A world-renowned expert in racing chemistry and toxicology, Dr. Richard Sams, after a review of the data packets provided to the New Jersey Racing Commission (NJRC) by Truesdail Laboratories regarding methocarbamol findings from Monmouth Park in 2019, opined that, “Based on my review of the data, I conclude that the quantitative analysis of horse blood for methocarbamol is likely both inaccurate and imprecise because Truesdail did not use the standard techniques that other racing laboratories use to assure that methods are fit for purpose.”
Background
I recently wrote a piece posted on Horse Racing Reform titled, “New Jersey drug finding: Bad regulation and poor judgment equals injustice for Monmouth Park trainer,” in which I described NJRC’s prosecution of a case despite a referee laboratory’s finding of a concentration of the therapeutic medication methocarbamol that was below the threshold adopted by the commission in its own regulation.
The race in question was the third race at Monmouth Park on July 5, 2019, for 2-year-old New Jersey-bred maidens. The owner and trainer of the horse, Shield of Faith, was Glenn Thompson.
Despite the finding by the referee laboratory, Kenneth Maddy Laboratory at the University of California at Davis (U.C. Davis), the board of stewards at Monmouth Park held a hearing and then issued a ruling (signed by only two of the three stewards) that fined Mr. Thompson $500, suspended him for 15 days, and disqualified the horse, thereby requiring the forfeiture of the purse. No mention was made in the ruling of any concentrations or the fact that the split sample result was under the regulatory threshold.
The concentration of methocarbamol in Mr. Thompson’s horse reported by Truesdail Laboratories was 5.0 nanograms per milliliter (ng/mL) but the U.C. Davis laboratory reported a concentration of only 0.75 ng/mL, a difference of more than fivefold. The concentration of methocarbamol in another horse at Monmouth Park decreased from 8.9 ng/mL to 1.5 ng/mL, another difference of more than fivefold. In addition, the methocarbamol concentration of a third horse (again at Monmouth Park) was reported at 19.0 ng/mL by Truesdail but 1.5 ng/mL, a difference of more than 13-fold.
I obtained Truesdail’s data packets on two of the findings in question through an official open records request and asked Dr. Sams to review. One packet was of the finding 5.0 ng/mL in Mr. Thompson’s horse, the other was the finding at 8.9 ng/mL. I was told that New Jersey data packets in the custody of the racing commission are limited to those instances when the board of steward’s decision are appealed.
Note: During my tenure as executive director of the Indiana Horse Racing Commission (IHRC), the IHRC discontinued its relationship with Truesdail Laboratories. (Click here for a link to article.)
Dr. Sams has served as a laboratory director responsible for equine drug testing at The Ohio State University, University of Florida, and LGC Laboratory in Lexington, Ky., where he oversaw the testing of all Kentucky racing samples. He has had more than 50 years of experience in measuring drugs and metabolites in biological samples for pharmacokinetic, diagnostic, and regulatory applications. He has also advocated for use of quantitative thresholds to regulate endogenous substances and therapeutic drugs in sports, including horse racing.
I asked Dr. Sams to explain the purpose of a data packet. He said, in part:
“The laboratory data packet (also known as the litigation packet) is the laboratory's documentation prepared in support of the findings reported on a certificate of analysis. The records in the data packet should contain sufficient information for an experienced analyst to review the record and determine whether appropriate methods had been used, ascertain who performed the analyses and whether they had adequate training and experience, the sources of the standards used to prepare calibrators (if a quantitative result was produced) and control materials, how the quantitative result was obtained, who had access to the sample after it was received by the laboratory, what test results were obtained for all tests run on the sample, and whether the total amount of sample can be accounted for since a discrepancy can indicate that a test was run but not reported.”
Dr. Sams’s review
A portion of Dr. Sams’s review focused on the 1:5 dilution of the test sample as part of the preparation process by Truesdail, which may have accounted for the laboratory’s findings of concentrations approximately five times greater than those of the split sample laboratories.
“Documentation provided by Truesdail indicated that sample M20406 (Mr. Thompson’s horse) from Monmouth Park was prepped “1:5” suggesting that analysts processed a smaller aliquot than normal (1 mL) because the sample concentration appeared to be high in the screening analysis. However, there is no documentation that analysts actually prepared a 1:5 dilution. Furthermore, documentation in chain of custody records indicates that two 1-mL aliquots were requested and used. If 1:5 dilutions were prepared, there is no documentation of this step in the records, so it is unclear whether these dilutions were prepared. The peak area ratios determined for both aliquots of the test sample were approximately the same as that for the QC sample prepared at a nominal concentration of 1 ng/mL. If the analyst failed to prepare a 1:5 dilution of the test sample, then the concentration determined from the analysis should have been approximately 1 ng/mL and not 5 ng/ml.
“If the samples were intended to be diluted by Truesdail personnel but were not, then the ratio of the concentration determined at Truesdail would be five times greater than the actual concentration. The ratio of the concentrations determined at Truesdail to those determined at the split sample laboratories were close to five (6.7 and 5.9, respectively) suggesting that Truesdail may have erred in applying a dilution factor to the calculations without having diluted the samples. Unfortunately, the documentation in the data packets is inadequate with regard to these specific questions.”
Dr. Sams opined on other factors that may have affected the accuracy of the findings, including the following three areas:
“Whereas published methods for determination of methocarbamol in horse plasma or serum describe the use of this internal standard (i.e., methocarbamol-d3), Truesdail used meperidine-d4. This substance is inappropriate for use as an internal standard because it elutes at a different retention time from methocarbamol and its chemistry is completely different from that of methocarbamol so it would be expected to experience different matrix effects compared to methocarbamol. Use of more appropriate internal standards would likely improve both accuracy and precision.
“The calibration line for determination of methocarbamol was based on five calibrators and was analyzed without appropriate weighting of the data. During my review of the data, I repeated the regression analysis using appropriate weighting of the calibration data … The slope and intercept for the calibration line determined with proper weighting are different from those reported by Truesdail. Consequently, concentrations of methocarbamol in official samples would be different if appropriate weighting had been used. Therefore, use of appropriate weighting of calibration data will likely improve both accuracy and precision.
“A standard practice is for laboratories to specify use of independent sources of the analyte used to prepare calibrators and that used to prepare control materials. If the purities of the material from the different sources are not the same, then bias will be detected during method validation and source materials can be investigated. If different sources are not used, then gross errors in purity or dilution will not be detected due to the common sourcing. There is no evidence in Truesdail’s records that they used different sources of methocarbamol to prepare calibrators and control materials. Since this information is not documented, it must be assumed that they used a single source. Use of independent sources of the analyte for preparation of calibrators and control materials will assure accuracy because discrepancies between target and measured concentrations in control materials will result in batch failures and rejection of data.”
Dr. Sams concluded:
“Comparison of the quantitative reports for two findings for methocarbamol from Truesdail and those from split sample laboratories reveals gross differences that are greater than those that can be attributed to imprecision. If it is assumed that the same samples were tested by the primary and split sample laboratories, then the differences must be attributed to gross inaccuracy due to improper preparation of calibrators or errors in calculation, such as the use of a dilution factor when dilution was not used.”
In my experience, differences in laboratory analysis can vary depending on what is specified in the contract between a racing commission and its equine testing laboratory. The contractual agreement between the state of New Jersey and Truesdail Laboratories was not part of the review undertaken by Dr. Sams.
What’s next?
In my more than 25 years as a regulator, I never experienced, nor heard of, a situation similar to this. New Jersey has had three findings for the same drug (methocarbamol) within a brief period of time that were five to more than 10 times higher than the split samples tested at two different laboratories.
I know what I would have done if this situation had occurred while I was executive director of the IRHC. I would have undertaken an immediate review. In order to get to the bottom of the issue and maintain the trust of the horsemen and the public, I would have ensured that the review was independent, thorough, and transparent. By transparent I mean that the complete findings of the review would have been made available to the public.
That’s what I would have done.
What will New Jersey do?