Letters to the Editor Improved Reference Measurement Method for Hemoglobin A1c by Use of Liquid Chromatography– Isotope Dilution–Tandem Mass Spectrometry
Table 1. LC-ID-MS/MS measurements of the samples in RELA 2013. Category
© 2014 American Association for Clinical Chemistry 1 Hb A1c, hemoglobin A1c; LC-ID-MS/MS, liquid chromatography–isotope dilution–tandem mass spectrometry; MU, measurement uncertainty; VE, -chain N-terminal Val-His-Leu-Thr-Pro-Glu; GE, 1-deoxyfructoxyl-Val-His-Leu-Thr-Pro-Glu; NGSP, National Glycohemoglobin Standardization Program.
Sample B
VE concentration, μmol/g (n = 8)
To the Editor: We have developed an improved reference measurement method for hemoglobin A1c (Hb A1c)1 based on liquid chromatography–isotope dilution–tandem mass spectrometry (LC-ID-MS/MS) with traceability to the SI unit. The method has a small measurement uncertainty (MU) and gives results in good agreement with the accuracy-based IFCC reference method. Kaiser et al. (1 ) demonstrated the possibility of using ID-MS for Hb A1c measurement, which involved the proteolysis of hemoglobins using endoproteinase Glu-C and hydrolysis of hexapeptide calibration standards using HCl. However, a large MU and significant difference in results compared with the IFCC reference method were observed. Using the same approach, we developed an improved procedure based on LC-ID-MS/MS and used the method for participation in an IFCC ring trial for reference laboratories (RELA 2013) for Hb A1c, in which our results for 2 lyophilized samples were compared against those from the IFCC reference method, including 6 approved IFCC network laboratories for Hb A1c. Key steps in our ID-MS procedure to preserve the unbroken traceability to the SI unit are the hydrolysis of the hexapeptide calibration standards [-chain N-terminal Val-HisLeu-Thr-Pro-Glu (VE) and 1-
Sample A
Mean
3.500
3.147
SD, μmol/g
0.024
0.034
CV, %
0.69
1.08
GE concentration, μmol/g (n = 8) Mean
0.12854
0.2909
SD, μmol/g
0.00075
0.0025
CV, %
0.58
Hb A1c value (obtained), mmol/mol
84.6
U, mmol/mola
0.98
2.4
% Ua
2.8
2.8
35.02
83.9
b
Hb A1c value (target), mmol/mol
a
b
0.86
35.42
Deviation from target value, mmol/mol
0.4
0.7
Relative deviation from target value, %
1.14
0.83
Expanded uncertainties were obtained by multiplying standard uncertainty by a coverage factor of 2. U, the expanded measurement uncertainty of the obtained Hb A1c value; % U, relative expanded measurement uncertainty. The means of the participating laboratories’ results in RELA 2013 were used as the target values. One outlier for sample A was excluded.
deoxyfructoxyl-Val-His-Leu-ThrPro-Glu (GE)] and the proteolysis of Hb A0 and Hb A1c. The use of hexapeptide calibration standards allows the traceability to amino acid certified reference materials, which is more reliable than the traceability to reference protein standards in the IFCC reference procedure (1 ). To ensure accurate measurement, complete hydrolysis of the hexapeptides must be achieved. We found that the use of 1% phenol in 6 mol/L HCl markedly reduced the hydrolysis time from the previously reported 65 h (1 ) to 24 h. In addition, amino acids that are stable in acidic environment, such as leucine and proline, were suitable for the quantification of the hexapeptides. This was demonstrated by the consistency between the results of leucine and proline (deviation ⬍0.6%) for both VE and GE. Valine was found to be unsuitable, possibly owing to its linkage to a deoxyfructoxyl group in GE. Because amino acid analysis is af-
fected by both peptide and amino acid impurities, these impurities would need to be quantified (by HPLC and LC-MS/MS, respectively) to ensure the accuracy of the analysis and subsequently the Hb A1c measurement. In our case, VE had satisfactory purity, but GE contained VE as an impurity. The concentration of GE calibration solution was therefore corrected by the quantification of VE in GE to remove the positive bias in the Hb A1c measurement. In proteolysis, incomplete reaction will result in significant deviations between the amounts of hexapeptides obtained and the true amounts of Hb A1c and Hb A0. We found that adding additional endoproteinase Glu-C was crucial to ensure complete proteolysis [125 g/ mg hemoglobin vs the reported amount of 10 g (2 )]. We believe that the complete proteolysis of hemoglobins markedly improves the imprecision and accuracy and further enhances the robustness of our ID-MS method. Clinical Chemistry 61:2 (2015) 435
Letters to the Editor Method imprecision (type A MU) is one of the major components in overall MU. It is possible to achieve a small MU by improving the imprecision of the sample preparation procedure. Other than ensuring complete proteolysis, we observed that adding the isotopelabeled internal standard before proteolysis provided better imprecision than after proteolysis. This may allow a better control of the isotope ratio throughout the proteolytic process and enhance method imprecision. In addition, the reconstituted sample was also weighed directly into the buffer solution (pH ⫽ 4). This ensured within-vial homogeneity of the sample blends and prevented the sample from forming aggregates before proteolysis. As shown in Table 1, with the improvement of peptide hydrolysis and hemoglobin proteolysis, the deviation between our LC-IDMS/MS method and the IFCC reference method was reduced to 0.4 – 0.7 mmol/mol [0.04%– 0.06% in National Glycohemoglobin Standardization Program (NGSP) units] compared with previously reported 2–3 mmol/mol (1 ). This demonstrates the good agreement between our LC-ID-MS/MS method and the IFCC reference method. Imprecision with CVs close to or ⬍1% was achieved for both VE and GE in our method. The relative MUs obtained were 2.8% for both samples, or 1.7% and 2.2% after converting to NGSP units (3 ), which were comparable to the desirable CVs of routine clinical laboratories [2% and 3.5% (NGSP) for intra- and interlaboratory CV, respectively] (4 ). Because our ID-MS method is traceable to the SI unit, it supports the accuracy of the current IFCC calibration system, which is based on purified protein standards. It may also be extended to the NGSP calibration system with the application of the master equation (5 ). 436
Clinical Chemistry 61:2 (2015)
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 revising the article for intellectual content; and (c) final approval of the published article. Authors’ Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
References 1. Kaiser P, Akerboom T, Ohlendorf R, Reinauer H. Liquid chromatography–isotope dilution–mass spectrometry as a new basis for the reference measurement procedure for hemoglobin A1c determination. Clin Chem 2010;56:750–4. 2. Kaiser P, Akerboom T, Molnar P, Reinauer H. Modified HPLC-electrospray ionization/mass spectrometry method for HbA1c based on IFCC reference measurement procedure. Clin Chem 2008;54:1018 –22. 3. Weykamp CW, Mosca A, Gillery P, Panteghini M. The analytical goals for hemoglobin A1c measurement in IFCC units and National Glycohemoglobin Standardization Program units are different. Clin Chem 2011;57:1204–5. 4. Sacks DB, Arnold M, Bakris GL, Bruns DE, Horvath AR, Kirkman MS, et al. Executive summary: guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Clin Chem 2011;57:793– 8. 5. Hoelzel W, Weykamp C, Jeppsson J-O, Miedema K, Barr JR, Goodall I, et al. IFCC reference system for measurement of hemoglobin A1c in human bood and the national standardization schemes in the United States, Japan, and Sweden: a method-comparison study. Clin Chem 2004;50:166 –74.
Lingkai Wong2 Hong Liu2 Sharon Yong2 Qinde Liu2* Tong Kooi Lee2 2
Chemical Metrology Division Applied Sciences Group Health Sciences Authority Singapore
* Address correspondence to this author at: 1 Science Park Rd #01-05/06, The Capricorn Singapore Science Park II Singapore 117528 E-mail [email protected]. Previously published online at DOI: 10.1373/clinchem.2014.231340
Cardiac Troponin Testing Is Overused after the Rule-In or Rule-Out of Myocardial Infarction To the Editor: No good studies have systematically evaluated appropriate clinical utilization of cardiac troponin testing in the clinical setting of the rule-in and rule-out of myocardial infarction (MI)1 (1, 2 ). Our collective 100plus years of clinical and laboratory experience suggested that provider test ordering and use of cardiac troponin has been excessive after a diagnosis of MI or no MI has been determined. There is no evidence that supports continuation of cardiac troponin testing after a diagnosis is made (2–5 ). The goal of this study was to determine whether clinicians appropriately use cardiac troponin I (cTnI) in the assessment of patients at moderate to high risk of acute coronary syndrome (ACS) and MI. After receiving institutional review board approval, we retrospectively reviewed electronic health/ medical records in 100 consecutive patients who had serial cTnI orders in the cardiac renal unit, where patients at moderate to high risk for MI are evaluated. Diagnoses were adjudicated by either an emergency medicine physician or a cardiologist as MI or no MI according to the 2012 Third Universal Definition of MI guidelines (2 ). A cTnI order set consisted of measurements obtained at 0, 3, 6, and 9 h (Ortho-Clinical Diagnostics Vitros ES, 99th percentile 0.034 g/L) (5 ). Clinicians were not limited to any number of order sets. Excessive cTnI measurements were defined as any cTnI ordered and measured after the provider made the diagnosis of MI or no MI. All cTnI results during hospitalization were tabulated.
© 2014 American Association for Clinical Chemistry 1 Nonstandard abbreviations: MI, myocardial infarction; cTnI,cardiactroponinI;ACS,acutecoronarysyndrome;PCI, percutaneous coronary intervention.
Table 1. LC-ID-MS/MS measurements of the samples in RELA 2013. Category
© 2014 American Association for Clinical Chemistry 1 Hb A1c, hemoglobin A1c; LC-ID-MS/MS, liquid chromatography–isotope dilution–tandem mass spectrometry; MU, measurement uncertainty; VE, -chain N-terminal Val-His-Leu-Thr-Pro-Glu; GE, 1-deoxyfructoxyl-Val-His-Leu-Thr-Pro-Glu; NGSP, National Glycohemoglobin Standardization Program.
Sample B
VE concentration, μmol/g (n = 8)
To the Editor: We have developed an improved reference measurement method for hemoglobin A1c (Hb A1c)1 based on liquid chromatography–isotope dilution–tandem mass spectrometry (LC-ID-MS/MS) with traceability to the SI unit. The method has a small measurement uncertainty (MU) and gives results in good agreement with the accuracy-based IFCC reference method. Kaiser et al. (1 ) demonstrated the possibility of using ID-MS for Hb A1c measurement, which involved the proteolysis of hemoglobins using endoproteinase Glu-C and hydrolysis of hexapeptide calibration standards using HCl. However, a large MU and significant difference in results compared with the IFCC reference method were observed. Using the same approach, we developed an improved procedure based on LC-ID-MS/MS and used the method for participation in an IFCC ring trial for reference laboratories (RELA 2013) for Hb A1c, in which our results for 2 lyophilized samples were compared against those from the IFCC reference method, including 6 approved IFCC network laboratories for Hb A1c. Key steps in our ID-MS procedure to preserve the unbroken traceability to the SI unit are the hydrolysis of the hexapeptide calibration standards [-chain N-terminal Val-HisLeu-Thr-Pro-Glu (VE) and 1-
Sample A
Mean
3.500
3.147
SD, μmol/g
0.024
0.034
CV, %
0.69
1.08
GE concentration, μmol/g (n = 8) Mean
0.12854
0.2909
SD, μmol/g
0.00075
0.0025
CV, %
0.58
Hb A1c value (obtained), mmol/mol
84.6
U, mmol/mola
0.98
2.4
% Ua
2.8
2.8
35.02
83.9
b
Hb A1c value (target), mmol/mol
a
b
0.86
35.42
Deviation from target value, mmol/mol
0.4
0.7
Relative deviation from target value, %
1.14
0.83
Expanded uncertainties were obtained by multiplying standard uncertainty by a coverage factor of 2. U, the expanded measurement uncertainty of the obtained Hb A1c value; % U, relative expanded measurement uncertainty. The means of the participating laboratories’ results in RELA 2013 were used as the target values. One outlier for sample A was excluded.
deoxyfructoxyl-Val-His-Leu-ThrPro-Glu (GE)] and the proteolysis of Hb A0 and Hb A1c. The use of hexapeptide calibration standards allows the traceability to amino acid certified reference materials, which is more reliable than the traceability to reference protein standards in the IFCC reference procedure (1 ). To ensure accurate measurement, complete hydrolysis of the hexapeptides must be achieved. We found that the use of 1% phenol in 6 mol/L HCl markedly reduced the hydrolysis time from the previously reported 65 h (1 ) to 24 h. In addition, amino acids that are stable in acidic environment, such as leucine and proline, were suitable for the quantification of the hexapeptides. This was demonstrated by the consistency between the results of leucine and proline (deviation ⬍0.6%) for both VE and GE. Valine was found to be unsuitable, possibly owing to its linkage to a deoxyfructoxyl group in GE. Because amino acid analysis is af-
fected by both peptide and amino acid impurities, these impurities would need to be quantified (by HPLC and LC-MS/MS, respectively) to ensure the accuracy of the analysis and subsequently the Hb A1c measurement. In our case, VE had satisfactory purity, but GE contained VE as an impurity. The concentration of GE calibration solution was therefore corrected by the quantification of VE in GE to remove the positive bias in the Hb A1c measurement. In proteolysis, incomplete reaction will result in significant deviations between the amounts of hexapeptides obtained and the true amounts of Hb A1c and Hb A0. We found that adding additional endoproteinase Glu-C was crucial to ensure complete proteolysis [125 g/ mg hemoglobin vs the reported amount of 10 g (2 )]. We believe that the complete proteolysis of hemoglobins markedly improves the imprecision and accuracy and further enhances the robustness of our ID-MS method. Clinical Chemistry 61:2 (2015) 435
Letters to the Editor Method imprecision (type A MU) is one of the major components in overall MU. It is possible to achieve a small MU by improving the imprecision of the sample preparation procedure. Other than ensuring complete proteolysis, we observed that adding the isotopelabeled internal standard before proteolysis provided better imprecision than after proteolysis. This may allow a better control of the isotope ratio throughout the proteolytic process and enhance method imprecision. In addition, the reconstituted sample was also weighed directly into the buffer solution (pH ⫽ 4). This ensured within-vial homogeneity of the sample blends and prevented the sample from forming aggregates before proteolysis. As shown in Table 1, with the improvement of peptide hydrolysis and hemoglobin proteolysis, the deviation between our LC-IDMS/MS method and the IFCC reference method was reduced to 0.4 – 0.7 mmol/mol [0.04%– 0.06% in National Glycohemoglobin Standardization Program (NGSP) units] compared with previously reported 2–3 mmol/mol (1 ). This demonstrates the good agreement between our LC-ID-MS/MS method and the IFCC reference method. Imprecision with CVs close to or ⬍1% was achieved for both VE and GE in our method. The relative MUs obtained were 2.8% for both samples, or 1.7% and 2.2% after converting to NGSP units (3 ), which were comparable to the desirable CVs of routine clinical laboratories [2% and 3.5% (NGSP) for intra- and interlaboratory CV, respectively] (4 ). Because our ID-MS method is traceable to the SI unit, it supports the accuracy of the current IFCC calibration system, which is based on purified protein standards. It may also be extended to the NGSP calibration system with the application of the master equation (5 ). 436
Clinical Chemistry 61:2 (2015)
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 revising the article for intellectual content; and (c) final approval of the published article. Authors’ Disclosures or Potential Conflicts of Interest: No authors declared any potential conflicts of interest.
References 1. Kaiser P, Akerboom T, Ohlendorf R, Reinauer H. Liquid chromatography–isotope dilution–mass spectrometry as a new basis for the reference measurement procedure for hemoglobin A1c determination. Clin Chem 2010;56:750–4. 2. Kaiser P, Akerboom T, Molnar P, Reinauer H. Modified HPLC-electrospray ionization/mass spectrometry method for HbA1c based on IFCC reference measurement procedure. Clin Chem 2008;54:1018 –22. 3. Weykamp CW, Mosca A, Gillery P, Panteghini M. The analytical goals for hemoglobin A1c measurement in IFCC units and National Glycohemoglobin Standardization Program units are different. Clin Chem 2011;57:1204–5. 4. Sacks DB, Arnold M, Bakris GL, Bruns DE, Horvath AR, Kirkman MS, et al. Executive summary: guidelines and recommendations for laboratory analysis in the diagnosis and management of diabetes mellitus. Clin Chem 2011;57:793– 8. 5. Hoelzel W, Weykamp C, Jeppsson J-O, Miedema K, Barr JR, Goodall I, et al. IFCC reference system for measurement of hemoglobin A1c in human bood and the national standardization schemes in the United States, Japan, and Sweden: a method-comparison study. Clin Chem 2004;50:166 –74.
Lingkai Wong2 Hong Liu2 Sharon Yong2 Qinde Liu2* Tong Kooi Lee2 2
Chemical Metrology Division Applied Sciences Group Health Sciences Authority Singapore
* Address correspondence to this author at: 1 Science Park Rd #01-05/06, The Capricorn Singapore Science Park II Singapore 117528 E-mail [email protected]. Previously published online at DOI: 10.1373/clinchem.2014.231340
Cardiac Troponin Testing Is Overused after the Rule-In or Rule-Out of Myocardial Infarction To the Editor: No good studies have systematically evaluated appropriate clinical utilization of cardiac troponin testing in the clinical setting of the rule-in and rule-out of myocardial infarction (MI)1 (1, 2 ). Our collective 100plus years of clinical and laboratory experience suggested that provider test ordering and use of cardiac troponin has been excessive after a diagnosis of MI or no MI has been determined. There is no evidence that supports continuation of cardiac troponin testing after a diagnosis is made (2–5 ). The goal of this study was to determine whether clinicians appropriately use cardiac troponin I (cTnI) in the assessment of patients at moderate to high risk of acute coronary syndrome (ACS) and MI. After receiving institutional review board approval, we retrospectively reviewed electronic health/ medical records in 100 consecutive patients who had serial cTnI orders in the cardiac renal unit, where patients at moderate to high risk for MI are evaluated. Diagnoses were adjudicated by either an emergency medicine physician or a cardiologist as MI or no MI according to the 2012 Third Universal Definition of MI guidelines (2 ). A cTnI order set consisted of measurements obtained at 0, 3, 6, and 9 h (Ortho-Clinical Diagnostics Vitros ES, 99th percentile 0.034 g/L) (5 ). Clinicians were not limited to any number of order sets. Excessive cTnI measurements were defined as any cTnI ordered and measured after the provider made the diagnosis of MI or no MI. All cTnI results during hospitalization were tabulated.
© 2014 American Association for Clinical Chemistry 1 Nonstandard abbreviations: MI, myocardial infarction; cTnI,cardiactroponinI;ACS,acutecoronarysyndrome;PCI, percutaneous coronary intervention.
