Amphotericin B enzyme-linked immunoassay for clinical use: comparison with bioassay and HPLC

OBJECTIVE: To evaluate a new enzyme-linked immunosorbent assay (ELISA) for amphotericin B in serum samples. Results are compared with those obtained by HPLC and bioassay. DESIGN: Comparison of results obtained by ELISA, HPLC, and bioassay. METHODS: We developed a new ELISA using a polyclonal rabbit antibody to measure serum amphotericin B concentrations. Blinded samples of amphotericin B in concentrations of 0.15-78 micrograms/mL were prepared in human serum and assayed simultaneously by the ELISA, HPLC, and bioassay. The results of each assay were derived from standard curves and evaluated by using the Table Curve 2D computer program. These data were compared by using correlation analysis with evaluation of Pearson's correlation coefficient by Student's t-test. RESULTS: ELISA and bioassay compared favorably at amphotericin B concentrations of 0.3-20 micrograms/mL with a correlation coefficient of r = 0.993, while ELISA and HPLC compared with a correlation coefficient of r = 0.944. The average coefficient of variation over the range 0.3-20.0 micrograms/mL was 28% +/- 7% for HPLC, 26% +/- 9% for ELISA, and 13% +/- 4% for bioassay. Comparison of all three assays revealed the highest correlation with the ELISA assay (r = 0.998) for the range of concentrations (0.3-20 micrograms/mL) routinely achieved. Samples containing concentrations in excess of 20 micrograms/mL could be diluted. Desiccation for concentrations less than 0.3 microgram/mL was not tested. CONCLUSION: The determination of serum amphotericin B concentrations by ELISA gave results similar to those obtained by a bioassay and HPLC technique. Although variability appears greater with ELISA, the ease of performing yjis assay expedites the evaluation of amphotericin B concentrations from lipid formulations without interference from coadministered antibacterials of azole antifungals.