1. Which different procedures have been used by the laboratory?

The laboratory used GC/MSD-SIM and GC/C-IRMS (Carbon 13 stable isotope ratio mass spectrometry)

2. For each procedure, describe briefly the important steps.

GC/MSD-SIM:

The sample goes through the injector into a capillary column, filled with adsorbent or stacionary phase. The sample vaporizes and goes through the column with the help of helium gas. Separation of the sample happens because every component has a different speed of traveling through the column – it depends on the retention of the components on the stationary phase. In the end, we use a detector, in this case SIM. It separates the ions due to their mass and charge. The components are being ionized and the ions come under the influence of the instrument’s magnetic field and go towards the detector. In the detector we get a signal that is proportional to the number of ions. In the end we get a chromatogram - chart of signals, dependant on the time.To calculate the T/E ratio the area under the testosterone peak is divided by the area under epitestosterone peak.

GC/C-IRMS: Sample of urine (with possible chemical processing) is injecteed into a capillary column GC. GC separetes the mixture into individual components that elute in different retention times. The components pass through a combustion furnace with the help of continuous stream of helium gas. This vapour stream passes through either a membrane filter or a cryogenic trap – the water is removed and the CO2 enters the isotope ratio mass spectrometer, where it is ionized. The singly charged ions come under the influence of instrument's magnetic field and the path they follow shows as a arc, as they go towards the detector. The interest is in ions that weigh 44 or 45 mass units (1C+ 2O or 1C13 + 2O). Then they use a endogenous internal standard and use GC/MSD with the same operational program. They can then identifiy the peaks by creating relative retention times.

3. For each procedure, define - as exactly as possible - what is the intended use.

GC/MS: It is ideal for the analysis of the hundreds of relatively low molecular weight compounds found in environmental materials. It is used for drugdetection (in this case to determine the testosterone/epitestosterone ratio),fireinvestigation, environmental analysis,explosivesinvestigation, and identification of unknown samples.

GC/C-IRMS: It is used to ascertain the realative ratio of light stable isotopes of carbon (13C/12C), hydrogen (2H/1H), nitrogen (15N/14N) or oxygen (18O/16O) in individual compounds separated from often complex mixtures of components. In the Floyd Landis case they say it is a “gold standard” to determine if sports doping was indicated, by examination of a urine sample from an athlete. They also use it for authenticity control of foodstuffs and determination of origin in archaeology, geochemistry, environmental chemistry and forensic science.

4. Identify and describe what are - in Bob Blackledge's opinion - the noncompliances when these procedures were applied.

  • Mismatching identification numbers of the sample and the laboratory -> was the sample really from Landis?
  • The original sample id number hs been whited out and a different number entered
  • The concentration of free testostreone and/or epitestosterone in the specimen exceeds 5% of the respective glucuroconjugates so the sample is either contaminated or degraded -> by WADA guideliness, results from such samples are unreliable, but they ignored the guideliness and proceeded on to IRMS.
  • Linearity checks were done not even once a month on the IRMS instruments, but should be done frequently.
  • Data files were deleted -> numerous time gaps of several hours were found. When technicians didn't like the results for a given injection, they did a repeat using the same file name and they overwrote the old file.
  • The peaks of testosterone and epitestosterone in GC/MSD were not well-separated from their nearest neighbours and the epitestosterone peak is unsatisfactory.
  • There is no proof that three ions were used for SIM, like there should be, considering the WADA requirenments, because using only one is untrustwothy as an identification of a specific compound.
  • Experimental results reported anywhere from two significant figures, all the way up to five.
  • The relative retention times fell outside the limits specifies in the LNDD SOP.
  • By trying to separate complex mixtures by GC, the forces that cause the separation of the components also cause some slight stable isotope fractionation ->the result of this is even with the perfect baseline separation of the peaks of interest the stable isotope ratio values will be in error if integration errors are made in selecting beggining and ending of the peak.
  • Identity of the athlete should be unknown, but the analyst could have figured it out which samples were from Landis, because of his elevated cortisone levels (he was taking cortisone for his arthritic hip).
  • Critical evidence, stored as electronic data files (EDF) had been erased from the hard drive and the original data was destroyed at the LNDD.
  • WADA and LNDD do not have the same criteria for a positive test of sports doping with IRMS.

5. For each of the procedures you identified, imagine that you would need to validate these in your laboratory. Can you then list the most important parameters to validate and briefly describe your reasoning.

The most important parameters for validation would be accuracy, precision, selectivity, stability, reproducibility, LOD, LOQand linearity.

(1) Accuracy: It is important that we get an accurate result when measuring. Accuracy is determined by replicate analysis of samples containing known amounts of the analyte. We should use a minimum of three to five determinations per concentration.

(2) Precision: Also, the method should be precise.The result of the measurements must be close one to another, when the procedure is applied repeatedly. Precision should be measured using a minimum of three to five determinations per concentration as well. An added internal standard would also provide better precision.

(3) Selectivity: A good selectivity is also important. Because we have a biological matrix – urine, there is a good possibility that all the other components of the sample could interfere with our analyte. Analyses of blank samples should be obtained from more sources and each blank sample should be tested for interference.

(4) Stability: When determining drugs in a biological sample, stability plays an important role. The analyte has to be stable until the measurements. Stability depends on the storage conditions, the chemical properties of the drug and the matrix. It should be defined how to collect the sample, how to handle, storage and analyse it. There could be differences in stability if we keep the sample on the room temperature or if we freeze it. The procedure should also include an evaluation of analyte stability in stock solution. All stability determinations should use a set of samples prepared from a freshly made stock solution of the analyte in the appropriate analyte-free, interference-free biological matrix. Stock solutions of the analyte for stability evaluation should be prepared in an appropriate solvent at known concentrations.

6) Reproducibility: If anything goes wrong with the experiment, it is important that reproducibility of the procedure is evaluated and analysis could be remade.

7) LOD, LOQ: This is the lowest analyte concentration that can be detected (LOQ - quantitatively determined) and identified with agiven degree of certainty and it is important that we know in which range it is possible to measure. If we have a procedure with a high LOD and LOQ, we could say that there are no drugs in the sample, because the instrument would not detect them, but the drugs would be present, just in lower concentration.

8) Linearity: Procedure should have a directly proportional relationship between the response and concentration of the analyte in the matrix over the working range, so we get a correct result. That is why a calibration curve should be prepared in the same biological matrix as the samples in the intended study by spiking the matrix with known concentrations of the analyte. Concentrations of standards should be chosen on the basis of the concentration range that is expected in a particular sample.