Letters - Clinical Chemistry

5 downloads 172127 Views 68KB Size Report
Panteghini M, Gerhardt W, Apple FS, Dati F,. Ravkilde J, Wu AH. ... Apple FS, Wu AH, Jaffe AS. European Society of ... based on our customer service records.
Letters Hemolysis Interference in the OrthoClinical Diagnostics Vitros ECi cTnI Assay

To the Editor: Assay specificity is an important requirement for cardiac troponin measurement (1 ). This study examines the effect of hemolysis on two troponin I (cTnI) assays: the OrthoClinical Diagnostics Vitros® ECi and the Beckman-Coulter Access 2 (AccuTnI). Erythrocyte hemolysates were prepared according to the procedure of Meites (2 ). Aliquots were added to unhemolyzed, heparinized plasma pools with three different cTnI concentrations (for the ECi, 0.05, 0.1, and 0.3 ␮g/L; for the AccuTnI, 0.1, 0.2, and 0.5 ␮g/L) to produce severely hemolyzed specimens. Each of these specimens was diluted in the respective unhemolyzed plasma pool to form nine samples. The hemolysis index (HI; Roche 917 analyzer) and cTnI (ECi and Access) were measured in all samples. The magnitude of hemolysis interference varied with the cTnI concentration. Multiple linear regression using the HI and measured cTnI in the hemolyzed sample as variables to predict true cTnI concentration (the original pool cTnI concentration) was performed with Sigmastat 2.0 (Jandel Corporation). A single equation (r ⫽ 0.91) was sufficient for the Access, whereas two equations (r ⫽ 0.90 and 0.92) covering different HI ranges were needed for the ECi. At

a HI of 9000 mg/L, models based on these equations predicted maximum differences in cTnI measurements of 0.33 ␮g/L for the ECi assay and 0.02 ␮g/L for the Access assay. All samples received by the laboratory for routine ECi cTnI analysis for 7 days were also analyzed for HI and with the Access cTnI. The classification of individual results at different cTnI cutoffs, with and without hemolysis correction using the above equations, was calculated. Samples with cTnI concentrations higher than or equal to the cutoff before hemolysis correction but below it after correction were considered to be false positives. We received 605 heparinized plasma samples for ECi cTnI analysis, of which 598 had sufficient volume for additional Access cTnI and HI measurements. Of these 598 samples, 6.2% were hemolyzed (4.2% with a HI of 1000 –2500 mg/L, 1.7% with a HI of 2500 –5000 mg/L, and 0.3% with a HI ⬎5000 mg/L). Table 1 shows the percentage of false positives (number of false positives/total number of uncorrected cTnI results above cutoff) at different cutoffs (3 ). Although there were no false-positive cases at any cutoff concentration with the Access cTnI assay, the false-positive rate began to increase above zero at cutoffs ⱕ0.22 ␮g/L with the ECi cTnI assay. This suggests that regardless of precision performance, cutoffs ⬍0.22 ␮g/L should not be used for the present formulation of the ECi cTnI assay. This study demonstrates that he-

molysis interference can potentially lead to false-positive cTnI results in the Vitros ECi assay. In practice, correction of cTnI results for hemolysis by use of mathematical models is not recommended; either analysis of the sample by a method unaffected by hemolysis or collection of a new, unhemolyzed sample is preferable. Other assays have shown negative interference (4, 5 ) and raise the possibility of false-negative results, which are arguably more dangerous. Laboratories should consider not only precision performance (1 ) but also the effect of factors such as hemolysis when selecting methods and setting cutoffs.

I thank Beckman-Coulter Singapore Pte. Ltd. for loan of the Access II instrument and the supply of reagents.

References 1. Panteghini M, Gerhardt W, Apple FS, Dati F, Ravkilde J, Wu AH. Quality specifications for cardiac troponin assays. Clin Chem Lab Med 2001;39:175–9. 2. Meites S. Reproducibly simulating hemolysis, for evaluating its interference with chemical methods. Clin Chem 1973;19:1319. 3. Apple FS, Wu AH, Jaffe AS. European Society of Cardiology and American College of Cardiology guidelines for redefinition of myocardial infarction: how to use existing assays clinically and for clinical trials. Am Heart J 2002;144:981– 6. 4. ver Elst KM, Chapelle JP, Boland P, Demolder JS, Gorus FK. Analytic and clinical evaluation of the Abbott AxSYM cardiac troponin I assay. Am J Clin Pathol 1999;112:745–52. 5. Dasgupta A, Wells A, Biddle DA. Negative interference of bilirubin and hemoglobin in the MEIA troponin I assay but not in the MEIA CK-MB assay. J Clin Lab Anal 2001;15:76 – 80.

Table 1. False-positive rates attributable to hemolysis at different cTnI cutoffs. OCDa ECi cTnI

Limit of detection 99th percentile reference limit Concentration corresponding to 10% CV WHO AMI cutoff a

Cutoff, ␮g/L

False-positive rate, %

Cutoff, ␮g/L

False-positive rate, %

0.038 0.08 0.12 0.4

16.0 7.5 4.1 0

0.01 0.04 0.06 0.5

0 0 0 0

OCD, Ortho-Clinical Diagnostics; AMI, acute myocardial infarction.

1226

Beckman-Coulter Access 2 cTnI

Clinical Chemistry 49, No. 7, 2003

Robert C. Hawkins Department of Pathology and Laboratory Medicine Tan Tock Seng Hospital 11 Jalan Tan Tock Seng Singapore 308433 Fax 65-6-2536507 E-mail [email protected]

Clinical Chemistry 49, No. 7, 2003

Ortho-Clinical Diagnostics offers the following reply: To the Editor: Interference by hemoglobin is clearly labeled in the VITROS® Troponin I assay’s “Instructions For Use” and Package Insert. The hemoglobin concentrations used and the differences measured are stated under “Limitations of the Procedure” (1 ). Our upper reference limit (URL) study used a total of 768 fresh heparin-plasma samples from healthy individuals, which were collected at four different centers and tested with the VITROS Troponin I assay to establish a reference interval for healthy individuals and to validate the product claims in the Package Insert and Instructions For Use. No samples were excluded because of hemolysis, and only two samples (0.25%) were above the URL of 0.08 ␮g/L (ng/mL). The incidence of hemolyzed samples in Dr. Hawkins’ study appears to be higher than our experience based on our customer service records. A hemoglobin concentration of 1000 mg/L (100 mg/dL) causes substantial discoloration of the sample, which can be easily observed by most laboratory technicians and therefore flagged for potential interferences. We recommend that customers continue to use the cutoffs stated in our labeling for the VITROS Troponin I, i.e., 0.08 ␮g/L as the URL and 0.4 ␮g/L as the cutoff for acute myocardial infarction. Use of the cutoff of 0.22 ␮g/L suggested by the author may lead to false negatives, which are clearly less desirable from a medical point of view than the false positives that may result from a small number of greatly hemolyzed samples that have not been excluded by good laboratory practice.

3.0. Raritan, NJ: Ortho-Clinical Diagnostics, December 11, 2002:9.

Rebecca Reeves Ortho-Clinical Diagnostics 1001 US Hwy 202 Raritan, NJ 08869

Reliability of IFCC Method for Lactate Dehydrogenase in Heparin Plasma

To the Editor: I wish to clarify the statement in our report (1 ) that the Roche lactate dehydrogenase (LD) method “does not provide reliable results” in heparin plasma. The report showed a problem observed only during primary-tube sampling from heparinplasma tubes from Becton Dickinson. Serum and EDTA plasma caused no problems, suggesting that the method otherwise performed well. We became aware of data from another laboratory that used Sarstedt sampling tubes and that showed none of the errors we had reported. From additional experiments (manuscript in preparation), we believe that the previously reported errors might be eliminated by adding a predilution step in the analytical sequence. This predilution step now is included in the IFCC method from Roche.

Reference 1. Bakker AJ, Mirchi B, Dijkstra JT, Reitsma F, Syperda H, Zijlstra A. IFCC method for lactate dehydrogenase measurement in heparin plasma is unreliable. Clin Chem 2003;49: 662– 4.

Andries J. Bakker Reference 1. Ortho-Clinical Diagnostics. VITROS Immunodiagnostic Products Troponin I Reagent Pack. Instructions for use. Publication No. J03812, Ver.

Department of Clinical Chemistry Klinisch Chemisch Laboratorium PO Box 850 8901 BR Leeuwarden, The Netherlands

1227

Mass Spectrometry-based Diagnostics: The Upcoming Revolution in Disease Detection Has Already Arrived

To the Editor: I read with great interest the editorial entitled “Mass Spectrometrybased Diagnostics: The Upcoming Revolution in Disease Detection” by Petricoin and Liotta (1 ). These authors provided provocative commentary on current and new mass spectrometry (MS)-based approaches in clinical chemistry related to detection and characterization of various diseases, with emphasis on proteomics and cancer diagnostics. They discuss MS as disruptive or nonlinear because of the excitement and fear that are generated by its use. The editorial correctly implies that the potential contributions of MS in clinical chemistry are staggering. I suggest that these authors have a somewhat narrow view of MS applications in clinical chemistry limited to their area of expertise. Furthermore, I would disagree somewhat with the statement that MS-based approaches are disruptive and nonlinear. Petricoin and Liotta omitted the extensive historic role of MS in clinical chemistry, including the numerous contributions and advancements in the previous two decades. Bruns et al. accurately describe the recent and important contributions of mass spectrometry and included timely manuscripts in the text, Molecular Testing in Laboratory Medicine (2 ). Dozens of other articles have appeared in Clinical Chemistry and other journals regarding MS applications in clinical chemistry, as evidenced in surveys of the literature that I and others have written recently (3, 4 ). The purpose of this letter is not to detract from the importance of the editorial or its content, but rather to perhaps provide the reader with a few more important ideas and to