| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Experimental and Clinical Medicine, Section of Nuclear Medicine University, 44100 Ferrara, Italy;
2
Hypertension and Clinical Physiology Center, Institute of Clinical Medicine University, 20100 Milan, Italy;
3
Endocrinology Section University, 60100 Ancona, Italy;
4
Central Laboratory RIA Section, S. Camillo Hospital, 00100 Rome, Italy;
5
Institute of Clinical Physiology, National Center of Research, 56100 Pisa, Italy
aaddress correspondence to this author at: Department of Experimental and Clinical Medicine, Section of Nuclear Medicine, Via Luigi Borsari, 46, 44100 Ferrara, Italy; fax 39-0532-236589, e-mail pif{at}unife.it
 |
Introduction |
|---|
 Top Introduction References |
|---|
A pilot study (11)(12) in a limited number of laboratories (eight Italian centers) found satisfactory indexes of precision for a direct immunoradiometric renin assay (Eria Diagnostics, Sanofi-Pasteur). In our previous experience, the inter- and intralaboratory reproducibility indices of this commercial immunoradiometric assay appeared to be better than those that had been achieved with an enzymatic assay (REN-CTK; Sorin Biomedica), with the differences probably being attributable to the greater complexity of the procedure for the latter assay (11).
Various reports have been published during the past 30 years describing useful tools to monitor the performance of immunoassays (13)(14). Regarding the active renin assay, it is of paramount importance to assess the methodologic accuracy, especially for "low renin" concentrations, which are clinically relevant (15)(16).
Here we describe the results of a supraregional quality-control program, open to laboratories performing direct renin assays with a commercially available immunoradiometric method. The program included 48 clinical laboratories studied over a 12-month period with results provided periodically to the participants.
Forty-eight Italian laboratories participated in the first phase of the program, which included the determination of nine reference samples of unknown renin content (one every 3 weeks) during a 6-month period. Once a month a report with the extrapolated "consensus mean" and bias (expressed as SD) of the laboratory was sent to every participant, and at the end of each program a final report carried "precision" and "accuracy" results for the nine sample determinations. No results of this multicenter survey were rejected as outliers.
For the first phase of the program (June 1998 through January 1999), reference samples covering a wide concentration range were prepared from three pools of plasma samples collected for routine assays of renin activity. The three pools contained the following theoretic concentrations: pool 1, 15 ng/L; pool 2, 50 ng/L; pool 3, 150 ng/L. Each vial contained 0.7 mL of lyophilized plasma. The expected renin concentration of each vial was not revealed to the participating laboratories.
Lyophilization (Edwards High Vacuum MFD 0.1) carried out under controlled conditions (pressure, 1 x 10-9 Pa; temperature, -20 °C; duration, 24 h) did not significantly influence the renin content of the samples.
Forty-nine laboratories participated in the second phase of the program (June 1999 through January 2000). The reference samples for this phase were prepared from normal plasma with or without the addition of the renin International Reference Preparation (IRP 68/356), which was established as the IRP for renin by the WHO Expert Committee on Biological Standardization (WHO ECBS–WHO TRS 565) (17). The pools were as follows: pool 1 contained normal plasma with 50 ng/L renin IRP [1 ng of IRP 68/356 is equivalent to 1.6 x 10-3 Goldblatt Units (GU), with 1 GU being identical to the Medical Research Council International Unit]; pool 2 contained normal plasma; pool 3 contained normal plasma with 110 ng/L renin IRP. The theoretic renin concentrations were 50, 1.7, and 110 ng/L for pools 1, 2, and 3, respectively.
As in the first phase of the quality-control program, each pool was divided into three identical series of vials for a total of nine vials per set (each vial contained 0.75 mL of lyophilized plasma). The lyophilization process was the same as for the first phase.
This quality-control program was open to laboratories performing renin assays with the commercial immunoradiometric assay provided by Eria Diagnostics (Sanofi-Pasteur). The first monoclonal antibody recognizes both the active and inactive forms of renin, whereas the second monoclonal antibody, labeled with 125I, recognizes specifically the active form of renin. The tubes were counted in a gamma counter (Cobra AutoGamma; Camberra Packard), and the renin concentration of unknown samples was calculated from a calibration curve generated with Medical Research Council human renin (1 ng = 1.6 x 10-3 GU; WHO 68/356). The detection limit of the assay was <1 ng/L for active renin at the half-life of the tracer. The interassay variation given by the manufacturer for renin concentrations of 8–269 ng/L was 14.5–2.7%. Each laboratory also periodically received its accuracy indices with respect to a working consensus mean.
Interlaboratory CVs during the first phase were 14–32% (Table 1
); on the other hand, mean intralaboratory CVs for the same round, calculated for every laboratory on the three concentrations assayed three times, were 14–19% (Table 1
). During the second phase, interlaboratory CVs were 16–46% (Table 1
), and intralaboratory CVs for the second phase, calculated in the same way as for the first phase, were 16–30% (Table 1
). The renin concentration range for the second phase included lower concentrations to allow evaluation of precision performance in this portion of the analytical interval.
|
The survey presented here involved a large group of participating laboratories (n = 48) applying the same methodology. The long period of time over which the program was applied (12 months over 2 years) increased our confidence in the reliability of the results achieved.
Although the present overall results confirm the previous good reproducibility data obtained with the same immunometric method (11), the data reveal marked imprecision for lower renin concentrations. Blood concentrations of renin are particularly valuable and useful at low concentrations (18) and when the renin-angiotensin system is activated (15). Because various immunometric assays for active renin are used for clinical research (15)(16)(19), it would be advisable to validate each technique with a thorough quality-control program.
|
Acknowledgments |
|---|
|
References |
|---|
|
TopIntroductionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
