A comparison of the polymerase chain reaction with standard laboratory methods for the detection of EHV-1 and EHV-4 in archival tissue samples

A detection system incorporating the polymerase chain reaction was compared with the use of histopathology and virus isolation to determine the presence of equid herpesvirus type 1 or equid herpesvirus type 4 in equine tissues submitted to a diagnostic laboratory. When the polymerase chain reaction was performed, these tissues had been stored for up to 3 years. Thirty-eight tissues representing 14 cases had been stored embedded in paraffin wax. Analysis of these tissues using the PCR gave predictive values of 1.0 and 0.91 for a positive and negative result respectively, and sensitivity and specificity values of 75% and 100% respectively. Fifty-three tissues representing 28 cases had been stored immersed in 10% formalin. Analysis of these tissues gave predictive values of 0.44 and 0.42 for a positive and negative result respectively, and sensitivity and specificity values of 28% and 57% respectively. The poor results obtained with this group of tissues was attributed to contamination of the samples during wax embedding. Viral DNA could not be amplified from older tissues. These results indicate that under appropriate conditions the polymerase chain reaction is reliable when applied to tissues collected for routine diagnosis. However, it is less reliable when samples for analysis are handled collectively. Also, storage of tissues in wax blocks for 14 or more years inhibits later amplification of viral DNA from these tissues. The implications of these results to the application of the polymerase chain reaction to routine laboratory diagnosis are discussed.