The following study was presented at the 49th meeting of the American Association of Veterinary Laboratory Diagnosticians in Minneapolis MN in October 2006.  The purpose of the study was to establish whether PCR testing of pooled ear notch samples from cattle to detect persistently infected animals might miss a proportion of animals.  In the experience of our virology laboratory, it does.  That is why we continue to offer the ear notch antigen capture ELISA, rather than offering a cheaper test which will miss some persistently infected animals.  Missing PI calves by bulk testing undercuts the purpose of trying to detect and eliminate ALL persistently infected calves in a herd. 

 

Please note the possible error rate: pooling 10 samples may miss as many as 1 in 10.  Pooling 100 or more samples may miss a PI animal as many as 50% of the time.

 

D. O'Toole

Director, WSVL

 

Evaluating stability, size requirements, viral load and pooling of ear notch samples in BVDV testing

 

J. F. Ridpath1, B. E. Hessman2, J. D. Neill1, R. W. Fulton3, D. L. Step3, A. Zimmerman4, and C.C.L. Chase5

 

Infections with bovine viral diarrhea viruses (BVDV) are a major source of economic loss for the U.S. cattle industry. BVDV control efforts in the U.S. are geared towards identifying and eliminating persistently infected (PI) animals. Several tests based on detection of either antigen or viral RNA in blood, serum, bulk milk or skin biopsies are currently in use for detecting PI’s.

 

While ear notches have become one of the samples of choice, there is little information available regarding sample size requirements and stability. Further, while pooling of ear notch samples has been proposed for reducing the cost of surveillance programs, the viral load available for detection from ear notch samples is largely undetermined. The purpose of this study was to establish working parameters for sample size, viral detection limit and sample storage conditions for real time PCR and antigen capture ELISA and to compare reproducibility of tests between three laboratories based on a blinded panel of pooled and unpooled samples.

 

For these experiments, notch samples were extracted by soaking for a minimum of 60 min in 2 mls PBS. There was no difference in the amount of virus detected in ear notch extractions using ear notch samples weighing between 0.75 and 0.05 gms. However, extracts generated using ear notches weighing 0.03 gms or less containing at least a log lower virus concentration. There were no significant differences between storage at -20, 4 and 25 C for 7 days. In contrast, detection was reduced in lyophilized samples and samples stored at 37 C for 7 days. The concentration of virus in standard size ear notch (avg. weight 0.75 gms) extractions averaged 452.3 virions/ml. Of the 153 samples evaluated for virus concentration, 16 (10.5%) had between 10 and 100 virions/ml, 86 (56.2%) had between 100 and 1000 virions/ml, 50 (32.6%) had between 1000 and 10,000 virions/ml and 1 (0.7%) had more than 10,000 virions per ml. The detection limit of the real time PCR test was determined to be 10 virions/ml.

 

The most consistent results between laboratories was observed using the ACE tests on non pooled tissues followed by ACE tests on samples pooled 1 to 5 and ACE tests on samples pooled 1:10. The least consistency was observed with PCR based tests based on sample pools of 1:100 or greater.

 

These results suggest that ear notch samples are relatively stable for at least 7 days if stored between -20 and 25 C. However, both exposure to higher temperature and drying reduced detection. Similarly the amount of virus extracted was not significantly affected by sample size over a wide weight range. The concentration range of virus in ear notch extractions and the detection limits of real time PCR suggest that pooling of samples in surveillance programs must be approached cautiously. Pooling of 10 samples, where a sample pool includes 1 positive and 9 negative samples, could result in the failure to detect 10% of the samples used in this study. Pooling of 100 samples, where the sample pool includes 1 positive and 99 negative samples, could result in failure to detect over 50% of the samples used in this study.

 

The comparison of the blind panel results was consistent with these assumptions.

 

1Virus and Prion Diseases of Livestock, National Animal Disease Center/ARS/USDA, Ames, IA

2Haskell County Animal Hospital, Sublette, KS

3Department of Veterinary Pathobiology, Oklahoma State University, Stillwater, OK

4Rural Technologies, Inc., Brookings, SD

5Department Veterinary Science, South Dakota State University, Brookings, SD