We report the occurrence of a novel immune-mediated polyradiculoneuropathy in swine abattoir workers. We identified 15 cases in Minnesota, seven in Indiana, and one in Nebraska (). These cases were defined primarily by the prominence of early sensory and motor neurologic symptoms, easily identified motor and reflex examination deficits on neurological examination, and specific electrophysiological findings. Although additional workers presented with neurologic symptoms, the observed symptoms were purely sensory and did not meet the epidemiologic case definition.
The case definition was developed at the onset of the investigation for the purpose of identifying risk factors associated with the illness. For this purpose, in the absence of a clear biologic marker, inclusion of only those individuals with the most reproducible signs and readily available diagnostics was important. All cases had some degree of sensory symptoms
, however if those with only sensory symptoms were included in the case definition, this would have increased the likelihood of misclassification bias where persons with sensory symptoms due to another cause would be included, given the high frequency of factors such as repetitive motion injury, prolonged standing, and persistently cold conditions in this occupational setting
. Similarly, excluding persons with mild neurologic symptoms that did not meet the case definition decreased the chance of misclassification bias within the controls. Excluding the mild or asymptomatic cases of disease is a common practice in epidemiologic studies and outbreak investigations.
The epidemiologic features of the cases are distinguished by association with exposure to porcine brain harvested by compressed air. Although onsite investigations were not conducted at abattoirs that removed brains whole or did not remove brains, none of those abattoirs reported knowledge of unusual neurologic illness among their employees. All cases in the implicated abattoirs reported porcine brain exposure. In the Minnesota abattoir, working at the brain-removal job or the job closest to it (backing heads), or working at a distance within 3.1 meters from the brain-removal operation, was associated with illness. Jobs of backing heads and removing brain at the head-table are considered preferred jobs. Workers with more seniority successfully bid for these jobs, possibly explaining why age was significant when analyzing by head-table controls. No additional cases have been identified with onset dates after the process ceased in any of the implicated abattoirs.
Autoimmunity appears to be the likely pathogenic mechanism of IP induced by exposure to aerosolized porcine brain matter. Mucous membrane contact, inhalation, or possibly through contact with broken skin are the most likely routes of entry based on the association of cases with the brain-removal operation, our observations of the mist of brain created by the removal process, and the presence of residual brain material on nearby workers. Exposure to neural tissue including sheep brain
, peripheral nerve myelin
and bovine gangliosides
has been previously epidemiologically linked with development of autoimmune neuropathy. While similar illnesses have been described following injection of neural proteins, to the authors' knowledge this is the first time that a respiratory, mucosal, or through broken skin exposure is the likely route of entry as we have implicated. We cannot exclude the possibility that the pathogenesis of IP involves either an infection or direct toxic effect related to a component of porcine brain. However, no toxic or infectious cause was identified despite a comprehensive exposure history interview, review of chemicals used in the abattoirs, and extensive laboratory testing for infectious agents.
An autoimmune mechanism in IP is supported by higher levels of IFNγ in cases than in controls, as elevated IFNγ has been observed among persons experiencing acute or chronic inflammatory demyelinating polyradiculoneuropathies (AIDP or CIDP)
. Further support for the importance of IFNγ-triggered cytokine cascades in IP derives from the parallel elevations of IFNγ and IFNγ-inducible protein of 10 kDa (IP-10) along with the Th2 cytokine, IL-4, and the chemokine, IL-8 (). IP-10 is known to be elevated in the CSF of patients with inflammatory neuropathies and in inflamed peripheral nervous system the distribution of IP-10 mirrors that of the chemokine receptor CXCR3, its cognate receptor
. Intriguingly, control subjects who were exposed to brain material but did not manifest illness were revealed by factor analysis to deviate the most from the exposed ill group in their cytokine and chemokine profiles relative to non-exposed, non-ill controls. In comparison with ill exposed subjects, the factor 2 and 3 scores of exposed non-ill individuals suggested greater proinflammatory drive (increased TNFα, IL1β); less skew toward Th2-type cytokines (decreased IL4, IL10); and decreased IFNγ production. It is possible that diminished IFNγ production in the face of exposure to brain material protected exposed controls from developing disease. Regardless, among those exposed to brain tissue in this way reduced IFNγ secretion may have utility as a marker for disease severity.
Spontaneous secretion of IFNγ by peripheral blood mononuclear cells is increased at the peak of clinical disease among patients in whom AIDP is diagnosed and decreases during recovery in parallel with rises in serum concentrations of neutralizing IgG autoantibodies to IFNγ
. Additionally, IFNγ knock-out mice are protected from development of AIDP-like illness, implicating IFNγ as a critical component in development of autoimmune inflammatory neuropathies
Changes in slaughter operations in the affected abattoirs might have affected the number of workers exposed to porcine brain or the intensity of exposure and could explain why illness occurred recently despite the fact that brain removal was occurring years prior. Workers from two abattoir reported being less efficient at removing brains after the line speed increased. These workers reported being unable to place the skulls completely on the brain removal device before triggering the compressed air, causing greater splatter of brain material. Plants B and C slaughtered fewer hogs per hour, and their compressed-air brain removal designs allowed for more control by the operator, potentially resulting in less brain splatter and fewer cases than in Plant A.
Proximity to brain removal was the strongest predictor of disease. However, certain workers positioned close to the brain-removal operation did not experience disease, indicating a potential role for genetic susceptibility or other host factors. This is supported by evidence among Semple rabies vaccine recipients where unique MHC class II alleles were identified among persons in whom neuropathy developed, compared with those who did not become ill
One potential limitation of this investigation is the possibility that despite intensive case-finding efforts employing multiple methods, other IP cases were not identified. The workers in these abattoirs were reported to be highly mobile, often terminating employment rather than taking medical leave. As there was no specific biological marker for IP and symptoms were nonspecific, potential cases could have been overlooked. All interview data, including work history, were self-reported and were subject to recall bias. Although cases were identified among former abattoir workers, controls only included the workers who were working at the abattoir at the time of the investigation.
In summary, IP was more likely to occur among workers who reported close contact with brains or the job of removing brains by using compressed air. No additional cases have been identified with onsets after cessation of brain harvesting by compressed-air methods in Plants A, B, and C. Our findings indicate that swine abattoirs and other animal commodity abattoirs should not use compressed air to remove brains and should avoid any procedures that aerosolize CNS tissue. This outbreak highlights the potential for respiratory or mucosal exposure to cause an immune-mediated illness in an occupational setting and the importance of health care providers taking a careful work place exposure history. The cooperation between human and animal health organizations provided an optimal framework for this investigation and demonstrates the synergy needed to address emerging issues at the human-animal interface.
Note: The findings and conclusions in this report are those of the author(s) and do not necessarily represent the views of the Centers for Disease Control and Prevention.