Measuring multidrug resistance expression in human malignancies: elaboration of consensus recommendations

Before the prognostic significance of P-glycoprotein (P-gp) expression can be properly evaluated in prospective clinical trials of P-gp modulators, standard techniques for the measurement of P-gp must be widely accepted. Several multicenter trials have demonstrated large discrepancies in the observed levels of P-gp expression in the same clinical samples evaluated at different centers. The greatest discrepancies occurred with samples that expressed low levels of P-gp. Although standardized procedures have dramatically increased the interlaboratory reproducibility of flow cytometry and polymerase chain reaction assays, data from immunocytochemistry remain difficult to interpret. Consensus recommendations are presented for improving data reproducibility. These recommendations emphasize the importance of using calibrated batches of antibodies and two different antibodies for immunocytochemistry, the need for an internal standard for reverse transcriptase-polymerase chain reaction (RT-PCR) assays, the need for the presentation of data as a continuous variable, and the need for setting standard parameters for flow cytometry. It is also extremely important for the success of clinical trials that multiple techniques be employed to insure accurate measurement of P-gp expression.