The aim of the present work was to compare the performances of HPLC-High Resolution Mass Spectrometry (HPLC-HRMS) in full scan and UPLC- Multiple mass spectrometry (UPLC-MS/MS) in multiple reaction monitoring (MRM) mode for the systematic, general unknown screening of drugs and pharmaceuticals in hair samples collected for clinical toxicology purposes at the local University Hospital.

10mg hair were washed and extracted by a rapid, micropulverized method already published. Extracts were analyzed in two different instruments, the former consisting of an Orbitrap coupled to a Finnigan Surveyor HPLC, the latter of a Xevo TQD triple quadrupole coupled to an Acquity UPLC. For HPLC, an Atlantis T3 (150 × 2,1mm, 3μm) was used while the UPLC employed a HSS C18 (150 × 2.1 × 1,8μm), with a gradient of ammonium acetate and acetonitrile (formic acid 0.1%) at 300μL/min and 400μL/min respectively. The Orbitrap was operated in positive ion electrospray mode at a resolution of 60000 with mass locking, in full scan mode, with a in-house developed library of 600 compounds; the identificatiion criteria were 2ppm accuracy of the exact mass of the protonated molecule, ± 0.1min of the retention time, a minimum area count of 10000 arbitrary units; the Xevo TQD worked in switching ion polarity electrospray mode, and in MRM using a Waters toxicology library of 800 (untargeted screening) or 180 compounds (targeted screening), a recognition window of ±0.2min and a minimum area count of 600 arbitraty units.

Compounds belonging to drugs of abuse and/or to pharmaceutical drugs could be identified with both analytical approaches in the same hair samples. In more than 100 hair samples collected in the clinical settings in a year, drugs belonging to the following therapeutic categories were identified: anaesthetics, anti-epileptics, antidepressants, benzodiazepines, sedatives, hypnotics, opioids, anti-psychotics, neuroleptics, barbiturates, stimulants, non steroidal analgesics, steroids, beta-blockers, diuretics, antibiotics, antiarrhythmic agents, antihypertensive drugs, anti-coagulants, hypoglycaemic agents. The screening results were in every case confirmed by injection of the tentatively identified analyte(s) in methanol standard mixtures. When possible, the compliance with prescribed therapies or alleged use/abuse was verified.

The UPLC-MRM method turned to be more rapid than the HPLC-HRMS (14min vs 30min total run time); the possibility of polarity switching could be applied in routine only in the Xevo TQD allowing to easily identify acidic drugs such as non steroidal analgesics. The HPLC-Orbitrap, working in full scan, resulted superior for implementing the library with new molecules, since no adjustment of the acquisition method was needed, and for the postacquisition, retrospective screening of the raw files. In the UPLCMRM method, limits of identification with targeted analytes were in the order 0.1 – 0.2ng/mg for basic drugs, with 10mg hair samples, whereas in the HPLC-HRMS limits were generally lower.