Hot Literature: Refining EPM diagnostics

May 02, 2011

Few infectious diseases are more confounding to diagnose than equine protozoal myeloencephalitis (EPM). Many horses exposed to the infectious agent, Sarcocystis neurona, have significant serum antibody titers yet never develop clinical infection. Since the clinical signs of EPM vary widely and mimic other neurologic conditions, diagnosis must be confirmed in the laboratory or at necropsy.

As plasma protein concentrations increase, there is a corresponding and proportional increase in the cerebrospinal fluid (CSF) concentration of those proteins. This phenomenon is also true of antigen-specific antibodies as the passage of antigen-specific immunoglobulins across the blood brain barrier from the peripheral blood to the CSF is predictable and proportional. Therefore, antibodies in both serum and CSF does not definitively diagnose infection. Using an antigen-specific ELIZA highly accurate for detecting antibodies to S. neurona, researchers have used a mathematical means for more accurately diagnosing EPM by differentiating samples from infected and noninfected, exposed horses and blood-contaminated CSF samples.

Paired blood and CSF samples were obtained from 14 confirmed cases of EPM and 15 horses with cervical vertebral compressive myelopathy (CVM). Two CSF indices (antibody index and Goldman-Witmer coefficient) were then calculated based on results of serum and CSF antigen-specific ELIZA testing and total protein, albumin, and immunoglobulin G (IgG) concentrations. These mathematical calculations differentiate the proportion of antibodies in the CSF that originate from the blood from those that are produced as a response to infection within the central nervous system. To further challenge the reliability of this test, the negative CSF samples were then spiked with increasing amounts of blood from a horse with a high S. neurona serum antibody titer, and the testing and calculations were repeated.

The CSF indices calculated ruled out S. neurona infection with 100% specificity and were fairly sensitive as well. The Goldman-Witmer coefficient (C-value) was positive in 12 of the 14 EPM horses (sensitivity of 86%), and the antibody index had 71% sensitivity. All of the spiked CSF samples were negative using these calculations. The fact that there were no false positive results in this study is noteworthy. Both false positive results and blood contamination of CSF samples are not uncommon problems with conventional antibody titer testing on paired serum and CSF samples, which can delay accurate diagnosis and may result in inappropriate and costly treatment. The study designers emphasize that these indices are easy to calculate from readily available ELISA testing, provide values that are easy to interpret, and eliminate the need to obtain repeat CSF samples because of blood contamination. Details on the mathematical calculations and suggested cut-off values are presented in the published study.

Furr M, Howe D, Reed S, et al. Antibody coefficients for the diagnosis of equine protozoal myeloencephalitis. J Vet Intern Med 2011;25(1):138-142.

Link to full text of article: