Immunochromatography screening devices for cannabinoids in oral fluid sample

Abstract Cannabis sativa L. is one of the most consumed drugs in the world and recent studies have associated its use with an increase in the number of traffic accidents in different countries. In many countries, like Brazil, simple and reliable methodologies are still needed for the detection of drugs on site, mainly cannabinoids, considering its prevalence of use and oral fluid (OF) has been proved as an appropriate biological matrix for this purpose. Considering that, this work aims to review previous studies on immunochromatographic devices for on-site detection of cannabinoids in OF, discussing their sensitivity, specificity, cut-offs values and confirmatory methods. This data shows the importance of choosing a screening device and it reinforces the need for its implementation in Brazil. The research was conducted on 5 databases and all original articles, published in the last 10 years, were selected. A total of 32 articles were found, providing data for 17 screening devices of distinct brands. Only 2 screening devices showed satisfactory sensitivity and specificity in the evaluated studies (≥80% and ≥90% respectively). However, it should be considered that the screening devices still have some limitations, such as a higher cut-off than those recommended by international guidelines (cut-off > 2 ng/mL), therefore demonstrating the need for more studies in the area and the importance of confirmatory analysis usually fulfilled by LC-MS/MS, GC-MS/MS or GC-MS. Thus, the screening analyzes should not be evaluated by itself, but in association with confirmatory results and observational traits (behavioral changes), for a better understanding of the traffic scenario.


INTRODUCTION
Traffic accidents (TA) are considered a major public health problem, at the expenses of approximately 3% of the national Gross Domestic Product (GDP) in most countries (WHO, 2020;WHO, 2018).Among the factors that contribute to the statistics are fatigue, speeding and the consumption of alcohol and/or drugs abuse, followed by negligence of safety devices (Leyton et al., 2012;WHO, 2018).The legal framework that regulates the use of psychoactive substances (PSs) varies according to the social, legal and economic characteristics of each country, so it is not clear what are the acceptable limits in biological samples from motor vehicle drivers (Herrera-Gómez et al., 2018).In Brazil, driving under the influence of alcohol and other PSs is not allowed (Brasil, 2006); however, the legislation does not clarify which are the PSs and does not recommend ways of monitoring them as well as it does not establish acceptable limits -or levels -of PSs in a biological sample (Pechansky et al., 2019;Saldanha et al., 2014).

Immunochromatography screening devices for cannabinoids in oral fluid sample
Page 2/17 Braz.J. Pharm.Sci.2023;59: e20555 Jonathaline Apollo Duarte, Roberta Petry Gorziza, Marina González, João Marcelo Astolfi Picanço, Renata Pereira Limberger (THC) is the mainly PS present in cannabis samples and is often detected in biological samples from drivers approached in police roadblocks and in drivers involved in TA with minor injuries and/or fatal victims (Lee, Huestis, 2014;Volkow et al., 2014).Cannabis exposure affects the ability to drive by promoting changes in perception and in the reaction time, altering the state of attention and compromising the motor skills of individuals (Hartman, Huestis, 2013).Considering the Brazilian traffic scenario and the constant discussion about the legalization and/or decriminalization of cannabis (CEE -FIOCRUZ, 2016;FIOCRUZ, 2013;Moreira et al., 2016), it becomes evident the need for simple and reliable methodologies that allow its detection in biological samples of vehicle drivers, as well as the implementation of more specific laws (Pelição et al., 2016;Saldanha et al., 2014).Among the biological matrices used for screening analysis of PSs, oral fluid (OF) stands out as a biological sample that has been achieving more space in routine analyzes in Toxicology, specially related to DUID (Driving Under the Influence of Drugs) scenarios.Due to its numerous advantages, OF has shown to be a promising matrix for on-site testing (Gentili et al., 2016).OF collection is easy and non-invasive, it can be performed under police supervision, without embarrassment to the driver, and its analysis provides a good correlation with blood, considering the recent use of several classes of PSs (Fiorentin et al., 2017;Gentili et al., 2016;Logan, Mohr, Talpins, 2014).Furthermore, OF can also be used for subsequent confirmatory analyzes, thus samples for both screening and confirmation can be collected at the same time, increasing the chances of consistent results.However, OF collection can be affected by some factors such as decreased salivary flow and dry mouth, attributed to lack of hydration or drug use (Logan, Mohr, Talpins, 2014).
The most common on-site screening devices chosen by several countries for the detection of cannabis, and other PSs, are usually immunochromatography tests.In general, these assays consist of collection pads attached to porous membrane strips, which are inserted into the donor's mouth.From the swab (pad), the OF migrates by capillarity, mobilizing the reservoir of colored antibodies that flow with the OF along the strip until the lines with the immobilized PSs are reached.In the presence of a positive sample for any PS in the cut-off concentration of the multi-drug device or above it, the binding sites of the respective colored antibody saturate and do not bind to the drug immobilized on the band (Souza et al., 2012).However, it is important to note that even with efficient screening devices for on-site use, the need for confirmatory analysis of the suspected drivers OF sample is not ruled out, considering that some devices still demonstrate a lack of specificity (≥ 90%) and of sensitivity (≥ 80%) for Δ 9 -THC (Blencowe et al., 2011;Musshoff et al., 2014;Strano-Rossi et al., 2012).International guidelines suggest a cut-off value of 2 ng/mL for confirmatory tests, while they recommend the value of 4 ng/mL for screening tests (Department of Health and Human Services, 2019;Walsh et al., 2008).
Considering that cannabis is among the main PSs related to TA in Brazil (DeBoni et al., 2014;Pelição et al., 2016;Saldanha et al., 2014), and that there is a need for a screening tool implementation in this country, this article aims to review previous studies regarding immunochromatographic devices for cannabinoids detection, focusing on its main advantages and disadvantages.

MATERIAL AND METHODS
Our research strategy involved a comprehensive systematic review in the following databases: PubMed, Google Scholar, Science Direct, Scopus and Science.gov.The descriptor used for databases was "cannabis or cannabinoids and oral fluid detection".The search was conducted in February 2020.Four independent researchers conducted the review, using the following inclusion criteria: a) original studies for immunochromatographic in loco devices, for cannabis detection in oral fluid; b) articles published in the last ten years (2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020), in order to present a recent panel of studies.Studies that do not include oral fluid as a biological matrix were excluded, as well as different types of screening tests.These studies could either include drivers or not, as long as it evaluated immunochromatographic screening test for cannabinoids in oral fluid.Figure 1 shows the results for the systematic review.The initial search retrieved 92 non-repeated articles from the total (Google Scholar -18.000;  and, after a detailed analysis, 32 papers met the inclusion criteria.Other studies that include guidelines and laws were only included in the discussion.

RESULTS AND DISCUSSION
Among the 32 selected articles, some studies used a device to collect the sample and other device to analyze it, while others use the same for both collection and analysis.The screening devices are multi-drug detection tests; however, only the detection of cannabinoids was approached in this review.All 32 studies evaluated the cannabinoid THC, the major psychoactive cannabis component, for drug detection, but Anzillotti et al.,(2014) and Newmeyer et al.,(2017) also included more cannabinoids, like cannabidiol (CBD), cannabinol (CBN), 11-Nor-9-carboxy-Δ 9 -tetrahydrocannabinol (11-THCOOH), Tetrahydrocannabivarin (THCV) and cannabigerol (CBG).Table I compiles the review results, considering the detected cannabinoids, the analyzed devices, and the confirmatory methods.Also, the cut-offs and the tests parameters, such as specificity and sensibility, are shown.
Sixteen different devices, from distinct manufacturers, were evaluated in these studies.However, some of their brands have different presentations of drug tests.For example, Dräger ® brand has four test devices (Drug Test 5000, DCH 5000, DCD 5000, 5000 STK) and DrugWipe ® brand have six types of devices (5/5+, 5A, 5, 5+, II Twin, 6S).DrägerDrugTest ® 5000 is the most used device in the studies -also presenting the lowest cut-off (5 ng/mL) -, followed by DrugWipe ® , with a cut-off at 30 ng/mL, and Alere DDS2, with a 25 ng/mL cut-off.For confirmatory methods in OF, Walsh et al. (2008) recommends a cut-off of 2 ng/mL for the analysis of cannabinoids, but they do not mention any screening devices (Walsh et al., 2008).Thus, from all studied brands, Dräger DrugTest ® 5000 presents the cut-off (5 ng/mL) which is closer to the limit recommended by the current guidelines (2 ng/mL).
Based on the review, the only two screening devices that showed good sensitivity and specificity were Dräger DrugTest ® 5000 (DDT5000) and Alere TM DDS ® 2 (DDS2).
For Dräger DrugTest ® 5000, sensitivity and specificity ranged from 76-95% and 71-99.3%,respectively, in the studies.As for Alere TM DDS ® 2 screening device, the sensitivity varied from 75-100% and the specificity ranged from 80-100%.However, these two devices present a higher cut-off than the recommended the aforementioned guides, and the DDT5000 has the closest value (5 ng/mL), followed by DDS2 (25 ng/mL).
Even after a decade of action for road safety, which it has started in 2011, the American region continues to record approximately 155.000 traffic deaths per year (ONU, 2019a).Brazil had an estimated traffic mortality rate of 19.7 per 100.000inhabitants, which is higher than the rate in the entire South America continent (15.6) and the average rate in the Southern Cone (18.4) (ONU, 2019b).Although Brazil has managed to reduce the number of deaths in traffic, the country still registers a high number of victims with minor and serious injuries, and this is relative to different factors, including the use of alcohol and/or drugs.Brazilian Law 11.343/2006 (Brasil, 2006) prohibits drug use, its cultivation and/or distribution, and also differentiates drug use and drug trafficking, with a social approach.Likewise, it defines the severity of the crime and its penalty, but it does not specify a law on driving under drug influence.Brazilian Traffic Code (Brasil, 1997) considers forbidden the use of any PS while driving; however, it does not specify a drug concentration, as it does for alcohol use (≥6 dg/L of ethanol in blood or ≥0.3 mg/L of ethanol in alveolar air), and it does not define ways of monitoring.There is no specific regulation in Brazil for testing devices for drug detection in drivers or a regulated device registered in the national health surveillance agency.
While in other countries drug screening is already performed on roadsides (EMCDDA, 2012), in Brazil there are still no screening tests available for routine monitoring in DUID scenarios, to help prevent TA.Since Cannabis sativa L. is the second most consumed drug in Brazil and in the world, along with its relation to TA (Callaghan et al., 2013;Perna et al., 2016), it should be considered a priority for drug screening, besides alcohol.Thus, recently, the Cannabis product cannabidiol (CBD) has been authorized as a controlled substance for therapeutic purposes in Brazil, and any product of Cannabis must not exceed the maximum concentration of 30 mg/mL of Δ 9 -THC and 30 mg/mL of CBD, considering health recommendations (ANVISA, 2019).Since driving under the influence of cannabis is an expressive risk factors that contribute to TA, the legalization of cannabis, even for medical and very punctual purposes, can be a risk on the road, demonstrating the importance of the understanding how sensitive, specific and accurate are the methods available in the market that make it possible to monitor the driver's impairment.
Compared to other biological matrices, OF is an advantageous matrix for on-site analysis.Due to the practicality of collection, donor acceptability and the noninvasive collection method (Anizan et al., 2015), OF has become increasingly popular in drug testing programs, mainly in the investigation of DUID (Newmeyer et al., 2014).Since the collection can be assisted by law enforcement officers without causing embarrassment to the donor, the possibility of adulteration of the matrix is reduced (Lund et al., 2011).
The Society of Forensic Toxicologists (SOFT/AAFS), through the guide "Oral Fluid as a Test Specimen for Roadside Studies: Guidelines for Implementing a Data Collection Program" addresses OF as a biological matrix for monitoring DUID cases, as well as the importance of the commitment of all parties in the implementation and the management of a roadside testing program.This guide also reinforces the advantages of using OF over blood, considering it the "gold standard" for alcohol and drugs detection in traffic -for instance, the possibility of sample collection and analysis onsite, without the need of a health professional,like with blood collection, and with the least possible embarrassment to the driver.Furthermore, the guide provides a brief protocol for the Oral Fluid Program, with recommended steps for the suspected drunk and/or drugged driver approach, as well as for on-site analysis, for sample collection and for sample storage, also pointing when to submit the sample for confirmatory analysis (SOFT, 2014).
It is important to point out here that the screening devices (multi-drug devices) for OF analysis are a tool that would help the police to investigate and control DUID cases in loco on the roads, allowing the monitoring of the use of psychoactive substances in different traffic scenarios (Gjerde et al., 2018).In forensic analyzes focused on the traffic scenario, the use of prohibited substances is considered a crime (HealthNewsReview.org,2020).In order to fit in this scenario, the chosen screening device must have high sensitivity and specificity, since the main objective is to identify drivers under the effect of PS, contributing in the future to compliance with the traffic safety law (Strano-Rossi et al., 2012).The screening device sensitivity is the proportion of true positive OF samples that have been correctly identified (Beirness, Smith, 2017).On the other hand, the screening device specificity is the proportion of true PS negative samples that have been correctly identified (Beirness, Smith, 2017).Both are calculated using the cut-off points adopted by a DRUID study, a fact that reflects the variations observed between studies, and they are calculated according to the equations presented in Table I.In addition, in order to determine the usefulness of a screening device, the true predictive value and the negative predictive value are calculated (Table I) and both are directly dependent on the PS prevalence in the studied population (Blencowe et al., 2010).Thus, the use of a screening device with low sensitivity and/or selectivity can result in an increase of false negative and/or false positive results.In both cases, it leads to wrong tests results and it can have serious legal consequences, considering that the legally controlled substances consumption has a close relationship with crime.False negative results lead to impunity for drivers who pose as a risk in traffic, demonstrating the failure of the evaluation system.As for false positives results, there are risks of condemnation and embarrassment provoked on an innocent driver.Thus, besides sensitivity and specificity, screening devices must be reliable and easy to handle (Strano-Rossi et al., 2012), with an easy and fast interpretation of the results, in order to allow that trained police officers can identify the use of psychoactive substances by motor vehicle drivers (Newmeyer et al., 2017).
Current programs that monitor the use of PS in traffic, such as the European Union's (Driving under the Influence of Drugs, Alcohol and Medicines -DRUID), suggest that OF screening devices should be evaluated for their analytical sensitivity, specificity and efficiency and that these values should be higher than 80% (Newmeyer et al., 2017).Other studies established that screening devices must present higher standards (sensitivity ≥ 80%, specificity ≥ 90% and accuracy ≥ 95%), so that it can be considered a satisfactory test (Strano-Rossi et al., 2012).Based on this information, and the data collected in our review, it is possible to verify that only Dräger DrugTest 5000 ® (DDT5000) and Alere TM DDS ® 2 (DDS2) devices fits the recommendation for the investigation of cannabis metabolites in OF, presenting sensitivity and specificity greater than 80% in a considerable number of studies (Beirness, Smith, 2017; Gjerde et al., 2018;Newmeyer et al., 2017;Rohrig et al., 2018;Strano-Rossi et al., 2012;Swortwood et al., 2017).However, it is important to note that some studies have found low specificity for the DDT5000 device, even though they have found satisfactory sensitivity (Domingo-Salvany et al., 2017;Lema-Atán et al., 2019;Logan, Mohr, Talpins, 2014).As for DDS2 device, Veitenheimer and Wagner (2017) have found a 100% of specificity, while the sensitivity was below the recommended values (75%).The Rapid STAT ® device showed 71-91% of sensitivity and specificity values ranging from 09-97, thought the studies in this review (Blencowe et al., 2011;Musshoff et al., 2014;Strano-Rossi et al., 2012;Wille et al., 2010).However Röhrich et al.(2010) have found > 80% values for both sensitivity and specificity for this device.
THC is traditionally known to be a problematic analyte for on-site screening devices (Fierro, González-Luque, Álvarez, 2014).DUID studies have shown low sensitivity of THC screening tests, and it may be associated with their higher cut-offs (Domingo-Salvany et al., 2017).Other factors that can contribute for the device's low sensitivity and its inadequate performance are the possibility of cross-reactivity and poor analyte recovery from the device (Mazina et al., 2015).These factors are particularly difficult to elucidate for cannabinoids (Domingo-Salvany et al., 2017) and it raises concerns about the use of these devices in police routine.
However, although the screening devices have some disadvantages that must be considered, they still are an important tool, and they are currently the method of choice for DUID cases in some countries.As an example, in 2015, the Norwegian Mobile Police Service (NMPS) have started DUID cases monitoring of PS, including cannabis metabolites, using the Dräger DrugTest5000 ® (DDT5000) screening device for OF (Gjerde et al., 2018).NMPS reported that DDT5000 did not correctly identified DUID offenders, but the screening device was helpful to assist in the identification of possible DUID suspects (Gjerde et al., 2018).
In a study conducted in Italy, trained police officers randomly approached drivers during road patrols, collecting their OFs for screening analysis using two different devices.At the end of the study, they found that only DDT5000 presented acceptable sensitivity for on-site investigation (Strano-Rossi et al., 2012).
In Spain, the Traffic Police is responsible for conducting on-site OF screening tests for alcohol and drugs (Herrera-Gómez et al., 2018).DDS2 and DDT5000 are among the screening devices chosen by the Spanish law (Herrera-Gómez et al., 2018).The device's performance was investigated in a study conducted by Lema-Atán et al. (2019) and they have found an appropriated sensitivity considering the devices' cutoffs for cannabis (Lema-Atán et al., 2019).In this study, samples were collected by police officers, investigating the use of psychoactive substances in Spanish drivers, between 2013 and 2015.
In Canada, the government have approved a legislation that would allow the use of screening devices for cannabis on-site investigation in suspected drivers (Canadian Centre on Substance Use and Addiction, 2018).According to this document, the screening device can help the police to decide which actions will be taken for suspected drivers at a risk of causing an accident (Canadian Centre on Substance Use and Addiction, 2018) and allows the traffic officer to decree flagrant and remove drugged drivers from the roads.The Dräger DrugTest ® 5000 is one of the devices approved for monitoring DUID cases by the Canadian government, as long as it is used in conjunction with the Dräger DrugTest 5000 ® STK-CA device.Other approved device includes SoToxa™, as long as used in conjunction with Abbott SoToxa™ test cartridge, and Abbott SoToxa™ OF collection device (Canada, 2019).
In the United States, some studies have been developed in partnership with specialist drug recognition officers (DRE) from the Tulsa Police Department (Oklahoma, US) such as the one developed by Veitenheimer and Wagner (2017).The DDS2 device was used in routine approaches of suspected DUID drivers in 2013.At the end of the study, it was found that the DDS2 screening device is a tool that can provide police a greater confidence in detecting drivers who are under the influence of PS such as cannabis (Veitenheimer, Wagner, 2017).
After on-site analysis, any positive result from a screening test must be confirmed by a validated analytical method (Fiorentin et al., 2017).For confirmatory analysis, it is necessary to collect an additional volume of OF, which is usually performed with a specific collection device.There are also many different manufacturers for those devices, such as Quantisal™, Oral-Eze ® , StatSure Saliva Sampler™, and this also require studies to define the better one for cannabinoids recovery, case by case, country by country, scenario by scenario.
For confirmation and quantification of cannabinoids in OF, the most commonly applied analytical tools are gas chromatography coupled to mass spectrometry (GC/ MS or GC-MS/MS) and liquid chromatography coupled to mass spectrometry (LC-MS/MS) (Table I).THC, as the major psychoactive cannabinoid, is also the chosen metabolite for confirmatory testing in OF, as it is for screening tests (Molnar et al., 2014), along with CBD and CBN.When detecting these cannabinoids in OF, it can be generally assumed that there was the recent consumption of cannabis (Vindenes et al., 2011).This information is also relevant, considering that the Brazilian Traffic Code (CTB) regulates the driving under drug influence, while an old drug consumption does not constitute a traffic crime (Baggio, 2017).When develop a confirmatory method for THC in OF, one should consider THC capacity of adherence to plastic (Anzillotti et al., 2014;Molnar, Lewis, Fu, 2013), leading to its poor recovery from collection devices and affecting the recommended cut-off (2 ng/mL) (Walsh et al., 2008).Considering this, special devices with elution/stabilization buffers are generally used to collect OF (Anizan et al., 2015) which should favor the elution of THC with minimal dilution.
Another important factor for clinical and for forensic purposes is the investigation of the analyte stability in the OF, considering its importance in the interpretation of concentrations (Lee et al., 2012) and to guarantee accurate and reliable results (Scheidweiler et al., 2017).Thus, the collection devices and the storage conditions of the sample are variables to be considered.Studies of the stability of cannabinoids in real OF collected by different collection devices are described in the literature.Leeet al.(2012) have found that THC and other cannabinoids (THCCOOH, CBD and CBN) are stable for 4 weeks when stored at 4°C, in the Quantisal™ device.StatSure Saliva Sampler™ and Oral-Eze ® devices also have presented stability for THC, THCCOOH, CBD and CBN in OF, under the same conditions of time and temperature.StatSure TM device also have been able to maintain samples stability within 24 weeks, if stored at -20°C (Anizan et al., 2015).The stability of minority cannabinoids in real OF was investigated by Scheidweiler et al.(2017).THC, THCV, 11-OH-THC, CBD and CBG had satisfactory stability when kept under refrigeration at 4°C during the period of 8 weeks, extracted from OF stored in Quantisal™ devices (Scheidweiler et al., 2017).Therefore, the stability of cannabinoids depends on the collection device, as well as the time and conditions of the storage of the sample, and this aspect is also something to be considered.
In Brazil, the Ministry of Justice and Security organized a work team composed of members of the National Drug Policy Secretariat (SENAD), the Federal Highway Police (PRF) and the National Public Security Secretariat (SENASP), with the purpose of assisting in the implementation of the use of screening devices to monitor other PS in addition to alcohol (MJSP, 2020).In order to be successful in the implementation of screening devices for PS monitoring in traffic, it is extremely important to have solid evidence-based information, considering local contexts (Scherer, 2017).In addition, special attention should be paid to costs for both the implementation of screening devices and the processing of confirmatory analyzes.For example, non-volatile analytes that that require LC-MS/MS analysis present a higher cost when compared to GC/MS analysis (Huestis et al., 2011;SENAD, 2014).Particularly in Brazil, only a few forensic laboratories have a LC-MS/MS instrument available for sample processing (SENAD, 2014).
In conclusion, this systematic review selected 32 articles for analysis of immunochromatographic devices and the analysis of cannabinoids in OF onsite.From all evaluated devices, only two have shown appropriated sensitivity and selectivity, as recommended by international guides.Some limitations that must be considered and improved, such as the devices cut-offs and the possibility of cross reactions, could lead to false positives results.Considering the legal and the emotional impact that false positive results can cause in the lives of drivers who are not under the influence of PS and the potential risk that false negative drivers represent for traffic, the improvement of screening devices by manufacturers is essential.Thus, confirmatory analyzes by GC/MS, GC-MS/MS and/or LC-MS/MS are also needed.However, drug screening tests results can assist the law enforcement in determining the offense of flagrante delicto, preventing possible traffic accidents.Besides the limitations, the implementation of screening tests for cannabinoids on the roadside still can be very helpful to reduce traffic accidents in Brazil, as it has been seen in some developed countries worldwide.

FIGURE 1 -
FIGURE 1 -Strategy used in the research of articles for review.

TABLE II -
Identification of cannabinoids in OF on the surveillance of traffic accidents

TABLE II -
Identification of cannabinoids in OF on the surveillance of traffic accidents

TABLE II -
Identification of cannabinoids in OF on the surveillance of traffic accidents Jonathaline Apollo Duarte, Roberta Petry Gorziza, Marina González, João Marcelo Astolfi Picanço, Renata Pereira Limberger

TABLE II -
Identification of cannabinoids in OF on the surveillance of traffic accidents

TABLE II -
Identification of cannabinoids in OF on the surveillance of traffic accidents