Chronic wasting disease (CWD): research

How well does antemortem testing work?

Piel and colleagues ran a blinded independent validation of RT-QuIC on mock biopsy samples of rectal mucosa from 284 white-tailed deer, taken from a frozen archive of intact rectoanal junctions collected when CWD-positive farmed herds were depopulated. All deer were preclinical, and infection status came from the regulatory record, which used immunohistochemistry on medial retropharyngeal lymph nodes and obex.

Peak overall diagnostic sensitivity was 76.8 percent with a commercially sourced recombinant substrate and 73.2 percent with a laboratory-produced substrate. A pre-analytic precipitation step improved the detection limit. Performance varied with the diagnostic cut points chosen, meaning the minimum number of positive wells and the assay run time, and with deer attributes, specifically preclinical infection stage and prion protein genotype.

The rectal mucosa near the rectoanal junction is used because prions accumulate there early and because pinch biopsy is safe in a living animal. The sensitivity figure is the ceiling for that sample type in preclinical animals, and it is well short of what a clearance test for an individual animal would need. Reported results for retropharyngeal lymph node are higher: Munster and colleagues report up to 96 percent sensitivity and 100 percent specificity on formalin-fixed nodes and up to 98 percent sensitivity and 99 percent specificity on fixed embedded nodes, but these are post-mortem samples.

How much does host genotype change?

Two findings need to be read together.

Wolfe and colleagues compared mule deer of PRNP genotypes 225SS and 225FF under natural environmental exposure. All three 225SS deer had positive tonsil biopsies by 710 days post-exposure and developed classic signs between 723 and 1200 days. All three 225FF deer also became infected, but the two that survived beyond 720 days had consistently negative biopsies, developed more subtle clinical signs and died at 924 and 1783 days post-exposure. Post-mortem, the 225FF deer were suspect by ELISA with negative or equivocal immunohistochemistry in lymphoid tissue, and both clinically affected animals had spongiform encephalopathy without immunohistochemical staining in brain tissue. Similar cases occurred among five further captive 225FF deer outside the experiment.

Seerley and colleagues sequenced PRNP in retropharyngeal lymph node samples from 358 wild Montana mule deer and identified 36 variants, many not previously reported, then tracked them by hunting district, year of harvest and CWD status. Computational structural prediction for S225F was consistent with the protective role previously proposed for that variant, by way of reduced fibril formation, while the novel variant V12F was predicted to alter protein interactions and was found in CWD-positive samples, which gave prion seeding activity by RT-QuIC.

The practical consequence runs in one direction: a genotype that lengthens incubation and suppresses lymphoid deposition also defeats the tests that rely on lymphoid tissue. Prevalence estimates built on lymph node immunohistochemistry will therefore undercount infection in resistant genotypes, and a protective variant is better described as delaying and altering the disease than as preventing it.

Is there one agent or several?

Strain typing in bank voles separated Norwegian CWD from North American CWD on incubation period, neuropathology, regional deposition and protease-resistant core size, and found that Norwegian reindeer and moose carried different strains. For risk assessment this means results are isolate-specific; for diagnostics it means assay substrates and cut points validated against North American isolates have not been validated against European ones.

How should surveillance be designed?

Mysterud and colleagues addressed the conflict between culling and conservation by modelling proactive hunting surveillance, which targets demographic groups with a higher likelihood of infection and a lower reproductive value rather than removing whole herds. Applied to reindeer populations to substantiate freedom from infection, the approach reached 99 percent probability of freedom, defined as fewer than 4 reindeer infected, within 3 to 5 years, against about 10 years with ordinary harvest surveillance. The authors state that implementation uncertainties linked to social issues remain challenging with this form of host culling.

Probability of freedom from infection is a statement about surveillance sensitivity, not a biological finding. It is also sensitive to the assumed test sensitivity, which the antemortem figures above show to be genotype-dependent.

Related: scrapie, BSE, biosafety and laboratory safety.