Regulating Metformin: More Questions Than Answers

Since the launch of the Horseracing Integrity and Safety Act's (HISA) drug testing program over a year ago, metformin has sucked up an awful lot of oxygen.

Under HISA, there have been seven resolved cases and four pending cases involving metformin—perhaps not an especially large number on its own, until framed against the number of metformin cases adjudicated before the program went into effect. An Association of Racing Commissioners International's (ARCI) review of their database “going back decades” found five metformin cases for Thoroughbreds. All were from 2017 onwards.

The resolved metformin cases under HISA include a two-year suspension, an 18-month suspension and a 75-day suspension. Four cases have been resolved with no sanctions to the trainer, though in one case, that was because the B sample had insufficient volume for analysis. Another trainer passed away before a formal hearing.

Until recently, if a metformin-related B sample returns positive–or is waived altogether–the Horseracing Integrity and Welfare Unit (HIWU) had the ability to impose an interim suspension until the case was formally heard, with the potential to drastically shake up someone's life and livelihood.

Earlier this month, HISA announced that it was deferring all interim suspensions involving metformin until the Racing Medication and Testing Consortium's (RMTC) Scientific Advisory Committee had conducted a review of the available science relating to metformin. The RMTC's review is expected to take months.

The thing is, the RMTC will find scant scientific information available regarding the substance's effect and relevant pharmacokinetics in horses. Nor is it easy to extrapolate the findings from human research to horses, or that from other animals. Indeed, several respected independent experts in the drug testing field declined when asked to comment for this story because of this information desert.

“All of the questions you've posed are certainly worthy, although current research may not provide answers to them,” wrote one. “Based on the limited information about this drug's behavior in horses, I am not sure I would be of much help,” wrote another.

Until more concrete evidence comes to light, therefore, this drug appears to sit in the crosshairs of a more philosophical debate.

Given racing's tenuous public standing, is a hardline stance on drugs a necessity for the survival of the sport—even if that approach sometimes results in collateral damage? Or should every effort be made to give the sport's participants discretion in the face of doubt—even if sometimes, a cheat slips through the cracks?

Pre-HISA

The ARCI classes the drug a category 2 penalty B substance, which comes with a 15-day suspension (or 60-days with aggravating circumstances) for a first-time offense. According to the ARCI, of the sanctions meted out for the metformin cases pre-HISA, the most severe was a 15-day suspension (and $2,000 fine) handed to Wesley Ward last year, along with a separate $5,000 fine (and 5-day suspension) handed to Ward in 2022.

HISA, on the other hand, makes it a banned substance that comes with a maximum two-year suspension and $25,000 fine for a first offense–sanctions meted out to trainer Jonathon Wong.

“There is no generally accepted use for metformin in the racehorse, and one of the [Anti-Doping and Medication Control] Program's goals is to protect equine welfare,” a HIWU spokesperson wrote.

The agency claims, too, that it has received whistleblower information that the drug is being administered purposefully to racehorses to enhance performance. When asked for specifics about these whistleblower complaints—including whether any of the accusations have been substantiated—a spokesperson responded that it “cannot comment on ongoing investigations.”

While HIWU officials have stressed that regulators are often looking at more than just the detected levels of a substance when adjudicating cases, this still leads to another wrinkle in the debate: What are permissible levels of this drug?

Using the information in this recent paper published in Pferdeheilkunde – Equine Medicine, the blood metformin level detected in one of Ward's two cases was 4.2 nanograms per milliliter (ng/ml). The paper includes no urine levels for that case.

With just the information provided in the same paper, the highest blood level detected under HISA is at 630 picograms per milliliter (pg/ml), found in Wong's case. One nanogram is 1000-times larger than one picogram.

As for the urine in Wong's case, a metformin level of 242.5 ng/ml was reported. In Ward's other metformin case, the horse tested positive with urine metformin levels of 577 ng/ml. The paper includes no blood levels for that case.

Trainer Guadalupe Munoz Elizondo's case was dismissed after HIWU established harmonized testing specifications for metformin, as was the metformin case against trainer Javier Morzan. Levels of metformin at 162 pg/ml had been detected in the blood of Munoz trainee, Quinton's Corner, before the case was dropped.

After a review of the current literature on metformin (which is primarily on humans), the same paper recommends an interim Screening Limit of Detection for metformin of 5 ng/ml in the blood, calling that amount an “irrelevant trace level.” That level would clear Wong. The methodology used to determine that level was for therapeutic use as opposed to prohibited substances.

When asked for her thoughts on this recommended threshold, former RMTC executive director and COO and current 1ST Racing chief veterinary officer, Dionne Benson, said that she didn't think “setting the levels just above the concentrations that have been found is the right approach.” She advocated, however, for more research into the matter.

“I think that's something worth investigating,” said Benson. “With the information we have right now, it's easy to say we don't know. And I think that should go into the consideration that HISA is taking into these suspensions. But a year from now, we might find out that you have to administer it to get these concentrations, and I would be wrong.”

Pathways of Exposure

Metformin is primarily prescribed in humans to help those who suffer type 2 diabetes more readily use the insulin produced in their bodies in order to better regulate blood glucose levels.

Though the drug has been available as a diabetes treatment for decades, researchers are still figuring out its other benefits, including as a possible defense against strokes, dementia and other age-related diseases. “We're still learning about all the different things that metformin does to the body,” explained Joseph Bertino, a human clinical pharmacologist.

With no discernable use in racehorses, metformin's primary use for horses, explained Benson, is to help ponies and older horses suffering insulin resistance, who are susceptible to foundering.

One key difference between humans and horses is that the drug is excreted much quicker and in larger quantities in horses than in humans. In other words, humans readily absorb a significantly greater amount of the drug than horses do. It's also a chemically stable drug excreted by humans largely unchanged into the environment.

Among the various experts consulted for the story, all said the likeliest route of possible environmental contamination is through a horse eating hay or straw soaked with the urine of someone who takes the drug.

HISA and HIWU have made public efforts to warn horsemen and women how best to protect themselves from inadvertent cross-contamination, including maintaining a strict policy against urination in the stables.

At the same time, stakeholders argue that's sometimes easier said than done given the poor and sometimes sparse bathroom amenities at certain backstretches and the fast pace of morning training.

A key detail in this particular argument is just how much straw a racehorse would need to consume to trigger the sorts of levels that have prompted regulatory action under HISA. “With the low bioavailability in horses, that's going to be challenging, especially with a human dose,” said Benson.

“I would love to do a calculation to figure out exactly how much straw they would have to actually eat that is covered in urine in order to actually create a dose that would significantly show up an hour after,” Benson added.

Another commonly-floated pathway of exposure is the transference of the drug from human to horse–a metformin-using groom, for example, whose hands are coated with the drug.

In a February TDN letter to the editor, Bertino poured cold water on this argument, arguing that the polymer coating around many of these typically long-acting pills makes such an occurrence unlikely. Are all metformin pills film coated in this way?

“I could not find metformin that was not film coated, and it's no surprise because generic companies have to be bioequivalent to an approved metformin tablet in order for the FDA to approve their tablets for use and sale,” Bertino wrote on an email. “There are immediate release metformin tablets but from what I can tell, they are also film coated to mask the bad taste of the drug.”

Bertino said that Latin American countries may manufacture non-film coated drugs. “But I can tell you that in many countries throughout the world, the metformin is film coated,” he added.

Robert Holland is an expert in respiratory and infectious diseases and is one of the authors on the paper recommending an interim screening limit of detection for metformin. He also sits on HISA's lasix advisory committee. In highlighting a general hygiene component to the issue, Holland explained how some tracks are equipped with portable toilets with no ready hand-washing facilities.

“You can see where even when you're thinking you're okay, you may not be,” Holland said, and pointed to a study from a few decades ago that found trace levels of urea on the handles of scoops used for children's confectionary.

Given how prodigiously metformin is prescribed, it is a widely found, near ubiquitous presence in rivers, lakes, and wastewater effluent across the country. It's even found in tap water. This includes the racetrack where two of the violations were found, according to the HBPA.

When asked about this phenomenon, Benson said how, given all horses stabled at a track drink from the same source, “I would think that if you had that as a true source of contamination causing positives, we would have so many it would overwhelm the system.”

Holland, however, suggested the answer is more complicated thanks to variables like the levels of metformin in the water over a period of time and the amount of water consumed by individual horses.

“Could that be constantly going at a low level and then trip a test? I don't know the answer to that,” he said. “I don't have dose.”

A few years ago, researchers conducted tests of Charles Town's ship-in stalls, to determine any presence in the environment of common substances. Metformin was detected in 2 of the 21 stalls swabbed, along with a host of other substances commonly detected in racehorse samples.

Which leads to a common gripe among horsemen and women: That receiving barns across the country are often contaminated with the remains of prior horses, and that they are in the sort of poor physical shape that makes thoroughly disinfecting them difficult.

The British Horseracing Authority (BHA) has four pages of strict requirements for the cleanliness of racetrack stables, including the removal of all bedding, “steam cleaning or pressure hosing all surfaces,” and chemical disinfectant before a new horse can enter the stable. Tracks in the UK face possible sanctions for not adhering to the rules.

HISA's regulations on the same topic are scant in comparison, with HISA's safety officers given a broad set of vaguely-written tasks, which includes requiring them to “monitor” common areas such as the receiving barns for cleanliness.

When asked, a HISA spokesperson wrote that no tracks have been officially sanctioned for receiving barn violations. However, “should the accreditation team identify deficiencies, the racetrack's accreditation could be impacted,” the spokesperson added.

HIWU is silent on how many metformin positives are from ship-in runners, writing that “HIWU does not track this data.” Only one published decision discloses this info—for Wong, whose horse had shipped into Horseshoe Indianapolis.

All of which leads to another possible feature of the environmental contamination argument: The accumulative effects from multiple exposures. In other words, could the combined exposure to metformin-using grooms who haven't washed their hands, to dirty ship-in stalls and to drinking water laced with metformin all combined result in a positive test?

“To me, if you look at these multiple stream arguments, most of those, with the exception of drinking water, you can control,” said Benson.

She added that metformin is one of many commonly prescribed drugs that are found in U.S. waterways, and is one of several common substances detected in receiving barns. She suggested that in using this argument, it would be fair to expect an array of common substances appearing more ubiquitously in tests.

Holland was more cautious when asked about this possibility. He said that metformin behaved differently to many other substances commonly found in the environment, and added how intentional illicit administration of a drug would often result in a rash of positives within the same barn.

“People recognize that we come into contact with different sources, and sometimes we just don't know all the sources,” said Holland. “We think that [people urinating in the stalls] is the likely source, but we just don't know. And at the same time, if you read our peer review paper, at those levels [which prompt regulatory action], it wouldn't have a biologic effect on the horse at all.”

Performance Enhancer?

The research into the possible performance enhancing effects of the drug is centered around metformin's impacts on humans. This recent blog post–shared on social media and forwarded to the TDN by a HISA executive during the reporting of this story–lists some of the key studies showing how it might impact human athletic performance.

In summarizing the study findings, the author makes the following claim: “Metformin may enhance exercise performance during short-high-intensity exercise, similar to what thoroughbred racing demands. However, metformin may also inhibit fitness gains if taken continually in the long term.”

   (Note: The TDN reached out to the author of the blog multiple times for a name and their credentials but received no response).

Does the scientific literature, however, support such a claim? HIWU declined to “speak to metformin as a performance-enhancing substance in racehorses.”

The authors of the paper advocating for a metformin screening limit of detection are more circumspect. “The scientific literature is mixed on whether Metformin has an effect on athletic performance on any species at any plasma concentration,” they write, before adding:

“A meta-analysis evaluating this group of scientific papers concludes that, overall, these studies failed to show any effect of Metformin on indices of athletic performance, other than an increased rating of perceived exertion.”

When asked the same question, Benson was similarly cautious. “We have a lot of mixed and unknown information,” she said, before raising the idea that trying to confirm one way or the other metformin's possible performance enhancing properties might be the wrong question to ask.

“We almost look at every drug the wrong way, at least on the regulatory side. It is difficult if not impossible to prove effect. What I mean is, I don't know that we can show even if a horse won by 30 lengths–so that's what, six seconds?–that there was a statistically significant effect of a drug,” said Benson.

“The problem we have is that horse races are won or lost by inches, not 30 lengths, and I don't know that I would be able to ever definitely prove what improved a horse a length or two,” she said.

Lab Variability

Towards the end of last year, metformin proved central to a key obstacle as HISA works towards drug testing uniformity—the issue of laboratory variability.

Internal reviews of the (at the time) six HIWU-contracted laboratories uncovered different limits of detection in blood for metformin. Indeed, all metformin positives up to that point originated from just the one lab.

The laboratory-related issues that HISA has faced in properly policing the drug were grandfathered in. Pre-HISA, every lab had different drug-testing sensitives, and a lot of discretion was given by the state regulators in how they approached enforcement, said Benson.

“There was an inherent inconsistency because the labs had different levels and the commissions potentially had the option to decide whether they wanted to pursue a finding,” said Benson.

At the end of last year, HIWU announced that it had harmonized “testing sensitivity” across its contracted labs, to ensure they were screening for metformin at the same levels in blood and urine.

A HIWU spokesperson wrote that since then, “more than one laboratory has reported the presence of metformin in an A Sample,” though added they could not disclose which ones, other than Industrial Laboratories in Colorado.

In the same vein, have any metformin positives been kept from public disclosure or dismissed before any regulatory action was taken? “HIWU cannot comment on cases or findings that have not been publicly disclosed in accordance with the ADMC Program rules,” the agency spokesperson wrote.

The Federation Equestre Internationale (FEI) lists metformin a “Prohibited Substance-Controlled Medication,” for use in the treatment of equine metabolic syndrome. There are no withdrawal guidelines, however. The FEI did not respond to multiple attempts for an interview on this story.

International efforts are underway to “harmonize” a regulatory approach to the drug, according to a renowned animal sports medication and doping control expert, who asked to remain anonymous.

“Obviously, the aim within Europe and North America is to achieve a harmonized approach at both a national and international level. Metformin, however, presents several challenges. Neither screening limits or international thresholds in the strictest sense of their definitions are appropriate applications and so the likes of harmonized, scientifically based reporting limits come more into consideration,” the expert wrote in an email.

All of which leaves this final question: Is there another way of testing for metformin to determine whether it was administered intentionally to a racehorse at a high dose? According to Holland, hair testing offers that possibility.

“If we are still learning about potential sources and if we are unsure, hair testing would be an avenue that would help us,” said Holland. “If you're seeing a horse that's getting a lot of a drug, you'll see it in the hair.”