A validation study of selected methods routinely used for measurement of cyclosporine

We compare four methods for measuring cyclosporine (CyA) in plasma and whole blood of transplant patients: HPLC, RIA with a polyclonal antibody, RIA with a monoclonal antibody, and fluorescence polarization immunoassay (FPIA). The monoclonal RIA procedure correlated acceptably with HPLC, with slope = 1.21, r = 0.97, and Sy,x = +/- 40.1. However, the FPIA, done in three separate instruments, correlated relatively poorly with HPLC, giving slopes of 1.67, 1.51, and 2.32; correlation coefficients of 0.72, 0.43, and 0.83; and Sy,x = +/- 205.4, +/- 334.5, and +/- 222.4. The polyclonal RIA correlated reasonably well with HPLC, with a slope = 1.15, r = 0.90, and Sy,x = +/- 72.6. Values for individual patients with increases both in gamma-glutamyltransferase and creatinine showed very poor correlation between FPIA and HPLC, which suggests that metabolite cross-reactivity with FPIA is significant and unpredictable in patients with liver dysfunction coexisting with renal dysfunction. Evidently, the monoclonal RIA can be substituted for HPLC, if the therapeutic range is adjusted for the 21% higher results obtained by RIA.