Vox Sanguinis (2014) © 2014 International Society of Blood Transfusion DOI: 10.1111/vox.12214
ORIGINAL PAPER
Routine bacterial screening of platelet concentrates by flow cytometry and its impact on product safety and supply B. M€ uller,1* G. Walther-Wenke,2,* M. Kalus,3 T. Alt,3 J. Bux,1 T. Zeiler4 & V. Schottstedt1 1
GRC-West GRC-West 3 GRC-West 4 GRC-West 2
Blood Blood Blood Blood
Donation Donation Donation Donation
Service, Service, Service, Service,
Central Laboratory Hagen, Hagen, Germany M€unster, Germany Bad Kreuznach, Germany Breitscheid-Ratingen, Germany
Background and Objectives Bacterial contamination represents the major infectious hazard associated with transfusion of platelet concentrates (PCs). As bacterial screening of PCs is not mandatory in Germany, the BactiFlow flow cytometry test has been introduced as a rapid detection method to increase product safety. Materials and Methods During a period of 25 months, a total of 34 631 PCs (26 411 pooled and 8220 apheresis-derived PCs) were tested at the end of day 3 of their shelf life using the BactiFlow system. PCs initially reactive in BactiFlow testing and expired PCs not reactive in BactiFlow on day 3 were also investigated by the BacT/ALERT system and by microbiological cultivation in order to identify the contaminating bacterial species and to confirm reactive BactiFlow results. Results Two hundred and twenty-eight PCs (07%) had an initially reactive result, 24 of them remained reactive in a second test run. Out of these reproducible reactive BactiFlow results, 12 could not be verified by parallel BacT/ALERT culturing, resulting in a confirmed false-positive rate of 003%. The bacterial species were identified as S. aureus, S. epidermidis, S. dysgalactiae ssp. equisimilis and B. cereus. In 10 out of 9017 expired PCs (011%), a confirmed-positive result was obtained in the BacT/ALERT system which had a negative result in the BactiFlow system.
Received: 13 May 2014, revised 29 August 2014, accepted 29 September 2014
Conclusion Testing of PCs by BactiFlow was successfully implemented in our blood donation service and proved sufficient as a rapid and reliable screening method. False reactive results are in an acceptable range since the transfusion of 12 bacterially contaminated PCs was prevented. Key words: bacterial contamination, blood donation testing, platelet concentrates.
Introduction Bacterial contamination of blood components, particularly of platelet concentrates (PCs), and its associated risk of septic reactions is the most frequent infectious complication in transfusion medicine in developed countries [1, 2]. Currently, bacterial contamination represents the Correspondence: Volkmar Schottstedt, DRK-Blutspendedienst West, Zentrallabor Hagen, Feithstrasse 180 – 186, 58097 Hagen, Germany E-mail: [email protected] *These authors contributed equally.
fourth leading cause of death from transfusion after TRALI, haemolytic transfusion reaction due to blood group incompatibilities and transfusion-associated circulatory overload [3, 4]. Numerous measures have considerably reduced the risk of bacterial contamination and their implications. However, a confirmed bacterial contamination rate of 1 in 1412 PCs was reported in a German multicentre study, [5] and international haemovigilance systems still report bacterial-related septic reactions and deaths [3, 6, 7]. Preservation of platelet function requires constant agitation at room temperature (20–24 °C) in gas-permeable bags. These storage 1
2 B. M€ uller et al.
conditions provoke bacterial proliferation even from extremely small bacterial inocula, present immediately after donation, to clinically relevant numbers during the storage period of the PCs [8, 9]. With respect to the fact that platelet-related septic complications in Germany have been observed in particular with older PCs [5, 10– 16], the shelf life of PCs in Germany was reduced in 2008 from 5 to 4 days after the day of production according to Vote 38 of the German Blood Working Party of the Federal Ministry of Health [17], to minimize the risk of a transfusion-associated bacterial sepsis. Nevertheless, in 2011, a fatal septic transfusion reaction occurred due to transfusion of a 4-day-old pooled PC (PPC) contaminated with streptococci group G bacteria [18]. Bacterial screening methods provide the opportunity to reduce the risk of septic transfusion reactions caused by PCs [19]. However, incubation or cultivation methods require time for detection, significantly depending on the bacterial growth kinetics. Therefore, this approach has to be combined with an early sampling strategy. The cultivation method with early sampling, applying the negative-to-date concept of PC release, has been implemented in several transfusion facilities [10], but demonstrated some drawbacks. Since the initial levels of bacteria in PC units are usually extremely low, the detection of bacterial contamination may fail due to sampling error [8, 20]. Several false-negative test results were reported in the Netherlands [16] and in the USA [12, 13, 21]. Studies on prerelease screening with the BacT/ALERT System reported severe or fatal septic reactions due to missed bacteria resulting in false-negative cultures [5, 22–24] The diagnostic sensitivity of BacT/ALERT culturing observed in different studies varies between 26% [25], 33% [26] and 40% (day 1 sampling) [27], mainly depending on the day of sampling and the sample volume. One way to address the problem arising from early sampling errors is the application of rapid detection methods, combined with a late sampling strategy at the end (day 3 or 4) of PC shelf life [4, 11]. Such methods comprise nucleic acid amplification techniques [11, 28– 31], flow cytometric analyses [32–35] or lateral-flow immunoassays [9]. In this study, we implemented the BactiFlow (BF) flow cytometric assay in a centralized routine test setting for bacterial screening of PCs still in the inventory on day 3 of shelf life. We present data out of 25 month of routine testing with a rapid detection method to block and exclude bacterially contaminated PCs, and we provide data demonstrating the application possibilities of this method. Based on the results obtained by BacT/ALERT culturing of PCs with initially or reproducible reactive BF testing and additionally of outdated platelets with negative BF results, we provide data on the capability of the BF assay to detect transfusion-relevant
bacterial titres in PCs. Finally, we discuss our concept of late screening with regard to an extension of platelet shelf life and bacterial safety. Furthermore, advantages and disadvantages of cultivation methods like the BacT/ ALERT test, which is considered as the gold standard for bacterial detection, compared to rapid detection methods like the BF system are discussed.
Materials and methods This study was integrated into the work schedule of five centres of the Blood Transfusion Service West of the German Red Cross (see Appendix) and comprised the implementation of the BactiFlow flow cytometric assay as a routine in-process control late in the shelf life (at the end of day 3) of platelet concentrates. Production of PCs, testing and analysis of data were performed between September 2011 and September 2013.
Platelet concentrate preparation A total of 120 460 PCs [89 814 pooled platelet concentrates (PPCs) and 30 646 apheresis-derived platelet concentrates (APCs)] were collected from volunteer donors and screened according to German guidelines for viral infections and syphilis. Donor arm disinfection was performed with a propanol-based disinfectant (Kodan, Schuelke and Mayr, Norderstedt, Germany). The minimal propanol residence time after the final swab was 30 s. The first proportion (30–40 ml) of both apheresis and whole blood collections was diverted in a diversion pouch. Whole blood donations were stored at room temperature for 10–15 h before centrifugation. After a resting period, four buffy coats were pooled and plasma or TPAS+ solution (Terumo BCT, Leuven, Belgium) was added. Apheresis PCs were prepared with different equipment (Trima Accel, Terumo BCT; Haemonetics MCS+, Haemonetics, Munich, Germany). PCs were stored at 20–24 °C under constant agitation. In Germany, the day of donation is described as day zero and shelf life of PCs is specified with four times 24 h counted from midnight of day zero (Fig. 1). According to German guidelines and in contrast to viral infections and syphilis, there is no need for PCs to undergo an initial testing for bacterial contamination or during their defined shelf life of 4 days.
Sampling and testing by BactiFlow flow cytometric assay Apheresis-derived platelet concentrates and PPCs that were still in the inventory on day 3 at 9 pm were put under quarantine during the period of sampling, transportation and testing of samples. A 10–15 ml aliquot of each © 2014 International Society of Blood Transfusion Vox Sanguinis (2014)
25 Months of BactiFlow screening of PCs 3
Fig. 1 Schematic overview of the time schedule of sample collection, PC expiry according to Vote 38 and extension of PC shelf life after negative BactiFlow (BF) testing (modified from [36]).
PC was transferred into a separate barcoded bag after mixing the content. In case of PPCs, small satellite bags were part of the blood pack system. In case of APCs, a sterile docking device (TSCD; Terumo Europe, Leuven, Belgium) was used to connect a sampling pouch. Samples from all centres were transferred to the central laboratory in Hagen for BF analysis (arrival times of the samples varied between 11 pm and 3 am). As soon as the results of BF testing were available (6 am to 7 am), PCs with a negative BF result were rereleased and labelled for extended storage time of 5 days (Fig. 1). The Paul-Ehrlich-Institut, as the responsible German authority, agreed with our advice of amendment and authorized to extend the storage period of PCs from 4 to 5 days after negative BF test results at the end of day 3. PCs with an initially reactive BF result remained in quarantine and were tested following the algorithm depicted in Fig. 2. The BF analysis was performed in a semi-automated manner by one technical assistant with minor modifications from the previously described protocol by Vollmer et al. [36]. The supernatant was discarded after the centri-
Fig. 2 Test algorithm for PC testing with BactiFlow flow cytometry and sterility testing by BacT/ALERT (dark grey shaded box).
© 2014 International Society of Blood Transfusion Vox Sanguinis (2014)
filtration step, 1 ml buffer ChemSol B24 was added, and samples were placed on the BactiFlow ALS system (bioMerieux, Nu¨rtingen, Germany). Subsequently, the samples were processed strictly following the manufacturer’s instructions for use. In brief, the different BF results were assessed as follows: (1) BF initial testing
ORIGINAL PAPER
Routine bacterial screening of platelet concentrates by flow cytometry and its impact on product safety and supply B. M€ uller,1* G. Walther-Wenke,2,* M. Kalus,3 T. Alt,3 J. Bux,1 T. Zeiler4 & V. Schottstedt1 1
GRC-West GRC-West 3 GRC-West 4 GRC-West 2
Blood Blood Blood Blood
Donation Donation Donation Donation
Service, Service, Service, Service,
Central Laboratory Hagen, Hagen, Germany M€unster, Germany Bad Kreuznach, Germany Breitscheid-Ratingen, Germany
Background and Objectives Bacterial contamination represents the major infectious hazard associated with transfusion of platelet concentrates (PCs). As bacterial screening of PCs is not mandatory in Germany, the BactiFlow flow cytometry test has been introduced as a rapid detection method to increase product safety. Materials and Methods During a period of 25 months, a total of 34 631 PCs (26 411 pooled and 8220 apheresis-derived PCs) were tested at the end of day 3 of their shelf life using the BactiFlow system. PCs initially reactive in BactiFlow testing and expired PCs not reactive in BactiFlow on day 3 were also investigated by the BacT/ALERT system and by microbiological cultivation in order to identify the contaminating bacterial species and to confirm reactive BactiFlow results. Results Two hundred and twenty-eight PCs (07%) had an initially reactive result, 24 of them remained reactive in a second test run. Out of these reproducible reactive BactiFlow results, 12 could not be verified by parallel BacT/ALERT culturing, resulting in a confirmed false-positive rate of 003%. The bacterial species were identified as S. aureus, S. epidermidis, S. dysgalactiae ssp. equisimilis and B. cereus. In 10 out of 9017 expired PCs (011%), a confirmed-positive result was obtained in the BacT/ALERT system which had a negative result in the BactiFlow system.
Received: 13 May 2014, revised 29 August 2014, accepted 29 September 2014
Conclusion Testing of PCs by BactiFlow was successfully implemented in our blood donation service and proved sufficient as a rapid and reliable screening method. False reactive results are in an acceptable range since the transfusion of 12 bacterially contaminated PCs was prevented. Key words: bacterial contamination, blood donation testing, platelet concentrates.
Introduction Bacterial contamination of blood components, particularly of platelet concentrates (PCs), and its associated risk of septic reactions is the most frequent infectious complication in transfusion medicine in developed countries [1, 2]. Currently, bacterial contamination represents the Correspondence: Volkmar Schottstedt, DRK-Blutspendedienst West, Zentrallabor Hagen, Feithstrasse 180 – 186, 58097 Hagen, Germany E-mail: [email protected] *These authors contributed equally.
fourth leading cause of death from transfusion after TRALI, haemolytic transfusion reaction due to blood group incompatibilities and transfusion-associated circulatory overload [3, 4]. Numerous measures have considerably reduced the risk of bacterial contamination and their implications. However, a confirmed bacterial contamination rate of 1 in 1412 PCs was reported in a German multicentre study, [5] and international haemovigilance systems still report bacterial-related septic reactions and deaths [3, 6, 7]. Preservation of platelet function requires constant agitation at room temperature (20–24 °C) in gas-permeable bags. These storage 1
2 B. M€ uller et al.
conditions provoke bacterial proliferation even from extremely small bacterial inocula, present immediately after donation, to clinically relevant numbers during the storage period of the PCs [8, 9]. With respect to the fact that platelet-related septic complications in Germany have been observed in particular with older PCs [5, 10– 16], the shelf life of PCs in Germany was reduced in 2008 from 5 to 4 days after the day of production according to Vote 38 of the German Blood Working Party of the Federal Ministry of Health [17], to minimize the risk of a transfusion-associated bacterial sepsis. Nevertheless, in 2011, a fatal septic transfusion reaction occurred due to transfusion of a 4-day-old pooled PC (PPC) contaminated with streptococci group G bacteria [18]. Bacterial screening methods provide the opportunity to reduce the risk of septic transfusion reactions caused by PCs [19]. However, incubation or cultivation methods require time for detection, significantly depending on the bacterial growth kinetics. Therefore, this approach has to be combined with an early sampling strategy. The cultivation method with early sampling, applying the negative-to-date concept of PC release, has been implemented in several transfusion facilities [10], but demonstrated some drawbacks. Since the initial levels of bacteria in PC units are usually extremely low, the detection of bacterial contamination may fail due to sampling error [8, 20]. Several false-negative test results were reported in the Netherlands [16] and in the USA [12, 13, 21]. Studies on prerelease screening with the BacT/ALERT System reported severe or fatal septic reactions due to missed bacteria resulting in false-negative cultures [5, 22–24] The diagnostic sensitivity of BacT/ALERT culturing observed in different studies varies between 26% [25], 33% [26] and 40% (day 1 sampling) [27], mainly depending on the day of sampling and the sample volume. One way to address the problem arising from early sampling errors is the application of rapid detection methods, combined with a late sampling strategy at the end (day 3 or 4) of PC shelf life [4, 11]. Such methods comprise nucleic acid amplification techniques [11, 28– 31], flow cytometric analyses [32–35] or lateral-flow immunoassays [9]. In this study, we implemented the BactiFlow (BF) flow cytometric assay in a centralized routine test setting for bacterial screening of PCs still in the inventory on day 3 of shelf life. We present data out of 25 month of routine testing with a rapid detection method to block and exclude bacterially contaminated PCs, and we provide data demonstrating the application possibilities of this method. Based on the results obtained by BacT/ALERT culturing of PCs with initially or reproducible reactive BF testing and additionally of outdated platelets with negative BF results, we provide data on the capability of the BF assay to detect transfusion-relevant
bacterial titres in PCs. Finally, we discuss our concept of late screening with regard to an extension of platelet shelf life and bacterial safety. Furthermore, advantages and disadvantages of cultivation methods like the BacT/ ALERT test, which is considered as the gold standard for bacterial detection, compared to rapid detection methods like the BF system are discussed.
Materials and methods This study was integrated into the work schedule of five centres of the Blood Transfusion Service West of the German Red Cross (see Appendix) and comprised the implementation of the BactiFlow flow cytometric assay as a routine in-process control late in the shelf life (at the end of day 3) of platelet concentrates. Production of PCs, testing and analysis of data were performed between September 2011 and September 2013.
Platelet concentrate preparation A total of 120 460 PCs [89 814 pooled platelet concentrates (PPCs) and 30 646 apheresis-derived platelet concentrates (APCs)] were collected from volunteer donors and screened according to German guidelines for viral infections and syphilis. Donor arm disinfection was performed with a propanol-based disinfectant (Kodan, Schuelke and Mayr, Norderstedt, Germany). The minimal propanol residence time after the final swab was 30 s. The first proportion (30–40 ml) of both apheresis and whole blood collections was diverted in a diversion pouch. Whole blood donations were stored at room temperature for 10–15 h before centrifugation. After a resting period, four buffy coats were pooled and plasma or TPAS+ solution (Terumo BCT, Leuven, Belgium) was added. Apheresis PCs were prepared with different equipment (Trima Accel, Terumo BCT; Haemonetics MCS+, Haemonetics, Munich, Germany). PCs were stored at 20–24 °C under constant agitation. In Germany, the day of donation is described as day zero and shelf life of PCs is specified with four times 24 h counted from midnight of day zero (Fig. 1). According to German guidelines and in contrast to viral infections and syphilis, there is no need for PCs to undergo an initial testing for bacterial contamination or during their defined shelf life of 4 days.
Sampling and testing by BactiFlow flow cytometric assay Apheresis-derived platelet concentrates and PPCs that were still in the inventory on day 3 at 9 pm were put under quarantine during the period of sampling, transportation and testing of samples. A 10–15 ml aliquot of each © 2014 International Society of Blood Transfusion Vox Sanguinis (2014)
25 Months of BactiFlow screening of PCs 3
Fig. 1 Schematic overview of the time schedule of sample collection, PC expiry according to Vote 38 and extension of PC shelf life after negative BactiFlow (BF) testing (modified from [36]).
PC was transferred into a separate barcoded bag after mixing the content. In case of PPCs, small satellite bags were part of the blood pack system. In case of APCs, a sterile docking device (TSCD; Terumo Europe, Leuven, Belgium) was used to connect a sampling pouch. Samples from all centres were transferred to the central laboratory in Hagen for BF analysis (arrival times of the samples varied between 11 pm and 3 am). As soon as the results of BF testing were available (6 am to 7 am), PCs with a negative BF result were rereleased and labelled for extended storage time of 5 days (Fig. 1). The Paul-Ehrlich-Institut, as the responsible German authority, agreed with our advice of amendment and authorized to extend the storage period of PCs from 4 to 5 days after negative BF test results at the end of day 3. PCs with an initially reactive BF result remained in quarantine and were tested following the algorithm depicted in Fig. 2. The BF analysis was performed in a semi-automated manner by one technical assistant with minor modifications from the previously described protocol by Vollmer et al. [36]. The supernatant was discarded after the centri-
Fig. 2 Test algorithm for PC testing with BactiFlow flow cytometry and sterility testing by BacT/ALERT (dark grey shaded box).
© 2014 International Society of Blood Transfusion Vox Sanguinis (2014)
filtration step, 1 ml buffer ChemSol B24 was added, and samples were placed on the BactiFlow ALS system (bioMerieux, Nu¨rtingen, Germany). Subsequently, the samples were processed strictly following the manufacturer’s instructions for use. In brief, the different BF results were assessed as follows: (1) BF initial testing
