Recent advances of drugs monitoring in oral fluid and comparison with blood

Eligibility criteria

Only studies written in the English language and published between January 2002 and December 2022 (20 years) were included. To satisfy the primary aim of the present review, the following additional inclusion criteria were applied: (1) studies reporting the detection of drugs in OF and (2) studies reporting the comparison of paired OF and blood specimens for the detection of drugs. All studies which do not satisfy the inclusion criteria were excluded.

Information sources and search strategy

A comprehensive literature research was conducted electronically via PubMed and Scopus bibliographic databases by two independent authors (S.C. and M.B.). A manual evaluation of the reference lists of all selected full-text articles was further conducted to complement the electronic search. The purpose was to identify all the available pertinent information on the detection of drugs of abuse in OF and their correlation with conventional matrices. In all databases, the date of coverage was from January 2002 to December 2022 (20 years). The most recent search was performed on 31/12/2022.

For the electronic search, specific keywords, medical subject headings [MeSH], and other terms not indexed as MeSH were combined to search all relevant studies. As such, publications were screened according to the following search query adapted to each database: (drugs of abuse OR drug abuse OR substance abuse OR abuse* OR psychotropic drugs) AND (detection OR finding OR investigat* OR detect* OR test OR test* OR detect*) AND (saliva OR saliv* OR spit OR spittle saliva OR oral fluid*) as either keywords or MeSH terms. Additional screening of the reference lists of all pertinent articles and of recent literature reviews on the topic was performed to identify further relevant studies.

Selection criteria

Primary screening of the titles and abstracts was performed by adding studies of any level of evidence published in peer-reviewed journals written in the English language. During this step, duplicates, abstracts, conference presentations, editorials, and expert opinions were also removed. Data regarding the type of analyte(s), the type of matrix(ces), the type of sampling, the analytical technique(s) and the timeline of sampling were included.

Recently proposed methods and applications in OF analysis

The general methods for analyzing therapeutic and illicit drugs in different biological fluids are based on a combination of efficient separation procedure with a sensitive detection technique. At present, numerous separation techniques, including high- or ultra-high-performance liquid chromatography (LC or UHPLC) [27], [28], [29], [30], [31], [32], [33], gas chromatography (GC) [34] and capillary electrophoresis (CE) [35, 36], have been employed for their analysis. Various detection methods have also been coupled to separation techniques in order to obtain an accurate and sensitive drugs determination and quantification such as: fluorescence [36, 37], diode array [26], UV adsorption [38, 39], and mass spectrometry (MS) [26], [27], [28], [29], [30], [31], [32], [33], [34]. Immunochemical methods, mainly enzyme-multiplied immunoassay technique (EMIT) and enzyme-linked immunosorbent assay (ELISA) [40], are generally used as screening tools. Positive samples from the screening test are then confirmed by a more specific technique such as GC-MS or LC-MS. According to internationally accepted criteria in the forensic toxicology field, all the data reported in this review were obtained by confirmatory analysis using chromatography coupled to mass spectrometry (GC-MS or LC-MS).

Cannabinoids

Among the substances of abuse, cannabinoids are the first for consumption both in Europe and United States, according to World Health Organization (WHO). This finding is ever increasing with the legalization in several states and after its introduction in clinics for medical purposes.

With regard to cannabinoids, the main analytes that have been researched in the cited works are: Δ9-Tetrahydrocannabinol (THC), 11-Nor-9-carboxy-Δ9-tetrahydrocannabinol (THC-COOH), 11-Hydroxy-Δ9-tetrahydrocannabinol (11-OH-THC), Tetrahydrocannabivarin (THCV), Cannabidiol (CBD), Cannabinol (CBN), Cannabigerol (CBG). In all manuscripts, a known dose of THC was administered via different routes such as: smoked, vaporized or eaten through brownies.

In particular, Spindle et al. [40] studied how concentrations of THC and THCCOOH vary, within biological matrices, depending on the method of administration (smoking, vaping) in 17 healthy adults who were infrequent users of cannabis. Cannabinoids were quantified by enzyme-linked immunosorbent assay (ELISA) and LC–MS/MS and OF was collected via expectoration. The highest concentrations of cannabinoids were detected both in OF and blood ten minutes after drug administration. THC was measurable in OF for a longer time compared to blood, whereas THC-COOH was fluctuating in OF. The higher detection sensitivity is obtained in vaporized settings for blood, and in smoked sessions for OF.

The same route of OF collection (expectoration) was used by Vandrey et al. [45]. They decided to administer three different THC doses (10, 25 and 50 mg) through brownies containing. Compared to other studies in which THC was inhalated, quantitative levels of cannabinoids in whole blood and OF were lower.

In two studies, Fabritius et al. [46] and Newmeyer et al. [47] administered the same dose of THC (about 10 mg in the first study and about 20 mg in the second one) to heavy and occasional smokers by inhalation and per os (cookies), respectively. Both studies aimed to observe how the concentrations of THC and its metabolites varied between the two groups and between OF and blood. Overall, these authors demonstrated the presence of a peak of concentrations in OF, followed by a rapid deceasing. THC detected in OF was higher (>300 μg/L) compared to blood (<100 μg/L) mainly due to oral exposure. Moreover, the studies highlighted discrepancies between regular and occasional smokers, especially for THCCOOH levels in both blood and OF. Differences between OF and blood THC and THC-COOH concentrations could be related to the used OF collection device, which was different in the two cited papers, in particular Newmeyer [47] evaluated the performance of Draeger Drug Test 5,000 (5 μg/L cutoff) and Alere DDS-2 (25 μg/L THC cutoff), while Fabritius [46] chose a Quantisal device for heavy smokers and Salivette for occasional ones. Unfortunately, similarly to other stimulants a possible limitation in cannabinoids testing in OF may be due to dry mouth after use, thus making difficult to collect sufficient samples for the evaluation.

The Quantisal device for OF collection was also used by Hubbard et al. [48], Odell [49] and Wille [50]. In the paper by Odell [49], twenty-one dependent cannabis users were recruited, allowing to provide once-daily blood, urine and OF samples for seven consecutive days after admission, involving abstinence from all cannabis use. In some subjects THC was detectable in blood for seven days, OF specimens were positive up to 78 h whereas in urine the THC metabolite (THC-COOH) exceeded 129 h. Hubbard [48] and Wille [50] administered different THC doses by smoking (inhalation), finding higher THC levels in OF than in blood possibly due to the OF collection device. Roadside testing was performed by Rohrich [51] and Laloup [52] as well, and OF was collected by two different devices, RapidStat and Intercept, respectively. Rohrich [51] pointed out that THC concentrations in OF were higher by using RapidSTat compared to those obtained through expectoration, underlying the crucial role of the collection procedure.

Opiates/opioids

Opiates and opioids that have been considered within the cited works are the following: oxycodone, noroxycodone, hydrocodone, norhydrocodone, codeine, norcodeine, morphine, tramadol, O-desmethyltramadol (ODMT), buprenorphine and methadone. Mostly, a known dose of exogenous was administered (oral or sublingual administration); then OF and blood samples were taken after different time intervals in order to study their pharmacokinetics and compared the two matrices. Codeine is currently used in cough relief and mild-moderate pain medication. Given several issues emerged in the compliance with the analgesic prescription, the monitoring of its concentrations is essential in some cases. Therefore, researchers have focused their attention on the identification of reproducible quantitative analysis useful to leverage various matrices. After the administration of 19.5 mg codeine phosphate in 12 healthy volunteers, Coucke and colleagues [53] examined its OF/plasma ratio and the variability in drug concentrations in OF, by using two different approaches, Saliva CollectionSystem (SCS) and Quantisal. Next, codeine levels were measured by GC-MS in the two different matrices. At 1 h, by using the two different types of collection system, the amount of codeine identified was higher in OF than blood (ratio OF/plasma ∼2), carrying a significant Pearson’s correlation coefficient between the two matrices. Both sampling procedures resulted efficient in codeine determination, however SCS may mirror better plasmatic fluctuations. The correlation elucidated between OF and plasma codeine levels is better than that identified in less controlled studies, in which the determinations are performed during epidemiological roadside surveys or in subjects who undergo treatments for addiction [50, 54]. For instance, Guinan et al. [54] tested the presence of methadone in OF and plasma of 13 adult subjects, under treatment for opioid dependence. Methadone is an opioid substance greatly diffused as replacement in the medication of heroine abusers, for its low excretion properties. The authors compared LC-MS and pSi SALDI-MS, establishing a good correlation between these two analytical techniques, testing either water or biological fluids and an excellent agreement among the two matrices.

Cocaine

Cocaine, a psychoactive drug obtained from coca leaf, is one of the most consumed illicit substances, after cannabis, according to United Nations Office on Drug and Crime. Only few studies have investigated the impact of cocaine-controlled administration in humans, and the majority of researches has been conducted in polydrug users [55], [56], [57]. Among them, Scheidweiler and colleagues studied the pharmacokinetics of cocaine and its metabolites in OF and plasma, after drug subcutaneous administration in 19 participants [58]. Levels of cocaine, and its metabolites produced by liver (benzoylecgonine (BE), and ecgonine methyl-ester (EME)) were assessed by GC-MS, demonstrating that cocaine was detectable in OF between 0.08 and 0.32 h, but it disappeared rapidly (1.1–3.8 h). Conversely, its metabolites have shown a longer half-life. These authors demonstrated for the first time that all analytic species assessed in OF displayed a good correlation profile with plasma levels. Other studies confirmed this data, showing that EME predominates the OF after repeated oral cocaine doses [58]. Hence, measuring metabolites in OF offers the possibility to extend the detectability window of cocaine.

Chantada-Vazquez et al. [57] compared cocaine quantification in polydrug users within OF and serum obtained by two different analytical techniques, LC-MS/MS and molecularly imprinted polymer – Mn-doped ZnS quantum dot (MIP-QD). The latter exploited the changes in luminescent properties of quantum dots (QD) nanoparticles to sense and quantify cocaine in OF. Indeed, when the analyte (cocaine) is present, the luminescence of QD on the surface of MIPs is quenched, to the point that these luminescence alterations are directly ascribable to analyte concentration. These authors reported that the alternative assessment of cocaine in both types of clinical samples by using the MIP-QD gained in versatility and reduced the costs of laboratory instrumentation, compared to conventional LC-MS/MS.

Amphetamines

Several studies analyze amphetamine, d,l,-methamphetamine, l-methamphetamine, 3,4-methylene-dioxymethamphetamine (MDMA), 3,4-methylenedioxyamphetamine (MDA) and 4-fluoroamphetamine (4-FA), mainly by GC-MS and LC-MS/MS methods. In almost all articles, a known dose of exogenous was administered (ingestion or inhalation) and subsequently quantified within the biological matrices at different times after administration.

Newmeyer and collaborators focused on the OF as alternative matrix for Methamphetamine testing, given its high abuse potential. Methamphetamine exists in two enantiomeric forms (destro-d and levo-l) and the d-methamphetamine isomer is the most powerful in stimulating the central nervous system, while the l-methamphetamine is contained in nasal decongestants. Thus, for more efficacious tests, the two enantiomers should be distinguished, to do not overinterpret the results. In this study, the authors administered 7 doses of the l-methamphetamine, through Vicks^® VapoInhaler™ (2 inhalations every 2 h) to healthy adults (n=16). Then, OF and plasma specimens were collected with two different devices (Quantisal™ and Oral-Eze^®), before and up to 32 h after the first dose. D and l-methamphetamine, as well as d and l-amphetamine were quantified by LC-MS/MS, applying a chiral derivatization. Positive OF samples to l-methamphetamine were produced by all participants after multiple doses, while no d-methamphetamine or d-amphetamine was detected. Conversely, only two patients were positive for l-methamphetamine in plasma samples, suggesting a huge inter and intra-individual variability in OF/plasma ratios. Notably, the use of the correct methodological approach which includes chiral analysis is essential to differentiate l-methamphetamine from the illicit one. Other studies addressed the assessment of amphetamine, d or l-methamphetamine, MDMA, MDA and 4-FA mainly by using mass spectrometry [43, 44, 51, 59, 60], with the pursuit to parallel OF and blood pharmaco-distribution. However, the contamination of the oral cavity during drug consume should be taken into account, explaining the great discrepancies between the two matrixes in the first time points of observation. Indeed, while in OF the substances can be detected in high concentrations during the first 3 h, probably because of oral contamination, in serum the maximal levels appeared after few hours.

Benzodiazepines

Among the Benzodiazepines, oxazepam is the most frequently identified in blood of drivers [61]. Since the sampling of the OF represents a valid tool for roadside testing, Smink and colleagues [62] performed pharmacokinetic studies to examine the correlation between the concentrations of oxazepam and its metabolite (oxazepam glucuronide), in OF and blood (whole blood and serum), after the oral administration of a single known dose (of 15 or 30 mg) in eight male healthy subjects. Samples were collected until 8.5 h after oral administration and quantifications were obtained by LC-MS/MS. As expected, similar concentration–time profiles have been identified in whole blood and serum. Accordingly, in OF both analytes have been detected, albeit at low levels, at least for 8.5 h. In details, oxazepam is detectable in the OF in dose-dependent manner and its concentration reflects its systemic availability. Thus, OF may represent a good methodological approach to detect recent ingestion of the drug in drivers, although further studies are required to better understand whether the saliva composition may impact on analyte concentrations and to estimate the real predictive value of OF.

Barbiturates

Currently, the abuse of barbiturates has been largely replaced with that of benzodiazepines and only few barbiturates remain in clinical setting as anesthetic, analgesics or anticonvulsant agents. However, given their low safety profile and the high risk of dependence, they remain a health concern in clinical and non-clinical settings.

In this context, the use of OF in monitoring barbiturates consumption is gained in popularity, since this means guarantee the possibility to exploit simple, non-invasive and supervised collection procedures. In a single center clinical study, Fritch and colleagues [33] have investigated the clinical effectiveness and the bioavailability of different barbiturates, among which Butalbital, Phenobarbital and Secobarbital, both in the OF and plasma. After the oral administration of a low dose of one of these three barbiturates (n=15 healthy individuals/group), OF and blood samples have been collected for the three consecutive days after the assumption at different timepoints and the quantitative analysis has been performed by LC-MS/MS and GC- MS/MS in OF and blood, respectively. Their intestinal absorption was fast and their effects on central nervous system appear quickly with only mild side effects recorded.

Butalbital and Phenobarbital were the most rapidly detected in both OF and plasma in the first 15 min and remained detectable until 52 h, while Secobarbital was dosable within 30–60 min after the administration. The mean concentrations measured in OF and plasma, are schematically listed in Table 1. Overall, the OF concentrations of the three barbiturates were higher in plasma compared to OF, although the timing of bioavailability was similar between these two biological fluids and their Cohen’s kappa coefficient of agreement was 0.876. Therefore, the authors claimed that the OF may be useful to predict barbiturates plasma concentrations, since their similar pattern of detection.