PerspectivesAre you interested in submitting a Perspective Article? Be sure to read The Science Advisory Board's Editorial Guides for Perspective Articles. Click here. Interference in Therapeutic Drug Monitoring Assays by Pradip Datta, Ph.D. Bayer Corp.- Diagnostics Division Tarrytown, NY 10590 USA Many of the therapeutic drugs used in medical care have toxic effects when present at increased levels in the blood. On the other hand, these drugs often need a minimum concentration to express their therapeutic potential. Since there are wide inter-individual (and even in the same individual) variation in drug absorption, distribution, or metabolism, Therapeutic Drug Monitoring (TDM) is an important aspect of clinical management of diseases. The most common patient matrix used for TDM analysis is blood (as serum or plasma); however, urine, saliva and sweat can also be used in some cases. TDM analysis is mostly performed in clinical laboratories. Based on the TDM results, the dosage of the drug is adjusted so that desired blood levels and optimum clinical management can be achieved. Because of the ease of use and increased automation, most TDM is now performed by immunoassays. In such analysis, the patient sample is treated with reagents containing specific antibodies raised against the target analyte. While new technology allows increased precision, specificity, and sensitivity in immunoassays, different types of interference may affect the assays to generate incorrect results. A second assay may not be affected by the same interference, thus generating a discordant result between the two. A clinical discordance may also result where an assay result disagrees with the clinical picture; e.g., some digoxin assays are interfered with endogenous digoxin-like immunoreactive factors (DLIF) (1). Incorrect lab results are unacceptable, because they lead to incorrect patient treatment, resulting in excess morbidity or mortality, and many legal issues as well. Definition and resolution of discordance: During method comparison between two assays, discordant results may be defined statistically (results out of the area defined by ± 2-3*Standard Estimate of Error from the least square fit regression line for the paired results in a scatter-plot) or arbitrarily (a percent or absolute difference between the assay results, depending on the assay range). After such results are identified, both of the assays need to be repeated. If the discordance still persists, then one or both results may be inaccurate because of possible assay interference. Several methods may be pursued to resolve such discordance:
Causes for discordance: Some of the causes of discordant results in an assay may be: specificity or precision of the assay, incorrect storage of the sample affecting analyte concentration (2), system issues, and matrix effect. The latter term may include interference from various constituents of the sample, e.g., protein, bilirubin, hemoglobin and blood substitutes. Cross-reactivity of a TDM assay, especially one measuring ng/mL or lower concentration of the analyte, causes the most cases of discordance observed. Assay cross-reactivity should be determined to various endogenous or exogenous substances with structure similar to the analyte, and which may be present in patient sera. For the most meaningful results, assay response to the metabolites of the analyte should match the metabolites’ physiological activity (3-6). Determination of cross-reactivity should be done both in presence and absence of the cross-reactant. Not only the extent of interference may depend on both the analyte and cross-reactant concentrations, under certain conditions interference may change from positive to negative in assays of certain formats. Thus, the interference of digitoxin in a microparticle enzyme immunoassay for digoxin changes from positive to negative as digoxin levels increase for the same concentration of digitoxin (7). In another example, the cross-reactivity of oxaprozin (a non-steroidal anti-inflammatory drug) in a fluorescence polarization immunoassay (FPIA) for phenytoin is significantly higher at higher oxaprozin concentrations (99% and 131% at 300 and 400 mg/L) (6). Furthermore, the crossreactivity was ~5% higher in presence of phenytoin than without phenytoin. Methods to reduce assay interference: Once a possible interfering agent is suspected for an immunoassay, several methods may be used to confirm the interference, or even eliminate it. If the interferent is of relatively large molecular weight (or, strongly bound to proteins of large molecular weight), and the analyte is not protein bound, the interferent may be removed by ultrafiltration, and the protein free filtrate assayed for the analyte. For example, ultrafiltration removed DLIF (~95% protein bound) interference by 89% in a RIA for digoxin (8) or in FPIA (9-11). Chromatographic separation of the drug from other interferents before detection is often the ‘gold’ method of analysis for a drug. For example, HPLC methods can separate digoxin from its metabolites, DLIF, and other interferents (12). Of course, the easiest method to reduce discordance is to use more specific assays. Many commercial kits for the common TDM assays using monoclonal antibodies of desirable specificity are now available. Examples are assays for digoxin (13-15), digitoxin (16-17), and phenytoin (18). In summary, when immunoassays are used for TDM, the possibility of various types of assay interference must be considered. Assays of superior specificity, whenever possible, should be used to generate accurate drug levels in patients’ sera. References:
### << Previous Next >> [ View All Perspectives ] |
|
