15.6E: Variant Creutzfeldt-Jakob Disease - Biology

15.6E: Variant Creutzfeldt-Jakob Disease - Biology

We are searching data for your request:

Forums and discussions:
Manuals and reference books:
Data from registers:
Wait the end of the search in all databases.
Upon completion, a link will appear to access the found materials.

Variant Creutzfeldt–Jakob Disease (vCJD) is a fatal neurological disorder which is caused by prions.

Learning Objectives

  • Generalize the role of prions in Creutsfeldt-Jakob disease

Key Points

  • Bovine spongiform encephalopathy (BSE) is believed to be the cause of variant Creutzfeldt–Jakob (vCJD); BSE is a prion disease that affects cattle. In both humans and cattle the disease causes large holes in the brain.
  • The prion the misfoled protein that causes vCJD has two conformations: one is the native form and is water soluble; the other is the disease form, which is water insoluble.
  • The misfolded prion proteins can cause other normally folded pre-prion proteins to become prions, which disrupts the native proteins disrupting function leading to cell death.
  • There is no known treatment for vCJD, except avoiding BSE contaminated meat.

Key Terms

  • Creutzfeldt–Jakob disease: a rare, progressive, currently fatal disease of the nervous system, characterized by dementia and loss of muscle control; a prion disease, apparently transmissible from animals to humans by eating infected tissue, as well as from tissue interchanges among humans
  • prion: A self-propagating misfolded conformer of a protein that is responsible for a number of diseases that affect the brain and other neural tissue.
  • transmembrane: traversing a cellular membrane

Creutzfeldt–Jakob disease, or CJD, is a degenerative neurological disorder (brain disease) that is incurable and invariably fatal. CJD is occasionally called a human form of mad cow disease (bovine spongiform encephalopathy or BSE), even though classic CJD is not related to BSE. However, given that BSE is believed to be the cause of variant Creutzfeldt–Jakob disease (vCJD) in humans, the two are often confused. In CJD, the brain tissue develops holes and takes on a sponge-like texture. This is due to a type of infectious protein called a prion. Prions are misfolded proteins which replicate by converting their properly folded counterparts.

Transmissible spongiform encephalopathy diseases are caused by prions. Thus, the diseases are sometimes called prion diseases. Other prion diseases include Gerstmann–Sträussler–Scheinker syndrome (GSS), fatal familial insomnia (FFI) and Kuru in humans; as well as bovine spongiform encephalopathy (BSE, commonly known as mad cow disease) in cattle, chronic wasting disease (CWD) in elk and deer, and Scrapie in sheep. Alpers’ syndrome in infants is also thought to be a transmissible spongiform encephalopathy caused by a prion.

The prion that is believed to cause Creutzfeldt–Jakob exhibits at least two stable conformations. One, the native state, is water-soluble and present in healthy cells. As of 2007, its biological function is presumably in transmembrane transport or signaling. The other conformational state is relatively water-insoluble and readily forms protein aggregates. People can also acquire CJD genetically through a mutation of the gene that codes for the prion protein (PRNP). This occurs in only 5–10% of all CJD cases.

The CJD prion is dangerous because it promotes refolding of native proteins into the diseased state. The number of misfolded protein molecules will increase exponentially and the process leads to a large quantity of insoluble protein in affected cells. This mass of misfolded proteins disrupts cell function and causes cell death. Mutations in the gene for the prion protein can cause a misfolding of the dominantly alpha helical regions into beta pleated sheets. This change in conformation disables the ability of the protein to undergo digestion. Once the prion is transmitted, the defective proteins invade the brain and are produced in a self-sustaining feedback loop.

Variant Creutzfeldt-Jakob disease

Variant Creutzfeldt-Jakob disease (vCJD) is a type of Creutzfeldt-Jakob disease (CJD) characterized by early psychiatric symptoms and cognitive decline. All forms of CJD belong to a rare family of progressive neurodegenerative disorders that affect both humans and animals, called prion diseases. The term "prion" refers to abnormal proteins within the brain, called prion proteins. vCJD, described primarily in the United Kingdom and France, accounts for less than 1% of cases of CJD, and tends to affect younger people. It can result when someone is exposed to contaminated products. The infection responsible for the disease in cows (bovine spongiform encephalitis) is believed to be the same one responsible for vCJD in humans. [1] [2] [3]

Another variant, called the panencephalopathic form, occurs primarily in Japan and has a relatively long course, with symptoms often progressing for several years. Scientists are trying to gain a better understanding about what causes these variations in the symptoms and course of the disease. [1] There is no specific treatment for CJD or vCJD, so the goal is to make a person as comfortable as possible. [1] [2] [3]

Cerebrospinal Fluid in Neurologic Disorders

Tainá M. Marques , . Marcel M. Verbeek , in Handbook of Clinical Neurology , 2018

Creutzfeldt–Jakob disease

CJD is a rapidly progressive neurodegenerative disorder belonging to the human prion diseases, also known as transmissible spongiform encephalopathies. Sporadic CJD is the most common form and accounts for 85% of all cases with disease onset, in general, at 70 years of age. CJD is caused by misfolding of the prion protein (PrP) into the disease form (PrP Sc , leading to prion plaques and spongiform changes in the brain (reviewed in Zanusso et al., 2016 , and Chapter 8 in this book).

CJD is characterized by rapidly progressive dementia, ataxia, and myoclonus. A final diagnosis is only possible with brain biopsy, but a probable diagnosis is reached by evaluating MRI, electroencephalogram, clinical symptoms, and CSF analysis of 14-3-3 protein for exclusion of other progressive neurodegenerative disorders ( Collins et al., 2006 Zerr et al., 2009 ).

The 14-3-3 protein is already being used as a biomarker in CJD diagnosis. Increased 14-3-3 protein levels in CSF discriminated CJD from controls and other neurodegenerative disorders with high accuracy of up to 97%, as shown by several groups ( Zerr et al., 2000 Wang et al., 2010 Zanusso et al., 2011 Stoeck et al., 2012 Tagliapietra et al., 2013 Dulamea and Solomon, 2016 Koscova et al., 2016 Leitao et al., 2016a, b ). However, in a large multicenter study, one group observed that the specificity of increased 14-3-3 levels for CJD decreases when comparing to acute neurologic disorders (accuracy only 87%) ( Stoeck et al., 2012 ).

The tau protein was also tested as a potential biomarker for CJD. Total tau levels were very strongly increased in CJD compared to controls or AD patients, and a significant increase was observed in patients with both the sporadic and inherited forms of CJD, with an accuracy up to 99% ( Wang et al., 2010 Hamlin et al., 2012 Stoeck et al., 2012 Tagliapietra et al., 2013 Skillback et al., 2014 Llorens et al., 2015, 2016b Dulamea and Solomon, 2016 Koscova et al., 2016 Leitao et al., 2016a, b ). Another study reported a positive correlation of total tau CSF levels with cognitive decline as well as disease severity ( Cohen et al., 2015 ). Moreover, total tau was indicated as a biomarker: (1) for discrimination of CJD from controls when the results of 14-3-3 are inconclusive ( Leitao et al., 2016a, b ) (2) for discrimination from AD ( Forner et al., 2015 Grangeon et al., 2016 ) (3) when the ratio of phosphorylated tau/total tau discriminates CJD from AD ( Zanusso et al., 2011 ) and (4) when combination of very high total tau levels and normal-to-mildly elevated phosphorylated tau discriminates CJD from controls ( Skillback et al., 2014 ). Finally, a combination of 14-3-3 and tau could further improve the diagnostic accuracy for CJD ( Zanusso et al., 2011 ).

Several attempts have been made to quantify the PrP Sc protein in CSF as a diagnostic measure for CJD, but the results were inconclusive, as equal or lower levels of this protein in CJD CSF compared to AD or controls were found ( Meyne et al., 2009 Torres et al., 2012 Llorens et al., 2013 Dorey et al., 2015 ). Of more recent date is the development of the RT-QuIC assay to quantify levels of PrP Sc ( Atarashi et al., 2011 ). Many groups have been testing PrP Sc using this assay in CJD all presented good results in comparison to controls, with high sensitivity up to 97% and specificity up to 100% (reviewed in Zanusso et al., 2016 ). Although not yet implemented in clinical practice, several centers are now incorporating this technique for PrP Sc quantification as this assay offers a good possibility for early diagnosis.

In summary, 14-3-3 protein is well established as a biomarker for CJD, and is currently used for diagnostic discrimination from controls and other neurodegenerative disorders. Total tau and phosphorylated tau/total tau ratio are also very useful biomarkers for CJD, particularly when 14-3-3 results remain inconclusive. In addition, the RT-QuIC assay has made it possible to quantify misfolded PrP, and has been shown to have extremely high accuracy levels for discrimination of CJD from controls.

Cause Cause

Some researchers believe an unusual 'slow virus ' or another organism causes Creutzfeldt-Jakob disease (CJD). However, they have never been able to isolate a virus or other organism in people with the disease. Furthermore, the agent that causes CJD has several characteristics that are unusual for known organisms such as viruses and bacteria . It is difficult to kill, it does not appear to contain any genetic information in the form of nucleic acids ( DNA or RNA ), and it usually has a long incubation period before symptoms appear. In some cases, the incubation period may be as long as 40 years. The leading scientific theory at this time maintains that CJD and the other TSEs are caused by a type of protein called a prion. [1]

Prion proteins occur in both a normal form, which is a harmless protein found in the body’s cells , and in an infectious form, which causes disease. The harmless and infectious forms of the prion protein have the same sequence of amino acids (the 'building blocks' of proteins) but the infectious form of the protein takes a different folded shape than the normal protein. Sporadic CJD may develop because some of a person’s normal prions spontaneously change into the infectious form of the protein and then alter the prions in other cells in a chain reaction. [1]

Once they appear, abnormal prion proteins aggregate, or clump together. Investigators think these protein aggregates may lead to the neuron loss and other brain damage seen in CJD. However, they do not know exactly how this damage occurs. [1]

About 5 to 10 percent of all CJD cases are inherited . These cases arise from a mutation , or change, in the gene that controls formation of the normal prion protein. While prions themselves do not contain genetic information and do not require genes to reproduce themselves, infectious prions can arise if a mutation occurs in the gene for the body’s normal prion protein. If the prion protein gene is altered in a person’s sperm or egg cells, the mutation can be transmitted to the person’s offspring. Several different mutations in the prion gene have been identified. The particular mutation found in each family affects how frequently the disease appears and what symptoms are most noticeable. However, not all people with mutations in the prion protein gene develop CJD. [1]


The pathological diagnostic features of vCJD are summarised in Table 5.

1. Multiple florid plaques in H&E sections numerous small cluster plaques in PrP-stained sections amorphous pericellular and perivascular PrP accumulation in the cerebral and cerebellar cortex. 2. Severe spongiform change perineuronal and axonal PrP accumulation in the caudate nucleus and putamen. 3. Marked astrocytosis and neuronal loss in the posterior thalamic nuclei and midbrain. 4. Reticular and perineuronal PrP accumulation in the grey matter of the brainstem and spinal cord. 5. PrP RES accumulation in lymphoid tissues throughout the body. 6. Predominance of diglycosylated PrP RES in central nervous system and lymphoid tissues.

Macroscopic features

By the end of October 2001, 89 cases of vCJD had been diagnosed on the basis of neuropathology. Of these, 3 were diagnosed on brain biopsy alone, 7 had material available from both brain biopsy and autopsy, and 79 cases were diagnosed on autopsy findings only. The brain weight after fixation ranged from 989–1530 g. Cerebral cortical atrophy and cerebellar atrophy (particularly involving the vermis) were identified in cases with a lengthy clinical history (>19 months). In these cases there was evidence of ventricular dilatation, with a corresponding reduction in myelinated axons in the white matter. One of the cases had a small focus of chronic inflammation within the pons in the absence of a widespread encephalomyelitis, and another had a small cavernous angioma in the right temporal pole. No other significant macroscopic abnormalities were detected in the CNS.

Microscopic features

Spongiform change (in the absence of plaques) was widespread in a patchy distribution within the cerebral cortex and involved all cortical layers (Fig. 1a). Confluent spongiform change was rare and most evident in the occipital and inferior frontal cortex, usually at the depths of the gyri. No spongiform change was detected in the hippocampus, but the entorhinal cortex showed patchy microvacuolar spongiform change.

a. Multiple florid plaques of varying size are present in the frontal cortex in a patient with a 27-month clinical history of variant CJD. The larger plaques (centre) have a dense eosinophilic core with a pale fibrillary periphery and are surrounded by spongiform change. b. The core of the florid plaques is positive on a periodic acid/Schiff stain (without diastase), but the fibrillary periphery is unstained. Although one of the plaques is located adjacent to a blood vessel, there is no evidence of an amyloid angiopathy. c. Immunocytochemistry for PrP in the cerebellum shows intense labelling of the larger plaques, but also demonstrates multiple smaller plaques and amorphous accumulations of PrP which are not visible on routine staining (KG9 antibody). d. Widespread astrocytosis is present in the pulvinar in variant CJD, with relatively little spongiform change and no plaque formation (immunocytochemistry for glial fibrillary acidic protein). e. Immunocytochemistry for PrP in the tonsil shows intense labelling of the germinal centres, with positivity appearing in follicular dendritic cells and tingible body macrophages (KG9 antibody). f. Immunocytochemistry for PrP shows intense labelling of the ganglion cells within the dorsal root ganglia adjacent to the spinal cord. Some of the satellite cells also appear positively labelled and there is evidence of neuronal degeneration (6H4 antibody).

Confluent spongiform change was always present in the basal ganglia (caudate nucleus and putamen) and was disproportionately severe in relation to the scanty numbers of amyloid plaques. Focal spongiform change involved many of the nuclei in the anterior and medial thalamus, often in the absence of plaques, but the posterior thalamic nuclei (including the pulvinar) were relatively spared. In the hypothalamus, spongiform change was most evident in the paraventricular and supraoptic nuclei, which also contained occasional amyloid plaques. Mild spongiform change was furthermore detected in the periaqueductal grey matter in the midbrain and in the grey matter of the pons and medulla, but was not present in spinal cord.

In the cerebellar cortex, spongiform change was a prominent feature throughout the hemispheres and vermis. Confluent spongiform change was observed in the molecular layer of the cerebellum in a patchy distribution, often associated with florid plaques.

Cerebral cortical neuronal loss was most marked in the occipital lobe (particularly in the primary visual cortex), but in cases with a lengthy clinical history there was widespread and severe loss of neurones, with accompanying astrocytosis. Even in these cases, the neuronal populations in the hippocampus were well preserved. The basal ganglia showed variable neuronal loss which was most evident in cases with severe and confluent spongiform change. In the medial and posterior regions of the thalamus there was extensive neuronal loss with marked astrocytosis, which was most severe (amounting to almost total neuronal loss) in the pulvinar. In the midbrain, neuronal loss and astrocytosis occurred in the periaqueductal grey matter superior and to a lesser extent in the inferior colliculi. Neuronal loss and astrocytosis were not prominent in the pons, medulla and spinal cord. The cases with a lengthy clinical course exhibited severe astrocytosis and neuronal loss involving all layers of the cerebellar cortex, particularly the granular cell layer.

Florid plaques were easily identified on haematoxylin and eosin stains (Fig. 1a) as a fibrillary structure with a dense eosinophilic core surrounded by a pale region of radiating fibrils, which in turn was encircled by a rim of microvacuolar spongiform change. The florid plaques measured up to 150 μm, and were particularly well visualised using the Congo red, periodic acid/Schiff (Fig. 1b) and Alcian blue stains, and the Gallyas silver impregnation technique. These amyloid plaques were often present close to blood vessels (Fig. 1b), but a true amyloid angiopathy was not identified. Florid plaques were not present in all cases which had undergone brain biopsy, but were present in all autopsy cases, being most numerous in the occipital and cerebellar cortex. They occurred in all layers of the cerebral cortex, often in a random focal distribution, but were most conspicuous at the bases of the gyri. Florid plaques were most easily identified in the molecular layer of the cerebellum, occasionally projecting into the subpial space, but kuru-type plaques without surrounding spongiform change were also present in aggregates in the granular layer.


CNS tissues. The florid plaques in the cerebral and cerebellar cortex showed strong staining on immunocytochemistry for PrP, which also revealed numerous smaller plaques, often arranged in irregular clusters within the neuropil (Fig. 1c). These “cluster plaques” were present in all cases, including those which had undergone brain biopsy. In addition, there was a widespread amorphous pericellular deposition of PrP around small neurones in the cerebral and cerebellar cortex (Fig. 1c). These amorphous deposits could also be identified on Gallyas silver impregnation. Occasional PrP deposits were identified around capillaries in the cerebral and cerebellar cortex, not as amyloid deposits in the blood vessel wall, but as loose amorphous deposits outside the basement membrane. In the hippocampus, the cornu ammonis showed little PrP accumulation, but there was a dense synaptic accumulation in the dentate fascia, subiculum and entorhinal cortex (which also contained small cluster plaques).

In the basal ganglia there was a noticeable perineuronal pattern of PrP accumulation, often with linear decoration of dendrites and axons, and accompanied by multiple small plaques. A synaptic pattern of immunoreactivity was detected in the thalamus and hypothalamus with occasional plaques, but no linear patterns of accumulation. Synaptic and perineuronal neuronal positivity was also present in the midbrain, pons and medulla, particularly in the pontine nuclei. In the spinal cord, PrP positivity was present in the grey matter regions at all levels, particularly in the substantia gelatinosa. No PrP accumulation was detected in either the dura mater or arachnoid granulations.

The severe astrocytosis in the posterior thalamus was best visualised on immunocytochemistry for glial fibrillary acidic protein (Fig. 1d). This technique also demonstrated astrocytosis in the midbrain. In the cerebral and cerebellar cortex, astrocytic proliferation was demonstrated in relation to areas of severe neuronal loss and occasionally around the margins of florid plaques.

Non-CNS tissues. Very few significant pathological features were detected in histology of non-CNS tissues most patients died from bronchopneumonia. Positive staining for PrP was identified in follicular dendritic cells and macrophages within many germinal centres in the pharyngeal, lingual and palatine tonsil (Fig. 1e), with a more restricted pattern of positivity within germinal centres in the appendix, Peyer's patches in the ileum, spleen and lymph nodes from the cervical, mediastinal, para-aortic and mesenteric regions. PrP positivity was also identified in dorsal root ganglia (Fig. 1f) and in the trigeminal ganglia. Peripheral nerves contained no detectable PrP on immunocytochemistry.

Synaptic positivity for PrP and a few small plaques were identified in the posterior pituitary gland the anterior pituitary gland showed no PrP accumulation. PrP immunocytochemistry in the other organs (heart, lung, liver, gall bladder, salivary gland, oesophagus, stomach, pancreas, skeletal muscle, kidney, adrenal gland, thyroid gland, parathyroid gland, bladder, testes, pelvic organs (vagina, cervix, uterus, Fallopian tubes and ovaries) and skin was negative.


SDS-PAGE Western blot profiles of PrP RES in the brain in CJD have revealed at least two PrP RES mobility variants, which correspond to degree of protease-mediated N-terminal truncation. PrP RES can also be subclassified according to the relative abundance of the nonglycosylated, monoglycosylated and diglycosylated glycoforms in any individual sample. We have found two different mobility PrP RES subtypes in the frontal cortex in sporadic CJD, one with a nonglycosylated PrP RES of ∼21 kDa (termed type 1) and another of ∼19 kDa (termed type 2). The PrP RES isoform pattern from patients with vCJD is indistinguishable in terms of mobility from that of the type 2 PrP RES isoform seen in sporadic CJD. However, the glycoform ratio in vCJD is different from that found in sporadic CJD, with a predominance of diglycosylated PrP RES (termed type 2B) (Fig. 2). We have found this characteristic hyperglycosylation in every case of vCJD (where frozen tissue was available, N=51). Only cases of vCJD have PrP RES detectable outside the central nervous system in lymphoid tissues. It too is characterised by a predominance of the diglycosylated form of PrP RES in Western blots after protease K digestion, although typically the glycoform ratio in these tissues is even higher than in the corresponding brain samples ( 13 ).

Western blot PrP RES isoform analysis of frontal cortex samples from cases of sporadic (S) and variant (V) Creutzfeldt-Jakob disease. Three glycoforms are seen in each sample corresponding to the di-, mono-, and nonglycosylated PrP RES . The samples are classified as type 1 with a nonglycosylated PrP RES of ∼21 kDa or type 2 with a nonglycosylated PrP RES of ∼19 kDa. The variant Creutzfeldt-Jakob disease sample is predominantly composed of the di-glycosylated form and is designated type 2B to distinguish it from type 2 typical of sporadic Creutzfeldt-Jakob disease samples in which the monoglycosylated form predominates (type 2A).

Predictors of survival in sporadic Creutzfeldt-Jakob disease and other human transmissible spongiform encephalopathies

A collaborative study of human transmissible spongiform encephalopathies has been carried out from 1993 to 2000 and includes data from 10 national registries, the majority in Western Europe. In this study, we present analyses of predictors of survival in sporadic (n = 2304), iatrogenic (n = 106) and variant Creutzfeldt-Jakob disease (n = 86) and in cases associated with mutations of the prion protein gene (n = 278), including Gerstmann-Sträussler-Scheinker syndrome (n = 24) and fatal familial insomnia (n = 41). Overall survival for each disease type was assessed by the Kaplan-Meier method and the multivariate analyses by the Cox proportional hazards model. In sporadic disease, longer survival was correlated with younger age at onset of illness, female gender, codon 129 heterozygosity, presence of CSF 14-3-3 protein and type 2a prion protein type. The ability to predict survival based on patient covariates is important for diagnosis and counselling, and the characterization of the survival distributions, in the absence of therapy, will be an important starting point for the assessment of potential therapeutic agents in the future.

Risk for Travelers

The current risk of acquiring vCJD from eating beef (muscle meat) and beef products produced from cattle in countries with at least a possibly increased risk of BSE cannot be determined precisely. If public health measures are being well implemented the current risk of acquiring vCJD from eating beef and beef products from these countries appears to be extremely small, although probably not zero. A rough estimate of this risk for the UK in the recent past, for example, was about 1 case per 10 billion servings. Among many uncertainties affecting such risk determinations are 1) the incubation period between exposure to the infective agent and onset of illness, 2) the appropriate interpretation and public health significance of the prevalence estimates of asymptomatic human vCJD infections, 3) the sensitivities of each country&rsquos surveillance for BSE and vCJD, 4) the compliance with and effectiveness of public health measures instituted in each country to prevent BSE contamination of human food, and 5) details about cattle products from one country distributed and consumed elsewhere. As of August 2006, despite the apparent exceedingly low risk of contracting vCJD through consumption of food in Europe, the US blood donor deferral criteria focuses on the time (cumulatively 5 years or more) that a person lived in continental Europe from 1980 through the present. In addition, these deferral criteria apply to persons who lived in the United Kingdom from 1980 through 1996.

Prion Diseases

II.A Neuroinvasion

Acquired prion infection such as scrapie is an oral infection iatrogenic CJD may be peripherally acquired. For invasion of the central nervous system, two possible routes are discussed: (1) transport in blood cells, possibly after amplification of prions in the lymphoreticular system (LRS), and (2) transport in peripheral nerves. Possibly a combination of amplification in the LRS and transport by peripheral nerves is what actually happens in acquired prion disease. The importance of the LRS has been known for a long time.

In experimentally infected mice, infectivity has been shown in the spleen 4 days after intraperitoneal and even after intracerebral inoculation. In these cases, prion replication in the spleen precedes intracerebral replication even after intracerebral inoculation. In nvCJD, PrP Sc accumulates in lymphoid tissues of the tonsils and in the appendix, in one case 8 months before the outbreak of clinical disease. The nature of the cells supporting prion replication in the LRS has not been established beyond doubt. Follicular dendritic cells (FDCs) would be prime candidates. Indeed, PrP Sc accumulation has been observed in these cells. In mouse scrapie, functional B lymphocytes are necessary for neuroinvasion, but PrP C expression in these cells is not required. B lymphocytes may indirectly influence neuroinvasion by allowing the development of mature spleen FDCs as sites of agent replication. Because lymphocytes do not normally cross the blood–brain barrier, it seems questionable, however, whether immune cells are sufficient to transport the agent from the periphery to the CNS.

Prion replication in the CNS first takes place in areas that relate to the site of peripheral inoculation or oral uptake. This implies that the agent spreads along the peripheral nervous system. The importance for neuroinvasion of PrP C positivity of peripheral nerves has been demonstrated in experiments where transgenic mice expressing PrP C only in neurons developed scrapie after oral or intraperitoneal infection with high doses of the infectious agent. A scenario in which the agent is first transported to FDCs by mobile immune cells, where it amplifies and spreads to the peripheral nervous system, also seems possible and may be of particular importance in cases of low-dose infection.


When the first vCJD cases were reported in the late 1990s, the small numbers combined with lack of knowledge of both the potential transmission routes and key epidemiological parameters meant that projections of the future epidemic were highly uncertain [24], [32]. Following the peak in cases in 2000, and their subsequent decline to low numbers, it has for several years been possible to characterise the oral transmission route in the MM genotype and estimate associated epidemiological parameters with a reasonable degree of precision [28], [31]. However, with small numbers of cases now arising in different genotypes and via other transmission routes (3 cases of MM genotype attributed to blood transfusions since 2003 and reports of a possible vCJD case in a person of MV genotype), there remain concerns about the potential for a second epidemic wave.

Building on previous work [10], [17], we have used a stochastic model in a Bayesian framework, combining the transmission via food-borne and red cell transfusion associated transmission with a differentiation with respect to genotype. Our results indicate that we can expect only a small number of future cases to arise in both the MV and VV genotypes through primary transmission. This is because the infection risk is assumed to have been very low indeed for several years and so efficient primary transmission to these genotypes is now only possible in combination with rather long incubation periods, such that a substantial proportion of those infected would reach the end of their natural life span before succumbing to clinical disease. Larger numbers of future cases are possible in all genotypes if efficient transmission occurs though red-cell transfusion. However, even these numbers are limited by the numbers of individuals and the age-profile of those that receive transfusions [10], [33]. Our results suggest that if a second epidemic does arise, this is likely to evolve over a number of decades. Our best estimate for the annual incidence is low with up to 10 cases occurring annually, although the credibility intervals are wide due to large uncertainties in many of the key parameters governing transmission.

Despite the rather long time-scale of this potential second wave, we did not find any scenarios which led to a self-sustaining epidemic as classified by the basic reproduction number . In fact, the values are so low that even if both leuko-depletion and the donor ban were totally ineffective they would reach values of less than 0.5. This is in contrast with previous work [10], which found the potential for a self-sustaining epidemic for some combination of parameter values in the absence of any control measures. These results were based on fitting to 2 transfusion associated cases up to 2006, whereas here we are fitting to 3 cases up to 2009, taking into account a number of years during which no transfusion associated cases have been observed. Furthermore, some of the assumptions underlying the earlier work were more pessimistic, whereas here we have refined the model to be more realistic, such as allowing for a delay between infection and the onset of infectivity, a lower susceptibility in non-MM genotypes and the use of several red cell units in a single transfusion, reducing the values of the basic reproduction number further.

One assumption implicit in our model simulations concerns the age dependence of susceptibility/exposure to infection. To fit the age distribution of the primary epidemic, as in previous work [27], [28], we fit a strong age dependence in susceptibility/exposure. To date there have been too few secondary cases to fit a distribution to this age profile and we have therefore assumed that for blood-borne transmission susceptibility is independent of age. This might be the case if the age-distribution of cases via primary transmission occurred due to differences in exposure rather than biological susceptibility per se, although evidence for this is limited [34]. Furthermore, animal studies have suggested that one mechanism for biological susceptibility may be age-related changes in the gut [35] and thus it is possible that all ages would be equally susceptible to transmission via blood transfusion. Regardless, even if biological susceptibility did occur via all transmission routes, this remains a reasonable assumption if the infectivity in a single red cell unit is very high. However, for lower transmissibility, including age-dependent susceptibility would reduce the secondary peak considerably given the lack of overlap between those that appear most susceptible to date (teenagers and young adults) and the age distribution of transfusion recipients [25].

We also investigated sensitivity to our assumptions regarding the effectiveness of the control policies in place (see Supplementary Material S1). None of the alternative assumptions investigated changed the overall dynamics significantly, however, the upper limit of the credibility intervals of the projected future epidemic size varied for different scenarios. If leuko-depletion is ineffective in preventing transmissions via red cell transfusions we would expect the secondary outbreak to be up to twice as large, whereas a less effective donor ban had very little effect on the outbreak size. This is because the majority of red cell transfusion cases are caused by people who were themselves infected via the oral transmission route and are therefore not subject to the donor ban. If the test sensitivity of the prevalence test is lower, the true population prevalence is higher than measured in the appendix study, leading to more secondary transmission and therefore potential for a larger secondary outbreak.

In summary, given that there are no further known transmission routes efficient enough that they could lead to a self-sustaining epidemic, the variant CJD epidemic in the UK is likely to continue with a low level annual incidence for a lengthy period of years to decades. Whilst any projections of future case numbers are highly uncertain, reflecting the current uncertainties in key transmission parameters for the genotypes and transmission routes in which we have not seen many cases as yet, the timescales involved are fairly insensitive to these highly uncertain parameters. Despite the inherent large uncertainty our results are important for public health planning: The current low level of annual incidence appears to suggest that the epidemic is nearly over. However, while this might be the case, a secondary peak remains a possibility, and this has to be taken into account when decisions are made about the introduction or withdrawal of control measures.