Volume 2: Science
3. The nature and cause of BSE
Dose
How much material would infect?

How much material would infect?

3.126 An important practical consideration when planning measures both to eradicate BSE in cattle and to prevent any possible transmission of BSE in food to other animals or to man was the question of what quantity of infective material would suffice to infect in this way. The quantity of infective material administered to, or eaten by, an animal does not of itself determine the dose of infective agent received by that animal. The effective dose depends on both the quantity of material and the infectivity of that material. ¹ In the case of BSE, the matter with the highest titre of infectivity would have been brain or spinal cord from an affected animal or one nearing the end of the incubation period. A basic question was how much of such material would suffice to transmit the disease orally.

3.127 From the outset there was information available which enabled a view to be formed as to the physical amount of tissue that had high infectivity (ie, brain and spinal cord) which might suffice to transmit the disease to cattle. This was:

1. the extent to which infectious tissue from an animal (be it sheep or cow) infected with a TSE would be diluted by being mixed with other matter in the course of manufacturing MBM;
2. the extent to which the resultant MBM would be further diluted in the course of manufacturing compound feed; and
3. the fact that cattle were succumbing to BSE as a result of eating the material that had been twice-diluted in this way.

3.128 We are not aware of attempts to determine infectivity titres of samples of MBM, or of any attempts to calculate the proportion that infectious material from a sheep with scrapie or a cow with BSE would bear to the overall mass of material processed to produced a batch of MBM. Plainly if tissues from a single infected animal went into the batch it would be likely to be a small proportion.

3.129 Evidence from a representative of the rendering industry indicated that the process of manufacturing MBM resulted in each batch being extremely well mixed so that the outturn was homogeneous and sifted to particles of 3 mm or less in diameter. This evidence was reinforced by written submissions to the Inquiry, which described various production processes in some detail. See, for example, the evidence provided by Mr Stephen Woodgate, who described the breaking, crushing and mixing system used in MBM production. ²

3.130 The proportion of MBM in feed was itself low, due to the low palatability of MBM to cattle. Typically, it made up between 2 and 4 per cent of the compound cattle feed, ³ with a maximum of 5 per cent ⁴ in feed used for weaning calves.

3.131 In certain products, for example, in compounds fed as part of midday rations to high yield cattle, and in concentrates supplied for on-farm mixing with cereals, the level was higher, up to 10 per cent and 26.7 per cent respectively. ⁵ However, mixtures made with these products as just one ingredient probably seldom contained more than 5 per cent of MBM.

3.132 Mr James Reed of the UK Agricultural Supply Trade Association (UKASTA) described how the performance of plants manufacturing compound feed is assessed to ensure the quality of feedstuffs as well as to monitor levels of potential carry-over between successive batches or production runs. He said that assessments are usually made by carrying out mixing trials at intervals of six months to show that the homogeneous distribution of trace ingredients is achieved.

3.133 In the case of MBM, distribution would have been assessed in terms of protein content. ⁶ Thus, both within individual batches of MBM itself, and in compound feedstuffs, the level of infectivity would appear to have been more or less homogeneous. Variation in infectivity between batches of MBM and compound feed could be expected, due to variation in the source of material being rendered.

3.134 This evidence points to the conclusion that cattle infected with BSE through feed are likely to have eaten only a small amount of infective material in that feed. As we understand the position, both Mr Wilesmith and Dr Kimberlin were of the view that it was possible that a small quantity of high titre material would suffice to transmit the disease orally.

The 'Packet' theory

3.135 Another important observation from the epidemiological analysis was the low within-herd incidence of BSE. By the end of 1989, for example, 4,329 out of 6,841 affected herds had only one case, ⁷ with the overall within-herd incidence increasing slowly over time. ⁸

3.136 Although the possibility of variation in uptake of feed per animal, and of differing genetic susceptibility and/or environmental factors (see paragraphs 3.151-3.161) could explain these findings, Mr Wilesmith and Dr Kimberlin favoured the theory of low level exposure to infection. They postulated that the situation was analogous to bioassays of scrapie at limiting doses (ie, at high dilution) when cases occur infrequently and unpredictably.

3.137 Mr Wilesmith and Dr Kimberlin suggested that recycling therefore had the effect of increasing the amount of MBM with a minimum 'threshold' dose of agent to cause disease, with only a modest increase in the average concentration of infectivity in contaminated batches of feed. ⁹ They acknowledged that for this to account for the low within-herd incidence there would also have to be an uneven distribution of 'packets' of highly infective feed. ¹⁰

3.138 The 'packet' theory of MBM distribution was first described in 1991, when Mr Wilesmith published several papers on the epidemiology of BSE. ¹¹ He postulated that the distribution of the most infective tissues within MBM was probably not homogenous, and that some animals might therefore have been exposed to higher levels of infective agent than others.

3.139 We have not found it easy to reconcile the 'packet' theory with the evidence referred to above as to the homogenisation produced by the rendering and feed compounding processes. If the theory is nonetheless valid, it underlines the fact that a very small quantity of high titre tissue must have been sufficient to infect.

BSE in cattle born after the ruminant feed ban (BABs)

3.140 As already discussed, the identification of MBM as the vector of the disease led directly to the introduction of the ruminant feed ban (RFB) in July 1988 and to the SBO ban in 1989. The RFB led to a rapid decline from 1992 in the number of cases of BSE. However, from March 1991 onwards, cases of BSE were confirmed in BABs, which in theory should not have been exposed to MBM. ¹² The occurrence of disease in these animals raised questions as to possible routes of transmission.

3.141 It was first assumed that the BABs were caused by carry-over of old stock of cattle feed containing MBM. Another strong possibility was maternal transmission, but few cases were found to be offspring of affected dams. However, as time went by and the number of BABs increased, these explanations became less tenable. The possibility that ruminant protein was still entering the cattle food chain was therefore considered. Indeed, there was suspicion that ruminant protein destined for pig and poultry feed was accidentally contaminating cattle feed that was manufactured at the same mill. ¹³

3.142 It had been noted that there was an increase in the proportion of cases born in the eastern region of the country after the introduction of the ban. As more pigs and poultry were kept in this area, this supported the possibility that ruminant feed produced after the introduction of the ban may have been inadvertently contaminated at feed mills with ingredients for use in feeds for other species. ¹⁴ Since the preliminary results of the attack rate study (see paragraphs 3.171ff) indicated that very small amounts of infective material were sufficient to cause BSE, the contamination of cattle feed became a major concern.

3.143 When giving oral evidence to the Inquiry, Mr Wilesmith described the different potential sources of cross-contamination - these included old feedstuffs that were still present on farms, accidental or deliberate cross-species feeding, and feed present in the bedding waste from pigs and poultry, which may have been fed to cattle. ¹⁵ However, he supported the view that the most important source of contamination was cross-contamination at feedmills.

3.144 Professor W.D. Hueston of the University of Maryland added that MBM marketed as fishmeal could have entered the cattle feed chain, and that nutritionists formulating feed at compounders could have accidentally incorporated dried MBM back into cattle feed. ¹⁶

3.145 Recent, as yet unpublished, analysis of epidemiological data by Professor Morris and co-workers in New Zealand has shed more light on the pattern of occurrence of BABs, and of the effectiveness of both the ruminant feed ban and the later Specified Bovine Offal (SBO) ban introduced in 1989. Although the ruminant feed ban was extremely effective in reducing the risk of infection, there were regional differences in the extent of this effect.

3.146 By analysing the change in the individual animal risk over time in each region, Professor Morris has illustrated that there was a greater improvement in the South West of England than in other areas. Dr Christl Donnelly proposed that this might be explained by the fact that people were 'more attuned' to the feed ban there, since more cases had been seen in that area, which in turn led to better enforcement. ¹⁷ This 'sociology of compliance' theory was also supported by Professor Hueston, who thought it was a 'very important point'. ¹⁸

3.147 However, another factor which may also explain the greater improvement in the South West is that the area known to have a lower proportion of pig and poultry farming than other areas, and so would suffer less from those potential sources of cross-contamination.

3.148 The effect of the animal SBO ban on the risk of infection is also being looked at by the New Zealand group headed by Professor Morris. Although the analysis is not yet complete, preliminary results indicate that the introduction of the legislation resulted in a further reduction of the number of BSE-affected animals. ¹⁹

3.149 In theory, at the time of the introduction of this ban, no ruminant protein should have been going to cattle, so the ban on SBO in animal feed should have made no difference to infection rates. The fact that it did adds weight to the cross-contamination theory.

Genetic resistance and susceptibility to BSE

3.150 Factors other than the dose of infectious agent might also be relevant to the pattern of disease seen in UK cattle herds. It was supposed, for example, that certain cattle might be more susceptible to disease than others. These factors are discussed in the following paragraphs.

3.151 As discussed in Chapter 2, it had long been established that polymorphisms within the prion protein gene were linked to susceptibility and resistance to TSEs. In sheep the PRNP locus had been termed sip, and in mice sinc. In humans too, certain genetic polymorphisms within PRNP had been associated with susceptibility and resistance to CJD. ²⁰

3.152 For this reason it was important to look at the question of whether such polymorphisms existed in the cattle PrP gene. Although the primary aim was to look at the possibility of eradicating BSE from the UK herd by selective breeding, the investigation of the low incidence within herds was also of interest. The need for the work was agreed early on, and preliminary studies took place at the CVL in 1988. ²¹ Characterisation of the bovine PrP gene was reported at the NPU/CVL research and development (R&D) meeting in November 1988. ²²

3.153 Following on from the initial characterisation of the gene, work to investigate polymorphisms in the bovine PrP gene continued at both the CVL and the NPU, starting in early 1989. ²³ Results were reported in 1991, and showed that two forms of the gene had been found, the more common one having six copies of a particular repeated DNA sequence (octapeptide repeat), the other five copies. ²⁴ However, no significant differences in the frequency of the two genotypes were found between BSE-affected cattle and normal controls, which suggested that this polymorphism was not related to susceptibility.

3.154 Further analysis of the PrP genotypes of healthy and affected cattle showed that approximately 90 per cent of cattle were homozygous for 6 octapeptide repeats (ie, 6:6), 10 per cent heterozygous (6:5) and less than 1 per cent homozygous for five repeats (5:5). The 5:5 genotype was not found in BSE-affected cattle, perhaps suggesting an association with resistance, though this remains to be tested. No apparent breed differences and no indication of difference in the age of onset of BSE were found in animals of different PrP genotypes. ²⁵

3.155 Work was also carried out on the analysis of the PrP gene of BABs. Results were communicated to veterinary officers in 1994, and again these provided no clear evidence for polymorphism, with no significant differences seen between animals born after the ruminant feed ban, and those born earlier in the epidemic. Nor did the study find differences between confirmed and unconfirmed cases. ²⁶

3.156 As well as polymorphisms in the PRNP gene itself, it was thought possible that susceptibility/resistance genes for BSE could exist in the bovine genome outside the PRNP gene. ²⁷ This eventuality was suggested for scrapie in 1993. ²⁸ Work to investigate the concept was undertaken by Dr John Williams at the Roslin Institute in Scotland, starting in November 1991. Results were published in collaboration with a group from the Texas University Center for Animal Genetics in 1994, and reported further evidence of a genotype that may be associated with susceptibility to BSE. ²⁹ However, this result awaits confirmation.

3.157 Overall, therefore, little evidence for a resistant genotype was found. This was significant in that it ruled out the possibility of selective breeding as a policy action. Although by 1992 the ruminant feed ban had already led to a reduction of BSE infection in younger animals, the appearance of BABs meant that, had a resistant genotype been identified, some kind of selective breeding strategy could have been significant in halting the epidemic.

3.158 The lack of an obvious genetic susceptibility factor also left unanswered the question why only certain individuals within a herd contracted BSE. Although in oral evidence to the Inquiry a member of the Spongiform Encephalopathy Advisory Committee (SEAC) confirmed that the packet theory is still thought to be the most likely model, ³⁰ recent modelling work and differences in disease incidence patterns among other animals suggest that other factors may be involved. TME, for example, normally affects 100 per cent of the mink on any affected farm, whereas feline spongiform encephalopathy (in cats) has only been seen in very low numbers. The possibility that TME may be spread via wounds inflicted by littermates during feeding time is one possible explanation for such differences, and others may exist. Individual non-genetic differences could also be important.

Environmental and other contributory factors

3.159 The potential for the involvement of other individual factors has been raised most recently with respect to vCJD victims. In a paper published in The Lancet in 1999, ³¹ Professor John Collinge wrote that the pattern of vCJD cases seen so far was suggestive either of one or more environmental risk factors or of unusually high innate sensitivity in certain individuals. He postulated that factors such as mouth lesions, and infection of the tonsils or gastrointestinal tract for example, could facilitate infection by the oral route, and hence lead to differences in individual susceptibility. In a 1995 paper, a similar factor, coexistent gut infection with nematodes, was suggested as predisposing sheep to scrapie. ³² Two experiments in 1982 and 1996 demonstrating the effects of gum and skin scarification on efficiency of infection lent weight to the argument that such conditions could be important. ³³ Another study published in 1998 described the involvement of the inflammatory response in skin cells in PrP expression, and again suggested that certain skin disorders may be relevant in disease acquisition. ³⁴ A more recent study has provided evidence suggesting that cells lining the gastrointestinal tract may represent a possible target for prion entry and replication, and in particular that PrPc is upregulated in the mucosa of patients with Heliobacter pylori gastritis. ³⁵ Transmission of scrapie to hamsters by the dental route has also been shown recently, raising questions about the possible risk of transmitting TSEs through dental procedures. ³⁶ As early as 1974, a possible protective factor, immunosuppression, had also been proposed in relation to scrapie. ³⁷ This list illustrates just how many different conditions could potentially influence overall disease susceptibility.

3.160 The use of organophosphate pesticides, proposed by Mr Purdey as a causative factor for BSE (see paragraphs 3.77-3.85), could also be relevant in this respect. Although these chemicals have largely been dismissed as the sole cause of BSE, it is possible that their presence in the body could be just one of many other contributory factors raising susceptibility to disease. Most of the results in favour of this theory have been obtained post-1996. In 1998 it was shown that phosmet applied to cultures of neuroblastoma cells caused a tenfold increase in the level of normal prion protein. ³⁸ More recently, in vitro experiments have shown that cellular PrP, which has a structure rich in -helices, can be converted to a structure rich in -sheets by reduction of the single disulphide bond in cellular PrP at a pH of 4.0. ³⁹ In addition, the studies which have shown that when manganese is bound to PrP the latter becomes resistant to protease digestion, indicate that chemical modification of PrP might render the individual more susceptible to infection (see paragraph 3.85). Although the full significance of these results in relation to BSE is not clear, such effects do indicate that OPs, and other chemicals, could potentially be involved.

3.161 Although these hypotheses are difficult to test, it is clear that a number of factors could influence whether or not an individual contracts a TSE. Future studies of risk factors for vCJD, and further investigation of TSE transmission in general, are likely to be important in furthering our understanding of what determined whether or not an individual animal contracted BSE following exposure to the agent.