| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1 National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA 30341
aauthor for correspondence; fax 770-488-4609, e-mail cpfeiffer{at}cdc.gov
Determination of total homocysteine (tHcy) in plasma is becoming an important diagnostic procedure in clinical chemistry because a slightly increased concentration of tHcy in plasma has been discussed as an important independent risk factor for atherosclerotic diseases (1). Many methods, mostly by HPLC, have been reported for measuring tHcy (2). However, these methods are relatively complex and require highly specialized equipment. Drew Scientific, Inc. (company named for identification purposes only; this evaluation does not constitute an endorsement by CDC) has developed the DS30 tHcy system for measuring tHcy in plasma. We evaluated this new system and compared it with our CDC reference HPLC method (3).
The DS30 tHcy system comprises a small HPLC system using a 5-cm
reversed-phase column, a tHcy assay reagent set that contains the
necessary reagents and calibrators [5 and 20 µmol/L homocystine
(concentration was equivalent to the free thiol)] and a
quality-control (QC) set containing two concentrations of Hcy. The
sample preparation requires 200 µL of plasma. A batch of 30 samples
can be processed within 90 min. After the addition of 10 µL of the
internal standard (IS; 2-mercaptoethylamine) to 200 µL plasma, the
mixture of disulfides, mixed disulfides, and protein-bound thiols is
reduced using 20 µL of tris(2-carboxyethyl)phosphine (TCEP). Protein
is precipitated from this solution with trichloroacetic acid, and 100
µL of the supernatant is then derivatized with a fluorescent
thiol-specific dye [ammonium
7-fluorobenzo-2-oxa-1,3-diazole-4-sulfonate (SBDF) in EDTA/borate
buffer] at 60 °C for 50 min. The thiol derivatives are separated in
a subsequent step by HPLC and detected by their fluorescence. The total
run time for a batch of 30 samples is 
5–6 h. Quantitative analysis
is achieved using a two-point calibration curve with homocystine in an
aqueous matrix.
We assessed the within-assay and between-assay imprecision, the recovery of homocystine added to plasma, the linearity of increasing tHcy concentration in plasma, the dilution linearity, and the limit of detection. We also compared the results of 260 plasma samples obtained by the DS30 system to the results we obtained previously for these samples by our CDC in-house HPLC method. These samples were a subset of the EDTA plasma samples for the National Health and Nutrition Examination Survey 1999+, which includes an omnibus informed consent and Human Subjects Review protocol.
The within-assay imprecision (CV) for five replicate measurements was <2% for the two concentrations of DS30 QC samples and the three concentrations of CDC QC pools. The between-assay imprecision was as follows: 5.8% for DS30 QC low (n = 21 days; mean tHcy, 13.4 µmol/L), 4.6% for DS30 QC high (n = 21 days; mean tHcy, 23.3 µmol/L), 3.8% for CDC QC low (n = 27 days; mean tHcy, 7.3 µmol/L), 3.2% for CDC QC medium (n = 27 days; mean tHcy, 14.7 µmol/L), and 4.8% for CDC QC high (n = 16 days; mean, tHcy, 31.8 µmol/L).
Homocystine was added to a plasma sample at six different concentrations in duplicate: 0, 6.25, 12.5, 25, 50, and 100 µmol/L (concentrations equivalent to the free thiol). The mean recovery (SD) was 101.7% ± 1%. The linearity of increasing tHcy concentrations in plasma was very good up to 100 µmol/L (y = 1.028x - 0.127; r2 = 1.000).
The DS30 HPLC system is programmed such that peak recognition of tHcy
and the IS depends on a minimum concentration of cysteine and
cysteinyl-glycine. We found that this minimum concentration was 50
µmol/L cysteine and 7.5 µmol/L cysteinyl-glycine. Thus, dilution of
plasma >1:4 with saline or water leads to a loss of recognition of the
peaks. Dilution of a plasma sample (tHcy concentration of 30
µmol/L) with a solution containing 200 µmol/L cysteine and 30
µmol/L cysteinyl-glycine [these concentrations of cysteine and
cysteinyl-glycine correspond to an average concentration found in the
population (4)] at 1:2, 1:4, 1:8, and 1:16 gave very good
linearity (y = 0.999x + 0.115;
r2 = 0.999). The difference between
the measured and the expected concentrations was <3%. Diluting the
same plasma with either saline or water at 1:2 and 1:4 also gave very
good linearity; however, the difference between the measured and the
expected concentrations was 5–10%. The limit of detection for tHcy
was 2 µmol/L.
The retention times for each compound were stable over the 24 assays we
performed over 4 months. The variation of the retention times for all
compounds was 3%. The variation (CV) of the IS heights for all
samples, except for the DS30 QC samples, over the period of evaluation
was 10.5%. The DS30 QC samples gave consistently lower heights
(25% lower) for the IS than all other samples. The reason for this
was that the DS30 QC samples were not EDTA plasma, but heparin plasma,
and apparently the anticoagulant influenced the obtained peak height.
However, because ratios between tHcy and the IS heights were used for
quantification, this should not have influenced the final result.
However, users should be alerted that we tested only the suitability of
the DS30 assay for EDTA plasma samples. We believe the use of this
assay for matrices other than EDTA plasma should be evaluated first.
An extensive method comparison was performed between the DS30 tHcy
system and the CDC in-house HPLC method. If samples were not analyzed
simultaneously with both methods, they were stored at -70 °C for
not >6 months between the assays. The performance of the CDC reference
HPLC assay and the validation results were described in detail in a
separate article (3). tHcy concentrations determined for 260
plasma samples from healthy subjects gave good correlation between the
two methods: y = 1.018x + 0.280
(r2 = 0.976). The Bland–Altman plot
in Fig. 1
is shown with the bias between the two methods by plotting for
each sample the difference of results from the two methods
(y axis) compared with the mean of results (x
axis). The mean difference for the DS30 showed a slight positive bias
of 0.39 µmol/L (95% confidence interval, 0.30–0.49). The central
0.95 interval (mean difference, ± 2 SD) indicates the agreement
between the two methods. Ninety-five percent of tHcy determinations by
DS30 were 1.10–1.89 µmol/L higher than concentrations determined by
the CDC HPLC method.
|
In conclusion, the DS30 tHcy system showed within-assay and between-assay imprecision (CV <6%) comparable to other frequently used HPLC assays (3)(4)(5) and the Abbott IMx assay (6). The DS30 tHcy system also showed complete recovery of added tHcy and a linearity up to 100 µmol/L. Dilution of plasma with a solution containing cysteine and cysteinyl-glycine produced very good linearity. However, dilutions of plasma with water or saline should not exceed 1:4 to avoid misidentification of peaks resulting from undetectable cysteine and cysteinyl-glycine peak heights. Samples with a tHcy concentration up to maximum 400 µmol/L can be measured after 1:4 dilution with water or saline. However, we recommend verifying the printout with regard to peak identification, retention times, and peak heights to avoid mislabeling peaks (especially with diluted samples). The comparison of this method with the CDC reference HPLC method on samples up to 25 µmol/L tHcy revealed only a minimal bias (0.39 µmol/L). Therefore, the DS30 tHcy system performed accurately and precisely, and thus might be well suited for routine measurement for tHcy where complex HPLC analysis is not feasible.
Acknowledgments
This study was supported by Drew Scientific, Inc., who provided the DS30 instrument as a loaner and all the reagents and columns needed to perform this evaluation. An abstract containing the summary of this evaluation has been submitted to Experimental Biology 2001 (published in FASEB J 2001;15:A613).
References
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  D. M. Hill, L. J. Johnson, P. J. Burns, A. M. Neale, D. M. Harmening, and A. C. Kenney Effects of Temperature on Stability of Blood Homocysteine in Collection Tubes Containing 3-Deazaadenosine Clin. Chem., November 1, 2002; 48(11): 2017 - 2022. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
