Sarcocystis neurona SAG2/3/4 ELISA Educational Bulletin August 2010

Diagnosis of equine protozoal myeloencephalitis (EPM), a neurologic disease in horses caused by the parasites Sarcocystis neurona and less commonly Neospora hughesi, has historically been a challenge. Tests currently available detect IgG antibodies against S. neurona in serum or CSF. However, due to the high exposure rate of horses to the parasite many asymptomatic animals carry moderate to high levels of serum antibodies, resulting in false positive blood tests. CSF testing is much more predictive of active disease, but results can be affected by blood contamination and normal passive transfer of antibodies across the blood-brain barrier.

How to Improve on EPM Diagnostics?

Two new quantitative enzyme-linked immunosorbent assays (ELISAs), designed to be run concurrently, were developed at the Maxwell H. Gluck Equine Research Center at the University of Kentucky and are now exclusively available at Equine Diagnostic Solutions, LLC (EDS). These assays incorporate three unique antigens of S. neurona, SAG2, SAG3 and SAG4, for the measurement of an IgG response to them (Hoane et al., 2005 and Yeargan and Howe, in review) and results are expressed as endpoint titers. They were validated against samples from 58 necropsy-characterized neurologic cases. From these, 83.3% sensitivity and 100% specificity were achieved using these ELISAs for EPM diagnosis. Blood contamination of CSF was determined to be confounding only with RBC counts greater than 10,000/ul.

While endpoint serum titers alone are not specifically predictive of an EPM diagnosis, endpoint CSF titers of ≥1:20 have a higher correlation with an EPM diagnosis. Using these new assays, the serum-to-CSF titer ratio is the most informative and predictive criterion for the differential diagnosis. This calculation is based on the premise of “proportionality” of antibodies across the blood-brain barrier (BBB). Movement of antibodies across the BBB occurs by normal passive transfer. However, with CNS infection, proportionally higher levels of antibodies are found in the CSF, reflecting intrathecal antibody production. Utilizing these three S. neurona specific antigens, this ratio becomes very predictive, with ratios of <100 strongly correlating to an EPM diagnosis. An antibody specific C-value or index can also be calculated using the serum and CSF endpoint titers in conjunction with either total IgG levels (C-value) or Albumin levels (Ab Index). These calculations are useful in determining intrathecal antibody production versus blood contamination or increased blood brain barrier permeability.

Test performance for field cases was conducted in collaboration with a large equine hospital and the data were presented at a national veterinary meeting (Reed, et. al, ACVIM 2010). Paired serum and CSF samples from 131 clinical cases were tested with the two ELISAs (at Gluck) and by standard western blot (at EDS). Neurologic status was monitored at presentation and on follow-up examination for most patients. Histopathological results (necropsy) were available for 15 cases. After all lab and clinical data were tabulated, cases were grouped by diagnosis and the data analyzed. As concluded from the necropsy data set, serum:CSF titer ratios of <100 were highly predictive of EPM, even when the serum titers were low (i.e. <1000). The results supported the clinical value of testing both CSF and serum.

How do these new tests compare to other EPM tests?

A test comparison study was performed with samples from 20 of the field study cases, including 12 of the necropsy cases and 7 other cases with well defined diagnoses. S. neurona ELISA Titer Ratio results were compared to 3 different test formats, IFA, SAG1 ELISA, and standard western blot. The sample submitted for testing was based on the recommendations of the testing laboratory for an EPM diagnosis: Serum for IFA and SAG1 ELISA, CSF for western blot, and serum plus CSF for S. neurona ELISA. The sample set consisted of 7 cases diagnosed with CVM (5 necropsy) and 13 cases diagnosed with EPM (7 necropsy).

• IFA serum results were interpreted as positive EPM for those samples with titers of >1:80 or >55% probability. 3 of the 7 CVM cases tested as false EPM positive and 6 of the 13 EPM cases tested as false EPM negative by IFA.

• SAG1 ELISA serum results were interpreted as positive EPM for those samples with titers of ≥1:32. 2 of the 7 CVM cases tested as false EPM positive and 9 of 13 EPM cases tested as false EPM negative by SAG1 ELISA

• Standard western blot CSF results were interpreted as positive EPM for those samples testing low positive and positive. 1 of the 7 CVM cases tested false EPM positive and 2 of the 13 EPM cases tested false EPM negative by standard western blot.

• S. neurona ELISA serum/CSF Titer Ratios were interpreted as positive EPM for those samples with titer ratios of <100. None of the 7 CVM cases tested false EPM positive while 2 of the 13 EPM cases tested as false EPM negative by S. neurona ELISA Titer Ratios.

• Two cases in this sample set tested as false EPM negative by all four assays with one outlying SAG1 positive result.

Evaluation of the three sets of data for clinical usefulness of the SAG2/3/4 ELISA’s and diagnostic capability concluded:

• Serum titers do not independently correlate with an EPM diagnosis, while serum:CSF titer ratios are highly predictive.

• Serum:CSF titer ratios were the most accurate diagnostically when compared to the three previously most common S. neurona IgG tests.

• The importance and diagnostic value of performing CSF taps is reaffirmed.