Skip to content

How many horses die from EPM? It’s a complicated answer.

A complicating factor is that horses can have neuroinflammation that is not due to EPM. A horse that has neuroinflammation is wobbly.  They have an unexplained lameness, and…well, looks just like a horse with EPM! And that is because EPM is a sub-set of these horses—they are EPM horses when we know S. neurona is the cause of the neuroinflammation. If we don’t know the cause then they are idiopathic, we call them idiopathic encephalomyelitis or IE.

How many horses have signs that look like EPM but are unrelated to S. neurona infections? We looked at sera from 4298 horses that had signs of EPM that were not treated with antiprotozoal drugs and found that 55% were seronegative to S. neurona. It’s important to point out that antiprotozoal drugs can alter the detection of S. neurona antibodies (Martin Furr published data to show that treating horses with Marquis delayed antibody production without altering the development of disease).

Exposure to antiprotozoal drugs is probably not the reason that the majority of the sera we tested were seronegative. We are aware that 20% of seronegative EPM suspect horses will show a rise in antibody titer shortly after treatment. Seroconversion takes place about 17 days after infection, faster in an “experienced” (has been infected before) animal and a bit slower in a “naive” (never had an S. neurona infection) animal. That is why it is important to test for serum antibodies against S. neurona 10 days after the end of treatment. Seroconversion supports the diagnosis of EPM.

Generally, 55% of animals being treated for EPM by veterinarians (suspect EPM due to ataxia and not tested) would be treating the wrong disease. If the treatment was only an anti-protozoal then the disease isn’t being treated. Could another serum test help identify the horses on the wrong treatment path?

A retrospective analysis of data obtained from serum C-reactive protein (CRP) concentration (from 1424 sera from animals with neuroinflammation tested for  before any treatment was initiated) was divided into groups: those with S. neurona antibodies (supporting the diagnosis of EPM) and those that were seronegative (supporting a diagnosis of idiopathic encephalomyelitis). The purpose of the analysis was to examine serum CRP levels from horses that died, presumably due to EPM, perhaps identifying CRP as predictor of survivability for EPM. Sixty seven horses died.  That means 5% of horses that were tested for EPM died before treatment.

The group of horses that succumbed to disease in this analysis died for various reasons and not all were treated (or treated with an anti-protozoal) therefore an analysis by treatment was not possible.

The results of the query indicated that 35% of the dead horses had a high serum CRP concentration (before treatment) in the EPM group while 61% of the horses that died had a high serum CRP concentration (before treatment) in the IE group.

Serum CRP is an indicator of non-specific inflammation due to bacteria, virus, and protozoa. The CRP value falls when the inciting cause of inflammation resolves with proper treatment. These data may indicate that a significant number of horses with suspect EPM actually have another etiology. Anti-protozoal agents are highly specific for pathogenic protozoa. The change in CRP, before and after treatment, may be useful in determining resolution of the infection. When the CRP remains elevated it’s time to look alternate causes of ataxia.

The original question was “How many horses die from EPM”? Retrospective analysis of our database indicated death in suspect EPM (seropositive for S. neurona) horses is 1.3% of cases. These horses have a history of multiple antiprotozoal treatments with diminishing returns.

How about idiopathic encephalomyelitis? We have the data to answer that question. Death in horses with IE (seronegative for S. neurona) that had a history of treatment with an antiprotozoal agent is 25%.  The death rate may be high because alternate causes of disease are not considered.

A definitive diagnosis for EPM remains elusive and a response to treatment is a logical approach. However, when there is no response to treatment the diagnosis is idiopathic encephalomyelitis and working up the case is important. C-reactive protein may be a parameter to monitor response to treatment when the diagnosis remains open.

Sarcocystis neurona and Toxoplasma gondii  are cousins.   Toxo is a parasitic protozoa  that causes life threatening conditions which includes encephalitis in immune-compromised individuals. Toxo is an extremely successful parasite, said to infect 25% of the worlds population.  Toxo is successful because it is known to manipulate the immune system of the host.  Sarcocystis  infections, like Infection by Toxoplasma gondii, down-regulates the host innate immune responses, such as proinflammatory cytokine production.

A lot is known about Toxoplasmosis.  The host initiates a broad range of immune responses upon T. gondii infection, and much is known about each molecule involved in the process. For example, when infected by T. gondii, cells belonging to the innate immunity system, such as macrophages and dendritic cells, produce a set of proinflammatory cytokines, including IL6, to induce the adaptive immune response or to directly eliminate the parasites.  The cytokine IL6 is an important molecule that can direct the immune system down alternate paths depending on the cellular  “environment”.

T. gondii is generally divided into three predominant clonal lineages (types I, II, and III) plus a few exotic strains.  Virulence is associated with the clonal lineage, type I is most virulent and type II is involved in congenital infections.  Similarly, Sarcocystis SAG 1 and 5 are predominant in the types that infect diseased animals.  Virulence is also linked to the ability of the parasite to manipulate the hosts responses right down to the type of cell that is infected.

Researchers injected human nerve cells with the three most common toxoplasma strains, each of which caused a different pattern of gene expression. Gene expression occurs when a gene is switched on to release a substance that tells cells what to do or not do, determining the cells' biologic behavior. Gene expression can be turned on and off, stepped up or down by various factors, including viral protozoal, and bacterial invasions.

Cells infected with toxoplasma type I -- the most virulent strain in mice -- had the greatest impact on gene expression, altering more than 1,000 genes, 59 of them linked to brain development, central nervous system function, and nerve impulse and signaling.

Cells injected with the less virulent types II and III had low and moderate levels of gene expression in nerve cells. Infection with toxoplasma type II affected 78 genes, while type III altered 344 genes, the majority of the expression was on the metabolism of the cells.  Sarcocystismay have conserved the ability to alter host gene expression a trait inherited  from its distant, common relative.  It is perhaps the manipulation of cytokine expression that produce clinical signs of EPM.  Disease may reflect how the horse responds to this manipulation.

In Toxo, specific activation of a cytokine is induced upon type I parasite infection, and this process was previously shown to be defective in type II parasites. Interestingly, there are molecular mechanisms which lead to suppression of innate immune responses, and is the basis for the strain differences between type I (or III) and II. The type II strain fails to suppress immune responses (including IL6 production) due to a potential defect in a parasite-derived enzyme. Using genetic engineering of this parasite enzyme, replacing a single amino acid, culminated in enhanced production of IL6 (and other cytokines) in the infected cells.  Virulence can be linked to one amino acid in a single enzyme.

The virulence factors involved in S. neurona infections are unknown.  However, one could speculate that virulence factors relate to biologic differences between of Sarcocystis (host specificity--neurona in mice, falcatula in birds) by the enzymes that target the hosts immune cells.  The enzymes that are adapted to mice (and horses) may not work in birds. A way to study this idea is to measure the host cell responses in infected and ataxic horses.

Due to the pivotal role of IL6 in innate responses to pathogenic protozoa it may be productive to measure IL6 in ataxic horses.  The concentration of IL6 in serum of horses was detected in horses that were experimentally challenged with S. neurona.  We found that as long as 42 days after experimental challenge 1/2 of the horses have measurable IL6 in the serum.  In field cases, we found that IL6 is quite high in some very sick animals with idiopathic encephalomyelitis.

We didn’t find IL6 in the serum of normal horses or horses before challenge infections (0 picograms/ml of serum) but found elevated IL6 in the serum of very sick horses (10.4 picograms/ml).  We know of no published values for serum IL6 in horses with EPM.  For a short time, we will run any serum sample from an ataxic horse that is submitted with a complete post mortem exam, at no charge. The normal charge for IL6 is $35.00.

The IL6 ELISA test is interpreted in conjunction with the SAG 1, 5, 6 ELISA, C-reactive protein, and clinical disease.  The acute phase proteins (CRP) are intricately associated and controlled by IL6.  While IL6 won’t be a diagnostic test that definitively identifies the horse with EPM, it may be part of a panel of tests that can.  We would like to correlate strain and virulence as they have in S. neurona’s relatives.  Based on work done with Toxo, finding IL6 in the serum of S. neurona infected horses may be a place to start.

Further reading: J. Exp. Med. 2009. 206(12); 2747-2760

The hunt for the Higgs boson exemplifies our work with EPM. The Higgs is the last undiscovered particle predicted by the Standard Model (of particle physics). The long sought Higgs has no spin, no electric charge, no color change, and is infinitely short lived. The search for the Higgs boson involves proton-proton collisions resulting in decay in a ridiculously short time, to predictably many, less energetic particles.

The data processers figured out the kind of decay products, on average, that should be there if Higgs didn’t exist. The task is to compare their theoretical expectation to the actual data, searching for that unpredicted “bump” that might indicate the decay of Higgs bosons. It’s important that they gather enough data to ensure the observation isn’t a statistical fluctuation. Before declaring victory, they further analyze the data and determine if the characteristics are consistent with Higgs or perhaps some new unpredicted physics.

To find a Higgs that doesn’t exist long enough to observe directly as a particle they look for a “signature” in the decay products.  Not unlike what we are doing with EPM.

For many years the insistence that parasites are in the CNS making EPM unrewarding to treat, that antibodies to various parts of Sarcocystis have significance if they are not-specific, and inflammation was unworthy of investigation led researchers to find data that fit their theoretical expectations. However, the disease caused by S. neurona infections in horses has a molecular signature and we are defining that signature. We too have theories and expectations and we are careful to analyze our data to determine if the characteristics evaluated indicate we are on the right track. We discovered something unpredicted but long recognized by many field veterinarians. We discovered that undiagnosed inflammation is a larger issue in horses that can cloud the general view of EPM.

To parse the difference between diseases we have a specific protocol. We expect specific results and then gather the data to ensure the observations are statistically relevant. To some, our interpretations and advice are confusing and frustrating and that is because you look at one animal. We look at data from hundreds of similarly afflicted animals. We will explain our take on the process for each and every animal, and that is why our phone is always busy. Shoot us an email, we will answer you within 24 hours.

The bottom line is that we are evaluating at least two different conditions that look the same. We evaluate horses with clinical signs of disease that 1-have antibodies to SAG 1, 5, and 6 and 2-those that don’t have antibodies. The terms we use are EPM (reserved for horses that have or had antibody, prior to treatment) and IE (idiopathic encephalomyelitis—encephalomyelitis with an unknown etiology-but might have been EPM). Recent treatment with anti-protozoals can affect immune responses delaying antibody production—it depends on past history of the horse with the organism. Treatment responses can help us evaluate the horse—so please, give us your feedback. The more data we can analyze, the more “bumps” we can identify and investigate.

It is important to you because horses that are not responding to treatment probably don’t have EPM. The earlier you can recognize that the diagnosis is open (IE) the better the outcome of future treatment. We expect that EPM horses that are treated with licensed drugs will recover, on average 63% improve and some will need some additional therapy to improve more or to address relapses (relapse is ineffectively treated inflammation). Horses that get additional therapy may recover to the owner’s satisfaction and get back to use. Horses with IE treated with our approach will respond to treatment most of the time, additional treatment and monitoring will improve our record. There are no published treatments for IE, you need to call us for our approach to these horses. When the CRP is high and remains high after our treatment, we have several suggestions that are based on the treatment response and the underlying condition. The underlying condition has not been addressed if the CRP remains high.

The elusive Higgs particle can’t escape identification—and neither can the mystery surrounding EPM. As someone once penned, “EPM is a mystery wrapped in enigma”—but we think it isn’t for much longer.

Pathogenes thanks Hal Hollis, consulting physics prof and mystic for the input.