Vox Sanguinis (2014) © 2014 International Society of Blood Transfusion DOI: 10.1111/vox.12158

ORIGINAL PAPER

Pathogen inactivation efficacy of Mirasol PRT System and Intercept Blood System for non-leucoreduced platelet-rich plasma-derived platelets suspended in plasma S. Y. Kwon,1,* I. S. Kim,2,* J. E. Bae,2 J. W. Kang,1 Y. J. Cho,1 N. S. Cho3 & S. W. Lee4 1

Blood Transfusion Research Institute, Korean Red Cross, Seoul, Korea Department of Biological Sciences and Biotechnology, Hannam University, Daejeon, Korea 3 Blood Service Headquarters, Korean Red Cross, Seoul, Korea 4 Division of Vaccine Research, Center for Infectious Disease, Korea National Institute of Health, Korea Center for Disease Control & Prevention, Osong, Korea 2

Background and Objectives This study was conducted to evaluate the efficacy of pathogen inactivation (PI) in non-leucoreduced platelet-rich plasma-derived platelets suspended in plasma using the Mirasol PRT System and the Intercept Blood System. Methods Platelets were pooled using the Acrodose PL system and separated into two aliquots for Mirasol and Intercept treatment. Four replicates of each viral strain were used for the evaluation. For bacteria, both low-titre (45–152 CFU/ unit) inoculation and high-titre (7
34–10
18 log CFU/unit) inoculation with two replicates for each bacterial strain were used. Platelets with non-detectable bacterial growth and platelets inoculated with a low titre were stored for 5 days, and culture was performed with the BacT/ALERT system. Results The inactivation efficacy expressed as log reduction for Mirasol and Intercept systems for viruses was as follows: human immunodeficiency virus 1, ≥4
19 vs. ≥4
23; bovine viral diarrhoea virus, 1
83 vs. ≥6
03; pseudorabies virus, 2
73 vs. ≥5
20; hepatitis A virus, 0
62 vs. 0
76; and porcine parvovirus, 0
28 vs. 0
38. The inactivation efficacy for bacteria was as follows: Escherichia coli, 5
45 vs. ≥9
22; Staphylococcus aureus, 4
26 vs. ≥10
11; and Bacillus subtilis, 5
09 vs. ≥7
74. Postinactivation bacterial growth in platelets inoculated with a low titre of S. aureus or B. subtilis was detected only with Mirasol.

Received: 12 November 2013, revised 14 March 2014, accepted 13 April 2014

Conclusion Pathogen inactivation efficacy of Intercept for enveloped viruses was found to be satisfactory. Mirasol showed satisfactory inactivation efficacy for HIV-1 only. The two selected non-enveloped viruses were not inactivated by both systems. Inactivation efficacy of Intercept was more robust for all bacteria tested at high or low titres. Key words: inactivation efficacy, Intercept Blood System, Mirasol PRT System, pathogen inactivation, platelet concentrates.

Correspondence: So-Yong Kwon, Blood Transfusion Research Institute, Korean Red Cross, 764, Sanggye 6.7-dong, Nowon-gu, Seoul 139-831, Korea E-mail: [email protected] Work was performed at the Blood Transfusion Research Institute, Korean Red Cross and the Hannam University. *These authors contributed equally to this work.

Introduction With the application of enhanced donor selection criteria and introduction of nucleic acid testing (NAT), the risk of transfusion-transmitted infections (TTI) has dramatically decreased. However, as no system is 100% perfect, breakthrough cases occur [1] and new pathogens threatening

1

2 S. Y. Kwon et al.

the safety of blood transfusion are continuously identified [2]. Moreover, bacterial contamination still remains the highest risk in blood transfusion [3]. NAT for HCV and HIV has been introduced in February 2005 in Korea, and in April 2009, screening for human T-lymphotropic virus-1/2 (HTLV-1/2) was implemented. In June 2012, HBV was included in NAT and the test method was switched from pool testing to individual testing. With improvement in testing systems and introduction of new test items, the safety of our national blood supply has markedly increased. However, known pathogens that are not screened for and new emerging pathogens represent a constant threat. In Korea, majority of platelets transfused are whole blood-derived platelets that are prepared by the platelet-rich plasma (PRP) method. These are not pooled and supplied to hospitals as individual units; therefore, culture for bacteria that require a minimum amount of sample of about 5 ml is not feasible. The limitations of our current system and the challenges that we face were evidenced by recently identified TTI cases. One case of sepsis due to transfusion of a unit of apheresis platelet that was contaminated with bacteria was identified in 2011 [4], and we also experienced two cases of hepatitis A virus (HAV) transmission in 2008 and 2011 [5]. Pathogen inactivation (PI) technologies have been introduced as a proactive approach to overcome these problems. So far two systems, the Mirasol PRT System (Terumo BCT Inc., Lakewood, CO, USA) and the Intercept Blood System (Cerus Corporation, Concord, CA, USA), are in clinical use for platelets mostly in European countries

[6–8]. Several studies about their inactivation efficacy have been reported [9–14], but except one recent study comparing inactivation efficacy of both systems for dengue and chikungunya viruses [15], no side-by-side comparison results have been published yet. This side-by-side comparison study was conducted to evaluate the efficacy of the two PI systems for the selected viruses and bacteria in non-leucoreduced PRP-platelets (PRP-PLTs) suspended in plasma.

Materials and methods General study design A schematic overview of the study design has been provided in Fig. 1. About 12 units of PRP-PLTs suspended in plasma were pooled using the Acrodose PL system (Pall Corporation, East Hills, NY, USA) to make a platelet pool of about 504 ml. The platelet pool was spiked with about 26 ml of the viral or bacterial suspension. Immediately after spiking, a 10 ml sample was taken and used to determine the total viral or bacterial load in the spiked platelet pool. About 225 ml aliquot of the spiked platelet pool was transferred to the Mirasol illumination bag and about 265 ml aliquot of the same spiked platelet pool was transferred to the Intercept illumination bag for inactivation, so that acceptable limits for product specifications defined by both manufacturers were met. Pathogen inactivation using the Mirasol PRT System and the Intercept Blood System was performed according to the

Fig. 1 Schematic overview of the study design. *The control spiked sample was kept only for the viral inactivation study.

© 2014 International Society of Blood Transfusion Vox Sanguinis (2014)

PI efficacy for PRP platelets 3

manufacturer’s instructions. Postinactivation total viral or bacterial load in the platelet pool was determined immediately after the inactivation procedure. During the viral inactivation study, a 10 ml aliquot of the spiked platelet pool was kept as a control spiked sample for the length of the process time and then the total viral load was determined again to ensure that the test virus was not inactivated by plasma components.

Viral inactivation study Based on the recommendations of Epstein and Vostal [16], five model viruses were selected (Table 1). For the propagation and titration of HIV-1 (HBX2 strain), bovine viral diarrhoea virus (BVDV, ATCC VR 534), pseudorabies virus (PRV, YS-400 strain), HAV (ATCC VT 1402) and porcine parvovirus (PPV, ATCC VR 742) C8166-45 cells (ECACC), Madin-Darby bovine kidney (MDBK) cells (ATCC CRL-22), Vero cells (ATCC CCL-81), FRhK-4 cells (ATCC CRL-1688) and minipig kidney (MPK) cells (ATCC CCL-166) were used, respectively. C8166 cells were grown in Roswell Park Memorial Institute 1640 medium containing 2% fetal bovine serum (FBS) and L-glutamine. Other cells were grown in high-glucose Dulbecco’s modified Eagle’s medium (HG DMEM) containing 2% FBS. An aliquot of each sample from the virus inactivation studies and appropriate control were titrated immediately after collection in sevenfold serial dilutions to the endpoint using a quantal 50% tissue culture infectious dose (TCID50) assay [17]. For the titration of HIV, suspensions of C8166 cells in 96-well culture plates were infected using at least eight 0
1 ml replicates of culture medium. The plates were incubated at 35°C for approximately 1 h, and the wells were fed with 0
1 ml of tissue culture medium. Wells were examined for cytopathic effects (CPE) and syncytial formation approximately 14 days later. Formation of a syncytium, a multinucleate cell resulting from multiple cell fusions of uninuclear cells, was observed by microscopy. For the titration of BVDV, PRV, HAV and PPV, indicator cell monolayers in 24-well culture plates were infected using at least eight replicates of 0
25 ml of the appropriate dilution of each sample or the Table 1 Model viruses used for the evaluation of pathogen inactivation efficacy

positive control. Negative control wells were mockinfected using at least eight replicates of 0
25 ml of culture medium. The plates were then incubated at 35°C for approximately 1 h, and the wells were fed with 1 ml of tissue culture medium. The wells were then examined for CPE after 7–14 days of incubation. As part of the virus validation protocol, cytotoxicity and interference tests were performed as previously described [18]. Cytotoxicity tests were performed on samples generated for virus titration in the virus spiking experiments to control for any possible cytotoxic effects on indicator cells that would interfere with virus titration. Interference tests were performed to determine whether or not the starting materials for the virus spiking studies exerted an inhibitory effect on the ability of the cell lines to permit detection of the virus. The virus log reduction factor was defined as the log10 of the ratio of the virus loads in the spiked preinactivation and postinactivation materials, in accordance with the World Health Organization (WHO) guidelines [19]. When viral infectivity was not detected after the inactivation procedure, the virus titre was calculated using a theoretical minimum detectable level of infectious virus, with a 95% upper confidence level. All the virus inactivation experiments were carried out for each virus using four different platelet pools, and mean values are given.

Bacterial inactivation study Based on the recommendations of Epstein and Vostal [16], three bacteria species were selected (Table 2). Culture stocks of each bacterial strain in the stationary phase were obtained to inoculate the platelets. Preinactivation titre for the high-titre evaluation ranged from 7
34 to 10
18 log CFU/unit. For the low-titre evaluation, preinactivation titre ranged from 45 to 152 CFU/unit. Two platelet pools were used for each bacterial strain. The log reduction factor was calculated according to the WHO guidelines [19]. When bacterial growth was not detected after the inactivation procedure, the log reduction factor was expressed as greater than or equal to the preinactivation titre. Platelets with a high bacterial titre with

Model virus

Representative virus

Genome

Envelope

Inactivation resistance

HIV-1 (HBX2 strain) BVDV (ATCC VR 534) PRV (YS-400 strain) HAV (ATCC VR 1402) PPV (ATCC VR 742)

HIV/HTLV HCV/WNV CMV HAV B19

RNA RNA DNA RNA DNA

Yes Yes Yes No No

Low Medium Low-medium High Very high

BVDV, bovine viral diarrhoea virus; PRV, pseudorabies virus; PPV, porcine parvovirus; HAV, hepatitis A virus.

© 2014 International Society of Blood Transfusion Vox Sanguinis (2014)

4 S. Y. Kwon et al.

utilized for HIV, BVDV and PRV. Average log reduction factors achieved were ≥4
23 for HIV-1, ≥6
03 for BVDV and ≥5
20 for PRV. However, it was not effective for the inactivation of HAV and PPV (Table 3).

Table 2 Bacterial species used for the evaluation of pathogen inactivation efficacy

Bacteria

Group

Spore formation

Inactivation resistance

Escherichia coli (ATCC 25922) Staphylococcus aureus (ATCC 25923) Bacillus subtilis (KCTC 1102)

Gram-negative

No

Low

Gram-positive

No

Low

Gram-positive

Yes

High

Bacterial inactivation efficacy Pathogen inactivation efficacy expressed as the average log reduction factor of the Mirasol PRT System and the Intercept Blood System for high-titre bacteria was as follows: Escherichia coli, 5
45 vs. ≥9
22; Staphylococcus aureus, 4
26 vs. ≥10
11; and Bacillus subtilis, 5
09 vs. ≥7
74 (Table 4). Bacterial growth was not detected in cultures performed on day 5 after storage of platelets with a high bacterial titre with non-detectable bacterial growth after inactivation. Postinactivation bacterial growth in platelets inoculated with a low bacterial titre was not detected in platelets treated with the Intercept Blood System (Table 5). Bacterial growth was detected in platelets with a low titre of S. aureus or B. subtilis and treated with the Mirasol PRT System.

non-detectable bacterial growth after inactivation and platelets with a low bacterial titre were stored for 5 days after inactivation in platelet agitators. On day five of storage, 5 ml samples were cultured for 5 days using the BacT/ALERT 3D system (Biomerieux Inc., Durham, NC, USA). Aerobic (BacT/ALERT SA; Biomerieux Inc.) and anaerobic (BacT/ALERT SN; Biomerieux Inc.) culture bottles were used.

Results Discussion Viral inactivation efficacy

As viral titres in the window period can reach levels of 108–1010 geq per ml, it is suggested that PI treatment should reduce the pathogen load by 6–10 logs to claim efficacy [16]. Since some viral particles might not be able to replicate for various reasons as described by Goodrich et al. [20], a viral genome does not necessarily constitute an infectious particle and hence the viral titre does not directly correlate with infectivity. Therefore, the level of pathogen reduction required for preventing disease transmission depends on the viral agent and the stage of the infection. In addition, as blood units with a high viral titre are most likely to be screened out by NAT, a reduc-

With an average log reduction factor of ≥4
19, the Mirasol PRT System inactivated HIV-1 to levels exceeding the limits of detection by the assay utilized. It was moderately effective for the inactivation of BVDV and PRV. Average log reduction factors achieved were 1
83 for BVDV and 2
73 for PRV. However, it was not efficient in the inactivation of HAV and PPV (Table 3, Supporting Information with data from each individual experiment has been provided). The Intercept Blood System showed an inactivation efficacy that exceeded the limits of detection by the assay

Table 3 Inactivation efficacy of Mirasol PRT System and Intercept Blood System for selected viruses Mirasol PRT System

Virus

Preinactivation total viral loada

HIV-1 BVDV PRV HAV PPV

7
44 8
97 8
22 8
47 7
87

– – – – –

0
04 0
27 0
09 0
14 0
16

Intercept Blood System Postinactivation total viral loada

Log reduction factorb

Preinactivation total viral load

≤3
25c 7
14 – 5
49 – 7
85 – 7
59 –

≥4
19 1
83 2
73 0
62 0
28

7
47 9
00 8
19 8
51 7
89

0
26 0
15 0
15 0
07

– – – – –

0
04 0
34 0
07 0
18 0
16

– – – – –

0
05 0
27 0
19 0
14 0
16

Postinactivation total viral load

Log reduction factor

≤3
24c ≤2
97c ≤2
99c 7
75 – 0
15 7
52 – 0
19

≥4
23 ≥6
03 ≥5
20 0
76 0
38

– – – – –

0
05 0
26 0
18 0
21 0
18

Four replicates were tested for each virus and results are reported as mean – SD. Viral load is expressed as the log TCID50 viral content of the treated platelet product. b Log reduction factor was calculated according to the WHO guidelines [18]. c No infectious virus was detected. These values were calculated using a theoretical minimum detectable level of infectious virus with a 95% confidence level. a

© 2014 International Society of Blood Transfusion Vox Sanguinis (2014)

PI efficacy for PRP platelets 5

Table 4 Inactivation efficacy of Mirasol PRT System and Intercept Blood System for selected bacteria with a high-titre inoculation Mirasol PRT System

Intercept Blood System

Preinactivation

Postinactivation

Log

Mean log

Preinactivation

Postinactivation

Log

Mean

total bacterial

total bacterial

reduction

reduction

total bacterial

total bacterial

reduction

log reduction

loada

loada

factorb

factor

load

load

factor

factor ≥9
22

Bacteria

Replicate

Escherichia

1

8
88

4
73

4
15

2

9
48

2
73

6
75

coli Staphylococcus

1

9
99

5
82

4
17

2

10
14

5
80

4
34

Bacillus

1

7
34

ND

≥7
34

subtilis

2

8
07

5
24

2
83

aureus

5
45 4
26 5
09

8
91

NDc

≥8
91

9
52

ND

≥9
52

10
03

ND

≥10
03

10
18

ND

≥10
18

7
38

ND

≥7
38

8
10

ND

≥8
10

≥10
11 ≥7
74

a

Bacterial load is expressed as the log CFU content of the treated platelet product. Log reduction factor was calculated according to the WHO guidelines [18]. c Bacterial growth was not detected. When bacterial growth was not detected after the inactivation procedure, the log reduction factor was expressed as greater than or equal to the preinactivation titre. b

Table 5 Inactivation efficacy of Mirasol PRT System and Intercept Blood System for selected bacteria with a low-titre inoculation Mirasol PRT System

Intercept Blood System

Bacteria

Replicate

Preinactivation total bacterial loada

Postinactivation bacterial growthb

Preinactivation total bacterial load

Postinactivation bacterial growth

Escherichia coli

1 2 1 2 1 2

82 84 152 145 89 46

NDc ND Dd D D D

85 90 147 142 94 45

ND ND ND ND ND ND

Staphylococcus aureus Bacillus subtilis

a

Bacterial load was expressed as the CFU content of the treated platelet product. Culture was performed on the BacT/Alert 3D system. c Bacterial growth was not detected. d Bacterial growth was detected. b

tion of 5 logs for B19. In the same study, Lin et al. [11] reported a zero reduction for PPV. Although PPV is used as a model virus for B19, PPV might not be an appropriate model virus for B19 as other disparate inactivation results have also been reported [23]. The reason for this was explained as a difference in the property of the viral capsid between the two viruses leading to a different sensitivity to inactivation [23]. The Mirasol PRT System had a similar inactivation efficacy for HIV-1 as the Intercept Blood System.

6 S. Y. Kwon et al.

HIV-1 was completely inactivated to undetectable levels by both systems. Ruane et al. [10] reported a 5
2 log reduction for WNV with the Mirasol PRT System. As BVDV is a model virus for WNV, similar levels of inactivation should be expected. However, in this study, BVDV was inactivated by only 1
83 log. A study comparing the inactivation efficacy of methylene blue plus light treatment and UVC irradiation for cell culture-derived HCV and BVDV, a model virus for HCV, also reported a lower susceptibility of BVDV to the PI procedures compared to HCV [24]. This again supports the notion that a model virus may not always reflect susceptibility of the corresponding virus for a certain inactivation procedure. The inactivation efficacy for PRV was similar to the result of a previous study [25]. As the log reduction factor for PRV in this study was