October 31, 2005

Bethesda, MD


Topic # 1

Progress Report on FDA’s Risk Assessment for Potential Exposure to Variant Creutzfeldt-Jakob Disease (vCJD) Agent in Human Plasma-Derived Antihemophilic Factor (FVIII) Products


Issue:  FDA is developing a risk assessment for vCJD exposure in U.S. licensed plasma-derived FVIII products.  We seek the Committee’s advice how best to refine the input parameters that are used in the risk assessment model.  Additionally, we seek the Committee’s opinion on the value of the model as a basis for risk communication in view of the large uncertainties.




Two cases of presumed transfusion transmission of vCJD were reported in the United Kingdom (U.K.), in December 2003 and July 2004 (Llewelyn, Peden).  Animal models of endogenous infection have shown that TSE infectivity is associated with both plasma and cellular blood elements (Brown).  Although vCJD in recipients of plasma derivatives has not been reported, TSE’s have long incubation periods, especially after low-dose exposures.  If there is vCJD exposure risk from plasma derivatives, it would be greatest in association with products that (1) were made from plasma pools containing a donation from an individual incubating vCJD; (2) have lower potential for TSE clearance by manufacturing; and (3) are administered repeatedly for prolonged periods to treat chronic conditions.


At their February 2005 meeting, the TSEAC agreed with the general design of a risk assessment model for vCJD in plasma derivatives proposed by FDA.  FDA has been developing the model and its inputs and is now undertaking to apply this model to FVIII products.  We intend subsequently to perform vCJD risk assessments for other plasma-derived products.  In this context, FDA seeks the Committee’s advice about its current thinking, and suggestions on possible refinements with respect to important but uncertain input parameters for the risk assessment.  We also seek an overall opinion from the Committee on the utility of the model as a basis for risk communication with patients and their care providers in view of the present uncertainties both with regard to the input parameters and the interpretation of the output of the risk assessment model. 


Risk Assessment Framework


In the setting of uncertainties, risk assessments cannot predict risk precisely, but rather supply a range of possible risks and elucidate the key factors that most affect the risk.  These ranges are refined over time, as more scientific information becomes known.  The risk assessment estimates the likelihood of potential exposure to an infectious dose (derived from animal studies) of the vCJD agent for recipients of plasma-derived FVIII, but cannot, due to limitations in currently available data, predict the risk that a recipient was actually infected by the exposure or will develop clinical illness.


The risk assessment approaches used by the FDA generally follow the four-part framework and guidelines of the National Academies of Science.[1]An assessment of vCJD exposure risk in U.S. products is intended to provide:


·        A framework for more precise risk assessments in the future as additional data becomes available

·        Analysis of product characteristics (such as manufacturing and plasma pool size) that may have greater or lesser influence on margins of safety.

·        Estimation of a range of likely, best-case, and maximum potential risks of exposure to vCJD via products

·        A model that can be used to help estimate the need for and potential effectiveness of additional risk reduction measures

·        Information that would be useful for policy considerations and for medical care decisions by physicians and consumers.

·        A basis for risk communication including to health care providers, patients, and their families


Variant CJD risk assessments are limited because some of the needed information is not available, is imprecise, or is based on extrapolations from indirect sources.  In these cases, the recommended approach is to use a range of estimates based upon current knowledge and expert opinion.  Because of the uncertainties associated with these components a probabilistic model using statistical distributions to represent the information used for each variable (called an “assumption”).  The prevalence of vCJD infection is not known among blood donors in the U.S. but is estimated to be very low.  The amounts and period of infectivity in blood of people incubating vCJD are also not known, but those can be estimated based upon extrapolation from observations in animal models.  The estimated TSE clearance capacity of various manufacturing processes is based upon limited animal model studies that have not been performed for all products and that used a variety of spiking material that may not accurately reproduce the form of the TSE agent in plasma—currently not known.  The determinants of susceptibility to vCJD are largely unknown, although to date clinical illness has only occurred in people homozygous for methionine at prion protein (PrP) codon 129.  However, vCJD infection without overt clinical illness was found in the second presumptive transfusion-transmission case.  The recipient, who died from unrelated causes, was heterozygous (methionine/valine) at PrP codon 129. Our risk model has assumed that all persons might be susceptible to infection with the agent of bovine spongiform encephalopathy (BSE).


The exposure estimate is the product of all risk assessment parameters, and takes into account a number of factors including:


·        The likelihood that a donation from a vCJD infected individual is contained in a plasma pool for fractionation

·        The amount of infectivity that might be contained in a single dose of a specific product made from a contaminated pool

·        The amount of product given to a typical recipient per course of treatment or per year


Risk Mitigation by Donor Deferral


Although no vCJD blood transmission had occurred at the time, as a prudent measure to reduce risk, in 1999 the FDA recommended deferring plasma donors who might have been exposed to the vCJD agent during travel or residence in the U.K. for any total period of six months or more between 1980 and 1996.  Based on FDA guidance that was issued in January 2002, these deferrals were expanded to include Source Plasma donors who resided in France for five years or longer, donors who traveled or resided in the U.K. for any total period of 3 months or more (1980-1996), members of the military who may have consumed British beef at military establishments in Europe during certain time periods (six or more months exposure between 1980-1990 or 1980-1996 depending on where they were based), and recipients of transfusions in the U.K. since 1980.  The same deferrals applied to donors of Whole Blood (including recovered plasma), with an additional deferral based on five years of exposure in any part of Europe since 1980 (not including exposures in the U.K. after 1996).  These deferrals were intended to significantly reduce the likelihood for collection of plasma from a donor infected with vCJD but could not totally eliminate that risk.


FDA proposes to estimate risk for at least two time periods: (1) current risk taking into account donor deferral policies to reduce vCJD risk that have been in place since October 2002, and (2) the risk prior to 1999, prior to the implementation of donor deferrals to reduce vCJD risk. 


Discussion:  Elements of vCJD Risk Assessment for Plasma-derived FVIII


Four main categories of information are needed to perform the vCJD risk assessment:  the likelihood of a contaminated donation, the potential amount of infectivity in the starting material used for fractionation, the amount of infectivity cleared during manufacturing, and the amount of product received by a recipient.  Within these categories, FDA requests the Committee’s advice on selection of appropriate input assumptions for the model:


·        Prevalence estimates for vCJD in the U.K., needed to estimate possible prevalence in U.S. donors taking into account the effectiveness of donor deferrals and foreign travel/residence history for U.S. donors of Source Plasma and recovered plasma

·        The time course (after initial infection and before the onset of overt illness) and level of infectivity that may be present in plasma of a donor who is incubating vCJD (in 50-percent infectious doses [ID50])

·        The strategy for estimating TSE infectivity clearance achieved by various steps in the manufacturing process (and the complete process), including whether the infected material selected for spiking experiments adequately represented the physical state of infectivity that might be present in human plasma

·        Data needed to estimate a recipient exposure based on clinical use of product, effects of plasma pool size on the risk of contaminating end products, and whether it is feasible to incorporate the potential cumulative effect of repeated exposures (especially multiple exposures to “sub-infectious” doses of infectivity)


Issue 1: Prevalence of vCJD in the U.K. and potential prevalence of vCJD in U.S. donors exposed to British beef products.  Estimates of vCJD prevalence in the U.K. must be used to estimate the prevalence of vCJD in U.S. donors, based upon time spent in the U.K. or opportunities to have eaten British beef while serving in the U.S. military in Europe.  However, there is uncertainty how recent U.K. tonsil and appendix surveillance data and/or epidemiological models of the clinical outbreak can best be used to estimate the prevalence of vCJD in the U.K.


Question: a) What estimate should be used to reflect prevalence of vCJD in the U.K.?




Surgical surveillance data from a tonsil/appendix survey of tissues, mainly from operations on people ages 20-29, demonstrated prevalence for disease-associated PrP (PrPTSE) immunohistochemical staining of 1:4224 (three positive appendices detected among samples from 12,674 individuals (Hilton) 95% confidence interval 49-692 cases per million people in the U.K.).  However, a recent epidemiological model predicted that the current U.K. epidemic of vCJD may be much smaller than the figure suggested by the surgical survey—final prevalence not to exceed 1:197,000-1:328,000people (Boelle).  The latter model, however, does not contain any adjustment for the possible emergence of vCJD cases in individuals with either the MV or VV genotype at codon 129 of PrP.


Use of the surgical tissue surveillance data to estimate prevalence has been controversial, because it yielded results so discrepant with epidemiological predictions of a much smaller epidemic.  The discrepancy could be explained by long incubation periods of vCJD, or by the presence of subclinical vCJD (infections without illness).  Clinical and preclinical (asymptomatic) cases that may eventually progress to clinical illness are addressed in many of the predictive models.  However, the possibility of indefinitely subclinical infections (a carrier state) is not reflected in most of the recent epidemiological predictions.


Limitations of the surgical tissue surveillance data in estimating vCJD prevalence include:

·        The tissues analyzed were mostly from people of ages 20-29, an age group that may be more susceptible to vCJD (mean age at onset about 28 years). Older and younger populations have had lower prevalences, based on epidemiological data. The reason for the peculiar age distribution of vCJD remains unknown; it may reflect either a more intensive exposure to the BSE agent (for example, due to a greater consumption of some high-risk beef product) by young people or some age-related increased susceptibility to vCJD (such as more frequent pharyngitis in childhood which might have facilitated infection); both kinds of explanation have been postulated by U.K. authorities.

·        Identification of PrPTSE in lymphoid tissues by immunostaining does not necessarily mean that blood from these subjects would be infectious (or that they will develop disease).

·        On the other hand, PrPTSE staining in a few appendices might not have detected all subclinical vCJD infections. There is reason to suspect that PrPTSE staining appears in the appendiceal lymphoid tissue rather late during the incubation period of vCJD; for example, in the second presumed transfusion-transmitted  vCJD infection, tonsils and appendix of the subject were both negative for PrPTSE staining while the spleen and a cervical lymph node were positive.

·        The epidemiology and clinical course of patients with positive PrPTSE staining in appendices is unknown, since the tissue samples in the study were anonymized.  Therefore other than age range, epidemiological information is unavailable from the tissue survey.


Limitations for the prediction of vCJD prevalence based on epidemiological modeling of the clinical outbreak of vCJD in the UK include:


·        Data needed for the model are limited or not available.  For example, information is not available on the probability or quantity of exposure to the vCJD agent, time of initial infection, minimum or maximum incubation period, or age-specific susceptibility.  Because these data are limited, predictive models rely on a number of simplifying assumptions.

·        Many of the current models may not adequately estimate the number of asymptomatic vCJD cases in the population.

·        Current models, to our knowledge, do not address the potential emergence of vCJD cases in individuals heterozygous (MV) at PrP codon 129, or homozygous for VV.



Proposal: We propose using the surgical tissue surveillance data as the assumed prevalence of vCJD in the U.K. to provide a more conservative estimate for possible exposure in the risk assessment than those from epidemiological models based observed clinical cases of vCJD.  However, we propose that the risk assessment also be done using epidemiologic predictions based on diagnosed clinical vCJD cases as an alternative assumption of  prevalence (with adjustments for possible latent infections during the incubation period of illness).



Issue 2:  Prevalence of vCJD in US plasma donors with a history of extended travel or residence in the U.K., France or elsewhere in Europe since 1980


Question: How effective are current donor deferrals for geographical risk of vCJD?



The 2002 geographic deferral criteria for vCJD currently in place recommend deferring donations from US blood and plasma donors with a history of three months total travel or residence between 1980 and 1996 in the U.K., or five years of travel or residence in France from 1980 until the present.  Donors of Whole Blood (but not of Source Plasma) who have a history of five years or more of travel or residence in Europe from 1980 until the present also should be deferred.  Donor deferrals are also recommended for members of the military (as well as civilian employees and dependents) residing at certain European bases where British beef comprised a large proportion of the commissary and other on-base beef supplies.  


There are two main categories of donors of recovered plasma and Source Plasma who may have vCJD risk, but remain in the donor pool: (1) donors with a deferrable history of travel or residence who did not give an accurate or complete travel history (or who were not adequately queried), and (2) donors with some geographical exposure risk whose exposure does not meet current criteria for deferral as specified in FDA’s 2002 guidance.


Estimating the current number of new donors presenting for blood donation that should be but are not deferred for potential vCJD travel/residence risk is especially difficult.  It was reported recently that, since the vCJD policies were implemented, 0.37% of Whole Blood donors have been deferred annually for vCJD risk history. The prevalence of deferrable risk for Source Plasma donors has not been estimated through travel surveys.  Absent this information, the prevalence has been estimated from travel survey data on Whole Blood donors, with an adjustment for age.


In previously published anonymous donor surveys that described an overall prevalence of deferrable risk among Whole Blood donors, unreported risk was found in 1.9% (Williams) and 3.1% (Sanchez) of donors.  These estimates are also supported by post-donation identification of risk factors in individuals who later test positive for HIV and other transfusion transmitted infections.  No survey to validate the efficiency of the more recent 2002 geographic deferral criteria for vCJD has been conducted.  Many former repeat donors and some potential new donors, aware of the current deferral criteria for vCJD risk, are thought to have self-deferred from donation.  Because they no longer present to collection centers, it is difficult to estimate the size of this population; that uncertainty confounds the estimation of the true total number of donors deferred for vCJD risk and also the number of deferrable donors who have continued to donate while at increased risk for vCJD.


Source Plasma and recovered plasma are collected largely from repeat donors who have undergone multiple rounds of donor screening.  The American Red Cross estimated that approximately 83 percent of all units collected from 1991 through 1993 were from repeat donors (Lackritz).  Consistent with policies of the plasma industry since 1991, all donations of Source Plasma are obtained from “Qualified Donors” who have donated on more than one occasion in the last six months.  Overall, less than 20 percent of plasma is collected from first-time donors who may be at higher vCJD risk than repeat donors since they have been screened less often.


Proposal: FDA is seeking additional information on the efficiency of donor deferrals for Source Plasma. Based on the currently available previous surveys of unreported risk for other conditions (and allowing for a margin of error), we propose to estimate that the FDA-recommended deferral policy has 90% to 95% efficiency for deferring donors with the specified increased vCJD risk. 


Risk amongst persons with exposure not meeting the deferral criteria is estimated in proportion to time spent, adjusted for the time period of exposure in relation to the BSE outbreak.  Age adjusted travel survey data on Whole Blood donors will be taken also to represent Source Plasma donors.



Issue 3: Level of vCJD infectivity in human plasma


Question (a) what intravenous (IV) infectivity range (in ID50) should be selected for plasma, based on animal studies




a.       Quantitation of infectivity in FVIII model

                           i.      A FXI risk assessment model (presented to TSEAC on Feb 8, 2005) incorporated the variability and uncertainty of the available scientific information about TSE infectivity in blood using a triangular statistical distribution with a minimum value of 0.1 intracerebral (ic) ID50, a most likely value of 10 ic ID50, and a maximum value of 1,000 ic ID50  (mean of ~ 343 ic ID50).  The latter value of 1,000 was based on a report of a study that failed to demonstrate vCJD infectivity in human blood, using an assay that had a limit of detection at about 1,000 ID50/ml (Wadsworth et al).

                         ii.      Participants at a meeting on TSE risks convened by Health Canada (in Feb 2005) suggested that a maximum infectivity of 1,000 ICID50 in human blood was likely to have been an overestimation.  This level is greater than any reported to date for an animal model.

                        iii.      Re-evaluation of scientific literature revealed that levels of infectivity measured in animal models were most often in the range of 10 to 20 IC ID50 per ml of blood.  In one animal experiment a maximum level of 310 ic ID50 was observed (Casaccia).   So far as we know, no published report has ever claimed to detect any greater concentration of TSE infectivity in blood.


                       iv.      Recent experiments in primates by Lasmezas and coworkers showed that the incubation periods were the same after inoculations of primate-adapted BSE agent by the IC and IV routes, suggesting that the efficiency of infection was equivalent by both routes (Herzog).  However, experimental TSE infections in rodent models have generally demonstrated that the IC route is 5 to 10 fold more efficient than IV. (Brown)


Proposal:  FDA proposes that the FVIII model will use a statistical distribution for infectivity, with a minimum value of 0.1 IC ID50, a most likely value of 10 IC ID50 and a maximum value of 310 IC ID50.  Since the BSE agent in primates may more closely approximate the human situation than rodent models, we propose to model the IC/IV ratio for infectivity over a range of 1 and 5.



Question (b) Is there sufficient evidence available to estimate when during the incubation period of vCJD human plasma is infectious?




Animal studies monitoring the appearance of infectivity after exposure have shown that blood infection can first be detected about halfway through the incubation period).  The incubation period for human vCJD is not known, which makes animal estimates difficult to extrapolate.  The year of travel-related exposure among donors is also not known.  The two transfusion transmitted cases of vCJD in the U.K. demonstrate that blood of one donor was infectious at least 3 years prior to the onset of symptoms and the second at least 18 months. 


Proposal: Because of uncertainties about the incubation periods of food-borne vCJD and the time during the incubation period at which infectivity appears, FDA proposes to adopt a conservative approach, and assumes plasma to be potentially infectious throughout the incubation period.



Issue 4:  Selection of appropriate values for clearance of TSE infectivity from plasma-derived FVIII

This parameter is used to estimate the clearance of vCJD by manufacturing processes and is one of the most important determinants of risk. 


Questions:  Does the Committee agree with FDA’s proposed approach for estimating clearance of vCJD infectivity from FVIII products by manufacturing processes?  What experiments might enable refinement of these estimates and allow comparison of clearance offered by various steps in the methods used to manufacture plasma-derived FVIII?



The potential removal of TSE infectivity clearance from products is estimated based upon evaluation of steps in manufacturing that have been shown by pilot studies with animal models to clear spiked or endogenous infectivity (Table).  Each FVIII product is unique in its details of manufacturing, and available generalizable data are limited, so that identical levels of clearance cannot automatically be assumed for different products, even those using similar manufacturing steps  (Foster).  


The estimates of TSE clearance for specific products are incomplete, because published, peer-reviewed detailed studies for specific steps in the manufacture of FVIII are not available for all products, nor is a comprehensive set of data available to FDA.  In addition, there remain uncertainties about the general relevance of studies in which various materials prepared from TSE-infected brain tissues are spiked into a pilot process for manufacturing plasma derivatives.  Clearance of TSE agents extracted from infected brain tissues may not precisely reflect partition of endogenous TSE agents, since the physical form of blood-borne form of infectivity is still unknown.  High-titer preparations from TSE-infected brain (homogenates, microsomal fractions, scrapie fibrils, or other) have been used, none of which is likely to completely reflect the properties of TSE agents in blood.  Clearance has been measured either as reduction in infectivity measured by bioassay in animals or as reduction in PrPTSE measured by various assays, which makes direct comparison across studies and across products difficult.


Until clearance data from laboratory studies of individual steps as well as the entire manufacturing process for each marketed product becomes available, we believe it might still be informative to estimate the range of clearance based on the presence of specific manufacturing steps that have been studied for FVIII.  Many of these studies have not been peer-reviewed or submitted to FDA.  A review of the literature, and of unpublished information available to FDA permits rough estimates of infectivity clearance values offered by specific steps in the manufacture of Factor VIII (Table). 

Several methods are commonly used to manufacture FVIII.  These include cryoprecipitation, PEG precipitation, ion exchange chromatography, and affinity chromatography using heparin or monoclonal antibodies to VWF (followed by elution of FVIII).  The table summarizes experimental data for various process steps that are common among many FVIII products.  Based on the available data, cryoprecipitation does not appear to be a robust clearance step  (one log10 or less prion removal). Other single steps such as PEG precipitation have consistently demonstrated 2-3 log10 of clearance for spiked TSE infectivity, and 3-4 log10 clearance for ion-exchange chromatography and monoclonal affinity steps, respectively.  Data currently available to FDA suggest, as an estimate, that FVIII products lacking affinity purification are likely to have less clearance, whereas plasma products manufactured using an affinity purification step are likely to have more TSE infectivity clearance.


Proposal:  FDA proposes to model three clearance values for the risk assessment model: 2, 5, and 8 logs, to represent likely minimal, likely mid-range and likely maximal clearance of the vCJD agent from products manufactured using a variety of methods. 



Issue 5:  Recipient use of FVIII


Several scenarios for use of FVIII by typical recipients over a given time period, in combination with assumptions, will be incorporated in the FDA model to estimate potential exposures to the vCJD agent.


Question (a) What data should be used to estimate how much FVIII is used by typical patients?


The CDC has provided summary data on patient utilization of  plasma-derived FVIII products.   CDC Hemophilia Treatment Center (HTC) data were collected for approximately 3000 hemophilia patients from 1993 through 1998.  Incorporation of the CDC data into the model should enhance the precision of the risk estimates.


Proposal:  We propose using CDC HTC data to estimate use of plasma-derived FVIII by subgroups of hemophilia patients with more severe disease and frequent prophylactic treatments, less severely affected patients treated intermittently, and people with severe von Willebrand’s disease.



Question (b) What is the effect of plasma pool size (i.e. number of donors per final product) for FVIII recipients treated repeatedly?


Use of larger plasma pools increases the risk that a lot of product will be infected with any infectious agent potentially present in the donor population if the agent is not cleared from the product during manufacturing.  Lynch et al. estimated such risks for a range of plasma pool sizes (Lynch).  They concluded that, while risk of exposure increases with pool size, repeated treatments attenuate this effect because exposure risk is also increased by repeated treatments.  In the case of treatment for hemophilia using plasma-derived FVIII, most patients received many doses over time, and thus the effect of plasma pool size on potential vCJD exposure should be minimal.


Proposal:  FDA proposes to estimate plasma pool size as a range, between 20,000 and 60, 000 donations, with a bimodal distribution to reflect expected Source Plasma and recovered plasma pool donor numbers.  FDA will seek additional data from plasma fractionators to clarify the distribution of pool sizes based on actual practices.



Question (c) Can a possible cumulative effect from repeated exposures to low doses of vCJD agent be incorporated into the risk model? 



Two hamster scrapie studies suggested that the risk of infection after repeated exposures to low doses of infectivity might be cumulative (Diringer, Jaquemot).  In the latter study, repeated intraperitoneal injections of low doses of infectivity were given once, twice, or 5 times weekly for 200 days.  The repeated low-dose injections resulted in higher disease frequency than when an equivalent dose or even a larger dose was given only once. 


A subset of people with Hemophilia A may receive, or may have in the past received prophylaxis with plasma-derived FVIII twice or three times a week.  Other patients may receive only intermittent treatments as needed.  Both kinds of patients might have repeated, low-dose exposures to the vCJD agent.  It is not feasible to model all possible scenarios since individual patient treatment regimens vary so much in their frequency and duration.  


FDA proposes that, to allow for the theoretical possibility of cumulative effects, the model provide a cumulative risk for a 1-year period.



Issue 6:  Risk communication based on risk assessment estimates of potential exposure to vCJD in US recipients of FVIII.


In the United Kingdom, where all plasma-derived FVIII was manufactured from U.K. donations until 1999, the risk to recipients was deemed sufficient to trigger risk communication and risk management measures.  These included direct communication between health care providers and their patients to explain risk, detailed website information for patients and healthcare providers, and precautionary recommendations to patients to inform their dentists and surgeons of their increased vCJD risk.  While the UK exposure and risk are likely higher, and the uncertainties in any risk estimate will be extremely large, FDA believes that the US FVIII risk assessment may be useful in providing risk estimates to consumers and health care providers, and in formulating policy.



Question to the Committee: Given the present scientific uncertainties in the underlying assumptions of the FVIII risk assessment, does the Committee believe that the risk assessment model could provide a useful basis for risk communication to patients, their families, and healthcare providers?



1.      Brown, P., et al.  Further studies of blood infectivity in an experimental model of transmissible spongiform encephalopathy, with an explanation of why blood components do not transmit Creutzfeldt-Jakob disease in humans.  Transfusion 39:1169-78, 1999.

2.      Boelle, P.-y. et al.  Modeling the epidemic of variant Creutzfeldt-Jakob disease in the UK based on age characteristics: updated, detailed analysis.  Stat. Meth. Med. Res. 12:221-3, 2003.

3.      Casacchia, P. et al.  Levels of infectivity in the blood throughout the incubation period of hamsters peripherally injected with scrapie.  Arch. Virol. 108:145-9, 1989.

4.      Cervenakova, L., et al. Factor VIII and transmissible spongiform encephalopathy: the case for safety.  Haemophilia 8:63-75, 2002.

5.      Diringer, H. et al.  Effect of repeated oral infection of hamsters with scrapie.  J. Gen. Virol. 79:609-612, 1998.

6.      Foster, PR et al.  Studies on the removal of abnormal prion protein by processes used in the manufacture of human plasma products.  Vox Sang 78:86-95, 2000.

7.      Hilton, D.A. et al.  Prevalence of lymphoreticular prion protein accumulation in UK tissue samples.  J. Pathol. 203:733-9, 2004.

8.      Jaquemot, C. et al.  High incidence of scrapie induced by repeated injections of subinfectious prion doses.  J. Virol. 79:8904-8, 2005.

9.      Lackritz, E.M. et al.  Estimated risk of transmission of the human immunodeficiency virus by screened blood in the United States.  NEJM 333:1721-5, 1995.

10.  Herzog, C. et al.  Tissue distribution of bovine spongiform encephalopathy agent in primates after intravenous or oral infection.  Lancet 363:422-28, 2004.

11.  Llewelyn, C.A. et al.  Possible transmission of variant Creutzfeldt-Jakob disease by blood transfusion.  Lancet 363:417-21, 2004.

12.  Lynch, T., et al.  Considerations of pool size in the manufacture of plasma derivatives.  Transfusion 36:768-9, 1996.

13.  Peden, A.H. et al.  Preclinical vCJD after blood transfusion in a PRNP codon 129 heterozygous patient.  Lancet 364:527-9, 2004.

14.  Sanchez, A.M. et al.   Prevalence, donation practices, and risk assessment of blood donors with hemochromatosis.  JAMA 286:1475-81, 2001.

15.  Williams, A.E. et al.  Estimates of infectious disease risk factors in US blood donors.  JAMA 277:967-72, 1997.


      Prion Removal Factors for FVIII


Removal Steps


PEG- precipitation

Ion Exchange Chromatography

Monoclonal Antibody Affinity Purification













2.2, 3.0










1.7 – 3.3








Estimated Reduction for each step






Contribution of each step

Not Effective

Contributes to Removal

Contributes to Removal

Significant Removal

Total contribution of the process steps

Manufacturing processes demonstrate a significant capacity to remove TSE infectivity or TSE-associated prions



1)      Foster et al., Vox Sang. (2000) 78: 86

2)      Lee et al. J. Virol Method (2000) 84:77

3)      Lee et al. Transfusion (2001) 41: 449

4)      Rohwer / Baxter & ARC (PPTA presentation)

5)      Vey et al., Biologicals (2002) 30: 187

6)      Brown et al., Transfusion (1999) 39: 1169

[1] Risk Assessment in the Federal Government: Managing the Process. 1983.  Washington DC, National Academy Press