Aims—To determine the conditions of photodynamic inactivation of vesicular stomatitis virus (VSV) added to pooled coumarin plasma and the effects of the photodynamic treatment on the prothrombin times and international normalised ratio (INR) in a Netherlands national external quality assessment scheme.
Methods—Pooled coumarin plasma samples were illuminated with visible light in the presence of 1 μM methylene blue. Inactivation conditions for VSV in pooled coumarin plasma were determined using an end point dilution assay. Plasma illuminated for 20 minutes was mixed with red blood cells and mailed to participants of the Netherlands external quality assessment (EQA) scheme. Prothrombin times and INRs were determined with various thromboplastin reagents.
Results—Photodynamic treatment using 1 μM methylene blue and 700 W/m2 caused 4.7 log inactivation of VSV in pooled coumarin plasma. Fibrinogen and coagulation factors II, V, VII, and X were decreased slightly by the treatment. These conditions caused prolongation of the prothrombin time in EQA surveys. The magnitude of the effect was different for various thromboplastin reagents. The increase of the INR was negligible when measured with the Thrombotest reagent. With other reagents, an approximately 5–16% increase of the INR was observed. Interlaboratory variation of the INR was not affected by photodynamic treatment.
Conclusions—Photodynamic treatment of pooled coumarin plasma is very effective for the inactivation of some enveloped viruses such as VSV, but has only a limited effect on the prothrombin time and INR. Photodynamic treatment can be used to improve the viral safety of coumarin plasma for EQA of the prothrombin time and INR.
- prothrombin time
- international normalised ratio
- external quality assessment
- photodynamic treatment
Statistics from Altmetric.com
External quality assessment (EQA) of the prothrombin time and international normalised ratio (INR) is an important component of quality assurance of oral anticoagulant treatment. Since 1974, the Federation of the Netherlands Thrombosis Centres has organised a national EQA scheme for its members.1 Each year, 10 surveys are performed with five liquid control blood samples in each survey. Many Dutch centres are using the Thrombotest reagent (Nycomed, Oslo, Norway), which is a modification of the prothrombin time test.2 It is used with citrated venous blood. As far as possible, survey materials should resemble test specimens, and should include a range of coagulation activities to be measured, especially at critical levels for decision making. The survey materials in the Netherlands scheme are prepared by mixing washed human red blood cells with various pooled plasma samples obtained from normal individuals and patients treated with oral anticoagulant drugs.3
There is a risk of contamination with blood borne viruses when a large number of patients' specimens are used for the preparation of pooled plasmas. Although the pooled plasmas are tested for hepatitis antigen and human immunodeficiency virus (HIV) antibodies, the risk of infection of laboratory workers cannot be eliminated completely. Leakage of survey material from the primary container may occur during transportation and postal delivery. EQA organisers might be held responsible for injury caused by survey materials. To improve the safety of the survey materials, we evaluated photochemical methods for the inactivation of viruses in pooled plasma samples.
In transfusion medicine, photodynamic treatment has been studied for the inactivation of blood borne viruses in blood components. A photodynamic virus inactivation method using methylene blue is being applied routinely in the production of virus inactivated plasma.4 Methylene blue was selected because it is used clinically and because of its known toxicological properties. The standard procedure for photodynamic treatment involves illumination with visible light at a methylene blue concentration of 1 μM in plasma.5 Upon illumination, methylene blue becomes excited, and energy transfer from the excited state to molecular oxygen dissolved in the plasma can cause the formation of singlet oxygen. This very reactive oxygen species can damage lipids, proteins, and viral nucleic acids. Photodynamic treatment with methylene blue effectively inactivates various enveloped viruses, including HIV. Polymerase chain reaction analysis revealed that hepatitis B virus, hepatitis C virus, HIV-1, and probably also the non-enveloped parvovirus B19 are sensitive to methylene blue/light treatment.6
The purpose of our study was to determine the conditions for effective photodynamic inactivation of vesicular stomatitis virus (VSV) added to pooled coumarin plasma, and to study the effect of photodynamic treatment on the prothrombin time and INR of the plasma to be used for preparation of survey samples in the Netherlands EQA scheme. We report the results of two surveys in which photodynamic treatment was applied.
Materials and methods
PREPARATION OF POOLED PLASMAS
Venous blood was collected from coumarin treated patients using evacuated tubes containing 0.105 M buffered sodium citrate (Becton Dickinson Vacutainer™ Systems; Becton Dickinson, Franklin Lakes, New Jersey, USA). Plasma was obtained by centifugation at 2000 ×g for 10 minutes. Samples with INRs in the range 1.5 to 3.0 or 2.5 to 4.0 were pooled and centrifuged for a second time, but now at 29 000 ×g for 30 minutes. After careful decantation the pooled plasmas were frozen in closed plastic containers at −70°C. After thawing in a waterbath at 37°C for 15 minutes, the pooled plasmas were mixed with HEPES buffer (final concentration in plasma, 0.05 M; pH 7.3). In some experiments, the oxygen dissolved in the plasma was removed by slowly bubbling nitrogen through the plasma before photodynamic treatment.
Methylene blue was obtained from Sigma Chemical Co (St Louis, Missouri, USA). A stock solution (1 mM in water) was prepared. Each plasma sample was mixed with methylene blue (1 μM final concentration). The plasma was then illuminated at room temperature, in 50 ml polystyrene tissue culture flasks, using a 500 W halogen lamp. The plasma was stirred during the illumination. A large glass vessel with running tap water was placed between the lamp and the flask to absorb the heat generated by the lamp. The distance between the lamp and flask was 12 cm. Irradiance at the site of the flask was 700 W/m2, as measured with a Gentec TPM-310 photometer.
VSV INFECTIVITY ASSAY
Measurement of the infectivity of VSV by an endpoint dilution assay was performed as described previously.7 Results were expressed as percentage of the control (no illumination).
The following assays were performed in the authors' laboratory. Fibrinogen was determined according to the method of Clauss.8 Factors II, V, VII, and X were determined with coagulation assays using human tissue factor and plasma samples deficient in the respective coagulation factors. Prothrombin times were determined with Innovin (Dade Behring, Marberg, Germany), Thrombotest (Nycomed), and PT-Fib HS (Instrumentation Laboratory, Breda, The Netherlands). Instrument specific international sensitivity index values were used to calculate INR values. Thrombotest dilution plots were made essentially as described by Hemker and co-workers.9 In these plots, Thrombotest clotting times t of diluted plasma samples are plotted against the dilution factor D. According to these authors,9 the quantity of a competitive inhibitor present in undiluted plasma can be estimated from the distance I on a horizontal line between tmin. uninh. and the intercept of the t-D line of the experimental plasma with that horizontal line (fig 1).
PREPARATION OF SURVEY SAMPLES
The preparation of survey samples was modified from the original procedure.3 Red blood cells of blood group O were obtained as packed cells in CPDA-1 (citrate phosphate dextrose adenine) from the Central Laboratory of the Netherlands Red Cross Blood Transfusion Service. After washing of the red blood cells with phosphate buffered saline and removal of the buffy coat, the cells were incubated with a solution containing sodium chloride (0.051 M), trisodium citrate (0.017 M), and sodium HEPES (0.1 M), pH 8.6. After two hours incubation at room temperature, the cells were centrifuged and the supernatant was discarded. The cells were then mixed with the pooled coumarin plasma samples, aiming at a haematocrit value of 0.4–0.45. A mixture of penicillin and streptomycin was added to prevent microbial growth. The pH of the artificial blood was approximately 7.5. The blood was distributed in 1.0 aliquots in capped polypropylene tubes. The blood samples were prepared on Monday and mailed to the participants of the EQA scheme on Tuesday. Participants were instructed to store the samples at room temperature until analysis on Friday—four days after preparation. Participants used their routine method for prothrombin time and INR determination. The homogeneity of the samples was determined in the authors' laboratory by prothrombin time testing of 10 samples, using Simplastin Excel-S as thromboplastin reagent and a Coagamate-MTX as coagulometer.
STATISTICAL EVALUATION OF EQA RESULTS
Prothrombin times and INRs reported by the participants of the Netherlands EQA scheme were evaluated for each group using the same brand of thromboplastin reagent, irrespective of the type of instrument (coagulometer) used. When a participant's result was greater than two times the mean value of the group using the same reagent, it was regarded as an outlyer and excluded from the final calculations. Interlaboratory variation was expressed as coefficient of variation (CV) in per cent.
The inactivation of VSV in pooled coumarin plasma was determined using an infectivity assay, based on the cytopathological effect of VSV on A549 cells, as scored by light microscopy after 72 hours. As shown in fig 2, complete inactivation of added VSV (4.7 log) could be obtained by treatment with 1 μM methylene blue and 20 minutes illumination with white light (700 W/m2). These conditions were used for the preparation of EQA samples.
The inactivation of fibrinogen and factor V in pooled coumarin plasma was determined as a function of the illumination time. As shown in fig 3, a biphasic decrease of the fibrinogen concentration occurred in the presence of air. Under a nitrogen atmosphere, there was practically no decrease of the fibrinogen concentration. Factor V activity decreased approximately 10% in the presence of air, but no change was seen under nitrogen (fig 4). Similarly, there was a slight decrease of factor II and factor X activities in the first 15 minutes of the illumination. In the following hour, no measurable change was seen. Factor VII activity remained constant in the first hour of the illumination. INR values measured with Innovin (recombinant human thromboplastin) increased rapidly in the first 15 minutes (with air), followed by a phase of slow increase (fig 5). Similar results were obtained with PT-Fib HS, a rabbit tissue thromboplastin (not shown). INR values measured with Thrombotest were only slightly increased (fig 6). In Thrombotest dilution plots (fig 1), the slope of the line obtained with illuminated plasma was greater than that with non-illuminated plasma. Furthermore, the distance I was slightly decreased after illumination.
EXTERNAL QUALITY ASSESSMENT OF PROTHROMBIN TIME AND INR
Two EQA surveys were performed in November and December 1998. In each survey, one sample was illuminated as described above, a second sample was prepared from the same coumarin pool with methylene blue but not illuminated, and a third sample was prepared from the same coumarin pool without methylene blue and not illuminated. The homogeneity of the samples was good (CV < 0.9%). The mean prothrombin times and interlaboratory CV are shown in table 1, and the mean INR values in table 2. The mean prothrombin times of the samples prepared from the same coumarin pool not illuminated with and without methylene blue were almost the same; the differences were not greater than 2%. The mean clotting times and INRs reported by participants using the Thrombotest reagent were almost the same before and after illumination of the pooled coumarin plasmas in the presence of methylene blue. In contrast, there was a substantial increase in clotting time and INR with the other reagents; the increase ranged from 5% to 16%. The interlaboratory variation of the prothrombin times and INRs was not different before and after illumination.
VSV is a very well characterised lipid enveloped virus that is used as a model for the study of photodynamic virus inactivation. The virus inactivating properties of methylene blue and other phenothiazine dyes in combination with visible light have been known for many years. Viruses vary in their sensitivity to methylene blue, but the susceptibility of HIV-1 to photoinactivation was similar to that of VSV.5
Under the conditions of photodynamic treatment used in our study (20 minutes illumination with white light (700 W/m2) in the presence of 1 μM methylene blue) there was complete inactivation of all VSV added to pooled coumarin plasma buffered with HEPES (fig 2). Our results are in agreement with those of Lambrecht et al who used fresh frozen plasma isolated from blood donations.5
Photodynamic treatment of pooled coumarin plasma resulted in a substantial decrease of the functional fibrinogen concentration as measured with the Clauss method. The photodynamic inactivation of fibrinogen was, as expected, dependent on the presence of oxygen in the solution (fig 3). Lowering the oxygen concentration by bubbling with nitrogen resulted in the complete absence of photodynamic damage. Previously, it has been shown that illumination in the presence of methylene blue resulted in a substantial reduction of histidine and tryptophan residues.10 This is probably caused by photooxidation of these residues by singlet oxygen, as shown in a study using scavengers of different reactive oxygen species.11 The relative decrease of factor II, V, VII, and X activities (fig 4) was smaller than that of fibrinogen, which is in agreement with the observations by other investigators.5,11,12 Furthermore, the activities of factors II, V, VII, and X remained practically constant after 15 minutes. The decrease in fibrinogen was biphasic and continued after 15 minutes. The biphasic kinetics might be the result of oxygen depletion occurring under our experimental conditions, which causes a slower rate of photodynamic damage in the second part of the illumination period. It seems that air bubbling before illumination cannot prevent the biphasic kinetics of activity loss. An alternative explanation for the observed effect might be that the different amino acid residues in the proteins undergo photooxidation at different rates. The partly oxidised proteins are apparently still able to function to some extent.
The photodynamically induced increase of the prothrombin time and INR is probably caused by inactivation of fibrinogen, and coagulation factors II, V, VII, and X.5,11,12 In the absence of oxygen the prothrombin time remained almost constant. Therefore, the increase of the prothrombin time in the presence of oxygen can be attributed to photooxidation of the proteins involved. The relative increase was not the same with the various thromboplastin reagents. The prothrombin times and INRs determined with plain rabbit and human thromboplastin increased relatively more than those determined with Thrombotest, a bovine thromboplastin combined with adsorbed bovine plasma. These results suggest that photooxidised coagulation factors and inhibitors react differently with bovine tissue factor than with human and rabbit tissue factor. In a Thrombotest dilution plot (fig 1) the slope of the pooled coumarin plasma was increased by photodynamic treatment and the distance I was decreased. This could be interpreted as a decrease of factor X activity and of competitive inhibitor protein induced by vitamin K antagonists (PIVKA).9 Because several coagulation factors are affected by the photodynamic treatment, it is not a simple task to determine which factors are responsible for the prolongation of the prothrombin time. Further experiments are needed to resolve this question.
Previous studies have shown that the liquid artificial blood samples used in the Netherlands EQA scheme are not stable.3 During storage at room temperature, a slow increase of the proththrombin time was seen. All participants were requested to perform the tests on the same day, so that the effect of deterioration was approximately the same in all samples. Differences in ambient temperature between the various parts of the Netherlands are small and the transit time is usually not greater than 24 hours. Furthermore, the homogeneity of the samples on the day of testing was very good (CV < 1%). The interlaboratory variation of the samples treated with methylene blue/light was not substantially different from the variation of the control samples when the variation was calculated for each reagent group separately (tables 1 and 2). In most of the reagent groups, an increase of the mean proththrombin time and INR was seen after photodynamic treatment of the pooled coumarin plama. In contrast, the mean clotting time and INR reported by the users of the Thrombotest reagent did not change greatly after photodynamic treatment. As a result, the differences between the reagent group mean values increased after photodynamic treatment. It should be noted that in the control samples that were not illuminated, there were also important differences in the INR between the reagent groups (table 2). For example, the mean INR reported by users of PT-Fib HS was 20% greater than the mean INR reported by Thrombotest and Hepato Quick users. After illumination, the difference increased to 35%. The differences in INR are the result of deterioration of the samples and a different response of the various reagents to deterioration. Therefore, the performance of each participant of the EQA scheme must be assessed by comparison with results from other laboratories using the same reagent. In agreement with previous reports,13 it is not possible to assign a single INR value to these EQA samples that would be valid for all reagents.
In conclusion, our results show that photodynamic treatment of pooled coumarin plasma is effective for viral decontamination, without important impairment of its usefulness for EQA of the prothrombin time and INR.
The authors thank Ms E Witteveen, Ms C van Rijn, Mrs H Schaefer-van Mansfeld, and Mrs J Meeuwisse-Braun for excellent technical assistance. Mr G van de Kamp (Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, Amsterdam) supervised the preparation of control blood specimens. Financial support was received from the Federation of the Netherlands Thrombosis Services.