By all accounts Graham Motion is one of the good guys in racing. He has often been cited as the example of how to be successful without violating medication or drug standards. Do it like Graham Motion does it, and we’ve cleaned up racing.

Graham Motion has been training for over 20 years, and during that time he’s had over 11,000 starters and until 2015 exactly zero drug or medication positives. That changed when Kitten’s Point, winner of the G3 Bewitch Stakes at Keeneland on April 24, 2015, tested positive for the therapeutic medication methocarbamol (sold under the brand name Robaxin).

A small overage for a common therapeutic is not news, but this was different. The first violation in an exemplary career was enough to make headlines, but more significant was Motion’s response to the positive. In an open letter published in the Thoroughbred Daily News, Motion said,

“I always felt that if the day ever came where, by some unforeseen circumstance, I was charged with a drug violation I would not lawyer up to defend myself, but rather would take my punishment and move on.”

But when the test for Kitten’s Point came back positive, Motion did decide to fight. He fully believed that he and his team had done everything they needed to do to comply with the methocarbamol threshold by withdrawing the drug a full seven days before the race, or five days longer than the published recommended withdrawal time.

Motion suspected something wasn’t right, either with the threshold or the withdrawal time, and set out to prove it.

The Methocarbamol Threshold

Horses are athletes, and like any athletes the stress of training and competing inevitably leads to everything from small aches and pains to injury. Horses are also individuals, with predispositions to everything from bleeding in the lungs to muscle cramps to heartburn. When a horse is ailing, it is the responsibility of the trainer to ensure his charges are given the right medical treatment.

In many cases horses with minor issues can continue racing through treatment with therapeutic medications. Racing jurisdictions are clear to show which therapeutics are allowable, and they are equally clear about how much of a respective therapeutic a horse is allowed to have in its system after a race.

Methocarbamol, sold under the brand name Robaxin, is commonly used in the prevention and treatment of painful muscle cramping, sometimes called “tying-up.” Horses predisposed to this condition will tie-up even during training. Like many medications, in order to be effective horses have to be given regular doses – in the case of methocarbamol once or twice daily to keep the cramping from disrupting the training schedule.

There are various methods by which a horse can be given methocarbamol: intravenous injection, tablets, or an oral paste, with tablets being by far the preferred method of administration during training. The general guideline for dosage is 15mg/kg, meaning for every kilogram the horse weighs it should be given 15 milligrams of the drug. In a 1,000 pound horse (454 kg) the dose would be 6.8 grams of the drug.

The Racing Medication and Testing Consortium (RMTC) publishes a schedule, known as the Controlled Therapeutic Medication Schedule (CTMS) that provides recommended dosing and withdrawal times for the approved therapeutic medications for race horses, as well as the post-race threshold. The current methocarbamol threshold was developed from a study done in 2013 by Marc Rumpler, et al.  and published in the Journal of Veterinary Pharmacology and Therapeutics in June 2013. Based on Rumpler’s work, in May 2013 the RMTC listed the methocarbamol threshold  as 1 nanogram/milliliter (ng/ml) through blood testing. It also listed the recommended withdrawal time as 48 hours, listed the administration method as either IV or oral, and suggested an IV dose of 15 mg/kg and an oral dose of 5 grams.

The RMTC has revised the CMTS for methocarbamol twice since the May 2013 publication. In December 2014 a footnote was added suggesting that when methocarbamol was administered with phenylbutazone there was a potential for reaction and the withdrawal time should be increased to 72 hours. Then, in February 2016, the dosing protocol was changed to IV only, with a one-time only dose of 15mg/ml. They also added a footnote saying an oral dose may be utilized but longer withdrawal time may be required to fall below the threshold. Trainers using methocarbamol orally for multiple days were encouraged to have the horse tested prior to entry. These modifications to the CTMS were made with no announcement to jurisdictions that had adopted the thresholds, no fanfare, no press release to warn unsuspecting horsemen; they were just quietly changed on the RMTC website.

The RMTC makes it clear that although withdrawal times and dosages are part of the therapeutic medication table, for compliance purposes the only number that is relevant is the residual threshold of 1 ng/ml. The withdrawal time and the dosing amounts and protocols are simply recommended guidelines. In other words, if a trainer gives his horse a 5-gram dose of Methocarbamol 48 hours before a race and the horse tests at 2.0 ng/ml, it is still a violation. As Graham Motion found out, even if you prove that you followed the dosing and withdrawal guidelines, it is no defense against a positive test.

While it may seem there is some unfairness in the way the residual thresholds and recommendations for dosing and withdrawal interplay, the RMTC, whether overtly or inferentially, was selling the idea that any trainer giving their horse the recommended dosage and allowing the recommended withdrawal time should not have a horse test positive.

Given that Motion had more than tripled the withdrawal time for Kitten’s Point, something didn’t seem right. The place to start was with the so-called Rumpler study.

We can start by explaining that in 2007 the well-known testing pharmacologists, Rick Sams (HFL Sport Science Laboratory in Lexington, KY) and Scott Stanley (University of California School of Veterinary Medicine in Davis, CA) designed a study to “determine the pharmacokinetics (disposition and clearance) of methocarbamol and to estimate the withdrawal time of methocarbamol after single intravenous administration of methocarbamol at a clinically relevant dose to athletically conditioned thoroughbred horses.”

Whether or not the study would be useful for standard setting would not be clear until after the results were published.

The research results were actually published in the Journal of Veterinary Pharmacology and Therapeutics (June 2013) with Marc Rumpler as the principal author.  Rumpler was part of the team headed by Sams and Stanley, although his actual listed role during the study was to do method validation studies and sample analyses. (Note: The original data gathering work was funded by both the RMTC and the Florida Department of Pari-Mutuel Wagering.)

Given that the RMTC was expecting to use the results from Rumpler’s paper in support of a methocarbamol threshold,  Rumpler slightly revised the purpose that Sams and Stanley had offered, stating that it was being published “to develop and validate a method for determining methocarbamol in horses and to investigate its disposition after intravenous and oral doses to exercise conditioned thoroughbred horses for the purpose of generating data that could be used to establish a regulatory threshold for use in horse racing.” 

Sams was careful to call it a pharmacokinetic study to determine how the drug was distributed and cleared in the horse, not overtly guaranteeing it was going to be conclusive enough to form the basis of a blood threshold. But Rumpler apparently believed there would be no problem adapting the data to setting a threshold.

Rumpler’s study had a number of less than ideal design aspects, starting with the test animals. The 2007 data collection phase used twenty thoroughbreds (9 mares and 11 geldings) ranging in age from 5 to 10 (with 15 of the horses actually between 7 and 10), and weighing between 468 and 605 kilograms (1031 to 1333 pounds), with the average weight being 531 kg. Compared to the typical population found at the racetrack, the study group was substantially older and heavier. At the race track most runners are under 5, and likely to be 400-500 kg, with very low percentages of body fat.

While the horses were exercised three times a week on a treadmill, that could not have simulated the exercise levels of racing stock – obvious when considering the weights of the animals. They also were not being fed a racehorse diet. Rumpler noted that, “horses were housed in grass paddocks at the University of Florida Veterinary Medical Center (in Gainesville, FL), maintained on a diet of commercially available grain mixture, and had open access to water and hay at all times.” As a researcher you take what is available, but clearly the study was not being done on the sort of active racehorses that would be tested after a race.

All of those characteristics could lead to very different results than if active racing stock and standard dosing were used. All 20 of the horses were given a single, 15 mg/kg intravenous dose of methocarbamol and then tested at various time intervals. Out of the 20 test subjects, 14 of the horses only received this single IV dose.  Why the investigators chose the IV method is not clear from the Rumpler paper, but it seems to have been at odds with standard veterinary practice if the dosing was meant to simulate how the drug is commonly used during training. On the other hand, as a means of assessing the pharmacokinetic component of the standard setting process, perhaps the IV administration was not that problematic.

After a ten-week washout period, six of the horses were also given five 5-gram doses of methocarbamol at 0, 12, 24, 36, and 48 hours. The Rumpler paper called this 5-gram dose “oral,” but it was actually 10 – 500 mg pills that were crushed, mixed with 60 ml of water and delivered to the stomach via a nasogastric tube, a subtle but significant difference from normal oral administration for three reasons. First, veterinarians do not administer the dose this way. They will crush pills and dissolve them in water, but then inject the mixture directly into the oral cavity; second, since many drugs are absorbed across the oral mucous membranes, use of the nasogastric tube does not mimic how the drug might be absorbed during normal dosing; third, even though the nasogastric tube was flushed with 60 mL (two ounces) of water, it was still possible that some of the dose remained in the tube and never made it to the stomach. In other words, there was a chance the dose was imprecise. In any case, it was interesting to substitute nasogastric administration for normal oral administration in order to ensure the horse got a full dose when there was no assurance of that at all.

Beyond that, the horses from the study group that received oral dosages of 5-grams were effectively given a lower dose than the 15 mg/kg a horse in training would get because they all weighed well above the weight for which 5-grams would be the proper dose. For example, the horse weighing 560 kg that received the 5-gram oral dose would have received an 8.4-gram dose if the dosing recommendations were followed. This means the study group could have lower residuals than would be expected for active race horses, since they were being under-dosed.  And given the discrepancy in body fat between the study group and active racehorses, the pharmacokinetics (disposition and clearance) could have also been affected.

The Rumpler analysis was flawed in other ways as well. Typically, thoroughbreds that are predisposed to “tying-up” (muscle cramping) will do so daily during training periods. This indicates that the medication would logically be used daily to prevent cramping in those horses.  While the orally dosed group was dosed over a 48-hour period, this hardly represents the kind of continuous dosing a horse in training might receive. The single IV dosing regimen used in the study certainly didn’t simulate how a veterinarian would normally use methocarbamol during the training period, and even if the oral dosing came closer, the dose of 5-grams was below the commonly used recommended therapeutic dose for all six horses treated that way.

Rumpler’s paper doesn’t effectively document why these choices were made, nor does he discuss the pitfalls of using a small sample size of older, heavier, minimally exercised thoroughbreds. He provides no discussion of the potential comparisons to actively-racing horses. The most obvious question: why the single intravenous dose at the recommended 15 mg/kg, but a seemingly less than optimal 5-gram dose for the oral administration? And while some might argue the 5-gram dose came from the United States Equestrian Federation (USEF), that is misleading. The USEF recommends a dose of 5 mg/Lb, OR 5 grams per thousand pounds, and allows methocarbamol up to 12 hours.  The horses were still underdosed using the oral method in the study considering their weights.

Even if one can ignore the fact that 20 horses is very close to the minimum number needed to produce a valid threshold, measuring the residual serum or plasma level based on testing ex-race horses after giving an uncommon oral dosage to some of them and an uncommon IV dose if you account for the training period  to all of them creates some uncertainty.

The odd part of Rumpler’s paper is that he only showed the results for the six horses that were given both the IV dose and the oral dose because according to the subject test table in the paper, the pharmacokinetics were only done on those six horses.   All six of the horses given the single IV dose were below the lower limit of quantitation, in this case 1.0 nanogram/milliliter after 48 hours, the withdrawal period listed by RMTC. The same six horses when given the oral dose also fell below the level of quantitation after 48 hours, although there was a single horse that demonstrated bioaccumulation of 2.7 fold. This was only mentioned once by Rumpler, in the discussion section of his paper. but never really explained. According to Dr. Clara Fenger, Secretary of the North American Association of Racetrack Veterinarians

“This single horse is problematic when trying to extrapolate these data to the entire racing population because the typical use of this substance is for daily use in horses which have problems with tying-up. If one in six horses accumulates methocarbamol (17% of the study population), the risk of a positive with dosing for more days than two would be very high.”

What we also don’t know is whether the horse that accumulated methocarbamol might have been the only one to receive the full dose by stomach tube, given the potential flaw with that delivery method.

Of more concern is that the 1 ng/ml threshold was not determined by a statistical analysis of values achieved by horses in the Rumpler analysis, but represents the limit of the mass spectrometer used to detect the drug. In other words, we don’t know exactly what the levels were at 48 hours for the six study subjects. This can be a dangerous method to use to set a threshold, since without appropriate statistical analysis we only know that the research horses for which results were given were below the threshold after 48 hours and nothing about how the threshold would behave when extrapolated to thousands of race horses. On the other hand, the values at 24-hours were accurately available.

Given what appears to be some questionable choices by the investigators, RMTC using this study to develop a threshold must create some uncertainty. Almost nothing about the investigators’ approach – the test herd, the dosing methodologies, the dosing levels – mimicked what we would expect to see on the racetrack. If horsemen used the veterinarian recommended dose, depending on withdrawal time there could be a much higher likelihood of seeing a violation, and if they used one of the RMTC recommended approaches at the time they ran the risk of lowering the drug’s effectiveness.

Subsequent to the Rumpler paper, there was a study done by Heather Knych et al. published in January 2016 in the same Journal of Veterinary Pharmacology and Therapeutics.

Knych points out some interesting background about dosing. While the RMTC standard dose is 15 mg/kg, she notes that the “label dose for intravenous administration is 4.4-22 mg/kg, and 22-55 mg/kg for moderate and severe conditions, respectively.” In other words, to do a study that mimicked dosing on active racehorses that were subject to tying-up, we would see a variety of administration amounts (based on weight and severity of tying-up) and dosing frequency. The obvious question would be, how can a threshold based on the sub-optimal dosing in the Rumpler analysis be considered valid for a horse that severely cramps and receives three or four times the study dosage during therapeutic treatment?

Knych further points out that, “elimination of methocarbamol is reportedly dose dependent (Muir et al., 1984) and therefore the potential for a positive regulatory finding in performance horses exists if doses in excess of those recommended by USEF (the United States Equestrian Federation) and ARCI (Association of Racing Commissioners International) are utilized.”

Unlike Rumpler, Knych makes it clear that dose size and timing can have a substantial effect on residual blood levels.

What Knych concludes is that further study on clinically used doses and dosing regimens for methocarbamol are warranted to assess whether a more prolonged withdrawal time recommendation would be advised when veterinarians are using higher doses and more frequent administration. In other words, Knych prompts the obvious question: what should the threshold and withdrawal time be considering how the drug is commonly used on the backside?

The Knych study used 16 horses (eight mares and eight geldings) between 4 and 7 years old, and weighing from 419 to 610 kilograms. This was still not ideal, but was slightly better than in the Rumpler analysis.

The Knych study administered methocarbamol using three methods: tablets dissolved in 40-60 mL of water using a syringe and administered directly into the oral cavity; a paste, also administered directly into the mouth; and an intravenous administration.

One major difference between Knych and Rumpler is that the Knych study used higher doses, more accurately reflecting the clinically useful dose. This led to something that didn’t emerge from the Rumpler study – the plasma clearance rate slows and the disposition curve flattens as it approaches the 1 ng/mL threshold, indicating that blood levels may actually remain close to the level of the threshold for many hours.

In the discussion section of the paper, Knych says this:

“The detection time and time above the ARCI-recommended regulatory threshold following administration of both single and multiple doses of 15 g of MCBL were more prolonged in the current study compared with the previous reports that utilized a lower dose (Rumpler et al., 2014). Fifteen of sixteen horses exceeded the ARCI threshold recommendation at 48 hours (the current recommended withdrawal time guideline) postadministration of the final dose. It has been suggested previously that the clearance of MCBL is dose dependent with decreased rates of clearance and a longer elimination half-life reported as intravenous doses increased from 4.4-17.6 mg/kg (Muir et al., 1984)”

In other words, if you are going to base a threshold on the clearance rate and residual concentration for a substance, using the common dosing and administration protocol makes a significant difference to where the threshold is set.

In 2011 Rick Sams, the principal investigator on the study on which the Rumpler paper was based, subsequently did a presentation to the Scientific Advisory Committee (SAC) of the Racing Medication and Testing Consortium (RMTC) stating that the same data used to generate the Rumpler paper did not support a threshold of 1 ng/mL if an oral administration of methocarbamol was given, instead suggesting that a 20 ng/mL threshold at 24-hours withdrawal would be more appropriate. In fact, based on documents subpoenaed by Graham Motion’s attorney, W. Craig Robertson III, in his words, “the SAC unequivocally recommended that the regulatory threshold for methocarbamol be set at 20 nanograms.” If anything, methocarbamol when first administered would have a slight sedating effect on a horse, but certainly at 20 ng/mL post-race level no one would rightfully argue that the drug was in any way performance-affecting.

Sams later provided deposed testimony in a subsequent Kentucky Racing Commission prosecution of a methocarbamol case that the threshold ultimately adopted by the RMTC Board was done so at the insistence of one Scientific Advisory Committee member and lab director to accommodate that member. Sams stated in his testimony that the lab director was unwilling to change the 1 ng/mL at 48 hours threshold already in place in his jurisdiction.

Amazingly, the RMTC dismissed the results of the study they funded in favor of an historical threshold in place in a single jurisdiction, despite the fact that there was no solid scientific basis for concluding there was a correlation between the 1 ng/mL concentration and any pharmacological effect of any kind. In essence, Sams concluded the threshold adopted by the RMTC was done at best for political reasons, since in his opinion 20 ng/mL – especially if a 48-hour withdrawal time was used – would have been sufficiently protective of the racing public. And while Sams didn’t know it at the time, the subsequent work done by Knych confirmed the threshold was arbitrarily low.

The final piece of information came from the actions of the RMTC in revising the Controlled Therapeutic Medication schedules. In May 2013, right after the RMTC had the Rumpler analysis, they established the methocarbamol threshold as

• 1 ng/mL (serum or plasma)
• 48-hours withdrawal time
• IV or oral administration
• 15 mg/kg dose for IV, 5 gram dose for oral

Despite the fact that they knew the 5 gram oral dose was going to be insufficient for most horses, the Rumpler analysis used that dose (apparently without strong documentation) and it was transferred to the RMTC schedule.

In December 2014 the RMTC added a footnote to the schedule

Once Graham Motion uncovered all this information, his decision was clear. He would fight the violation.

The KHRC Decision

No one was more surprised at the positive for Kitten’s Point than Graham Motion. He didn’t dispute she was given methocabamol – in fact, 7.5 grams twice a day for the period between her last race, the Orchid, and the Bewitch Stakes, as prescribed by Dr. Jeff Blea, coincidentally a current member of the RMTC Scientific Advisory Committee and someone very familiar with the methocarbamol threshold. According to veterinary records, the medication was withdrawn seven days before the Bewitch Stakes and she was not given any kind of dose 48-hours before the race.

Blea had  employed that dosing routine on Motion’s horses in California without suffering a positive, and even given the change of venue, a seven-day withdrawal should have worked.

Motion’s attorney argued that since Motion withdrew the medication well before the time specified in the CTMS, as well as pointing out potential problems with how the sample was maintained – the room that handled the samples doubled as the outriders’ tack room until two days before the start of the Keeneland meet – there was enough mitigating evidence to dismiss the case.

It is rare for racing authorities to find a trainer not responsible after a medication positive, and Kentucky was no exception in Motion’s case. Kitten’s Point was disqualified, purse money was taken back, and Motion was fined $500 and given a five-day suspension.

Motion responded to the decision by saying, “while I always realized the possibility of a violation despite the highest degree of diligence and understand the need for penalties, in these circumstances I am convinced that my staff and I followed all the required protocols and I plan to appeal the decision for a variety of reasons, all of which were presented at the stewards hearing.”

Motion’s most significant concern was directed to the published RMTC guidelines. “We have to, as trainers, have guidelines. If we can’t follow the guidelines that are issued then I’m not quite sure how we’re supposed to operate. Forty-eight hours is clearly not enough to withdraw from this medication–I withdrew seven days out and it still wasn’t enough. It makes it very hard to operate under those conditions.”

This was not the only time that the RMTC was accused of using bad or absent science. Todd Pletcher beat a positive for Princess of Sylmar in the 2014 Delaware Handicap for the corticosteroid betamethasone because the Deputy Attorney General advising the Delaware Racing Commission identified that the lack of published and recognized science to support the RMTC recommendations would not withstand a court challenge.

Motion’s appeal of the initial stewards’ ruling to the KHRC only resulted in them dropping the five-day suspension but leaving everything else intact. But because Motion was not allowed the opportunity to defend himself by speaking to the KHRC, he published an open letter stating that the threshold for methocarbamol in place at the time “was not supported by good science, including going completely against the recommendation set by the head of KHRC’s testing lab, Dr. Sams.”

Motion saw the elimination of the five-day suspension as at least a partial admission there were concerns with the case.

Motion’s letter prompted Dionne Benson, Executive Director and CEO of the RMTC to respond. Benson fired back against Motion’s claim of a lack of good science.

“Quite simply, the science is not lacking. All existing published scientific research supports the current RMTC-recommended threshold and withdrawal guidelines for methocarbamol, when used in conjunction with the route of administration, total number of doses, and total dose provided in the recommendation.”

What Benson was saying was that if Motion had followed the RMTC guideline instead of the advice of his veterinarian, he would have been fine. At the time of Motion’s violation the RMTC recommendation included two approved routes of administration (IV and oral), a recommendation still in effect in Kentucky, and there was nothing in the CTMS about total number of doses or the frequency of dosing. Benson was asking Motion to follow a protocol that was clearly ambiguous. How could anyone read the May 2013 CTMS and conclude that the RMTC was only specifying a single dose by IV or oral, especially because Rumpler’s paper had used a twice daily “oral” administration of 5 g? It’s impossible for Benson not to have seen how the the publication of the threshold could cause confusion.

Benson, in fact, dismisses the argument that the dosing specified in the CTMS was uncommon by frequency and method of administration by saying,

“Veterinary practitioners desire numerous administration options. The RMTC, however, must focus its finite research and funding resources on thresholds that protect against the route, frequency and dose that represents an administration used closest in proximity to racing. For methocarbamol, that is a single intravenous dose.”

Clearly, Benson was suggesting that a 15 mg/kg intravenous dose was the standard in the industry, despite evidence to the contrary and despite the obvious contradiction in dose size between IV and oral. But remember that at the time of Motion’s violation the RMTC’s recommended dosing included IV and oral. The RMTC must have seen the problem after Motion’s positive because they took action by revising the CTMS in February 2016. It is difficult to not see Benson as misleading, since she knew the dosing recommendation was different in 2015 and she knew that the Knych study come to a very different conclusion when horses were dosed more in line with standard veterinary practice, but chastised Motion for not following a guideline that was not in place in that form in 2015. In other words, they revised the CTMS and then blamed Motion for not following it.

She continues,

“Contrary to Mr. Motion’s assertions, the studies published by both Dr. Sams and Dr. Knych included single intravenous administrations of methocarbamol. Under each research protocol, all results were below the RMTC proposed threshold at 48 hours. Thus, all existing published research supports the RMTC’s recommended 48-hour withdrawal guideline for a single intravenous methocarbamol administration.”

Benson’s rationale is once again that Motion used a different methodology than the RMTC recommendation, and that was the cause of his problems, completely ignoring the conclusion Knych came to that, “The detection time and time above the ARCI-recommended regulatory threshold following administration of both single and multiple doses of 15 g of MCBL were more prolonged in the current study compared with the previous reports that utilized a lower dose (Rumpler et al., 2014). Fifteen of sixteen horses exceeded the ARCI threshold recommendation at 48-hours (the current recommended withdrawal time guideline) postadministration of the final dose.”

Remarkably, Benson concludes by saying,

“It is simply impractical to perform research and provide guidance for every route of administration, dosage, and duration of administration that each trainer or veterinarian may prefer. This is cost prohibitive and, frankly, a misuse of resources. The RMTC recommendations are based on common dosages and routes of administrations from practicing veterinarians and veterinary pharmacologists. If used conscientiously as recommended, they will allow the use of therapeutic medications in training without fear of a post-race violation.”

Even if you grant her point about practicality and cost, there is no reason that the RMTC should have ignored the clear recommendation of the SAC and Rick Sams to set the threshold at 20 ng/mL, and seemingly ignore the breadth of the work done by Knych. Given the RMTC’s problems with regard to the flunixin, betamethasone and methocarbamol standards, it is becoming more difficult for horsemen to trust them. The message to horsemen seems to be, when the CTMS is in conflict with what your veterinarian recommends, ignore your veterinarian if you want to stay out of trouble, regardless of what he tells you is necessary for the protection of the horse.

Graham Motion is committed to seeing this case to a final conclusion. While he remains overall supportive of the efforts of the RMTC to develop fair, science based thresholds, Motion believes it is important for trainers to have confidence that following the RMTC guidance will not lead to violations of the thresholds.

Motion spoke for many horsemen when he said, “It’s a real problem when you’re trying to follow the rules and still have a positive.”

It’s a problem that can and should be solved, at least for methocarbamol.