B-185 Why chromatography still matters: the case of an interference in two benzodiazepines LC-MS/MS methods from hospital and reference laboratories

Background In clinical toxicology testing, interferences causing false-positive or negative results are usually associated with immunoassays used for drug screening. Liquid chromatography tandem mass spectrometry (LC-MS/MS) as toxicology’s gold-standard, on the other hand, is often praised for being free of such interferences. This is not always true in practice, however, especially when it comes to using short LC-MS/MS methods for analyzing complex specimens from patients on multiple medications. In this work we report on the observation of the inability of two independently developed LC-MS/MS benzodiazepines methods (from hospital and national reference laboratories) to quantify the metabolite of the CNS depressant clonazepam, 7-amoniclonazepam, due to an unknown interfering substance; and how the third method’s chromatographic performance was likely responsible for successful analyte quantification. Methods Random urine from a patient prescribed 0.5 mg oral clonazepam BID with the last dose of the medication taken 4 hours prior to specimen collection was obtained. The specimen was initially analyzed by Geisinger medical laboratory (GML) on a Sciex 5500 triple quadrupole mass spectrometer and later referred to Quest Diagnostics’ Nichols Institute (QD) and University of California San Diego’s clinical toxicology laboratory (UCSD) for reanalysis. The reporting cutoff levels for 7-aminoclonazepam for GML, QD and UCSD were, respectively, 20, 25 and 20 ng/mL. Additionally, the specimen was analyzed by high resolution MS at UCSD on a Waters Xevo G2 LC-QTOF-MS with the hope of identifying possible interfering substances by spectral library matching. Results At GML, the processed LC-MS/MS batch for the specimen was found acceptable. Two transitions (286.1/222 and 286.1/121) were monitored for quantification and qualification of 7-amoniclonazepam at 0.9 min. Upon inspection of the extracted ion chromatograms (XICs) for the specimen, a small secondary peak was observed for the 286.1/222 transition (∼5% of the base peak, chromatographically resolved) and larger (∼70% of base peak, incompletely resolved) for 286.1/121 transition. The quantifier/qualifier ion ratio (even with manual integration) was outside of the acceptability range of 7-aminoclonazepam indicating the presence of an interferent. Although reporting a “negative” result for this specimen was possible, reported recent clonazepam use by the patient and retention time match led to the request for specimen reanalysis at QD (“unable to report quantitative result due to drug/chemical interference”) and UCSD (concentration of 7-aminoclonazepam 111 ng/mL). Based on UCSD’s quantitative results, the patient’s compliance status was determined as compliant. Supplementary LC-QTOF-MS analysis at UCSD identified several analytes, including sertraline, levetiracetam and quetiapine (listed in the patient chart), providing a list of candidate interferents to subsequently evaluate in future studies. Conclusion Mass spectrometry is one of the most selective detection methods available in the clinical laboratory due to its ability to measure an analyte’s unique characteristics, such as mass-to-charge ratio and fragmentation patterns. However, the wide range of endogenous and exogenous interfering substances found in complex clinical specimens still requires robust chromatography and sample preparation to fully harness the power of tandem mass spectrometry for quantitative clinical analysis.