False-positive phencyclidine immunoassay results caused by venlafaxine and O-desmethylvenlafaxine.

To the Editor: We recently observed three occurrences of false-positive results for phencyclidine (PCP) on urine samples tested with the Syva RapidTest d.a.u. 9 Test Panel (Syva Company, subsidiary of Dade Behring Inc.). The RapidTest d.a.u. 9 device is a singleuse, one-step, solid-phase immunochromatographic assay for the qualitative, discrete detection of several drugs/drug metabolites in human urine (1 ). The urine samples were collected from three different patients who were seen in our Emergency Department within 3–4 weeks after we had implemented the RapidTest device in our laboratory (Danbury Hospital). Because PCP is not a commonly used drug in our locale and the rare positive results have usually been confirmed as false positives, three positive PCP results in such a short time period immediately aroused suspicion. Patient A was a 52-year-old male with schizoaffective disorder who was admitted to the psychiatry service. Patient B was a 93-year-old female from a nursing home who was admitted with a hip fracture. Patient C was a 50-year-old female admitted for opiate and benzodiazepine overdose. None of these patients had a history of PCP use or presented with symptoms consistent with use of this drug. Patient A’s urine gave negative results for all other drugs tested with the RapidTest. Patient B’s urine also tested positive for benzodiazepines and barbiturates; these results could be attributed to medications documented in this patient’s medical record: lorazepam, clonazepam, and phenytoin (the latter is listed in the RapidTest package insert as an interferent in the barbiturates assay). Patient C’s urine also tested positive for benzodiazepines and opiates; again, these results were consistent with this patient’s known medications (lorazepam and hydrocodone). None of the patients was receiving dextromethorphan, diphenhydramine, ibuprofen, imipramine, meperidine, mesoridazine, or thioridazine—drugs known to produce false-positive results with other PCP immunoassays. The only drug that appeared on the medication lists of all three patients was venlafaxine, a relatively new antidepressant marketed as Effexor (Wyeth-Ayerst Pharmaceuticals, Inc.) (2 ). We therefore suspected that this drug and/or one or more of its metabolites were the cause of these false-positive results via cross-reactivity with the anti-PCP antibodies used in the RapidTest devices (cutoff concentration for PCP, 25 g/L). Negative results were obtained when the same urine samples from these three patients were tested with the Syva Emit II Plus PCP assay (25 g/L cutoff) on the Roche Cobas MIRA analyzer (performed at Danbury Hospital). Similarly, analysis by two other single-use immunochromatographic devices, OnTrak TesTstik PCP (Roche Diagnostics) and Biosite Triage 8 DOA (Biosite Diagnostics) also gave negative results for PCP (both 25 g/L cutoff; performed at Hartford Hospital), and gas chromatographic–mass spectrometric analysis did not detect PCP (5 g/L limit of detection; performed at Hartford Hospital). Of note, we also encountered a “true-positive” PCP sample from an 18-year-old male patient admitted with head trauma; this urine tested positive for PCP by the RapidTest, Emit, and gas chromatographic–mass spectrometric methods and thus served as a “positive control”. Venlafaxine undergoes both Oand N-demethylation, with the major metabolite, O-desmethylvenlafaxine (ODV), also exhibiting antidepressant activity. An average of 87% of a labeled oral dose is excreted in a 48-h urine, with 5% excreted as parent drug, 29–48% as ODV, 6–19% as di-N-desmethylvenlafaxine, and 0.2– 7.4% as mono-N-desmethylvenlafaxine (2 ). To further test our hypothesis that venlafaxine and/or its metabolites were responsible for these falsepositive results, we obtained pure samples of venlafaxine and the ODV metabolite from Wyeth-Ayerst Research, the manufacturer of this drug. We prepared solutions of venlafaxine and ODV in drug-free urine at final concentrations of 10, 10, 10, 10, and 10 g/L venlafaxine or ODV and tested these solutions with both the Emit II and RapidTest PCP assays. The RapidTest gave a clearly positive PCP result (no signal line indicates positive result) at a concentration of 10 g/L of either venlafaxine or ODV and equivocal or borderline results (extremely faint, barely visible line) at 10 g/L of either venlafaxine or ODV, whereas the Emit II assay gave negative results at all concentrations. Additional testing narrowed the concentration where the RapidTest became clearly positive for PCP to somewhere between 1 10 and 2 10 g/L of either venlafaxine or ODV, indicating similar cross-reactivities for these two compounds. From these data, we calculated a cross-reactivity of between 0.0125% and 0.025% for both the parent drug and metabolite, using the 50% displacement method described by Miller and Valdes (3 ). Although at face value this appears to be a very low degree of crossreactivity, it becomes clinically significant if combined concentrations of venlafaxine and ODV of 1 10 g/L are present in a urine sample. According to the literature, predicted steady-state plasma concentrations of venlafaxine and ODV in healthy individuals receiving daily 150-mg doses are estimated as 70 and 254 g/L, respectively (2 ). These concentrations are approximately three orders of magnitude lower than those that were found to give positive results in our addition experiments. Although we could find no published data on urine concentrations of venlafaxine and ODV, we learned that combined concentrations of these two substances on the order of 1 10 g/L have indeed been measured in the urine of patients taking therapeutic doses of venlafaxine (K-T. J. Yeo, personal communication). Combined with the results of our addition experiments, this provides strong evidence that the false-positive results we obLetters