Evaluating a Statistical Method for Comparison of Mass Spectra of Novel Psychoactive Substance

Identification of novel psychoactive substances (NPS) based on comparisons of electron-ionization (EI) mass spectra continues to pose challenges in seized drug analysis. The continual evolution of NPS analogs, along with their related structural and positional isomers, results in EI spectra that visually appear very similar and are difficult to distinguish. 

A method to statistically compare EI mass spectra was developed to provide a more objective means of comparison. The method is based on the unequal variance form of the t-test, which is used to compare the mean intensities of corresponding mass-to-charge ratio (m/z) values in two spectra. At each m/z value, the null hypothesis (H₀) states that the difference in mean intensity is equal to zero. If H₀ is accepted for all m/z values, the two spectra are statistically indistinguishable, and in these cases, the two spectra represent the same substance. In contrast, if H₀ is not accepted at one or more m/z value, the two spectra are statistically distinguishable. In these cases, the spectra represent two different substances.  

Current work focuses on evaluating the robustness and ruggedness of the statistical comparison method. For this evaluation sets of synthetic cathinone structural isomers, fluorobutyrylfentanyl (FBF) positional isomers, and fluoroisobutyrylfentanyl (FIBF) positional isomers were selected. Spectra for each compound were collected over time, at various concentrations, and on different instruments. The association of corresponding isomers and distinction of different isomers were demonstrated at the 99.9% confidence level. In instances of discrimination, the m/z values responsible for discrimination were determined, and in many cases, were defined as low-intensity ions that are not readily visible in the spectra. The statistical comparisons will be discussed in more detail, ions that are considered reliable for discrimination of each set of isomers will be identified, and recommendations to ensure accurate comparisons will be provided. 

Detailed Learning Objectives

1. Attendees will be introduced to the statistical theory behind the comparison method. 
2. Attendees will be provided with example applications of the method to distinguish structural and positional isomers of NPS analogs.
3. Attendees will be provided with recommendations for accurate comparisons of mass spectral data.