Newly Emerging Drugs of Abuse and Their Detection Methods

An ACLPS Critical Review

Li Liu, MD, PhD; Sarah E. Wheeler, PhD; Raman Venkataramanan, PhD; Jacqueline A. Rymer, MT(ASCP); Anthony F. Pizon, MD; Michael J. Lynch, MD; Kenichi Tamama, MD, PhD


Am J Clin Pathol. 2018;149(2):105-116. 

In This Article

Overview of Urine Drug Testing

Drug Testing Methods

Urine is the preferred biological specimen for drug screening because drugs are more concentrated in urine than plasma, which prolongs the drug detection window. Other biological specimens, including blood, saliva, and sweat, can also be used for drug testing.[35] For most clinical and forensic applications, initial testing is conducted with immunoassay panels specific for classes of drugs with similar structures. They are qualitative or semiquantitative tests to evaluate the presence or absence of a substance based on a preestablished cut-off. Definitive identification of a specific drug and/or its metabolite(s) requires more sophisticated tests with mass spectrometry (MS), coupled with either gas or liquid chromatography (GC and LC, respectively).

Immunoassays use antibodies developed to react with epitopes in the target compounds to detect drugs and/or their metabolites. The major advantages of immunoassays include fast turnaround time, simplicity of the assay procedures, wide availability of immunoassay platform (either analyzers and/or point of care), and ability to detect multiple drugs within the same class, whereas the major disadvantage is their limited specificity and sensitivity potentially leading to false positive or negative results. That is why the positive results of antibody-based drug screens are considered "presumptive" or "unconfirmed" positive until the test results are confirmed by more specific MS-based assays. Likewise, the absence of a positive test does not definitively eliminate the potential presence of a drug in the same class or with similar pharmacologic activity.

The classes of drugs commonly covered in the urine drug screening by immunoassay include amphetamines, barbiturates, benzodiazepines, buprenorphine, cannabinoids (THC), cocaine metabolite, methadone, opiates, oxycodone, and phencyclidine assays.[36–38] An immunoassay kit specific to 6-monoacetylmorphine (heroin-specific metabolite) (Immunalysis, Pomona, CA) has been approved recently by the US Food and Drug Administration (FDA) for clinical use. This kit detects heroin use without cross-reactivity to morphine, morphine metabolites, and many common analgesics. These assays are available as kits that can be applied on different automated analyzers. Generally, most of these immunoassays are offered as one immunoassay panel.

GC-MS or LC-MS(-MS)-based assays are analytical techniques regarded as more definitive in identifying specific drugs. In these methods, the mixture of compounds within the specimen is separated first by chromatography and then further interrogated by MS. The direct coupling of MS with GC was first developed in the 1970s, dramatically improving both sensitivity and specificity of the analysis of the mixture of compounds.[39]

GC-MS has long been used as a gold standard method for toxicology testing.[39] As the name suggests, the gaseous phase chromatographic separation takes place in a heated oven. Thus, the analytes must be small and nonpolar in order to be thermostable and volatile. That means any compounds that are nonvolatile and/or unstable at high temperatures cannot be analyzed easily by GC-MS without modification. To overcome this limitation, chemical modification with derivatizing agents such as pentafluoropropionic anhydride (PFPA) are required to mask the polar groups, thereby improving volatility. The sample preparation is, therefore, laborious with multiple steps (extraction, derivatization, clean-up, etc) before running GC-MS-based assays.

LC-MS emerged later as an alternative analytical method for drug screening. LC-MS has the advantage of eliminating the requirement for volatility, thus simplifying the sample preparation as well as improving sensitivity for larger and nonvolatile molecules. LC-MS is now commonly equipped with two quadrupole detectors in tandem (LC-MS-MS). The first detector generates precursor (or parent) ions that are in turn selectively allowed to enter the second detector, where further fragmentation occurs and product (or daughter) ions are produced. Generally, LC-MS-MS has higher analytical specificity than GC-MS. With its higher sensitivity and shorter sample preparation process, LC-MS-MS has been used in place of GC-MS in drug screening.[40]

MS-based drug screening can be classified into untargeted or targeted screening. The untargeted drug screening uses the full scan analysis in which the entire mass spectra, including both unfragmented and major fragmented ions, are scanned. Unknown analytes in the specimen are identified by their retention times (comparison of the observed retention time with the ones previously recorded of the known compounds) in the total ion chromatogram and mass spectra (software-assisted library matching of mass spectra of the unknown analytes with preestablished reference mass spectra of the known analytes.) (Please review Figure 5 for the detailed information of these steps.) It can potentially detect any compounds, as long as their mass spectra are available. This method is especially suited for detection of infrequent or newly emerging drugs of abuse, although compounds might be missed at low concentrations due to reduced sensitivity. With targeted drug screening on the other hand, a selected ion monitoring (in GC-MS) or selected reaction monitoring (in LC-MS-MS) mode is used to monitor only preselected ions (or their ion transition) to detect only preselected compounds of interest. The targeted method attains better sensitivity than the untargeted method and is suitable for the detection of frequently abused drugs and monitoring of prescription compliance.[40,41]

Figure 5.

Total ion chromatogram (TIC) (A) and software-assisted library matching of mass spectra of the unknown compounds eluted at 17.52 min and 25.78 min (B-E) in the specimen of Case 2. The urine specimen underwent liquid-liquid extraction with activated charcoal. The extracts were dissolved in methanol and injected into gas chromatography–mass spectrometry (Agilent Technologies 5973 mass spectrometer, Santa Clara, CA) operated in full scan using electron ionization. The mass spectra of the unknown peaks at 17.52 min (shown in the inlet) and 25.78 min in the TIC (A) were shown in (B) and (D), respectively.
These mass spectra were identified as methylnorfentanyl, a metabolite of 3-methylfentanyl and fentanyl through software-assisted library matching of mass spectra with the preestablished reference mass spectra of methylnorfentanyl (C) and fentanyl (E) in the Mass Spectra of Designer Drugs 2012 (Wiley). The limit of detection of fentanyl analogs spiked in the blank urine is around 100 ng/mL, making this detection system adequate for overdose cases of fentanyl and its analogues. Note that the retention times (RT) of methylnorfentanyl and fentanyl are predicted to be 17.5 min and 26.5 min, respectively, comparable to the actual retention times of these unknown peaks (17.52 min and 25.78 min) in TIC (A).

The sensitivity of MS-based drug testing is also influenced by various factors, including sample preparation method (eg, liquid extraction and solid phase extraction), type and size of chromatography columns, and parameter setting (eg, voltage and frequency) and specifics (eg, accuracy, resolution, and scanning speed) of the MS instrument.[41,42] Consequently, the sensitivity of drug testing might be different in each laboratory, even if they utilize a similar methodology.

Challenges for Clinical Toxicology Laboratories

The ongoing emergence of designer drugs is a great challenge for toxicology laboratories in several ways. First, limited information about the chemical structure of these newly emerging drugs makes laboratory detection difficult. The new compounds and their metabolites usually do not cross react with immunoassays that target the existing classes of drugs of abuse. Availability of mass spectrum of the compound is a prerequisite for the detection by MS-based screening assays. Second, a lack of information about the metabolism and pharmacokinetics of these compounds complicates their detection in urine. Third, illicit drugs are often "rebranded" in the underground market (eg, bath salts circulate as "Molly," and fentanyl analogs are sold as "heroin"), making the clinical histories unreliable and the targeted drug screening less useful. Untargeted drug screening is required in these situations but is limited to the existence of a library match for the emerging drugs.

Laboratory Tests for New Emerging Drugs

Immunoassays. Synthetic cathinones have some structural similarities with amphetamine (Figure 1), which could cause cross-reactivity in some of the commercially available amphetamine immunoassay kits such as CEDIA Amphetamine/Ecstasy Drugs of Abuse Assays (ThermoFisher Scientific, Waltham, MA).[43] MDPV has also been reported to cross react on SYNCHRON System(s) Phencyclidine Drugs of Abuse Testing (Beckman Coulter, Brea, CA).[44] The following synthetic cathinones (Ethylone, 3-FMC, 4-FMC, MDP, methedrone, methylone, pyrovalerone, and α-PVP) do not cross-react with Syva EMIT II Plus Amphetamine Assay (Siemens, Munich, Germany) up to 5 μg/mL.[45]

Synthetic cannabinoids are not expected to cross react with THC immunoassays due to structural differences (Figure 2). At present, immunoassays have been developed by several manufactures for rapid detection of some designer drugs. Neogen Corporation (Lexington, KY) launched enzyme-linked immunosorbent assays (ELISA) for synthetic cathinones (bath salts) and synthetic cannabinoids (Spice or K2). Randox Toxicology (Crumlin, UK) offers several (ELISA) kits for synthetic cannabinoids, synthetic cathinones, and mitragynine. Immunalysis (Pomona, CA) has developed three distinct homogeneous enzyme immunoassay K2 Spice kits for the detection of synthetic cannabinoids. But none of these immunoassay kits are approved by the FDA for clinical use as of this writing.

GC-MS Based Assays.Bath Salts: Previous publications have reported GC-MS identification of synthetic cathinones in biological samples, including MDPV and α-PVP with different detection limits.[46,47] Consistent with these studies, we have also identified a series of synthetic cathinones (Ethylone, ethylpentylone, MDPV, methedrone, methylone, and pentylone) in clinical specimens using GC-MS–based untargeted comprehensive drug screening after liquid-liquid extraction, as exemplified by case 1.

Synthetic cathinones are extractable from urine specimen by liquid-liquid extraction with organic solvents and detectable at 500 ng/mL by GC-MS even without derivatization, but PFPA-based derivatization further improves the limit of detection to 50 ng/mL for the synthetic cathinones with secondary amine (Methylone, methedrone, ethylone, 3-FMC, and 4-FMC).[45]

Spice: Contrary to synthetic cathinones, synthetic cannabinoids are more difficult to detect in clinical specimens using GC-MS. One reason is their rapid and extensive metabolism. For example, JWH-073 and JWH-018, the prototypal synthetic cannabinoids, are quickly metabolized to monohydroxylated or carboxyl metabolites, and the monohydroxylated metabolites are further glucuronidated before urinary excretion, and these compounds are not excreted in a parental form in urine.[48–50] These polar and hydrophilic metabolites, even after glucuronidase treatment, are not only more difficult to extract by liquid-liquid extraction, but also to analyze by GC-MS than parental compounds. Their relatively large molecular size is another factor that makes these compounds less compatible with GC-MS. AB-CHMINACA, a newer synthetic cannabinoid with different chemical structure, should be even more difficult to analyze by GC-MS than JWH-073 and JWH-018, presumably because of the presence of more polar groups within the molecule. Due to structural similarity, other newer synthetic cannabinoids such as XLR-11 and AKB48 (APINACA) are expected to be as difficult as AB-CHMINACA to detect by GC-MS (Figure 2). For these reasons, LC-MS(-MS) is preferable for detection of synthetic cannabinoids.[41,51]

Fentanyl Analogs: The FDA has recently cleared the Immunalysis SEFRIA fentanyl urine immunoassay for qualitative determination of fentanyl in human urine at a cutoff of 1 ng/mL. Fentanyl analogs, acetyl fentanyl and butyryl fentanyl, can be detected in this assay with 100% cross-reactivity. This addition will greatly facilitate the detection of fentanyl and its analogs in the clinical specimens. Fentanyl, as well as multiple fentanyl analogs (acetyl fentanyl, 3-methylfentanyl, butyryl fentanyl, butanoyl-4-fluorofentanyl, and para-fluoroisobutyrylfentanyl) (Figure 3), are detectable by GC-MS–based untargeted comprehensive drug screening, as shown in case 2 (Figure 5). Many of these specimens were collected from patients of self-reported "heroin" overdose cases, but no opiates were detected (see case 2).

Other Synthoid Opioids: Literature indicates that the synthetic opioids U-47700 and AH-7921 are detectable from urine specimen by GC-MS.[29,52] Consistently, U-47700 has been detected in urine specimens from multiple patients with GC-MS–based untargeted comprehensive drug screening in our laboratory (data not shown).

Mitragynine: Mitragynine and its metabolites have been detected in human urine specimens using GC-MS with solid phase extraction and derivatization, with a limit of detection of 100 ng/mL.[53] Chemical derivatization is not an absolute requirement for mitragynine detection by GC-MS.

LC-MS-MS and LC-High Resolution-MS Based Assay. GC-MS has been the gold standard for toxicology testing, but GC-MS has a limited utility for detection of polar compounds. This could be problematic for toxicology testing, because drugs are often metabolized in the liver and become more polar before being excreted in urine. Sample preparation such as glucuronidase treatment or derivatization is often needed for GC-MS-based analysis, but detection gains from this processing are often not enough to detect compounds of interest (see the discussion about Spice).

In contrast, LC-MS (-MS) is able to analyze polar compounds with lower limits of detection. This simplifies the sample preparatory procedures for LC-MS-MS, minimizing the burden on technical staff and reducing the turnaround time. LC-MS-MS-based assays have been developed for newly emerging drugs.[40,41] Indeed, LC-MS(-MS)-based testing has been reported for synthetic cathionones,[54,55] synthetic cannabinoids,[49,51] fentanyl analogs,[56,57] U-47700,[56,58] and mitragynine.[59]

LC-MS-MS still requires mass spectral libraries for compound identification; that means newly emerging drugs without mass spectral information cannot be identified by LC-MS-MS. The advent of LC-high-resolution (HR)-MS has provided a solution for this problem. Both LC-time of flight-MS or LC-orbitrap MS are able to resolve molecular mass to 0.001 atomic mass units, compared to the 1 atomic mass unit for conventional MS. This allows for tentative identification of unknown compounds by deducing the molecular formula from accurate mass databases without using mass spectral libraries.[41] This is a very powerful system for toxicology laboratories to detect newly emerging drugs, but the cost of LC-HR-MS system is a major hindrance for standard clinical laboratories, reducing its utility to a specialty instrument held at only a handful of reference laboratories.