Development of LC-MS/MS toxicology screening methods for common poisoning agents in South Africa

Tiya, Luthando Lukhanyo (2020-12)

Thesis (MSc)--Stellenbosch University, 2020.

Thesis

ENGLISH ABSTRACT: Background: Hospitals admit many patients suffering from acute poisoning on a regular basis. Over a single year period the Tygerberg Poison Information Centre (TPIC) dealt with 4 771 consultations related to human exposure to various poisonous substances. Of these consultations, exposures to medicines and pesticides accounted for the majority of cases. As a response to this medical burden, this project aims to develop effective screening tools for both medicines and pesticides in human plasma. It is believed that such a tool could enhance how medical practitioners accurately treat and respond to serious poisonings. There are many toxicological screening methods employed today. In the past immunoassays (IAs) and gas chromatography coupled to mass spectrometry (GC-MS) were seen to be the gold standard for the identification of compounds. Due to false positive results associated with immunoassays and the labour intensive sample preparation associated with GC-MS, it is clear that there is a need for faster, more robust and more sensitive bioanalytical methods. This study’s objective therefore is to develop a cost effective, fast and robust liquid chromatography tandem mass spectrometry (LC-MS/MS) screening tools for the identification of drugs and pesticides. Relevant drug and pesticide selection was done by consulting with the TPIC, a centre which deals with poisoning cases on a daily basis, as well as clinicians in the Division of Clinical Pharmacology who deal with patients admitted at Tygerberg Hospital daily and CropLife SA, an organisation that has the knowledge of which pesticides are currently in use in South Africa. A total of 37 drugs and 10 pesticides were identified as those often suspected to be involved in poisoning/overdose cases. Two methods were developed: An untargeted liquid chromatography quadrupole time-of-flight mass spectrometry (LC-TOFMS) method for the determination of therapeutic drugs; and a targeted LC-MS/MS method for the determination of pesticides. Method: Analytes were extracted from CPD plasma using protein precipitation (PP). For the untargeted method, a Waters Synapt G2 Quadrupole Time-of-flight (QTOF) mass spectrometer (MS) fitted with an ultra-high performance liquid chromatography (UHPLC) was used. The separation was done using a Waters HSS T3 (1.8 μm, 2.1 ×150 mm) column. A gradient consisting of a mobile phase of 0.1% formic acid in both water (A) and in acetonitrile (B) at a flow rate of 0.25 mL/min. Ionisation mode was set for both positive and negative (ESI +/-); Nitrogen as the desolvation gas was set at a flow rate of 650L/h; Desolvation temperature was set at 275˚C. The instrument was operated in the MSE mode. The total run time was 37 min. Four concentration levels (10, 50, 100, 1000 ng/mL) were determined to cover the pharmacokinetic peak concentrations of all drugs. This range was determined according to sub-therapeutic, therapeutic and overdose type concentrations. Furthermore, the data generated from the untargeted method development process was used to accurately identify analytes using online search libraries. For the targeted method, two liquid chromatographic separation methods were developed on a Shimadzu 8040 LC-MS/MS instrument. One gradient method using the Agilent Poroshell 120 EC-C18 (2.7 μm, 4.6 ×100 mm) column was developed for atrazine, MCPA, fipronil, methomyl and aldicarb, while a second isocratic method utilizing of the Restek Raptor Biphenyl (2.7 μm, 2.1 x 100 mm) column was also developed for imidacloprid. Both methods used mobile phase consisting of 0.1% formic acid in water (A) and in acetonitrile (B) with a flow rate of 0.5 mL/min. The LC program on the Poroshell method starts at 30% (B) for 2 min followed by a 0.5 minute ramp to 90% (B); 2.5 min at 90% (B); 0.5 min to 30% (B), equilibrate at 30% (B) for 2.5 min. The total run time is 9 min. The biphenyl method isocratic run comprises of (65%:35%, v/v) for a run time of 1.75 min. A semi-quantitative method with a quadratic curve fit with 1/x regression over the range of 20 to 600 ng/mL was chosen. The two targeted methods were subjected to a partial validation to ensure that the methods were accurate, robust and reliable. Results: The untargeted method was able to detect and identify 92% of the therapeutic drugs included in the screen. Due to matrix effects and differences in compound ionisation some could not be detected at lower concentrations. Polarity and thus lack of retention on the columns, difficulty in ionisation, and the inability to produce stable fragments made the detection of valproate, levetiracetam and metformin not possible on the method. A screening tool for simultaneous determination of pesticides was optimized. Due to several challenges associated with the LC-MS/MS determination of glyphosate, paraquat and deltamethrin, the study resorted to excluding these three pesticides from the current methods. The targeted method was developed for the determination of seven pesticides. The USA Food and Drug Administration (FDA) and the European Medicines Agency (EMA) guidelines were followed for a partial validation of the semi-quantitative method. Three independent runs were assessed to measure accuracy and precision. The calibration curve was linear over the ranges 20 to 600 ng/mL. Matrix effects, recovery, process efficiency, on-bench stability, freeze/thaw stability, storage stability, and whole blood stability for the methods was successfully determined for the seven pesticides. Conclusion: To respond to the medical burden of acute poisoning cases faced by our hospitals, this study has developed efficient screening tools for both relevant medicines and relevant pesticides in human plasma. The latter method was subjected to a partial validation, ensuring accuracy and reliability of results that will be applied to the clinical management of poisoning patients.

AFRIKAANSE OPSOMMING: Agtergrond: Hospitale sien baie gevalle van pasiente wat vergiftig is. Tygerberg Gifstof Inligtingsentrum (TPIC) het 4771 gevalle in een jaar gehad waar mense aan verskillende gifstowwe blootgestel is. Die meeste van hierdie gevalle was die medisyne en pestisiede. Hierdie projek het gepoog om metodes te ontwikkel waarmee daar vir hierdie stowwe getoets kan word in menslike plasma. So ‘n metode kan ‘n hulpmiddel vir dokters wees om die beste behandeling vinnig aan pasiente met vergiftiging te kan gee. Daar bestaan heelwat toksikologiese toetsmetodes waar daar vir meervuldige stowwe in een analise getoets kan word. In die verlede is immuumgebaseerde metodes en gaschromatografie tesame met massaspetrometrie (GC-MS) gesien as die goue standaarde vir die identifisering van gifstowwe. Daar is egter vals-positiewe resultate as die immuungebaseerde metodes gebruik word en die GC-MS metodes is lank en die monstervoorbereiding omslagtig. Daar is ‘n behoefte aan vinniger, meer betroubare en sensitiewe bio-analitiese metodes. Die doel van hierdie studie is om ‘n koste-effektiewe, vinnige en robuuste metode te ontwikkel vir die identifisering van farmaseutiese middels en pestisiede deur van vloeistof chromatografie tandem massa spektrometrie (LC-MS/MS) gebruik te maak. ‘n Aantal van hierdie middels is gekies in konsultasie met beide TPIC, wat op ‘n daaglikse basis met vergiftigings werk, en dokters in die Divisie van Kliniese Farmakologie by Tygerberg wie gereeld met pasiente werk wat vergiftig is. CropLife SA het baie kennis oor watter pestisiede in Suid-Afrika gebruik word en Dr Gerhard Verdoorn is gekonsulteer oor die keuse van pestisiede wat ingesluit is in die studie. Daar is besluit op 37 farmaseutiese middels en 10 pestisiedes wat gereeld verdink word as oorsake van vergiftigings en oordosisse. Twee metodes is ontwikkel: ‘n ongeteikende (LC-TOFMS) metode vir die farmaseutise middels en ‘n geteikende LC-MS/MS metode vir die pestisiede. Metodes: Die middels is uit die plasma ge-ekstrareer deur van proteïen presipitasie (PP) gebruik te maak. ‘n Waters Synapt G2 Kwadrupool Vlugtyd (QTOF) massaspektrometer (MS) met ‘n Acquity uiters hoë werkverrigting vloeistofchromatograaf (UHPLC) is gebruik met ‘n Waters HSS T3 (1.8 μm, 2.1 ×150 mm) kolom. Die mobiele fase het bestaan uit 0.1% mieresuur in beide water (A)en in asetonitriel (B) teen ‘n vloeispoed van 0.25 mL/min. Beide positiewe andnegatiewe ionisasie modus was gebruik (ESI +/-); Stikstof was die desolvasie gas teen 650 L/h; die desolvasie temperatuur was 275˚C. Data was gegenereer in MSE mode. Die metode was 37 minute lank. Daar is vier verskillende konsentrasies ingespuit om al die moontlike vlakke van farmakokinetiese topvlakke van die verskillende middels in te sluit (10, 50, 100, 1000 ng/mL). Daar is ook bewys hoe hierdie metode in die studie van aanlyn biblioteke gebruik kan maak om onbekende middels in die plasma te identifiseer, Daar is twee skeidingsmetodes gebruik vir die geteikende metode op die Shimadzu 8040 LC-MS/MS instrument. Die eerste gebruik ‘n Agilent Poroshell 120 EC-C18 (2.7 μm, 4.6 ×100mm) kolom met die gradient. Die tweede is ‘n isokratiese metode waar ‘n Restek Raptor Biphenyl (2.7 μm, 2.1 x 100 mm) kolom gebruik word. Beide metodes gebruik dieselfde mobiele fase as die ongeteikende metode (0.1% mieresuur water (A) en in asetonitriel (B)) met ‘n vloeispoed van 0.5 mL/min. Die LC program van die Poroshell metode begin met 30% (B) vir 2 minute gevolg deur ‘n 0.5 minute oorgang na 90% (B); 2.5 minute by 90% (B); en oor 0.5 minute terug na 30% (B) om te ekwilibreer vir 2.5 minute. Die metode is 9 minute lank. Die biphenyl metode gebruik (65% A:35% B, v/v) vir 1.75 minute. Die kalibrasiekurwe was kwadraties met 1/x regressie vir al die middels oor ‘n konsentrasiegebied van 20 tot 600 ng/mL. Die metodes was gedeeltelik gevalideer om te verseker dat dit toepaslik en akkuraat is. Resultate: Die ongeteikende metode kon 92% van die farmaseutiese middels waarneem waarna gekyk is. Matrikseffekte en verskillende ionisasiepotensiale het veroorsaak dat sommige middels nie by die laagste vlakke gesien kon word nie. Drie middels kon nie met sukses geanaliseer word nie agv van hul polariteit en dus onvermoë om teruggehou te word deur die kolom of swak ionisasie: epilizine (valproate), epikepp (levetiracetam) en diaphage (metformin). A metode om tergelyke tyd ‘n aantal pestisiede (7) te analiseer is opgestel. Drie middels is laat vaar omdat dit nie bepaal kon word met die tegniek nie (glifosaat, paraquat en deltamethrin). ‘n Ander tipe kolom en mobiele fase is nodig vir die analise daarvan en van hulle (glifosaat) sal eers gederivatiseer moet word om dit te kan meet. Die FDA en EMA voorskrifte is gevolg om die metode gedeeltelik te valideer. Drie onafhanklike lopies was gedoen om die akkuraatheid en presisie te bepaal. Matriks effekte, herwinning, proses effektiwiteit, op toonbank stabiliteit, vries en ontvries stabiliteit, storingsstabiliteit en heelbloed stabiliteit was suksesvol bepaal vir die sewe pestisiede. Die kalibrasie kurwe was reglynig oor die konsentrasiegebied van 20 tot 600 ng/mL. Die helling van die kurwe se standaardafwyking was minder as 5% vir al 7 verbindings. Gevolgtrekking: In reaksie op die las wat vergiftigings op die medici in hospitale veroorsaak het hierdie studie effektiewe skandeermetodes opgestel vir relevante farmaseutiese middels en pestisiede in menslike plasma. Die pestisied metode is gedeeltelik gevalideer om sy betroubaarheid en akkuraatheid te bevestig sodat dit gebruik kan word om die gifstowwe wat gebruik is in vergiftigde pasiente te bepaal.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/109098
This item appears in the following collections: