm disinfection and diversion (predonation sampling), in which the first 30 to 40 mL of donor blood donation is diverted away from the collection bag.7-9 In addition, many countries have introduced the screening of platelet (PLT) concentrates using automated blood culture systems (e.g. BacT/ALERT, BACTEC, or VersaTrek).10-12 In the United States, screening has reduced bacterial adverse transfusion reactions between 50 and 75%, and in England, no confirmed septic reactions have
ABBREVIATIONS: AP(s) 5 apheresis platelet(s); BC 5 buffy coat; PI 5 pathogen inactivation; WB 5 whole blood. €rttemberg-Hessen From the 1DRK Blutspendedienst Baden-Wu gGmbH, Institute of Transfusion Medicine and Immunohematology, Goethe University Frankfurt am Main, and the 2Institute for Medical Microbiology and Infection Control, Hospital of Goethe-University, Frankfurt am Main, Germany; and the 3National Bacteriology Laboratory, NHS Blood and Transplant, Colindale, London, UK. Contract grant sponsor: Cerus Corporation. Address reprint requests to: Michael Schmidt, DRK Blut€rttemberg-Hessen, Sandhofstreet 1, spendedienst Baden-Wu 60529 Frankfurt, Germany; e-mail: [email protected]. Received for publication February 26, 2015; revision received April 9, 2015; and accepted April 10, 2015. doi:10.1111/trf.13171 C 2015 AABB V
TRANSFUSION 2015;55;2104–2112 2104 TRANSFUSION Volume 55, September 2015
EFFICIENCY OF PATHOGEN INACTIVATION
been reported since implementation in February 2011.9 However, limited rescreening of outdated PLTs with certain protocols suggest that more than 50% of contaminated PLT components are not detected.6,13-16 In England, more than 3000 outdated screened PLT components have been tested with no positives (C.P. McDonald, personal communication, 2014). However, based on more than 10 years of experience with this screening procedure, there are limitations. It is possible that there may be low numbers of bacteria present in the PLT component at the time of sampling. Bacteria may therefore not be present in the test sample and the contamination in the PLT bag will remain undetected.17-19 The low numbers of bacteria may then proliferate to higher levels during shelf life and cause patient morbidity and mortality. Several cases of “false-negative” screening results have been reported in the literature.16 A further limitation is that true positives may be detected by these systems after the PLT components have been transfused.14,16 As a consequence, transmission of bacteria by transfusion is not prevented. The most frequent cause of this is the detection of the slow-growing skin commensal Staphylococcus epidermidis, which can lead to serious infections such as, for example, endocarditis.20 An alternative is the use of pathogen inactivation (PI) techniques. The available systems can be classified into photochemical systems (INTERCEPT),21,22 photodynamic systems (Mirasol),23-25 and physical techniques (Theraflex).26-28 The purpose of the technologies of PI systems is to prevent bacterial replication. In the majority of bacterial PI studies effectiveness has been determined by measurement of logarithmic reduction.29-33 This is inappropriate for bacteria as they have the ability to multiply after treatment if not inactivated. A few noninactivated bacterial cells then have the potential to proliferate to sufficient number to cause patient morbidity and mortality. Our study investigated the effectiveness of the INTERCEPT Pathogen Inactivation System against a panel of transfusion relevant bacterial strains. This was performed in accordance with the manufacturers’ instructions in PLT components prepared from whole blood (WB)-, apheresis-, and buffy coat (BC)-derived minipools. PI was performed after a time delay (35.5 hr for WB and 12 hr for BC pools and apheresis-spiked components), to mimic the delay that may occur between PLT collection and PI treatment in routine practice. Sterility, as a characteristic value for an effective inactivation process, was examined at regular intervals and at the end of shelf life.
MATERIALS AND METHODS
TABLE 1. Analysis time intervals Time after bacterial spiking (hr) Time interval T0 5 spiking of blood products T1 5 time before BC production T2 5 time after BC production T3 5 time before PI T4 5 time after PI and before CAD T5 5 time after CAD T6 5 5 days after blood donation T7 5 7 days after blood donation
WB to BC PLTs
BC PLTs
APs
0 23 23.5 35.5 35.75
0 NA NA 12 12.25
0 NA NA 12 12.25
41.75 120 168
18.25 120 168
18.25 120 168
NA 5 not applicable.
forming units (CFUs)/bag. One species was spiked per blood unit at each concentration in replicates of four. The presence of bacteria was tested at various time intervals dependent on the component using the BacT/ALERT system (Table 1). Bacterial numbers were determined and species identification was performed to confirm the presence of the inoculated species.
Blood donation and PLT component preparation The blood donations were obtained at the Blood Transfu€rttemberg-Hessen in accordance sion Service Baden-Wu with the guidelines of the German Medical Association for the preparation of blood and blood components. The donor’s venipuncture sites were doubly disinfected by wiping with a preparation approved by the German Soci€ lke & ety of Hygiene and Microbiology (Octeniderm, Schu Mayr, Norderstedt, Germany) containing 30% 1-propanol, 45% 2-propanol, and 0.1% octenidine. The WB donations were collected over a period of 10 to 15 minutes. A diversion system was in operation directing 30 mL of each WB donation into a sample pouch.
Preparation of BC PLT concentrates WB units were centrifuged (4400 3 g for 30 min) and separated into three fractions (plasma concentrate, red blood cells [RBCs], and BC; 70 6 10 mL) using an automated extractor (Compomat G5, Fresenius Kabi, Bad Homburg, Germany). Four BC fractions were pooled and supplemented with 200 mL of plasma, followed by a soft-spin centrifugation at 540 3 g for 10 minutes. After centrifugation, the PLT-rich plasma was obtained using a separator. All of the BC-derived minipool PLT concentrates had a PLT concentration of more than 2 3 1011 PLTs per bag and a volume of approximately 270 to 280 mL.
Study protocol Blood units (WB-, apheresis-, and BC-pooled PLT concentrates) were inoculated with bacterial species in two concentrations of approximately 100 and 1000 colony-
Production of AP concentrates AP concentrates were produced according to the manufacturer’s instructions in the blood donation service using Volume 55, September 2015 TRANSFUSION 2105
SCHMIDT ET AL.
Bacterial identification TABLE 2. Bacterial strains Species
Strain designation
B. cereus K. pneumoniae K. pneumoniae Serratia marcescens S. aureus Staphylococcus epidermidis Streptococcus pyogenes Yersinia enterocolitica
PEI-B-07-23 PEI-B-08-09 PEI-B-05-01 ATCC 43862 PEI-B-23-07 PEI-B-13-03 PEI-B-20-05 DSM-11502
Bacteria were identified to the subspecies level using automated biochemical-based identification (Vitek2, biorieux), matrix-assisted laser desorption ionization-time Me rieux), and of flight mass spectrometry (Vitek MS, bioMe polymerase chain reaction of 16S rRNA gene with subsequent sequencing at each time point.
Intercept PI technique
an automatic separator (Haemonetics MCS, Haemonetics, Munich, Germany). After production, all of the PLT concentrate samples contained more than 2 3 1011 PLTs/bag and fewer than 1 3 106 and 1 3 109 white blood cells and RBCs, respectively. All AP concentrates have a volume of approximately 300 mL.
Bacterial reference strains Bacteria used in the study were chosen from those reported by the German hemovigilance data from the Paul-Ehrlich Institute to have caused severe patient complications. Gram-positive and Gram-negative bacterial strains were characterized by and obtained from the Department of Bacteriology of the Paul-Ehrlich-Institute (Agency of the German Federal Ministry of Health). Table 2 lists the bacterial strains used in this study.
PI with the INTERCEPT Blood System (Cerus Corporation, Concord, CA) was performed in accordance with the manufacturer’s instructions in two steps. Initially, the PLT concentrates were mixed with the psoralen derivate amotosalen (150 lmol/L). Then, the UV-A irradiation was performed (3 J/cm2 for 4-6 min). Subsequently, the remaining free amotosalen and the resulting photoproducts were absorbed using a component absorber (CAD phase, 6 hr). After the CAD phase, the inactivated PLT concentrates were made available for blood transfusion.
Statistical analysis Standard deviations and coefficients of variation were calculated using a computer spreadsheet (Excel 2013 for Windows, Microsoft Corp., Redmond, WA). p values were calculated with Fisher’s exact test or the t test. A p value of less than 0.05 was interpreted as significant.
Bacterial growth kinetics The growth kinetics of each of the selected bacterial strains was investigated in WB, AP concentrates, and BCderived PLT concentrates. Each sample was spiked with 100 CFUs/bag and stored under standard PLT storage conditions at 20 to 248C with agitation. The bacterial concentrations were determined every 6 hours up to 72 hours after spiking by serial dilution and plating onto blood agar plates. The growth kinetics of each bacterial strain was examined in replicates of 5.
Bacterial detection At designated time points in the study, qualitative detection of bacteria was performed using the BacT/ALERT rieux, Nu € rtigen, automated blood culture system (bioMe Germany). Aerobic and anaerobic culture bottles were each filled with 7.5 mL of sample volume and incubated for a maximum of 7 days at 35.58C in the BacT/ALERT 3D incubator.
Bacterial quantitation Bacterial CFUs were determined by plating serial dilutions of each sample on blood agar plates (Merck Millipore, Darmstadt, Germany) and subsequent manual counting after 24 to 48 hours cultured at 378C. 2106 TRANSFUSION Volume 55, September 2015
RESULTS Bacterial growth kinetics All bacterial strains in the study were shown to grow rapidly (n 5 5) in all component types. Figure 1 illustrates the growth kinetics of Klebsiella pneumoniae and Bacillus cereus, respectively. The initial postspiking concentrations are shown in Table 3. The 1000 CFUs/bag spiking concentrations range from 876 to 1143 CFUs/bag (mean inocula, 1019 CFUs/bag) and the 100 CFUs/bag range from 83 to 145 CFUs/bag (mean inocula, 103 CFUs/bag) in all component types.
PI evaluation Figure 2 shows the bacterial mean concentrations at each time point for all four independent experiments per bacterial strain, per spiking concentration and type of product inoculated before and after treatment. Six of the eight bacterial strains (75%) tested were completely inactivated by the INTERCEPT System. The bacterial concentrations at the 100 CFUs/bag spike just before treatment for WB were in the range of 104 to 107 CFUs/mL, apheresis 102 to 106 CFUs/mL, and pooled 103 to 106 CFUs/mL. Bacterial concentrations at the 1000 CFUs/bag per spike were just before treatment in the range of 103 to 108 CFUs/mL for
EFFICIENCY OF PATHOGEN INACTIVATION
TABLE 3. Bacterial concentrations (CFUs/bag) at Time Point T0 WB samples
Bacterial strain Target concentration B. cereus (PEI-B-07-23) K. pneumoniae (PEI-B-08-09) K. pneumoniae (PEI-B-05-01) S. marcescens (ATCC 43862) S. aureus (PEI-B-23-07) S. epidermidis (PEI-B-13-03) S. pyogenes (PEI-B-20-05) Y. enterocolitica (DSM 11502)
Fig. 1. (A) Growth kinetics of K. pneumoniae. (B) Growth kinetics of B. cereus. Mean growth kinetics of five independent experiments in WB (—), in APs ( ), and in BC PLTs (– – –).
WB, apheresis 102 to 106 CFUs/mL, and BC pools 103 to 107 CFUs/mL, respectively. Two bacterial strains were not inactivated: K. pneumoniae (PEI-B-08-09) and B. cereus (PEI-B-07-23). K. pneumoniae was not inactivated at an initial spiking concentration of 100 CFUs/bag in BC-derived PLTs (Fig. 2B) in one of four replicates tested. The concentration before treatment was 106 CFUs/mL and on Day 5 (3 days posttreatment) and Day 7 (5 days posttreatment) growth was observed at 102 and 106 CFUs/mL, respectively. K. pneumoniae (PEI-B-08-09) was also not inactivated with WB spiked at 1000 CFUs/bag in one replicate. The bacterial concentrations before treatment were 108 CFUs/mL and posttreatment on Day 5 104 CFUs/mL and 108 CFUs/mL on Day 7. In BC pools spiked with 1000 CFUs/bag breakthrough occurred in two replicates with K. pneumoniae (PEI-B-08-09). Bacterial concentrations
BC PLTs
APs
1000 100 1000 100 1000 100 981 89 1045 112 841 91 1104 94 1047 96 994 105 1044 108
979
104 1143
83
1087 117 1106 116 1075 145 891 955
96 945 103 1067
88 92
1021 111 942 107
1045 167 1023 106 876 1037 84 986 121 1141
97 83
for the breakthrough replicates were 107 and 106 CFUs/ mL before treatment, 102 on Day 5, and on Day 7, 106 and 105 CFUs/mL, respectively. B. cereus spiked in the vegetative form were not inactivated at the 1000 CFUs/bag spike in two replicates of the WB and two replicates of the BC PLT components. In WB for the breakthrough replicates the bacterial concentrations before treatment were 106 and 105 CFUs/mL and after treatment on Day 5 103 and 102 CFUs/mL and Day 7 105 CFUs/mL, respectively, for each replicate. In the BC pool replicates in which breakthrough was observed in two replicates, the bacterial concentrations before treatment were 104 CFUs/mL and after treatment on Day 5 102 and 101 CFU/mL and on Day 7 104 CFUs/mL, respectively, for each replicate. Cross-contamination within all experiments was excluded by identification of the bacterial strain at each time point in which bacteria were detected. In all cases, the bacterial strains detected later were identical to the spiked bacterial strains identified by DNA fingerprinting.
DISCUSSION Bacteria are microorganisms that can replicate in blood components without any additional target cells in contrast to transfusion relevant viruses like hepatitis B virus or human immunodeficiency virus (HIV)-1. PLT concentrates are at higher risk for bacterial contamination due to being stored at room temperature under aerobic conditions with agitation, which provides a welcoming environment for a broad range of bacterial species.34-36 Procedures to improve blood safety such as donor selection, improved donor arm disinfection, and diversion have already been successfully implemented in blood donation services worldwide. In recent years, PI technologies have been developed to irreversibly prevent the proliferation of bacteria, for example, by permanent cross-linking additives with double-stranded DNA and RNA. Volume 55, September 2015 TRANSFUSION 2107
SCHMIDT ET AL.
Fig. 2. Growth curves for each time point before and after inactivation. (A-F) Eight different transfusion-relevant bacteria strains (from left to right: B. cereus PEI-B-07-23; K. pneumoniae PEI-B-08-09; K. pneumoniae PEI-B-05-01; S. marcescens ATTC 43862; S. aureus PEI-B-23-07; S. epidermidis PEI-B-13-03; S. pyogenes PEI-B-20-05; Y. enterocolitica DSM 11502) were spiked into WB samples in replicates of four. Bacterial concentration was investigated at each time point by counting CFUs on blood agar plates 24 hours after culturing at 378C.
Our study investigates transfusion-relevant bacteria strains and the efficiency of the INTERCEPT PI system (Cerus Corporation). A key observation was that the bacterial concentration within the blood components has a direct influence on the efficiency of the PI technique. Complete inactivation to 100% was not achieved for the fast-growing K. pneumoniae strain PEI-B-08-09 in WB and BC PLTs, as growth before treatment exceeded the PI capacity of the system (Fig. 2). The manufacturer claims an inactivation capability for K. pneumoniae of log of more than 5.6. However, in our study this rapidly growing 2108 TRANSFUSION Volume 55, September 2015
organism in which breakthrough occurred was at a level of higher than 106 CFUs/mL at the time of treatment with a low initial spike of 100 and 1000 CFUs/bag. Bacterial number after treatment on Day 5 and Day 7 after spiking was substantial for this pathogenic Gram-negative organism. Transfusion of these PI-treated units potentially would have resulted in patient morbidity or mortality. This is of particular importance as the amount of multidrug-resistant K. pneumoniae is increasing continuously worldwide in recent years. The dramatic effect of transfusion of a PLT unit contaminated with this organism
EFFICIENCY OF PATHOGEN INACTIVATION
Fig. 2. (Continued)
is documented in the literature.37-40 In a study performed by Jacobs and colleagues41 it could be shown that a bacterial concentration exceeding 105 CFUs/mL would cause a transfusion reaction; the concentrations of the breakthrough replicates were in excess of these levels. In view of this finding, the time period between blood donation and the application of a PI technology needs to be as short as possible by damaging nucleic acids to prevent their replication.28,42 There are bacterial species that can exist in a spore or a vegetative form. A representative of this group of bacteria is B. cereus, which was included in our study. As in the case of K. pneumoniae, complete inactivation was not achieved for this organism in all replicates in WB or BC PLTs. Although the preparation of bacterial standards at
the Paul-Ehrlich Institute was of a vegetative form, isolated spores cannot be excluded and could be an explanation for the observed test results.43 The manufacturer makes no claim for the INTERCEPT System to inactivate bacterial spores. The inactivation capability of Mirasol44 and Thera45 flex has been demonstrated in several experimental studies to be lower than that of the INTERCEPT System. Breakthrough was observed in our study and is therefore possible with other PI systems. Kwon and colleagues46 compared the experimental inactivation potential of INTERCEPT and Mirasol with each other especially for PLT-rich plasma. The inactivation efficacy expressed as log reduction was at least 4.19 versus at least 4.23 for HIV-1; 1.83 versus at least 6.03 for bovine viral diarrhea virus; Volume 55, September 2015 TRANSFUSION 2109
SCHMIDT ET AL.
Fig. 2. (Continued)
2.73 versus at least 5.2 for porcine parvovirus; 5.45 versus at least 9.22 for Escherichia coli; 4.26 versus at least 10.11 for Staphylococcus aureus; and 5.09 versus at least 7.74 for Bacillus subtilis for Mirasol and INTERCEPT PI systems, respectively. Kwon and colleagues46 reported that bacterial growth in PLTs inoculated with low concentrations of S. aureus or B. subtilis was detected only with Mirasol. Based on the time period between blood donation and the implementation of a PI technology, we believe that there is an advantage to APs because the timeconsuming preparation of BCs, the pooling process and the additional centrifugation steps are not necessary. This position is supported by our finding that breakthrough infection was not observed within the study for APs. It is possible that differences in leukoreduction, 2110 TRANSFUSION Volume 55, September 2015
PLT concentration, and volume may have played a role in rendering PI more effective in APs; however, conclusions regarding different residual infectious risks between APs and BC PLTs cannot be drawn from our data. Funk and coworkers47 summarized the hemovigilance data between 1997 and 2010 and described four bacterial fatalities in AP and four in BC PLTs. In addition, Schrezenmeier and colleagues6 examined the risks of bacterial infection between BC PLTs and APs. After screening more than 50,000 PLT concentrates, no significant difference in the infectious risk for both groups of blood components was observed. This study underscores our understanding that the residual transfusiontransmitted risk of bacterial infection is a very complex mechanism with many variables.
EFFICIENCY OF PATHOGEN INACTIVATION
In principle, PI systems offer an alternative to bacterial screening by culture, but neither are 100% effective. Theoretically PI may be effective as the initial bacterial concentration within the PLT concentrate is low, most likely between 10 to 100 CFUs/bag, but there are no experimental data to support this claim. The K. pneumoniae breakthrough occurred with one replicate spiked at the low inoculum of 100 CFUs/bag in BC pooled. Potentially a higher concentration will induce clinical symptoms and will usually exclude donors from blood donation. Nevertheless, all preventive techniques and technologies will have advantages and disadvantages. We have demonstrated that the INTERCEPT PI system has limitations for fast-growing bacteria and spore-forming bacteria. These findings may suggest that treatment must ideally be performed as soon as practical after collection. If this is not operationally practicable, PI-treated PLT components transfused near to the end of their shelf life could be screened for bacteria using rapid detection techniques.48-51 This combination of different procedures would be useful to combine the advantages of various technologies and reduce or prevent potential transmission. However, although the combination of a PI technology together with a rapid bacterial detection method will improve blood safety, the costs for both procedures have to be significantly reduced to enable the implementation of both technologies in blood donor services worldwide. ACKNOWLEDGMENTS
€ ller TH, et al. Bacte6. Schrezenmeier H, Walther-Wenke G, Mu rial contamination of platelet concentrates: results of a prospective multicenter study comparing pooled whole bloodderived platelets and apheresis platelets. Transfusion 2007; 47:644-52. 7. de Korte D, Curvers J, de Kort WL, et al. Effects of skin disinfection method, deviation bag, and bacterial screening on clinical safety of platelet transfusions in the Netherlands. Transfusion 2006;46:476-85. 8. de Korte D, Marcelis JH, Verhoeven AJ, et al. Diversion of first blood volume results in a reduction of bacterial contamination for whole-blood collections. Vox Sang 2002;83:13-6. 9. McDonald CP, Roy A, Mahajan P, et al. Relative values of the interventions of diversion and improved donor-arm disinfection to reduce the bacterial risk from blood transfusion. Vox Sang 2004;86:178-82. 10. Riedel S, Siwek G, Beekmann SE, et al. Comparison of the BACTEC 9240 and BacT/Alert blood culture systems for detection of bacterial contamination in platelet concentrates. J Clin Microbiol 2006;44:2262-4. 11. Espasa M, Salvado M, Vicente E, et al. Evaluation of the VersaTREK system compared to the Bactec MGIT 960 system for first-line drug susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol 2012;50:488-91. 12. Martinez RM, Bauerle ER, Fang FC, et al. Evaluation of three rapid diagnostic methods for direct identification of microorganisms in positive blood cultures. J Clin Microbiol 2014; 52:2521-9. 13. Dumont LJ, Kleinman S, Murphy JR, et al. Screening of
We thank Ms Borschu for technical assistance by preparing the
single-donor apheresis platelets for bacterial contamination: the PASSPORT study results. Transfusion 2010;50:589-99.
blood components and Dr Thomas Montag (2012) for preparing
14. Fang CT, Chambers LA, Kennedy J, et al. Detection of bacte-
ready-to-use bacterial standards.
rial contamination in apheresis platelet products: American Red Cross experience, 2004. Transfusion 2005;45:1845-52. 15. Ramırez-Arcos S, Kou Y, Mastronardi C, et al. Bacterial
CONFLICT OF INTEREST The authors have disclosed no conflicts of interest.
screening of outdated buffy coat platelet pools using a culture system and a rapid immunoassay. Transfusion 2011;51: 2566-72.
REFERENCES 1. Kal EF. Conflicted caregivers. Am J Psychiatry 2000;157:305. 2. Dodd RY. Emerging pathogens and their implications for the blood supply and transfusion transmitted infections. Br J Haematol 2012;159:135-42. 3. Offergeld R. [Blood - surely but safely!]. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2012;55:9056. 4. Sireis W, Hourfar MK, Seifried E, et al. How to improve blood safety with respect to bacterial contaminations? Vox Sang 2012;103:87-8. 5. Hourfar MK, Jork C, Schottstedt V, et al. Experience of German Red Cross blood donor services with nucleic acid testing: results of screening more than 30 million blood donations for human immunodeficiency virus-1, hepatitis C virus, and hepatitis B virus. Transfusion 2008;48:1558-66.
16. te Boekhorst PA, Beckers EA, Vos MC, et al. Clinical significance of bacteriologic screening in platelet concentrates. Transfusion 2005;45:514-9. 17. Montag T. Strategies of bacteria screening in cellular blood components. Clin Chem Lab Med 2008;46:926-32. 18. Murphy WG, Coakley P. Testing platelet components for bacterial contamination. Transfus Apher Sci 2011;45:69-74. 19. Owusu-Ofori S, Temple J, Sarkodie F, et al. Predonation screening of blood donors with rapid tests: implementation and efficacy of a novel approach to blood safety in resourcepoor settings. Transfusion 2005;45:133-40. 20. Polewczyk A, Janion M, Podlaski R, et al. Clinical manifestations of lead-dependent infective endocarditis: analysis of 414 cases. Eur J Clin Microbiol Infect Dis 2014;33:1601-8. 21. Irsch J, Lin L. Pathogen inactivation of platelet and plasma blood components for transfusion using the INTERCEPT Blood System. Transfus Med Hemother 2011;38:19-31. Volume 55, September 2015 TRANSFUSION 2111
SCHMIDT ET AL.
22. Osselaer JC, Messe N, Hervig T, et al. A prospective observational cohort safety study of 5106 platelet transfusions with
contamination based on unconfirmed positive results. Vox Sang 2014;106:23-30.
components prepared with photochemical pathogen inactivation treatment. Transfusion 2008;48:1061-71. 23. Jocic´ M, Trkuljic´ M, Jovicic´ D, et al. Mirasol PRT system inac-
38. Eder AF, Kennedy JM, Dy BA, et al. Bacterial screening of apheresis platelets and the residual risk of septic transfusion
tivation efficacy evaluated in platelet concentrates by bacteria-contamination model. Vojnosanit Pregl 2011;68:
Transfusion 2007;47:1134-42. 39. Niu MT, Knippen M, Simmons L, et al. Transfusion-transmit-
1041-6. 24. Schubert P, Culibrk B, Coupland D, et al. Riboflavin and ultraviolet light treatment potentiates vasodilator-stimulated phosphoprotein Ser-239 phosphorylation in platelet concentrates during storage. Transfusion 2012;52:397-408. 25. Vanlandingham DL, Keil SD, Horne KM, et al. Photochemical inactivation of chikungunya virus in plasma and platelets using the Mirasol pathogen reduction technology system. Transfusion 2013;53:284-90. 26. Lobo SL, Mehta N, Forgione AG, et al. Use of Theraflex-TMJ topical cream for the treatment of temporomandibular joint and muscle pain. Cranio 2004;22:137-44. 27. Politis C, Kavallierou L, Hantziara S, et al. Quality and safety of fresh-frozen plasma inactivated and leucoreduced with the Theraflex methylene blue system including the Blueflex filter: 5 years’ experience. Vox Sang 2007;92:319-26. € ller TH. UVC irradiation for pathogen reduc28. Seltsam A, Mu tion of platelet concentrates and plasma. Transfus Med Hemother 2011;38:43-54. 29. Dodd R, Kurt Roth W, Ashford P, et al. Transfusion medicine and safety. Biologicals 2009;37:62-70.
reactions: the American Red Cross experience (2004-2006).
ted Klebsiella pneumoniae fatalities, 1995 to 2004. Transfus Med Rev 2006;20:149-57. 40. Sazama K. Bacteria in blood for transfusion. A review. Arch Pathol Lab Med 1994;118:350-65. 41. Jacobs MR, Good CE, Lazarus HM, et al. Relationship between bacterial load, species virulence, and transfusion reaction with transfusion of bacterially contaminated platelets. Clin Infect Dis 2008;46:1214-20. € ller TH. Update on the use of pathogen42. Seltsam A, Mu reduced human plasma and platelet concentrates. Br J Haematol 2013;162:442-54. € rmer M, Arroyo A, Brachert J, et al. Establishment of the 43. Sto first international repository for transfusion-relevant bacteria reference strains: ISBT working party transfusiontransmitted infectious diseases (WP-TTID), subgroup on bacteria. Vox Sang 2012;102:22-31. 44. Goodrich RP, Edrich RA, Li J, et al. The Mirasol PRT system for pathogen reduction of platelets and plasma: an overview of current status and future trends. Transfus Apher Sci 2006; 35:5-17. 45. Seghatchian J, Tolksdorf F. Characteristics of the theraflex
30. Postma MJ, van Hulst M, De Wolf JT, et al. Cost-effectiveness
UV-platelets pathogen inactivation system—an update.
of pathogen inactivation for platelet transfusions in the Netherlands. Transfus Med 2005;15:379-87.
Transfus Apher Sci 2012;46:221-9. 46. Kwon SY, Kim IS, Bae JE, et al. Pathogen inactivation efficacy
31. Rock G. A comparison of methods of pathogen inactivation
of Mirasol PRT System and INTERCEPT Blood System for
of FFP. Vox Sang 2011;100:169-78. 32. Ruane PH, Edrich R, Gampp D, et al. Photochemical inacti-
non-leucoreduced platelet-rich plasma-derived platelets suspended in plasma. Vox Sang 2014;107:254-60.
vation of selected viruses and bacteria in platelet concentrates using riboflavin and light. Transfusion 2004;44:877-85. 33. Stramer SL. Current perspectives in transfusion-transmitted infectious diseases: emerging and re-emerging infections. ISBT Sci Ser 2014;9:30-6. 34. Mejia YX, Mao H, Forde NR, et al. Thermal probing of E. coli RNA polymerase off-pathway mechanisms. J Mol Biol 2008; 382:628-37. 35. Tassou CC, Panagou EZ, Samaras FJ, et al. Temperatureassisted high hydrostatic pressure inactivation of Staphylococcus aureus in a ham model system: evaluation in selective and nonselective medium. J Appl Microbiol 2008;104: 1764-73. 36. Wong TL, Whyte RJ, Graham CG, et al. Growth of enterotoxigenic Staphylococcus aureus in povi masima, a traditional Pacific island food. J Appl Microbiol 2004;97:1185-91. 37. Benjamin RJ, Dy B, Perez J, et al. Bacterial culture of apheresis platelets: a mathematical model of the residual rate of
2112 TRANSFUSION Volume 55, September 2015
47. Funk MB, Lohmann A, Guenay S, et al. Transfusion-transmitted bacterial infections—haemovigilance data of German blood establishments (1997-2010). Transfus Med Hemother 2011;38:266-71. 48. Dreier J, Vollmer T, Kleesiek K. Novel flow cytometry-based screening for bacterial contamination of donor platelet preparations compared with other rapid screening methods. Clin Chem 2009;55:1492-502. € ster B, Daiss C, et al. Extension of platelet shelf 49. Sireis W, Ru life from 4 to 5 days by implementation of a new screening strategy in Germany. Vox Sang 2011;101:191-9. 50. Vollmer T, Dreier J, Schottstedt V, et al. Detection of bacterial contamination in platelet concentrates by a sensitive flow cytometric assay (BactiFlow): a multicentre validation study. Transfus Med 2012;22:262-71. 51. Vollmer T, Knabbe C, Dreier J. Novel flow cytometric screening method for bacterial contamination of red blood cells: a proof-of-principle evaluation. Transfusion 2014;54:900-9.
BACKGROUND: To increase blood safety, various procedures are currently implemented, including donor selection, optimized donor arm disinfection, and diversion. In addition, pathogen inactivation (PI) techniques can be used for platelets (PLTs) and plasma concentrates. STUDY DESIGN AND METHODS: This study investigated the clinical efficacy of an inactivation technique for different blood components at two time points (12 and 35.5 hr). Eight transfusion-relevant bacterial strains were spiked at two different concentrations (100 and 1000 colony-forming units [CFUs]/bag) into whole blood (WB), apheresis PLTs (APs), and buffy coat (BC)-derived minipool PLTs. RESULTS: The bacterial concentrations were higher than 106 CFUs/mL within 24 hours after spiking depending on the particular bacterial strain. PI was absolute for all of the APs performed 12 hours after inoculation, but the bacterial strains of Klebsiella pneumoniae and Bacillus cereus were not completely inactivated in WB or BC PLTs, performed 35.5 and 12 hours after inoculation, respectively. CONCLUSION: The INTERCEPT PI system was not 100% effective for high concentrations of certain K. pneumoniae strains or spore-forming B. cereus. A critical observation was that the period between blood donation and inactivation needs to be minimal to enable efficient PI. In the case where PI cannot be performed immediately after preparation, a combination of a PI technology after the production of blood components with a rapid bacterial screen test on Day 4 or 5 after donation may offer a solution to further prevent the risk of bacterial transmission by transfusion.
T
he prevention of transfusion-related infectious diseases is one of the highest goals in transfusion medicine. The Hippocratic principle (primum non nocere) is important for transfusion medicine as well as for other disciplines.1 Blood safety with respect to bacterial contamination has improved significantly over the past decade, although the residual bacterial infection risk is still one to two logarithmic steps higher than the residual risk of a viral infection.2-6 Blood services have implemented the interventions of improved donor arm disinfection and diversion (predonation sampling), in which the first 30 to 40 mL of donor blood donation is diverted away from the collection bag.7-9 In addition, many countries have introduced the screening of platelet (PLT) concentrates using automated blood culture systems (e.g. BacT/ALERT, BACTEC, or VersaTrek).10-12 In the United States, screening has reduced bacterial adverse transfusion reactions between 50 and 75%, and in England, no confirmed septic reactions have
ABBREVIATIONS: AP(s) 5 apheresis platelet(s); BC 5 buffy coat; PI 5 pathogen inactivation; WB 5 whole blood. €rttemberg-Hessen From the 1DRK Blutspendedienst Baden-Wu gGmbH, Institute of Transfusion Medicine and Immunohematology, Goethe University Frankfurt am Main, and the 2Institute for Medical Microbiology and Infection Control, Hospital of Goethe-University, Frankfurt am Main, Germany; and the 3National Bacteriology Laboratory, NHS Blood and Transplant, Colindale, London, UK. Contract grant sponsor: Cerus Corporation. Address reprint requests to: Michael Schmidt, DRK Blut€rttemberg-Hessen, Sandhofstreet 1, spendedienst Baden-Wu 60529 Frankfurt, Germany; e-mail: [email protected]. Received for publication February 26, 2015; revision received April 9, 2015; and accepted April 10, 2015. doi:10.1111/trf.13171 C 2015 AABB V
TRANSFUSION 2015;55;2104–2112 2104 TRANSFUSION Volume 55, September 2015
EFFICIENCY OF PATHOGEN INACTIVATION
been reported since implementation in February 2011.9 However, limited rescreening of outdated PLTs with certain protocols suggest that more than 50% of contaminated PLT components are not detected.6,13-16 In England, more than 3000 outdated screened PLT components have been tested with no positives (C.P. McDonald, personal communication, 2014). However, based on more than 10 years of experience with this screening procedure, there are limitations. It is possible that there may be low numbers of bacteria present in the PLT component at the time of sampling. Bacteria may therefore not be present in the test sample and the contamination in the PLT bag will remain undetected.17-19 The low numbers of bacteria may then proliferate to higher levels during shelf life and cause patient morbidity and mortality. Several cases of “false-negative” screening results have been reported in the literature.16 A further limitation is that true positives may be detected by these systems after the PLT components have been transfused.14,16 As a consequence, transmission of bacteria by transfusion is not prevented. The most frequent cause of this is the detection of the slow-growing skin commensal Staphylococcus epidermidis, which can lead to serious infections such as, for example, endocarditis.20 An alternative is the use of pathogen inactivation (PI) techniques. The available systems can be classified into photochemical systems (INTERCEPT),21,22 photodynamic systems (Mirasol),23-25 and physical techniques (Theraflex).26-28 The purpose of the technologies of PI systems is to prevent bacterial replication. In the majority of bacterial PI studies effectiveness has been determined by measurement of logarithmic reduction.29-33 This is inappropriate for bacteria as they have the ability to multiply after treatment if not inactivated. A few noninactivated bacterial cells then have the potential to proliferate to sufficient number to cause patient morbidity and mortality. Our study investigated the effectiveness of the INTERCEPT Pathogen Inactivation System against a panel of transfusion relevant bacterial strains. This was performed in accordance with the manufacturers’ instructions in PLT components prepared from whole blood (WB)-, apheresis-, and buffy coat (BC)-derived minipools. PI was performed after a time delay (35.5 hr for WB and 12 hr for BC pools and apheresis-spiked components), to mimic the delay that may occur between PLT collection and PI treatment in routine practice. Sterility, as a characteristic value for an effective inactivation process, was examined at regular intervals and at the end of shelf life.
MATERIALS AND METHODS
TABLE 1. Analysis time intervals Time after bacterial spiking (hr) Time interval T0 5 spiking of blood products T1 5 time before BC production T2 5 time after BC production T3 5 time before PI T4 5 time after PI and before CAD T5 5 time after CAD T6 5 5 days after blood donation T7 5 7 days after blood donation
WB to BC PLTs
BC PLTs
APs
0 23 23.5 35.5 35.75
0 NA NA 12 12.25
0 NA NA 12 12.25
41.75 120 168
18.25 120 168
18.25 120 168
NA 5 not applicable.
forming units (CFUs)/bag. One species was spiked per blood unit at each concentration in replicates of four. The presence of bacteria was tested at various time intervals dependent on the component using the BacT/ALERT system (Table 1). Bacterial numbers were determined and species identification was performed to confirm the presence of the inoculated species.
Blood donation and PLT component preparation The blood donations were obtained at the Blood Transfu€rttemberg-Hessen in accordance sion Service Baden-Wu with the guidelines of the German Medical Association for the preparation of blood and blood components. The donor’s venipuncture sites were doubly disinfected by wiping with a preparation approved by the German Soci€ lke & ety of Hygiene and Microbiology (Octeniderm, Schu Mayr, Norderstedt, Germany) containing 30% 1-propanol, 45% 2-propanol, and 0.1% octenidine. The WB donations were collected over a period of 10 to 15 minutes. A diversion system was in operation directing 30 mL of each WB donation into a sample pouch.
Preparation of BC PLT concentrates WB units were centrifuged (4400 3 g for 30 min) and separated into three fractions (plasma concentrate, red blood cells [RBCs], and BC; 70 6 10 mL) using an automated extractor (Compomat G5, Fresenius Kabi, Bad Homburg, Germany). Four BC fractions were pooled and supplemented with 200 mL of plasma, followed by a soft-spin centrifugation at 540 3 g for 10 minutes. After centrifugation, the PLT-rich plasma was obtained using a separator. All of the BC-derived minipool PLT concentrates had a PLT concentration of more than 2 3 1011 PLTs per bag and a volume of approximately 270 to 280 mL.
Study protocol Blood units (WB-, apheresis-, and BC-pooled PLT concentrates) were inoculated with bacterial species in two concentrations of approximately 100 and 1000 colony-
Production of AP concentrates AP concentrates were produced according to the manufacturer’s instructions in the blood donation service using Volume 55, September 2015 TRANSFUSION 2105
SCHMIDT ET AL.
Bacterial identification TABLE 2. Bacterial strains Species
Strain designation
B. cereus K. pneumoniae K. pneumoniae Serratia marcescens S. aureus Staphylococcus epidermidis Streptococcus pyogenes Yersinia enterocolitica
PEI-B-07-23 PEI-B-08-09 PEI-B-05-01 ATCC 43862 PEI-B-23-07 PEI-B-13-03 PEI-B-20-05 DSM-11502
Bacteria were identified to the subspecies level using automated biochemical-based identification (Vitek2, biorieux), matrix-assisted laser desorption ionization-time Me rieux), and of flight mass spectrometry (Vitek MS, bioMe polymerase chain reaction of 16S rRNA gene with subsequent sequencing at each time point.
Intercept PI technique
an automatic separator (Haemonetics MCS, Haemonetics, Munich, Germany). After production, all of the PLT concentrate samples contained more than 2 3 1011 PLTs/bag and fewer than 1 3 106 and 1 3 109 white blood cells and RBCs, respectively. All AP concentrates have a volume of approximately 300 mL.
Bacterial reference strains Bacteria used in the study were chosen from those reported by the German hemovigilance data from the Paul-Ehrlich Institute to have caused severe patient complications. Gram-positive and Gram-negative bacterial strains were characterized by and obtained from the Department of Bacteriology of the Paul-Ehrlich-Institute (Agency of the German Federal Ministry of Health). Table 2 lists the bacterial strains used in this study.
PI with the INTERCEPT Blood System (Cerus Corporation, Concord, CA) was performed in accordance with the manufacturer’s instructions in two steps. Initially, the PLT concentrates were mixed with the psoralen derivate amotosalen (150 lmol/L). Then, the UV-A irradiation was performed (3 J/cm2 for 4-6 min). Subsequently, the remaining free amotosalen and the resulting photoproducts were absorbed using a component absorber (CAD phase, 6 hr). After the CAD phase, the inactivated PLT concentrates were made available for blood transfusion.
Statistical analysis Standard deviations and coefficients of variation were calculated using a computer spreadsheet (Excel 2013 for Windows, Microsoft Corp., Redmond, WA). p values were calculated with Fisher’s exact test or the t test. A p value of less than 0.05 was interpreted as significant.
Bacterial growth kinetics The growth kinetics of each of the selected bacterial strains was investigated in WB, AP concentrates, and BCderived PLT concentrates. Each sample was spiked with 100 CFUs/bag and stored under standard PLT storage conditions at 20 to 248C with agitation. The bacterial concentrations were determined every 6 hours up to 72 hours after spiking by serial dilution and plating onto blood agar plates. The growth kinetics of each bacterial strain was examined in replicates of 5.
Bacterial detection At designated time points in the study, qualitative detection of bacteria was performed using the BacT/ALERT rieux, Nu € rtigen, automated blood culture system (bioMe Germany). Aerobic and anaerobic culture bottles were each filled with 7.5 mL of sample volume and incubated for a maximum of 7 days at 35.58C in the BacT/ALERT 3D incubator.
Bacterial quantitation Bacterial CFUs were determined by plating serial dilutions of each sample on blood agar plates (Merck Millipore, Darmstadt, Germany) and subsequent manual counting after 24 to 48 hours cultured at 378C. 2106 TRANSFUSION Volume 55, September 2015
RESULTS Bacterial growth kinetics All bacterial strains in the study were shown to grow rapidly (n 5 5) in all component types. Figure 1 illustrates the growth kinetics of Klebsiella pneumoniae and Bacillus cereus, respectively. The initial postspiking concentrations are shown in Table 3. The 1000 CFUs/bag spiking concentrations range from 876 to 1143 CFUs/bag (mean inocula, 1019 CFUs/bag) and the 100 CFUs/bag range from 83 to 145 CFUs/bag (mean inocula, 103 CFUs/bag) in all component types.
PI evaluation Figure 2 shows the bacterial mean concentrations at each time point for all four independent experiments per bacterial strain, per spiking concentration and type of product inoculated before and after treatment. Six of the eight bacterial strains (75%) tested were completely inactivated by the INTERCEPT System. The bacterial concentrations at the 100 CFUs/bag spike just before treatment for WB were in the range of 104 to 107 CFUs/mL, apheresis 102 to 106 CFUs/mL, and pooled 103 to 106 CFUs/mL. Bacterial concentrations at the 1000 CFUs/bag per spike were just before treatment in the range of 103 to 108 CFUs/mL for
EFFICIENCY OF PATHOGEN INACTIVATION
TABLE 3. Bacterial concentrations (CFUs/bag) at Time Point T0 WB samples
Bacterial strain Target concentration B. cereus (PEI-B-07-23) K. pneumoniae (PEI-B-08-09) K. pneumoniae (PEI-B-05-01) S. marcescens (ATCC 43862) S. aureus (PEI-B-23-07) S. epidermidis (PEI-B-13-03) S. pyogenes (PEI-B-20-05) Y. enterocolitica (DSM 11502)
Fig. 1. (A) Growth kinetics of K. pneumoniae. (B) Growth kinetics of B. cereus. Mean growth kinetics of five independent experiments in WB (—), in APs ( ), and in BC PLTs (– – –).
WB, apheresis 102 to 106 CFUs/mL, and BC pools 103 to 107 CFUs/mL, respectively. Two bacterial strains were not inactivated: K. pneumoniae (PEI-B-08-09) and B. cereus (PEI-B-07-23). K. pneumoniae was not inactivated at an initial spiking concentration of 100 CFUs/bag in BC-derived PLTs (Fig. 2B) in one of four replicates tested. The concentration before treatment was 106 CFUs/mL and on Day 5 (3 days posttreatment) and Day 7 (5 days posttreatment) growth was observed at 102 and 106 CFUs/mL, respectively. K. pneumoniae (PEI-B-08-09) was also not inactivated with WB spiked at 1000 CFUs/bag in one replicate. The bacterial concentrations before treatment were 108 CFUs/mL and posttreatment on Day 5 104 CFUs/mL and 108 CFUs/mL on Day 7. In BC pools spiked with 1000 CFUs/bag breakthrough occurred in two replicates with K. pneumoniae (PEI-B-08-09). Bacterial concentrations
BC PLTs
APs
1000 100 1000 100 1000 100 981 89 1045 112 841 91 1104 94 1047 96 994 105 1044 108
979
104 1143
83
1087 117 1106 116 1075 145 891 955
96 945 103 1067
88 92
1021 111 942 107
1045 167 1023 106 876 1037 84 986 121 1141
97 83
for the breakthrough replicates were 107 and 106 CFUs/ mL before treatment, 102 on Day 5, and on Day 7, 106 and 105 CFUs/mL, respectively. B. cereus spiked in the vegetative form were not inactivated at the 1000 CFUs/bag spike in two replicates of the WB and two replicates of the BC PLT components. In WB for the breakthrough replicates the bacterial concentrations before treatment were 106 and 105 CFUs/mL and after treatment on Day 5 103 and 102 CFUs/mL and Day 7 105 CFUs/mL, respectively, for each replicate. In the BC pool replicates in which breakthrough was observed in two replicates, the bacterial concentrations before treatment were 104 CFUs/mL and after treatment on Day 5 102 and 101 CFU/mL and on Day 7 104 CFUs/mL, respectively, for each replicate. Cross-contamination within all experiments was excluded by identification of the bacterial strain at each time point in which bacteria were detected. In all cases, the bacterial strains detected later were identical to the spiked bacterial strains identified by DNA fingerprinting.
DISCUSSION Bacteria are microorganisms that can replicate in blood components without any additional target cells in contrast to transfusion relevant viruses like hepatitis B virus or human immunodeficiency virus (HIV)-1. PLT concentrates are at higher risk for bacterial contamination due to being stored at room temperature under aerobic conditions with agitation, which provides a welcoming environment for a broad range of bacterial species.34-36 Procedures to improve blood safety such as donor selection, improved donor arm disinfection, and diversion have already been successfully implemented in blood donation services worldwide. In recent years, PI technologies have been developed to irreversibly prevent the proliferation of bacteria, for example, by permanent cross-linking additives with double-stranded DNA and RNA. Volume 55, September 2015 TRANSFUSION 2107
SCHMIDT ET AL.
Fig. 2. Growth curves for each time point before and after inactivation. (A-F) Eight different transfusion-relevant bacteria strains (from left to right: B. cereus PEI-B-07-23; K. pneumoniae PEI-B-08-09; K. pneumoniae PEI-B-05-01; S. marcescens ATTC 43862; S. aureus PEI-B-23-07; S. epidermidis PEI-B-13-03; S. pyogenes PEI-B-20-05; Y. enterocolitica DSM 11502) were spiked into WB samples in replicates of four. Bacterial concentration was investigated at each time point by counting CFUs on blood agar plates 24 hours after culturing at 378C.
Our study investigates transfusion-relevant bacteria strains and the efficiency of the INTERCEPT PI system (Cerus Corporation). A key observation was that the bacterial concentration within the blood components has a direct influence on the efficiency of the PI technique. Complete inactivation to 100% was not achieved for the fast-growing K. pneumoniae strain PEI-B-08-09 in WB and BC PLTs, as growth before treatment exceeded the PI capacity of the system (Fig. 2). The manufacturer claims an inactivation capability for K. pneumoniae of log of more than 5.6. However, in our study this rapidly growing 2108 TRANSFUSION Volume 55, September 2015
organism in which breakthrough occurred was at a level of higher than 106 CFUs/mL at the time of treatment with a low initial spike of 100 and 1000 CFUs/bag. Bacterial number after treatment on Day 5 and Day 7 after spiking was substantial for this pathogenic Gram-negative organism. Transfusion of these PI-treated units potentially would have resulted in patient morbidity or mortality. This is of particular importance as the amount of multidrug-resistant K. pneumoniae is increasing continuously worldwide in recent years. The dramatic effect of transfusion of a PLT unit contaminated with this organism
EFFICIENCY OF PATHOGEN INACTIVATION
Fig. 2. (Continued)
is documented in the literature.37-40 In a study performed by Jacobs and colleagues41 it could be shown that a bacterial concentration exceeding 105 CFUs/mL would cause a transfusion reaction; the concentrations of the breakthrough replicates were in excess of these levels. In view of this finding, the time period between blood donation and the application of a PI technology needs to be as short as possible by damaging nucleic acids to prevent their replication.28,42 There are bacterial species that can exist in a spore or a vegetative form. A representative of this group of bacteria is B. cereus, which was included in our study. As in the case of K. pneumoniae, complete inactivation was not achieved for this organism in all replicates in WB or BC PLTs. Although the preparation of bacterial standards at
the Paul-Ehrlich Institute was of a vegetative form, isolated spores cannot be excluded and could be an explanation for the observed test results.43 The manufacturer makes no claim for the INTERCEPT System to inactivate bacterial spores. The inactivation capability of Mirasol44 and Thera45 flex has been demonstrated in several experimental studies to be lower than that of the INTERCEPT System. Breakthrough was observed in our study and is therefore possible with other PI systems. Kwon and colleagues46 compared the experimental inactivation potential of INTERCEPT and Mirasol with each other especially for PLT-rich plasma. The inactivation efficacy expressed as log reduction was at least 4.19 versus at least 4.23 for HIV-1; 1.83 versus at least 6.03 for bovine viral diarrhea virus; Volume 55, September 2015 TRANSFUSION 2109
SCHMIDT ET AL.
Fig. 2. (Continued)
2.73 versus at least 5.2 for porcine parvovirus; 5.45 versus at least 9.22 for Escherichia coli; 4.26 versus at least 10.11 for Staphylococcus aureus; and 5.09 versus at least 7.74 for Bacillus subtilis for Mirasol and INTERCEPT PI systems, respectively. Kwon and colleagues46 reported that bacterial growth in PLTs inoculated with low concentrations of S. aureus or B. subtilis was detected only with Mirasol. Based on the time period between blood donation and the implementation of a PI technology, we believe that there is an advantage to APs because the timeconsuming preparation of BCs, the pooling process and the additional centrifugation steps are not necessary. This position is supported by our finding that breakthrough infection was not observed within the study for APs. It is possible that differences in leukoreduction, 2110 TRANSFUSION Volume 55, September 2015
PLT concentration, and volume may have played a role in rendering PI more effective in APs; however, conclusions regarding different residual infectious risks between APs and BC PLTs cannot be drawn from our data. Funk and coworkers47 summarized the hemovigilance data between 1997 and 2010 and described four bacterial fatalities in AP and four in BC PLTs. In addition, Schrezenmeier and colleagues6 examined the risks of bacterial infection between BC PLTs and APs. After screening more than 50,000 PLT concentrates, no significant difference in the infectious risk for both groups of blood components was observed. This study underscores our understanding that the residual transfusiontransmitted risk of bacterial infection is a very complex mechanism with many variables.
EFFICIENCY OF PATHOGEN INACTIVATION
In principle, PI systems offer an alternative to bacterial screening by culture, but neither are 100% effective. Theoretically PI may be effective as the initial bacterial concentration within the PLT concentrate is low, most likely between 10 to 100 CFUs/bag, but there are no experimental data to support this claim. The K. pneumoniae breakthrough occurred with one replicate spiked at the low inoculum of 100 CFUs/bag in BC pooled. Potentially a higher concentration will induce clinical symptoms and will usually exclude donors from blood donation. Nevertheless, all preventive techniques and technologies will have advantages and disadvantages. We have demonstrated that the INTERCEPT PI system has limitations for fast-growing bacteria and spore-forming bacteria. These findings may suggest that treatment must ideally be performed as soon as practical after collection. If this is not operationally practicable, PI-treated PLT components transfused near to the end of their shelf life could be screened for bacteria using rapid detection techniques.48-51 This combination of different procedures would be useful to combine the advantages of various technologies and reduce or prevent potential transmission. However, although the combination of a PI technology together with a rapid bacterial detection method will improve blood safety, the costs for both procedures have to be significantly reduced to enable the implementation of both technologies in blood donor services worldwide. ACKNOWLEDGMENTS
€ ller TH, et al. Bacte6. Schrezenmeier H, Walther-Wenke G, Mu rial contamination of platelet concentrates: results of a prospective multicenter study comparing pooled whole bloodderived platelets and apheresis platelets. Transfusion 2007; 47:644-52. 7. de Korte D, Curvers J, de Kort WL, et al. Effects of skin disinfection method, deviation bag, and bacterial screening on clinical safety of platelet transfusions in the Netherlands. Transfusion 2006;46:476-85. 8. de Korte D, Marcelis JH, Verhoeven AJ, et al. Diversion of first blood volume results in a reduction of bacterial contamination for whole-blood collections. Vox Sang 2002;83:13-6. 9. McDonald CP, Roy A, Mahajan P, et al. Relative values of the interventions of diversion and improved donor-arm disinfection to reduce the bacterial risk from blood transfusion. Vox Sang 2004;86:178-82. 10. Riedel S, Siwek G, Beekmann SE, et al. Comparison of the BACTEC 9240 and BacT/Alert blood culture systems for detection of bacterial contamination in platelet concentrates. J Clin Microbiol 2006;44:2262-4. 11. Espasa M, Salvado M, Vicente E, et al. Evaluation of the VersaTREK system compared to the Bactec MGIT 960 system for first-line drug susceptibility testing of Mycobacterium tuberculosis. J Clin Microbiol 2012;50:488-91. 12. Martinez RM, Bauerle ER, Fang FC, et al. Evaluation of three rapid diagnostic methods for direct identification of microorganisms in positive blood cultures. J Clin Microbiol 2014; 52:2521-9. 13. Dumont LJ, Kleinman S, Murphy JR, et al. Screening of
We thank Ms Borschu for technical assistance by preparing the
single-donor apheresis platelets for bacterial contamination: the PASSPORT study results. Transfusion 2010;50:589-99.
blood components and Dr Thomas Montag (2012) for preparing
14. Fang CT, Chambers LA, Kennedy J, et al. Detection of bacte-
ready-to-use bacterial standards.
rial contamination in apheresis platelet products: American Red Cross experience, 2004. Transfusion 2005;45:1845-52. 15. Ramırez-Arcos S, Kou Y, Mastronardi C, et al. Bacterial
CONFLICT OF INTEREST The authors have disclosed no conflicts of interest.
screening of outdated buffy coat platelet pools using a culture system and a rapid immunoassay. Transfusion 2011;51: 2566-72.
REFERENCES 1. Kal EF. Conflicted caregivers. Am J Psychiatry 2000;157:305. 2. Dodd RY. Emerging pathogens and their implications for the blood supply and transfusion transmitted infections. Br J Haematol 2012;159:135-42. 3. Offergeld R. [Blood - surely but safely!]. Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz 2012;55:9056. 4. Sireis W, Hourfar MK, Seifried E, et al. How to improve blood safety with respect to bacterial contaminations? Vox Sang 2012;103:87-8. 5. Hourfar MK, Jork C, Schottstedt V, et al. Experience of German Red Cross blood donor services with nucleic acid testing: results of screening more than 30 million blood donations for human immunodeficiency virus-1, hepatitis C virus, and hepatitis B virus. Transfusion 2008;48:1558-66.
16. te Boekhorst PA, Beckers EA, Vos MC, et al. Clinical significance of bacteriologic screening in platelet concentrates. Transfusion 2005;45:514-9. 17. Montag T. Strategies of bacteria screening in cellular blood components. Clin Chem Lab Med 2008;46:926-32. 18. Murphy WG, Coakley P. Testing platelet components for bacterial contamination. Transfus Apher Sci 2011;45:69-74. 19. Owusu-Ofori S, Temple J, Sarkodie F, et al. Predonation screening of blood donors with rapid tests: implementation and efficacy of a novel approach to blood safety in resourcepoor settings. Transfusion 2005;45:133-40. 20. Polewczyk A, Janion M, Podlaski R, et al. Clinical manifestations of lead-dependent infective endocarditis: analysis of 414 cases. Eur J Clin Microbiol Infect Dis 2014;33:1601-8. 21. Irsch J, Lin L. Pathogen inactivation of platelet and plasma blood components for transfusion using the INTERCEPT Blood System. Transfus Med Hemother 2011;38:19-31. Volume 55, September 2015 TRANSFUSION 2111
SCHMIDT ET AL.
22. Osselaer JC, Messe N, Hervig T, et al. A prospective observational cohort safety study of 5106 platelet transfusions with
contamination based on unconfirmed positive results. Vox Sang 2014;106:23-30.
components prepared with photochemical pathogen inactivation treatment. Transfusion 2008;48:1061-71. 23. Jocic´ M, Trkuljic´ M, Jovicic´ D, et al. Mirasol PRT system inac-
38. Eder AF, Kennedy JM, Dy BA, et al. Bacterial screening of apheresis platelets and the residual risk of septic transfusion
tivation efficacy evaluated in platelet concentrates by bacteria-contamination model. Vojnosanit Pregl 2011;68:
Transfusion 2007;47:1134-42. 39. Niu MT, Knippen M, Simmons L, et al. Transfusion-transmit-
1041-6. 24. Schubert P, Culibrk B, Coupland D, et al. Riboflavin and ultraviolet light treatment potentiates vasodilator-stimulated phosphoprotein Ser-239 phosphorylation in platelet concentrates during storage. Transfusion 2012;52:397-408. 25. Vanlandingham DL, Keil SD, Horne KM, et al. Photochemical inactivation of chikungunya virus in plasma and platelets using the Mirasol pathogen reduction technology system. Transfusion 2013;53:284-90. 26. Lobo SL, Mehta N, Forgione AG, et al. Use of Theraflex-TMJ topical cream for the treatment of temporomandibular joint and muscle pain. Cranio 2004;22:137-44. 27. Politis C, Kavallierou L, Hantziara S, et al. Quality and safety of fresh-frozen plasma inactivated and leucoreduced with the Theraflex methylene blue system including the Blueflex filter: 5 years’ experience. Vox Sang 2007;92:319-26. € ller TH. UVC irradiation for pathogen reduc28. Seltsam A, Mu tion of platelet concentrates and plasma. Transfus Med Hemother 2011;38:43-54. 29. Dodd R, Kurt Roth W, Ashford P, et al. Transfusion medicine and safety. Biologicals 2009;37:62-70.
reactions: the American Red Cross experience (2004-2006).
ted Klebsiella pneumoniae fatalities, 1995 to 2004. Transfus Med Rev 2006;20:149-57. 40. Sazama K. Bacteria in blood for transfusion. A review. Arch Pathol Lab Med 1994;118:350-65. 41. Jacobs MR, Good CE, Lazarus HM, et al. Relationship between bacterial load, species virulence, and transfusion reaction with transfusion of bacterially contaminated platelets. Clin Infect Dis 2008;46:1214-20. € ller TH. Update on the use of pathogen42. Seltsam A, Mu reduced human plasma and platelet concentrates. Br J Haematol 2013;162:442-54. € rmer M, Arroyo A, Brachert J, et al. Establishment of the 43. Sto first international repository for transfusion-relevant bacteria reference strains: ISBT working party transfusiontransmitted infectious diseases (WP-TTID), subgroup on bacteria. Vox Sang 2012;102:22-31. 44. Goodrich RP, Edrich RA, Li J, et al. The Mirasol PRT system for pathogen reduction of platelets and plasma: an overview of current status and future trends. Transfus Apher Sci 2006; 35:5-17. 45. Seghatchian J, Tolksdorf F. Characteristics of the theraflex
30. Postma MJ, van Hulst M, De Wolf JT, et al. Cost-effectiveness
UV-platelets pathogen inactivation system—an update.
of pathogen inactivation for platelet transfusions in the Netherlands. Transfus Med 2005;15:379-87.
Transfus Apher Sci 2012;46:221-9. 46. Kwon SY, Kim IS, Bae JE, et al. Pathogen inactivation efficacy
31. Rock G. A comparison of methods of pathogen inactivation
of Mirasol PRT System and INTERCEPT Blood System for
of FFP. Vox Sang 2011;100:169-78. 32. Ruane PH, Edrich R, Gampp D, et al. Photochemical inacti-
non-leucoreduced platelet-rich plasma-derived platelets suspended in plasma. Vox Sang 2014;107:254-60.
vation of selected viruses and bacteria in platelet concentrates using riboflavin and light. Transfusion 2004;44:877-85. 33. Stramer SL. Current perspectives in transfusion-transmitted infectious diseases: emerging and re-emerging infections. ISBT Sci Ser 2014;9:30-6. 34. Mejia YX, Mao H, Forde NR, et al. Thermal probing of E. coli RNA polymerase off-pathway mechanisms. J Mol Biol 2008; 382:628-37. 35. Tassou CC, Panagou EZ, Samaras FJ, et al. Temperatureassisted high hydrostatic pressure inactivation of Staphylococcus aureus in a ham model system: evaluation in selective and nonselective medium. J Appl Microbiol 2008;104: 1764-73. 36. Wong TL, Whyte RJ, Graham CG, et al. Growth of enterotoxigenic Staphylococcus aureus in povi masima, a traditional Pacific island food. J Appl Microbiol 2004;97:1185-91. 37. Benjamin RJ, Dy B, Perez J, et al. Bacterial culture of apheresis platelets: a mathematical model of the residual rate of
2112 TRANSFUSION Volume 55, September 2015
47. Funk MB, Lohmann A, Guenay S, et al. Transfusion-transmitted bacterial infections—haemovigilance data of German blood establishments (1997-2010). Transfus Med Hemother 2011;38:266-71. 48. Dreier J, Vollmer T, Kleesiek K. Novel flow cytometry-based screening for bacterial contamination of donor platelet preparations compared with other rapid screening methods. Clin Chem 2009;55:1492-502. € ster B, Daiss C, et al. Extension of platelet shelf 49. Sireis W, Ru life from 4 to 5 days by implementation of a new screening strategy in Germany. Vox Sang 2011;101:191-9. 50. Vollmer T, Dreier J, Schottstedt V, et al. Detection of bacterial contamination in platelet concentrates by a sensitive flow cytometric assay (BactiFlow): a multicentre validation study. Transfus Med 2012;22:262-71. 51. Vollmer T, Knabbe C, Dreier J. Novel flow cytometric screening method for bacterial contamination of red blood cells: a proof-of-principle evaluation. Transfusion 2014;54:900-9.
