Science-Based Solution: The Regulatory Problem of Trace-Level Tests of Environmental Substances

The following article is from the Spring 2024 issue of The Horsemen’s Journal:

By Kimberly Brewer, DVM; Clara Fenger, DVM, PHD, DACVIM; Andreas Lehner, MS, PHD; and Thomas Tobin, MRCVS, PHD, DABT

The Horseracing Integrity and Safety Authority (HISA) Anti-Doping and Medication Control regulations went into effect May 22, 2023, changing the face of Thoroughbred racing overnight. Heralded as the long-awaited uniformity in medication regulations and penalties for which all Thoroughbred racing stakeholders have hoped, the regulations have had disappointing results.

Their application began with a group of trace-level methamphetamine identifications at levels far below any possible physiological effect, putting a number of trainers with decades of spotless records out of business. Metformin, a medication perennially second or third on the list of the most commonly prescribed human medications in the world, was next. Recent identifications of metformin in post-race samples at trace levels have sullied the reputation of racing and called into question whether there is anything uniform about HISA.

The classification of substances irrespective of the concentration found in the horse is not new, as the Association of Racing Commissioners International (ARCI), under the guidance of the Racing Medication and Testing Consortium, has long classified substances without regard to the level at which they might be found. The difference now is that the penalties for innocent trace-level identifications rival the criminal justice system, with fines exceeding the annual incomes of the affected trainers and suspensions that guarantee their permanent exit from the racing industry.

Protecting horses from trace-level exposures seems to be an obvious answer, but that is more easily said than done. Environmental substances are all around them and long accepted by the industry to be unavoidable, as evidenced by the long-held screening limits for substances such as caffeine. The problem of environmental substances is readily solved by applying scientific evidence—employing calculations of irrelevant plasma concentrations (IPCs) where relevant, setting no-effect thresholds and/or creating screening limits that have been derived from no-effect dose determinations.

Science is easy to apply in most cases. With funding from the National Horsemen’s Benevolent and Protective Association and its affiliates, the Equine Health and Welfare Alliance (EHWA) and the North American Association of Racetrack Veterinarians (NAARV), our research group has been conducting research, reviewing available data and scientific papers and publishing in peer-reviewed journals. While adoption of IPCs by HISA remains an elusive goal, peer review is a platform through which other scientists review our conclusions and agree to their scientific validity before publication. The research provides a guide to the scientifically correct and appropriate regulation of substance identifications in racehorses.

The Matter of High-Sensitivity Testing

The first published account of drug testing in racehorses was in czarist Russia, where the racing chemist marched into the saddling paddock with a basket of frogs. Saliva was taken from a horse trained by American Jack Keene and administered to a frog, which subsequently “began behaving in a suspiciously irregular manner.” Announcements banning the use of stimulants in horses were made in quick succession in the U.S., England and France, officially starting a long history of drug testing, now in place for more than 120 years.

Drug testing in racehorses is therefore the longest established and most sensitive routine chemical testing on earth. The racehorse is the most rigorously tested animal on the planet, tested for more substances at lower levels than its human counterpart in sport. And drug testing technology has advanced exponentially in sophistication beyond the use of frogs in a basket.

Highly advanced separation and identification methods using supercomputers have made the identification of femtogram per milliliter concentrations (parts per quadrillion) a reality. Sensitivity is a good thing, as it enables detection of a wide range of substances in post-race samples. Where such sensitivity gets racing in trouble is when the substance being identified is an environmental substance that was unknowingly transferred at an irrelevant trace level to the horse.

This possibility of environmental transfer is well recognized in the arena of human drug testing, in which screening limits are established that account for passive environmental transfer. This phenomenon is well established in the scientific literature for both humans and horses, yet despite this, horse racing is well behind human drug testing in that screening limits are in place for only a few substances.

Environmental substances like metformin and methamphetamine require science-based screening limits or cutoffs in order to both protect the integrity of horse racing and move past the indiscriminate penalizing of innocent horsemen for trace-level identifications beyond their control.

Metformin: A Commonly Prescribed Human Medication

Metformin is the third most widely prescribed oral antihyperglycemic agent and a first-line medication in the treatment of human type 2 diabetes. Prescribed to approximately one in 12 U.S. adults, metformin is a classic example of a human medication with a high probability of inadvertent transfer to horses for several reasons. First, the dose to a human is large, at times as much as 2,500 milligrams (mg) per day. This daily human dose contains about 12 septillion molecules. For context, a septillion has 24 zeroes, a number approaching the number of stars in the known universe.

What is unusual about metformin is that it is excreted unmetabolized in human urine, so every human taking metformin contributes their daily 2.5-gram dose of 12 septillion metformin molecules unchanged into waterways in their environment. Metformin is also chemically stable in the environment, so what each human passes into the environment persists and also accumulates therein.

Consistent with these high-dose chemical and pharmacological characteristics of metformin, it is found at readily measurable concentrations in city wastewater, stream water and even tap water. In a U.S. Geological Survey study, Dr. Paul Bradley and his colleagues detected metformin in 57 of 59 streams tested and found metformin in 89% of the samples tested.

They described metformin in these streams as having a “pervasive presence” and occurring at “near ubiquity,” stating that “metformin is reported widely in wastewater effluent, increasingly in environmental samples and even in tap water.”

Metformin is therefore an environmental substance presenting randomly to horses, and environmental-origin blood levels in horses are virtually always below pharmacologically significant concentrations. Consequently, protecting horses from exposure to metformin is essentially impossible.

Metformin identifications under HISA rules have varied from 160 picograms per milliliter (pg/mL) to 630 pg/mL in plasma, with one at 40 nanograms (ng)/mL in urine with no plasma concentration reported, most likely not detectable. Penalties have ranged from dismissal owing to the lack of uniformity among laboratories to a 75-day suspension with a $2,500 fine to a two-year suspension and $25,000 fine for the highest level identified to date. The question then arises as to what—if any—the pharmacological and regulatory significance is of a 630 pg/mL identification of metformin in a post-race plasma sample.

The correct solution to this problem based on scientific evidence is to identify the blood level of metformin required for pharmacological effect, also known as the effective plasma concentration (EPC). Upon scientific review, the EPC is around 2.5 micrograms per milliliter or 2,500 ng/mL in humans. This EPC value can be divided by a safety factor to provide an IPC. Essentially, any fraction of the EPC is, by definition, ineffective, but just in case there is an effect of the drug, a more conservative value is chosen.

French scientist Pierre-Louis Toutain, who initially proposed this method, starts with 1/500th of the EPC. In the case of metformin, this figure is around 10 ng/mL in plasma. The highest reported regulatory trace-level identification in horses was around 4.2 ng/mL. Based on these determinations, in a recent scientific publication, we laid out the argument for 5 ng/mL in plasma as a conservative screening limit of detection for metformin in equine blood/plasma/serum samples, and our scientific peers who reviewed our conclusions concurred.

Methamphetamine: Human Recreational Substance, Classic Environmental Substance

If metformin is the human prescription medication poster child for environmental contamination, methamphetamine has become the poster child for human recreational substance environmental contamination. Methamphetamine is a potent central nervous system stimulant, producing an intense euphoria in humans and leading not surprisingly to its widespread human recreational use in the U.S. and a related illegal market. At low doses, methamphetamine can increase focus, wakefulness and general well-being, not unlike high-caffeine energy drinks, and has a legal role in the treatment of attention-deficit/hyperactivity disorder (ADHD).

The principal source of methamphetamine in the U.S. comes from south of the border in Mexico. The drug’s synthesis is relatively straightforward, and recent years have seen increased synthesis in Mexico, from where it travels into the U.S. and is marketed as street methamphetamine. This increased availability has led to a 50-fold increase in human methamphetamine overdose-related deaths in the U.S. from 1991 to 2021 and a more recent parallel increase in trace-level transfers of methamphetamine to racehorses.

An early cluster of trace-level detections of methamphetamine in racehorses occurred in Texas in 2016. These detections were at low concentrations in urine with no detectable blood concentrations, leading the Texas racing authorities to treat them as inadvertent environmental transfer events, disqualifying the horses but not penalizing the trainers. In the Texas cluster of methamphetamine positives, several of the assistant starters tested positive for methamphetamine, with one being arrested for possession and trafficking. Per the ARCI, methamphetamine identifications in racehorses peaked in 2019 at about 10 identifications and then remained under five per year until 2023, when HISA testing activated and about eight methamphetamine identifications were called along with an approximately similar number in Ohio harness racing.

A review of the analytical data associated with these 2023 HISA and Ohio harness racing calls shows that these identifications are all below 1 ng/mL in blood plasma/serum. Two questions immediately arise concerning these methamphetamine identifications—their likely source and their pharmacological and/or regulatory significance.

One likely source was described by a prominent equine regulatory veterinarian, cited in a Fox News report as saying, “I think it is probably an incidental transfer from a human substance abuser, likely through contact with the human hands to the horse’s mucous membranes.” This mechanism has high probability given that the doses of methamphetamine self-administered by substance-dependent recreational users can be many times larger than medically prescribed doses and repeated relatively frequently. As a result, the potential for inadvertent transfer from human recreational users to horses is far from insignificant.

A calculation of the IPC for methamphetamine is simple because an actual equine methamphetamine administration study has been done. Using data from that study, an IPC can be calculated—10 mg of methamphetamine placed onto a horse’s gums can give rise to 88,000 pg/mL of jugular blood methamphetamine. In theory, 1/100th of that dose or a 0.1 mg dose of methamphetamine administered in a similar manner can give rise to an 880 pg/mL concentration of jugular blood methamphetamine.

The mucosal exposure of a horse to 0.1 mg of methamphetamine is well below any dose that can produce a pharmacological effect but has the potential to give rise to an 880 pg/mL jugular blood detection of methamphetamine in equine jugular blood samples. Therefore, an IPC of 880 pg/mL could be considered fully consistent with mucosal exposure to very small and pharmacologically irrelevant amounts of methamphetamine. This level is also well below the expected therapeutic blood level for a human taking methamphetamine for ADHD, which is 30 ng/mL (30,000 pg/mL).

Aminorex and Pemoline: Synthesized Medications and Plant Substances

Our research is not limited to determining screening limits for substances transferred in trace amounts from humans to horses. Some substances can be found both in drugs that are produced synthetically and as natural components of plants. Recent studies we have conducted provide interim screening limits for both aminorex and pemoline.

Aminorex is produced as a metabolite of the plant alkaloid barbarin, which we were able to show by feeding the common contaminant of hay, Barbarea vulgaris (yellow rocket), to horses. Pemoline is closely related to aminorex and has a widespread distribution across the world, including Europe and North and South America. Studies are ongoing to determine the natural source of pemoline.

Aminorex is an amphetamine-like substance and, in the U.S., a Drug Enforcement Administration (DEA) schedule 1 controlled substance. Aminorex was at one time marketed as a prescription medication in Europe but was withdrawn when it was linked to fatal pulmonary hypertension. In horse racing, aminorex first attracted attention in Ohio in 2004 and soon thereafter in Pennsylvania and Ontario, where about 80 horses were identified as positive for the substance. In 2007, aminorex was identified as a metabolite of levamisole, an anthelmintic and immune stimulant medication. Curtailing levamisole use led to a sharp reduction in the incidence of aminorex identifications but did not eliminate them completely.

In 2018, researchers at the English LGC laboratory identified aminorex in sport horse urine samples with no history of exposure to levamisole. Review of the urine samples in question led these researchers to suggest that these aminorex identifications were associated with consumption of certain plants containing the substance barbarin, plants that were also shown to contain aminorex.

Picking up on these leads, we administered the Barbarea vulgaris, a well-known Kentucky barbarin-containing plant, to horses. These horses tested urine positive for aminorex, establishing a naturally occurring plant source for a DEA substance previously known only as a human synthesized medication. Given that the urinary concentrations identified in our administration samples were on the order of 75 ng/mL and that some field cases were less than 20 ng/mL, we proposed 30 ng/mL as an interim screening limit of detection for aminorex in equine urine.

Fast forward to December 11, 2023, when a University of Pennsylvania/HISA positive call for pemoline in plasma was at 143 pg/mL. Pemoline is closely related chemically to aminorex and is also a central nervous system stimulant. It was marketed in Europe in the 1960s and approved in 1975 for use in the U.S. for ADHD. However, it was soon linked to serious liver toxicity, and pemoline has long been withdrawn from clinical use in both Europe and the U.S.

In 2016, a number of European racing analysts communicated that, in their experience, low nanogram per milliliter urinary concentrations of pemoline are not uncommon in French, German and also South African post-race urine samples, and these jurisdictions currently regard pemoline at low concentrations in equine urine as a naturally occurring substance. It is likely that similar low-concentration pemoline occurs in North American racing, although no official communication has been issued from the laboratories. Application of the Toutain method of determining an IPC for pemoline yields a screening limit of 2 ng/mL in blood.

Substances in the Water Supply

The environmental science literature clearly shows that many human therapeutic medications are so common in the environment as to be present in measurable concentrations in the water supply. This includes surface water, tap water, well water and essentially water everywhere we look.

In our research, we are continuing to investigate the substances that are present in water as a potential risk to our horses, including tramadol, gabapentin and fentanyl. We have determined that an application of the Toutain method to tramadol yields a urinary threshold of 50 ng/mL of its metabolite O desmethyltramadol.

Similar application to gabapentin yields an IPC of 8 ng/mL and for fentanyl an IPC of 50 pg/mL.

Conclusion

Our research group is committed to applying science to the problem of environmental substances and submitting our work to be published in the scientific literature, showing not only that this work can be done but that it can be done openly for the scrutiny of the scientific community.

Environmental exposures that give rise to trace-level identifications either by consumption of naturally occurring substances in hay or inadvertent transfer from humans are generating a blemish on horse racing.

The only solution to this problem is determining scientifically valid screening limits that protect the integrity of racing. By drawing these lines high enough, we can eliminate most, if not all, environmental transfer events while also ruling out any possibility of a pharmacological effect.

This goal is readily achievable as demonstrated by the research outlined in this article, initiated by the National HBPA and its affiliates, the EHWA and NAARV. There is a path forward, and it is to follow the science.