Selectivity and sensitivity of urine fentanyl test strips to detect fentanyl analogues in illicit drugs

Abstract

Background:

Urine fentanyl test strips have been employed to check street drugs for fentanyl and fentanyl analogue contamination, but there is limited evidence for the applicability of fentanyl strips for this purpose. We examined the ability of four commercially-available fentanyl test strips to detect fentanyl and a range of fentanyl analogues currently on the recreational drug market.

Methods:

Four brands of fentanyl test strips (Rapid Response, One Step, Nal van Minden, and Rapid Self Test) were examined using single-component drug solutions containing fentanyl, 28 fentanyl analogues, four non-fentanyl synthetic opioids, or eight traditional drugs of abuse. The effect of co-presence of heroin or ascorbic acid on test results was also examined.

Results:

All test strips detected fentanyl as well as 21-24 of the 28 fentanyl analogues tested. One of the test strips gave false positive results in the presence of ascorbic acid.

Conclusions:

Fentanyl test strips successfully detected the majority of fentanyl analogues tested. Drug solutions for testing should not be overly dilute, since the test results are highly concentration dependent. Fentanyl test strips have utility as a harm reduction tool, but they are no panacea for overdose since certain fentanyl analogues are not detected.

Introduction

Since 2013, there has been a marked increase in the number of overdose deaths caused by highly potent opioids such as fentanyl and its analogues, collectively termed fentanyls (UNODC, 2019a). The number of fentanyls on the illicit drug market is continuously growing, with 34 new substances reported since 2009 (EMCDDA, 2019). Whereas fentanyl is approximately 100 times more potent than morphine, certain analogues have an even higher potency, e.g. carfentanil, which is reported to be 10,000 times more potent. At the present time, fentanyls are used as adulterants in illicit heroin or cocaine and sold as counterfeit prescription medications, exposing unsuspecting users to an increased risk of overdose due to the high potency of the compounds (Prekupec, Mansky, & Baumann, 2017; UNODC, 2019a, 2019b).

Fentanyl test strips, originally developed to detect fentanyl and its major metabolite, norfentanyl, in urine, have been used for testing illicit drug products for the presence of fentanyls. Most fentanyl test strips are based on competitive lateral flow immunoassays on simple paper-based devices and are user-friendly, inexpensive, and give immediate test results (McGowan, Harris, Platt, Hope, & Rhodes, 2018; Sherman, et al., 2019). However, there is a considerable lack of knowledge about the selectivity and sensitivity of urine fentanyl strips when used for drug solution testing.

In the present study we evaluated the ability of four commercially-available brands of urine fentanyl test strips to identify fentanyl and 28 different fentanyl analogues (Supplementary Material; Figure 1) at different drug concentrations. The cross-reactivity to non-fentanyl synthetic opioids and commonly abused drugs was also assessed, as well as the effect of co-presence of heroin and ascorbic acid.

Methods

Chemicals

Fentanyl, 28 fentanyl analogues, four non-fentanyl synthetic opioids, including the opioid maintenance treatment drug methadone, and eight commonly abused drugs were acquired from commercial distributors (see Table 1 and Supplementary Material).

Table 1:

Selectivity and sensitivity of four brands of urine fentanyl strips when testing single-component drug solutions (100, 1000 and 10 000 ng/mL) of fentanyl and 28 fentanyl analogues.

Compound	Rapid Responsea			Rapid Self Testa			One Stepa			Nal Van Mindena
ng/mL	100	1000	10 000	100	1000	10 000	100	1000	10 000	100	1000	10 000
Fentanyl	+	++	N.M.	+	++	N.M.	++	++	N.M.	++	++	N.M.
2-Fluorofentanyl	(+)	++	N.M.	−	++	N.M.	−	−	++	(+)	++	N.M.
3-Fluorofentanyl	−	++	N.M.	−	++	N.M.	++	++	N.M.	−	++	N.M.
4-Fluorofentanyl	(+)	++	N.M.	(+)	++	N.M.	++	++	N.M.	+	++	N.M.
3-Methylcrotonylfentanyl	(+)	++	N.M.	−	++	N.M.	−	(+)	++	(+)	++	N.M.
4-Chloroisobutyrfentanyl	−	++	N.M.	−	+	N.M.	−	+	N.M.	−	++	N.M.
4-Fluorobutyrfentanyl	−	++	N.M.	−	++	N.M.	++	++	N.M.	(+)	++	N.M.
4-Fluorocyclopropylbenzylfentanyl	−	+	N.M.	−	(+)	++	+	++	N.M.	−	+	N.M.
4-Fluoroisobutyrfentanyl	+	++	N.M.	+	++	N.M.	++	++	N.M.	+	++	N.M.
4-Methoxybutyrfentanyl	−	(+)	++	−	(+)	++	−	−	+	−	+	N.M.
Acetylfentanyl	+	++	N.M.	++	++	N.M.	++	++	N.M.	++	++	N.M.
Acrylfentanyl	+	++	N.M.	+	++	N.M.	++	++	N.M.	++	++	N.M.
Alfentanil	N.M.	−	−	N.M.	−	−	N.M.	−	−	N.M.	−	−
Benzodioxolefentanyl	−	++	N.M.	−	+	N.M.	−	−	−	−	++	N.M.
Butyrfentanyl	(+)	++	N.M.	(+)	++	N.M.	++	++	N.M.	(+)	++	N.M.
Carfentanilb	−	(+)	++	−	(+)	++	−	−	−	−	(+)	++
cis-3-Methylfentanyl	−	+	N.M.	−	+	N.M.	−	(+)	++	−	++	N.M.
Crotonylfentanyl	(+)	++	N.M.	(+)	++	N.M.	(+)	++	N.M.	+	++	N.M.
Cyclopropylfentanyl	(+)	++	N.M.	−	++	N.M.	++	++	N.M.	+	++	N.M.
Despropionyl-2-fluorofentanyl	N.M.	−	−	N.M.	−	−	N.M.	−	−	N.M.	−	−
Furanylbenzylfentanyl	−	+	N.M.	−	(+)	++	++	++	N.M.	−	++	N.M.
Furanylfentanyl	(+)	++	N.M.	−	++	N.M.	++	++	N.M.	(+)	++	N.M.
Methoxyacetylfentanyl	++	++	N.M.	++	++	N.M.	++	++	N.M.	++	++	N.M.
Norcarfentanil	N.M.	−	−	N.M.	−	−	N.M.	−	−	N.M.	−	−
Ocfentanil	(+)	++	N.M.	(+)	++	N.M.	−	−	++	+	++	N.M.
Remifentanil	N.M.	−	(+)	N.M.	−	(+)	N.M.	−	−	N.M.	−	(+)
Sufentanil	N.M.	−	+	N.M.	−	+	N.M.	−	−	N.M.	−	+
Tetrahydrofuranylfentanyl	+	++	N.M.	+	++	N.M.	(+)	++	N.M.	++	++	N.M.
Valerylfentanyl	−	++	N.M.	−	++	N.M.	−	−	++	(+)	++	N.M.
Clearly Positive ++/ Positive +	5	21	24	5	19	24	12	15	21	9	22	24
Weak Positive/Ambiguous (+)	8	2	1	4	4	1	2	2	0	6	1	1
Negative −	15	5	3	19	5	3	14	11	7	13	5	3

^aThe summed results at 10,000 ng/mL are based on accumulation of the results at 1000 and 10,000 ng/mL,

^bCarfentanil was tested at 4000 ng/mL and generated a clearly positive result for all test strip brands except One Step. The test results were graded as follows: ++; clearly positive (no T-band), +; positive (barely visible T-band), (+); weak positive/ambiguous (faint T-band) and; − negative (solid T-band). N.M.; not measured.

Fentanyl test strips

Four different brands of commercially-available fentanyl test strips were purchased for the study: Rapid Response Fentanyl (FYL) Test Strips (cut-off 20 ng/mL) from BTNX Inc. (Markham, ON, Canada), Rapid Self Test (RST) - Fentanyl FYL20 (cut-off 20 ng/mL) from Rapid Self Test Inc. (Mississauga, ON, Canada), Nal van Minden Drug-Screen ^® FYL 10 (cut-off 10 ng/mL) from Nal van Minden (Moers, Germany), and One Step FYL20, Rapid Test Dipstick (Urine) (cut-off 20 ng/mL) from Hangzhou Alltest Biotech Co., Ltd (Hangzhou, China).

Preparation of drug solutions for fentanyl test strip experiments

Individual drug solutions of fentanyl analogues (listed in Table 1), non-fentanyl synthetic opioids (U-47700, U-48800, and U-50488), traditional drugs of abuse (amphetamine, cocaine, heroin, LSD, MDMA, methamphetamine, morphine, and THC), and methadone at 100 and 1000 ng/mL were prepared in Type 1 water. Additional highly concentrated individual solutions of selected fentanyls (10 000 ng/mL) and carfentanil (4000 ng/mL) were prepared in Type 1 water (Table 1). The drug concentrations were chosen to examine whether the test strips could detect all fentanyls at lethal and sub-lethal concentration levels, based on a previous report stating that 20 μg carfentanil (~4000 ng/mL when dissolved in 5 mL) is potentially life-threatening (Casale, Mallette, & Guest, 2017). Solutions of fentanyl at 10-1000 ng/mL were prepared in Type 1 water. Individual drug solutions of fentanyl, carfentanil, and remifentanil (1000 ng/mL) were prepared in a solution of ascorbic acid (75 mg/mL) or heroin (50 000 ng/mL, dissolved in Type 1 water), respectively. The concentration of ascorbic acid was chosen to reflect the concentration used when preparing heroin for injection (Scott, Winfield, Kennedy, & Bond, 2000). Individual drug solutions (10 ng/mL) of acetylfentanyl, acrylfentanyl, furanylfentanyl, and butyrfentanyl, and a combined mixture of the four fentanyls (each 10 ng/mL) were prepared in Type 1 water to assess additive effects on test results.

Fentanyl test strip experiments

The test strips were dipped into the drug solutions for 13-20 seconds according to the instructions for urine testing provided by the manufacturer. For more information, see Supplementary Material.

Results

Interpretation of test results

According to the manufacturers, the test should be considered negative if two colored bands appear, regardless of the intensity of the color band in the test region (Figure 1A). When testing a positive (fentanyl 1000 ng/mL) and negative control (water), a clearly positive and negative test result could be observed for all four fentanyl test brands examined (Figure 1B). However, when testing different concentrations of fentanyl (10-250 ng/mL) it became evident that the test results were highly concentration dependent (Figure 1C). In the present study, we therefore chose to grade the results as follows: ++; clearly positive (no T-band), +; positive (barely visible T-band), (+); weak positive/ambiguous (faint T-band) and; − negative (solid T-band). Tests displaying ++ /+ results are hereafter referred to as positive.

Figure 1:

A: Instructions for interpretation of fentanyl test strip results provided by the manufacturers. B: Four brands of fentanyl test strips displaying positive (fentanyl 1000 ng/mL) and negative (Type 1 water) results. From the top: Rapid Response, One Step, Nal van Minden Drug-Screen, and Rapid Self Test. C: Fentanyl test strips tested in descending concentrations of fentanyl (left: Rapid Response; right: One Step). The test results were graded as follows: ++; clearly positive (no T-band), +; positive (barely visible T-band), (+); weak positive/ambiguous (faint T-band), and − ; negative (solid T-band).

Selectivity of fentanyl test strips

As expected, all tests were positive for fentanyl (Table 1). Rapid Response, Rapid Self Test, and Nal van Minden showed identical selectivity with a positive result for all fentanyl analogues except alfentanil, despropionyl-2-fluorofentanyl, norcarfentanil and remifentanil, the latter being weak positive/ambiguous (Table 1). The test from One Step differed in selectivity, producing negative test results for alfentanil, benzodioxolefentanyl, carfentanil, despropionyl-2-fluorofentanyl, norcarfentanil, remifentanil, and sufentanil.

All of the fentanyl test strip brands generated negative results when tested for the non-fentanyl synthetic opioids U-47700, U-48800, and U-50488, the opioid dependence treatment methadone, as well as common drugs of abuse such as amphetamine, cocaine, heroin, LSD, MDMA, methamphetamine, morphine, and THC.

Sensitivity of fentanyl test strips

To examine the sensitivity of the four different brands of fentanyl test strips, we tested all strips for fentanyl analogue solutions at 100 ng/mL, i.e. a concentration 5–10 times higher than the reported limit of detection for fentanyl. The One Step test was superior with respect to sensitivity, resulting in a positive test result for 12 different fentanyl analogues, compared to 5 for Rapid Response and Rapid Self Test, and 9 for Nal van Minden (Table 1).

Of major importance, none of the test strips gave a positive test result for the ultra-potent fentanyl analogue carfentanil when tested at 1000 ng/mL. Rapid Response, Rapid Self Test, and Nal van Minden gave a positive test result for carfentanil at concentrations ≥4000 ng/mL (Table 1), while One Step gave a negative result for all concentrations tested. These findings indicate that the test strips could give a negative result for a potentially deadly dose of carfentanil (20 μg in 5 mL; 4000 ng/mL).

Additive effect of fentanyl analogues for test strip results

Test strips were tested using single-component solutions of acetylfentanyl, acrylfentanyl, furanylfentanyl, and butyrfentanyl at low concentrations (10 ng/mL), generating a negative or weak positive/ambiguous test result for the four individual fentanyl analogues alone. When tested as a mixture of the four fentanyl analogues (each at 10 ng/mL), a positive result was achieved (Supplementary Material; Figure 2). This experiment was only demonstrated by using the Nal van Minden test, but this additive effect is expected to be a general principle for all test strip brands.

The effect of ascorbic acid and heroin on test results

For fentanyl, carfentanil, and remifentanil, we examined the test results in drug solutions containing ascorbic acid to mimic the preparation procedure employed by heroin users prior to injection of heroin base. For all test strips dipped in solution containing ascorbic acid at ≥ 25 mg/mL, the reading area and bands displayed a smeared, pale pink appearance. Despite the smeared appearance, the test result was easy to interpret and gave the same result as previously, except for the One Step test which now displayed a clearly positive result also for carfentanil and remifentanil (Supplementary Material; Figure 3).

To examine whether the presence of ascorbic acid alone generated false positive test results, all fentanyl test strips were tested with drug-free solutions of increasing concentrations of ascorbic acid (0-75 mg/mL). Ascorbic acid at concentrations of 10-75 mg/mL gave false positive test results for the One Step test (Figure 2) and ascorbic acid solution had to be diluted to 1.5 mg/mL to not affect the results. None of the other tests gave false positive test results in the presence of ascorbic acid. The presence of heroin had no effect on the test results for any of the fentanyl strips examined.

Figure 2:

A: The effect of ascorbic acid on fentanyl strips results. Fentanyl strips (left: Rapid Response; right: One Step) were tested in drug-free solutions of ascorbic acid (0-75 mg/mL). Ascorbic acid at concentrations ≥ 10 mg/mL gave false positive results for the One Step test, while the test results for the three other tests were not affected by the presence of the acid.

Discussion

The risk of overdose has markedly increased in recent years, partly because highly potent fentanyls are being sold as heroin or counterfeit opioid prescription medicines. This situation calls for user-friendly methods for rapid self-testing of street drugs (Ciccarone, 2017; Jannetto, et al., 2019). Urine fentanyl test strips are one potential method to examine contamination of illicit drugs with fentanyls, but there is a major lack of knowledge about the ability of such tests to recognize the numerous fentanyl analogues present on the current drug market.

Here we show that four different urine fentanyl test strips successfully recognized a broad panel of fentanyls, including several analogues recently implicated in many overdose deaths (e.g. acetylfentanyl, butyrfentanyl, cis-3-methylfentanyl, cyclopropylfentanyl, furanylfentanyl, and methoxyacetylfentanyl) (Drummer, 2019; Kraemer, Boehmer, Madea, & Maas, 2019). Three of the tests, Rapid Response, Rapid Self Test and Nal van Minden, gave a positive test result for as many as 19-22 of the 28 fentanyl analogues tested at 1000 ng/mL and 24 analogues at 10 000 ng/mL, whereas the One Step test detected 15 and 21 compounds at 1000 and 10 000 ng/mL, respectively. Alfentanil, despropionyl-2-fluorofentanyl, and norcarfentanil were not recognized by any of the tests, while the One Step test additionally generated negative test results for benzodioxolefentanyl, carfentanil, remifentanil, and sufentanil. In contrast to the manufacturer guidelines, we recommend a less rigorous interpretation of the test results, i.e. interpreting tests with a faint T-band as indicative of presence rather than absence of fentanyls. This was based on our finding that the test results were highly concentration dependent.

A major weakness of the fentanyl test strips examined was the low sensitivity to the highly potent fentanyl analogue carfentanil. The concentration of carfentanil had to be increased to a potentially deadly concentration (4000–10 000 ng/mL) to give a positive result, and for the One Step test, the result was negative for all concentrations examined. Web stores selling fentanyl strips, as well as harm reduction organizations, recommend testing street drugs by adding water to the equipment used for preparing the user dose or to the container employed for drug storage (BTNX, 2019; Dancesafe, 2019; Toledo-Lucas County Health Department, 2018). Using this procedure, the concentration of the drug solution for testing would be considerably lower than the solution for injection, which again would increase the chances of generating a negative test result for potentially deadly doses of carfentanil.

The presence of common drugs of abuse and other non-fentanyl synthetic opioids did not give positive test results; however, ascorbic acid at concentrations ≥ 10 mg/mL gave false positive test results for the One Step test. Four fentanyls at concentrations too low to give a positive test individually (10 ng/mL) generated a clearly positive test result when examined as a mixture. This additive effect might be especially important in cases where a drug dose contains low levels of two or more highly potent fentanyl analogues.

Conclusion

In summary, the fentanyl test strips tested displayed a broad selectivity, detecting the vast majority of the fentanyl analogues present on the recreational drug market. The test results were highly concentration dependent; hence, the test result should be interpreted less rigorously than recommended by the manufacturers. None of the fentanyl strips were affected by the presence of heroin; however, one of the test strip brands gave false positive results when exposed to ascorbic acid. A major limitation for all the test strips was the low sensitivity for detecting the ultra-potent fentanyl analogue carfentanil, even at potentially life-threatening concentrations. In contrast to the instructions provided by user organizations and web stores, we recommend to test drug injection solutions without any further dilution. Overall, the urine fentanyl test strips can be a useful tool for harm reduction initiatives when accompanied by the pertinent information about test limitations and how to interpret test results.