Printer Friendly

Quality control in the new environment: ligand assay and TDM; part II.

Quality control in the new environment: Ligand assay and TDM

Ligand assay and extensive theapeutic drug monitoring are relatively recent additions to the clinical laboratory armamentarium. Both have undergone significant evolutionary change within the past decade, as have the quality control methods for insuring their reliability.

Ligand assay was initially described by Berson and Yalow in their report of an insulin assay using anti-insulin-antibody and labeled insulin as the key reagents. This was the radioimmunoassay form of ligand assay. The most commonly used early ligand assay procedures, however, were based on competitive protein binding using serum carrier proteins for thyroxine and cortisol. These assays were initially performed with homemade reagents, before commercial reagents became widely available.

Competitive protein binding assays were gradually replaced by radioimmunoassays. Until recently, the RIAs used polyclonal antibodies of animal origin, with significant lot-to-lot variation. These variations were also prevalent with changes in kit components other than the antibody, contributing to imprecision in long-term quality control.

Now that the use of monoclonal antibodies is gaining momentum, we can expect major improvements in stability of antibody characteristics. In some cases, sandwich techniques have already led to improved sensitivity, while use of two monoclonal antibodies directed at different antigenic sites on a single molecule--himan chorionic gonadotropin, for example--has led to greatly improved specificity. In addition to isotopic labeling, enzyme labeled immunoassays and such new technologies as fluorescence polarization immunoassay have become available to laboratorians.

Therapeutic drug monitoring has undergone a similarly impressive transformation over the past decade. From crude spectrophotometric determinations of barbiturates, theophylline, and other drugs, the state of the art has evolved to more widespread use of chromatographic procedures for drug analysis in larger labs. Most drugs are currently monitored by ligand assay techniques, particularly enzyme and fluorescence polarization immunoassays.

Thus in recent years the two fields have come to use similar techniques and major assay systems. Each step in the technological evolution of therapeutic drug monitoring and ligand assay has beenaccompanied by increased ease of performance, generally decreased skill requirements for test performance, and improved precision and accuracy demonstrable in both internal quality control programs and surveys.

Data reduction for ligand assay is now widely accomplished by the selection of appropriate algorithms preprogrammed into current instrumentation. Some instruments also help prevent laboratory blunders by automatically checking for short samples and positively identifying reagent containers to monitor assay performance.

Early ligand assay quality control was performed by using point-to-point plotting of standard curve data on appropriate graph paper. With the introduction of calculator- and microprocessor-based data reduction, a variety of curve fitting techniques--most often logit-log with various weighting schemes--were developed. Early in the course of mathematical treatment of ligand assay data, multiple curve parameters were used to define assay component status and in some cases to provide added QC parameters.

Two related developments--improved quality of kit reagents and the resulting loss of ability to adjust the assay component concentration in the laboratory--greatly diminished the usefulness of studying curve parameters in the individual lab using commercial reagents. The critical decision in the clinical laboratory today is selection of the appropriate algorithm for the individual ligand assay to reduce data exhibiting a nonlinear dose response. The continuing evolution of monoclonal antibody techniques and dual anti-body procedures, as well as immunoassay and instrument-based data reduction techniques, will continue to improve the quality of assays.

Initial therapeutic drug monitoring survey results were appalling, both in the first U.S. survey coordinated by Pippenger and in an early British drug assay survey program. However, performance here, as with ligand assay, improved with simpler, more automated techniques. The manufacturers of commercial TDM reagents and instrument systems have played a large role in improving the quality of assays available to laboratorians.

Quality assurance is not an entirely statistical process. It represents a constellation of techniques to assure the quality of patient results, estimate assay imprecision, and to a lesser degree estimate the inaccuracy that can affect patient results. As new reagents and instrument systems develop, quality assurance can best be improved by careful selection of products. Data from the College of American Pathologists' Survey and Quality Assurance Service programs provide good clues to reagent and system performance.

We will conclude with recommendations for 1) routine statistical quality control of ligand assay and therapeutic drug monitoring and 2) quality assurance. First, the QC recommendations:

* Control each procedure at two clinically relevant levels, usually near the upper and lower limits of the reference or therapeutic range. Routine use of tri-level quality control materials serves largely to increase manufacturer sales of control products and to slightly increase reagent utilization.

Periodic testing of samples that contain very high or low levels, particularly in classic RIA techniques, is useful since imprecision increases at extremes of the assay range. Similar periodic testing of upper assay range limits for enzyme immunoassays is also apprpriate. Occasional linearity studies, perhaps annually, are warranted. The current emphasis by some inspecting agencies on frequent linearity studies seems to lack a reasonable basis.

* Test your data reduction system in an ongoing fashion by reviewing the concentrations of the standards as determined from the standard curve that they generate. Assign acceptable deviation from each of the standard concentrations according to your experience with the assay. Reject individual points on a standard curve based upon such review and recalculate the standard curve.

* Participate in regional or national quality control programs to compare your precision with that of peer laboratories. I have found the CAP's Quality Assurance Service especially useful in establishing benchmark precision targets and demonstrating bias in our laboratory's performance. The CAP Survey is useful for periodic assessment of accuracy, and its summar data performance by

all laboratories helps in the selection of better methods.

The modern precise laboratory methods now available are capable of producing results in high quantity that in some cases may be totally misleading clinically. Preventing this anomalous situation must be the goal of all laboratorians who are concerned with quality assurance. Here are several recommended steps to take:

* Develop a fail-safe system relating drug dosing times to peak and/or trough collection times. Preferably, quantity and time of administration should accompany the drug level data and collection time in t he laboratory report.

* Educate the clinical staff concerning the proper timing of specimen collection for therapeutic drug monitoring.

K Know your assays. Understand interactions of drugs; collection device artifacts; the effects of storage conditions on specimens; and pathophysiologic states that may affect assay results--such as pseudo-digoxin in neonatal serum and pseudo-theophylline in uremic serum when measured by some immunoassays. Know which TSH techniques are useful for the detection of both hyperthyroidism and hypothyroidism.

* Many hormones demonstrate frequent bursts of secretory activity, which may lead to a clinically misleading result on a single specimen. In that event pooled specimens may be useful for the study of endogenous basal levels.

Thanks to the help of instrument and reagent system manufacturers, the statistical battles of quality control in ligand assay and therapeutic drug monitoring have largely been woon for common analytes. Now it is up to concerned laboratorians to win the war by providing interpretable data free of artifactual blunders.
COPYRIGHT 1986 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1986 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:therapeutic drug monitoring
Publication:Medical Laboratory Observer
Date:Oct 1, 1986
Previous Article:Strategic management in blood banking.
Next Article:A timed blood ordering system; requisition forms with a new vocabulary, linking turnaround time to patient need for transfusion, eliminated blood...

Related Articles
A single-instrument approach to TDM toxicology.
Quality control in the new environment.
Quality control in the new environment: QC materials; an ideal control closely simulates patient specimens, is stable, and comes in large homogeneous...
Starting a pharmacokinetic dosing service.
Future challenges for the clinical immunology laboratory.
3-Dimensional Pharmaceuticals and Scriptgen settle patent dispute.
Arterial versus venous reference ranges.
Vertex Pharmaceuticals receives United States patent covering assay technology to accelerate drug discovery targeting hepatitis C protease.
Resolving discordant samples in clinical laboratory practice.
Trends in therapeutic monitoring of immunosuppressive drugs.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters