| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Laboratory Medicine, Municipal Ho-Pin Hospital, 33 Sec 2, Chung-Hwa Rd., Taipei, Taiwan, R.O.C.
2
School of Medical Technology, College of Medicine,
National Taiwan University, 7 Chung-Shan South Rd., 10016 Taipei,
Taiwan, R.O.C.
3
Department of Laboratory Medicine, College of Medicine,
National Taiwan University Hospital, National Taiwan University, 7
Chung-Shan South Rd., 10016 Taipei, Taiwan, R.O.C.
a Author for correspondence. Fax 886-2-3224263;
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-cyclodextrin sulfate
(PEGME) assay (Kyowa), and y = 0.84x +
106.51 mg/L, r = 0.984,
Sy|x = 26.10 mg/L (n = 152)
for the polyanion–polymer/detergent (PPD) assay (Daiichi). The
specificity of the PEGME method seemed better than that of the PPD
method, as the PPD method was markedly interfered with by supplemental
LDL-C. Addition of 20 g/L triglycerides produced a negative error of
~18% in both homogeneous assays. Bilirubin and hemoglobin had little
influence on the PEGME method; hemoglobin had little effect on the PPD
method. Bilirubin, however, markedly decreased the readings by the PPD
method. We found the PEGME assay superior to the PPD assay for routine
HDL-C testing, because the PPD assay is relatively inaccurate and not
specific.
Key Words: indexing terms: methods comparison • triglycerides • atherosclerosis
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
600 mg/L has been defined in the NCEP ATP II
as a negative risk factor (3)(5).
Concentrations of LDL-C are strongly associated with cardiovascular
risk (6)(7), which is calculated in general
practice with the Friedewald formula: LDL-C = total
cholesterol - HDL-C - TG/5 (8). This requires
that both HDL-C and TG be determined in a laboratory separately. As a
result, the demand for HDL-C determination is increasing in clinical
practice worldwide. Several techniques for determining HDL-C in serum have been described, including ultracentrifugation (9), electrophoresis (10), HPLC (11), and precipitation-based methods (12). Of these, the precipitation-based methods are widely used in the clinical laboratories. Most commonly, HDL-C is measured in the supernatant after use of dextran sulfate or phosphotungstic acid (PTA)/MgCl2 to precipitate the apolipoprotein (apo) B-containing lipoproteins (13)(14). These methods, however, require relatively large volumes of sera, are time-consuming, not suitable for automated analysis, and are interfered with by high TG concentrations (13). Recently, two new methods, one homogeneous assay (15) and one magnetic dextran sulfate assay (16), without a centrifugation step, have been developed. The former needs four different reagents, and thus can only be used on a limited number of automated analyzers; the latter needs a large sample volume (500 µL) and is not fully automated. For these reasons, convenient and reliable methods without any pretreatment are needed to determine HDL-C.
This study evaluated the linearity, detection limit, precision,
accuracy, and specificity of two convenient homogenous assays, i.e.,
the polyethylene glycol-modified enzymes/-cyclodextrin sulfate
(PEGME) assay (17) and the polyanion–polymer/detergent
(PPD) assay for HDL-C and compared them with the PTA/MgCl2
method. We also assessed the interference of TG, hemoglobin, and
bilirubin.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To assess within-run and between-run precision, we used three sets of pooled human sera with HDL-C concentrations of 280, 560, and 840 mg/L (PTA/MgCl2 method) as controls, which were stored in aliquots at -20 °C.
hdl-c assays
PTA/MgCl2 assay.
In this assay
(12)(18), 200 µL of sample is mixed with 500
µL of solution containing 0.44 mmol/L PTA and 20 mmol/L
MgCl2. After a 10-min incubation and precipitation at room
temperature, samples were centrifuged at 7000g for 10 min.
The supernatant was removed manually for assays, and then HDL-C was
determined with an enzymatic end point assay, by using cholesterol
oxidase and peroxidase and then a chromogenic reaction with
4-aminophenazone (CHOD-PAP) on a Hitachi (Tokyo, Japan) 7450 analyzer.
homogeneous assays
The PEGME method (Kyowa, Tokyo, Japan) involved two reagents in
solution form. Reagent 1 contained 0.5 g/L dextran sulfate, 2 mmol/L
MgCl2, 0.96 mmol/L sodium
N-(2-hydroxy-3-sulfopropyl)-3,5-dimethoxyaniline, and 0.5
mmol/L -cyclodextrin sulfate, which reduces the reactivity of
cholesterol, especially in chylomicrons and VLDL (17).
Reagent 2 contained 1 kU/L PEG-modified cholesterol esterase, 5 kU/L
PEG-modified cholesterol oxidase, 30 kU/L peroxidase, and 0.5 g/L
4-aminoantipyrine. These PEG-modified enzymes showed selective
catalytic activities towards different lipoprotein fractions, with the
reactivity increasing in the following order: LDL < VLDL 
chylomicrons < HDL (17). In the first step, 4 µL
of serum was incubated with 300 µL of reagent 1 at 37 °C for 5
min. Then 100 µL of reagent 2 was added and the absorbance was
measured bichromatically at 600 nm (main) and 700 nm (subsidiary).
The PPD method (Daiichi, Tokyo, Japan) also involved two reagents. Reagent 1, in solution form, contained polyanion and synthetic polymer. Reagent 2 was reconstituted before use and consisted of lyophilized enzyme containing 0.8 kU/L cholesterol esterase, cholesterol oxidase, peroxidase, and 2 mmol/L 4-aminoantipyrine, and solvent containing 1 mmol/L N,N-bis(4-sulfobutyl)-m-toluidine disodium (DSBmT), 60 mmol/L phthalic acid buffer (pH 5.7), and detergent. To assay, 3 µL of serum was combined with 300 µL of reagent 1 and incubated at 37 °C for 5 min, thereby forming LDL–polymer–polyanion and HDL–polymer complexes. Then, 100 µL of reagent 2 was added; the detergent broke down the HDL structure and HDL-C was measured enzymatically. The absorbance was measured at 37 °C at a main wavelength of 546 nm and a subsidiary wavelength of 660 nm.
These two homogeneous assays for HDL-C were performed with a Hitachi 7450 automated analyzer. The two sets of calibrators, as well as their Hitachi 7450 application protocol, were provided by the manufacturers. Each homogeneous method included the calibrating system specified by the original manufacturer.
ldl-c assay
For this analysis, we used the polyvinyl sulfate method
(Boehringer Mannheim, Mannheim, Germany) (19). Briefly, a
mixture of 200 µL of serum and 100 µL of reagent containing
polyvinyl sulfate was incubated for 15 min at room temperature, and
then centrifuged for 15 min at 1500g. After centrifugation,
the cholesterol content of the supernatant was determined by the
CHOD-PAP method. The LDL-C concentration was calculated from the
difference between the total serum cholesterol and the cholesterol in
the supernatant.
lipid measurements
Concentrations of cholesterol and TG were determined enzymatically
with the CHOD-PAP (Amresco, Solon, Ohio) and glycerophosphate
oxidase–peroxidase-4-aminophenazone (GPO-PAP) methods (Boehringer
Mannheim), respectively, on a Hitachi 7450 analyzer. The interassay CV
for determinations of total cholesterol and total TG varied between
1.02% and 2.71% and between 2.01% and 3.35%, respectively. Control
sera were purchased from Baxter Diagnostics (Deerfield, IL) and Bio-Rad
(Anaheim, CA).
isolation of vldl, ldl, and hdl
Lipoproteins were isolated by sequential ultracentrifugation on a
Beckman L8-M ultracentrifuge at 125 000g for 20 h. The
lipoproteins were identified at the following densities: <1.006 kg/L
for VLDL, between 1.030 and 1.050 kg/L for LDL, and between 1.063 and
1.21 kg/L for HDL (20).
linearity, interference, and specificity
We added various amounts of HDL fraction isolated by sequential
ultracentrifugation to lipoprotein-deficient serum obtained by
ultracentrifugation to examine the linearity capacity of both
homogeneous assays. Interferences from TG, hemoglobin, and bilirubin
were also analyzed. The serum pool, containing total cholesterol 1880
mg/L, HDL-C 600 mg/L, LDL-C 1040 mg/L, VLDL-C 130 mg/L, and TG 1180
mg/L, was supplemented with IntralipidTM (Kabi Pharmacia,
Stockholm, Sweden), hemoglobin, or bilirubin at various concentrations
and was also used for the specificity study. Reagents for evaluating
interference of hemoglobin and bilirubin were purchased from
International Reagent Corp. (Kobe, Japan). The specificity of both
homogeneous assays was evaluated by adding various amounts of VLDL and
LDL, isolated by sequential ultracentrifugation, to a serum pool. The
serum pool contains total cholesterol 1128 mg/L, HDL-C 360 mg/L, LDL-C
680 mg/L, VLDL-C 80 mg/L, and TG 708 mg/L after diluting with 9 g/L
saline.
statistical methods
The mean, SD, and CV were calculated with Microsoft Excel Version
5.0 (Microsoft, Redmond, WA). Parametric paired t-test and
linear regression analyses were calculated with Statistica (Statsoft,
Tulsa, OK). The paired t-test was significant at
P <0.05.
We calculated the total errors by adding the systematic errors and the random errors (21)(22). At an HDL-C concentration of xc, systematic error was equal to yc - xc, where yc = bxc + a (linear regression equation). Random error was 1.96 SD from the between-run precision study (22)(23). Sy|x represented the standard deviation of the regression line.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). Therefore, both assays fulfilled the performance criteria for
HDL-C imprecision of <5% established by both the CDC and the NCEP
(23).
|
The linearity studies (Fig. 1
) showed the PEGME method to be linear up to 1500 mg/L HDL-C,
with a detection limit (mean of blanks plus 3 SD of blank readings) of
4.5 mg/L HDL-C. The PPD method was linear up to 1200 mg/L HDL-C, and
the detection limit was 4.9 mg/L HDL-C. Moreover, recovery studies were
performed after adding known amounts of HDL-C to six serum samples with
HDL-C concentrations of 280–770 mg/L. Percentages of recovery for both
homogeneous methods were 90.7% to 112.5% (Table 2
).
|
|
With the 152 fresh samples, the accuracy of the two homogeneous HDL-C methods was established by comparison with the conventional PTA/MgCl2 precipitation method. The mean total cholesterol and total TG concentrations were 2232 mg/L (range 1420–4100 mg/L) and 2522 mg/L (range 300–21 480 mg/L), respectively, for these samples. Mean HDL-C concentrations were 447 mg/L by the PTA/MgCl2 method, 512 mg/L by the PEGME method, and 483 mg/L by the PPD method, with respective ranges of 105–1260, 145–1255, and 162–1096 mg/L. The paired t-test showed significant difference (P <0.05) between these two homogeneous methods and the precipitation method, indicating the existence of a systematic error.
To compare the HDL-C values obtained by the two homogeneous methods and
the PTA/MgCl2 method, samples were separated into two
groups according to their TG concentration: group I, TG <4000 mg/L,
n = 135; group II, TG 4000 mg/L, n = 17. The correlation
coefficients between the PEGME method and the PTA/MgCl2
method in both groups were excellent, but the intercepts were
relatively large: y = 0.99x + 71.53,
r = 0.987, slope = 0.99 (95% confidence interval
0.965–1.018), Sy|x = 28.35 mg/L in
group I and y = 0.98x + 56.57,
r = 0.953, slope = 0.98 (95% confidence interval
0.768–1.092), Sy|x = 20.81 mg/L in
group II (Fig. 2
A, D).
|
In group I, readings obtained by the PPD method and the
PTA/MgCl2 method showed significant deviation from the
line of identity: y = 0.84x + 105.33,
r = 0.986, slope = 0.84 (95% confidence interval
0.816–0.864), Sy|x = 24.49 mg/L
(Fig. 2B
). In group II, the PEGME method seemed to correlate better
with the precipitation method than did the PPD method
(r = 0.95 vs 0.86) (Fig. 2D
, E). We also compared the
HDL-C values of hypertriglyceridemic samples (group II, TG range
4690–21 480 mg/L, mean 13 085 mg/L) using each method before and
after ultracentrifugation with a Beckman Airfuge. The data showed that
the PEGME assay was less interfered with by lipemia than the PPD assay:
y = 1.05x - 25.23, r =
0.975, slope = 1.05 (95% confidence interval 0.928–1.172),
Sy|x = 16.30 mg/L and
y = 0.85x + 58.62, r =
0.945, slope = 0.85 (95% confidence interval 0.698–1.002),
Sy|x = 21.98 mg/L (Fig. 3
).
|
Table 3
shows the calculated total errors of both homogeneous HDL-C
assays. This calculation was based on (a) the interassay
variation presented in Table 1
, as the random errors, and
(b) the estimates of the regression lines of the 152 samples
(groups I plus II) measured in parallel by the homogeneous assays and
the PTA/MgCl2 procedure, as the systematic errors. At 280
mg/L both homogeneous assays failed to meet the forthcoming 1998 NCEP
total error goal of <13% (24) because of large
systematic errors. At 560 and 840 mg/L, these criteria were just met,
while the total error estimated was highly borderline (13.29%) for the
PEGME assay at 560 mg/L (Table 3
).
|
With the PEGME method, there were negligible changes in HDL-C readings
even after addition of 410 mg/L VLDL-C to yield a final concentration
of 490 mg/L (+3%) or after addition of 2300 mg/L LDL-C to yield a
final concentration of 2980 mg/L (+10%) (Fig. 4
). With the PPD method, HDL-C readings increased from 346 to 388
mg/L (+12%) after adding 410 mg/L VLDL-C and increased from 344 to 453
mg/L (+32%) after adding 2300 mg/L LDL-C, indicating a poorer
specificity of the PPD assay for HDL-C (Fig. 4
).
|
We plotted the differences between the results of the PPD and PEGME
methods on the one hand and that of the PTA/MgCl2
method on the other hand against the ratio of LDL-C to HDL-C (Fig. 5
). This analysis confirmed that the PPD but not the PEGME method
was susceptible to interference by LDL-C. The higher the LDL-C/HDL-C
ratio of sera, the larger the bias between the PPD method and the
PTA/MgCl2 (P <0.0001, Fig. 5B
). The same trend
did not show statistical significance with the PEGME method
(P = 0.21, Fig. 5A
).
|
Finally, analysis of the interference from TG, hemoglobin, and
bilirubin showed that addition of 20 g/L TG to the pooled serum,
baseline HDL-C concentration of 366 mg/L after diluting with 9 g/L
saline, produced a negative error in the results of both homogeneous
assays (18%). Hemoglobin up to 12 000 mg/L had little influence on
both homogeneous methods (<5%). Bilirubin up to 400 mg/L showed no
effect on the PEGME method, whereas a decrease of HDL-C up to 42% was
seen in the PPD method (Fig. 6
).
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Both homogeneous assays were precise but had a large systematic error
compared with the PTA/MgCl2 method. Warnick et al.
found that the HDL-C concentration measured by the
PTA/MgCl2 precipitation method was consistently 5% to 10%
lower than that by ultracentrifugation method, the reference method
(14). This discrepancy may be caused by the incomplete
capability to measure all the HDL-C fraction by the
PTA/MgCl2 method (25), since the precipitation
method would fail to measure the cholesterol molecules in the
high-Mr apo E-rich HDL (26). We
also found that both homogeneous assays were calibrated falsely high
(Table 4
). This could also partly explain the reasons for the
overestimation of HDL-C by the two new assays.
|
The PPD method showed a systematic deviation over the whole range, with
a positive y-intercept and a slope being significantly <1,
resulting in falsely high results for HDL-C at concentrations ~<600
mg/L and falsely low concentrations at ~>600 mg/L (Fig. 2B
). With
this method, there would be disadvantages in calculating the LDL-C
with the Friedewald formula (8). The underestimation of
the HDL-C concentrations >600 mg/L by the PPD method would result
in a reciprocal error in the LDL-C estimate and produce a considerable
shift in the ratio of LDL-C to HDL-C.
We found the PPD assays not as specific for HDL-C as desired (Fig. 4
).
This lack of specificity may be a result of incomplete wrapping of
non-HDL with the reagents, so that the enzymes used in the cholesterol
assays may also measure the cholesterol in the chylomicrons, VLDL, and
LDL, resulting in falsely high HDL-C concentrations. Only with the PPD
method did we find a statistically significant relation (P
<0.0001) between serum LDL-C/HDL-C ratios and the difference between
the results of the homogeneous assays and the PTA/MgCl2
method (Fig. 5
). This relation implies that in samples with a high
LDL-C/HDL-C ratio, falsely high HDL-C concentrations will be obtained
with the PPD method. For these samples, a lower ratio of total
cholesterol/HDL-C will ensue and give a falsely low risk of coronary
heart disease with the PPD assay.
Hemoglobin seems to have little influence on either homogeneous assay.
However, 20 g/L TG decreased the HDL-C readings by ~18% in both
assays. Its volume effect may have contributed to the negative
interference. In contrast to the PEGME method, the PPD method was
markedly interfered with by bilirubin (Fig. 6
). This could be caused by
the reaction between bilirubin and H2O2, and
the subsequent reduction of the amount of peroxide available for the
formation of the colored complex.
Both the PEGME and PPD assays have the advantages of requiring small sample volumes, convenience, and rapidity. Both also need no pretreatment and are ready for full automation. However, the systematic deviation between the two homogeneous assays and the PTA/MgCl2 assay were substantial. One of the reasons seemed to be in the calibration systems of both assays, which needed to be improved. We found the PPD assay not as specific and accurate as desired, and the PEGME method superior in many aspects. Thus, before they can be generally applied in clinical laboratories, restandardization of the calibration systems are mandatory and the specificity of the PPD assay has to be improved.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
-cyclodextin sulfate; PPD, polyanion–polymer/detergent; CHOD-PAP, cholesterol oxidase–peroxidase, 4-aminophenazone; and apo, apolipoprotein. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-cyclodextrin. Clin Chem 1995;41:717-723.
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  G. R. Warnick, M. Nauck, and N. Rifai Evolution of Methods for Measurement of HDL-Cholesterol: From Ultracentrifugation to Homogeneous Assays Clin. Chem., September 1, 2001; 47(9): 1579 - 1596. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Gomez, J. Camps, J. M. Simo, N. Ferre, and J. Joven Agreement Study of Methods Based on the Elimination Principle for the Measurement of LDL- and HDL-Cholesterol Compared with Ultracentrifugation in Patients with Liver Cirrhosis Clin. Chem., August 1, 2000; 46(8): 1188 - 1191. [Full Text] [PDF]
|
|
|
|
|
|
J. C. Escola-Gil, O. Jorba, J. Julve-Gil, F. Gonzalez-Sastre, J. Ordonez-Llanos, and F. Blanco-Vaca Pitfalls of Direct HDL-Cholesterol Measurements in Mouse Models of Hyperlipidemia and Atherosclerosis Clin. Chem., September 1, 1999; 45(9): 1567 - 1569. [Full Text] [PDF]
|
|
|
|
|
|
J. Camps, J. M. Simo, S. Guaita, N. Ferre, and J. Joven Altered Composition of Lipoproteins in Liver Cirrhosis Compromises Three Homogeneous Methods for HDL-Cholesterol Clin. Chem., May 1, 1999; 45(5): 685 - 688. [Full Text] [PDF]
|
|
|
|
|
|
M.-L. Arranz-Pena, J. Tasende-Mata, and F. J. Martin-Gil Comparison of two homogeneous assays with a precipitation method and an ultracentrifugation method for the measurement of HDL-cholesterol Clin. Chem., December 1, 1998; 44(12): 2499 - 2505. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. J. Lackner and G. Schmitz ß-VLDL of Patients with Type III Hyperlipoproteinemia Interferes with Homogeneous Determination of HDL-Cholesterol Based on Polyethylene Glycol-modified Enzymes Clin. Chem., December 1, 1998; 44(12): 2546 - 2548. [Full Text] [PDF]
|
|
|
|
|
|
M. Nauck, W. Marz, and H. Wieland New immunoseparation-based homogeneous assay for HDL-cholesterol compared with three homogeneous and two heterogeneous methods for HDL-cholesterol Clin. Chem., July 1, 1998; 44(7): 1443 - 1451. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Rifai, T. G. Cole, E. Iannotti, T. Law, M. Macke, R. Miller, D. Dowd, and D. A. Wiebe Assessment of interlaboratory performance in external proficiency testing programs with a direct HDL-cholesterol assay Clin. Chem., July 1, 1998; 44(7): 1452 - 1458. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. M a Simo, I. Castellano, N. Ferre, J. Joven, and J. Camps Evaluation of a homogeneous assay for high-density lipoprotein cholesterol: limitations in patients with cardiovascular, renal, and hepatic disorders Clin. Chem., June 1, 1998; 44(6): 1233 - 1241. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
, PEGME method; •, PPD method.
