Should we attempt to detect bacteria in red blood cells?

In this issue, Chen and colleagues1 describe the use of a culture-based system, the Pall eBDS (Pall Corp., East Hills, NY), for detection of bacteria in red blood cell (RBC) concentrates. The eBDS system, originally developed for detection of bacteria in platelet (PLT) concentrates, involves transfer of approximately 2 to 3 mL of PLT-rich plasma or RBCs into an incubation pouch containing growth-enhancing substances.2, 3 When used for PLT concentrates, the sampling pouch is incubated at 35°C for at least 24 hours. After incubation, the oxygen content in the pouch headspace is measured with an oxygen analyzer, and a level of 15.5 percent or lower is indicative of bacterial growth. The eBDS system is in use for bacterial detection in apheresis and pooled PLT-rich plasma (PRP) PLT concentrates. Sensitivity has been found to be in the order of 1 to 10 colony-forming units (CFUs) per mL for a variety of bacterial species.2, 3 In the study by Chen and colleagues, leukoreduced RBCs prepared by the PRP or buffy-coat method were inoculated with 1 to 15 or 100 CFUs per mL of 12 bacterial species. Bacterial growth and oxygen concentration were measured by sampling eBDS pouches after various periods of storage at 1 to 4°C. Autosterilization occurred in some RBC units; bacterial growth led to an oxygen reading below 14.4 percent after 48 hours of incubation of the sampling pouch, with clear separation of oxygen content levels between contaminated and control units. The authors concluded that the eBDS system was suitable for detection of bacteria in RBC units, with a sensitivity of 1 CFU per mL. Sampling could be performed 1 to 3 days after collection and oxygen tension could be read after 48 to 72 hours of incubation. Other systems of detection of bacteria initially developed for PLT concentrates have been applied to RBCs. The Scansystem (Hemosystem, Marseille, France) is a solid-phase laser cytometry method requiring approximately 90 minutes to perform.4 Sensitivity levels of 1 to 10 CFUs per mL have been shown for detection of bacteria in both PLT and RBC concentrates.4, 5 Nucleic acid–based methods under development may also be applicable to detection of bacteria in both PLT and RBC concentrates.6 As has been seen with PLTs, screening alone is unlikely to completely resolve the problem of septic transfusion reactions. In addition to test characteristics, factors to consider in evaluating possible screening utility include the frequency of RBC septic transfusion reactions, mechanisms of RBC contamination, and impact of recent measures taken to limit and detect bacterial contamination of PLT components. Bacterial contamination of refrigerated whole blood was one of the earliest recognized complications of blood transfusion.7 With the introduction of sterile, single-use plastic collection sets and component therapy, the frequency of septic reactions decreased dramatically. It was subsequently recognized that PLT concentrates, which were stored at room temperature, provided a more hospitable environment for a wide range of organisms, compared to RBCs. However, rare case reports of septic reactions due to RBC contamination continue to be reported.8, 9 In the French hemovigilance system, from 1999 to 2004, the incidence of septic transfusion reactions was 1 in 25,540 apheresis PLT units transfused, compared to 1 in 336,790 RBC units transfused.10 In the UK SHOT program, 25 severe septic reactions were reported from 1999 to 2006; only 2 of these involved RBC concentrates, whereas 23 involved PLT components.11 The Canadian Transfusion Transmitted Injuries Surveillance System (TTISS) hemovigilance data for 2002 and 2003 indicate a 9-fold higher incidence of possible, probable, or definite reactions related to bacterial contamination of PLT concentrates compared to RBCs.12 Finally, from 1998 to 2000, the US BaCon study demonstrated a 15-fold higher fatality rate due to contamination of PLT concentrates compared to RBCs.13 Therefore, data from the mid-1990s until 2006 demonstrate that the majority of septic transfusion reactions and related fatalities are due to contamination of PLT concentrates, and on a per-unit basis, the frequency of both reactions and fatalities is at least 10-fold higher with PLT concentrates compared to RBCs. The majority of septic reactions associated with contaminated RBC concentrates involve organisms associated with bacterial contamination of PLT concentrates, including Gram-positive organisms such as Staphylococcus aureus and Staphylococcus epidermidis and Gram-negative organisms such as Escherichia coli and Serratia liquefaciens.8, 9 Mechanisms of contamination are presumably similar, with organisms originating from the donor's skin, low-level donor bacteremia, or storage container defects. In addition to these organisms, Yersinia enterocolitica has been particularly associated with severe septic reactions due to RBC contamination.14, 15 This organism survives well during cold storage and requires iron for growth, making RBC concentrates a particularly good growth environment. Transient bacteremia may occur during the incubation or recovery phase of gastrointestinal illness or from bacterium in asymptomatic cases. Cases of Y. enterocolitica–associated septicemia seem to occur in clusters. Frequency has varied greatly between countries and over time, with rare sporadic cases reported in France and the United Kingdom and especially high frequency noted in New Zealand in the 1990s and in the United States between 1991 and 1996.9 In vitro spiking experiments suggest that leukoreduction may afford significant protection against low-level contamination with Y. enterocolitica, due to phagocytosis and subsequent removal of intracellular organisms.16 It is uncertain if the introduction of universal leukoreduction in many countries has contributed to decreased reports of Y. enterocolitica–associated septic reactions. Other psychrophilic species of bacteria such as Pseudomonas fluorescens are frequent contaminants in refrigerated environments and have been isolated from disinfectant solutions. These organisms were implicated in one-third of RBC septic reactions reported in the 1970s and 1980s, but have been rarely reported in more recent times.8, 9 Collection practices such as the use of a cool cloth contaminated with the organism have been associated with contamination of RBC units.17 It is possible that improved skin disinfection practices, including use of individually packaged applicators, have contributed to the decreased rate of contamination with these microorganisms. With increased recognition of the importance of bacterial contamination of PLT concentrates, many countries have recently introduced prevention and detection strategies.18 Visual inspection is currently the only measure routinely used specifically to reduce RBC-related septic reactions. However, optimization of skin disinfection and diversion of the first few milliliters of blood collected should both contribute to decreased contamination of RBC concentrates. Culture of whole blood–derived PLTs may result in recall of RBCs associated with a contaminated PLT unit.19 Because the mean age of RBCs transfused is well over 7 days, most RBC units would still be in inventory at the time of detection of a positive PLT culture. Severe septic reactions with RBC concentrates have mainly been associated with units stored for longer periods of time. Studies from Holland and Denmark involving cultures of buffy coat PLT pools have demonstrated that in 40 to 48 percent of cases, at least 1 RBC unit associated with a positive buffy coat pool also had a positive culture.19, 20 Similarly, in pooled cultures performed on PRP PLTs in Hong Kong, 9 RBC units associated with 14 confirmed-positive PLT pools were found to be contaminated with the same organism.21 Increased use of the Pall Acradose system or improved point-of-care testing for whole blood–derived PLTs may similarly result in detection of contaminated RBC cocomponents. Because of the relatively recent adoption of these prevention and detection strategies, it is uncertain if the frequency of septic reactions associated with RBCs reported by various hemovigilance systems will decrease still further. After many years when industry attention appeared to be focused almost exclusively on the development of tests for viral pathogens, it is encouraging that test development is proceeding for bacterial detection in RBCs. The use of the same testing platform for bacterial detection in RBCs and PLTs increases the feasibility of test implementation. However, contamination of RBC concentrates accounts for 10 percent or less of severe septic reactions, and the frequency of these reactions may decline further due to measures implemented to enhance bacterial safety of PLT concentrates. Therefore, it is difficult to estimate the actual contribution that testing of RBCs could make to further reduce risks of septic transfusion reactions.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.