By continuing to browse this site you agree to us using cookies as described in About Cookies
Notice: Wiley Online Library will be unavailable on Saturday 7th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 08.00 EDT / 13.00 BST / 17:30 IST / 20.00 SGT and Sunday 8th Oct from 03.00 EDT / 08:00 BST / 12:30 IST / 15.00 SGT to 06.00 EDT / 11.00 BST / 15:30 IST / 18.00 SGT for essential maintenance. Apologies for the inconvenience.
This study was sponsored by the Military Major Special Project (AWS11J007-08).
A flow-based treatment device using riboflavin and ultraviolet (UV) light was developed to inactivate viruses in fresh-frozen plasma (FFP). The objective of this study was to evaluate the in vitro effectiveness of virus inactivation and changes in protein quality in FFP treated with this device.
Study Design and Methods
FFP-contaminating viruses were treated with riboflavin and UV light using a one-pass linear flow device. The infectivity of viruses was measured using established biologic assays. Real-time polymerase chain reaction (PCR) was performed to detect damage to viral nucleotides after treatment. Treated plasma was analyzed using standard coagulation assays.
FFP treated at the UV dose of 3.6 J/cm2 (J) exhibited a mean reduction of virus titer of more than 4 logs. The effectiveness increased significantly at higher doses. Real-time PCR showed that the cycle threshold values for both complete inactivation and virus recultivation were higher than that of the untreated sample. At doses of 3.6, 5.4, and 7.2 J, the protein recovery rates were 60.2 ± 8.6, 46.6 ± 9.4, and 28.0 ± 1.0% for fibrinogen; 67.0 ± 3.1, 57.3 ± 8.0, and 49.2 ± 3.8% for Factor VIII; 93.6 ± 2.8, 89.6 ± 6.1, and 86.5 ± 5.3% for antithrombin-III; and 72.1 ± 5.6, 59.8 ± 14.2, and 49.2 ± 8.4% for Protein C, respectively.
The effectiveness of virus inactivation was enhanced, but total activity of plasma factors was reduced, in a UV dose-dependent manner.
Generally, physicians perform blood transfusions in the clinic using single units of plasma. In these procedures, plasma derived from different donors has variable functional factors. Plasma with undefined levels of factors and functional doses could influence therapeutic effectiveness. To preserve the functional stability of plasma for transfusion, several batches of plasma might be prepared as pooled plasma so that these different plasma units could achieve functional complementarity. Accordingly, a virus inactivation procedure would be needed to prepare the pooled plasma to mitigate the risk of failing to detect infection in single plasma samples. The solvent/detergent (S/D) method has generally been applied to treat pooled plasma in the past, but this method failed to inactivate nonenveloped virus.
Compared with currently available methods of S/D, methylene blue, and amotosalen additives for pathogen inactivation in plasma, a treatment method using a combination of riboflavin and broadband ultraviolet (UV) has been developed as a novel method for virus inactivation in plasma, platelets, and whole blood.[3, 4] It has the safety advantages of riboflavin addition and allows a high rate of retention of plasma proteins. Furthermore, the nucleic acid damage induced by riboflavin was irreversible because of the guanine base modifications, in contrast to the reversible damage by UV light treatment alone.[5, 6] Studies have also shown that an extensive range of transfusion-transmitted viruses and bacteria were obviously reduced in fresh-frozen plasma (FFP) with treatment by riboflavin photochemical method.[7, 8] Moreover, this method had a remarkable ability to inactivate certain parasites.[9-11] Based on this method, the Mirasol system was developed by CaridianBCT Biotechnologies (Lakewood, CO) and has been used for the treatment of various blood components; its effectiveness has been validated by several institutions.[12-15]
The Mirasol system was developed to treat single units of plasma; it is not suitable for the demands of pooled or large-scale plasma treatment. It cannot be used to accomplish coinstantaneous treatment by flow, especially for the treatment of pooled plasma such as the virus-inactivated universal plasma developed by our group. Given the current status of the use of this method, we have made reasonable modification to the design of a flow-based treatment device for FFP based on the special mechanism and technologic characteristics of riboflavin in combination with UV. This newly established technique is designed to balance both safety, which involves virus inactivation, and efficacy, which denotes the loss of proteins after treatment. Therefore, we performed this study to evaluate in vitro the effectiveness of virus inactivation and retention of plasma proteins in FFP using the modified flow-based treatment device to determine an optimal treatment procedure.
Materials and Methods
Cell culture and viruses
The indicator viruses vesicular stomatitis virus (VSV), pseudorabies virus (PRV), Sindbis virus (Sindbis), porcine parvovirus (PPV), and encephalomyocarditis virus (EMCV), and the susceptible cells African green monkey kidney-E6 cells for VSV, porcine kidney-15 cells for PRV and PPV, baby hamster Syrian kidney cells for EMCV, and African green monkey kidney-E6 cells for Sindbis, were supplied by The Institute of Transfusion Medicine, The Academy of Military Medical Sciences. Cells were cultured as a monolayer in Dulbecco's modified Eagle's medium (Hyclone, Logan, UT) containing penicillin and streptomycin supplemented with 5% to 10% inactivated newborn calf serum (Gibco, Grand Island, NY) and were grown in an incubator at 37°C with 5% CO2. Viruses were propagated in respective susceptible cells and could be used for experiments when the virus titer was greater than 6 logs after propagation.
Preparation of plasma and riboflavin solutions
Plasma units were prepared from whole blood by 4000 × g centrifugation at 22°C for 10 minutes using a centrifuge (Biofuge Primo R, Heraeus, Waltham, MA). The separated plasma was stored at −30°C or cooler as FFP. Before use, plasma was thawed in a shaking water bath (MultiTemp III, GE Healthcare Europe GmbH, Freiburg, Germany) at 37°C. For virus inactivation, FFP was further heated to 56°C for 30 minutes to inactivate any neutralizing antibodies. Riboflavin with the purity of more than 98% was supplied by Sigma Corporation (St Louis, MO) and dissolved in 0.9% sodium chloride solution with pH 4.5 to 5 adjusted by acidification with hydrochloric acid (1 mol/L). The solution was then agitated and continuously heated at 60°C for 5 minutes, and a final solution with a riboflavin concentrate of 500 μmol/L was stored at 4°C and used as soon as possible, at longest within a week. All procedures were carried out in an opaque foil pouch to protect from ambient light.
The microenvironment of plasma contaminating viruses and the addition of riboflavin
Plasma units were spiked 1:10 with virus suspension and then treated. One mL of virus suspension was filtered by using a 0.45-μm filter (Pall, Port Washington, NY) and was added to 10 mL of FFP after thawing, treating, and filtering, thereby creating a virus-contaminated microenvironment. Finally, 1.5 mL of riboflavin solution was filtered with a 0.45-μm filter and added so that riboflavin was diluted to an approximate concentration of 60 μmol/L in plasma for virus inactivation. The three components were mixed and incubated for less than 10 minutes for device treatment (DT).
Establishment of a virus inactivation system
The device for riboflavin and broadband UV was assembled using a UV illumination and a sample-flow tube. The illumination had two parallel rows of UV lamps—one row was positioned above and one row was positioned below. Four 313-nm UVB and 365-nm UVA lamps were arranged by turns for every row. The S-shape quartz sample-flow tube was located parallel to the middle of two lamp rows and the interval to each lamp row was 1 cm. Moreover, a ventilator was arranged to cool both the lamps and the sample flow tube. In the entrance, a peristaltic pump was installed to send impulses to get the plasma to flow; in the exit, an electromagnetic valve was installed to control plasma switching to outflow or off. Otherwise, the sample flow in the quartz tube could be observed using the visible window, and a door on the outside made it convenient to disassemble lamps and the quartz tube (Fig. 1).
Measurement of the device variables
The UV spectrum of UVA- and UVB-combined irradiation was measured by using a spectrometer (iHR550, Horiba Jobin Yvon, Longjumeau, France) and the intensity of combined irradiation was measured directly by a UV light meter (YK-35UV, Lutron, Taiwan). Depending on the average intensity of UV light and the treatment time, the UV dose was calculated based on the formula:
where D is the UV dose (mJ/cm2), I is the average intensity of UV (mW/cm2), and t is the irradiation time (sec).
The sample-flow tube was made of quartz and the transmittance of UV through the quartz tube was measured using a UV light meter. The diameter or thickness of the quartz tube was measured by vernier caliper.
Virus inactivation with DT
To determine the optimal treatment dose, four different treatment times were selected for each virus inactivation experiment. Different treatment times for sample irradiation were regulated by increasing or decreasing the velocity of plasma flow in the sample-flow tube using a peristaltic pump. With a sterile connection, the samples were ejected into the sample tube and treated by flow at a specified velocity for each respective dose. At the exit, effluent samples were collected as virus-inactivated plasma. Before and after treatment, the flow was sampled in Eppendorf tubes and stored at −70°C or cooler until they were assayed. Treatments at four UV doses were carried out for each indicator virus experiment.
Virus titration and determination of the effectiveness of inactivation
The samples untreated and treated with DT at different doses were drawn and tested. Virus titers were determined by endpoint titration microtiter plate assays (serial dilutions, eight parallel samples per dilution), using the cells mentioned by microscopy for virus-induced changes in morphology (cytopathic effect, syncytia formation). Titers were calculated using the method of Kaerber and Spearman and are expressed as log of tissue culture infectious doses (log 50% tissue culture infectious dose).[16, 17] The effectiveness of inactivated viruses was calculated based on the virus titer in untreated plasma minus the virus titer in plasma with DT at different doses.
Measurement of variables plasma factors
The measurements were performed in the laboratories of the Department of Clinical Laboratory, Chinese PLA General Hospital. Factor (F)VIII, fibrinogen, antithrombin-III (AT-III), and Protein C were selected to be measured. Plasma samples were treated with different UV doses and control samples without any treatment were immediately used in the assays mentioned above. FII (fibrinogen) activity tests were performed on a coagulation analyzer (STA Compact, Diagnostica Stago, Asnieres, France) and the IL coagulation system (Instrumentation Laboratory Company, Bedford, MA) was used to analyze FVIII, AT-III, and Protein C concentrations. Fibrinogen was measured with the STA Compact using the Clauss clotting method and a fibrinogen kit (STA-fibrinogen 5 kit, Diagnostica Stago). FVIII, AT-III, and Protein C were measured on the IL coagulation system using the activated partial thromboplastin time assay with a FVIII-deficient plasma kit (Hemosil, IL Company, Bedford, MA), automated chromogenic assay with a liquid antithrombin kit (HemosIL, IL Company), and automated chromogenic assay with a Protein C kit (HemosIL, IL Company), respectively.
Real-time polymerase chain reaction
RNA was extracted from samples containing VSV using TRIZOL (Sigma) and RNA was converted into cDNA for amplification (GoScript reverse transcription system, Promega, Madison, WI). Two primer pairs, F1/R1 (TGATACAGTACAATTATTTTGGGAC/GAGACTTTCTGTTACGGGATCTGG) and F2/R2 (TGGCTTCCTATTTACATCCAT/TATCCACTGCAGCAACTATTT) were designed to specifically amplify 227- and 202-bp genomic fragments, respectively. Five microliters of cDNA was added to 45 μL of master mix containing 10 μL EveGreen, 300 nmol/L primers, 500 nmol/L dNTP, 30 μL ddH2O, and 2 μL Taq polymerase (Invitrogen, Carlsbad, CA). Polymerase chain reaction (PCR) amplification was performed in a cycler (MiniOpticon, Bio-Rad, Munich, Germany) as follows: initial incubation for 5 minutes at 94°C and 40 cycles of 10 seconds at 94°C and 30 seconds at 72°C. The mean cycle threshold (Ct) values were recorded.
The experiments related to virus inactivation, plasma protein assays, and RT-PCR were each performed three times. The results are presented as the arithmetic mean ± standard deviation (SD).
System variables of the light source
The variables of the light source were regarded as possible factors that influenced the effectiveness of virus inactivation and retention of proteins. The determination of variables could potentially ensure the stability and reproducibility of the device for long-term use. The actual measurements of certain main variables were as follows: the spectrum of illumination was a broadband with overlaying peaks of 313 and 365 nm (Fig. 2), and the intensity of illumination was 6 mW/cm2.
The UV irradiation doses calculated were 1.8, 3.6, 5.4, and 7.2 J for treatment times of 5, 10, 15, and 20 minutes, respectively. The capacity of the sample-flow tube was 20 mL, the transmittance of quartz tube was 95%, the diameter was 5 mm, and thickness was 1 mm. The regulated velocities of peristaltic pump were 4 mL/min at 2 RPM, 2 mL/min at 1 RPM, 1.3 mL/min at 0.75 RPM, and 1 mL/min at 0.5 RPM for 5, 10, 15, and 20 minutes treatments, respectively.
Effectiveness of indicator virus inactivation with DT
The device was evaluated for efficacy of inactivation against five viruses: VSV, PRV, Sindbis, EMCV, and PPV, which are indicator viruses for human immunodeficiency virus (HIV), hepatitis B virus (HBV), hepatitis C virus (HCV), hepatitis A virus (HAV), and parvovirus B19 (B19V), respectively. The results are shown in Table 1. At the low dose of 1.8 J, enveloped VSV and PRV and nonenveloped EMCV and PPV could all be inactivated, resulting in a reduction of virus titers by approximately 3 logs. The low-dose treatment was less effective than higher dose. At 3.6 J, we found that, with the exception of Sindbis virus, the other four viruses were inactivated more than 4 logs, which met the criteria for virus inactivation in blood products according to the guidelines in China. The most effective dose was 7.2 J; all viruses were inactivated more than 5.5 logs.
Table 1. Log reduction of virus titers in FFP with DT at different doses*
Log reduction titer (J)
* Data are reported as mean ± SD. n = 3.
ss = single stranded.
3.84 ± 0.36
5.74 ± 0.45
≥6.00 ± 0.00
≥6.00 ± 0.00
3.15 ± 0.12
4.25 ± 0.35
4.90 ± 0.14
5.40 ± 0.14
2.00 ± 0.00
3.31 ± 0.43
5.35 ± 0.49
5.73 ± 0.32
3.23 ± 0.25
4.83 ± 0.15
6.00 ± 0.62
6.73 ± 0.46
3.53 ± 0.04
4.38 ± 0.18
≥5.00 ± 0.00
≥5.00 ± 0.00
Complete inactivation of low-titer virus with DT
The complete inactivation of low-titer virus, adjusted by dilution, was also evaluated. Complete inactivation of 2.3 logs of virus was accomplished at a dose of 3.6 J. After recultivation, virus titers were still not detectable by 50% tissue culture infectious dose (Table 2).
Table 2. The effectiveness of inactivated low titers of virus in pooled plasma with DT*
Log virus titer
After 3.6 J DT
* Data are reported as mean ± SD. n = 3.
2.27 ± 0.25
Using real-time PCR, the Ct values of samples for both primer pairs after complete inactivation with 3.6 J DT and after recultivation were all similar at approximately 30 cycles, compared with a Ct value before DT of approximately 20 cycles (Table 3). Higher Ct values represent lower copies of target fragments and vice versa.
Table 3. Ct values of samples treated using real-time PCR*
After 3.6 J DT
* Data are reported as mean ± SD. n = 3.
22.54 ± 0.25
30.14 ± 0.18
31.93 ± 0.43
21.08 ± 0.59
31.53 ± 0.30
31.27 ± 0.52
Retention of proteins in plasma with DT
Doses from 3.6 J, the lowest dose that efficiently inactivated viruses, to higher doses were evaluated for the retention of plasma proteins. At the low dose of 3.6 J, the activities of coagulation proteins were reduced by different amounts: the minimum retention was approximately 60% for fibrinogen and the maximum retention was more than 90% for AT-III. Among them, fibrinogen was the most susceptible to DT. At the highest dose of 7.2 J, abundant loss of proteins was generated and the minimum retention was only 28% for fibrinogen (Table 4).
Table 4. Retention of plasma variables in FFP with DT at different doses*
* Data are reported as mean ± SD. n = 3.
2.36 ± 0.32
1.42 ± 0.28
1.12 ± 0.39
0.66 ± 0.10
60.2 ± 8.6
46.6 ± 9.4
28.0 ± 1.0
0.72 ± 0.20
0.48 ± 0.13
0.42 ± 0.16
0.36 ± 0.13
67.0 ± 3.1
57.3 ± 8.0
49.2 ± 3.8
0.97 ± 12.3
0.91 ± 0.14
0.87 ± 0.16
0.84 ± 0.15
93.6 ± 2.8
89.6 ± 6.1
86.5 ± 5.3
Protein C (IU/mL)
0.90 ± 0.15
0.66 ± 0.16
0.55 ± 0.21
0.45 ± 0.15
72.1 ± 5.6
59.8 ± 14.2
49.2 ± 8.4
Among the techniques used, in China, the methylene blue method for treating single plasma samples is still applied extensively. Furthermore, the riboflavin photochemical method for virus inactivation is not a new approach for the treatment of blood products abroad. Using this method, however, the flow-based treatment device can be used to treat pooled or large volumes of plasma for a practical purpose. The flow-based treatment approach could be directly used for subsequent safe treatment of pooled plasma or the universal plasma supply of China that was developed by our group for emergency military use. We chose broadband UV combined with the photosensitizer riboflavin. The emission spectra of UVA and UVB sources peaked at 365 and 312 nm, respectively, while UVA had some overlap on UVB (Fig. 2). Because of the photosensitization induced by riboflavin, nucleotides underwent extensive degradation under both UVA and UVB. In addition, photochemical treatment with long-wavelength UV light yielded experimental evidence that reflected effective inactivation of certain relevant viruses, for example, PPV.[20-22] We needed to design a protocol to inactivate viruses effectively and maximally activate the riboflavin reaction. To overcome some difficulties, such as plasma container and the UV transmittance, a device was assembled from a quartz material sample tube that had the characteristics of high transmittance, a thin wall, and other incidental added modifications (Fig. 1). After fixing the main variables of the device, next we wanted to determine the optimal treatment time using our machine for which a treatment dose of 5 minutes was equal to 1.8 J, 10 minutes was equal to 3.6 J, and so on. Two aspects must be carefully considered: the effectiveness of virus inactivation and the retention of protein level or activity.
We demonstrated the effectiveness of virus inactivation and the retention of proteins using the device. Based on the virus inactivation results, the data showed that the enhancement of effectiveness of virus inactivation was achieved, when other variables were fixed, with the increasing UV dose. These indicator viruses showed different sensitivities against DT. It is known that a model virus may not always reflect the susceptibility of the corresponding clinically relevant virus to a certain treatment. The detection limit was reached when the original virus titer was not high enough, such as PPV, or the inactivated titer was greater than the original titer, such as VSV—in this case we expressed the inactivated results not as a definite data, but as “>” symbol (Table 1). Enveloped VSV is a model virus of HIV that seemed to be more easily inactivated than others, perhaps because it contains a lipid envelope and RNA genome. By contrast only moderate inactivation of another RNA virus, Sindbis, was observed at 1.8 and 3.6 J, but increased markedly when the dose increased to 4.8 or 7.2 J. Another enveloped virus, PRV, was similarly inactivated, as were nonenveloped EMCV and PPV. It appeared that whether an envelope was present on the surface of a virus did not obviously influence the inactivation efficiency with our DT. In the aspect of genome, earlier findings indicated that DNA viruses were generally more sensitive to UV light than RNA viruses.[24, 25] In our study, however, there were no remarkable differences between viruses with DNA or RNA genomes. Our virus inactivation results were different than the Mirasol system, which generated nonuniform inactivation of each indicator virus. To account for this difference, we would need additional studies. According to the China guidelines for virus inactivation in blood products, an available virus inactivation method applied in manufacture of blood products must meet the standard that the average inactivated indicator virus titer reduction is greater than 4 logs. As a result, the minimum treatment dose that could meet the criteria of available inactivation was 3.6 J, corresponding to a treatment time of 10 minutes. With the flow-based treatment of at least 10 minutes, the safety of plasma or other blood products could be ensured. Our investigation of virus inactivation should be extended, especially to the complete inactivation of low-titer viruses and the long-term safety consequences of using our device to treat pooled plasma by flow. It was known that no routine laboratory technique could test for total pathogens and a condition of low virus titers exists in pooled plasma because some single units of plasma would contain viruses that were not detected.[26, 27] It was essential to understand that the titer reduction of virus and the complete inactivation of virus were in two different cases. To simulate the microenvironment of containing a low-titer virus in plasma, VSV was diluted by plasma to a low titer that was less than 3 logs. The complete inactivation of VSV could be performed with DT at the dose of 3.6 J and no titer was detected after virus recultivation (Table 2). Real-time PCR was suggested as a simple method for detection in this case.[28-30] By comparing the Ct values obtained by RT-PCR (Table 3), we made the preliminary conclusion that the VSV genome was randomly fractured into fragments because of damage induced by UV radiation resulting from the increasing of Ct value from samples with no treatment to treatment. The damage was irreversible as the genome could not be successfully repaired after virus recultivation because of the embedded action of riboflavin resulting from the similar Ct values between samples with treatment and recultivation. The inactivation with irreversible damage achieved by this approach promotes long-term safety.
Meanwhile, in addition to guaranteeing safety from virus infection, further functionalities of FFP must be also considered. Determination of protein retention should be conducted using an efficient minimum dose for virus inactivation, which we had defined as 3.6 J. It was known that the penetrating and destructive abilities of UV radiation could decrease or abolish the bioactivities of the main functional factors in plasma. Four major plasma proteins that have functional therapeutic properties in plasma were selected to evaluate in vitro the loss of activities after DT. Fibrinogen as the basal raw materials is involved in the coagulation process of whole blood. Deficiency of coagulation FVIII leads to hemophilia A, an inherited sex-linked recessive disorder that causes severe bleeding. Liquid antithrombin is a major inhibitor of blood coagulation and its deficiency is associated with a high risk of thromboembolic disorders. Protein C is found as a zymogen in plasma that can be activated by thrombin to promote hemolysis.[31-33] The data showed that the different plasma factors responded differently with treatment at each dose of UV irradiation (Table 4). Fibrinogen was most vulnerable to the treatment, with retention declining from 60% at 3.6 J to 28% at 7.2 J. AT-III seemed to be the least affected of the four with retention changing from 93% to 86%. We concluded that the total level of plasma factor activities is reduced with increasing UV dose. Therefore, we should select the lowest effective dose when considering the functions of plasma. To balance the effectiveness of virus inactivation with retention of plasma proteins, in our study, we selected a “compromise” dose of 3.6 J or a treatment time of 10 minutes. At this dose, the retention of these proteins all exceeds 60%; however, additional studies will need to be performed to evaluate whether treated plasma is as efficacious as untreated plasma. Because of the heat of the UV lamps and the small distance between the lamps and the flow tube, it is possible that samples are heated during the treatment process and that some declines in protein activity may be caused by protein denaturation. Protein stability in frozen stored plasma had been confirmed and known for many years.[34, 35] However, our protein retention results at the determined dose were inferior to the Mirasol system in total level. In a future study, a sealed-in cold chain system was applied to the device, such as a homothermal cooled thermoconditioner.
Our study represents an initial development of a new method, and we show slight differences between our device and the Mirasol system in the evaluation of virus inactivation and protein retention. In conclusion, the novel flow-based treatment device using the riboflavin photochemical method that we designed has been demonstrated to be an efficient system to safeguard blood at a treatment dose of 3.6 J. More modes that differ from classical treatment methods for single plasma samples should lead to the spread of new ideas to improve blood safety. Furthermore, as described, we will continue to verify this flow-based treatment device for the inactivation of other viruses and to develop additional plasma protein assays, to optimize it for large-volume industrialization. Now the device is small scale, if just to treat plasma with volume of less than the capacity, 10 minutes of treatment time is needed. If the volume is larger than the capacity, prolongation of total time is required. When the capacity is enlarged, the flow speed also needs to be increased. Further, we intend to enlarge the capacity to 5 L; the flow speed needs to be increased to 500 mL/min. So to treat 500 mL, 2 L, or 5 L of plasma, 10 minutes would be required, but to treat 10 L of plasma, a total of 20 minutes would be required. The way of changing the other variables such as widely increasing the light intensity is not suggested.
The experiment support of Yuyuan Ma, Rui Wang, and Jianguo Wang from the Institute of Transfusion Medicine, Academic of Military Medical Sciences, is acknowledged. We also thank the Department of Clinical Laboratory, Chinese PLA General Hospital where the plasma variables were assayed and the Beijing Jing-Meng Stem Cell Technology Company where the real-time PCR was done.
Conflict of Interest
The authors have disclosed no conflicts of interest.