SAMPLE PRETREATMENT TO MINIMIZE INTERFERENCES FROM WHOLE BLOOD IN THE RADIOIMMUNOASSAY FOR CYCLOSPORINE

Publisher Website
Google Scholar
Abstract
There is much controversy as to whether the analysis of cyclosporine (CsA) should be performed by radioimmunoassay (RIA) or high-performance liquid chromatography (HPLC), and wheter the specimen should be serum or whole blood. Whole-blood specimens present specific advantages, but the presence of hemoglobin (Hgb) and other endogenous compounds can produce major errors in the RIA by "quenching" the analytical signal or by interfering with the antigen-antibody binding in the assay. We have developed a simple pretreatment step to remove the Hgb and other proteins responsible for this error. Red cells in whole blood are hemolyzed with a mixture of acetonitrile and water, the protein precipitated with acetonitrile, and the supernatant assayed by RIA. In a controlled study in which CsA concentration was kept constant and the Hgb concentration varied, the errors in measurement were directly proportional (r = 0.999) to the Hgb concentration. CsA values were spuriously deflated or inflated by 22.7 .mu.g/L for each gram per 100 milliliters that the Hgb deviated from the 9.2 g/100 ml Hgb in the CsA calibration standards. In a similar study in which patient samples (n =57) were assayed with and without pretreatment, the fractional error induced by Hgb was compounded in some patients by additional interferences that also appear to be removed by sample pretreatment. Without the pretreatment, CsA values could be in error by 33% when the Hgb varied 4 g/100 ml, thus providing potentially misleading results to the clinician. An I-125-labeled CsA tracer (purported not to be affected by the "quenching" interference of Hgb) produced consistently higher results when it was substituted for the tritiated CsA tracer contained in the Sandoz kit. In summary, sample pretreatment appears to be the simplest method of effectively removing endogenous interferences and minimizing erroneous results from whole blood submitted to the Sandoz RIA for CsA analysis.

This publication has 0 references indexed in Scilit: