Multivariate drug analysis

A simple analytical approach to identifying drugs of abuse would be a boon to forensic scientists and law enforcement agencies. A collaboration between researchers in the US and Europe demonstrates how an assessment of different methods using chromatography coupled to mass spectrometry reveals that multivariate selectivity can take into account the degree of resolution between nominally unresolved peaks due to the presence of various drugs in a forensic sample and so allow quicker identification.

Sarah Rutan of the Department of Chemistry, at Virginia Commonwealth University, Richmond, Virginia, USA, working with Melanie Dumarey and Yvan Vander Heyden of the department of Analytical Chemistry and Pharmaceutical Technology, at the Vrije Universiteit Brussel, in Belgium describe details in the journal Analytical Chemistry.

The researchers explain that, "The main goal [of drug forensics] is to trace and identify as many drugs as possible in the shortest possible time preferably with a simple analysis method." They add that one option is simply to screen samples using a liquid chromatography-diode array detection (LC-DAD) system. Several drugs might then be analysed at once and the only problem is to select the most appropriate columns. Column selection has previously been carried out using various correlation measures to find the most dissimilar columns. However, these measures do not take into account the identification power of the individual systems, the team says. The problem would then be that many drugs might not be identified on such a system.

"In forensic laboratories, powder samples, liquids or biological fluids often need to be screened for a large number of drugs in a short time," the team says.

The researchers have now looked at three other measures to see which can optimise identification power on a parallel screening system of two columns or by comprehensive two-dimensional LC (LC × LC). They explain that the simplest approach to assessment would be to simply count the number of compounds separated based on retention times above a cut-off point, but this would ignore co-elution pattern of unidentified drugs. The second criterion would look at differentiate co-elution patterns and identify systems giving the same number of compounds. That approach also has disadvantages so the team has turned to multivariate selectivity.

"The third tested parameter takes into account the degree of separation of all compounds and has the advantage that it reflects the gain in identification power achieved by introducing DAD data," the team explains.

To test the three different measures, the team recorded retention times of 25 drug compounds (methapyrilene, zolpidem, aminoflunitrazepam, pyrilamine, chloripheniramine, chlodiazepoxide, bromazepam, desipramine, oxazepam, nitrazepam, amitriptyline, estazolam, perphenazine, clonazepam, desalkylflurazepam, nordazepam, triazolam, flunitrazepam, temazepam, clobazam, lormetazepam, diazepam, halazepam, prazepam and buclizine) on 13 different chromatography columns. All retention times were normalized before analysis and assessment using various new approaches as well as the more conventional correlation coefficient (R²and Snyder's F_s value.

The team emphasises that (dis)similarity measures such as F_s and R² do not necessarily find the most dissimilar chromatography systems having best identification power. Moreover, they don't distinguish between parallel one-dimensional separations nor comprehensive two-dimensional separations. The F_s metric does have one advantage in that columns can be selected without undertaking the separations of the target compounds, which can be used as a pre-assessment before more informative experimental measurements are made.

They add that information theory allows them to distinguish between two separations with the same number of identifiable compounds, but with a difference in co-elution pattern but it is multivariate selectivity that is the only approach capable of quantifying the degree of co-elution as well as the degree of separation.

"The latter can evaluate the number of compounds additionally identified by applying multivariate curve resolution methods when incorporating spectral information obtained with a DAD," the team conclude. "We feel that this metric offers the most potential for comparing the performance of different analytical methods."