- Email: [email protected]
Performance evaluation of a third party enzymatic total bile acid assay for use on the Roche Integra 800 analyzer
Abstracts b
Clinical Biochemistry, The Hospital for Sick Children, Toronto, ON, Canada
Objectives: In the diagnosis of central nervous system germ cell tumours (CNS GCTs), alpha-fetoprotein (AFP) in cerebrospinal fluid (CSF) above 10 ug/L distinguish non-germinomatous germ cell tumours from germinomas while human chorionic gondatropin (hCG) above 5 IU/L confirm a germinoma. During follow-up, AFP or hCG above these concentrations suggest recurrence. To determine suitability of the Abbott Architect AFP and total beta-hCG assays for these purposes, we validated their analytical performance in CSF. Design and methods: Evaluation was performed with CSF pools spiked with AFP or beta-hCG from serum samples. For imprecision, pools were measured in duplicate over 10 days. To determine values in absence of disease, measurements were made in 30 individuals free from pregnancy, trophoblastic disease or malignancy. Results: Recovery ranged from 81–90% for beta-hCG and 87–101% for AFP in CSF compared to serum. Functional sensitivity was 1.59 IU/L for beta-hCG and 0.26 ug/L for AFP. Linearity was confirmed from 2.8 to 27.8 IU/L for beta-hCG and 2.4 to 45.2 ug/L for AFP. For imprecision, CV values were from 4.1–7.7% for beta-hCG and 1.8–2.6% for AFP. Measurement of samples from individuals free from pregnancy, trophoblastic disease or malignancy were below the measuring range of 1.2 IU/L for beta-hCG and ranged from 0–0.05 ug/L for AFP, confirming their absence in normal CSF. Conclusions: The Abbott Architect assays for beta-hCG and AFP can be used in CSF without significant matrix effects and with good precision, linearity and functional sensitivity and are, thus, suitable for use in the management of CNS GCTs. doi:10.1016/j.clinbiochem.2015.07.054
312 Development of a rapid LC–MS/MS method for hair cortisol determination to assess the HPA axis Laura Smya, Michael Leadleyb, Katarina Aleksab, Gideon Korenb a University of Toronto and The Hospital for Sick Children, Toronto, ON, Canada b The Hospital for Sick Children, Toronto, ON, Canada Objectives: Hair cortisol is an increasingly utilized biomarker to assess the hypothalamic-pituitary-adrenal (HPA) axis. Often, hair cortisol is measured using enzyme-linked immunosorbent assays (ELISA). Other LC–MS methods detect multiple steroid compounds leading to long run times. To improve detection over ELISA, and LC– MS turn-around-times and costs, we are developing a rapid UHPLC– MS/MS method to measure hair cortisol in 2 min. Design and methods: Cortisol was extracted with methanol from hair samples collected from the head vertex posterior region. The supernatants were treated with C18 solid phase extraction (SPE) or filtered, dried down, reconstituted with methanol:water, and analyzed using UHPLC with a C18 cartridge and a mobile phase gradient. The mass spectrometry cortisol transition was 407.2 to 335.1 m/z and the cortisol-D4 transition was 411.2 to 339.1 m/z. Potential interference from the presence of cortisone (405.1 to 329.1 m/z) and prednisone (403.1 to 327.1 m/z.) were assessed. Results: With this method, hair cortisol results are obtained in a 2.0-minute run time. The retention time for cortisol and cortisol-D4 was 0.82 ± 0.01 min. The calibration curve is linear with a correlation coefficient ≥ 0.998. The LOD is 0.33 pmol/g for samples pretreated with SPE, or 14.9 pmol/g for filtered samples. The intra-day coefficient of variation was 2.5%. Further, cortisone and prednisone do not contribute to the cortisol peak with this method.
1013
Conclusions: Hair cortisol levels can successfully be determined in a short run time of 2 min without interference from the cortisone, or any potential prednisone, present in the hair. Keywords: Hair cortisol, rapid LC–MS/MS, hypothalamus–pituitary– adrenal doi:10.1016/j.clinbiochem.2015.07.055
313 Performance evaluation of the epoc® point-of-care blood analysis system Yu Chena, Monique Gormanb, Bill O'Reillyb a Horizon Health Network, Fredericton, NB, Canada b Saint John Regional Hospital, Saint John NB, Canada Objectives: The aim of this study was to evaluate the analytical performance of the new Alere epoc® point-of-care blood analysis system. Design and methods: The precision study was conducted on 3 epoc® blood analysis systems using 5 levels of quality control materials twice per day for 5 days. A blood gas syringe, a gold top tube, a grey top tube, and a lavender top tube were collected for the comparison study on cardiac perfusion patients (n = 40). The blood gas syringe samples were tested on epoc® (all 3 m), Instrumentation Laboratory GEM4000, Abott iSTAT, and Nova CCX meters. Gold top tube and grey top tube samples were tested on Roche Modular P for electrolytes, glucose and lactate. Lavender samples were tested on Beckman Coulter LH-780 for hemoglobin. Results: The epoc® blood analysis systems demonstrated clinically acceptable precision for all analytes (from 0.07%, 0.07%, and 0.13% for pH 7.6, 7.4, and 7.0 levels; to 3.87%, 3.74%, and 7.56% for pO2 197, 103, and 56 mm Hg levels). Comparison studies yielded a correlation coefficient R from 0.9201 (sodium) to 0.9969 (pO2) with the GEM4000; from 0.9071 (sodium) to 0.9965 (potassium) with the iSTAT; from 0.8793 (sodium) to 0.9957 (pO2) with the CCX; from 0.8463 (sodium) to 0.9942 (potassium) with Modular; and 0.9557 (hemoglobin) with LH-780. Average biases for all analytes were within the total allowable error limits. Conclusions: The Alere epoc® blood analysis system is acceptable for point-of-care testing in the hospital setting. doi:10.1016/j.clinbiochem.2015.07.056
314 Performance evaluation of a third party enzymatic total bile acid assay for use on the Roche Integra 800 analyzer Yu Chen, Hilary Smith, Heather Tarr Department of Laboratory Medicine, Dr. Everett Chalmers Regional Hospital, Horizon Health Network, Fredericton, NB, Canada Objectives: The aim of this study was to evaluate the analytical performance of a third party enzymatic total bile acid assay (Diazyme Laboratories) for use on the Roche Integra 800 analyzer. Design and methods: The Diazyme total bile acid assay was evaluated using the Clinical and Laboratory Standards Institute evaluation protocols. The method comparison study included the Bio Quant total bile acid kit (supplied by Diazyme Laboratories) on Integra 400 (n = 48), and the LC–MS/MS assay (n = 30). Results: The Diazyme total bile acid assay demonstrated an excellent precision in that with-in run imprecision and total imprecision were 1.0% and 2.5% at 10.8 μmol/L level; 0.5% and 1.0% at 30.7 μmol/L level; and 0.5% and 1.0% at 111 μmol/L level. There is
1014
Abstracts
no obvious carry-over effect. This assay correlated well with the Bio Quant total bile acid kit on Integra 400 with a slope of 0.9429, intercept of 1.46, R = 0.9853, and average bias of 0.4 μmol/L (2.4%). In addition, this assay correlated well with the LC-MS/MS assay with a slope of 0.8589, intercept of 1.94, R = 0.9952, and average bias of −2.8 μmol/L (22.1%). Conclusions: The Diazyme total bile acid assay is acceptable for use on the Roche Integra 800 analyzer for clinical laboratories.
doi:10.1016/j.clinbiochem.2015.07.057
315 Laboratory performance on reporting monoclonal gammopathy during cerebrospinal fluid oligoclonal banding analysis Yu Chen Horizon Health Network, Fredericton NB Canada Objective: Cerebrospinal fluid (CSF) oligoclonal banding analysis is a sensitive routine test to mainly aid multiple sclerosis diagnosis. Monoclonal gammopathy is usually an incidental finding during the CSF oligoclonal banding analysis. The aim of this study was to assess the laboratory performance on reporting monoclonal gammopathy pattern during CSF oligoclonal banding analysis. Methods: The CSF oligoclongal banding surveys from the College of American Pathologists (CAP) of the last 5 years were reviewed. The UK National External Quality Assessment Service (NEQAS) for CSF oligoclonal banding 2014 survey was also reviewed. All monoclonal gammopathy patterns were confirmed by serum protein electrophoresis followed by immunofixation on Sebia Hydrasys analyzer. Results: There were 9 monoclonal gammopathy cases identified in the CAP Oligoclonal banding survey in 2010–2014 (Table 1). The average rate of North American laboratories that correctly reported the pattern was 24.1% (range 2.4–66.7%). The most common pattern incorrectly reported as was the systemic inflammation pattern, followed by the oligoclonal bands present/positive pattern. The NEQAS oligoclonal banding survey (2014–145) indicated a higher number (89.1%) of European laboratories to detect the monoclonal gammopathy. Conclusion: The monoclonal gammopathy is still an under recognized pattern in the CSF oligoclonal banding analysis and warrants more endeavors on education. Keywords: monoclonal gammopathy; multiple myeloma; CSF oligoclonal banding; missed diagnosis; laboratory error
Table 1: Laboratory performance on interpretation of monoclonal gammopathy competency test samples (correct reporting in bold). CAP, College of American Pathologists; NEQAS, UK National External Quality Assessment Service; n, number of the clinical laboratories.
doi:10.1016/j.clinbiochem.2015.07.058
316 Validation of a LC–MS/MS method for urinary free cortisol Lufang Yang, Gordon Ball, Gordon Hoag Vancouver Island Health Authority, Victoria BC Canada Objectives: To develop a selective and rapid routine LC-MS/MS method for urinary free cortisol analysis. Design and methods: A 250 μl urine sample with internal standard (Cortisol-9,11,12,12-d4) was extracted using solid-phase extraction tubes (Strata-X 33 μm, polymeric sorbent). A 10 μl extract was injected for analysis by a Shimadzu Prominence HPLC system and AB SCIEX 4000 QTRAP mass spectrometer with electrospray ionization in positive polarity. The multiple reaction-monitoring transitions used for the detection of cortisol and cortisol-d4 were 363.1 N 121.1 vs 97 and 367.1 N 121.1, respectively. The cortisol was separated using gradient elution at 30 °C with a 0.5 ml/min flow rate. The method validation included analysis for precision, linearity, sensitivity, ion suppression, interference from structurally similar steroids and a reference range verification. Forty-five patient samples were assayed to compare with an established LC–MS/MS method. Results: Total run time was 2.6 min. Total imprecision in two levels of QC and one level of patient pool was less than 7%. The calibration curves exhibited linearity and reproducibility in the range 2.5–5000 nmol/L. The limit of quantification was 2.5 nmol/L. The regression equation for our method (y) and established LC–MS/MS method (x) method in 45 patient samples was: Y = 1.0879X − 7.9782 (R2 = 0.9947). No interference from structurally similar steroids was noted. Ion suppression effects were noticeable for both cortisol and cortisol −d4. The reference value was verified at 10– 166 nmol/d in 20 normal subjects. Conclusion: This is a rapid and selective LC–MS/MS method suitable for routine quantification of urinary free cortisol. doi:10.1016/j.clinbiochem.2015.07.059
317 Quantification of carbohydrate-deficient transferrin by HPLC Pierre-Olivier Hétu, Anaïck Lagana-Teyssier Pharmacology–Toxicology, Clinical Biochemistry Dept., CHUM, Montreal, QC, Canada Objectives: Carbohydrate-deficient transferrin (CDT), and more specifically the glycoform containing two sialic acid residues (disisialotransferrin or DST), is a marker of chronic alcohol abuse. We have validated and implanted an HPLC based method for the quantification of DST in serum. Design and methods: Serum transferrin was first saturated with iron ions and lipoproteins were precipitated with a CaCl2/dextran solution. The different transferrin glycoforms were then separated by HPLC using anion exchange chromatography. Results (%DST) were calculated as a percentage of the total transferrin. Results: The HPLC method showed good and reproducible separation of the major transferrin glycoforms. The total area under the curve in the chromatograms correlated well with the transferrin concentration in the samples (R2 = 0.91). The %DST measures were precise (inter-assay imprecision = 7.1%, n = 28) and accurate when compared to target values in the external quality assurance program (97.5% accuracy, n = 2). The reference values for the method (b2.0%) were confirmed by analyzing 65 serum samples from patients with normal GGT levels. Conclusion: An HPLC method for %DST quantification was validated and implanted at CHUM and is currently used for identification of chronic alcohol abuse. doi:10.1016/j.clinbiochem.2015.07.060
Clinical Biochemistry, The Hospital for Sick Children, Toronto, ON, Canada
Objectives: In the diagnosis of central nervous system germ cell tumours (CNS GCTs), alpha-fetoprotein (AFP) in cerebrospinal fluid (CSF) above 10 ug/L distinguish non-germinomatous germ cell tumours from germinomas while human chorionic gondatropin (hCG) above 5 IU/L confirm a germinoma. During follow-up, AFP or hCG above these concentrations suggest recurrence. To determine suitability of the Abbott Architect AFP and total beta-hCG assays for these purposes, we validated their analytical performance in CSF. Design and methods: Evaluation was performed with CSF pools spiked with AFP or beta-hCG from serum samples. For imprecision, pools were measured in duplicate over 10 days. To determine values in absence of disease, measurements were made in 30 individuals free from pregnancy, trophoblastic disease or malignancy. Results: Recovery ranged from 81–90% for beta-hCG and 87–101% for AFP in CSF compared to serum. Functional sensitivity was 1.59 IU/L for beta-hCG and 0.26 ug/L for AFP. Linearity was confirmed from 2.8 to 27.8 IU/L for beta-hCG and 2.4 to 45.2 ug/L for AFP. For imprecision, CV values were from 4.1–7.7% for beta-hCG and 1.8–2.6% for AFP. Measurement of samples from individuals free from pregnancy, trophoblastic disease or malignancy were below the measuring range of 1.2 IU/L for beta-hCG and ranged from 0–0.05 ug/L for AFP, confirming their absence in normal CSF. Conclusions: The Abbott Architect assays for beta-hCG and AFP can be used in CSF without significant matrix effects and with good precision, linearity and functional sensitivity and are, thus, suitable for use in the management of CNS GCTs. doi:10.1016/j.clinbiochem.2015.07.054
312 Development of a rapid LC–MS/MS method for hair cortisol determination to assess the HPA axis Laura Smya, Michael Leadleyb, Katarina Aleksab, Gideon Korenb a University of Toronto and The Hospital for Sick Children, Toronto, ON, Canada b The Hospital for Sick Children, Toronto, ON, Canada Objectives: Hair cortisol is an increasingly utilized biomarker to assess the hypothalamic-pituitary-adrenal (HPA) axis. Often, hair cortisol is measured using enzyme-linked immunosorbent assays (ELISA). Other LC–MS methods detect multiple steroid compounds leading to long run times. To improve detection over ELISA, and LC– MS turn-around-times and costs, we are developing a rapid UHPLC– MS/MS method to measure hair cortisol in 2 min. Design and methods: Cortisol was extracted with methanol from hair samples collected from the head vertex posterior region. The supernatants were treated with C18 solid phase extraction (SPE) or filtered, dried down, reconstituted with methanol:water, and analyzed using UHPLC with a C18 cartridge and a mobile phase gradient. The mass spectrometry cortisol transition was 407.2 to 335.1 m/z and the cortisol-D4 transition was 411.2 to 339.1 m/z. Potential interference from the presence of cortisone (405.1 to 329.1 m/z) and prednisone (403.1 to 327.1 m/z.) were assessed. Results: With this method, hair cortisol results are obtained in a 2.0-minute run time. The retention time for cortisol and cortisol-D4 was 0.82 ± 0.01 min. The calibration curve is linear with a correlation coefficient ≥ 0.998. The LOD is 0.33 pmol/g for samples pretreated with SPE, or 14.9 pmol/g for filtered samples. The intra-day coefficient of variation was 2.5%. Further, cortisone and prednisone do not contribute to the cortisol peak with this method.
1013
Conclusions: Hair cortisol levels can successfully be determined in a short run time of 2 min without interference from the cortisone, or any potential prednisone, present in the hair. Keywords: Hair cortisol, rapid LC–MS/MS, hypothalamus–pituitary– adrenal doi:10.1016/j.clinbiochem.2015.07.055
313 Performance evaluation of the epoc® point-of-care blood analysis system Yu Chena, Monique Gormanb, Bill O'Reillyb a Horizon Health Network, Fredericton, NB, Canada b Saint John Regional Hospital, Saint John NB, Canada Objectives: The aim of this study was to evaluate the analytical performance of the new Alere epoc® point-of-care blood analysis system. Design and methods: The precision study was conducted on 3 epoc® blood analysis systems using 5 levels of quality control materials twice per day for 5 days. A blood gas syringe, a gold top tube, a grey top tube, and a lavender top tube were collected for the comparison study on cardiac perfusion patients (n = 40). The blood gas syringe samples were tested on epoc® (all 3 m), Instrumentation Laboratory GEM4000, Abott iSTAT, and Nova CCX meters. Gold top tube and grey top tube samples were tested on Roche Modular P for electrolytes, glucose and lactate. Lavender samples were tested on Beckman Coulter LH-780 for hemoglobin. Results: The epoc® blood analysis systems demonstrated clinically acceptable precision for all analytes (from 0.07%, 0.07%, and 0.13% for pH 7.6, 7.4, and 7.0 levels; to 3.87%, 3.74%, and 7.56% for pO2 197, 103, and 56 mm Hg levels). Comparison studies yielded a correlation coefficient R from 0.9201 (sodium) to 0.9969 (pO2) with the GEM4000; from 0.9071 (sodium) to 0.9965 (potassium) with the iSTAT; from 0.8793 (sodium) to 0.9957 (pO2) with the CCX; from 0.8463 (sodium) to 0.9942 (potassium) with Modular; and 0.9557 (hemoglobin) with LH-780. Average biases for all analytes were within the total allowable error limits. Conclusions: The Alere epoc® blood analysis system is acceptable for point-of-care testing in the hospital setting. doi:10.1016/j.clinbiochem.2015.07.056
314 Performance evaluation of a third party enzymatic total bile acid assay for use on the Roche Integra 800 analyzer Yu Chen, Hilary Smith, Heather Tarr Department of Laboratory Medicine, Dr. Everett Chalmers Regional Hospital, Horizon Health Network, Fredericton, NB, Canada Objectives: The aim of this study was to evaluate the analytical performance of a third party enzymatic total bile acid assay (Diazyme Laboratories) for use on the Roche Integra 800 analyzer. Design and methods: The Diazyme total bile acid assay was evaluated using the Clinical and Laboratory Standards Institute evaluation protocols. The method comparison study included the Bio Quant total bile acid kit (supplied by Diazyme Laboratories) on Integra 400 (n = 48), and the LC–MS/MS assay (n = 30). Results: The Diazyme total bile acid assay demonstrated an excellent precision in that with-in run imprecision and total imprecision were 1.0% and 2.5% at 10.8 μmol/L level; 0.5% and 1.0% at 30.7 μmol/L level; and 0.5% and 1.0% at 111 μmol/L level. There is
1014
Abstracts
no obvious carry-over effect. This assay correlated well with the Bio Quant total bile acid kit on Integra 400 with a slope of 0.9429, intercept of 1.46, R = 0.9853, and average bias of 0.4 μmol/L (2.4%). In addition, this assay correlated well with the LC-MS/MS assay with a slope of 0.8589, intercept of 1.94, R = 0.9952, and average bias of −2.8 μmol/L (22.1%). Conclusions: The Diazyme total bile acid assay is acceptable for use on the Roche Integra 800 analyzer for clinical laboratories.
doi:10.1016/j.clinbiochem.2015.07.057
315 Laboratory performance on reporting monoclonal gammopathy during cerebrospinal fluid oligoclonal banding analysis Yu Chen Horizon Health Network, Fredericton NB Canada Objective: Cerebrospinal fluid (CSF) oligoclonal banding analysis is a sensitive routine test to mainly aid multiple sclerosis diagnosis. Monoclonal gammopathy is usually an incidental finding during the CSF oligoclonal banding analysis. The aim of this study was to assess the laboratory performance on reporting monoclonal gammopathy pattern during CSF oligoclonal banding analysis. Methods: The CSF oligoclongal banding surveys from the College of American Pathologists (CAP) of the last 5 years were reviewed. The UK National External Quality Assessment Service (NEQAS) for CSF oligoclonal banding 2014 survey was also reviewed. All monoclonal gammopathy patterns were confirmed by serum protein electrophoresis followed by immunofixation on Sebia Hydrasys analyzer. Results: There were 9 monoclonal gammopathy cases identified in the CAP Oligoclonal banding survey in 2010–2014 (Table 1). The average rate of North American laboratories that correctly reported the pattern was 24.1% (range 2.4–66.7%). The most common pattern incorrectly reported as was the systemic inflammation pattern, followed by the oligoclonal bands present/positive pattern. The NEQAS oligoclonal banding survey (2014–145) indicated a higher number (89.1%) of European laboratories to detect the monoclonal gammopathy. Conclusion: The monoclonal gammopathy is still an under recognized pattern in the CSF oligoclonal banding analysis and warrants more endeavors on education. Keywords: monoclonal gammopathy; multiple myeloma; CSF oligoclonal banding; missed diagnosis; laboratory error
Table 1: Laboratory performance on interpretation of monoclonal gammopathy competency test samples (correct reporting in bold). CAP, College of American Pathologists; NEQAS, UK National External Quality Assessment Service; n, number of the clinical laboratories.
doi:10.1016/j.clinbiochem.2015.07.058
316 Validation of a LC–MS/MS method for urinary free cortisol Lufang Yang, Gordon Ball, Gordon Hoag Vancouver Island Health Authority, Victoria BC Canada Objectives: To develop a selective and rapid routine LC-MS/MS method for urinary free cortisol analysis. Design and methods: A 250 μl urine sample with internal standard (Cortisol-9,11,12,12-d4) was extracted using solid-phase extraction tubes (Strata-X 33 μm, polymeric sorbent). A 10 μl extract was injected for analysis by a Shimadzu Prominence HPLC system and AB SCIEX 4000 QTRAP mass spectrometer with electrospray ionization in positive polarity. The multiple reaction-monitoring transitions used for the detection of cortisol and cortisol-d4 were 363.1 N 121.1 vs 97 and 367.1 N 121.1, respectively. The cortisol was separated using gradient elution at 30 °C with a 0.5 ml/min flow rate. The method validation included analysis for precision, linearity, sensitivity, ion suppression, interference from structurally similar steroids and a reference range verification. Forty-five patient samples were assayed to compare with an established LC–MS/MS method. Results: Total run time was 2.6 min. Total imprecision in two levels of QC and one level of patient pool was less than 7%. The calibration curves exhibited linearity and reproducibility in the range 2.5–5000 nmol/L. The limit of quantification was 2.5 nmol/L. The regression equation for our method (y) and established LC–MS/MS method (x) method in 45 patient samples was: Y = 1.0879X − 7.9782 (R2 = 0.9947). No interference from structurally similar steroids was noted. Ion suppression effects were noticeable for both cortisol and cortisol −d4. The reference value was verified at 10– 166 nmol/d in 20 normal subjects. Conclusion: This is a rapid and selective LC–MS/MS method suitable for routine quantification of urinary free cortisol. doi:10.1016/j.clinbiochem.2015.07.059
317 Quantification of carbohydrate-deficient transferrin by HPLC Pierre-Olivier Hétu, Anaïck Lagana-Teyssier Pharmacology–Toxicology, Clinical Biochemistry Dept., CHUM, Montreal, QC, Canada Objectives: Carbohydrate-deficient transferrin (CDT), and more specifically the glycoform containing two sialic acid residues (disisialotransferrin or DST), is a marker of chronic alcohol abuse. We have validated and implanted an HPLC based method for the quantification of DST in serum. Design and methods: Serum transferrin was first saturated with iron ions and lipoproteins were precipitated with a CaCl2/dextran solution. The different transferrin glycoforms were then separated by HPLC using anion exchange chromatography. Results (%DST) were calculated as a percentage of the total transferrin. Results: The HPLC method showed good and reproducible separation of the major transferrin glycoforms. The total area under the curve in the chromatograms correlated well with the transferrin concentration in the samples (R2 = 0.91). The %DST measures were precise (inter-assay imprecision = 7.1%, n = 28) and accurate when compared to target values in the external quality assurance program (97.5% accuracy, n = 2). The reference values for the method (b2.0%) were confirmed by analyzing 65 serum samples from patients with normal GGT levels. Conclusion: An HPLC method for %DST quantification was validated and implanted at CHUM and is currently used for identification of chronic alcohol abuse. doi:10.1016/j.clinbiochem.2015.07.060