Pall Medical has developed a prion reduction filter set, the Pall^® Leukotrap Affinity Prion Reduction Filter (LAPRF^§) System, for the reduction of infectious prion as well as residual leucocytes from red cell units. This report provides a characterization of filter performance with respect to prion reduction, quality of the red cells after filtration and subsequent storage, and the further lowering of residual white blood cells in a previously leukoreduced product.^§Pall Leukotrap Affinity Prion Reduction Filter System, is abbreviated LAPRF System and, represents a system designed for prion reduction intended for use with leukoreduced packed red cells.

Prions are infectious proteins believed to be responsible for a variety of transmissible spongiform encephalopathies (TSEs) which are progressive neuro-degenerative diseases characterized by sponge-like morphology of the brain upon gross physical examination at autopsy.^1,2 TSEs in humans and other animals are invariably fatal. During the 1980s in the United Kingdom, citizens became aware that cattle were infected with prions and succumbed to “mad cow” disease also called bovine spongiform encephalopathy (BSE).³ In 1996 it was recognized that ingestion of beef from BSEs infected cattle could lead to a variant of TSE in humans called variant Creutzfeldt-Jakob Disease (vCJD).⁴

It quickly became apparent that vCJD was different from sporadic CJD in that those afflicted were of a considerably younger age (median age of death 28 years vs. 67 years, respectively). Since the latency from infection to onset of symptoms was known to be long in many TSEs, and there is no accepted ante-mortem test for prions,⁵ concern was expressed over the potential that aysmptomatic blood donors could transmit vCJD. Blood borne prion transmission had previously been demonstrated in animal models.⁶ These fears were realized in 2004 with the first report of probable transfusion-transmitted vCJD, which developed over 6 years after a patient received blood from an aysmptomatic donor who subsequently died of prion disease.⁷ Not long after, signs of vCJD at autopsy were detected in yet another patient who died of unrelated causes 5 years after receiving a blood transfusion from a donor who subsequently developed vCJD.⁸

Sensitive and specific ante-mortem tests for either BSE or vCJD are not available nor has it been possible to provide an accurate estimate of individuals who may be incubating vCJD.9 As a result, a system of donor deferral is the strategy being employed to minimize the risk of transfusion-transmitted vCJD. Donor deferral relies upon the donor’s best recollection of times spent in the UK and is, consequently, not very reliable. In addition, the impact of donor deferral has the potential to threaten blood availability.10 Pathogen inactivation is not yet seen as a plausible strategy because most pathogen inactivation approaches rely on the fact that the infectious agent is dependent upon nucleic acids for growth and reproduction, unlike prions. Leucofiltration is employed to reduce leukocyte-associated prion.

Although leukofiltration is a necessary step for reducing leukocyte-associated prion infectivity, this process is necessary but not singularly sufficient for eliminating the plasma-associated prions and the risk they present from a transfused infected unit of blood. The pathogenic prion is distributed within the cellular (white blood cells) and plasma component of blood nearly equally. Many blood services introduced universal leukocyte reduction (ULR) as a precaution against transmission of white cell-associated vCJD. Gregori et al.¹¹ collected whole blood from scrapie-infected hamsters to provide a pool of 450 mL, which was then filtered using a commercially available whole blood leukocyte-reducing filter. Leucofiltration of the whole blood was found to remove only 42% of the initial prion infectivity, an observation consistent with others. Nevertheless, the safety of the blood supply has been, and continues to be, improved with ULR.

Due to the protracted interval of asymptomatic infection and the inability to detect pathogenic prions in blood, reduction of infectious risk through prion filtration is an attractive option, especially since blood leucofiltration is a common practice already employed in blood centers.

The primary response measures, shown in Table 1, were evaluated using the Pall LAPRF System.

Table 1

Primary response measures

prion reduction

white blood cell removal

plasma free hemoglobin (pfHgb), % hemolysis

RBC recovery (total Hgb)

filtration time

filter holdup volume

hemo-/bio-compatibility (band 3, neoantigenicity)

 

 

 

 

Overview of Prion Reduction Testing

Pathogenic prions (PrP*) are novel infectious agents and the experimental methods to study prion removal are evolving; however, current methods for characterizing filter performance vary and are worth defining. Three important considerations attend discussion of testing prion removal or reduction technology (Table 2). These are:

1. the source of the pathogenic prion,
2. the manner in which the blood becomes infected exogenously or endogenously), and
3. the response measures including: Western blot assay, infection without dilution of the blood product (Infectivity Study; Figure 1) and infection following serial dilution of the blood product (Bioassay; Figure 2).

 

Infectivity Defined

The best animal model is one that simulates the risk of transfusion-transmitted disease. The blood of animals should be infected endogenously, that is naturally from within the body (Figure 1). The human pathogenic prion is of very limited availability; however, the prion protein responsible for the sheep disease (PrP^Sc), furnishes an appropriate model for screening a filtration device. It is appropriate because symptoms of disease can be induced following intracerebral injection (IC) of prion and since numerous animals are required for study, small animals, like hamsters, are manageable. As brain is the affected target organ, IC injection of infected brain material containing a hamster-adapted sheep scrapie (or PrP^Sc strain 263K) into the sensitive Syrian hamster represents an established model of endogenous infection producing disease within 60-90 days.¹⁴ The blood of animals exhibiting symptoms of disease can be recovered, pooled to sizes representing a unit of human donated blood and then processed into components and filtered. Approximately 4 - 5 mL of blood can be recovered from a hamster thereby requiring over 100 animals to give enough blood to represent a single unit. Although sensitive to PrP^Sc, the latency from intracerebral injection to disease is still months. Time and the number of animals combine to preclude this type of testing for all except the final tests of filter performance.

Table 2

Definitions of models of prion reduction

Exogenous/Western blot	Exogenous/Bioassay	Endogenous/Infectivity
A screening method used to assess prion removal capability	Quantitative method to determine log reduction of infectious prion protein	Animal model most clinically relevant to determine the efficacy of prion removal

                          

*PrP=prion protein(s); PrPSc=PrP from Scrapie; PrPC=normal (non-pathogenic) form of PrP found on cells and is sensitive to digestion with proteinase K; PrPres= refers to the pathogenic form of PrP after limited digestion with proteinase K.

 

Figure 1

Pictorial representation of infectivity study

Bioassay Defined

Alternatively, human blood can be infected or ‘spiked’ with PrP^Sc scrapie infected hamster brain homogenate (SIHBH) making the source of blood infection exogenous (Figure 2). The blood, filtered or not, is then injected intracerebrally into normal hamsters and symptoms of disease can be monitored over time to assess the efficacy of prion removal. Since the exogenous administration of PrP^Sc can be a high titer, the filtrate can be diluted serially and each injected into a group of hamsters. When compared with unfiltered and serially-diluted blood, an estimate of the log reduction of prion by filtration can be determined. This dramatically reduces the time to conduct studies, since the necessity of a lengthy latency period from IC injection to symptom onset in order to allow distribution of PrPSc into the blood (i.e., endogenous) is avoided.

 

Figure 2

Pictorial representation of bioassay

Western Blot Assay

For the screening of filtration media and testing filters under a variety of filtration conditions, the latency to disease as a limitation in the conduct of studies can be obviated by measuring the amount of PrP^Sc using the Western blot assay. Briefly, the Western blot assay for pathogenic PrP relies upon the fact that digestion of the sample with proteinase K results in the lysis of normal protein found on cells (PrP^C); however, the pathogenic PrP (PrP^res) is more resistant and the two can be separated by electrophoresis in a polyacrylamide gel. The proteins remaining on the gel are transferred to a membrane by ‘electroblotting’ (which is the transfer of proteins from a gel to a membrane with an electric current). The membrane is incubated with antibody to PrP and this antibody is targeted by another molecule that emits light under appropriate conditions. The light is captured by photographic film or a charge coupled device (CCD) camera, the intensity of the light is quantified by densitometry, and is proportional to the amount of PrPres on the gel. This approach of pathogenic PrP spiking with Western blot quantification of results allows for a greater number of studies to be conducted than would otherwise be feasible.

A proper characterization of prion removal should employ a combination of approaches because each has strengths and weaknesses.

Strengths and Weaknesses of Each Technical Approach

The infectivity model most closely emulates the clinical situation. The blood is not directly contaminated with pathogenic prion from an exogenous source but rather transmission is endogenous via the nervous system or the immune system or both. The limitation is that the low bioburden of prion in blood (1) does not allow for easy detection by the Western blot assay, and (2) is not sensitive enough to test the limits of highly efficacious prion removal methods. The endpoint, is therefore, simply the presence or absence of the disease after a prescribed interval.

The bioassay has its advantage in being quantitative. Unfortunately, the source of infection is exogenous and may not represent the same form of pathogenic prion as might accrue from an endogenous infectious route.

The Western blot assay has value in being able to separate normal from pathogenic prion. Given the strengths and weaknesses of these investigative tools, a combination of them is most appropriate.

Preparation of SIHBH 

Weanling Syrian hamsters were inoculated IC with 1% (w/v) brain homogenate from symptomatic hamsters infected with scrapie strain 263K. The animals were sacrificed after 70 days at an advanced stage of disease, and the brains were removed to prepare 10% suspensions in phosphate buffered saline (PBS), pH 7.4, homogenized with an Ulter-Turrax T25 homogenizer. The homogenates were then centrifuged at 3000 g for 3 minutes at room temperature in a Beckman GPR tabletop centrifuge (Beckman Coulter, Fullerton, CA). The supernatants were recovered, pooled and separated into aliquots (50 mL) and stored at -70 ºC until required.

Variables Tested

Three filter set lots were compared for prion removal using blood spiked with SIHBH at levels approximating 10⁹ infectious units (IU)/unit of red cells as confirmed using a bioassay (BioReliance¹⁵). The effect of anticoagulant was studied comparing CPD with and without SAGM additive for red cells and CPDA-1. The filtration temperature was generally 22±2 ºC for all except one study of 4ºC temperature (n=8). Head height was also studied.

Blood Products

Whole blood (450 mL±10%) was collected from healthy donors into collection sets (Leukotrap^® Whole Blood Collection and Filtration System; Pall Corporation) for CPD and CPD/SAGM or for CPDA-1 that incorporated Pall WBF leukocyte reduction filters, and contained either 63 mL of CPD or CPDA-1 solution. Filtration took place within 24 hours (time indicated where applicable) from the time of collection. After filtration, the whole blood was centrifuged (hard spin) and red cells were expressed into satellite bags. Thereafter, where appropriate, 100 mL of SAGM was added. The number of units varied and is reported in results.

Inoculation of Pathogenic Prion

The volume of the red cell concentrate (RCC) was adjusted to 270 mL and 10 mL of 10% (w/v) brain homogenate were added so that the final concentration of the SIHBH was 0.36% (v/v) (that is 1:28 dilution of the stock suspension of SIHBH). The unit was mixed end-over-end for about 2 minutes (15-20 rotations). Twenty milliliters (20 mL) of RCC was transferred into 50 mL plastic centrifuge tube. The blood bag containing the remaining RCC was attached to the Leukotrap Affinity filter set and then filtered at 22±2 ºC at a filtration height of 30 inches unless specified otherwise. 

Prion Extraction

Prion extraction from blood was performed by phosphotungstic acid (PTA) precipitation followed by sarkosyl resuspension, in the manner described by Wadsworth et al.16 with minor variations. Initially, 2.5 mL of red cell concentrate was added to 6 mL of 4% sarkosyl in phosphate buffered saline (PBS). This was incubated with continuous mixing at 37ºC for 10 minutes.

Next, 700 µL of 4% PTA in 170 mM MgCl2 at pH 7.4 were added; and, the resulting mixture was incubated with continuous mixing at 37ºC for 30 minutes. Centrifugation at 19,600 g for 60 minutes at room temperature followed. The supernatant was then removed and discarded; and, the pellet was resuspended in 2.5 mL of PBS containing 0.1% sarkosyl (PBS-S). The suspension was then centrifuged at 19,600 g for 60 minutes at room temperature. Finally, the supernatant was removed and the pellet resuspended in 100 µL of PBS-S.

Western Blot: Prion Log Reduction Determination

The Western blot assay was performed by an enhanced chemiluminescence method. First, 20 µL (15 µL of sample and 5 µL of loading buffer) of sample were applied to each well of 4-12% SDS Bis-Tris polyacrilamide gels in NuPage running buffer (MES) (Novex, Invitrogen, San Diego, CA) for 35 minutes at 200 constant volts. This was followed by electroblotting onto PVDF membrane (Immobilon) in Novex NuPage transfer buffer at constant 25 V for 60 minutes. The membrane was blocked using 50 mL of 5% non-fat dry milk in PBS containing 0.2% Tween-20 (PBST) for 1 hour in a hybridization plastic bag with gentle agitation. For each 8 x 8 cm blot, 20 mL of solution were used. The membrane was then incubated in blocking solution containing 1:5,000 (1 mg/mL) dilution of primary monoclonal antibody to PrP, known as 3F4 for 2.5 hours in a hybridization bag with gentle agitation. Blots were then washed 4 times for 5 minutes each, and then rinsed in distilled water for at least 5 minutes. At this point the blots were ready for the addition of substrate.

In order to prepare a working solution of substrate, components of the SuperSignal West Dura Extended Signal Substrate (Pierce Biotechnology, Inc., Rockford, IL) were mixed in a 1:1 ratio just prior to use. For each 8 x 8 cm blot, 8 mL of solution were prepared with minimal exposure to light. The blots were then placed in a hybridization bag, and 8 mL of the working solution of substrate were added to each blot. The bag was sealed, and incubated for 5 minutes with gentle agitation. The hybridization bag was then opened, and all of the excess substrate solution was pressed out. Blots were placed in a clear plastic folder, and the edges sealed using a radiofrequency sealer in order to maintain a moist environment. The folder was placed on a stage in a dark room cabinet (UVP ChemiDoc-it, Upland, CA). Images were then obtained with a Hamamatsu CCD camera (Hamamatsu Photonics, Hamamatsu City, Japan), using Labworks™ version 4.8 program (UVP Inc., Upland, CA), with a set exposure time of 2 minutes. Blots were analyzed using the program for 1D gel analysis.

Quantification of prions was performed by developing a standard curve of the log dose of prions (in infectious units confirmed by Bioreliance Inc.) compared with log densitometric readings. Densitometric readings from gels of blood samples after spiking with SIHBH before and after filtration were then converted into infectious doses per sample by interpolation from the standard curve. Samples were adjusted for the volume of blood filtered and the ratio was converted to log reduction values.

Bioassay: Log Reduction Determination

Summary-General Principle

For the endpoint titration of prions in hamsters, ten serial dilutions of a prion sample are prepared, with each more diluted by a factor of 10. Each dilution is then injected (IC) into six hamsters. The titration is terminated when all or most of the hamsters destined to become ill show signs of CNS dysfunction. Upon neuropathologic examination, vacuolation of neurons, widespread astrocytic gliosis, and PrP accumulations are found. Reliable data from an endpoint titration require that no hamsters are ill at the highest dilution; otherwise, a valid endpoint cannot be calculated.

Bioassay Method

All the samples were frozen at -70ºC and then thawed at 37ºC prior to serial dilution. Briefly, the pre-filtration control red cell concentrates samples were diluted with phosphate buffered saline (PBS) 10 fold to produce samples with serial dilutions between 10^-5 and 10^-9. The post-filtration red cell concentrates were also serially diluted 10 fold to obtain samples between 10^-2 to 10^-8. 

Five to six hamsters were injected IC with 40 µL of either pre-filtration or filtered serial dilution of red cell concentrate. The animals were examined at weekly intervals and the titration was terminated when the hamsters showed clinical signs of scrapie disease. The concentration of infectious prions in the sample is calculated from the score at the

highest dilution at which there is a positive result. Titers were calculated using the Reed-Muench method.¹⁷ The observed titers were adjusted to values per mL of sample. The results are calculated and expressed as log reduction.

Infectivity Method: Estimating Clinical Impact of Prion Removal

Aliquots of 50 µL of the pre- and post-filtration red cell samples were injected intracerebrally to both sides of the brain of normal hamsters (300 each for pre- and post-filtration samples are planned for the final product evaluation). Maintenance and monitoring of the animals biweekly for 300 days and those that develop clinical symptoms of scrapie are sacrificed, and the brains of all animals (including survivors) are tested for the presence of PrP^res by Western blot assay using 3F4 monoclonal antibody.

Red Cell Storage Study 1

Whole blood was collected into anticoagulants and processed as described earlier for prion reduction with some exceptions noted below. Units were held at ambient temperature during transportation from mobile blood collection sites, and then filtered as described above within 8 hours of donation (Day 0). The resulting red cell units were held at 4ºC overnight. Thereafter, on Day 1 (sample size indicated in results): 

• some units were removed and filtered immediately with the LAPRF System (RC4 C/RC4 C),
• other units were removed, allowed to warm to room temperature, then filtered with the LAPRF System (RC 4C/RCRT),
• and still other units were not filtered (control; RC4 C/C)

The remaining 9 units were held at ambient temperature during transportation from the mobile blood collection sites, then held at 4ºC overnight as whole blood units. After this hold period, they were filtered prior to processing into red cell units.

Thereafter, on Day 1: 

• some units were allowed to warm to room temperature, and then filtered immediately using the LAPRF System (WB4 C/RCRT),
• other units were placed at 4ºC, and then filtered with the LAPRF System once the blood reached 4ºC (WB4 C/RC4 C),
• and still other units were unfiltered (control; WB4 C/C)

All red cell units were subsequently stored at 4ºC until the end of storage at 28, 35, and 42 days for CPD, CPDA-1, and CPD/SAGM or CP2D units, respectively. Samples were taken for analysis on Day 1, Day 7, and at the end of storage.

Red Cell Storage Study 2

Whole blood units were collected and then stored overnight on cooling plates. Units were processed into leukocyte depleted red cells day 1. Half of these units were secondary filtered using LAPRF on day 1 (red cells were stored at 4ºC for 3-4 hours prior to filtration) and half on day 2 (after red cells stored overnight at 4ºC).

Red cell units were subsequently stored at 4ºC until the end of storage (35 days). Samples were taken for analysis at the beginning and end of storage (35 days).

Red Cell Survival Methods

The objective of this study was to evaluate the 24-hour in vivo recovery of leukocyte reduced CP2D/AS-3 red cells that were processed with the LAPRF System and stored for 42 days.

Blood Collection

Units of whole blood (450 mL) were collected into the Leukotrap^® WB System with CP2D/AS-3 anticoagulant/storage solution at the American Red Cross, Mid-Atlantic Research Facility, Norfolk, VA.

Blood Processing and Leukoreduction

Of the twelve units that were collected, six units were processed to leukoreduced (LR) AS-3 RBCs at room temperature (RT) within 8 hours of donation and then placed in the cold for 24 hours prior to being filtered with the Affinity filter (RT group). The remaining units were collected and held as whole blood at 1-6ºC for 24 hours and then processed

to LR AS-3 RBCs and filtered with the LAPRF System (4 C group). 

LAPRF System Filtration

The day following collection, a LAPRF System was sterile connected to the leukoreduced AS-3 red cell unit and the unit was filtered and then stored until 42 days. Samples were taken pre- and post-filtration for cell count and hemolysis determinations.

Post-Filtration Blood Storage and RBC Survival

The filtered units were stored at 1-6ºC for 42 days and then sampled to evaluate the in vivo 24-hour recovery using standard single and double label methods.1-3,18,19, 20 Post-storage supernatant hemoglobin and hemolysis levels were also determined.

Statistical Methods

Comparison between means of 2 samples were conducted with a t-test unless assumptions of normal distribution or homogeneity of variance were violated wherein the Mann Whitney U test was employed. For comparisons of greater than 2 samples, ANOVA with Newman-Keuls multiple comparison follow up tests were used for parametric data. Kruskall-Wallis with Dunn’s multiple comparison was employed for non-parametric data. Mean and standard deviations (SD) are shown in all figures. For comparisons between more than two means or medians, a line spanning two bars in the figures designates a statistically significant difference (P<0.05) between the data represented by the two bars over which the line ends, as determined by the appropriate ad-hoc multiple comparison test.

Residual WBC were analyzed by counting the number of pre-filtration samples below the limit of detection by flow cytometry compared with post-filtration samples. Contingency tables were analyzed with Fisher exact test. WBC probabilities were calculated using the binomial distribution function (Excel; Microsoft, Redmond, WA). All statistical analyses, with the exception of binomial distribution analyses, were performed with Prism (Graph Pad, Inc.; San Diego, CA) as were all graphs prepared.

Prion Reduction

Log prion reduction was assessed for filters representing three different filter production lots; and, the values were compared by ANOVA. No statistically significant difference were found between the lots tested (Figure 3). The average log reduction value, as determined by the Western blot assay, is 2.75±0.72 (SD) for the combined lot numbers with the 25th and 75th percentiles indicated in the figure.

Log prion reduction was measured for units representing each of 3 storage solutions: CPD (N=8), CPDA-1 (N=10), and CPD/SAGM (N=9). Statistically significant differences were only observed between values for CPDA-1 and CPD as well as CPDA-1 and CPD/SAGM (Figure 4); however, CPD did not differ significantly from CPD/SAGM.

Figure 3

Lot-to-lot prion reduction

 

Figure 4

Effect of anticoagulant/additive solutions on prion reduction

Figure 5

Effect of head height on prion reduction

Log prion reduction was assessed in CPD/SAGM for 3 different head heights: 15" (N=6), 30" (N=8), and 40" (N=8). Statistically significant were only noticed between 15" and 30" (Figure 5).

Figure 6

Effect of temperature on prion reduction

Prion reduction was compared in CPD/SAGM for filtration at 4ºC versus 22ºC. Values for log prion reduction were compared by Mann Whitney test. In this analysis, a significantly greater reduction was noted with filtration at 4ºC (Figure 6).

Infectivity and Bioassay Preliminary Data

Although both studies are ongoing, preliminary infectivity and bioassay studies were performed with a prototype of the LAPRF System. The bioassay showed log reduction of 3.7. Infectivity attenuation with a single filter tested showed 0 of 35 animals exhibiting symptoms of prion disease in the group receiving filtered blood, whereas among controls, 6 of 43 showed evidence of disease (P=0.0384; Fisher exact test).²¹

Red Cell Quality Response Measures —

Red Cell Storage Study 1

Hemolysis and Plasma Free Hemoglobin (pfHgb)

Percent hemolysis and pfHgb were measured over time following filtration for red cells stored in CPDA-1, CPD, CP2D, and CPD/SAGM (Figure 7).

As expected, percent hemolysis and pfHgb were demonstrated to increase sharply on day 1, followed by a plateau in these values to outdate in CPDA-1 and CPD. Minimal increases were noted for CP2D and CPD/SAGM. This effect appeared to be consistent for each of the holding and filtration temperature conditions tested. No test condition resulted in values exceeding 0.8% hemolysis, which represents the limit defined by the Council of Europe.

Figure 7

Effect of filtration with LAPRF System and subsequent storage of blood products on pfHgb and hemolysis at varying hold and filtration temperatures.

(WB4 C/RC4 C) = whole blood overnight 4 ºC hold, red cells filtered day 1 close to 4 ºC; (WB4 C/RCRT) = whole blood overnight 4 ºC hold, red cells filtered day 1 room temperature; (RC4 C/RC4 C) = red cells overnight 4 ºC hold, red cells filtered day 1 close to 4 ºC; (RC4 C/RCRT) red cells overnight 4 ºC hold, red cells filtered day 1 room temperature.

Figure 8

Influence of hematocrit on post-filtration plasma free hemoglobin

In addition, filtrate pfHgb was evaluated as a function of HCT for each of three storage solutions: CPDA-1, CPD, and CPD/SAGM (Figure 8). It was noted that pfHgb appeared to rise with increasing HCT, particularly for values of HCT > 65% for both CPDA-1 and CPD without SAGM.

Flow Time

The duration of filtration or flow time (in minutes) was also evaluated as a function of HCT in CPDA-1, CPD, and CPD/ SAGM (Figure 9). Flow time was demonstrated to increase with increasing HCT, particularly for values of HCT >70 for CPDA-1. The effect was less pronounced for CPD. In general, temperature does not impact filtration times as much as increasing hematocrit. With hematocrit values less than 65%, the average filtration time is approximately 13 minutes.

Figure 9

Effect of hematocrit on the duration of filtration or flow time.

Measures of Bio- and Hemo-Compatibility

The effect of filtration upon red cells stored to outdate, was evaluated for a number of factors related to hemo- and bio-compatibility (ATP, K+, pfHgb, pH and C3a) for each of 12 filtered and 6 control units in each of the storage solutions (CPDA-1, CPD and CPD/SAGM) (Figures 10-12, respectively). Values for lactate, pO₂, pCO₂, sodium and glucose were measured, found unremarkable and are not shown here.

Values for filtered red cells were assessed at: Day 1 pre-filtration, and days 1, 7, and at outdate (35 post-filtration for CPDA-1, 28 for CPD and 42 for CP2D or CPD/SAGM), and compared to values for control (unfiltered) RBCs assessed at Days 1 (pre-storage), 7, and outdate.

These values were compared with one-way ANOVA and Newman-Keuls multiple comparisons tests. Statistically significant differences between filtered and control values were noted at outdate only for: pfHgb, wherein values for filtered product were higher. White blood cells were reduced to below the limit of detection of flow cytometry in significantly greater frequency when compared with pre-filtration samples. (Figures 10-12).

Potassium values were significantly higher in filtered units in CPD (Figure 11), but not CPDA-1 or CPD/SAGM (Figures 10 & 12).

Figure 10

Effect of CPDA-1 anticoagulated blood filtration with LAPRF System on indices of hemo- and bio-compatibility as well as residual leukocyte counts below the limit of detection of the flow cytometric assay (approximately 10⁴ WBC/unit).

N=12 for LAPRF System, N=6 for controls and residual WBC counts shown in corresponding figure (N=39)

 

Figure 11

Effect of CPD anticoagulated blood filtration with LAPRF System on indices of hemo- and bio-compatibility as well as residual leukocyte counts below the limit of detection of the flow cytometric assay (approximately 10⁴ WBC/unit).

N=12 for LAPRF System, N=6 for controls and residual WBC counts shown in corresponding figure (N=72).

 

Figure 12

Effect of CPD anticoagulated blood filtration with LAPRF System on indices of hemo- and bio-compatibility as well as residual leukocyte counts below the limit of detection of the flow cytometric assay (approximately 10⁴ WBC/unit).

N=12 for LAPRF System, N=6 for controls and residual WBC counts shown in corresponding figure. (N=34).

 

Figure 13

Effect of prion filtration on Band 3 protein

In addition, Band 3 protein was measured in 7 filtered and unfiltered units in CPDA-1. No significant difference in this measurement was found (Figure 13).

Filtration time was assessed for units in each storage solution held and filtered at either RT or 4ºC. In each case, filtration times were longer for units filtered at 4ºC (Figure 9). This effect was statistically significant for CPD and CPD/SAGM, but not for CPDA-1.

The hematocrit of red cells prior to filtration averaged 66.3+1.9% (n=24) for the combined levels within CPDA-1 and CPD with no statistically significant difference between the two; whereas, CPD/SAGM cells had a lower hematocrit averaging 60.0+1.7% (n=12).

The volume of blood filtered was not significantly different comparing CPDA-1, CPD and CPD/SAGM and, in total, averaged 266.8±19.0 (N=36) wherein the set hold-up volume averaged 57.0±3.6 mL.The loss of red cells is a function of the hold-up volume of the filter, which approximates 40 mL, and the set as a whole. Total hemoglobin averages 40 gm/unit with percent recoveries approximately 80%. (Figure 14).

Figure 14

Red cell recovery in total hemoglobin per unit and percent following filtration with the LAPRF system

Red Cell Quality Response Measures —

Red Cell Storage Study 2

 

Data for day 1 and day 2 filtration were combined because there was no significant difference (T-Test) between day 1 and day 2 as measured at the end of storage. ATP, % haemolysis, K+ and Na+ were all at acceptable levels at the end of storage for day 1 and day 2 LAPRF filtered red cells. Data is shown in Table 4.

Table 4

ATP, % haemolysis, K+ and Na+ during storage (mean and range)

	Day 1/Day 2	End of storage
ATP (µmol/gHb)	4.46 (2.85-6.68)	3.48 (2.27-7.74)
Haemolysis (%)	0.06 (0.00-0.16)	0.33 (0.25-0.47)
K+ (mmol/L)	2.3 (1.3-3.8)	34.2 (29.7-39.8)
Na+ (mmol/L)	102.6 (98.6-108.7)	80.0 (76.2-84.0)

 

                        

Red Cell Survival

After signing an IRB approved consent form, twelve volunteers meeting normal AABB donor criteria were entered into the study. A serum pregnancy test was performed on all fertile females at the time of collection and within 24 hours before re-infusion. A negative result was required to continue with the study. 

The mean single-label 24-hour in vivo recoveries were 84.4% and 85.7% for the RT and 4ºC units, respectively (Figure 15). The mean double label recoveries were 81.9% and 84.1% for the RT and 4ºC units, respectively. The overall 24-hour recoveries were 85.0% and 82.9% for the single label and double label recoveries for the twelve units.

Figure 15

Effect of storage condition prior to LAPRF System filtration on the in-vivo red cell survival after LAPRF System filtration.

The mean pre-LAPRF System leukocyte counts were <0.54 and <0.28 x 105/unit for the RT and 4ºC units, respectively. The mean post-LAPRF System yields were <0.12 and <0.14 x 105 for the RT and 4ºC units, respectively. Individual results not shown.

The mean pre-LAPRF System pfHgb values were 50 and 22 mg/dL for the RT and 4ºC units. The mean post-LAPRF System pfHgb values were 70 and 37 mg/dL for the RT and 4ºC units. The pre-storage pfHgb values were 82 and 60 mg/dL for the RT and 4ºC units. The resulting mean pre-LAPRF System hemolysis values were 0.11% and 0.03% for the RT and 4ºC units, respectively. The mean post-LAPRF System (pre-storage) hemolysis values were 0.14% and 0.08% for the RT and 4ºC units. The post-storage hemolysis values were 0.17% and 0.13% for the RT and 4ºC units. Individual results not shown.

Pre-LAPRF System K+ levels were 3.76 and 1.25 mEq/L for the RT and 4ºC units. The overall pre-LAPRF System K+ level was 2.51 mEq/L. Pre-storage post-LAPRF System K+ levels were 4.23 and 1.39 mEq/L for the RT and 4ºC units. Overall post-LAPRF System K+ level was 2.81 mEq/L. Post-storage K+ levels were 46.60 and 42.85 mEq/L for the RT and 4ºC units. The overall post-storage K+ level was 44.73 mEq/L. Individual data not shown.

This report characterizes the safety and efficacy of prion reduction of packed red cells by filtration. The data support the view that exogenous pathogenic PrP may be reduced by an average of 2.9±0.7 (mean ± standard deviation) logarithmic units or nearly 99.9% on average. Packed red cells varying in their hematocrit over the normal range for each anticoagulant/additive solution studied, can be used without untoward effect on the integrity of the red cells and with customary flow rates comparable to that seen with leukoreduction by filtration.

A consequence of the use of the Western blot assay is the requirement for a high spiking dose because the limit of detection of the infectious PrP is high. The concentration of pathogenic PrP in blood (at about 10 IU/mL) is believed to be far lower than present in brain (about 10⁹ infectious units per gram).²² This study employed levels of approximately 1 million infectious units per mL and higher. This was done because the dynamic range of the Western blot is about 3 orders of magnitude and the limit of detection approximates 1,000-10,000 IU/mL, representing a difference of about 3 logarithmic units. Indeed the average log reduction approximates our ability to detect this separation, suggesting that greater log reduction may be possible but not easily measured in this system. Infectivity experiments which are currently underway will therefore presumably reveal a higher log reduction.

The performance is not altered from lot-to-lot (Figure 3). There were significant differences between anticoagulants with CPDA-1 having greater prion reduction compared with either CPD or CPD/SAGM, however the magnitude of the differences are small (Figure 4). It is apparent that with CPDA-1 is less variable based upon a few outliers, but the medians are quite close nonetheless. Both lower filtration head height (Figure 5) and lower temperatures (Figure 6) are associated with better prion reduction and this may be due to the increase in time the blood is in contact with the filtration media. Overall, prion reduction in the range of 3 log removal may be sufficient to abrogate the transmission of prion disease because the levels in blood of diseased patients are so low.

The process of filtration should leave the blood product intact after prion removal and not subject to greater sensitivity to storage lesions. It is well known that red cells will leak hemoglobin upon storage and filtration itself may exacerbate the release of hemoglobin.²³  However from a practical perspective there are levels of hemolysis considered acceptable by the Council of Europe that is <0.8% hemolysis. These studies show that prion reduction filters do not yield excessive values (Figures 7 & 8, Table 4). Filtration time too is unaffected except in excessively high hematocrit challenges not likely to be seen in routine blood banking practice. (Figure 9)

Safety of the blood appears unaffected for all the indices of the integrity of the blood and plasma appear little, if at all, affected as a consequence of filtration and subsequent storage. This is obvious when comparing filtration with the LAPRF System with controls stored to outdate regardless of anticoagulant and additive solution (Figures 10-12). Moreover, the biocompatibility is further established by the observed comparability of band 3 protein in the presence and absence of exposure to the prion removal filter (Figure 13).

One potential added benefit relates to the observation that filtration results in a further reduction in the leukocyte count. Blood banking practices in North America, as well as Europe, recognize that levels of residual leucocytes below a prescribed value (1 million cells per unit in Europe), confers benefit to at least selected patients by averting leukocyte-mediated morbidity. However, the exact level required to abrogate all white cell mediated adverse affects in all patients is not known. Therefore, there may be an additional benefit to be ascribed to a prion reduced and previously leukoreduced blood product because the levels of residual leucocytes are reduced further by passage through the LAPRF System (Figure 10-12).

In summary these data support the position that the LAPRF System may be expected to reduce pathogenic prion from red cells by a factor of 2.9 log or 99.9% when combining the data for lot-to-lot variability and the use of all anticoagulants studied. The variability in the performance of the filter may be explained by the known variation in the extent of PrP^Sc aggregation, a phenomenon well described in the literature.²⁴ Prion removal by filtration using the LAPRF System does not impair the quality of the red cells as reflected by measures of hemo- and bio-compatibility (Figures 10-12, 13).

The 24-hour in-vivo red cell recovery means (Figure 15) are well above the FDA and Council of Europe’s requirement of achieving a mean post-transfusion survival of no less than 75% of the transfused red cells, and they are comparable to results of red cells filtered in a previously licensed Leukotrap^® Red Cell Collection and Filtration System with CP2D/AS-3 (Pall Medical).²⁵  The post-storage hemolysis values are well below the European maximum limit of 0.8%. Thus, the results of this study indicate that leukoreduced red cells filtered through the LAPRF System will have acceptable quality for transfusion.

The residual white cells were shown to be lowered from a leukoreduced blood product with the LAPRF System by counting the number of units below the limit of assay detection. However, another descriptor of these data result from applying a binomial distribution to the data from all conditions of filtration with 30" head height wherein it can be stated with 95% confidence that 97.9% of filtered products will be below 1x10⁵ WBC/unit. Similarly it can be claimed with 99% confidence that 96.9% of filtered product will contain residual leucocytes below 1x10⁵/unit.

Although the Western blot assay has limitations, so too do all approaches that can reasonably be used to determine the clinical impact of prion reduction technology. The preliminary favorable findings of a 3.7 log pathogenic prion reduction and abrogation of infectivity suggests that the efficacy calculated from the Western blot data may under-estimate the free prion removal potency of the system. The ongoing studies to corroborate the preliminary data are expected to be available by the end of 2005. In the interim, these data provide confidence that a capable prion removal technology is available that can be safely integrated into current practices of blood banking.

1. Glatzel M, Stoeck K, Seeger H, Luhrs T, Aguzzi A. Human prion diseases: molecular and clinical aspects. Arch Neurol. 2005 Apr;62(4):545-52.
2. DeArmond SJ, Prusiner SB. Perspectives on prion biology, prion disease pathogenesis, and pharmacologic approaches to treatment. Clin Lab Med. 2003 Mar;23(1):1-41.
3. Collins SJ, Lawson VA, Masters CL. Transmissible spongiform encephalopathies. Lancet. 2004 Jan 3;363(9402):51-61.
4. Collinge J. Prion diseases of humans and animals: their causes and molecular basis. Annu Rev Neurosci. 2001;24:519-50.
5. Heikenwalder M, Prinz M, Heppner FL, Aguzzi A. Current concepts and controversies in prion immunopathology. J Mol Neurosci. 2004;23(1-2):3-12.
6. Ironside JW, Head MW. Variant Creutzfeldt-Jakob disease: risk of transmission by blood and blood products. Haemophilia. 2004 Oct;10 Suppl 4:64-9.
7. Llewelyn CA, Hewitt PE, Knight RS, Amar K, Cousens S, Mackenzie J, Will RG. Possible transmission of variant Creutzfeldt-Jakob disease by blood transfusion. Lancet. 2004 Feb 7;363 (9407):417-21.
8. Peden AH, Head MW, Ritchie DL, Bell JE, Ironside JW. Preclinical vCJD after blood transfusion in a PRNP codon 129 heterozygous patient. Lancet. 2004 Aug 7;364(9433):527-9.
9. Serban A, Legname G, Hansen K, Kovaleva N, Prusiner SB. Immunoglobulins in urine of hamsters with scrapie. J Biol Chem. 2004 Nov 19;279(47):48817-20.
10. Dodd RY. Bovine spongiform encephalopathy, variant CJD, and blood transfusion: beefer madness? Transfusion. 2004 May;44(5):628-30.
11. Gregori L, McCombie N, Palmer D, Birch P, Sowemimo-Coker SO, Giulivi A, Rohwer RG. Effectiveness of leukoreduction for removal of infectivity of transmissible spongiform encephalopathies from blood. Lancet. 2004 Aug 7;364(9433):529-31.
12. Hornsey V, Drummond O, MacGregor I, Bethel H, Krailadsiri P, Smith K, Seghatchian J, Prowse C. Leucofiltration, retention/generation of soluble prion and annexin V and storage of blood components. Transfus Sci. 2000 Feb-Apr;22(1-2):75-6.
13. MacGregor I, Drummond O, Turner M, Barclay R, Prowse C. Distribution of normal prion protein in blood. Transfus Sci. 2000 Feb-Apr;22(1-2):51.
14. Zhang J, Chen L, Zhang BY, Han J, Xiao XL, Tian HY, Li BL, Gao C, Gao JM, Zhou XB, Ma GP, Liu Y, Xu CM, Dong XP. Comparison study on clinical and neuropathological characteristics of hamsters inoculated with scrapie strain 263K in different challenging pathways. Biomed Environ Sci. 2004 Mar;17(1):65-78.
15. BioReliance Corporation; 14920 Broschart Road; Rockville, MD 20850
16. Wadsworth JD, Joiner S, Hill AF, Campbell TA, Desbruslais M, Luthert PJ, Collinge J. Tissue distribution of protease resistant prion protein in variant Creutzfeldt-Jakob disease using a highly sensitive immunoblotting assay. Lancet. 2001 Jul 21;358(9277):171-80.
17. WHO Expert Committee on Biological Standardization. Guidelines on viral inactivation and removal procedures intended to assure the viral safety of human blood plasma products. Geneva; 26 to 30 November 2001: page 65.
18. Moroff G, Shomer P, Buttn L: Proposed standardization of methods for determining 24-hour survival of stored cells. Transfusion. 24:109-14, 1984.
19. Heaton WAL, Keegan T, Hanbury CM, Holme S, et al: Studies with nonradioisotopic sodium chromate. II Single and double-label 52Cr/51Cr posttransfusion recovery estimations. Transfusion. 29:703-7, 1989.
20. Heaton W, Keegan T, Holme S, Momoda G: Evaluation of 99mTechnetium/51Chromium post-transfusion recovery of red cells stored in saline, adenine, glucose, mannitol for 42 days. Vox Sang. 57:37-42, 1989.
21. Sowemimo-Coker S, Kascak R, Kim A, et al. Removal of exogenous (Spiked) and endogenous prion infectivity from red cell concentrates using a new prototype of leucocyte reduction filter. Transfusion. 2005 (Accepted for publication)
22. Fagge T, Barclay GR, Macgregor I, Head M, Ironside J, Turner M. Variation in concentration of prion protein in the peripheral blood of patients with variant and sporadic Creutzfeldt-Jakob disease detected by dissociation enhanced lanthanide fluoroimmunoassay and flow cytometry. Transfusion. 2005 Apr;45(4):504-13. 
23. Sowemimo-Coker SO. Red blood cell hemolysis during processing. Transfus Med Rev. 2002 Jan;16(1):46-60.
24. Masel J, Jansen VA. The measured level of prion infectivity varies in a predictable way according to the aggregation state of the infectious agent. Biochim Biophys Acta. 2001 Feb 14;1535(2):164-73.
25. Heaton W, Holme S, Smith, Brecher M, Pineda A, AuBuchon J, Nelson E: Effects of 3-5 log10 pre-storage leukocyte depletion on red cell storage and metabolism. Brit. J. Haem. 87:363-368, 1994.
    

Leukotrap® Affinity Filter System is Not Currently Available for Sale in the U.S. or Canada