Skip to content

The EPM Society is meeting this month in California.  Pathogenes is presenting 3 papers, one is of particular interest to the Society and that is Development and validation of assays offered by Pathogenes, these are tests used in our consulting services.  The nuts and bolts of our tests are published, including test format, dilution factors, assay development, samples used for validation and testing, and results interpretation.  A big part of testing is the sensitivity and specificity of the test.  We have enough experimental and field data to allow veterinarians to use a predictive value for EPM.  The predictive value depends on the amount of disease in the population.  For example, disease is more prevalent in Texas, the predictive value of a positive test is more meaningful in Texas compared to Idaho, we found no EPM in Idaho.   Yet, some details interest the group, such as strains of organisms and the validation protocols.

Background for non-test developers: The USDA,  under animal plant health inspection service (APHIS) directs animal health and veterinary  biologics.  The Veterinary Biologics Program implements provisions assuring that effective diagnostics are available and appropriate standards are developed.  They also issue licenses and inspect products and facilities.  They issue memoranda that relate to licensing diagnostic tests.  These directives outline the steps to provide sensitivity and specificity data, ruggedness,  receiver operating characteristic curves (ROC) that test sensitivity versus a false positive rate. The ROC curve is useful visualizing the compromise between sensitivity and specificity for different cutoff values and for selecting a cutoff value.

Our mission is to develop patent protected technologies for licensing and we use USDA’s requirements to guide our work.  For example, developing assays.  A first step is understanding what you want to measure, the analyte (that can be an antigen-a protein related to a specific organism, an antibody, or a genetic sequence). The next steps are  deciding on a range of concentration, sample type, and the potential for cross-reactions.

A big hurdle is the Gold Standard selected for a diagnostic test.  You will read about “Gold Standards” and “EPM” testing making this topic worth understanding.

USDA’s context for a Gold Standard is to provide “an accepted means of determining the diagnostic status (positive versus negative) of a diseased animal”.  These same animals should be evaluated by the test and the Gold Standard samples are used to determine the accepted “true diagnosis”.  Using EPM as an example, USDA requires reference standards (serum and CSF for example) from 20 animals.  They accept samples from field cases.  USDA requires confirmation of disease: organisms in the brain of the animal-the accepted definition of EPM and, here is the kicker, enough of the samples (serum and CSF) to provide the reference standard over the life of the test.  Our discussions with USDA confirm this is not an achievable goal.

Gold standard samples from 20 natural cases of EPM are difficult to obtain. The horses must not be treated prior to sampling because treatment changes the diagnostic analyte. S neurona has three serotypes (each analyte is unique to each serotype) that must be detected by the test. Sixty horses with natural disease are needed for SAG 1, 5, and 6 testing.  Common analytes SAG 2, 3, and 4 are not specific for the disease EPM because these SAG’s are present in non-EPM causing protozoa.

A licensed EPM test is not happening.

A core issue with the disease caused by S neurona is the contribution of inflammation to the EPM syndrome.  The organisms can be eliminated leaving some horses with treatable clinical signs of disease.  Detecting analytes that identify S neurona in conjunction with inflammation is clinically useful.  Each veterinarian must weigh the “Gold Standard” that was used for the test and it’s usefulness in their treatment decisions.

USDA will license an antibody test that detects a specific antigen.  The Gold Standard for these tests can be obtained by vaccination. For example, we vaccinated 50 animals with recombinant SAG 1 to obtain enough reference material for submission to USDA.  In addition, USDA requires the “master seed”, a number of vials of the analyte used in test development.  The master seed is validated from serial passages of the recombinant organism, all tested and documented, so a 3rd generation seed (serial)  is proven to be unchanged by passage. The requirements to license our SAG 1, 5, 6 tests are 3 master seeds (with validated serials) and 20 vaccinated animals.  The test is “for the detection of antibody against serotype 1 (or 5 or 6) of S neurona in the serum or CSF of horses”. We do not diagnose the disease EPM. We report the level of antibody against the analyte SAG 1, 5, or 6 of S neurona. We also use a serum test to evaluate the inflammatory reaction to infections.  This panel of tests forms the basis of our consultation.

We are pursuing a license for the ante-mortem diagnosis of equine muscular sarcocystosis due to S. fayeri  in horses. The assay will detect disease (the presence of muscle cysts) so the USDA Gold Standard for disease applies.  Reference standards are produced from field cases of EMS.  EMS is also easy to produce experimentally, a source of material for reference standards.


Another disease that warrants a licensed diagnostic test is polyneuritis equi, a demyelinating polyneuropathy that results in ataxia (classical) or cranial nerve deficits (atypical).  The USDA Gold Standard for polyneuritis equi is met with field cases.   We identify cases of polyneuritis equi by clinical signs and antibody against myelin protein.  The disease process, a response to an infectious organism, results in demyelination of nerves followed by remyelination of that nerve.  The diagnostic we are developing identifies the demyelinating process.

S. neurona, electron micrograph

Apicomplexan parasites are intracellular protozoa that are responsible for a great range of diseases in man and animals. The family Sarcocystidae contain the cyst-forming coccidia, cysts form in muscle tissues of the prey-host. Carnivores (prey) eat the infected muscle tissue and that completes the parasites lifecycle. It is generally thought that muscle cysts (sarcocysts) cause no pathology in the prey-host, except perhaps in debilitated animals. A compromised host may lack a robust immune response that can limit the infection. Exposure to protozoa, and the resulting immunity, builds resistance to infections. It is well known that overuse of antimicrobials and antiparasitic agents was an unwise strategy, it is no different in the fight against Sarcocystis. An unintended consequence of chemical prophylaxis is an animal with no natural immunity.

Sarcocystis fayeri produces sarcocysts in horse muscles and like most Sarcocystis infections, this finding is considered incidental. Early research dismissed S. fayeri as a factor in equine protozoal myeloencephalitis (EPM). Experimentally infecting ponies with S. fayeri and evaluating the immune responses convinced the researchers that infections by two species neurona and fayeri, were indistinguishable using the IFAT test. Important molecular tests identifying S. neurona resulted in 22,076 nucleotide sequences. In contrast, S. fayeri has 15 reported sequences-all are the small subunit ribosomal RNA gene. Variability in the small subunit ribosomal gene is useful to identify Sarcocystis species. Incorrectly, S. neurona and S. falcatula were reported as the same organism based on synonymous regions of this gene.

Our investigations reveal the need for a reassessment of the pathogenesis of S. neurona infections in horses and a need evaluating the role of the immune responses in equine disease. There are reasons that S. fayeri should get a more serious look.

Sarcocystis neurona causes muscle weakness in horses but the parasite isn’t thought to develop cysts in horses. There are four horse-related species that develop cysts in horses: S. asinus, S. bertrami, S. equicanis, and S. fayeri. Canids, including dogs, are the reported definitive hosts for these organisms.

Generally sarcocysts are not associated with inflammation in horse muscles (examined by histopathology), although in some debilitated horses, muscle degeneration is reported. Equine sarcocystosis, considered a mild disease, but can cause fever, apathy, anorexia, myositis, difficulty chewing, muscle weakness, autoimmune disease and sometimes hair loss. Surprisingly the profound muscle weakness exhibited clinically doesn’t correlate with the mild lesions observed by histopathology. And this leads some to hypothesize that there is a toxin associated with muscle infections.

The toxin idea isn’t new. There were reports of Sarcocystis-cyst toxins, called sarcocystine, in 1899. A toxin was isolated from cattle muscle cysts and characterized one hundred years later. Yet the possibility of muscle toxins causing disease in horses hasn’t been evaluated. Sarcocystine causes disease in people. A toxin found in raw horsemeat was associated with human food poisoning. The toxin was isolated from S. fayeri sarcocysts and toxic effects were evaluated in rabbits. The protein toxin, histopathological lesions, animal feeding experiments, rabbit enterotoxin assays, enzymatic digestion experiments, and heat/acid lability assays are similar between S. cruzi and S. fayeri-cyst toxins.

So far, initial research on the enterotoxin from S. fayeri sarcocysts is in humans. It would be interesting to explore a relationship between a S. fayeri sarcocystine and myositis in horses.

There is enough molecular information to investigate sarcocystine as a cause of muscle weakness in horses. The proteomics suggest the S. fayeri sarcocystine is homologous to proteins of Eimeria tenella and Toxoplasma gondii. Protein similarity, if high enough, would indicate conservation of the protein and a role in parasite survival. Protein similarity between organisms would also sink the protein as a good diagnostic to implicate an organism. The best it could be is a diagnostic for protozoal myositis.

It is possible that detecting the sarcocystine would benefit the treatment of sarcocystosis in horses.

Coccidiosis in horses is complicated. It is important to examine many factors before initiating practices that have unintended consequences. It took many animal infection studies to correct the false claim that S. neurona and S. falcatula were synonymous. False assumptions have plagued EPM research for 25 years. This has cost many horses their lives. Studies that examine the effect of equine coccidial infections and the immune response to infection should dominate the conversation. Initially clarifying the definition of EPM and the pathogenesis of disease are important.

Recent attention to daily prophylaxis to reduce antibodies against S. neurona may have unintended consequences in the disease EPM. One must weigh the need for antibody prevention (the consequence of prophylaxis) against the risk of neurologic disease and the consequences to the reduction of natural protective immunity against coccidiosis in the horse. Natural immunity holds S. fayeri in check and probably minimizes the effect of cyst toxins on the infected horse.

Overuse of antimicrobials led to superbugs. The human pharmaceutical industry will spend the many millions of dollars to develop new antimicrobials—if they can. The animal pharmaceutical industry will not spend any dollars on developing new anti-protozoals for the treatment of EPM, especially if there is a resistant superbug. Unintended consequences of prophylaxis may be the release of toxins from S. fayeri cysts. A veterinarian may misdiagnose a toxic event for an active infection—how can one distinguish these cases?

Lyme disease (infection with Borrelia bugdoferi) can confound EPM research because the diseases may have similar presentations. Clinically, there is no clear distinction that indicates a horse with Lyme disease from a horse with EPM.  Researchers are making an effort to describe the typical Lyme case.  Signs consistent with Lyme disease include ataxia, peripheral neuropathies, cranial nerve inflammation, muscle wasting, skin sensitivity, stiff neck, and possibly uveitis.  Changes may be observed in CSF fluid prompting collection of fluid by a spinal tap to look for organisms or antibodies against B bugdorferi.

There is no good diagnostic test to define the actively infected horse with neuroborelliosis (organisms in the central nervous system).  Detecting the organism in the central nervous system isn’t easy and requires molecular analysis and reasonable certainty that there is antibody production in the CSF, not antibody contamination from the periphery.  Similar to a supportive diagnosis of EPM, some clinicians use the serum/CSF ratio—generally a referral facility.  A test-positive interpretation may necessitate detection of antibodies against several Borrelia antigens as well as paying close attention to dilution factors used in the laboratory.


CDC Map of Lyme Disease in the US

Where the horse lives is an important consideration when including Lyme as the cause of neurologic disease because Borreliosis is a regional disease.  Check out the CDC’s map, shown above. Serum testing for Borrelia antibodies by ELISA is helpful to support a diagnosis, most Borrelia infected horses have serum antibodies.  This is another disease in which ruling out other causes of the signs is key.  And a negative serum test supports another cause of neurologic dysfunction and maybe helpful to rule in other diseases on the differential diagnosis list.

Several treatment protocols for Lyme are published, however, the ability to eliminate the organism by antibiotic therapy (in some infections) may not be possible.  Chronic disease is suspected in horses just like the syndrome in people.  Lyme disease does cause inflammation increasing the expectation that inflammatory markers like C-reactive protein will be elevated. Likewise, treating the inflammation can alleviate some signs of the disease and is useful.

Because horses can have antibodies against S. neurona, horses with and without EPM can have Lyme.  Lyme and EPM are a serious combination and can be lethal.  In our research it is important to include horses that have disease exclusively due to EPM.  It helps our final FDA submission that we are conducting laboratory studies because we know the horses don’t have Lyme disease.  But we’d like to avoid the  issues that would complicate our analysis of field cases.  That is why we ask you if Lyme disease is a consideration, you’ll find it on our submission form.  Being aware of Lyme, the regional nature of the disease, and ruling it in or out in a horse with ataxia is important to us.

Research developments newly reported for Sarcocystis neurona may impact horse owners their veterinarians.  A novel genotype XIII was reported by Barbosa et al in the International Journal for Parasitology (2015).  This novel genotype is a sea mammal-virulent SAG 1 strain supporting SAG 1 and 5 antigen types dominate animal disease. This strain is vertically transmitted, from the mother to the fetus indicating S. neurona is more like than unlike other pathogenic protozoa.  Our pending publications, reviewed in our last two blogs, report new tests for horses with recurrent or residual signs of EPM that seek to clarify the role of inflammation in suspect-EPM horses.  The bottom line is that the key to maintaining a healthy horse is management through testing and examinations and understanding the pathogenesis of disease.

Sarcocystis neurona possesses one of six major surface antigen genes, SAG’s 1-6, on its outer surface.  The horse makes antibodies to these SAG’s and the antibodies are detected in the serum by ELISA testing.  Minor differences within the SAG genes allows classification into genotypes, or antigen types.  For example a SAG 1 S. neurona may be antigen type II or XIII.  The horse can only distinguish between SAG’s 1, 5, or 6 (serotypes) not antigen types.  The SAG’s 2, 3, and 4 are genetically variable between serotypes, are present in all Sarcocystis, and allow molecular biologist to examine differences between SAG genes.  Geneticists look at allelic variation within the SAG genes and that allows them to sub-classify S. neurona into genotypes or antigen types.

We developed three SAG specific ELISA tests based on recombinant SAG 1, 5, and 6, the strains that infect horses .  The specificity of these tests allows us to distinguish between serotypes by the antibodies made in response to infection. The majority of all disease caused by S. neurona in animals is due to SAG 1 and SAG 5 serotypes.  There may be virulence differences between the S. neurona SAG 1: antigen type II or XIII (discussed in Barbosa’s paper).  What is clinically relevant in the sick horse is recognizing the  serotype.  Measuring specific antibodies allows the veterinarian to identify resistant infections, determine the response to treatment, and distinguish relapse versus re-infection.

Our newest work identifies horses that have chronic inflammation.  Inflammatory responses cause the clinical signs often associated with EPM.  Some horses won’t respond to antiprotozoal agents because the protozoa are gone.  A frustrating clinical presentation is identifiable with our new serum testing, MPP and IL6 ELISA’s.  Our approach to managing these horses has not changed, we still measure SAG antibodies pre- and post-treatment.  We assess the horses by gait score before and after treatment.  We monitor the CRP serum concentration.  What has changed is that we can identify horses that will relapse and give the veterinarian an explanation why and a management program.

It is well known that equine serum samples show variation in reactivity to different surface antigens of S. neurona.  The most useful clinical point: it is not the level of antibody (titer) present in a horse’s serum that is important, but noting that the levels rise with duration of infection.  Another general rule is that the first experience with infection (naïve horse) will induce antibody production. The levels are minimal and short lived (8 weeks or so).  A horse experienced with infections will get and maintain a higher antibody level up to 5 months in some animals.  Management of EPM cases requires multiple serum analysis.  A single point test can’t decipher a new infection or a relapse. Multiple tests can suggest it the animal has naive infection or chronic exposure.  The horse with chronic exposure is more likely to experience abnormal immune responses that may look like EPM but really suffer from chronic polyneuritis.  It is important to distinguish these infections because the clinical management differs.

There is a report for a new trivalent SAG chimera ELISA test for efficient detection of antibodies against S. neurona .  This is an ELISA test that seeks to reduce the time, materials, and cost associated with running multiple ELISA’s using SAG 2, 4/3.  The diagnostic protocol involves using the the SAG ELISA to determine a consensus serum-to-CSF ratio, ratios less than 100 suggest that antibodies against S. neurona are being produced in the CNS and therefore parasites are suspected in the CNS.  Diagnosis of EPM based on CSF results is still confounded by normal passive transfer of antibodies across the blood-brain barrier.  The changes to detection of SAG 2, 4/3 antibodies by the third generation test don’t identify the issues concerning non-specific testing, it can’t discern serotype, doesn’t indicate a treatment failure due to strain resistance, or point the clinician in the direction of inflammation when parasites aren’t there. It remains to be seen if the reduction in costs for time and materials will transfer to the client.

The most exciting new information is in the Barbosa paper.  They report vertical transmission in S. neurona in a sea lion, a harbor porpoise, five harbor seals, and a pygmy sperm whale. We suspected and reported S. neurona in the lung tissue of a fetus from a mare experimentally infected with S. neurona in 2004. We suggest that there is a unique window of opportunity for fetal infection, before the fetus gains cellular immunity.  The observations of Barbosa and sea mammal infections may change the opinion that S. neurona is not vertically transmitted in horses (Dubey).

The possibility that mares can transmit infections to the fetus may stimulate management changes on farms with a high incidence of EPM.  It would be a very rare condition and the veterinarian is the best source to analyze risks on a case-by-case basis.

Give us a call if you have questions or concerns about EPM .  We outline management protocols for horses as part of our consulting service.  We haven’t seen any new evidence that prods us to change our approach to the diagnosis of sarcocystosis or inflammatory mediated neuropathy.  We advise multiple exams, even in a recovered horse, once healthy let’s keep them that way!  We are committed to testing for SAG 1, 5, and 6 in independent ELISA tests, we won’t combine our three tests for convenience or price.  Confirming the presence of inflammation and distinguishing peripheral from central neuropathy are current goals.

We are committed to developing diagnostic tests and effective treatments for parasitic disease.


Sarcocystis neurona in host cells 100x

Results from experimentally infecting horses with Sarcocystis neurona supports the notion that equine protozoal myeloencephalitis is a syndrome caused by cell damage due immunological responses to protozoal infections. The disease has two phases. In the first phase, parasites turn off the horse immune responses (Witonsky 2008) and that allows the parasite to spread from the gut to other organs (Elitsur 2007). The second phase occurs when antibodies are produced to the protozoal infection and that is when clinical signs are apparent. Not all horses succumb to damaging immune responses, they don’t show signs of EPM, despite the presence of antibodies produced during infections.

Critically, this view of the pathogenesis of disease prevents a “diagnostic test for EPM” using tests to detect antibodies to protozoan parasites. Horse infections can be detected by antibody tests, however diagnosing EPM requires the develop biomarkers detecting pathologic cell processes associated with protozoal infections. We recognize that protozoal infections and EPM are not the same thing. This discussion touches three issues that are the forefront of our research: the pathogenesis of sarcocystis infections, diagnosis of EPM (a syndrome) must include inflammatory markers, and inflammation associated with protozoal infections in horses is detected by C reactive protein.

There are several pathogenic protozoa that are implicated in EPM, most commonly S. neurona and rarely Neospora hughesi, and perhaps the most successful parasite on the planet, Toxoplasma gondii. Diagnostic antibody tests have to accommodate detection of all these contenders. We assert the terminology should be “idiopathic” until a specific etiology is determined. Each of these parasites may, or may not, enter the CNS. However clinical signs are found in conjunction with inflammatory cells in neural tissues of infected horses and inflammation + parasites define the disease syndrome EPM. Inflammation as the cause of clinical signs is not a new idea, nor a hypothesis held by exclusively by us. Researchers in Germany (Olias) identified the same scenarios in pigeon sarcocystosis (proposed as a model for equine protozoal myeloencephalitis) and recently Do Carmo (2015) reports immunological responses and markers of cell responses in equine toxoplasmosis that include C reactive protein, CRP.

We take issue with the notion that parasites invade the central nervous system (CNS) to cause EPM—and we reject that documentation of these interlopers by CSF antibody is necessary for diagnosis and treatment of the clinical signs of EPM in a horse. We challenge the dogma that states “parasites that enter and remain in the CNS to cause disease”. This position assumes that disease and parasites go hand-in-hand ignoring the inflammatory part of the EPM syndrome. No doubt parasites are related to EPM. Occupation in the CNS tissue is not necessary. A marker for pathologic cell involvement is needed to address the disease EPM.

Consider this: parasites aren’t recovered from very many animals with terminal disease attributed to EPM. Animal experiments support inflammation, and not parasites, cause signs of EPM. Histological evidence of inflammation is used to make the diagnosis in the majority of cases because parasites aren’t found by any detection method. And most profoundly-- animals with long-term disease are treated with the right protocol. Inflammatory lesions in the CNS are treatable.  Reconciling these facts put more weight on inflammation in the pathogenesis of EPM. Logically, more emphasis should be placed on diagnostic tests that include inflammatory markers of cell damage over antibody detection. A panel of tests may be necessary. Releasing the grip on old dogma is necessary to design new tests to diagnose and treat EPM.

The terminology used in various publications confounds understanding EPM. Parasites in the CNS of a horse, supported by antibodies in the cerebrospinal fluid (CSF), would be more appropriately called sarcocystosis. By definition EPM requires clinical signs and those signs are related to inflammation present often in the absence of protozoa. Sarcocystosis requires the presence of parasites. Unfortunately, when the “presumptive diagnosis” of EPM was used (that included samples from horses with EPM and not sarcocystosis) to “validate EPM tests”, understanding EPM took an errant path.

Tests that detect antibody to a specific organism (ie sarcocystosis) are available. Simply calling parasite infections diagnosed by antibodies as sarcocystosis, followed by serotype, would allow researchers the ability to re-frame their view of the pathogenesis of disease in the EPM syndrome. A small caveat is that detecting active protozoa would be desirable. We have a pretty good idea how long antibodies linger in a horse that eliminated parasites. Likewise post-treatment success is measured by a reduction in serum antibodies if the animal isn't chronically exposed to parasites. There is substantial evidence that CSF antibodies found in challenged horses is transient despite continued progressive signs of EPM. That data has profound implications on the value of CSF antibodies for determining EPM.

Wendte (2010) pointed out that highly conserved parasite proteins are similar (he was discussing SAG’s 2, 3, and 4) to S. falcatula and the implications for the lack of specificity of diagnostic tests based on PCR and antibodies. The mutual exclusiveness of SnSAG1, 5, and 6 present an interesting unexplained phenomenon that requires more research. If we knew the function of SnSAG1, 5, and 6 proteins, we may more fully understand the pathogenesis of EPM. We propose these proteins function in host inflammatory pathways leading to pathology associated with sarcocystosis in turn, that leads to EPM. Parasites in the CNS aren’t required to cause signs in this scenario. Concentrating on antigens unique to pathogenic strains would allow researchers the ability to re-frame their view on the pathogenesis of disease in the EPM syndrome.

A recent paper (Do Carmo, 2015) gives us hope that others out there understand parasitic protozoal infections as we do. Toxoplasma affects many warm-blooded animals, horses included. Toxo uses horses as intermediate hosts and these infections are generally asymptomatic. Fulminant toxoplasmosis is often associated with immunosuppression. Sometimes signs are mild. Signs can indicate involvement of the central nervous system that include ataxia or sometimes excessive irritability. In South America antibodies to Toxoplasma occur in 32% of the animals. Do Carmo and co-workers suggest a hypothesis that equine immune responses against T. gondii are lasting, variable, and a contributing factor for the disease pathogenesis and cellular lesions. These researchers investigated the levels of several immunological variables and markers of cell damage in Toxoplasma-seropositive horses. They found higher levels of immunoglobulins, pro-inflammatory cytokines, and CRP when compared to seronegative horses. They found a correlation between high antibody levels and inflammatory mediators. They conclude that as a consequence of chronicity of disease, cellular lesions may lead to tissue damage with the appearance of clinical disease.

Our starting list for immunological variables for S. neurona included TNFα, IFNγ, IL1, IL4, and IL6 (Spenser 2004). We also investigated serum amyloid A (Schwab 2010) and CRP. After years of testing serum from experimental and suspect clinical cases, we have settled on the most useful serological markers for S. neurona that are: antibodies against SAG 1, 5, 6 and CRP. We currently investigate responses (down regulating the IL6 receptor) to levamisole HCl and the direct effects levamisole HCl has on Sarcocystis neurona and Toxoplasma.

The diagnosis and treatment of EPM is a changing field, effecting the predicted outcome for so many horses. The EPM dogma will have to change simply by correcting the terminology used in infections and disease. Inflammation will occupy a position that is front and center to the discussion.