- Email: [email protected]
Laboratory validation of a low density lipoprotein apolipoprotein-B assay
Clinical Biochemistry 47 (2014) 211–215
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Laboratory validation of a low density lipoprotein apolipoprotein-B assay David E. Kelsey a, Jessica L. Toher a, Michael T. Foster b, James A. Boulanger b, Mark A. Cervinski b,c,⁎ a b c
Maine Standards Company, Windham, ME, USA Department of Pathology, Dartmouth-Hitchcock Medical Center, Lebanon, NH, USA Geisel School of Medicine at Dartmouth, Hanover, NH, USA
a r t i c l e
i n f o
Article history: Received 12 June 2014 Received in revised form 16 July 2014 Accepted 18 July 2014 Available online 28 July 2014 Keywords: Low density lipoprotein particle Apolipoprotein B Assay validation Atherosclerosis
a b s t r a c t Objectives: Numerous publications have shown strong association between CHD risk and either apolipoprotein B (Apo-B) or low density lipoprotein (LDL) particle number (LDL-P). It is however unknown if Apo-B or LDLP has a stronger predictive ability for future CHD. This uncertainty may be due to the inability of current Apo-B assays to separate the contribution of very low-density lipoprotein particles from the total Apo-B concentration. As such we have performed a laboratory validation of the Maine Standards® LDL Apo-B assay on the Roche Cobas 6000 analyzer. Design and methods: Imprecision, linear range, and limit of quantitation studies were performed using quality control materials. Plasma samples collected for lipid profile analysis were analyzed via the LDL Apo-B assay and compared to the LDL cholesterol (LDL-C) concentration determined via direct LDL assay and Friedewald equation. Results: The LDL Apo-B within-run imprecision was 2.3% at 62 mg/dL and 2.2% at 109 mg/dL. The withinlaboratory imprecision was 9.7% at 57 mg/dl and 6.1% at 104 mg/dL. Linear regression analysis of LDL Apo-B versus calculated and measured LDL-c resulted in equations of LDL Apo-B = 0.620 ∗ (LDL) + 45.4, R = 0.9063 and LDL-Apo-B = 0.607 ∗ (LDL) + 38.8, R = 0.9393, respectively. Bias plot analyses revealed that at low LDL-C concentration, there was a tendency for a higher than anticipated LDL Apo-B concentration. Conclusions: The Maine Standards LDL Apo-B assay is a precise automated assay and comparison of LDL ApoB to LDL-c concentration demonstrates that low LDL-C concentrations may still carry residual risk of CHD due to increased concentration of small dense LDL particles. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction It is well appreciated that the relationship between apolipoprotein-B (Apo-B) concentration and the risk of developing cardiovascular disease (CVD) is more robust than the relationship between low density lipoprotein cholesterol (LDL-C) concentration and CVD. Despite this knowledge the current National Cholesterol Education Program Adult Treatment Panel (ATP) III guidelines continue to direct medical providers to quantify LDL-C as part of patient's risk assessment for CVD. The ATP IV guidelines expected in late 2013 have yet to be published and thus their recommendations are yet unknown. Perhaps the new guidelines will take into account the stronger association between
Abbreviations: Apo-B, apolipoprotein B; LDL, low density lipoprotein; LDL-C, low density lipoprotein cholesterol; LDL-P, low density lipoprotein particle number; VLDL, very low density lipoprotein; IDL, intermediate density lipoprotein; Lp(a), lipoprotein (a); CHD, coronary heart disease; CVD, cardiovascular disease; NCEP ATP III, National Cholesterol Education Program Adult Treatment Panel III; NMR, nuclear magnetic resonance; LoQ, limit of quantitation. ⁎ Corresponding author at: Department of Pathology, Dartmouth-Hitchcock Medical Center, One Medical Center Dr., Lebanon, NH 03756, USA. Fax: +1 603 650 4845. E-mail address: [email protected] (M.A. Cervinski).
atherogenic lipoprotein particle number and CVD risk and provide guidance on treatment goals for Apo-B or LDL particle number. A large body of work has demonstrated that the total Apo-B concentration, or the total number of atherogenic particles has a stronger association with subclinical atherosclerosis and the risk of developing CVD than LDL-C [1–6]. The VLDL, IDL, LDL and Lp(a) particles all contain Apo-B but they are not all equally atherogenic. LDL the particle at the highest concentration is also the most atherogenic. The atherogenicity of lipoprotein particles is also dependent on particle size and the ease with which they enter the sub-endothelial space of the vasculature to induce an inflammatory response. LDL particles are heterogeneous in size and density such that at any given concentration of LDL-C the actual number of LDL particles and thus by extension atherogenicity for a given LDL-C concentration is not fixed. The most cost effective method to quantify all of the atherogenic lipoproteins has been to measure the Apo-B concentration. The measurement of Apo-B has been repeatedly shown to have superior predictive ability for future CVD risk as compared to measurement of LDL-C [1,3–7]. The estimation of LDL particle concentration as well as LDL particle size can also be achieved via NMR [8]. Determination of the LDL particle concentration using NMR has been shown to be as good or better than
http://dx.doi.org/10.1016/j.clinbiochem.2014.07.016 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
212
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
LDL-C for the prediction carotid plaque size, development of CVD or CHD events [2,8–12]. A meta-analysis examining the predictive abilities of Apo-B and LDLP quantification has demonstrated that these two markers are highly correlated with LDL-P quantitation showing a stronger association with CVD events in some studies, while in others the Apo-B concentration has been a better predictor [13]. None of these studies have done a head-to-head comparison of the two techniques to determine the superior assay [13]. Some of the variability between the LDL-P and Apo-B and their ability to predict future CVD may reside with the fact that the available ApoB assays cannot differentiate between LDL Apo-B and VLDL/IDL Apo-B. Previous studies have shown that ~90% of the serum Apo-B immunoreactivity in the serum of normo- and hypertriglyceridemic patients comes from the circulating LDL particles [7]. Thus an assay that specifically blocks the Apo-B present on VLDL particles may improve the predictive ability of the Apo-B measurement. To that end, Maine Standards LLC has developed an investigational use only (IUO) automated immunoturbidimetric assay for the Roche Cobas 6000 analyzer that specifically quantifies the Apo-B present on the surface of LDL particles. Specificity for LDL Apo-B is achieved in an automated preincubation step in which anti-Apo-B antibodies are prevented from binding to VLDL Apo-B via chemical inhibition. In this report we highlight the laboratory validation of this IUO assay on the Roche Cobas 6000 analyzer. Included in this report is a summary of assay precision, linear range and comparison of the LDL Apo-B concentration to LDL-C concentration as determined by both Friedewald calculation and direct assay.
Materials and methods Clinical samples A total of 300 lithium heparinized patient plasma samples obtained for routine physician ordered assessment of fasting lipid profile (total cholesterol, HDL, triglycerides and calculated LDL cholesterol) were included for analysis via both the LDL Apo-B assay as well as a Direct LDL assay on the Cobas 6000 analyzer. Each sample came from separate individual patients above the age of 18. All samples were de-identified, aliquoted and frozen prior to testing. The testing protocol was reviewed and approved by the Committee for the Protection of Human Subjects at Dartmouth-Hitchcock Medical Center.
Maine Standards LLC LDL Apo-B assay The Maine Standards LLC LDL Apo-B assay for use on the Roche Cobas 6000 analyzer is an immunoturbidimetric assay employing a polyclonal antiserum against Apo-B. In order to accurately quantify only the Apo-B content on LDL particles the patient samples are first automatically diluted via the instrument. The diluted samples are then mixed with a proprietary reagent solution that in addition to containing the anti-Apo-B antibodies specifically inhibits the reaction of Apo-B present on VLDL particles with the anti-Apo-B antibodies. The incubation of the sample with the inhibitory reagent and anti-Apo-B antibodies produces light absorbing complexes that are measured at 340 nm. The absorbance due to the turbidity is compared to a calibration curve set by using a set of known calibrators. The calibrators were manufactured and assigned values using a protocol similar to that used to prepare the ApoB reference material SP3-07 [14]. Briefly, LDL (density = 1.030 − 1.050 kg/L) was purified by ultracentrifugation from fresh plasma pools, and assigned a mass value using a sodium dodecyl sulfate — Lowry procedure [15]. This material was used as a calibrator in the assay to assign values to the master calibrator lot.
Imprecision, limit of quantitation and linear range In order to determine within-run and within-laboratory imprecision of the Maine Standards LDL Apo-B method, two levels of quality control material manufactured by Maine Standards were analyzed ten times in a single run and once a day for twenty non-consecutive days on the Cobas 6000 analyzer. The limit of quantitation for this assay was defined as the concentration at which the CV% exceeded 20. The limit of quantitation (LoQ) for the assay was determined by repeated measurement of five aliquots of materials containing decreasing concentrations of LDL Apo-B produced from assay quality control materials in duplicate, once daily, for ten non-consecutive days. Two of the five levels were targeted to be at concentrations below 30 mg/dL. The LoQ was established as the concentration of LDL Apo-B at which the assay had a CV equal to 20% and was calculated using the appropriate module of EP Evaluator Release 9 (Data Innovations, LLC; Burlington, VT). The linear range for this assay labeled for investigational use only was determined via mixing of patient samples containing either a very low concentration of LDL Apo-B or diluent with samples containing a high concentration of LDL Apo-B and measuring the resulting concentration with the LDL Apo-B assay.
Method concordance LDL concentrations determined either via direct LDL assay or by Friedewald equation (LDL Cholesterol = total cholesterol − (HDL cholesterol + VLDL-C/5)) on the Roche Cobas 6000 were plotted against the LDL Apo-B concentration and analyzed via Deming regression. Additionally percent bias plots comparing the calculated or direct LDL cholesterol as determined via the Roche LDL-C plus assay on the Cobas 600 to the LDL Apo-B cholesterol were also generated. The percent bias was calculated as ((LDL Apo-B) − LDL-C) / LDL-C. All data analysis and plots were generated using EP Evaluator.
Results The LDL Apo-B assay imprecision was assessed using two levels of QC material as outlined in the Materials and methods section. The withinrun imprecision of the assay was 2.3% and 2.2% at LDL Apo-B concentrations of 62 mg/dL (1127 nmol/L) and 109 mg/dL (1982 nmol/L), respectively. The total assay imprecision determined via analysis of the same QC materials over 20 nonconsecutive days was 9.7% and 6.1% at LDL Apo-B concentrations of 57 mg/dL (1036 nmol/L) and 104 mg/dL (1980 nmol/L), respectively. The LoQ for this assay (CV = 20%) was determined to be 33 mg/dL (600 nmol/L) and was linear to a concentration of 284 mg/dL (5162 nmol/L), the highest concentration tested. In addition to testing the performance characteristics of the assay we also compared the LDL Apo-B concentration to both the calculated and measured LDL on 300 remnant patient plasma samples from physician ordered lipid profiles. As anticipated there was no strong linear relationship between the calculated (Fig. 1A) or measured (Fig. 1B) LDL cholesterol concentration and the measured LDL Apo-B concentration. The linear regression equation for the calculated LDL cholesterol LDL Apo-B pair was y = 0.62x + 45.4, R = 0.9063 while the regression equation for the measured LDL cholesterol, LDL Apo-B pair was similar at y = 0.60x + 38.8, R = 0.9393. For both plots there is a distinct positive bias with elevated LDL Apo-B at low LDL cholesterol concentrations. Within this positive bias there are a few patient samples that display a much greater bias with significantly higher LDL Apo-B concentrations. Analysis of the same data in percent bias plots reveals a significant spread of LDL Apo-B concentrations surrounding an LDL cholesterol concentration of approximately 50 mg/dL (Fig. 2).
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
213
Fig. 1. Linear regression analysis of: A) LDL Apo-B concentration vs. calculated LDL-C concentration or B) LDL Apo-B concentration vs. measured LDL-C concentration.
Discussion In this report we have detailed the laboratory validation of a LDL Apo-B assay developed for the Roche Cobas 6000 analyzer. As expected this assay demonstrates laboratory acceptable precision, low LoQ and a wide measurement range. This assay that specifically measures Apo-B
on LDL particles demonstrates a positive correlation with LDL-C that is not unlike the relationship between LDL-C and total Apo-B. While there is a positive correlation between the two measures, a visual interpretation of the dispersion of data indicates that for any individual patient that one cannot infer a LDL Apo-B concentration from the LDL-C or vice versa. As outlined in the Results section there is also
214
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
Fig. 2. Bias plot demonstrating the percent bias between the: A) LDL Apo-B and calculated LDL-C concentration or B) LDL Apo-B and measured LDL-C concentration.
a significant positive bias in the Apo-B concentration at low LDL-C concentrations and a negative bias in Apo-B concentration at high LDL-C concentrations. This relationship between the Apo-B and LDL-C concentrations demonstrates that for any given LDL-C concentration the Apo-B concentration for any individual patient can vary significantly. Although as yet the American College of Cardiology and the American Heart Association have not endorsed the use of LDL-particle number or Apo-B for CVD risk assessment there is a substantial body
or literature that demonstrates that the Apo-B concentration and LDLparticle number are better predictors of future CVD risk [1,3–12,16]. The current body of literature on the subject of which measure, Apo-B or LDL-P, is the superior marker is split with some publications showing Apo-B to be the better marker, and others showing LDL-P to be better. Part of the discrepancy in these studies may be related to the specificity of the current total Apo-B assays that measure not only LDL but also VLDL and IDL. A comparison of the performance of this LDL specific
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
Apo-B assay to LDL-P measurements may serve to clarify the issue. Further, as the LDL Apo-B assay does not measure the Apo-B bound to VLDL and IDL particles we can indirectly infer the LDL-particle number from the molar concentration of LDL Apo-B as measured by this assay, however this assertion was not explored in this study. Our present study does have certain limitations. Notably due to the nature of this validation study we could not directly compare the performance of the LDL Apo-B assay vs. LDL-C for assessment of future CVD risk. Similarly this validation study did not test the association between the LDL Apo-B measurement and LDL-particle number as determined by NMR. These additional studies are needed to further delineate the future role of this assay in clinical practice. Should future studies verify the relationship between the LDL Apo-B concentration and the LDL-particle assay, this precise and easy to use automated turbidimetric assay for use on the Roche Cobas 6000 analyzer may serve as either a low cost alternative to the high-complexity LDLparticle assay by NMR, or as a screen to identify which patients should have a follow up LDL-particle assessment performed in a reference laboratory. Either possible eventuality, replacement of the NMR assay or screening of patients via the automated turbidimetric assay would prove valuable in restraining health care costs by reducing the cost of testing as the cost of this assay would likely be less financially burdensome to health care facilities. References [1] Barter PJ, Ballantyne CM, Carmena R, Castro Cabezas M, Chapman MJ, Couture P, et al. Apo-B versus cholesterol in estimating cardiovascular risk and in guiding therapy: report of the thirty-person/ten-country panel. J Intern Med 2006;259(3):247–58. [2] Virani SS, Pompeii L, Lincoln AE, Dunn RE, Tucker AM, Nambi V, et al. Association between traditional cholesterol parameters, lipoprotein particle concentration, novel biomarkers and carotid plaques in retired National Football League players. Atherosclerosis 2012;222(2):551–6. [3] Kastelein JJ, van der Steeg WA, Holme I, Gaffney M, Cater NB, Barter P, et al. Lipids, apolipoproteins, and their ratios in relation to cardiovascular events with statin treatment. Circulation 2008;117(23):3002–9.
215
[4] Gotto Jr AM, Whitney E, Stein EA, Shapiro DR, Clearfield M, Weis S, et al. Relation between baseline and on-treatment lipid parameters and first acute major coronary events in the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS). Circulation 2000;101(5):477–84. [5] Gigante B, Leander K, Vikstrom M, Frumento P, Carlsson AC, Bottai M, et al. Elevated ApoB serum levels strongly predict early cardiovascular events. Heart 2012; 98(16):1242–5. [6] Lamarche B, Tchernof A, Mauriege P, Cantin B, Dagenais GR, Lupien PJ, et al. Fasting insulin and apolipoprotein B levels and low-density lipoprotein particle size as risk factors for ischemic heart disease. JAMA 1998;279(24):1955–61. [7] Sniderman A, Vu H, Cianflone K. Effect of moderate hypertriglyceridemia on the relation of plasma total and LDL apo B levels. Atherosclerosis 1991;89(2–3):109–16. [8] Otvos JD, Collins D, Freedman DS, Shalaurova I, Schaefer EJ, McNamara JR, et al. Lowdensity lipoprotein and high-density lipoprotein particle subclasses predict coronary events and are favorably changed by gemfibrozil therapy in the Veterans Affairs High-Density Lipoprotein Intervention Trial. Circulation 2006;113(12):1556–63. [9] Mora S, Otvos JD, Rifai N, Rosenson RS, Buring JE, Ridker PM. Lipoprotein particle profiles by nuclear magnetic resonance compared with standard lipids and apolipoproteins in predicting incident cardiovascular disease in women. Circulation 2009; 119(7):931–9. [10] Parish S, Offer A, Clarke R, Hopewell JC, Hill MR, Otvos JD, et al. Lipids and lipoproteins and risk of different vascular events in the MRC/BHF Heart Protection Study. Circulation 2012;125(20):2469–78. [11] Blake GJ, Otvos JD, Rifai N, Ridker PM. Low-density lipoprotein particle concentration and size as determined by nuclear magnetic resonance spectroscopy as predictors of cardiovascular disease in women. Circulation 2002;106(15):1930–7. [12] Cromwell WC, Otvos JD, Keyes MJ, Pencina MJ, Sullivan L, Vasan RS, et al. LDL particle number and risk of future cardiovascular disease in the Framingham Offspring Study —implications for LDL management. J Clin Lipidol 2007; 1(6):583–92. [13] Cole TG, Contois JH, Csako G, McConnell JP, Remaley AT, Devaraj S, et al. Association of apolipoprotein B and nuclear magnetic resonance spectroscopy-derived LDL particle number with outcomes in 25 clinical studies: assessment by the AACC Lipoprotein and Vascular Diseases Division Working Group on Best Practices. Clin Chem 2013;59(5):752–70. [14] Marcovina SM, Gaur VP, Albers JJ. Biological variability of cholesterol, triglyceride, low- and high-density lipoprotein cholesterol, lipoprotein (a), and apolipoproteins A-1 and B. Clin Chem 1994;40:586–92. [15] Markwell MAK, Haas SM, Bieber LL, Tolber NE. A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal Biochem 1978;87:206–10. [16] Stone NJ, Robinson JG, Lichtenstein AH, Merz NB, Blum CB, Eckel RH, et al. 2013 ACC/ AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults. Circulation 2014;129:S1–S45.
Contents lists available at ScienceDirect
Clinical Biochemistry journal homepage: www.elsevier.com/locate/clinbiochem
Laboratory validation of a low density lipoprotein apolipoprotein-B assay David E. Kelsey a, Jessica L. Toher a, Michael T. Foster b, James A. Boulanger b, Mark A. Cervinski b,c,⁎ a b c
Maine Standards Company, Windham, ME, USA Department of Pathology, Dartmouth-Hitchcock Medical Center, Lebanon, NH, USA Geisel School of Medicine at Dartmouth, Hanover, NH, USA
a r t i c l e
i n f o
Article history: Received 12 June 2014 Received in revised form 16 July 2014 Accepted 18 July 2014 Available online 28 July 2014 Keywords: Low density lipoprotein particle Apolipoprotein B Assay validation Atherosclerosis
a b s t r a c t Objectives: Numerous publications have shown strong association between CHD risk and either apolipoprotein B (Apo-B) or low density lipoprotein (LDL) particle number (LDL-P). It is however unknown if Apo-B or LDLP has a stronger predictive ability for future CHD. This uncertainty may be due to the inability of current Apo-B assays to separate the contribution of very low-density lipoprotein particles from the total Apo-B concentration. As such we have performed a laboratory validation of the Maine Standards® LDL Apo-B assay on the Roche Cobas 6000 analyzer. Design and methods: Imprecision, linear range, and limit of quantitation studies were performed using quality control materials. Plasma samples collected for lipid profile analysis were analyzed via the LDL Apo-B assay and compared to the LDL cholesterol (LDL-C) concentration determined via direct LDL assay and Friedewald equation. Results: The LDL Apo-B within-run imprecision was 2.3% at 62 mg/dL and 2.2% at 109 mg/dL. The withinlaboratory imprecision was 9.7% at 57 mg/dl and 6.1% at 104 mg/dL. Linear regression analysis of LDL Apo-B versus calculated and measured LDL-c resulted in equations of LDL Apo-B = 0.620 ∗ (LDL) + 45.4, R = 0.9063 and LDL-Apo-B = 0.607 ∗ (LDL) + 38.8, R = 0.9393, respectively. Bias plot analyses revealed that at low LDL-C concentration, there was a tendency for a higher than anticipated LDL Apo-B concentration. Conclusions: The Maine Standards LDL Apo-B assay is a precise automated assay and comparison of LDL ApoB to LDL-c concentration demonstrates that low LDL-C concentrations may still carry residual risk of CHD due to increased concentration of small dense LDL particles. © 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
Introduction It is well appreciated that the relationship between apolipoprotein-B (Apo-B) concentration and the risk of developing cardiovascular disease (CVD) is more robust than the relationship between low density lipoprotein cholesterol (LDL-C) concentration and CVD. Despite this knowledge the current National Cholesterol Education Program Adult Treatment Panel (ATP) III guidelines continue to direct medical providers to quantify LDL-C as part of patient's risk assessment for CVD. The ATP IV guidelines expected in late 2013 have yet to be published and thus their recommendations are yet unknown. Perhaps the new guidelines will take into account the stronger association between
Abbreviations: Apo-B, apolipoprotein B; LDL, low density lipoprotein; LDL-C, low density lipoprotein cholesterol; LDL-P, low density lipoprotein particle number; VLDL, very low density lipoprotein; IDL, intermediate density lipoprotein; Lp(a), lipoprotein (a); CHD, coronary heart disease; CVD, cardiovascular disease; NCEP ATP III, National Cholesterol Education Program Adult Treatment Panel III; NMR, nuclear magnetic resonance; LoQ, limit of quantitation. ⁎ Corresponding author at: Department of Pathology, Dartmouth-Hitchcock Medical Center, One Medical Center Dr., Lebanon, NH 03756, USA. Fax: +1 603 650 4845. E-mail address: [email protected] (M.A. Cervinski).
atherogenic lipoprotein particle number and CVD risk and provide guidance on treatment goals for Apo-B or LDL particle number. A large body of work has demonstrated that the total Apo-B concentration, or the total number of atherogenic particles has a stronger association with subclinical atherosclerosis and the risk of developing CVD than LDL-C [1–6]. The VLDL, IDL, LDL and Lp(a) particles all contain Apo-B but they are not all equally atherogenic. LDL the particle at the highest concentration is also the most atherogenic. The atherogenicity of lipoprotein particles is also dependent on particle size and the ease with which they enter the sub-endothelial space of the vasculature to induce an inflammatory response. LDL particles are heterogeneous in size and density such that at any given concentration of LDL-C the actual number of LDL particles and thus by extension atherogenicity for a given LDL-C concentration is not fixed. The most cost effective method to quantify all of the atherogenic lipoproteins has been to measure the Apo-B concentration. The measurement of Apo-B has been repeatedly shown to have superior predictive ability for future CVD risk as compared to measurement of LDL-C [1,3–7]. The estimation of LDL particle concentration as well as LDL particle size can also be achieved via NMR [8]. Determination of the LDL particle concentration using NMR has been shown to be as good or better than
http://dx.doi.org/10.1016/j.clinbiochem.2014.07.016 0009-9120/© 2014 The Canadian Society of Clinical Chemists. Published by Elsevier Inc. All rights reserved.
212
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
LDL-C for the prediction carotid plaque size, development of CVD or CHD events [2,8–12]. A meta-analysis examining the predictive abilities of Apo-B and LDLP quantification has demonstrated that these two markers are highly correlated with LDL-P quantitation showing a stronger association with CVD events in some studies, while in others the Apo-B concentration has been a better predictor [13]. None of these studies have done a head-to-head comparison of the two techniques to determine the superior assay [13]. Some of the variability between the LDL-P and Apo-B and their ability to predict future CVD may reside with the fact that the available ApoB assays cannot differentiate between LDL Apo-B and VLDL/IDL Apo-B. Previous studies have shown that ~90% of the serum Apo-B immunoreactivity in the serum of normo- and hypertriglyceridemic patients comes from the circulating LDL particles [7]. Thus an assay that specifically blocks the Apo-B present on VLDL particles may improve the predictive ability of the Apo-B measurement. To that end, Maine Standards LLC has developed an investigational use only (IUO) automated immunoturbidimetric assay for the Roche Cobas 6000 analyzer that specifically quantifies the Apo-B present on the surface of LDL particles. Specificity for LDL Apo-B is achieved in an automated preincubation step in which anti-Apo-B antibodies are prevented from binding to VLDL Apo-B via chemical inhibition. In this report we highlight the laboratory validation of this IUO assay on the Roche Cobas 6000 analyzer. Included in this report is a summary of assay precision, linear range and comparison of the LDL Apo-B concentration to LDL-C concentration as determined by both Friedewald calculation and direct assay.
Materials and methods Clinical samples A total of 300 lithium heparinized patient plasma samples obtained for routine physician ordered assessment of fasting lipid profile (total cholesterol, HDL, triglycerides and calculated LDL cholesterol) were included for analysis via both the LDL Apo-B assay as well as a Direct LDL assay on the Cobas 6000 analyzer. Each sample came from separate individual patients above the age of 18. All samples were de-identified, aliquoted and frozen prior to testing. The testing protocol was reviewed and approved by the Committee for the Protection of Human Subjects at Dartmouth-Hitchcock Medical Center.
Maine Standards LLC LDL Apo-B assay The Maine Standards LLC LDL Apo-B assay for use on the Roche Cobas 6000 analyzer is an immunoturbidimetric assay employing a polyclonal antiserum against Apo-B. In order to accurately quantify only the Apo-B content on LDL particles the patient samples are first automatically diluted via the instrument. The diluted samples are then mixed with a proprietary reagent solution that in addition to containing the anti-Apo-B antibodies specifically inhibits the reaction of Apo-B present on VLDL particles with the anti-Apo-B antibodies. The incubation of the sample with the inhibitory reagent and anti-Apo-B antibodies produces light absorbing complexes that are measured at 340 nm. The absorbance due to the turbidity is compared to a calibration curve set by using a set of known calibrators. The calibrators were manufactured and assigned values using a protocol similar to that used to prepare the ApoB reference material SP3-07 [14]. Briefly, LDL (density = 1.030 − 1.050 kg/L) was purified by ultracentrifugation from fresh plasma pools, and assigned a mass value using a sodium dodecyl sulfate — Lowry procedure [15]. This material was used as a calibrator in the assay to assign values to the master calibrator lot.
Imprecision, limit of quantitation and linear range In order to determine within-run and within-laboratory imprecision of the Maine Standards LDL Apo-B method, two levels of quality control material manufactured by Maine Standards were analyzed ten times in a single run and once a day for twenty non-consecutive days on the Cobas 6000 analyzer. The limit of quantitation for this assay was defined as the concentration at which the CV% exceeded 20. The limit of quantitation (LoQ) for the assay was determined by repeated measurement of five aliquots of materials containing decreasing concentrations of LDL Apo-B produced from assay quality control materials in duplicate, once daily, for ten non-consecutive days. Two of the five levels were targeted to be at concentrations below 30 mg/dL. The LoQ was established as the concentration of LDL Apo-B at which the assay had a CV equal to 20% and was calculated using the appropriate module of EP Evaluator Release 9 (Data Innovations, LLC; Burlington, VT). The linear range for this assay labeled for investigational use only was determined via mixing of patient samples containing either a very low concentration of LDL Apo-B or diluent with samples containing a high concentration of LDL Apo-B and measuring the resulting concentration with the LDL Apo-B assay.
Method concordance LDL concentrations determined either via direct LDL assay or by Friedewald equation (LDL Cholesterol = total cholesterol − (HDL cholesterol + VLDL-C/5)) on the Roche Cobas 6000 were plotted against the LDL Apo-B concentration and analyzed via Deming regression. Additionally percent bias plots comparing the calculated or direct LDL cholesterol as determined via the Roche LDL-C plus assay on the Cobas 600 to the LDL Apo-B cholesterol were also generated. The percent bias was calculated as ((LDL Apo-B) − LDL-C) / LDL-C. All data analysis and plots were generated using EP Evaluator.
Results The LDL Apo-B assay imprecision was assessed using two levels of QC material as outlined in the Materials and methods section. The withinrun imprecision of the assay was 2.3% and 2.2% at LDL Apo-B concentrations of 62 mg/dL (1127 nmol/L) and 109 mg/dL (1982 nmol/L), respectively. The total assay imprecision determined via analysis of the same QC materials over 20 nonconsecutive days was 9.7% and 6.1% at LDL Apo-B concentrations of 57 mg/dL (1036 nmol/L) and 104 mg/dL (1980 nmol/L), respectively. The LoQ for this assay (CV = 20%) was determined to be 33 mg/dL (600 nmol/L) and was linear to a concentration of 284 mg/dL (5162 nmol/L), the highest concentration tested. In addition to testing the performance characteristics of the assay we also compared the LDL Apo-B concentration to both the calculated and measured LDL on 300 remnant patient plasma samples from physician ordered lipid profiles. As anticipated there was no strong linear relationship between the calculated (Fig. 1A) or measured (Fig. 1B) LDL cholesterol concentration and the measured LDL Apo-B concentration. The linear regression equation for the calculated LDL cholesterol LDL Apo-B pair was y = 0.62x + 45.4, R = 0.9063 while the regression equation for the measured LDL cholesterol, LDL Apo-B pair was similar at y = 0.60x + 38.8, R = 0.9393. For both plots there is a distinct positive bias with elevated LDL Apo-B at low LDL cholesterol concentrations. Within this positive bias there are a few patient samples that display a much greater bias with significantly higher LDL Apo-B concentrations. Analysis of the same data in percent bias plots reveals a significant spread of LDL Apo-B concentrations surrounding an LDL cholesterol concentration of approximately 50 mg/dL (Fig. 2).
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
213
Fig. 1. Linear regression analysis of: A) LDL Apo-B concentration vs. calculated LDL-C concentration or B) LDL Apo-B concentration vs. measured LDL-C concentration.
Discussion In this report we have detailed the laboratory validation of a LDL Apo-B assay developed for the Roche Cobas 6000 analyzer. As expected this assay demonstrates laboratory acceptable precision, low LoQ and a wide measurement range. This assay that specifically measures Apo-B
on LDL particles demonstrates a positive correlation with LDL-C that is not unlike the relationship between LDL-C and total Apo-B. While there is a positive correlation between the two measures, a visual interpretation of the dispersion of data indicates that for any individual patient that one cannot infer a LDL Apo-B concentration from the LDL-C or vice versa. As outlined in the Results section there is also
214
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
Fig. 2. Bias plot demonstrating the percent bias between the: A) LDL Apo-B and calculated LDL-C concentration or B) LDL Apo-B and measured LDL-C concentration.
a significant positive bias in the Apo-B concentration at low LDL-C concentrations and a negative bias in Apo-B concentration at high LDL-C concentrations. This relationship between the Apo-B and LDL-C concentrations demonstrates that for any given LDL-C concentration the Apo-B concentration for any individual patient can vary significantly. Although as yet the American College of Cardiology and the American Heart Association have not endorsed the use of LDL-particle number or Apo-B for CVD risk assessment there is a substantial body
or literature that demonstrates that the Apo-B concentration and LDLparticle number are better predictors of future CVD risk [1,3–12,16]. The current body of literature on the subject of which measure, Apo-B or LDL-P, is the superior marker is split with some publications showing Apo-B to be the better marker, and others showing LDL-P to be better. Part of the discrepancy in these studies may be related to the specificity of the current total Apo-B assays that measure not only LDL but also VLDL and IDL. A comparison of the performance of this LDL specific
D.E. Kelsey et al. / Clinical Biochemistry 47 (2014) 211–215
Apo-B assay to LDL-P measurements may serve to clarify the issue. Further, as the LDL Apo-B assay does not measure the Apo-B bound to VLDL and IDL particles we can indirectly infer the LDL-particle number from the molar concentration of LDL Apo-B as measured by this assay, however this assertion was not explored in this study. Our present study does have certain limitations. Notably due to the nature of this validation study we could not directly compare the performance of the LDL Apo-B assay vs. LDL-C for assessment of future CVD risk. Similarly this validation study did not test the association between the LDL Apo-B measurement and LDL-particle number as determined by NMR. These additional studies are needed to further delineate the future role of this assay in clinical practice. Should future studies verify the relationship between the LDL Apo-B concentration and the LDL-particle assay, this precise and easy to use automated turbidimetric assay for use on the Roche Cobas 6000 analyzer may serve as either a low cost alternative to the high-complexity LDLparticle assay by NMR, or as a screen to identify which patients should have a follow up LDL-particle assessment performed in a reference laboratory. Either possible eventuality, replacement of the NMR assay or screening of patients via the automated turbidimetric assay would prove valuable in restraining health care costs by reducing the cost of testing as the cost of this assay would likely be less financially burdensome to health care facilities. References [1] Barter PJ, Ballantyne CM, Carmena R, Castro Cabezas M, Chapman MJ, Couture P, et al. Apo-B versus cholesterol in estimating cardiovascular risk and in guiding therapy: report of the thirty-person/ten-country panel. J Intern Med 2006;259(3):247–58. [2] Virani SS, Pompeii L, Lincoln AE, Dunn RE, Tucker AM, Nambi V, et al. Association between traditional cholesterol parameters, lipoprotein particle concentration, novel biomarkers and carotid plaques in retired National Football League players. Atherosclerosis 2012;222(2):551–6. [3] Kastelein JJ, van der Steeg WA, Holme I, Gaffney M, Cater NB, Barter P, et al. Lipids, apolipoproteins, and their ratios in relation to cardiovascular events with statin treatment. Circulation 2008;117(23):3002–9.
215
[4] Gotto Jr AM, Whitney E, Stein EA, Shapiro DR, Clearfield M, Weis S, et al. Relation between baseline and on-treatment lipid parameters and first acute major coronary events in the Air Force/Texas Coronary Atherosclerosis Prevention Study (AFCAPS/TexCAPS). Circulation 2000;101(5):477–84. [5] Gigante B, Leander K, Vikstrom M, Frumento P, Carlsson AC, Bottai M, et al. Elevated ApoB serum levels strongly predict early cardiovascular events. Heart 2012; 98(16):1242–5. [6] Lamarche B, Tchernof A, Mauriege P, Cantin B, Dagenais GR, Lupien PJ, et al. Fasting insulin and apolipoprotein B levels and low-density lipoprotein particle size as risk factors for ischemic heart disease. JAMA 1998;279(24):1955–61. [7] Sniderman A, Vu H, Cianflone K. Effect of moderate hypertriglyceridemia on the relation of plasma total and LDL apo B levels. Atherosclerosis 1991;89(2–3):109–16. [8] Otvos JD, Collins D, Freedman DS, Shalaurova I, Schaefer EJ, McNamara JR, et al. Lowdensity lipoprotein and high-density lipoprotein particle subclasses predict coronary events and are favorably changed by gemfibrozil therapy in the Veterans Affairs High-Density Lipoprotein Intervention Trial. Circulation 2006;113(12):1556–63. [9] Mora S, Otvos JD, Rifai N, Rosenson RS, Buring JE, Ridker PM. Lipoprotein particle profiles by nuclear magnetic resonance compared with standard lipids and apolipoproteins in predicting incident cardiovascular disease in women. Circulation 2009; 119(7):931–9. [10] Parish S, Offer A, Clarke R, Hopewell JC, Hill MR, Otvos JD, et al. Lipids and lipoproteins and risk of different vascular events in the MRC/BHF Heart Protection Study. Circulation 2012;125(20):2469–78. [11] Blake GJ, Otvos JD, Rifai N, Ridker PM. Low-density lipoprotein particle concentration and size as determined by nuclear magnetic resonance spectroscopy as predictors of cardiovascular disease in women. Circulation 2002;106(15):1930–7. [12] Cromwell WC, Otvos JD, Keyes MJ, Pencina MJ, Sullivan L, Vasan RS, et al. LDL particle number and risk of future cardiovascular disease in the Framingham Offspring Study —implications for LDL management. J Clin Lipidol 2007; 1(6):583–92. [13] Cole TG, Contois JH, Csako G, McConnell JP, Remaley AT, Devaraj S, et al. Association of apolipoprotein B and nuclear magnetic resonance spectroscopy-derived LDL particle number with outcomes in 25 clinical studies: assessment by the AACC Lipoprotein and Vascular Diseases Division Working Group on Best Practices. Clin Chem 2013;59(5):752–70. [14] Marcovina SM, Gaur VP, Albers JJ. Biological variability of cholesterol, triglyceride, low- and high-density lipoprotein cholesterol, lipoprotein (a), and apolipoproteins A-1 and B. Clin Chem 1994;40:586–92. [15] Markwell MAK, Haas SM, Bieber LL, Tolber NE. A modification of the Lowry procedure to simplify protein determination in membrane and lipoprotein samples. Anal Biochem 1978;87:206–10. [16] Stone NJ, Robinson JG, Lichtenstein AH, Merz NB, Blum CB, Eckel RH, et al. 2013 ACC/ AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults. Circulation 2014;129:S1–S45.