Printer Friendly

Proposed serum cholesterol reference measurement procedure by gas chromatography-isotope dilution mass spectrometry.

For more than 50 years, the Centers for Disease Control and Prevention has standardized cholesterol measurements by use of a spectrophotometric reference measurement procedure (RMP) [3] (1). This procedure is a modified version of the Abell-Levy-Brodie-Kendall (AK) colorimetric method and is accepted as the accuracy point to which clinical cholesterol measurements are traceable. In addition, the National Cholesterol Education Program established medical decision points that were determined from epidemiologic studies and clinical trials that had been standardized by the AK RMP (2). The AK RMP has historically demonstrated imprecision consistent with a CV [less than or equal to] 1.0% for total cholesterol. AK cholesterol measurements also have a well-defined bias to the NIST gas chromatography-isotope dilution mass spectrometry (GC-IDMS) primary RMP. This has been shown in published comparisons as well on NIST SRM certificates for SRM 1951a and 1951b (3-5). The source of the bias in AK RMP is nonspecificity for cholesterol, because interfering compounds such as noncholesterol sterols, cholesterol precursors, and oxidation products produce chromophores (6) that are also measured at the same wavelength as cholesterol.

In recent years, there has been a push toward establishing traceability to higher-order RMPs and reference materials (7) for analytes of clinical significance. Higher-order RMPs have a well-defined analyte and lack any influence from interfering substances. IDMS is generally accepted as the highest-order analytical procedure (8) because of its high analyte specificity and selectivity. For these reasons, we evaluated an IDMS method as potential replacement for the AK RMP in the CDC Reference Laboratory.

Several chromatography and IDMS assays have been proposed and used as primary (9-11) and secondary (12-16) RMPs for cholesterol. Many of these procedures employ GC-IDMS or liquid chromatography-IDMS with stable isotope ([sup.2]H, [sup.13][C.sub.2], and [sup.13][C.sub.3]) enriched cholesterol analogs as an internal standard (IS). Whereas deuterated analogs are readily available from many suppliers, C-13-labeled analogs are more suitable for use as an IS because they coelute with unlabeled cholesterol. In contrast, deuterated analogs separate from cholesterol on the capillary column, which may affect precision (9, 11).

Many of the reported IDMS procedures use a bracketing technique to quantify cholesterol in serum (9, 11, 12). In this technique, the sample is spiked with labeled IS so that the unlabeled-to-labeled isotope ratio is approximately 1.0. The bracketing pair of standards is prepared with an isotopic ratio of unlabeled-to-labeled cholesterol [+ or -] 15% of that of the ratio in the sample. Consequently, a preliminary analytical run is required to estimate the cholesterol level in a sample before the matching pair of bracketing standards can be prepared. Bracketing compromises efficiency and increases workload and overall cost when analyzing many serum samples. Additionally, the bracketing technique is not feasible when analyzing samples of varying cholesterol concentrations; using a single multilevel linear calibration scheme is a better option. Multilevel linear calibration eliminates the need for preliminary experiments and permits accurate and efficient quantification of cholesterol in samples of varying concentrations.

We have developed a GC-IDMS measurement procedure that incorporates multilevel linear calibration for use as a secondary RMP to quantify cholesterol in serum. We evaluated the method's linearity, imprecision, accuracy, and specificity for cholesterol in the presence of a few interfering compounds. Furthermore, we examined the statistical relationship between the AK and GC-IDMS methods and propose a reliable linear equation that defines the correlation between these 2 RMPs for cholesterol. We modeled our approach on that used for the standardization of hemoglobin [A.sub.1c] by the IFCC RMP and the designated comparison methods (17, 18). This linear equation permits the conversion of analytical results between GC-IDMS and AK measurement as needed.

Materials and Methods


We purchased the following standard reference materials (SRMs): pure cholesterol (SRM 911b) from NIST (Gaithersburg, MD), lipids in frozen (liquid) human serum (SRM 1951b), and lyophilized human serum (SRM 909b). Serum pools were prepared at Solomon Park Research Laboratories following Clinical Laboratory Standards Institute (CLSI) C37A guidelines (19). Solomon Park Research Laboratories also provided a set of 10 native serum samples from unidentified individuals. We obtained additional samples through participation in the IFCC External Quality Assessment Scheme for Reference Laboratories in Laboratory Medicine (RELA) (20). These samples are lyophilized serum. We used [3,4-[sup.13][C.sub.2]] cholesterol supplied by CDN Isotopes as the IS. Aaper Alcohol & Chemical Company supplied absolute ethanol (200 proof). Fischer Scientific supplied potassium hydroxide and hexanes (HPLC grade). We obtained lanosta-8,24-dien-3-ol; cholest-5en-3[beta]-ol (98%); [beta]-sitosterol; 5[alpha]-cholesten-7-ene-3[beta]-ol (95%); 5[alpha]-cholestan-3[beta]-ol; ergosta-5,7,22-trien-3[beta]-ol (75%), and N,0-bis(trimethylsilyl)acetamide/pyridine (trisil BSA) from Sigma-Aldrich.


We used a Digiflex TP (Titertek) automatic pipettor equipped with dual microprocessor-controlled, direct drive stepper motors for aspirating and dispensing and a Turbovap LV evaporator (Caliper Life Sciences) for drying samples. Sartorius supplied an analytical balance. The analytical balance is gravimetrically calibrated annually by Rite-Weight using NIST certified weights. Calibrate Inc. calibrates the Digiflex TP annually by gravimetric measurements of specific volumes ofdeionized water. We used an Agilent 6890 GC (Agilent Technologies) with an HP series 6890 autosampler and an Agilent autosampler controller. A J&W brand DB5-ms (5% phenyl)-methylpolysiloxane capillary column (Agilent Technologies) was installed between the GC and an Agilent 5973 mass selective detector (MSD), and we used ChemStation software for instrument control, data acquisition, and processing.


We prepared calibration standard solutions from SRM 911b by making a primary stock solution of pure cholesterol standard after accurately weighing 2.59 mmol (1.0 g) of 911b on an analytical balance and transferring it to a dry, 200-mL volumetric flask with 150 mL warm (60 [degrees]C) absolute ethanol. We sonicated the contents in an ultrasonic water bath for 10 min then placed the flask into another water bath maintained at 25 [degrees]C to equilibrate before diluting with absolute ethanol, also maintained at 25 [degrees]C. We prepared 7 working standard solutions for calibration (0.065, 0.13, 0.26, 0.39, 0.52, 0.78, and 1.03 mmol/L cholesterol) and stored them at 4 [degrees]C. These working standard solutions were prepared from the stock standard solution using a Digiflex TP automatic pipettor to aspirate exact volumes of stock standard solutions (0.5, 1.0, 2.0, 3.0, 4.0, 6.0, and 8.0 mL) and dispensing it into clean, 100-mL volumetric flasks along with 20 mL absolute ethanol. We brought all standard solutions to volume with ethanol maintained at 25 [degrees]C. We maintain at least 3 sets of standards and use them on a rotating basis for all of our analytical runs.


We used [3,4-[sup.13][C.sub.2]]cholesterol as the IS for IDMS measurements. The IS solution was prepared by dissolving 0.26 mmol (0.1 g) [3,4-[sup.13][C.sub.2]]cholesterol with 200 mL ethanol in a 200-mL, class A, volumetric flask. The final concentration was 1.29 mmol/L. We stored individual 8-mL portions at 4 [degrees]C until use.


We rehydrated lyophilized serum samples (SRM 909b; levels I and II; IFCC-RELA samples) according to the manufacturer's instructions. Frozen native serum samples and serum pools were allowed to thawand homogenize on a hematology mixer. We used a Digiflex TP automatic pipettor to aspirate 0.25 mL serum that was dispensed with 4.75 mL Tris-HCL buffer (0.05 mol/L, pH 7.5, 0.25% vol/vol Triton X-100) into clean, 8-mL screwcap vials, then mixed it on a hematology mixer. We placed 100 [micro]L diluted serum into a culture tube with 1.0 mL ethanolic potassium hydroxide (KOH) (0.32 mol/L) hydrolyzing mixture. We then added 100 [micro]L [3,4-[sup.13][C.sub.2]]cholesterol (1.29 mmol/L) and 1.0 mL ethanol/KOH to each culture tube and hydrolyzed the contents for 1 h at 50[degrees]C on a heating block. After hydrolysis, we added 3 mL water and 5 mL hexanes and extracted the cholesterol byliquid-liquid extraction in a mechanical shaker (15 min). We removed 4.5 mL of the hexanes layer to a clean culture tube and evaporated it to dryness in a Turbovap LV evaporator that was connected to a high-pressure liquid nitrogen dewar flask.

We used trisil BSA as a derivatizing agent. The dried extract was reconstituted in 100 [micro]L trisil BSA for derivatization at 70 [degrees]C.


We performed GC-MS analyses on an Agilent 5973 benchtop MSD that was connected to an Agilent 6890 gas chromatograph. The GC was equipped with a J&W DB5-ms capillary column (30 m by 0.5 mm by 0.1 [micro]m film thickness) and an autosampler. We operated the GC inlet in splitless mode. The operating temperature was 295 [degrees]C, and the carrier gas velocity was 40 cm/s. The GC oven was programmed as follows: initial temperature 250 [degrees]C, held for 1 min; temperature increased to 300 [degrees]C at 30 [degrees]C/min and held at 300 [degrees]C for 3 min. The transfer line between the GC and the MSD was 250 [degrees]C, and the MS source and quadrupoles were maintained at 230 [degrees]C and 150 [degrees]C, respectively. The MS was operated at 70 eV and in selected ion monitoring (SIM) mode.

We monitored masses m/z 368.4 and 370.4, corresponding to unlabeled and labeled cholesterol, respectively, with a dwell time of 30 ms. All calibration standards were analyzed in duplicate, and the corrected ratios of ion intensities [368.4/370.4] and the ratio of molar concentrations of cholesterol and IS were used to create a linear calibration curve to quantify cholesterol in serum samples. We corrected the calibration ion intensity ratios according to the theoretical model proposed by Siekmann (21) to compensate for deviations from linearity with increasing concentration resulting from MS signal overlapping between the labeled IS and the unlabeled native cholesterol, and vice versa.


To evaluate the specificity for cholesterol in the presence of a few possible interferences, we selected a mixture of sterol compounds representing plant sterols and precursors in the cholesterol biosynthesis pathway. We prepared a standard mixture of 7 sterols [lanosta-8,24-dien-3-ol; cholest-5en-3[beta]-ol (98%); [beta]-sitosterol; 5[alpha]-cholesten-7-ene-3[beta]-ol (95%); 5[alpha]-cholestan-3[beta]-ol; and ergosta-5,7,22-trien-3[beta]-ol (75%)] in toluene at 1.03 x [10.sup.-2] mmol/L sterol. We acquired the mass spectrum for each analyte in the mixture and selected the unique mass ion with the greatest intensity as the base peak for SIM analysis. We spiked 100 [micro]L sterol mixture into 1.0 mL native serum sample and analyzed the spiked serum as described earlier. We analyzed trimethylsilyl (TMS) derivatives of the dried hexanes extract with the MS set to monitor m/z 368.4 and 370.4, specifically. We determined the presence of interference if cholesterol values consistently increased or decreased in the spiked sample vs a sample without spiking.


We performed all statistical analyses with Analyze-it software and Microsoft Excel. We evaluated the imprecision of the GC-IDMS procedure by performing statistical analysis on the results from 4 analytical runs with quadruplicate measurements for SRM 1951b, levels I and II. In brief, we diluted 2 independent 250-[micro]L aliquots with Tris-HCl buffer solution and carried out duplicate GC-MS measurements from each aliquot for 4 measurements per analysis. We performed a nested ANOVA on the final calculated cholesterol concentrations to determine the intraassay, interassay, and total percent CVs.


We compared our GC-IDMS RMP with the NIST primary RMP to determine the accuracy of the cholesterol measurements. We prespecified targets as measureable goals to assess method accuracy. Briefly, the relative biases between the proposed GC-IDMS cholesterol measurements and the target concentrations on NIST-certified SRMs had to be [less than or equal to] 1.0%, and the slope and correlation coefficient of the regression had to be >0.95. We analyzed for cholesterol in SRM 1951b, levels I and II, and in SRM 909b, levels I and II. We obtained the GC-IDMS cholesterol target values for SRM 1951b and SRM 909b from the NIST SRM Certificates of Analysis (4, 5).


To arrive at a reliable linear relationship between AK cholesterol measurements and GC-IDMS, we compared the AK RMP cholesterol results for SRM 1951b, SRM 909b, 10 native serum samples, and 17 CDC serum pools vs the GC-IDMS measurements. We analyzed the samples in quadruplicate in 10 independent runs by both methods. The AK RMP was performed as described (22). To ensure comparability between the 2 procedures, the calibration standards used were prepared from the same batches of 911b. We used Analyze-it for Microsoft Excel to perform linear regression analysis on the averages of the run means for each sample.



Fig. 1 illustrates a typical linear regression of the corrected peak area ratios for mass ions m/z [368.4]/ [370.4] corresponding to unlabeled cholesterol and [3,4-[sup.13][C.sub.2]]cholesterol vs the ratio of molar concentrations of unlabeled to [3,4-[sup.13][C.sub.2]]cholesterol. The inset shows the mass chromatogram for unlabeled and labeled cholesterol. We evaluated the reproducibility of the slope and intercept from 10 analytical runs. The mean slope and intercept were 1.03 and 3.7 x [10.sup.-3], respectively, and the imprecision (percent CVs) of the slope and intercept were 0.9% and 12.4%, respectively.


Fig. 2A shows the GC-MS SIM chromatogram for a sample spiked with a standard mixture of 8 sterols, including cholesterol, found in serum. Fig. 2B shows a chromatogram for the TMS derivatives of a serum sample processed as described above. Despite similarities in the chemical structure and molecular weight of the various sterols, none of these compounds eluted at the same time as cholesterol (retention time 4.62 min). Moreover, they were undetected in the hexanes extract from serum samples, and cholesterol concentrations did not change in spiked serum vs samples without spiking.



We compared cholesterol values from the proposed GC-IDMS RMP to NIST assigned targets for SRMs. The mean relative bias from NIST reference values for SRM 1951b and SRM 909b was <0.6%. Table 1 summarizes the intraassay, interassay, and total SDs and percent CVs for SRM 1951b and 909b levels I and II. The total percent CVs were 0.61%, 0.73%, 0.8%, and 0.92% for the 2 levels of SRM 1951b and 909b, respectively. The table also includes the bias from the NIST reference values expressed as a percentage. Table 2 summarizes our results from the IFCC ring trial interlaboratory comparison of RMPs (20). We compared our cholesterol measurements to the group means for the IDMS methods participating in 2 survey periods.


Fig. 3 is a bias plot that summarizes observed relative biases of AK cholesterol vs the CDC's GC-IDMS cholesterol measurements for SRMs, native serum samples, and CDC pools. The relative biases ranged from 0.4% to 2.9%, with a mean bias of 1.6% for all samples investigated. The correlation equation represents a constant relationship between the AK RMP and GC-IDMS:

[AK.sub.[CHOL]] = 1.0181([GC-IDMS.sub.[CHOL]])-0.002426

The correlation coefficient ([r.sup.2]) equaled 0.9996, and the 95% CIs for the slope and intercept were 1.011 to 1.025 (P < 0.0001) and -0.0365 to 0.0317, respectively. The SE were 0.0034, 0.018, and 0.0280 mmol/L for the slope, intercept, and predicted AK cholesterol values, respectively. Table 3 shows actual and calculated AK cholesterol values in frozen (1951b) and lyophilized serum (909b) SRMs, native serum samples, and pooled sera. The mean difference between cholesterol values measured by the AK RMP and the calculated values obtained by the regression equation and the GC-IDMS data was 0.1% (range -0.74% to 1.03%) (Table 3). The mean difference for 7 serum pools that were not included in the batch of samples used to derive the equation was -0.4%.




Several sterol-type compounds circulate along with cholesterol in serum (23), and previous studies have shown that [beta]-sistosterol, campesterol, desmosterol, and lathosterol interfere with AK measurement of cholesterol (4). Plant sterols such as [beta]-sitosterol and campesterol in serum originate from oral consumption, whereas a few endogenous sterols such as lathosterol, desmosterol, and ergosterol are precursors in the biosynthesis of cholesterol and vitamin D (24, 25). Because their chemical structures are similar to cholesterol, they may interfere with MS analysis of cholesterol. Specifically, during data acquisition, mass ions originating from noncholesterol sterols that are identical to fragment masses in the cholesterol mass spectrum are not excluded, despite the high selectivity and specificityofMS analysis. However, the concentrations of many of these sterols such as lathosterol, campesterol and ergosterol are very low in nondiseased serum relative to cholesterol, and maybe undetectable in normal samples. The chromatogram in Fig. 2B confirms that none of the sterols we examined were detectable in diluted serum samples and further indicates that this MS RMP ensures an interference-free assay for cholesterol.


We achieved suitable imprecision with this GC-IDMS RMP. This was enabled by using instruments such as automatic pipettors and a stable MS system. Using the automatic pipettors at critical steps in the procedure complemented the reproducibility of GC-MS measurements and minimized systematic errors that would adversely affect precision. Consequently, the intraassay, interassay, and total percent CVs were maintained at < 1% as demonstrated for the lyophilized and fresh frozen serum samples (Table 1). Furthermore, analysis of the expanded uncertainty (U) associated with the IDMS measurements and transformation by the regression is important for RMP measurements (26, 27). The data presented in Supplemental File 1, which accompanies the online version of this article at http://, for the 7 serum pools demonstrates how the %U changed when the equation was used to transform the data set.

The proposed GC-IDMS RMP satisfied our set of criteria for acceptable accuracy. There was good enough agreement with the certified values for SRMs, with relative biases ranging from -1.15% to 0.21%. More specifically, there was better agreement for fresh frozen SRM (1951b) than for lyophilized SRM (909b). This difference may result from degradation or increased oxidation of cholesterol in the lyophilized SRM with aging. Despite the observed difference in relative biases, the accuracy makes the procedure suitable for use as a secondary RMP for cholesterol standardization.


In a few early reports, AK RMP measurements were subject to interferences from exogenous and endogenous noncholesterol sterols found in serum (6, 14). These sterols do not interfere in IDMS cholesterol measurements because IDMS procedures have high specificity for the analyte of interest. In contrast to the AK RMP, labeled and unlabeled cholesterol can be completely isolated from surrounding interferences, as illustrated in Fig. 2. In this procedure, the mean bias (1.6%) between AK and GC-IDMS cholesterol measurement is consistent, with relative biases reported by other researchers who have compared the procedures (3, 6, 14).

Because many studies rely on cholesterol measurements to normalize and stratify the results for easy interpretation, we took special consideration for how adopting this RMP in the lipid standardization program (LSP) will affect ongoing clinical and epidemiologic studies that are standardized by the AK RMP. Implementing this GC-IDMS RMP induces a minor shift of the accuracy base for cholesterol measurements, and ongoing studies using data standardized to the AK cholesterol RMP may detect or lose statistical significance when normalized to GC-IDMS cholesterol RMP results. To make up for this change, we propose using the linear equation represented above to convert cholesterol measurements between the 2 RMPs. This equation is suitable because the difference between the AK estimates calculated with this equation (for samples not included in the batch of samples used to derive the equation) and actual AK measurement, as shown in Supplemental File 1, were sufficiently small (mean difference -0.4%). Clinical and epidemiology studies that rely on cholesterol values that are traceable to the cholesterol AK RMP may not be severely affected by conversion from IDMS to AK cholesterol values.

In summary, the CDC's effort to standardize cholesterol measurements ensures continued equivalency and comparability of clinical data. Current metrological recommendations are for RMPs based on IDMS, which assure metrological traceability to procedures with a well-defined analyte and specificity. It is appropriate for the CDC to establish a higher-order RMP for standardizing cholesterol in the LSP. This RMP offers several advantages over the AK RMP and other IDMS procedures that employ a standard bracketing technique. These include improved specificity for cholesterol over the AK RMP, reduced analysis time (3 h vs 5.5 h by AK RMP), and increased efficiency without compromising the accuracyofmeasurements. The imprecision (%CV) maybe greater than values reported by other researchers (11, 14) for lyophilized and frozen SRMs; however, it is within the CLSI goals for RMPs (28). Furthermore, this GC-IDMS RMP demonstrates excellent agreement with other IDMS RMPs for cholesterol measurements, further confirming that this RMP is suitable for standardizing cholesterol measurements. The CDC plans to implement the GC-IDMS method as the RMP to assign cholesterol values for commutable reference materials used in its standardization programs.

Author Contributions: All authors confirmed they have contributed to the intellectual content of this paper and have met the following 3 requirements: (a) significant contributions to the conception and design, acquisition of data, or analysis and interpretation of data; (b) drafting or revisingthe article for intellectual content; and (c) final approval of the published article.

Authors' Disclosures or Potential Conflicts of Interest: Upon manuscript submission, all authors completed the Disclosures of Potential Conflict of Interest form. Potential conflicts of interest:

Employment or Leadership: G.L. Myers, AACC.

Consultant or Advisory Role: None declared.

Stock Ownership: None declared.

Honoraria: None declared.

Research Funding: Interagency Agreement with the National Heart, Lung and Blood Institute (NHLBI) and funding from the Division for Heart Disease and Stroke Prevention, National Center for Chronic Disease Prevention and Health Promotion (NCCDPHP), CDC.

Expert Testimony: None declared.

Other remuneration: G.L. Myers, travel reimbursement for speaking at an Asian-Pacific Congress of Clinical Biochemistry (APCCB) meeting.

Role of Sponsor: The funding organizations played no role in the design of study, choice of enrolled patients, review and interpretation of data, or preparation or approval of manuscript.


(1.) Abell LL, Levy BB, Brodie BB, Kendall FE. A simplified method for the estimation of total cholesterol in serum and demonstration of its specificity. J Biol Chem 1952;195:357-66.

(2.) Current status of blood cholesterol measurement in clinical laboratories in the United States: a report from the Laboratory Standardization Panel of the National Cholesterol Education Program. Clin Chem 1988;34:193-201.

(3.) Ellerbe P, Myers GL, Cooper GR, Hertz HS, Sniegoski LT, Welch MJ, White E. A comparison of results for cholesterol in human serum obtained by the reference method and by the definitive method of the National Reference System for cholesterol. Clin Chem 1990;36:370-5.

(4.) Wise SA, Watters RL. NIST certificate of analysis: standard reference material 1951b, lipids in frozen serum. view_cert.cfm?srm=1951B (Accessed October 2010).

(5.) Wise SA, Watters RL. NIST certificate of analysis: standard reference material 909b; lipids in frozen serum. cfm?srm=909B (Accessed October 2010).

(6.) Bernert JT Jr, Akins JR, Cooper GR, Poulose AK, Myers GL, Sampson EJ. Factors influencing the accuracy of the national reference system total cholesterol reference method. Clin Chem 1991; 37:2053-61.

(7.) Panteghini M. Traceability, reference systems and result comparability. Clin Biochem Rev 2007;28: 97-104.

(8.) Tate J, Panteghini M. Standardisation: the theory and the practice. Clin Biochem Rev 2007;28: 93-6.

(9.) Cohen A, Hertz HS, Mandel J, Paule RC, Schaffer R, Sniegoski LT, et al. Total serum cholesterol by isotope dilution/mass spectrometry: a candidate definitive method. Clin Chem 1980;26:854-60.

(10.) Ellerbe P, Meiselman S, Sniegoski LT, Welch MJ, White E. Determination of serum cholesterol by a modification of the isotope dilution mass spectrometric definitive method. Anal Chem 1989;61: 1718-23.

(11.) Pelletier O, Wright LA, Breckenridge WC. Isotope-dilution mass-spectrometry of serum-cholesterol with [3,4-C-13]cholesterol: proposed definitive method. Clin Chem 1987;33:1403-11.

(12.) Thienpont LM, Deleenheer AP, Stockl D, Reinauer H. Candidate reference methods for determining target values for cholesterol, creatinine, uric-acid, and glucose in external quality assessment and internal accuracy control. 2. Method transfer. Clin Chem 1993;39:1001-6.

(13.) Derks HJ, van Heiningen A, Koedam HC Gaschromatographic determination of cholesterol in serum: candidate reference method. Clin Chem 1985;31:691-4.

(14.) Chen WX, Li PY, Wang S, Dong J, Li JZ. Serum cholesterol determined by liquid chromatography with 6-chlorostigmasterol as internal standard. Clin Chem 1993;39:1602-7.

(15.) Siekmann L Reference methods for total cholesterol and total glycerol. Eur J Clin Chem Clin Biochem 1991;29:277-9.

(16.) Siekmann L, Huskes KP, Breuer H Determination of cholesterol in serum using mass fragmentography: a reference method in clinical chemistry. Z Anal Chem 1976;279:145-6.

(17.) Geistanger A, Arends S, Berding C, Hoshino T, Jeppsson JO, Little R, et al. Statistical methods for monitoring the relationship between the IFCC reference measurement procedure for hemoglobin A1c and the designated comparison methods in the United States, Japan, and Sweden. Clin Chem 2008;54:1379-85.

(18.) Weykamp C, John WG, Mosca A, Hoshino T, Little R, Jeppsson JO, et al. The IFCC reference measurement system for HbA1c: a 6-year progress report. Clin Chem 2008;54:240-8.

(19.) NCCLS. Preparation and validation of commutable frozen serumpoolsassecondaryreferencematerials for cholesterol measurement procedures; approved guidelines. NCCLS document C37A, Wayne (PA): NCCLS; 1999.

(20.) IFCC. RELA: IFCC External Quality Assessment Scheme for Reference Laboratories in Laboratory Medicine. (Accessed December 2010).

(21.) Siekmann L. Determination of steroid hormones by the use of isotope dilution-mass spectrometry: a definitive method in clinical chemistry. J Steroid Biochem 1979;11:117-23.

(22.) Cooper GR, Smith SJ, Duncan IW, Mather A, Fellows WD, Foley T, et al. Interlaboratory testing of the transferability of a candidate reference method for total cholesterol in serum. Clin Chem 1986;32:921-9.

(23.) Ahmida HSM, Bertucci P, Franzo L, Massoud R, Cortese C, Lala A, Federici G. Simultaneous determination of plasmatic phytosterols and cholesterol precursors using gas chromatography-mass spectrometry (GC-MS) with selective ion monitoring (SIM). J Chromatogr B 2006;842:43-7.

(24.) Holick MF. Vitamin D deficiency. N Engl J Med 2007;357:266-81.

(25.) Miettinen TA, Tilvis RS, Kesaniemi YA. Serum plant sterols and cholesterol precursors reflect cholesterol absorption and synthesis in volunteers of a randomly selected male population. Am J Epidemiol 1990;131:20-31.

(26.) Hoelzel W, Weykamp C, Jeppsson JO, Miedema K, Barr JR, Goodall I, et al. IFCC reference system for measurement of hemoglobin A1c in human blood and the national standardization schemes in the United States, Japan, and Sweden: a method-comparison study. Clin Chem 2004;50: 166-74.

(27.) EURACHEM/CITAC Guide CG 4: quantifying uncertainty in analytical measurement, 2nd ed. Ellison SLR, Rosslein M, Williams A. http://www. (Accessed December 2010).

(28.) NCCLS. Method comparison and bias estimation using patient samples; approved guidelines, 2nd ed. NCCLS document EP9-A. Wayne (PA): NCCLS; 1995.

Selvin H. Edwards, [1] * Mary M. Kimberly, [1] Susan D. Pyatt, [1] Shelton L. Stribling, [2] Kara D. Dobbin, [2] and Gary L. Myers [1]

[1] Clinical Chemistry Branch, Division of Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention, Atlanta, GA; [2] Battelle Memorial Institute, Atlanta, GA.

[3] Nonstandard abbreviations: RMP, reference measurement procedure; AK, Abell-Levy-Brodie-Kendall RMP; GC-IDMS, gas chromatography-isotope dilution mass spectrometry; IS, internal standard; SRM, standard reference material; CLSI, Clinical Laboratory Standards Institute; RELA, IFCC External Quality Assessment Scheme for Reference Laboratories in Laboratory Medicine; trisil BSA, N,O-b/s(trimethylsilyl)acetamide; MSD, mass selective detector; SIM, selected ion monitoring; TMS, trimethylsilyl; LSP, lipid standardization program.

* Address correspondence to this author at: Centers for Disease Control and Prevention, 4770 Buford Hwy., MS-F25, Chamblee, GA 30341. Fax 770-4887030; e-mail

Received November 5, 2010; accepted January 27, 2011.

Previously published online at DOI: 10.1373/clinchem.2010.158766

The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.
Table 1. Imprecision and accuracy of serum cholesterol
measurements by GC-IDMS for SRMs.

 SRM 909b

 Level I Level II

Mean (SD), mmol/L 3.76 (0.03) 6.01 (0.06)
 Intraassay SD (%CV) 0.016(0.43) 0.020 (0.33)
 Interassay SD (%CV) 0.025 (0.66) 0.051 (0.85)
 Total SD (%CV) 0.03 (0.81) 0.055 (0.92)
 Mean NIST target value (SD) 3.79 (0.05) 6.08 (0.08)
 Bias vs NIST, % -0.77 -1.15

 SRM 1951b

 Level I Level II

Mean (SD), mmol/L 4.81 (0.03) 6.88 (0.05)
 Intraassay SD (%CV) 0.021 (0.44) 0.032 (0.47)
 Interassay SD (%CV) 0.019(0.41) 0.039 (0.56)
 Total SD (%CV) 0.029(0.61) 0.050 (0.73)
 Mean NIST target value (SD) 4.80 (0.014) 6.89 (0.022)
 Bias vs NIST, % 0.21 -0.15

Table 2. Interlaboratory comparison of GC-IDMS RMP for
cholesterol in lyophilized serum, IFCC ring trial.

Sample ID Mean cholesterol Group mean Difference,
 (SD), mmol/L (SD), mmol/L %

IFCC RELA 2007, 5.188(0.011) 5.207 (0.028) 0.36
 level I
IFCC RELA 2009, 3.564 (0.03) 3.579 (0.028) 0.4
 level I
IFCC RELA 2007, 5.675 (0.011) 5.724 (0.034) 0.9
 level II
IFCC RELA 2009, 4.697 (0.042) 4.701 (0.023) 0.08
 level II

Table 3. Mean cholesterol values for serum measured by GC-IDMS,
AK RMP, and the corresponding calculated AK cholesterol value,
and percent difference between actual AK and calculated values.

Serum Cholesterol, mmol/L

 (actual) (calculated) (actual
 - calculated)

 909b level I 3.791 3.875 3.835 1.03
 909b level II 6.069 6.179 6.176 0.05
 1951b level I 4.789 4.846 4.873 -0.56
 1951b level II 6.872 6.962 6.993 -0.45
Native serum
 1 3.494 3.556 3.555 0.03
 2 3.907 3.993 3.995 -0.05
 3 4.359 4.442 4.436 -0.15
 4 4.459 4.562 4.537 0.55
 5 4.692 4.768 4.775 0.15
CDC pools
 1 1.334 1.356 1.356 0.0
 2 3.947 4.013 4.016 -0.07
 3 4.445 4.525 4.526 -0.02
 4 6.706 6.775 6.825 -0.74
 5 7.085 7.217 7.211 0.08
COPYRIGHT 2011 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2011 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Lipids, Lipoproteins, and Cardiovascular Risk Factors
Author:Edwards, Selvin H.; Kimberly, Mary M.; Pyatt, Susan D.; Stribling, Shelton L.; Dobbin, Kara D.; Myer
Publication:Clinical Chemistry
Date:Apr 1, 2011
Previous Article:Maternal serum visfatin at 11-13 weeks of gestation in gestational diabetes mellitus.
Next Article:Newborn screening for isovaleric acidemia using tandem mass spectrometry: data from 1.6 million newborns.

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