Performance of Two Fentanyl Immunoassays against a Liquid Chromatography-Tandem Mass Spectrometry Method

Abstract

Rapid and automated fentanyl screening assays are in need due to the prevalence of fentanyl abuse. In the present study, we evaluated the clinical performance of two FDA-cleared automated fentanyl immunoassays, the Immunalysis SEFRIA fentanyl assay and the ARK fentanyl assay. Liquid chromatography-tandem mass spectrometry (LC-MS-MS) was used as a gold standard. Two groups of urine specimens were tested, including 225 specimens from patients presenting to the emergency department (ED) for whom urine drugs of abuse screens were ordered and 57 specimens from patients in chronic pain management programs. The SEFRIA assay generated higher assay imprecision than ARK assay (intraday CV%, 7.15 vs. 4.7%; interday CV%, 6.6 vs. 5.3%). Clinical sensitivity and specificity for detection of fentanyl exposure were 100 and 96% for the ARK assay and 95 and 80% for the SEFRIA assay. An 'auto-repeating' issue was observed for some validation specimens flagged with high absorbance values (OD > 3.0), generating false repeat results. The frequency of auto-repeating was lower in the ARK assay than SEFRIA (0.7 vs. 15.5%). Auto-repeating occurred for only previously frozen specimens in the ARK assay, but 9% of fresh specimens were also flagged and repeated in the SEFRIA assay. Positive predictive value (PPV) of the ARK assay was 73% in the ED population and 67% in the non-ED populations. The concentrations of fentanyl and norfentanyl were higher in specimens from ED patients than patients from pain management programs. High prevalence of morphine, methamphetamine, benzoylecgonine and 6-MAM was observed in specimens positive for fentanyl in both populations.