Efavirenz Treatment and False-Positive Results in Benzodiazepine Screening Tests

The use of rapid testing with immunoassays to detect illicit drugs with specific antibodies is well established in medical practice. The interpretation of results, however, requires careful consideration, because positive test results may have a serious impact for individuals [1, 2]. The assessment of sample integrity and medication history plays a crucial role for correct interpretation of results. All positive results from rapid testing should be confirmed by specific semiquantitative or quantitative methods [1].

Efavirenz, a nonnucleoside reverse-transcriptase inhibitor, is used as part of highly active antiretroviral therapy (HAART), the current standard treatment for patients infected with human immunodeficiency virus (HIV) [3, 4]. Cross-reactivity of efavirenz in drug screening tests for tetrahydrocannabinol has been previously observed [5–8]. Recently, our group reported incidental findings from a study of healthy volunteers who were receiving efavirenz. In the Triage 8 (Biosite), a well-accepted drug screening test [9], all individuals tested positive for tetrahydrocannabinol and benzodiazepines [10]. To our knowledge, this cross-reactivity between efavirenz and benzodiazepine has not been described elsewhere.

We therefore investigated, in a study involving 100 HIV-positive patients, the effect of efavirenz treatment on the results of drug screening tests and identified the compound responsible for the false-positive result.

Materials and methods. This open, controlled study was approved by the ethics committee of the medical faculty of the University of Heidelberg, and relevant authorities were notified. After receiving full information about the nature of the trial, 100 patients were recruited from 2 centers, and they gave written informed consent. One-half of the patients were receiving efavirenz treatment. Spot urine samples were collected, visually inspected, and immediately frozen at −20°C. Data on medication history within the previous 2 weeks were collected. Drug screening tests and laboratory measurements were performed by individuals who were blinded with regard to the efavirenz intake of patients.

Efavirenz, 8-hydroxy-efavirenz (8-OH-efavirenz), and 8,14-dihydroxy-efavirenz (di-OH-efavirenz) concentrations in urine were determined by high performance liquid chromatography coupled to tandem mass spectrometry (LC/MS/MS) after hydrolysis of glucuronides. The LC/MS/MS method was validated according to the recommendations of the FDA [11] and fulfilled all criteria of required analytical quality. Within-batch accuracies ranged from −13.3% to +9.4%, and batch-to-batch accuracies ranged from −5.5% to +1.8%. Precision (within-batch and batch-to-batch) was not higher than 12.1%. For all urine samples, creatinine concentrations were measured (Jaffe method) to assess possible dilution. Urine samples were analyzed with the following drug screening tests, according to each manufacturer's instructions: Triage 8 (Biosite; package insert from 27 April 2007), Drug Screen Multi 5 (von Minden; package insert from 8 May 2006), DrugControl 008A444 (Ultimed Products; package insert from July 2007), and bioFAST addiTest (testing for benzodiazepines only; Protzek; package insert from February 2005). These tests represent the 2 different methods of drug screening—dip and drip—and were chosen because they are used in the University Hospital Heidelberg.

Samples that yielded positive results when screened for drugs were retested for benzodiazepine with the CEDIA benzodiazepine assay (Microgenics; package insert from 2002) and for tetrahydrocannabinol with the CEDIA multilevel tetrahydrocannabinol assay (Microgenics; package insert from 2002). Hydrolyzed urine samples obtained from 9 patients (6 of whom were receiving efavirenz and 3 of whom were not) as well as blank urine spiked with pure efavirenz and metabolites were screened for drugs with the Triage 8. Pure efavirenz and purified metabolites from urine dissolved in methanol and H₂O (1:1) were diluted in blank urine to yield concentrations comparable with those measured in patients' urine for 8-OH-efavirenz (mean concentration, 238.4 µg/mL). Concentrations used were 270 µg/mL for efavirenz, 264 µg/mL for 8-OH-efavirenz, and 200 µg/mL for di-OH-efavirenz. Blank urine with buffer was used as a control.

Results. One hundred eight patients were screened to recruit 100 patients, 50 of whom were receiving efavirenz in their treatment regimen. Eight patients were excluded because of inappropriate concomitant medication or illegal drug use (n=5), withdrawn consent (n=2), or insufficient amount of urine (n=1). Antiretroviral treatment was prescribed for 88 patients (50 received efavirenz and 38 did not), and 78 received HAART (44 received efavirenz and 34 did not). The use of concomitant medication other than antiretrovirals was well balanced between both groups. Visual inspection of urine samples revealed no abnormal findings. Mean (± standard deviation) creatinine concentration was 9.63 ± 4.95 µg/mL in the efavirenz group and 9.58 ± 5.74 µg/mL in the control group.

In the efavirenz group, efavirenz concentrations (as determined by LC/MS/MS) were below the limit of quantification. Measurement of metabolites showed mean 8-OH-efavirenz concentrations of 238.4 µg/mL (range, 33.7–678.7 µg/mL) and mean di-OH-efavirenz concentrations of 36.9 µg/mL (range, 3.7–69.8 µg/mL). One sample from a patient in the efavirenz group did not have any detectable efavirenz or metabolite concentrations, presumably because of the patient's noncompliance. This sample was excluded from the calculation of mean concentrations. All drug screening tests performed on this sample were negative for benzodiazepines and tetrahydrocannabinol. In the control group, no efavirenz or metabolites were detected at all.

In the efavirenz group, the samples obtained from all 49 patients with proven efavirenz intake tested positive for benzodiazepines with the Triage 8. Three samples tested positive with ⩾1 additional test. Only 1 of the 50 control subjects tested positive for benzodiazepines in the Drug Screen Multi 5 (table 1).

View larger version:

• In this window
• In a new window

• Download as PowerPoint Slide

Table 1

Results of drug screening and confirmation tests for benzodiazepines and tetrahydrocannabinol for 100 patients.

In the efavirenz group, all 49 patients tested positive for tetrahydrocannabinol with the Triage 8, and 46 tested positive in the Drug Screen Multi 5 also (table 1). In the control group, 3 patients tested positive for tetrahydrocannabinol in ⩾2 drug screening tests. All samples with positive findings were subjected to confirmation testing by the Microgenics CEDIA test system. The single sample that had a positive test result for benzodiazepines in all drug screening tests (which was obtained from a patient who was receiving efavirenz) had a positive result for the confirmation test with the Microgenics CEDIA test for benzodiazepines. All other samples tested negative for benzodiazepines (table 1). For tetrahydrocannabinol, 3 samples from patients in the efavirenz group and 3 samples from patients in the control group that had ⩾1 positive screening test result were also confirmed positive with Microgenics CEDIA (table 1). Drug screening tests further exhibited sporadic positive findings for amphetamines, methamphetamines, methadone, barbiturates, opiates, and tricyclic antidepressants. These findings were balanced between both treatment groups and did not show any pattern related to the intake of efavirenz.

Nine randomly selected urine samples (obtained from 6 patients who were receiving efavirenz and 3 control subjects) were hydrolyzed, and the samples from the 6 patients who were receiving efavirenz were still positive for benzodiazepines with the Triage 8, whereas the cross-reaction for tetrahydrocannabinol disappeared. The samples from the 3 control subjects remained negative. Blank urine spiked with efavirenz and 8-OH-efavirenz resulted in positive results for the Triage 8 for benzodiazepines but not for tetrahydrocannabinol. Urine spiked with di-OH-efavirenz showed no positive result for either benzodiazepines or tetrahydrocannabinol on the Triage 8.

Discussion. Drugs for which there is anecdotal information about cross-reactions in drug screening tests sometimes show no such reaction when larger prospective studies are performed [12]. We therefore performed a prospective study in a population of 100 HIV-positive patients who were receiving pharmacotherapy regimens—one-half of which included efavirenz—to examine presumed cross-reactions from efavirenz in drug screening tests.

Our study confirmed that patients with measurable concentrations of efavirenz metabolites showed false-positive drug screening results for benzodiazepines and tetrahydrocannabinol with the Triage 8, an otherwise accepted and reliable drug screening test [9]. Three other drug screening tests did not show more than an incidental number of cross-reactions with the benzodiazepine test; however, 1 of them exhibited false-positive results for tetrahydrocannabinol.

Given the structural similarity between efavirenz and benzodiazepines, the persisting positive test reaction for benzodiazepines after hydrolyzation of urine samples suggested that the efavirenz moiety of the metabolite is responsible for the cross-reaction with benzodiazepines. This is also supported by positive results for benzodiazepines from blank urine spiked with efavirenz and 8-OH-efavirenz metabolite tested by the Triage 8. Because efavirenz is not present in urine in unchanged form, the observed response in patients' urine is therefore likely caused by 8-OH-efavirenz. In contrast to these findings, the cross-reaction of efavirenz with tetrahydrocannabinol has been attributed to the glucuronidated structure of efavirenz metabolites [5], and this attribution is supported by our data.

We examined 4 commercially available drug screening tests that represent an enormous variety of such tests. It is reasonable to assume that the described findings might also occur in some other drug screening tests; however, evaluation of each test would be necessary to allow for a final assessment of each test. Manufacturers should be obliged to evaluate their tests in this respect. On the basis of our results, however, the Triage 8 and the Drug Screen Multi 5 should definitely be avoided when screening patients who are receiving efavirenz for the presence of illicit drugs. Because efavirenz is an antiretroviral drug that is often given to HIV-infected patients, it is therefore of utmost importance that clinicians and patients know about this cross-reactivity to avoid incorrect consequences after false-positive urine drug screening test results.

• Received December 11, 2008.
• Accepted January 29, 2009.

• © 2009 by the Infectious Diseases Society of America