- Email: [email protected]
Evaluation of the Abbott PCx® point of care glucose analyzer
POSTER ABSTRACTS
Reportedly, about a quarter of all cases of hyperprolactinemia is associated with or due to macroprolactinemia (MW ⬎ 100 kD). Objectives: To evaluate a simple assay for macroprolactin as described in the literature (Lancet 1999;353:720), and to compare the cross-reactivity of macroprolactin in commonly used automated prolactin (PRL) assays. Methods: Macroprolactin in serum was precipitated in a buffer containing 25 mM phosphate pH 7.4 and 10% polyethylen glycol (PEG) 6000, redissolved, and assayed on the Bayer Immuno 1 for PRL. Some sera were sizefractionated by FPLC on a Pharmacia Superose 12 column, and fractions assayed for PRL. Sera with and without macroprolactin were then assayed on the Abbott AxSYM, Beckman Access, and Roche Elecsys. Results: The PEG precipitation assay is simple and reproducible (CVs ⬍ 15%), and we established a preliminary normal range of ⬍ 20% precipitation of total PRL by PEG. The assay correlates well with the amount of macroprolactin seen by FPLC as a peak with a MW of 180 –200 kD. Sera with confirmed macroprolactin showed the following cross-reactivities in commonly used PRL assays: Roche Elecsys ⬎ Bayer Immuno 1 ⬎ Abbott AxSYM ⬎ Bayer Centaur ⬃ Beckman Access. Thus, some of those sera appeared highly abnormal by one assay but normal by another assay. While macroprolactin was previously thought to be innocuous, this might not always be the case (Al-Shammari et al., Abstract, Endocrine Society Meeting, Toronto 2000). Conclusion: Macroprolactin can be easily quantitated by the Immuno 1 PRL assay after PEG precipitation. It cross-reacts to different degrees with common prolactin assays, and is a major source of variability between them. 42 CLONING OF A HUMAN GENE, ENCODING A MEMBER OF THE SER/ARG-RICH FAMILY OF PROTEINS THAT INTERACT WITH RNA POL-II Scorilas, A., Kyriakopoulou, L., and Diamandis, E.P. Department of Pathology and Laboratory Medicine, Mount Sinai Hospital, and Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada Objectives: Identification and characterisation of the human homologue of the rat A1 gene which may be involved in the pathogenesis of various human cancers. Methods: Positional cloning gene analysis was employed to characterize in detail a large region of chromososme 19q13.3 that has been shown to contain known oncogenes. Screening of expressed sequence tags (ESTs) was used to delineate the genomic organization of a novel gene and determine its splicing sites. Results: The novel gene spans 16.7 Kb of genomic sequence it is formed of 11 coding exons and 10 intervening introns and it is transcribed from telomere to centromere. The predicted amino acid sequence showed an extensive identity (78%) to the rat A1 protein which is a member of the Ser/Arg-rich protein family. Members of this family in CLINICAL BIOCHEMISTRY, VOLUME 33, APRIL 2000
rat, have been shown to interact with the C-terminal domain of the large subunit of the RNA polymerase II. Thus we speculate that the newly cloned gene is the human homologue of the rat A1 gene. RT-PCR analysis has shown that hA1 gene transcripts are expressed in prostate, thymus, lung, testis, uterus, colon and thyroid but not ovaries. Preliminary data suggest that this gene is overexpressed in a subset of ovarian tumors. Furthermore, studies with the steroid hormone receptor-positive breast carcinoma cell line BT-474 show that this gene is upregulated by estrogens and progestins and to a lesser extend by androgens. Conclusions: The new gene is the human homolog of the rat A1 gene. It is up-regulated in ovarian cancer and it is also regulated by steroid hormones.
43 EVALUATION OF THE ABBOTT PCX威 POINT OF CARE GLUCOSE ANALYZER Soldin, S.J1,2 Devairakkam, P.D.,1 and Agarwalla, P.K.1 Department of Laboratory Medicine1, Children’s National Medical Center, Washington, DC, USA, Departments of Pediatrics and Pathology2, The George Washington University School of Medicine, Washington, DC, USA Objectives: To assess (1) the effect of hematocrit on glucose measurement by the PCx威 analyzer and (2) the accuracy and precision for glucose measurement on this analyzer. Methods: 20 mL blood samples were obtained from three healthy volunteers. Each sample was separated into 3 aliquots a, b and c. These were centrifuged and plasma from a was added to c to artificially change the hematocrit values. Glucose measurements were immediately performed on both the PCx威 (after remixing) and Vitros 威analyzers. Within and between day precision studies were performed using PCx liquid controls at low, medium and high glucose Concentrations. Accuracy was evaluated using both known control values and by comparing PCx patient results with those obtained on the Vitros analyzer. Results: In the hematocrit range of 32–51% the effect of hematocrit on PCx glucose results was minimal (differences ⬍8%). For hematocrits ⬎60% the glucose results were substantially affected (decreased by 15–20%). The between day precision (%CV) for glucose concentrations in the range of 2.6 –14.5 mmol/L was 5.8 – 4.0 % on the PCx analyzer while the within day precision ranged from 3.8 –1.8%. A comparison of 40 patient samples run on both the PCx and Vitros analyzers gave a correlation coefficient of 0.92 and the equation PCx ⫽ 1.07 Vitros ⫺0.47. Conclusions: For neonates the 2.5–97.5th percentiles for hematocrit are 38 –57%. Hematocrits within these ranges cause only small discrepancies in the glucose results measured on the PCx. Patient results for glucose within the range 2.2–25.0 mmol/L on the PCx and Vitros analyzers compared well. 239
Reportedly, about a quarter of all cases of hyperprolactinemia is associated with or due to macroprolactinemia (MW ⬎ 100 kD). Objectives: To evaluate a simple assay for macroprolactin as described in the literature (Lancet 1999;353:720), and to compare the cross-reactivity of macroprolactin in commonly used automated prolactin (PRL) assays. Methods: Macroprolactin in serum was precipitated in a buffer containing 25 mM phosphate pH 7.4 and 10% polyethylen glycol (PEG) 6000, redissolved, and assayed on the Bayer Immuno 1 for PRL. Some sera were sizefractionated by FPLC on a Pharmacia Superose 12 column, and fractions assayed for PRL. Sera with and without macroprolactin were then assayed on the Abbott AxSYM, Beckman Access, and Roche Elecsys. Results: The PEG precipitation assay is simple and reproducible (CVs ⬍ 15%), and we established a preliminary normal range of ⬍ 20% precipitation of total PRL by PEG. The assay correlates well with the amount of macroprolactin seen by FPLC as a peak with a MW of 180 –200 kD. Sera with confirmed macroprolactin showed the following cross-reactivities in commonly used PRL assays: Roche Elecsys ⬎ Bayer Immuno 1 ⬎ Abbott AxSYM ⬎ Bayer Centaur ⬃ Beckman Access. Thus, some of those sera appeared highly abnormal by one assay but normal by another assay. While macroprolactin was previously thought to be innocuous, this might not always be the case (Al-Shammari et al., Abstract, Endocrine Society Meeting, Toronto 2000). Conclusion: Macroprolactin can be easily quantitated by the Immuno 1 PRL assay after PEG precipitation. It cross-reacts to different degrees with common prolactin assays, and is a major source of variability between them. 42 CLONING OF A HUMAN GENE, ENCODING A MEMBER OF THE SER/ARG-RICH FAMILY OF PROTEINS THAT INTERACT WITH RNA POL-II Scorilas, A., Kyriakopoulou, L., and Diamandis, E.P. Department of Pathology and Laboratory Medicine, Mount Sinai Hospital, and Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, Ontario, Canada Objectives: Identification and characterisation of the human homologue of the rat A1 gene which may be involved in the pathogenesis of various human cancers. Methods: Positional cloning gene analysis was employed to characterize in detail a large region of chromososme 19q13.3 that has been shown to contain known oncogenes. Screening of expressed sequence tags (ESTs) was used to delineate the genomic organization of a novel gene and determine its splicing sites. Results: The novel gene spans 16.7 Kb of genomic sequence it is formed of 11 coding exons and 10 intervening introns and it is transcribed from telomere to centromere. The predicted amino acid sequence showed an extensive identity (78%) to the rat A1 protein which is a member of the Ser/Arg-rich protein family. Members of this family in CLINICAL BIOCHEMISTRY, VOLUME 33, APRIL 2000
rat, have been shown to interact with the C-terminal domain of the large subunit of the RNA polymerase II. Thus we speculate that the newly cloned gene is the human homologue of the rat A1 gene. RT-PCR analysis has shown that hA1 gene transcripts are expressed in prostate, thymus, lung, testis, uterus, colon and thyroid but not ovaries. Preliminary data suggest that this gene is overexpressed in a subset of ovarian tumors. Furthermore, studies with the steroid hormone receptor-positive breast carcinoma cell line BT-474 show that this gene is upregulated by estrogens and progestins and to a lesser extend by androgens. Conclusions: The new gene is the human homolog of the rat A1 gene. It is up-regulated in ovarian cancer and it is also regulated by steroid hormones.
43 EVALUATION OF THE ABBOTT PCX威 POINT OF CARE GLUCOSE ANALYZER Soldin, S.J1,2 Devairakkam, P.D.,1 and Agarwalla, P.K.1 Department of Laboratory Medicine1, Children’s National Medical Center, Washington, DC, USA, Departments of Pediatrics and Pathology2, The George Washington University School of Medicine, Washington, DC, USA Objectives: To assess (1) the effect of hematocrit on glucose measurement by the PCx威 analyzer and (2) the accuracy and precision for glucose measurement on this analyzer. Methods: 20 mL blood samples were obtained from three healthy volunteers. Each sample was separated into 3 aliquots a, b and c. These were centrifuged and plasma from a was added to c to artificially change the hematocrit values. Glucose measurements were immediately performed on both the PCx威 (after remixing) and Vitros 威analyzers. Within and between day precision studies were performed using PCx liquid controls at low, medium and high glucose Concentrations. Accuracy was evaluated using both known control values and by comparing PCx patient results with those obtained on the Vitros analyzer. Results: In the hematocrit range of 32–51% the effect of hematocrit on PCx glucose results was minimal (differences ⬍8%). For hematocrits ⬎60% the glucose results were substantially affected (decreased by 15–20%). The between day precision (%CV) for glucose concentrations in the range of 2.6 –14.5 mmol/L was 5.8 – 4.0 % on the PCx analyzer while the within day precision ranged from 3.8 –1.8%. A comparison of 40 patient samples run on both the PCx and Vitros analyzers gave a correlation coefficient of 0.92 and the equation PCx ⫽ 1.07 Vitros ⫺0.47. Conclusions: For neonates the 2.5–97.5th percentiles for hematocrit are 38 –57%. Hematocrits within these ranges cause only small discrepancies in the glucose results measured on the PCx. Patient results for glucose within the range 2.2–25.0 mmol/L on the PCx and Vitros analyzers compared well. 239