Validation of a quantitative screening method by liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) from a dried blood spot (DBS)

Aim

To develop and validate a quantitative screening method for a panel of 300 compounds by LC-MS/MS using a DBS.

Method

A drop of blood (10 μL) is applied to HemaXis DB 10 filter paper (HMX-DB-10) and left to dry for 3 hours. The compounds of interest are then extracted by a two-step liquid-liquid extraction: first with 2 mL of a hexane/ethyl acetate mixture (50/50) in an alkaline environment (pH 9.7), followed by 2 mL of a chloroform/isopropanol mixture (80/20). After evaporation of the supernatants, the residue is reconstituted with 80 μL of a mixture of 2 mM ammonium formate (0.1% formic acid) and acetonitrile (50/50). In all, 10 μL is then injected into the system. Separation is performed on a HypersilGold PFP column (100 × 2.1 mm, 1.9 μm) using a gradient mode with 2 mM formate buffer (0.1% formic acid) and acetonitrile. Detection is performed in MRM mode using a triple quadrupole mass spectrometer (TSQ Altis, ThermoFisher®).

Results

The method was validated for several panels of compounds, including traditional drugs of abuse (cocaine and its metabolites, 13 amphetamines, 28 opioids), NPS (12 synthetic benzodiazepines, 29 cathinones, 26 synthetic cannabinoids, 26 new synthetic opioids, other NPS), as well as several drug classes (30 benzodiazepines, 37 antipsychotics, 30 antidepressants, 12 antiepileptics, 18 β-blockers, 8 anesthetics, 8 sartans, etc.). The limits of quantification (LOQ) ranged from 5 ng/mL (86% of compounds) to 10 ng/mL (14% of compounds), with bias and CV ≤ 20%. Linearity was validated over a range of 5–10 to 500 ng/mL with satisfactory performance in terms of accuracy and precision (bias and CV ≤ 15%). Analysis of a 200 ng/mL quality control showed accuracy (bias + precision ≤ 15%). No carry-over effect was observed in blanks injected after the high point (500 ng/mL). Selectivity was verified using negative blood samples, with no significant interferences. Patient samples were quantified, and results were compared to those obtained by the routine blood screening LC-MS/MS method used in the laboratory. The concentrations obtained were similar between the two techniques. Studies on stability, recovery, and reproducibility are on going.

Conclusion

The use of DBS for toxicological analyses or pharmacological therapeutic drug monitoring (TDM) is gaining popularity. This is due to several advantages: a simplified sampling procedure, a small sample volume, and good analyte stability on this type of substrate.

The adoption of DBS could also improve patient compliance, thanks to the simplicity of the collection (which can be done by the patient themselves), easy transport (by post), and storage without constraints (at room temperature). Additionally, the small volume of blood required could be a solution in specific contexts, such as pediatric sampling or multiple analyses, where samples are often limited.

However, the analysis of very small blood volumes (10 μL) poses a significant challenge, requiring the use of instruments with very high sensitivity. Another important constraint is the nature of the matrix being analyzed: whole blood. Since TDM is primarily performed on plasma, models for predicting plasma concentrations from DBS results need to be developed and validated. Blood viscosity also affects the results, as it can impact sample homogeneity or extraction efficiency. Hematocrit determines blood viscosity, and its effect should be evaluated during analytical validation.

Advances in analytical techniques enable increasingly higher sensitivity. The results demonstrate the potential of DBS for therapeutic drug monitoring applications or forensic toxicology.