Background and Objectives This paper reports the results to 1 March 2006 of an ongoing UK study, the Transfusion Medicine Epidemiological Review (TMER), by the National CJD Surveillance Unit (NCJDSU) and the UK Blood Services (UKBS) to determine whether there is any evidence that Creutzfeldt–Jakob disease (CJD), including sporadic CJD (sCJD), familial CJD (fCJD), and variant CJD (vCJD) is transmissible via blood transfusion.
Materials and Methods Sporadic CJD and fCJD cases with a history of blood donation or transfusion are notified to UKBS. All vCJD cases aged > 17 years are notified to UKBS on diagnosis. A search for donation records is instigated and the fate of all donations is identified by lookback. For cases with a history of blood transfusion, hospital and UKBS records are searched to identify blood donors. Details of identified recipients and donors are checked against the NCJDSU register to establish if there are any matches.
Results CJD cases with donation history: 18/31 vCJD, 3/93 sCJD, and 3/5 fCJD cases reported as blood donors were confirmed to have donated labile components transfused to 66, 20, and 11 recipients respectively. Two vCJD recipients have appeared on the NCJDSU register as confirmed and probable vCJD cases. The latter developed symptoms of vCJD 6.5 years and 7.8 years respectively after receiving non-leucodepleted red blood cells (RBCs) from two different donors who developed clinical symptoms approximately 40 and 21 months after donating. A third recipient, given RBC donated by a further vCJD case approximately 18 months before onset of clinical symptoms, had abnormal prion protein in lymphoid tissue at post-mortem (5-years post-transfusion) but had no clinical symptoms of vCJD. CJD cases with history of transfusion: Hospital records for 7/11 vCJD and 7/52 sCJD cases included a history of transfusion of labile blood components donated by 125 and 24 donors respectively. Two recipients who developed vCJD were linked to donors who had already appeared on the NCJDSU register as vCJD cases (see above). No further links were established.
Conclusion This study has identified three instances of probable transfusion transmission of vCJD infection, including two confirmed clinical cases and one pre- or sub-clinical infection. This study has not provided evidence, to date, of transmission of sCJD or fCJD by blood transfusion, but data on these forms of diseases are limited.
Until 2004, it was generally accepted that Creutzfeldt–Jakob disease (CJD) had not been transmitted by blood transfusion. Preliminary findings from sheep studies indicate that bovine spongiform encephalopathy (BSE) and scrapie can be transmitted by blood transfusion [1,2]. It is vital to find out whether this also applies to human transmissible spongiform encephalopathies (TSEs) and, in particular, variant CJD (vCJD). The UK is the only country where a significant outbreak of vCJD has occurred and is in a unique position to study this question which has important implications for public health policy. The results reported in this paper are from a study which is being carried out, with ethical approval, to investigate whether or not there is any evidence for the transmission of any type of CJD (sporadic, familial and variant) by blood transfusion.
Materials and methods
A surveillance system for CJD, the National CJD Surveillance Unit (NCJDSU), was established in the UK in 1990 with the aim of identifying all cases of CJD in the UK. The methodology of this study has been described previously , but in brief involves referral of suspected cases to the Unit from targeted professional groups, including neurologists and neuropathologists, review of suspects by a neurologist from the Unit and review of investigation results and neuropathological material when available. Cases are classified according to standard diagnostic criteria [4,5]. Onset of clinical symptoms for vCJD cases are estimated to the nearest month by NCJDSU on the basis of available clinical information. Details of past medical history, including blood donation or transfusion, are obtained from the family of suspected cases. Following the identification of vCJD in 1996 a collaborative study, the Transfusion Medicine Epidemiological Review (TMER), was established between the NCJDSU and UK Blood Services (UKBS) to search for evidence of transfusion transmission of CJD. The study was granted ethical approval by the local Research Ethics Committee.
Notification of CJD cases with a history of donation
Sporadic CJD (sCJD) and familial CJD (fCJD) cases with a history of blood donation are notified to UKBS retrospectively. For vCJD, all patients who are old enough to have donated blood (> 17 years of age) are notified to UKBS at diagnosis, whether or not there is a known history of blood donation. Upon receipt of notification from the NCJDSU, a search is made for donor records. Current computer databases and archived records (computerized and paper-based records where appropriate) based at individual blood centres are searched using name, date of birth, and previous addresses as identifiers. For CJD cases reported as blood donors, information on dates and places of donation is also used to help locate past donor records. Where donor records are found, all components produced and issued to hospitals are identified and their fate determined, as recorded in hospital blood transfusion laboratory records. Recipient details are then checked against the NCJDSU register to establish if there is a match between these individuals and patients who have developed CJD. Recipients details are also flagged with the Office for National Statistics (ONS) to establish date and certified cause of death.
Plasma used for UK fractionation
Independent of this study, and for regulatory reasons, all plasma derived from donations made prior to a diagnosis of vCJD is notified to UK fractionators so that appropriate actions can be taken.
Notification of CJD cases with a history of transfusion
Information provided by relatives of CJD cases with a previous transfusion history is also passed to UKBS, who then liaise with appropriate hospital blood transfusion laboratories. The laboratories identify whether transfusions took place at the time and place indicated, and, if so, identify the components transfused to the case. Details of donation numbers, component type and date transfused are passed back to the local blood centre and this information is used to identify the donors. Donor details are checked against the NCJDSU register and with ONS. For both donors and recipients, searches before 1980 are now impractical as most hospital records are no longer extant.
Related public health measures
The Department of Health (England) set up a committee (the CJD Incidents Panel) in 2000 to advise health authorities on the management of incidents where patients may have been put at risk of CJD through medical procedures. Cases where a blood donor or blood recipient has later developed vCJD have been referred to the Incidents Panel for consideration and further actions have been recommended and implemented. These further actions are outside the scope of this study and are not reported in this paper.
vCJD cases with history of blood donation
Identification of donors
A total of 150 vCJD cases (out of a total of 160 cases on the NCJDSU register) who were old enough to have been potential blood donors have been notified to UKBS as of 1 March 2006. Of these, 31 of 150 (21%) were reported to have been blood donors at various times in the past, although there is variation in the details of available information and the confidence of families in donation history.
Donor records were found for 24 vCJD cases, comprising 20 reported by relatives as blood donors and four additional cases with no reported donation history. Of these, 18 vCJD cases (12% of the total eligible to donate blood) were confirmed to have donated labile blood components, with the number of components made and issued for use in UK hospitals ranging from 1 to 14 per donor. Six vCJD cases were registered as donors, but had not donated labile blood components. Two of these had never attended sessions, three were deferred (due to past medical history, low haemoglobin value and illness, respectively) and one case had donated plasma for fractionation only (made from a single donation from which the red cells were discarded).
The search for donor records was negative in 11 of 31 (35%) vCJD cases reported as putative donors (three of whom allegedly donated well before the onset of the BSE epidemic in the 1980s). The information provided in these negative cases was minimal, except in one case where relatives were confident that regular donations (up to 50) had been made in the years leading up to 1993. Despite extensive searches no records were found; moreover, blood collection sessions had never been made at the purported venue. No explanation has been found for the lack of records, although discrepancies in some of the details given suggest that the history was not as certain as initially thought.
Labile components issued to hospitals
Sixty-six labile components originating from 18 donors were issued to UK hospitals over the period 1981–2004 and transfused to patients according to blood transfusion laboratory records. A further nine components issued between 1982 and 1996 could not be traced by the relevant hospital. Table 1 gives the number of recipients transfused by year and the type of blood component transfused. Fifty-six recipients (85%) received red cells or whole blood, seven (11%) were transfused with labile plasma components or derivatives and three (4%) received pooled platelets made according to UK specifications in which the buffy-coat preparation containing platelets from the implicated vCJD donor was pooled with buffy coats from three other donors and resuspended in plasma from one of the four donations. Nearly half of the red cell recipients received red cells that had been leucocyte-depleted by pre-storage filtration to < 5 × 106 leucocytes per unit (in 99% of units with 95% statistical confidence according to UK guidelines ) after the introduction of universal leucocyte depletion of the UK blood supply in 1999.
Table 1. Recipients of blood donated by variant Creutzfeldt–Jakob disease cases by year and blood component transfused (n = 66)
Year of transfusion
Blood component transfused
Number of recipients
Red cells with buffy-coat (containing most of the platelets and white cells) removed by centrifugation and physical separation.
Red cells leucocyte-depleted by pre-storage filtration to < 5 × 106/unit according to UK guidelines .
Patient identifiers are available for 66 recipients who received blood from 18 different donors who went on to develop vCJD. None of the 66 recipients had themselves donated blood between receiving their transfusion and early 2004 when the UKBS implemented a policy of excluding all donors transfused in the UK since 1 January 1980. It is of note that 41 (62%) recipients were aged over 60 years at the time of transfusion and were not eligible to donate. All living recipients (n = 26) have been informed of their risk and advised not to donate blood, tissues or organs. Three instances of probable transfusion transmitted vCJD infection have occurred, including two confirmed clinical cases and one pre- or subclinical infection. Of these, two cases have died, and one is still alive (see succeeding discussion). Figures 1 and 2 show the survival period for dead (transfusion to death) and live recipients (transfusion to 1 March 2006) of vCJD components, respectively, according to the interval between transfusion and onset of clinical symptoms in the donor.
Forty recipients (61%) are known to be dead, with mean age at death 66 ± 19 years. Table 2 gives the time and cause of death as stated on death certificates for the recipients known to have died. Around half (n = 21) of the dead recipients died within a year of receiving their transfusion, with only seven surviving for more than 5 years. Two recipients, who died 4 months and 14 months, respectively, after transfusion had ‘dementia’ recorded on the death certificate, but examination of case notes indicated that neither case had features to suggest vCJD. All the other recipients were certified as dying of causes unrelated to vCJD, except for a recipient whose cause of death on the death certificate was recorded as ‘1A dementia and II. prostate cancer’ and was later confirmed neuropathologically as suffering from vCJD . This patient, who had received a transfusion of red cells 6.5 years before onset of clinical symptoms, was a methionine homozygote at codon 129 of the human prion protein gene (PRNP). The case that donated to this individual also had a neuropathological diagnosis of vCJD, with clinical onset approximately 40 months after donating. In a second red cell recipient (of a different donor who developed clinical symptoms approximately 18 months after donating and was later diagnosed with neuropathologically confirmed vCJD), protease-resistant prion protein (PrPres) was detected in the spleen and one lymph node (but not in the brain) at post-mortem . This recipient, who died 5 years after transfusion without any clinical symptoms of vCJD, was a codon 129 PRNP heterozygote and is thought to represent pre- or subclinical infection.
Table 2. Cause of death of variant Creutzfeldt–Jakob disease recipients known to have died (n = 40)
Interval from transfusion to death
Number of recipients
Cause of death
Confirmed variant Creutzfeldt–Jakob disease case .
PrP positivity in lymphoid tissue, pre- or subclinical vCJD infection .
Twenty-six recipients (39%) are alive as of 1 March 2006 with a mean age of 63 ± 19 years. Table 3 shows the number of live recipients according to the time elapsed since transfusion, along with their current age, component transfused and the interval between donation and onset of clinical symptoms of vCJD in the donor. Fifty per cent of live recipients were transfused with components from vCJD donors whose donations were made within 20 months of clinical onset, in seven cases around the time of development (n = 3) or shortly after (n = 4) the first signs of clinical illness. These cases would have appeared healthy when attending donor sessions and passed the normal medical checks as being fit to donate. Sixteen recipients have survived longer than 5 years, with six surviving > 10 years (one for over 18 years). These patients, mean age currently 61 ± 19 years, were given blood from donors who developed vCJD symptoms at intervals ranging from around 5 months to 191 months after making the donation (see Table 3). Recently, a diagnosis of probable vCJD has been made in one of these surviving recipients who had received a transfusion of red cells 7 years and 10 months before onset of clinical symptoms . The donor of this third probable transfusion-transmitted vCJD infection developed vCJD approximately 21 months after the donation, and the recipient is a codon 129 PRNP methonine homozygote.
Table 3. Live recipients of labile blood components donated by variant Creutzfeldt–Jakob disease cases (n = 26)
Interval between blood donation and onset of clinical symptoms in donor (months)b
As at 1 March 2006.
A negative interval denotes that donation was made by individual while (retrospectively recognized) clinical symptoms were present.
Probable variant Creutzfeldt–Jakob disease case .
1 – < 2 years
Red cells (leucodepleted)
Red cells (leucodepleted)
2 – < 3 years
Red cells (leucodepleted)
3 – < 4 years
Red cells (leucodepleted)
Red cells (leucodepleted)
Red cells (leucodepleted)
4 – < 5 years
Red cells (leucodepleted)
Red cells (leucodepleted)
Red cells (leucodepleted)
5 – < 6 years
Red cells (leucodepleted)
Red cells (leucodepleted)
Red cells (leucodepleted)
Red cells (leucodepleted)
6 – < 7 years
7 – < 8 years
8 – < 9 years
9 – < 10 years
> 10 years
Plasma for UK fractionation
Twenty-five units of plasma originating from 11 different donors, bled between 6 months and 17 years, 11 months before onset of clinical vCJD symptoms, were supplied for UK fractionation during the period 1986–1998. Product batches manufactured from 23 plasma units derived from nine donors have been traced. The fate of batches of product derived from the two remaining plasma donations, from two different donors, has not yet been traced, and this search is still ongoing. Table 4 lists the plasma products derived from the 23 traced donations and the number of batches implicated, divided into risk categories as used in the plasma product notification exercise (http://www.hpa.org.uk/infections/topics_az/cjd/Recommendations.pdf). The fate of batches of products has not been traced to individual recipients as part of this study. It is known, however, that haemophilia centres have traced the ultimate fate of the batches of factor VIII. It is also known that no case of vCJD has been identified in a patient with haemophilia in the UK.
Table 4. Product batches made by UK fractionators derived from plasma donated by individuals who later developed variant Creutzfeldt–Jakob diseasea,b
Ninety-three cases of sCJD identified between 1980 and 2000 were reported to have been blood donors, with only 38 reported to have donated from 1980 onwards. Donation records for most sCJD cases were untraceable since most dated back many years before 1980, in some cases to the 1940s. Donation records were found for eight sCJD cases, but only three had actually donated labile blood components for hospital use (one with 18 recipients, and one each with one recipient) which could be traced to recipients. A total of 20 recipients were transfused between 1995 and 1999 with components from these three donors who went on to develop sCJD between 1 and 5 years after donation. Of these, 11 (55%) received red cell components, eight recipients (40%) received platelets and one (5%) received fresh frozen plasma.
As of 1 March 2006, 12 recipients are confirmed dead with a mean age at death of 74 ± 15 years. Of these, five died soon after transfusion (four within a week, and one 2 months later) and seven survived for between 1 and 8 years after receiving their transfusion before dying of a variety of non-CJD-related causes (cerebrovascular accident/stroke, n = 3; acute myeloid leukaemia, n = 3, general debility/old age, n = 1). Seven recipients are not known to be dead from ONS flagging to date, and are therefore presumed to be alive. The mean age of these seven recipients is 58 ± 19 years. The time elapsed since their transfusion ranges from 7 to 9 years. The fate of a further recipient is unknown. None of the sCJD recipients identified as having received blood from donors who went on to develop sCJD have appeared on the NCJDSU register to date.
fCJD cases with history of blood donation
Donation records were found for three out of five cases of fCJD identified between 1992 and 2000, all reported to have donated blood after 1980. These three cases had all donated labile blood components (one with five recipients, one with four recipients and one with two recipients) for hospital use which could be traced to individual recipients. A total of 11 recipients were transfused between 1977 and 1992 with labile components from these three donors who went on to develop fCJD between 1 and 15 years later. Nine of the 11 (82%) recipients received red cell components (whole blood n = 6, red cells n = 3) while two received platelets.
Five of 11 recipients identified have since died with a mean age at death of 75 ± 6 years. Three of these survived for 3, 10 and 17 years after transfusion before dying of non-CJD-related causes (cancer, n = 2; bronchopneumonia, n = 1); and two died of cancer shortly after receiving their transfusion. Three recipients are not known to be dead from ONS flagging to date, and are therefore presumed to be alive. The mean age of these three recipients is 44 ± 20 years. The time elapsed since their transfusion ranges from 13 to 21 years. The fate of a further three recipients is not known. None of the fCJD recipients identified as having received blood from donors who went on to develop fCJD have appeared on the NCJDSU register to date.
vCJD cases with history of transfusion
Eleven vCJD cases were reported to have received past blood transfusions between 1962 and 1999. A further case received a blood transfusion after onset of illness. This case is excluded from further analysis. For two cases, hospital records showed that they had not been transfused. No hospital records could be found for another two cases reported to have been transfused in 1962 and 1971, respectively. Hospital transfusion records were found for seven vCJD cases (64% of those reported as transfused) who had been transfused with components donated by 125 donors (121 identified), with one vCJD case, who also received a solid organ transplant, receiving components from 103 donors. The identity of four donors who donated red cell/whole blood components to two cases (case 2 and case 7, see Table 5) is unknown. Table 5 shows the transfusion date, number of donors and blood components donated, and the interval from transfusion to onset of clinical symptoms of vCJD in these seven recipients. These cases had been exposed to between two and 103 donors, respectively (NB search for donors to case 6 is incomplete). To date, one donor who gave red cells to case 5 and another donor who gave red cells to case 6 are also registered on the NCJDSU database as vCJD cases. These are the donors of the two clinical cases of transfusion-transmitted vCJD referred to previously (see vCJD cases with history of donation).
Table 5. Donors (n = 125) of labile blood components given to variant Creutzfeldt–Jakob disease casesa (n = 7) with identifiable past hospital transfusion records
Number of donors of labile blood components transfused
Blood component donated to vCJD recipient
Interval from transfusion to onset of illness
Two of these cases linked to donors already on the National CJD Surveillance Unit (NCJDSU) register as vCJD cases [7,9].
Component details traced, but donors not identifiable.
Timing of clinical illness excludes blood transfusion as the source of infection in this case.
One of the donors already on NCJDSU register as vCJD case, others presumed not to be source of infection.
One donor already on NCJDSU register as vCJD case. Search for 40+ donors to Case 6 not complete, as of 1 March 2006.
Fifty-two cases of sCJD identified between 1980 and 2000 were reported to have received a blood transfusion, of which 28 received a transfusion after 1980. Transfusion records were found for seven sCJD cases transfused between 1984 and 1997. Donor details were found for 24 donors who donated components transfused to these seven sCJD cases. One of these donors is known to have died, with a cause of death not related to CJD. Twenty donors are not known to have died from ONS flagging to date, and are therefore presumed to be alive. The fate of a further three donors is not known. The mean age of the donors presumed still alive is 51 ± 9 years. None of the traced donors who gave blood to patients who were subsequently diagnosed with sCJD have appeared on the NCJDSU register to date.
fCJD cases with history of transfusion
One case of fCJD identified in 1992 was reported to have received three blood transfusions in 1965, 1970, and 1987 none of which could be traced.
This study has identified three instances in which a recipient of a transfusion derived from a ‘vCJD’ donor has developed infection with vCJD, including two clinical cases and one pre- or subclinical infection [7–9]. These are three different donor/recipient pairs. In view of the small size of the total at-risk recipient population (n = 66) and the background mortality rate for vCJD in the general UK population (0·24/million/annum), these observations provide strong evidence that vCJD can be transmitted from person to person through blood transfusion. This finding has had important implications for public health policy nationally and internationally.
The risk of developing vCJD infection in the surviving recipient population is significant but cannot be precisely estimated because of variables including the timing of blood donation in relation to clinical onset in the donor, the influence of the codon 129 genotype of donor and recipient and the effect of the introduction of leucodepletion in 1999. Furthermore, the currently observed number of infections in the recipient population may be an underestimate as some surviving recipients may yet develop vCJD and there is limited available information on the outcome in the cohort of deceased recipients; a significant proportion of these individuals may not have survived long enough to express clinical disease even if infected. The minimum incubation period in CJD transmitted from person to person by a peripheral route is 4·5 years in kuru and growth-hormone-related CJD [10,11] and only nine deceased recipients survived for longer than this period. An investigation of the hospital records of the deceased recipients is underway, and to date, none had clinical features of vCJD pre-mortem. However, the identification of the individual with ‘preclinical’ vCJD infection was dependent on post-mortem examination of peripheral lymphoreticular tissues, and, to date, no equivalent tissues have been available in the deceased transfusion recipients. Extrapolating from the three observed infections in the total recipient population is likely to lead to an underestimate of the overall risk of transfusion transmission of vCJD, although the introduction of leucodepletion in 1999 may have reduced the risk to recipients transfused after this date.
A further important variable in estimating individual risk is the time from blood donation to clinical onset in the donor and, although evidence from animal studies in relation to this issue is conflicting [12–14], it is likely that an extended gap between blood donation and clinical onset in the donor will reduce the risk of transfusion transmission. All tested clinical cases of vCJD have been methionine homozygotes at codon 129 of PRNP, but the individual with ‘pre-clinical’ transfusion transmitted infection was heterozygous at this locus , indicating that individuals with this genotype are susceptible to secondary infection with vCJD. Except for the three cases infected through blood transfusion, the codon 129 genotypes of the recipient population are not known. Although the relative risk of secondary infection in relation to the codon 129 genotype is uncertain, a recent study in a transgenic mouse model suggests that individuals with all human codon 129 genotypes may be susceptible to secondary infection with vCJD, with a hierarchy of risk from methionine homozygotes to heterozygotes to valine homozygotes . Risk may vary according to genetic background, but it cannot be assumed that some recipients will possess an absolute genetic barrier to infection.
The analysis of vCJD cases with a history of blood transfusion has identified over 100 donors to these cases, although the great majority were linked to one vCJD case who had undergone an organ transplant. A risk assessment has suggested that these donors are themselves at significant risk of developing vCJD and these individuals have been informed of this risk and have been advized not to act as blood or organ donors. To date, none of these individuals have developed vCJD, with the exception of the two donors linked to the two clinical cases of vCJD described above.
Plasma derived from vCJD cases has been used in the production of plasma derived products, including clotting factors and immunoglobulin. To date, there is no evidence that vCJD has developed in a recipient of these products. However, the potential incubation period from a presumed low dose exposure by a peripheral route may be prolonged and current observational data cannot exclude the possibility of transmission of vCJD through plasma products. A risk assessment carried out in the UK suggested that, on worst case assumptions, some plasma products could be associated with an additional risk of developing vCJD in relation to the background population risk through exposure to BSE and since 1999 plasma for the production of plasma products has been imported to the UK from other countries. In contrast to labile blood components, plasma products are manufactured using a production process, some steps of which may reduce TSE infectivity . Together with the estimated relatively low levels of initial infectivity in plasma used in fractionation, the risks from plasma products are probably much lower than the risks from transfusion of labile blood components.
An important question raised by the evidence of transfusion transmission of vCJD is whether other human TSEs may be transmissible through this mechanism. Cumulative evidence in sCJD over many years does not suggest that sCJD is transfusion transmitted . Case-control studies have not demonstrated an increased risk of sCJD through a past history of blood transfusion [18–20] and lookback studies have not linked blood transfusions derived from sCJD blood donors to sCJD cases [21,22]. However, the case-control methodology may be compromised by control selection (some studies of CJD used hospital based controls) and this type of research cannot exclude rare transfusion transmission events. The lookback study of sCJD (and fCJD) from the TMER study provides only limited evidence, not least because of the difficulties in tracing blood donations made years or decades in the past, and there is only one study of sCJD equivalent to the TMER study of vCJD . This has not provided evidence of transfusion transmission of sCJD, despite prolonged periods of follow-up in a proportion of cases. There is however a need to continue the study of blood transfusion and sCJD, and this will become of particular importance should highly sensitive tests for infectivity in blood be developed . There is recent evidence of prion protein deposition in peripheral tissues in sCJD, including muscle . However, the increased lymphoreticular involvement in vCJD in comparison to sCJD may be associated with an increased risk of transfusion transmission of vCJD.
The TMER study has provided compelling evidence that vCJD is transmissible through blood transfusion, representing the first evidence of transmission of human TSEs through this route and via material sourced from a peripheral tissue rather than the high level infectivity tissues of the central nervous system. The manifest implications for public health of transmission transfusion of vCJD have led to measures to minimize the risk from blood transfusion  and plasma products derived from cases incubating vCJD and many of these actions were taken years in advance of the evidence for transfusion transmission, both in the UK and many other countries. Although there is uncertainty about the potential for transfusion transmission of vCJD to lead to a self-sustaining epidemic, the introduction of a policy of deferring transfusion recipients as blood donors in the UK has minimized this possibility. Other actions to reduce the risk, such as the introduction of filtration devices, are under consideration in the UK, but the identification of vCJD cases with a history of blood donation in France, Ireland, Spain and Saudi Arabia indicates that this issue has an international dimension. One determinant of the risk is the population prevalence of vCJD infection, which is almost certainly highest in the UK, but the identification of secondary transmission of vCJD underlines the importance of international surveillance systems both for human and animal prion diseases.
We thank our colleagues at the NCJDSU for their support for this project. The study would not have been possible without the collaboration of the National Blood Service (NHS Blood and Transplant), Scottish National Blood Transfusion Service, the Welsh Blood Service and the Northern Ireland Blood Transfusion Service. We are grateful to the relatives of CJD cases for their assistance in providing information, and the many clinicians and other staff at UK hospitals and UKBS who have helped to trace records. We thank Dr Nicky Connor and Dr Anna Molesworth of the Health Protection Agency for the data on plasma product batches. The study was funded by the NBS and the Department of Health.