Printer Friendly

Compatibility of the Abbott IMx homocysteine assay with citrate-anticoagulated plasma and stability of homocysteine in citrated whole blood.

Hyperhomocysteinemia is widely regarded as a risk factor for arterial thrombosis (1, 2), and it is also implicated as a risk factor for venous thrombosis (3-6). Therefore, homocysteine (Hcy) often is included in hypercoagulability evaluations (7). The Abbott IMx Hcy fluorescence polarization immunoassay instructions recommend EDTA- or lithium heparin-anticoagulated plasma or serum, whereas other coagulation tests are generally performed on citrate-anticoagulated specimens. To simplify specimen collection and avoid unnecessary phlebotomy, we investigated the compatibility of the Abbott IMx Hcy assay with citrate-anticoagulated plasma. Previous reports have suggested that acidic citrate stabilizes the Hcy concentration in whole-blood specimens for at least 6 h at room temperature (8, 9). Therefore, we also studied the stability of Hcy in whole-blood specimens collected in sodium citrate.

To evaluate the correlation between citrate and EDTA Hcy values, 114 sets of paired specimens were concurrently obtained from 96 nonfasting individuals (87 healthy volunteer platelet donors and 9 patients undergoing hypercoagulability evaluation) in Becton Dickinson Vacutainer lavender-top [tripotassium EDTA ([K.sub.3]EDTA)] and blue-top (3.2% sodium citrate) tubes. For each patient, the citrate tube was drawn immediately before the EDTA tube, in accordance with the NCCLS guidelines. The conditions and timing for specimen collection, processing, storage, and assay performance were identical for the two anticoagulants. The specimen pairs were separated from the cells within 30 min of phlebotomy unless specimen transportation would be longer than 30 min, in which case specimen pairs were placed immediately on ice and separated from cells within 4 h. Plasma samples were then kept frozen at -20[degrees]C or lower until blinded analysis using a standard Abbott IMx analyzer and a single lot of Abbott Hcy reagents (kindly donated by Dr. Jessie Shih, Abbott Diagnostics, Abbott Park, IL) (10). Five pairs of specimens from healthy donors were omitted because of errors or delays in processing or collection, leaving 109 specimen pairs for analysis.

To compare the stability of Hcy in citrate vs EDTA-whole-blood specimens, five paired [K.sub.3]EDTA and citrate specimens were concurrently drawn from each of seven healthy volunteers. Plasma was separated from the cells within 30 min of phlebotomy and after 2, 4, 8, and 24 h at room temperature. Storage and analysis were as above. Informed consent from all volunteer donors and Institutional Review Board approval were obtained for this study. The paired Student t-test (two-tailed) was used to calculate statistical significance.

Among the 109 paired samples in the correlation study, Hcy concentrations were 4.8-17.8 [micro]mol/L in EDTA samples and 4.0-16.7 [micro]mol/L in citrate samples. One obvious outlier point was eliminated from further analysis. For the healthy donors, the mean and median Hcy concentrations and nonparametric reference interval (2.5th to 97.5th percentile) were 8.1, 7.9, and 5.0-12.3 [micro]mol/L, respectively, in EDTA and 6.9, 6.9, and 4.3-10.3 in citrate (P <0.000001 for the difference between EDTA and citrate). The proportional bias can be seen in Fig. 1A, and in the following equation calculated by standard linear regression:

[Hcy.sub.citrate] = 0.85([HCY.sub.EDTA]) + 0.13 [micro]mol/L

([r.sup.2] = 0.839; 95% confidence intervals (CIs) for slope and intercept, 0.78-0.92 and -0.48 to 0.73 [micro]mol/L, respectively).

Some difference between citrate and EDTA results was expected because citrate tubes contain a larger volume of anticoagulant than do EDTA tubes, which produces a slight dilutional effect on the citrate results. The [K.sub.3]EDTA tubes draw 10 mL of whole blood and contain 0.117 mL of EDTA solution (150 g/L), whereas the citrate tubes draw 4.5 mL of whole blood and contain 0.5 mL of sodium citrate (105 mmol/L). To correct for the difference in dilution of whole-blood volume, citrate results were multiplied by a correction factor, [C.sub.B]:

[C.sub.B] (blood volume, EDTA) (blood + anticoagulant volume, citrate)/(blood volume, citrate)(blood + anticoagulant volume, EDTA) =

(10)(5)/(4.5)(10.117) = 1098

Multiplication of the citrate results by 1.098 improved the agreement between citrate and EDTA results. However, the difference between EDTA and corrected citrate results remained significant (P = 0.0000017), and some proportional bias remained, as seen in Fig. 1B and in the following linear regression equation:

[Hcy.sub.citrate] = 0.93([Hcy.sub.EDTA]) + 0.14 [micro]mol/L

([r.sup.2] = 0.839; 95% CIs for slope and intercept, 0.85-1.01 and -0.52 to 0.80 [micro]mol/L, respectively).

The correction factor, [C.sub.B], corrects for the dilution of whole blood by the anticoagulant in the tube. However, if the citrate and EDTA anticoagulant solutions remain extracellular, then the correction factor should only reflect the centrifuged specimen plasma volume and not the whole blood volume:

[C.sub.P] = (plasma volume, EDTA)(plasma + anticoagulant volume, citrate)/(plasma volume, citrate)(plasma + anticoagulant volume, EDTA)

Assuming a plasma volume of 57% (hematocrit, 43%), the revised correction factor, [C.sub.P], is 1.171. When we used this correction factor, the differences between citrate and EDTA results were no longer significant (P = 0.4), and the proportional bias was essentially eliminated [[Hcy.sub.citrate] = 0.99([HCY.sub.EDTA]) + 0.15 [micro]mol/L; [r.sup.2] = 0.839; 95% CIs for slope and intercept, 0.91-1.08 and -0.56 to 0.86, respectively; not shown].

Hematocrit values for the specimens were not available to determine individual specimen plasma volumes, but gender and hemoglobin concentrations were known. Therefore, two additional types of plasma correction factors were calculated: gender- and hemoglobin-based. The gender-based correction factors, assuming average hematocrits of 40% for females and 46% for males, were 1.163 and 1.18, respectively (Fig. 1C). The hemoglobin-based correction factor, calculated assuming that hematocrit is three times the hemoglobin (g/dL), was 1.177 on average, with a range of 1.157-1.218 (Fig. 1D). Both types of correction essentially eliminate the proportional bias: [Hcy.sub.citrate] = 1.00([HCY.sub.EDTA]) + 0.10 [micro]mol/L ([r.sup.2] = 0.839) for gender; [Hcy.sub.citrate] = 0.99([HCY.sub.EDTA]) + 0.20 [micro]mol/L ([r.sup.2] = 0.839) for hemoglobin (95% CIs for slope and intercept, 0.92-1.08 and -0.62 to 0.81 [micro]mol/L for gender-based and 0.90-1.07 and -0.52 to 0.92 [micro]mol/L for hemoglobin-based corrections, respectively). There was no significant difference between these corrected citrate and EDTA results (P = 0.3 and 0.2, respectively). The Hcy reference interval (nonparametric range from 2.5th to 97.5th percentile) in citrate when the [C.sub.P] (hematocrit, 43%) or the hemoglobin- or gender-based correction factor was used was 5.0-12.0 [micro]mol/L. We recommend the gender-based correction factor for citrate Hcy measurements because patient gender is generally provided with all clinical specimens, the residual bias is negligible and perhaps slightly less than when the single plasma volume correction factor [C.sub.P] is used, and its use is simpler than the hemoglobin-based approach. We suspect that such a correction factor would be appropriate for correlating most analyses performed in both citrate and EDTA, but each assay should be evaluated for specific interactions of the anticoagulant with analytes or reagents (11).

Several authors have compared chemistry assay performance in sodium citrate and EDTA, and generally a negative proportional bias has been noted in which citrate results underestimate EDTA results, similar to the unadjusted results in the present study. Among the explanations for this effect are whole blood dilution (12, 13), osmotic effects (14), and specific interactions with the analytes (15). In the present study, the proportional bias was best eliminated by correcting for differences in plasma dilution.

In the stability study, no significant difference was evident in the stability of Hcy in blood over 24 h in the two anticoagulants (Table 1). Initial Hcy concentrations were 6.1-10.8 [micro]mol/L in EDTA and 5.5-10.1 [micro]mol/L in citrate (uncorrected). In both anticoagulants, an upward trend in Hcy concentration was evident starting at 4 h. Acidic citrate prevented this upward trend in previous studies (8,9). In contrast to the standard sodium citrate tubes used in the present study, acidic citrate tubes contain 0.5 mL of 0.5 mol/L citrate acidified to a pH of 4.3.


In summary, Hcy concentrations measured by the Abbott method in EDTA- and citrate-anticoagulated plasma agree well after a plasma-volume correction factor is taken into consideration. Specifically, citrate Hcy results agree well with EDTA results if the citrate value is multiplied by 1.163 for females and 1.18 for males. No difference in specimen stability over time was evident in citrate tubes compared with EDTA tubes.

We wish to thank the platelet donors and the staff of the Massachusetts General Hospital Blood Transfusion Service and Amino Acid Laboratory for their invaluable assistance in this study. We thank Dr. Jessie Shih (no relation to author) of Abbott Diagnostics for kindly donating the Abbott reagents used in this study.


(1.) Cattaneo M. Hyperhomocysteinemia, atherosclerosis and thrombosis. Thromb Haemost 1999;81:165-76.

(2.) Welch GN, Loscalzo J. Homocysteine and atherothrombosis. New Engl J Med 1998;338:1042-50.

(3.) den Heifer M, Rosendaal FR, Blom HJ, Gerrits WB, Bos GM. Hyperhomocysteinemia and venous thrombosis: a meta-analysis. Thromb Haemost 1998; 80:874-7.

(4.) den Heifer M, Blom HJ, Gerrits WBJ, Rosendaal FR, Haak HL, Wiiermans PW, Bos GM. Is hyperhomocysteinemia a risk factor for recurrent venous thrombosis? Lancet 1995;345:882-5.

(5.) den Heifer M, Koster T, Blom HJ, Bos GM, Briet E, Reitsma PH, et al. Hyperhomocysteinemia as a risk factor for deep-vein thrombosis. New Engl J Med 1996;334:759-62.

(6.) Simioni P, Prandoni P, Burlina A, Tormene D, Sardella C, Ferrari V, et al. Hyperhomocysteinemia and deep-vein thrombosis. A case control study. Thromb Haemost 1996;76:883-6.

(7.) Van Cott EM, Laposata M. Laboratory evaluation of hypercoagulable states. Hematol Oncol Clin North Am 1998;12:1141-66.

(8.) Willems HP, Bos GM, Gerrits WB, den Heifer M, Vloet S, Blom HJ. Acidic citrate stabilizes blood samples for assay of total homocysteine. Clin Chem 1998;44:342-5.

(9.) Salazar JF, Herbeth B, Siest G, Leroy P. Stability of blood homocysteine and other thiols: EDTA or acidic citrate? Clin Chem 1999;45:2016-9.

(10.) Shipchandler MT, Moore EG. Rapid, fully automated measurement of plasma homocyst(e)ine with the Abbot IMx analyzer. Clin Chem 1995;41: 991-4.

(11.) Chuang CK, Lin SP, Lin YT, Huang FY. Effects of anticoagulants in amino acid analysis: comparisons of heparin, EDTA, and sodium citrate in Vacutainer Tubes for plasma preparation. Clin Chem 1998;44:1052-6.

(12.) Rodriguez-Mendizabal M, Lucena MI, Cabello MR, Blanco E, Lopez-Rodriguez B, Sanchez de la Cuesta F. Variations in blood levels of aminoglycosides related to in vitro anticoagulant usage. Ther Drug Monit 1998;20:88-91.

(13.) Holten-Andersen MN, Murphy G, Nielsen HJ, Pedersen AN, Christensen IJ, Hoyer-Hansen G, et al. Quantitation of TIMP-1 in plasma of healthy blood donors and patients with advanced cancer. Br J Cancer 1999;80:495-503.

(14.) Lippi G, Giampaolo L, Guidi G. Effects of anticoagulants on lipoprotein(a) measurements with four commercial assays. Eur J Clin Chem Clin Biochem 1996;34:251-5.

(15.) Wiese J, Didwania A, Kerzner R, Chernow B. Use of different anticoagulants in test tubes for analysis of blood lactate concentrations. Part 2. Implications for the proper handling of blood specimens obtained from critically ill patients. Crit Care Med 1997;25:1847-50.

Darryl E. Palmer-Toy, * Zbigniew M. Szczepiorkowski, Vivian Shih, and Elizabeth M. Van Cott ([dagger])

(Massachusetts General Hospital, Department of Pathology, Division of Laboratory Medicine, Boston, MA 02114; * Current address: Johns Hopkins School of Medicine, Department of Pathology, 600 N. Wolfe St./Meyer B-125, Baltimore, MD 21287; ([dagger]) address correspondence to this author at: Coagulation Laboratory, Massachusetts General Hospital, Department of Pathology, Division of Laboratory Medicine, Gray-Jackson 235, 55 Fruit St., Boston, MA 02114; fax 617-726-7758, e-mail
Table 1. Stability of Hcy in EDTA- and citrate-
anticoagulated whole blood.

 Fold increase in Hcy, mean t SD

Hours post
collection (a) EDTA (b) Citrate (c)

 0 1.0 1.0
 2 1.1 [+ or -] 0.1 1.0 [+ or -] 0.2
 4 1.3 [+ or -] 0.2 1.2 [+ or -] 0.2
 8 1.5 [+ or -] 0.2 1.6 [+ or -] 0.5
 24 1.9 [+ or -] 0.2 1.8 [+ or -] 0.4

(a) For each time point, n = 7 healthy volunteers.

(b) Relative to EDTA result from immediate analysis.

(c) Relative to citrate result from immediate analysis.
COPYRIGHT 2001 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2001 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Technical Briefs
Author:Palmer-Toy, Darryl E.; Szczepiorkowski, Zbigniew M.; Shih, Vivian; Van Cott, Elizabeth M.
Publication:Clinical Chemistry
Date:Sep 1, 2001
Previous Article:Anodic stripping voltammetry compared with graphite furnace atomic absorption spectrophotometry for blood lead analysis.
Next Article:Serum C-reactive protein in Canadian Inuit and its association with genetic variation on chromosome 1q21.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters