Our findings support that RT-QuIC of individual and pooled third eyelids from white-tailed deer can be used to detect the CWD prion in white-tailed deer in a diagnostic setting. This demonstrates that RT-QuIC is reliable to detect CWD from naturally infected captive and wild white-tailed deer across laboratories building upon previous results from inoculated white-tailed deer in a research facility27.
The RT-QuIC assay did not identify any false positives (100% specificity) in the Wildlife Futures laboratory but was unable to detect 7 third eyelids from deer that had lymph node and/or obex previously diagnosed as CWD-positive by ELISA/IHC (sensitivity 94%). Genetic analysis was performed to understand if there was a genetic bias among these discrepant samples, and results showed that 3/7 carried either QH or GS heterozygosity (considered more resistant) but the remainder were homozygous (considered less resistant). Overall, the number of discrepancies between ELISA/IHC of lymph node and/or obex and RT-QuIC on third eyelids were rather low in our study (8/250 or 3%, including the discrepancy for sample 97). As 3/7 deer did harbor heterozygous genetic variations in the PRNP gene that are believed to delay the disease progression of CWD (one 95 H/Q and two 96G/S) compared to homozygous variations that are more susceptible (QQ/GG/AA/QQ)53, it is possible that the genotype could have contributed to these false negative results. However, in previous studies it was found that the more susceptible homozygous genotypes for PRNP were most prevalent in Pennsylvania white-tailed deer and that the few individuals that were heterozygous (HQ or GS) and thus carrying a more resistant PRNP genotype still tested positive indicating that these genotypes likely do not have a significant impact on RT-QuIC test performance40,53. However, we consider this sample size to be too small to make any conclusions on whether QH or GS would be more likely to test negative for CWD by RT-QuIC on third eyelids. An expanded study in wider geographic regions can further tease out discrepancies encountered with testing and use of RT-QuIC.
The discrepancy between the positive ELISA/IHC results of the lymph node and/or obex and the negative RT-QuIC results of third eyelids could be because the infectious prion protein had not yet aggregated in the third eyelid of these individuals, or that the 50 ± 5 mg piece of third eyelid we tested did not contain lymphoid tissue or detectable concentration of infectious prions, similar to what has been previously reported for IHC of lymph nodes54. This could also explain the discrepancy which was observed for the CWD-positive deer where one third eyelid was considered negative by RT-QuIC but the contralateral eyelid was considered positive by IHC. Additional studies could further determine differences in timing of when different lymphatic tissues test positive during the course of a CWD infection. Nevertheless, the high specificity of this assay yields a confident positive result, which is particularly important when evaluating results from samples in an area close to and/or outside an established CWD endemic zone.
The inter-laboratory cross validation supports that results are reproducible in separate laboratories and by different technicians, with minor differences. The discrepancy between the two laboratories’ results was that one third eyelid from a deer that previously tested negative by ELISA/IHC on the lymph node tested positive by RT-QuIC in the CWD Evolution laboratory, while it was considered negative by RT-QuIC in the Wildlife Futures RT-QuIC laboratory. Reevaluation of the IHC staining for that individual’s lymph node was consistent with a weak CWD-positive result. Thus, it is likely that this individual was in the early phase of the disease with a low number of stained follicles in the lymph node that were missed in the initial screening of the histology slide. This finding emphasizes the high sensitivity of RT-QuIC as individual 97’s third eyelid had multiple wells cross the threshold within 10–49 h in both laboratories but was not considered positive at initial IHC screening of the lymph node associated with that deer. Despite this discrepancy between laboratory results, the sensitivity and specificity were very similar (Table 4) which illustrates how RT-QuIC of third eyelids can produce reliable results across laboratories and technicians.
By following the plate reader manufacturer’s recommendations of adjusting the gain by the wells with the highest expected fluorescence value (known positive control), we were able to significantly reduce the time it took for a sample to cross the set threshold to be considered positive without compromising the sensitivity and specificity. For individual samples, we were able to reduce the time to threshold by 8 h and 9 h for pooled samples which is beneficial as it can facilitate faster result reporting and the possibility to overall reduce the assay length.
The lower sensitivity that was observed in the pools of 10 samples occurred in pools that included only 1 or 2 third eyelids from deer with lymph node and/or obex that had previously tested CWD-positive. For the initial testing where the gain was manually set to 1350, 3/5 pools that included 1 third eyelid from a CWD-positive deer and 4/5 pools that included 2 third eyelids from CWD-positive deer were considered positive. However, by adjusting the gain to a known positive control, we were able to increase the sensitivity for the pools of 10 samples which contained 2 third eyelids from CWD-positive deer from 4/5 to 5/5 positive results, which increased the overall sensitivity of pools of 10 samples from 90 to 93%. Combining 10 samples in a pool was less ideal due to a sensitivity of 75% (6/8) for pools which only contained one third eyelid from a CWD-positive deer, whereas combining 5 samples per pool offered 100% sensitivity (21/21) and 100% specificity (11/11), including all pools which only contained one third eyelid from a CWD-positive deer (7/7 positive). Pooling 10 samples increases the possibility of false negatives in situations where only 1/10 samples is from a CWD-positive deer, possibly due to dilution or inhibitory factor of pooling this many samples together. In contrast, the pooling of 5 samples was promising and would allow laboratories to test 150 individuals in 30 pooled samples, which is 5 times the number of individuals that can be tested individually by RT-QuIC using 3 replicates per sample on a 96-well plate. This demonstrates the ability of RT-QuIC testing on third eyelids to increase the total number of individuals that are screened for CWD without reducing sensitivity and specificity.
As RT-QuIC is currently not a validated test for diagnosing CWD, immunohistochemical staining of third eyelids showed to be a promising follow-up or confirmatory test following a positive RT-QuIC result. However, interpretation of the third eyelids by IHC was challenging for several reasons, including the overall paucity of visible lymphoid follicles compared to lymph nodes, the cartilage hindering smooth and consistent cutting of the microtome resulting in tearing and shattering of the section placed on the slide, and a wide variety of immune staining intensity including unexpectedly common false positive or non-specific staining. Considering third eyelids are unique compared to lymph node and obex with respect to tissue size, shape, and texture, further optimization of this tissue is required during all stages of testing. This includes the collection phase to minimize crushing of the tissue upon extraction of the deer, fixation stage to ensure the tissue is flat to maximize the surface area to view on the slide, slide processing stage to reduce the cartilage hindering the ability to smoothly and consistently cut tissue sections with the microtome, and staining stage to ensure appropriate duration of antigen-antibody binding to minimize non-specific or false positive staining. Additionally, determining a minimum number of lymphoid follicles that need to be present and examined to ensure a non-detection similar to lymph nodes would be a valuable step in validation and should be established prior to widespread use. As lymphoid tissue in the third eyelid can be less developed in juveniles55, age of the animal was considered as one reason for the paucity of interpretable lymphoid tissue. However, all three fawns included in this analysis were correctly identified as positive (n = 1) or negative (n = 2) suggesting factors other than age are more likely to affect the required amount of lymphoid tissue available to interpret, including the aforementioned factors related to fixation, trimming, and slide preparation. Nonetheless, IHC was considered useful as a confirmatory test and often (case from Fig. 6), but not always, showed strong, intense immunoreactivity that was easily interpretable by diagnosticians experienced in interpreting IHC slides. Further optimization of IHC-staining third eyelids is necessary before the authors would like to recommend it as part of a diagnostic tool for CWD.
Our results support that third eyelids could be used as an alternative or in addition to the current accepted tests for detecting CWD, with three major management implications. First, as the tissue is easier to collect than lymph nodes, and does not require deep dissection, hunters could more easily collect this sample, which would greatly reduce the financial cost and workload of wildlife agencies. Further investigations that include third eyelid and lymph nodes and/or obex submitted by hunters should be conducted to evaluate the assay’s sensitivity and specificity. Secondly, since third eyelids can be tested in pools, it enables laboratories to screen a higher number of samples at once, decreasing the costs of running each sample individually, and increasing the turnaround time for reporting results. Third, easier sample collection and the ability to screen a higher number of samples at a time enable wildlife agencies to expand their surveillance capabilities.
While IHC is considered the gold standard by the USDA to diagnose CWD, it needs to be appreciated that samples from wild cervids are not always in an ideal condition for this assay, where delays in submission, decomposition, and freezing can make disease detection, particularly in early stages, challenging to assess by a pathologist. In addition, many states have areas that are under-surveyed for CWD (such as areas thought to have no CWD prevalence) where providing a highly sensitive (94%) assay with the above considerations can improve a state’s disease surveillance plan.
We demonstrated that third eyelids from white-tailed deer can be tested by RT-QuIC individually, and in pooled samples with high sensitivity and specificity. IHC of third eyelids can be used as a follow-up test to confirm a sample that tested positive by RT-QuIC but should not yet be used in isolation. Together, these results provide support for further investigation of third eyelids as a potential target sample for CWD diagnostics in an effort to reduce the cost of testing while expanding surveillance capacity.
