Clinical evaluation of an authorized medical equipment based on high performance liquid chromatography for measurement of serum voriconazole concentration

BACKGROUND: Therapeutic drug monitoring for voriconazole is recommended for its optimum pharmacotherapy. Although the feedback of the measurement result of serum voriconazole concentration by outsourcing needs a certain time (days within a 1 week), there was no medical equipment for the measurement available in clinical practice. Recently, a medical equipment based on high performance liquid chromatography, named LM1010, has been developed and authorized for clinical use. In this study, to validate the clinical performance of LM1010, we compared the measured serum voriconazole concentrations by LM1010 with those by outsourcing measurement using liquid chromatography-tandem mass spectrometry. METHODS: We conducted the observational study approved by the institutional review board of Kumamoto University Hospital (No. 1786). Residual serum samples harvested for therapeutic drug monitoring were separated. Measured concentrations by LM1010 by the standard filter method (needs serum volume of > 400 μL) or the dilute method (needs serum volume of 150 μL) were compared with those by outsourcing, respectively. Acceptable measurement error range of 0.72–1.33 was considered. There were 69 serum samples, where the 35 or 34 samples were employed for evaluation of the standard filter method or the dilute method, respectively. RESULTS: The measured concentration using the standard filter method/outsourcing was 2.22/2.10 μg/mL as the median, 1.57–3.40/1.53–3.62 as the interquartile range, < 0.2–10.76/< 0.2–11.46 μg/mL as the range, while those using the dilute method/outsourcing was 2.36/2.29 μg/mL as the median, 1.08–2.94/1.03–3.06 as the interquartile range, 0.24–10.00/< 0.2–10.85 μg/mL as the range. The regression line for the standard filter method or the dilute method were y = 0.935x + 0.154 or y = 0.933x + 0.162, respectively. The standard filter method or the dilute method showed 11.4% samples (4/35, 95%CI 3.2–26.7%) or 8.8% samples (3/34, 95%CI 1.9–23.7%) out of the acceptable measurement error range, respectively. CONCLUSION: Measurement of serum voriconazole concentration by LM1010 can be acceptable in clinical TDM practice. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s40780-021-00225-8.