Search tips
Search criteria 


Logo of clinbiorevLink to Publisher's site
Clin Biochem Rev. 2003 August; 24(3): S1–S35.
PMCID: PMC1853342

Proceedings of the Australasian Association of Clinical Biochemists’ 41st Annual Scientific Conference

Invited SpeakersPage
S1Automated Assays for Glycohaemaglobin/HbA1c
I Goodall
S2The Use of NMR Spectroscopy in Metabolic Disease
F Bowling
S3Diabetes: Tracking, Treating and Preventing the Epidemic
P Zimmet
S4Continuous Glucose Monitoring In Diabetic Patients
T W Jones
S5Current Awareness in Alzheimer’s Disease Research, with an Emphasis on Glutamate Exitoxicity
H Scott and P. Dodd
S6Infertility Diagnosis and Treatment
KL Harrison
S7Troponin: On-going Issues
J Tate and G. Aroney
S8Evidence-Based Laboratory Medicine
AR Horvath
S9Evidence for Self-Measurement of Glucose
S. Sandberg, AR Horvath
S10The Technology of Glucose Self-Monitoring
A St John
S11Diabetic Dyslipidaemia
M d'Emden
S12Diseases in Pregnancy - Gestational Diabetes
RH Mortimer
C Boothroyd
S14Key Performance Indicators Derived from Quality Assurance Data - A Valuable Tool for all Laboratories?
L Penberthy, P Stewart, K Sikaris, J Gill, D Graves, M Meerkin, D Chesher, G Jones, R Bais, I Farrance.
S15The Fading Male
T Gaffney
B McCall
S17Molecular Mechanisms in Cancer - An Overview
HA Morris
S18Future Directions in Diabetes
J Prins
S19First Trimester Screening for Down Syndrome
D Kanowski
S20The Genetic Basis of Early Familial Colorectal Cancer
R Scott
S21Viral Hepatitis
D Siebert
Oral PresentationsPage
O1Macroprolactin and the Abbott Architect
MP Metz, T Petrou
O2Clinical Uses of Serum Free Light Chain Measurements
J Tate, P. Mollee, D Gill, R Cobcroft, PE Hickman
O3Measurement of Free Light Chains in Serum - Analytical and Clinical
J Tate, D. Gill, R. Cobcroft, P.E. Hickman
O4Quality Management - The Six Sigma Approach
G Koerbin
O5Effect of Fenofibrate on Plasma Lipids and Apolipo-Proteins
MJ Whiting, M Pejakovic, JA Leach, A Keech
O6Genetic Testing for Gilbert Syndrome by DHPLC
C Mamotte, L Langan, S Vasikaran
O7Quantitation of LDL-Receptor and LRP mRNA Expression in Human Cells by Real-Time PCR
A Pocathikorn, RR Taylor, C Mamotte
O8A Simple Protocol for Identification of Heterophilic Antibody Interferences in the Abbott AXSYM hCG Assay
P Giannopoulos, G Jones, EM Lim
O9Evaluation of Samples Collected in Becton Dickinson PST II and SST II Tubes
P Graham, B Martin, M Roser, G Jones
O10Use of BVBLUE® To Confirm Sialidase Contamination of Samples for B2 Transferrin Determination
LM Boscato
O11Vitamin D Deficiency and Abnormal PTH-Calcium in Lipid Clinic Patients
T Yen, DR Sullivan
Poster Presentations
P1Comparsion of Bilirubin Determination With the Beckman Synchron CX5-ce and the Radiometer ABL-735
MP Metz, L Nesci, F Ku
P2Comparison of Results from Sweat Tests with Thirty Minutes to Two Hours Collection Times
E Whitham
P3Reference Intervals for TFTS in the First Trimester of Pregnancy
A Read and M Arumanayagam
P4Nutritional Assessment of People with Diabetic Foot Ulcers
J.A. Burgess, S. Daniells and F San Gil
P5Method Performance of HbA1C Measurement Using the Roche Integra 400 Plus
W. Ferguson, J. Tate, P. Hickman.
P6Some Anomalous Analytical Findings Using a Commercial Soluble Transferrin Receptor Assay
K. Lee, B Barr, J. Chang, S Bazeley, S Klingberg, J Tate
P7Potential For False Positive Tau (b2)- Transferrin Using Salivary Samples
K Sanders, J Tate, J Chang, PE Hickman.
P8The Lack of Harmonisation of Methods for Total Urine Protein (TUP) Measurement
J Tate, W Ferguson, L Callaghan, S Klingberg, K. Marshall, G. Koerbin
P9Analytical Issues with Low Level Cardiac Troponin T Measurement
B Russell, B Jones, J Tate, G. Koerbin, JM Potter, PE Hickman
P10Evaluation of the Dade Behring HbA1c Assay for the Dimension RxL
V Miranda, G Koerbin, W Ferguson
P11Evaluation of the Dade Behring Emit 2000 Mycophenolic Acid Assay on the Dimension RxL and its Comparison to a Simple HPLC Method
V Miranda, L Chan, G Koerbin
P12Plasma Free Metanephrine analysis by HPLC with Colourmetric Detection
B McWhinney, P Hickman, S Edwards
P13A HPLC Method for the Analysis of Amprenavir, Saquinavir, Indinavir, Lopinavir, Ritonavir, Efavirenz, Nevirapine, Nelfinavir and its Actibe Mertabolite (M8)
S Edwards, B McWhinney
P14Simultaneous Determination of 6-Thioguanine, 6- Mercaptopurine and 6-Methyl Mercaptopurine in Whole Blood by HPLC
B McWhinney, P Hickman, C Davies
P15A Serum Angiotensin Converting Enzyme (ACE) Method for the Roche Modular P Analyser
G Dimeski, W Ferguson, J Tate
P16An Automated Competitive Immunoassay for Androstenedione on the Immulite® and Immulite® 2000
Z Wang, Z Bostanian, M McKenna, E Whitters, JD Lei, T Ito, P Hsia, N Panosian-Sahakian, A El Shami.
P17An Automated Chemiluminescent Immunoassay of Cross-Linked Fibrin Degradation Products (XL- FDP) Containing D-Dimer on the Immulite® 2000 Analyser
Z Bostanian, E Gudipati, J Wen, E Whitters, JD Lei, D Sustarsic, E Unver, A S El Shami
P18An Automated Immunoassay for Gastrin on the Immulite® 2000 Analyser
E Whitters, JD Lei, N Figueroa, R Laroya, N Panosian- Sahakian, AS El Shami.
P19Adaptive Local Regression Methods for Age-Related Reference Range Analysis Applied to Immulite® IGF-I and IGFBP-3 Data
M Elmlinger, W Kühnel, A Durham
P20Automated Chemiluminescent Immunoassays for Pregnancy-Associated Plasma Protein A ( PAPP-A) on the Immulite® and Immulite 2000 Analysers
Z Bostanian, G Hall, E Whitters, JD Lei, K Pregger, D Sustarsic, E Unver, A S El Shami
P21Development of a Third-Generation Allergen-Specific IgE Assay on the Immulite ® 2000
T Li, T Chuang, S Tse, D Hovanec-Burns, A El Shami
P22Salivary Cortisol Assay-Adaptation of the Routine Serum Assay.
M Goodwin, M Harrop and P Colman.
P23Plasma Renin Activity and the Advantage® Direct Renin Method - Method Comparison and Correlation Studies
LG Sampson, R. Rappel, G. Ward, D Jessop, P Hickman, L Price
P24Comparison between Plasma Renin Activity and the Advantage® Direct Renin Method - Precision Studies
LG Sampson, R Rappel, G. Ward, N Avsenev, P Hickman, L Price
P25Investigations into changes in Plasma Renin Activity Assay Conditions
LG Sampson, G Ward, R Rappel, B Sipinkoski, P Hickman, L Price
P26Plasma Renin Activity Assay and Precision
LG Sampson, G Ward, R Rappel, J McArthur, P Hickman, L Price
P27Evaluation of the Microgenics Cedia®Plus Cyclosporin Assay on the Olympus AU2700
RJ.Czajko, PA Shevenan, RX Davey
P28A Comparison of Serum FLC by the Freelite Immunoassay Kit with Urine FLC by Conentional IFE and Densitometry Techniques.
M Lovrincevic, C Appleton
P29Pseudohypernatraemia in Samples from Intensive Care Units
CG Clark, JP Galligan
P30Technical Evaluation of D-10, a New Small Laboratory Glycated Haemaglobin Testing System
CMF Yin, SK Ong, S Saw and S Sethi
P31Performance Characteristics for the Assay of mRNA Using Real Time Reverse Transcriptase PCR
HA Morris, PH Anderson and BK May
P32Free Light Chain Measurement in Amyloidosis
GI Raines, M Cole-Sinclair
P33Real-Time PCR Detection Of PiZ and PiS Alpha-1- Antitrypsin Mutations
K Namdarian, A Wootton and M Dobos
P34Comparison of the Clinical Suitability of the Roche Whole Blood HbA1c on the Cobas Integra 800 with the Bio-Rad Variant II
M Lovrincevic, M. Blakey
P35A Novel 15 Basepair Deletion in Keratin 5 in Epidemolysis Bullosa Simplex-Weber Cockayne (EBS- WC)
MW Kemp, S Klingberg, L Lloyd, P Marr, L Vonthethoff, Y Wang, GAC Murrell, DF Murrell
P36Combined Mutations in Keratin 5 and Keratin 14 Genes Causing Epidermolysis Bullosa Simplex
N Trisnowati, S Klingberg, L Lloyd, Y Wang, GAC Murrell, DF Murrell
P37Hereditary Hyperferritinaemia Cataract Syndrome (HHCS) An Example of Translational Control of Protein Synthesis.
S Ratnaike, P Thurlow, D Macgregor, V Gurtler, P Martinello and J McLeod
P38Glomerular Filtration Rate (GFR) In Diabetes
R MacIsaac, C Tsalamandris, S Panagiotopoulos, T Smith, D Lowe, S Ratnaike, M Jenkins, M Thomas, G Jerums
P39A Novel Approach to Pre-Analytical Whole Blood Lysis Using an Olympus AU2700
RJ Czajko, PA Shevenan, RX Davey
P40Intact PTH and ‘BIO INTACT PTH’ Assays on the Nichols Advantage and DPC Immulite Immunoassay System
R Rappel, L Sampson, G Ward, PE Hickman, L Price, D Jessop, J McArthur
P41Evaluation of Roche Stat and 18 Minute Troponin T Assays
V Wood and R David
P42Evaluation of Immulite 2000 TNI Assay in the Detection of False Positive AXSYM Results.
Q. Lam, G. Raines and HG Schneider
P43Tau Transferrin Detection Using Sebia Hydrasys System Using IF Hydragels
LLloyd, D Travers and S Klingberg.
P44Proficiency Testing Alone Has Little Effect on the Performance of POCT Glucose Meters
M Charan and R Bais
P45An Evaluation of ACTH and IGF-I Assays on the Nichols Advantage Immunoassay System
R Rappel, L Sampson, G Ward, PE Hickman, L Price, D Jessop, J McArthur, D Nicholas
P46Spironolactone and Canrenone - Assessment of Interference with the Advia-Centaur Digoxin Assay
R Rappel, J McArthur, L Sampson, G Ward, PE Hickman, J. Potter, L Price, R Mason, B McWhinney,
P47Molecular Diagnostic Assessment of Cytochrome P450 CYP2C9 Polymorphisms in Managing Warfarin Dosing
F Gray, A Dear, K Byron, L Paiman.
P48Comparison of β Trace Protein and β2 - Transferrin for the Detection of Cerebrospinal Fluid.
R Scarff, A Musk, P Sheehan, J Hill, K Smith
P49What is a Normal Lactose Intolerance Test?
KA Sikaris, AR McNeil, G Wilcox
P50Oestradiol Assays by RIA and The Immulite 2000 - What are we Measuring?
R Greaves, RW Hunt, J Frederiksen
P51The Power of Error Detection of Westgard Multi-Rules: A Re-Evaluation
GRD Jones
P52The Effect of Setting Quality Control Standard Deviations Greater than Actual Standard Deviations
GRD Jones
P53Hepatic Iron Loading in HFE Compound Heterozygote Patients
J Joseph, EM Lim, E Rossi, B De Boer, WD Reed , GP Jeffrey
P54Lead Levels in Pre-School Children
C Macaulay, J Joseph, E Rossi.
P55Hypnotraemia and Treatment with Intravenous Gamma Globulin
AR McNeil and P Panaya
P56Folate and Prostate Cancer: The Busselton Health Study
J Beilby, E Rossi, J Hung, M Knuiman, M Divitini
P57Effect of Paraprotein on Serum/Plasma Gel Seperator Tubes
H. Fong and M. Black
P58Interference by the S65C Mutation HFE Genotyping On the Lightcycler
C Chin, G Arscott, T Nguyen and J Beilby
P59Towards a Routine Genetic Screening Method for Porphyrias
N Al Hafid, V Poulos


I Goodall, Special Chemistry Unit, Austin and Repatriation Medical Centre, Heidelberg, VIC, 3084.

The ever increasing demand for glycated haemoglobin (GHb) assays for monitoring long term average glycaemia in both type I and type II diabetes has necessitated widespread automation of this assay. Only a very small percentage of assays are now performed by manual techniques.

The round table will cover all current automated assays for glycohaemoglobins. These include automated HPLC (cation exchange and affinity chromatography), LPLC cation exchange chromatography, immunological based assays and the newest development, immuno / enzymatic HbA1c from Arkray (Japan). Several Point of Care HbA1c assays are now available and will be included.

The discussion will encompass aspects such as specificity of assays, analyser throughput, turnaround time for clinic use, and interference from non GHb components which can significantly affect some assay results.

The Diabetes Control and Complication Trial has shown the need for accurate and precise GHb assays. Interassay and interlaboratory standardisation and precision requirements, and the significance of the HbA1c level in the evaluation of diabetic control will be discussed.


F Bowling Mater Children’s Hospital, Mater Hill QLD 4101.

The number of recognised inborn errors of metabolism is still increasing. The incidence of individual diseases is low but the overall prevalence of metabolic disease is considerable. For clinicians, metabolic diseases are often hard to recognise as the clinical phenotypes are rather variable. Any organ or specific cell type in the human body may be affected. Biochemically, any pathway of human metabolism may be involved. Over the years this has placed increasing demands on the number of specialised analytical techniques that are necessary to obtain a diagnosis.

NMR spectroscopy has now become an important tool in the diagnosis of inborn errors of metabolism. NMR spectroscopy can be applied in vitro to body fluids and tissue samples but also can be applied in vivo to specific organ systems, in particular the brain. NMR spectroscopy can be used to quantitate the presence of specific molecules and in this era of genomics and proteomics, NMR spectroscopy also provides a means of studying metabolomics.

The most applied form of NMR is proton spectroscopy which gives an overview of human metabolism by looking at proton-containing metabolites. However, spectrometric techniques also exist based on phosphorus, carbon, nitrogen, sodium and other elements.

In this paper I will briefly review the principles behind NMR spectroscopy and continue with specific examples of its application, both in vivo and in vitro in various disease states. A review of a selected number of key recent papers which report the successful application of a metabolomics approach to the diagnosis and to the monitoring of disorders will also be presented.


P Zimmet, International Diabetes Institute, Melbourne VIC

"Globalization" and "Coca-colonization" are having profound effects on health in many developing countries all over the world. They face a huge public health challenge from Type 2 diabetes and obesity (diabesity), particularly as they are now affecting younger people ("Nintendonization"). Our projections suggest that the current figure of 180 million people worldwide with Type 2 diabetes will climb to over 320 million by the year 2025. This represents an epidemic of major proportions and possibly the largest epidemic in human history.

The Australian Diabetes, Obesity and Lifestyle Study (AusDiab) was the first ever study to provide national estimates of the prevalence of glucose intolerance and its risk factors in Australia. The study showed the number of people with diabetes in Australia has increased by 300% over the past 20 years - from around 250,000 to 900,000. 1 in 4 Australians now have diabetes or are at high risk. 7.5% of Australian adults have diabetes, and a further 16.3 percent are at high risk with abnormal blood sugar levels not yet diagnostic of diabetes. AusDiab also showed that 60% of Australian adults are overweight or obese, an alarming 51% of adults have dyslipidaemia, and 29% have hypertension. These conditions form a cluster of cardiovascular disease (CVD) risk factors called the Metabolic Syndrome or "Deadly Quartet".

The main cause of mortality in Type 2 diabetes is CVD accounting for at least 70% of mortality. A key strategy in reducing the burden of CVD lies in the better understanding of the Metabolic Syndrome, a condition invariably associated with insulin resistance. This clustering of these CVD risk factors represents a "time bomb" for public health internationally as it provides the driving force for an escalating CVD epidemic. Treating Type 2 diabetes in this millennium will involve polypharmacy. The health costs to nearly every nation, including Australia, will be a huge burden. This highlights the need for urgent action in preventing the epidemic from accelerating.

Our studies using Psammomys obesus (the Israeli sand rat) suggest hyperinsulinaemia/insulin resistance is the initial metabolic lesion in the development of obesity and Type 2 diabetes. Psammomys, when placed on an ad libitum laboratory diet, develops hyperinsulinaemia, obesity and diabetes. It is likely that a similar sequence and process occurs in the transition from the prediabetic state to Type 2 diabetes in humans. We have discovered several novel genes including Beacon and Tanis which may have important aetiological roles in obesity and diabetes. These genes, their receptors and pathways become important targets for drug discovery for treatment of obesity, Type 2 diabetes and the Metabolic Syndrome.

The epidemiological studies on diabetes have also prepared the way for a better understanding of the potential for prevention with lifestyle interventions. Healthy nutrition along with exercise resulting in reduced energy intake and increased energy expenditure provide the first line of prevention. There is also a potential role for pharmacotherapy with drugs such as merformin, acarbose, the glitazoncs, orlistat and possibly ACEI and angiotensin 2 receptor antagonists.


TW Jones, Department of Endocrinology & Diabetes, Princess Margaret Hospital for Children, Perth WA 6000

Type 1 diabetes is the commonest, chronic, medical condition in children and adolescents and its incidence continues to increase. It is now definitively established that the rate of progression of the long term degenerative complications of diabetes can be slowed by improving glucose control. Unfortunately, the achievement of normal glucose values in patients with Type 1 diabetes has proved difficult because of the unphysiological nature of insulin replacement therapy. Both hypoglycaemia and hyperglycaemia are frequent in all insulin treated patients. Recently new insulins and new modes of insulin therapy have made improved glycaemic control a more realisable goal. An exciting additional development has been the development of novel glucose sensing technologies. These allow for continuous glucose sensing over a complete time period compared to the intermittent blood glucose values previously achieved with home glucose monitoring. This additional information allows more rational adjustments of insulin treatment. The ultimate goal of "closing the loop" between a continuous glucose sensor and an insulin infusion pump, although under development, has not yet been achieved.


H Scott, and P. Dodd, Biochemistry Department, University of Queensland, Brisbane, QLD 4072

Alzheimer's disease (AD) is the most common form of dementing illness, and is characterised by progressive clinical abnormalities, including symptoms such as memory loss, personality and behavioural changes, unsociability and a decrease in alertness. The pathological features of AD include senile plaques (β-amyloid deposition), neurofibrillary tangles (aggregation of hyperphosphorylated tau) and neuronal cell loss. The exact cause of Alzheimer's disease is unknown, although it seems that the biological events leading to the behavioural and clinical problems in AD centre around the destruction of certain nerve cells. Many theories have been proposed as to what initiates this process, and several lines of ongoing investigation include genetic factors (eg. apolipoprotein E), β-amyloid and senile plaques, tau protein and neurofibrillary tangles, nerve growth factors, neurotransmitter deficiencies and dysfunction, effects of inflammation and oxidative stress, and programmed cell death. Research in our laboratory primarily focuses on defects in glutamate neurotransmission in AD brain, which may result in an excitotoxic mechanism of neuronal cell death. We are examining the distribution and expression of both the glutamate receptors and transporters in cortical regions that are either susceptible to or spared from pathology. These and other studies may lead to a greater understanding of the underlying mechanism responsible for the pathology of AD and will significantly advance research into diagnostics, treatment, prevention and care. A few years ago, there were few treatment options available for AD patients. There are now a number of experimental drugs undergoing trial that may help in clearing amyloid deposits from the brain, enhancing nerve cell communication, and protecting and repairing nerve cells. However, there is still a great deal of research that needs to be done into early and accurate diagnosis of AD, which will give affected individuals a greater chance of benefiting from any treatment available.


KL Harrison, Queensland Fertility Group, 225 Wickham Terrace, Brisbane QLD 4000

Infertility will complicate the lives of around 20% of couples wishing to have children and half of these will require some level of assisted conception. Both natural and assisted conceptions then have a pregnancy loss rate ranging from 15% to 50% depending upon age.

Causes of infertility include:

  • ♦ Ovulation defects
  • ♦ Tubal blockages
  • ♦ Uterine problems
  • ♦ Oocyte quality
  • ♦ Sperm quality
  • ♦ Embryo quality

and are diagnosed by a combination of endocrinology testing, semen analysis, genetic testing, ultrasound and X-ray, and various endoscopic procedures.

The treatment of infertility commences with the simplest approach to the particular aetiology and progresses to the highest levels of the assisted reproductive technologies. Treatments include:

  • ♦ Ovulation induction
  • ♦ Controlled ovulation and timed intrauterine insemination
  • ♦ In vitro fertilisation and embryo transfer (IVF-ET)
  • ♦ Gamete intrafallopian transfer (GIFT)
  • ♦ Assisted fertilisation of oocytes by intracytoplasmic sperm injection (ICSI)
  • ♦ Pre-implantation embryo genetic diagnosis (PGD)
  • ♦ Assisted embryo hatching.

Success rates have risen to over 50% per treatment cycle and cumulative pregnancy rates mean that a vast majority of couples can now achieve their dream of a normal pregnancy.

Around 60% of products of conception following miscarriage will be found to have chromosomal abnormalities, this being significantly related to maternal age. Other causes of pregnancy loss include luteal phase endocrinological defects, hyperthyroidism and immunological causes.


J Tate1 and G Aroney2. 1Chemical Pathology QHPS, Princess Alexandra Hospital, Brisbane and 2Cardiology Department, Gold Coast Hospital, Southport QLD.

The introduction of cardiac troponin testing for the identification of myocardial ischaemia has changed the way clinicians manage patients with chest pain. The consensus document from the American College of Cardiology (ESC/ACC) in 2000 further redefined the criteria for diagnosis of myocardial infarction (MI) with the proposal that any increase in troponin concentration above the 99th percentile of a reference population, in the clinical setting of ischaemia, indicates myocardial necrosis. By contrast, the National Academy of Clinical Biochemistry proposed that the concentration of troponin corresponding to an analytical precision (CV) of 10% be used as the clinical diagnostic cutpoint. However, the clinical evidence base from TIMI and FRISC trials demonstrates that very low levels of troponin are associated with adverse cardiac outcomes. Precisely what constitutes a "very low level" remains controversial. Some studies have used a decision cutpoint corresponding to the troponin concentration at 20% CV and demonstrated a worse outcome compared with undetectable troponin concentrations. Not until troponin assays are optimised for improved analytical sensitivity, and the question of whether healthy persons have detectable troponin in their blood is answered, can the relevance of using the 99th percentile be determined.

Meanwhile, there are a number of on-going issues that may affect the laboratory measurement of troponin. These include pre-analytical factors such as time of sample collection and specimen type, and analytical factors such as the lack of troponin I method standardisation, differences in the analytical and functional sensitivity of troponin assays, non-ischaemic troponin elevations, and false positive results.

There are practical issues also facing the clinician as a result of the redefinition of MI. It is important that the laboratory understands the clinical impact of reporting troponin to lower levels and the effect of false positive results on patient treatment.


AR Horvath, Chair of the IFCC Committee on Evidence Based Laboratory Medicine; Department of Clinical Chemistry, University of Szeged, Hungary.

Patients and society expect physicians to base their approach to any type of clinical problem on informed diagnostic reasoning, for which we need high quality and reliable scientific evidence. Despite the crucial importance of the appropriate use of diagnostic tools in clinical decision-making, many diagnostic investigations have never been subjected to rigorous evaluation using modern standards of clinical epidemiology. The causes of this paradox in laboratory medicine are manifold: (a) the diagnostic accuracy of laboratory investigations is seriously over-estimated, due to study-design related biases (e.g. spectrum-, reviewer-, lead time-, volunteer- and verification bias); (b) lack of appropriate gold standards; (c) lack of randomised controlled trials, outcome studies and technology assessments; (d) publication bias; (e) lack of reliable data on disease prevalence and diagnostic thresholds; (f) difficulties to summarize the results of primary studies, due to heterogeneity of patient populations, the analytical methods of measurement, rapid changes in, or lack of standardisation of technologies, and changes in the definition of diseases over time (e.g. diabetes mellitus, myocardial infarction), etc. Summarizing data in systematic reviews or meta-analysis needs special skills and the methods of pooling data, without seriously distorting conclusions, are still under development in the field of diagnostic testing.

The lack of good quality evidence in laboratory medicine not only contributes to inappropriate utilization of laboratory services but also may cause harm to patients and wastes significant resources. Therefore it is the responsibility of the profession to set, implement, maintain and improve evidence-based methodological, diagnostic and clinical standards in laboratory medicine. To achieve evidence-based medicine, and thus higher efficiency in laboratory medicine, more effective collaboration between laboratory professionals, clinicians, patients, epidemiologists, biostatisticians, industry, quality-, technology assessment- and government agencies, and an international harmonization of these approaches are needed.


S Sandberg1, AR Horvath2, 1Chair of the IFCC Task Force of the Global Campaign of Diabetes Mellitus; Laboratory of Clinical Biochemistry, Haukeland University Hospital, Bergen, Norway;2 Chair of the IFCC Committee on Evidence Based Laboratory Medicine; Department of Clinical Chemistry, University of Szeged, Hungary.

The IFCC has decided to focus on the laboratory aspects of diabetes mellitus and has therefore announced a Global Campaign on Diabetes Mellitus and established a task force to address this theme. The Task Force will provide scientific evidence on how to diagnose and monitor diabetes mellitus by laboratory parameters and it will use this evidence to educate the profession. In more detail the Task Force will: (a) examine the best diagnostic strategies for diagnosing diabetes mellitus; (b) review and study the current use of laboratory tests, and educate doctors and patients on how to interpret the results; (c) provide quality specifications for analytes used in the diagnosis and monitoring of diabetes mellitus; (d) give recommendations on the minimum criteria for the use of glucometers; (e) co-operate with relevant groups or persons (including manufacturers and patients).

Patients with diabetes mellitus are encouraged to measure their own blood sugar several times a day. Taking into account that diabetes mellitus is a disease with increasing prevalence, this practice has large economic and practical implications. Two important questions can be formulated: (a) What are the driving forces for this practice, and (b) Can it be documented that self-measurement of blood glucose reduces the long term complications or improves the quality of life in patients with diabetes mellitus? For self-measurement to be beneficial for patients, three important prerequisites have to be fulfilled: (a) the instruments must have good enough quality for the intended purpose; (b) the instrument must produce reliable results in the hands of the patient; and (c) reasonable actions must be taken when the results are produced. The presentation will look into each of these aspects, and will examine whether self-measurement of glucose in diabetes mellitus is based upon evidence.


A St John, ARC Consulting, Perth, WA 6003

The glucose test strip has come a long way since the seminal paper of Clark and Lyons in the 1960’s which described the inclusion of an enzyme transducer into the oxygen electrode to measure glucose. The first commercial version of this biosensor appeared 1973, and 30 years later it is estimated that the glucose monitoring industry is worth US$ 3 billion.

Major developments have taken place in the design of both the glucose strip and the measurement device or meter with the aim of improving precision/accuracy as well as making the devices easier to use. The original strips were reflectance based and used glucose oxidase as the enzyme. Smaller meters and perhaps more crucially for performance, the development of non-wipe strips, was facilitated by the introduction of electrochemical sensors using ferrocene as a mediator in combination with glucose dehydrogenase which is not subject to oxygen interferences.

Other technological improvements include much faster reaction times with some strips producing results in 5 secs. Recently meters have been developed which sample from sites such as the arm and leg as an alternative to the often painful sampling from finger pricks. The analytical performance of meters has often been assessed by means of the Clark Error Grid which attempts to classify different areas of accuracy depending upon their clinical importance.

While the analytical performance of later generation meters has undoubtedly improved, various reports have documented that many meters do not meet the quality specifications set by the American Diabetes Association and other professional bodies. More interesting has been the recent report that the analytical performance of meters in patient hands is poorer than when used by laboratory personnel. Also identified in this report was the lack of professional input into decisions concerning choice of meter and related issues. Manufacturers will continue to strive for better analytical performance from meters but it is likely that patient education and support via software together with interactive health technologies will have equal if not greater benefit on patient outcomes.


M d'Emden, Department of Endocrinology, RBH, Brisbane, QLD 2029

It is accepted that Type 2 diabetes is a major risk factor for cardiovascular disease. Epidemiological studies have suggested that the rate of CVD may be 4 times greater in persons with diabetes. Women with diabetes lose their cardiovascular protection. Studies have suggested that diabetes may be as significant a risk factor as established CVD itself. However, persons with impaired glucose tolerance also have an increased risk of CVD and the risk of CVD doesn't correlate with duration of diabetes, leading some to believe that it is the state of insulin resistance (the metabolic syndrome, Reaven's syndrome, Syndrome X) associated with diabetes that is the major factor involved. The recent changes to the biochemical definition of diabetes and impaired fasting glucose has also created confusion as to whether these new conditions are accurately defining the at-risk population.

Studies have demonstrated that diabetes is associated with a characteristic lipid profile - normal LDL cholesterol, higher triglycerides and lower HDL levels. The high TG/low HDL profile has been shown to be associated with smaller, denser LDL particles which are thought to be more prone to oxidation, more atherogenic and likely to explain, in part, the higher rate of CVD.

The appropriate definition of diabetes and impaired glucose, mechanisms behind these biochemical changes found in diabetes and the appropriate management of these conditions will be discussed.


RH Mortimer, Department of Endocrinology, Royal Brisbane and Women's Hospital and Department of Obstetrics and Gynaecology, The University of Queensland, Brisbane, QLD.

Gestational diabetes (GDM) is defined as glucose intolerance that appears during pregnancy and regresses upon delivery. The cause appears to be a combination of increasing insulin resistance as pregnancy progresses and inability of the pregnant woman to achieve the increase in insulin secretion normally seen in pregnancy. A strong family history of Type 2 diabetes is common in GDM and women with GDM tend to be older and heavier and more often from non-Caucasian families than those without GDM. Up to 50% of women with GDM go on in later life to develop Type 2 diabetes.

The major obstetric problems associated with GDM are large babies and neonatal hypoglycaemia. Fetal size correlates with maternal weight gain and blood glucose control. Most guidelines recommend screening for GDM in pregnancy at 26–28 weeks gestation but there is debate about whether all women should be studied or whether screening should be restricted to a higher risk group. There appears to be a continuum between normality and GDM and biochemical definitions of GDM are somewhat arbitrary. Most commonly screening is done by ingestion of 50 g glucose and measurement of venous plasma glucose 1 hour later. Levels of ≥7.8 mmol/L are an indication for a glucose tolerance test where fasting levels of ≥5.5 mmol/L or 2 hour levels of ≥8.0 mmol/L are regarded as indicative of GDM. In a Caucasian population about 4% of pregnant women screened in this way will have GDM.

In our clinic all women with GDM receive dietary advice and monitor finger prick blood glucose levels four times daily. Failure to maintain fasting levels below 5.5 mmol/L or post prandial levels below 6.5 mmol/L is usually an indication for insulin treatment. There is however increasing interest in the use of oral hypoglycaemic drugs.


C Boothroyd, Departments of Endocrinology and Gynaecology, Greenslopes Private Hospital, Brisbane, QLD.

The investigation of infertility relies on a number of endocrinological assays as supplements to clinical assessment. This presentation will cover the expectations of clinicians when requesting endocrine assays of couples with infertility. In principle the adequacy of ovulation, tubal transfer of gametes and seminal quality require evaluation in all cases of infertility. Assessing ovulation can be performed clinically, ultrasonographically, by measurement of luteal phase progesterone and/or endometrial histology. The luteal phase progesterone is the most commonly used test but relies on correct timing of collection and this is often defined retrospectively.

Where oligo-ovulation or anovulation is present evaluation of oocyte reserve, hyperandrogenaemia and pituitary function (particularly hyperprolactinaemia) is required. Thyroid and adrenal disorders contribute to subfertility occasionally but are reversible and require evaluation. Polycystic ovarian syndrome is a common cause of reduced ovulation and the diagnostic criteria for PCOS are debated. Reference ranges for the frequently evaluated androgens and LH/FSH are ill defined. Assessing ovarian reserve relies on the determination of a low or relatively low oestradiol level with elevation of the FSH. Reversible or treatable pituitary pathology is occasionally present as the aetiology of male factor infertility and with severe male factor infertility karyotypic analysis is frequently appropriate.


L Penberthy, P Stewart, K Sikaris, J Gill, D Graves, M Meerkin, D Chesher, G Jones, R Bais, I Farrance. RCPA Quality Assurance Programs Pty Limited, c/o FMC, Bedford Park SA 5042

Following the Evaluation of the Australian Pathology Laboratory Accreditation Arrangements the Commonwealth of Australia, Department of Health and Ageing sought a mechanism for more quickly and efficiently identifying poorly performing pathology laboratories. As the routine quality assurance (QA) programs offer the most regular contact, assessing at least one aspect of laboratory performance, it is not surprising that these data were suggested as a means of identifying poor performance.

RCPA Quality Assurance Programs Pty Limited was concerned that the use of QA data in this way could change their relationship with participants and may change the way some laboratories approach their participation in QA programs. Nevertheless, as it was probable that the mandatory use of QA data for this reason would occur, the Company felt it was essential that they were involved with the profession, in establishing mechanisms for using these data. The Commonwealth Government has encouraged the professional development and discussion of tools using QA data and has supported this initiative by funding a project to develop Key Performance Indicators (KPIs).

It was decided that the KPI project would be initiated using chemical pathology QA data. The development of the project and prototype reporting systems will be discussed. Considerations such as the selection of analytes for assessment and selection of relevant statistical data will be addressed. The wide-ranging use of this reporting system to all participants in the QA programs rather than confining the tool to only identifying poor performers will also be considered. In addition it is emphasised that KPIs are only a tool to assist in identifying the analytical performance of participants in the QA programs.

While not within the scope of the KPI project the involvement of the profession in reviewing a laboratory identified as a possible poor performer will be discussed.


T Gaffney, Chemical Pathologist, Sullivan Nicolaides Pathology, Taringa Brisbane, QLD.

Male virility is associated with serum testosterone levels, with a cutoff less than 8 nmol/L used for the diagnosis of hypogonadism. Subtle progressive diminution in androgen levels have been documented in ageing men. In its position paper, the Endocrine Society of Australia (ESA) states that there is no evidence at present that the modest decrease in circulating blood testosterone commencing in mid-life has any clinical significance. Promotion of androgen therapy for 'male menopause' is based on inadequate evidence and is inappropriate. While it is understandable that the ESA adopts a conservative approach, many other authors suggest that one must wonder why mainstream medicine has been so slow to incorporate androgen deficiency in ageing men into its therapeutic armamentarium. The answer may lie in the titillating history of early testosterone replacement, and the notoriety it brought to the field. Testosterone has many effects; not only sexuality, but on muscle strength, osteoporosis, cognition, cardiovascular system, lipids and the prostate. With over 15% of humans greater than 65 years being reached by 2025, the effects on male independence will be enormous. Since the long term effects of testosterone replacement need to be determined, a Men's Health Study comparable to that in females needs urgent implementation.

Erectile dysfunction (ED) is a common condition affecting around 50% of the male population over the age of 40 years. Some 15 years ago the intracavernosal injection of alprostidil was first used to treat ED. Since then oral agents such as Sildenafil, Vardenafil and Tadalfil have been introduced, with a high degree of efficiency and safety. As a group they act by the selective inhibition of the enzyme phosphodiesterase type 5 (PDE5), thereby potentiating the action of nitric oxide (NO) on penile and cavernosal smooth muscle relaxation, so necessary for erectile efficiency. To compare the effectiveness of these drugs an International Index of Erectile Function (IIEF) has been developed, which allows for some quality control in this difficult area. Sildenafil has now been evaluated for over eight years, and it has been proven that it can be used safely and effectively to treat ED. Graphic newspaper advertisements ensure that the general public is bombarded with information regarding the increase in male satisfaction with the use of these drugs. It would seem that the fading male may become a myth, or at least will fade at a rate comparable with the diminution of other organ systems.


B McCall, Communicable Disease Control Institute: Brisbane Southside Public Health Unit, PO Box 333, Archerfield Qld 4108

Bioterrorism has emerged as a key public health issue in the 21st Century. The deliberate use of micro-organisms in warfare has been known for centuries, however the threat to civilian populations was emphatically demonstrated by the anthrax attacks of October 2001. These attacks, in the form of mailed anthrax spores designed to be easily aerosolised and inhaled, caused 22 cases of anthrax infection with five deaths. The resulting community alarm, confusion, pressure on the health care system, demand for chemoprophylaxis and loss of confidence in the mail system demonstrated the effectiveness of this agent for terrorist purposes.

The extensive literature on bioterrorism has focused on the biological agents that if released are likely to cause major impact in terms of mortality, morbidity and cost. The US Centers for Disease Control has nominated anthrax, smallpox, plague, tularemia, viral haemorrhagic fevers and botulinum toxin as the principal (Category A) agents of concern in terms of clinical and public health impact. However, it is equally important to recognise that Salmonella typhimurium and Shigella sp. have been used in deliberate food contamination incidents that were not initially suspected as intentional. While not having the clinical or terrorist impact of anthrax or smallpox, they are perhaps easier to obtain and disseminate. The health system must prepare for the most dangerous agents, but it must not lose sight of the potential for other agents to be used for bioterrorism.

This presentation will provide an overview of bioterrorism and a synopsis of the current biological agents of concern.


HA Morris, Hanson Institute, Frome Road, Adelaide SA 5000

Proliferation of healthy human cells is dependent on a complex interplay between proliferating cells and other cells in their immediate environment. Cell reproduction requires the recognition and activation of the appropriate signals received from other cells in the vicinity. In cancer this control is dysregulated. A tumour descends from one ancestral cell that at some point has commenced a program of unregulated proliferation through the serial accumulation of mutations in specific classes of genes that maintain and/or stimulate a growth advantage.

Hundreds of genes are involved in cell growth and proliferation as well as genes that ensure the maintenance of DNA integrity during replication. Other important genes regulate the destruction of unwanted cells. The development of cancer can result from mutations in any of these types of genes. Cancer cell survival also requires down regulation of the immunological surveillance system.

Proto-oncogenes guide the cell through the cell cycle sequence. When mutated an oncogene may drive excessive proliferation by producing increased levels or increased activity. Alternatively a downstream protein such as a receptor, may be constantly activated. When tumoursuppressor genes are mutated growth inhibitory signals are lost. Genes that check the integrity of DNA replication are also often mutated in cancer resulting in the rapid introduction of further mutations into the cell. Final backups to growth control are the genes controlling cell death (apoptosis). These are initiated in health when the cell is damaged. Oncogenes and tumour suppressor genes account for a large proportion of cancers.

Identification of these numerous mutations in the many different kinds of cancers have provided a wide array of potential targets for their specific treatment. Recently we have seen significant success with the development of molecular agents to treat specific cancers based on identification of specific mutations in certain cancers.


J Prins, Princess Alexandra Hospital & University of Queensland Brisbane, QLD

The incidence of Type 1 diabetes and, particularly, Type 2 diabetes is increasing world wide with enormous associated personal and community costs. This rising disease burden emphasises the need to develop and implement effective and tailored prevention and treatment strategies.

A huge research effort is in place to unravel the genetics and biochemistry of diabetes and to characterise, at a molecular level, the influence of environmental factors such as obesity, nutrition and gestation. In terms of disease prevention, this research has identified a number of putative factors that may contribute to the risk of the development of diabetes. It now remains to prove causation and then trial interventions.

The biochemical basis of insulin resistance is becoming more understood, and this has led to the development of new classes of therapeutic agent that are currently under trial. It appears likely that in the foreseeable future, patients will be biochemically and/or genetically "characterised" to facilitate choice of therapy. Such tailoring of therapy has potential benefit in terms of treatment efficacy, reduction on treatment side-effects or interactions and reduction in cost.

Finally, "lifestyle modification" has a proven role in the prevention and management of diabetes, but the major barrier to its widespread implementation is low compliance. Future strategies here will involve identification of specific components of diet or physical activity or environment that contribute to diabetes. Such components may be beneficial or detrimental and many will be highly amenable to manipulation. For example, specific dietary fatty acids influence insulin signal transduction via defined and characterised biochemical pathways. For this and other reasons "nutra-peutics" may have a prominent role in future diabetes management.


D Kanowski, Biochemistry Department Sullivan Nicolaides Pathology , Indooroopilly, QLD

Screening for Down Syndrome in the second trimester has become an accepted part of obstetric practice during the last decade, despite less than optimal sensitivity (approx. 67% detection rate for a 5% false positive rate using 3 markers). Several advances since this time have documented markers which have now allowed antenatal screening to be moved earlier into the first trimester. These discoveries are the finding of increased nuchal translucency in the 10–14 week period associated with aneuploidy, and the identification of elevated levels of maternal free beta hCG and reduced levels of maternal serum PAPP-A (pregnancy associated plasma protein - A) in Down Syndrome pregnancies. This has led to the possibility of combined ultrasound and biochemical screening in the first trimester in which 90% of Down Syndrome cases could be identified at a 5% false positive rate. However, this combination test poses some logistic and interpretation issues which must be addressed for it to be successful.


RJ Scott, Discipline of Medical Genetics, University of Newcastle, NSW.

Colorectal cancer affects approximately 1:20 persons and represents a significant burden to our healthcare system. Until about 1986 little was known about the molecular mechanisms that underlay disease development. Studies into inherited forms of colorectal cancer, however, have revealed much information about the molecular mechanisms, which underpin this devastating disease. Initial studies focused on familial adenomatous polyposis (FAP), which is a rare condition that occurs with a frequency of approximately 1:10,000 births. The disease is characterized by the presence of hundreds to thousands of premalignant adenomas. If left untreated colorectal cancer will develop with virtual certainty. The molecular basis of this disease has been shown to primarily be associated with mutations in the APC gene whose protein product is involved in the Wnt/Wingless signaling pathway.

A more frequently inherited predisposition is that of hereditary non polyposis colorectal cancer (HNPCC). This condition is characterized by familial aggregations of epithelial cancers (primarily colorectal cancer), which occur at unusually early ages. There is, unlike FAP, no premalignant marker that can be used to identify persons at risk of the disease. The genetic basis of this disease has been shown to be defective DNA mismatch repair. Defective DNA mismatch repair can be identified as it leaves within tumours a characteristic signature of DNA microsatellite instability.

There are distinct differences between FAP and HNPCC and these have lead to significant breakthroughs in our understanding of how colorectal cancer develops not only in these highly specialized disease entities but also in the general population. Studies of familial forms of colorectal cancer have been the paradigm for our understanding of the molecular mechanisms associated with colorectal cancer development and will remain so for some considerable time to come.


D Siebert, Queensland Health Pathology Service Royal Brisbane Hospital, QLD.

The single common feature of human hepatitis viruses is that they cause an inflammatory disease that is largely confined to infected hepatocytes. The most reliable chemical marker of this process is ALT. While serology is the most valuable tool in the diagnosis of viral hepatitis, the role of nucleic acid testing is increasing.

In Australia the most prevalent hepatic virus infections are caused by the blood borne agents (Hepatitis B) HBV and Hepatitis C (HCV). Risk factors include injecting drug use, exposure to contaminated blood and blood products and in the case of HBV, sexual and neonatal exposure to a person with actively replicating HBV. Both viruses have the capacity to establish persistent infection but HBV is a DNA virus while HCV has an RNA genome with a distinct replication strategy. The respective markers of infection and disease can be used to predict the outcome of both acute and chronic infection.

Hepatitis D (HDV or Delta) is a satellite virus or Virusoid that can cause either a co-infection or a super-infection of people who acquire HBV.

Hepatitis A (HAV) and E (HEV) are enterically transmitted agents with distinct geographic and socioeconomic demography. They are RNA viruses from unrelated families. While HAV is a sporadic largely food-borne disease of adults in affluent communities, HEV causes epidemic outbreaks in communities with poor water supplies and quality. Except for fatal HEV in pregnant women both are acute diseases from which almost all sufferers recover.

Vaccination interrupts the transmission of HAV and HBV (and HDV).


MP Metz1, T Petrou2

1 Clinical Biochemistry, IMVS, Adelaide, SA, Australia

2 Clinical Pathology, IMVS, Adelaide, SA.

Introduction The presence of macroprolactin and its detection by automated immunoanalysers have been the subject of extensive recent discussion in both the clinical chemistry and endocrinology literature. In order to better understand the performance of the Abbott Architect in regard to macroprolactin we undertook this study of macroprolactin with the Architect using the PEG precipitation method.

Methods 41 specimens of serum which had been analysed for prolactin at the IMVS on the Abbott Architect were studied. 21 had values greater than 25 μg/L and 20 had values less than 25 μg/L. Re-analysis following PEG precipitation took place one day after initial analysis. PEG precipitation was achieved by mixing 250 μl of a PEG 6000 25% solution to 250 μl of serum. This was vortexed for 1 minute then centrifuged for 5 minutes at 10,000 g. The supernatant was analysed for prolactin in the usual fashion.

Results The initial prolactin concentrations ranged from 7 to 257 μg/L with a mean of 45 and SD of 53. The plot of percent recovery vs. frequency revealed two populations. One had a distribution with a mean of 76% and SD of 12. The other population consisted of three specimens with recovery less than 20%. These had pre-PEG precipitation values of 47, 51, and 105. The post-PEG values were 4, 9, and 14 respectively. Passing-Bablok analysis of the pairs, excluding the three with low recovery, yielded a slope of 0.76 and intercept of 0.3. The Pearson coefficient of correlation was 0.99.

Conclusion In this population, macroprolactin was present in 3/41 (7.3%) of all specimens and 3/21 (14.3%) of specimens with increased prolactin concentrations as detected by the Architect. Samples with macroprolactin, when analysed using the Abbott Architect yield increased values for prolactin. PEG precipitation is an effective method to determine macroprolactin presence with the Abbott Architect.


J Tate1, P Mollee2, D Gill2, R Cobcroft2, PE Hickman1.

Chemical Pathology1 and Haematology Departments2, QHPS, Princess Alexandra Hospital, Brisbane, QLD 4102

Introduction Current electrophoretic and immunofixation methods fail to detect non-secretory myeloma (NSM), detect only 76–79% of monoclonal free light chains (FLC) in AL-amyloidosis, and may misclassify light chain myeloma (LCMM) during monitoring. FLC measurement in serum may prove useful for the diagnosis and/or monitoring of these diseases.

Methods FLC concentration was measured in patients with NSM, AL- and AA-amyloidosis, multiple myeloma (MM), peripheral blood stem cell transplantation (PBSCT)-treated myeloma, renal disease, monoclonal gammopathy of undetermined significance (MGUS), and in healthy subjects using a kit assay (The Binding Site Ltd., Birmingham, UK) and the IMMAGETM (Beckman Coulter, Brea, CA) nephelometer.

Results Kappa/lambda (K/L) FLC ratio was abnormal in 1 of 2 NSM, 5 of 8 AL-amyloidosis, all LCMM (n=9), all MM (n=15), 2 of 8 plateau phase PBSCT-myeloma with residual monoclonal protein, and 10 of 33 MGUS. FLC ratios were within the reference interval for healthy subjects, patients with renal disease or AA-amyloidosis, and PBSCT-myeloma in complete remission. Changes in the FLC ratio in LCMM corresponded to changes in urinary Bence Jones protein (BJP) output, whereas the ratio did not always correspond to an increasing monoclonal intact immunoglobulin concentration in relapsed PBSCT-myeloma. Monitoring of FLC levels in myeloma and AL-amyloidosis patients post-chemotherapy or transplantation allows assessment of disease response and may identify disease reoccurrence earlier especially if monoclonal FLC is not detected on immunofixation. We speculate that MGUS patients with an abnormal K/L ratio and urine BJP are likely to progress to MM.

Conclusions FLC measurement in serum can detect and monitor NSM and monoclonal FLC in AL-amyloidosis where traditional electrophoretic and immunofixation methods are ineffectual. In LCMM, FLC measurement is useful where 24-hour urine collection is difficult or unsatisfactory.


J Tate1, D Gill2, R Cobcroft2, PE Hickman1.

Chemical Pathology1 and Haematology Departments2, QHPS, Princess Alexandra Hospital, Brisbane, QLD. 4102

Introduction Measurement of free light chains (FLC) in serum is reported to be useful for the diagnosis and monitoring of monoclonal light chain diseases and non-secretory myeloma. We have identified some potential analytical and clinical pitfalls of the assay.

Methods The FLC kit assay (The Binding Site Ltd., Birmingham, UK) was assessed for precision, sample type and stability, and method harmonisation (n=37) using the IMMAGETM (Beckman Coulter, Brea, CA) and BN ProSpecTM (Dade-Behring, Marburg, Germany) assay systems. FLC concentration was measured in various subjects: myeloma (n=25); apparently healthy subjects (n=10), to determine biological variation; monoclonal gammopathy of undetermined significance (MGUS, n=33); transplanted myeloma (n=19); renal disease (n=13).

Results Serum and lithium-heparin plasma samples gave comparable FLC values and were stable for 1 week at 2–8°C, −20°C and −70°C. Intra-individual CVs for FLC concentration were generally <2.5%. Between run imprecision (CV, n=13–21) was 10.5–6.0% at kappa (K) FLC concentration, 7.1–39.7 mg/L, and 11.4–7.6% at lambda (L) FLC concentration, 10.6–53.4 mg/L. For the K/L FLC ratio, the mean difference between methods was −0.06 (95% C.I. −1.5 to 1.4). K/L FLC ratios were clinically concordant for 34 of 37 samples (quoted Reference Interval: 0.26–1.65). In myeloma, monoclonal FLC concentrations were elevated and ratios abnormal (0.003–223); in renal disease, FLC concentrations were elevated but K/L ratio was normalised (0.35–1.39); in MGUS, K/L ratio was outside the reference range in 10 of 33 patients (0.02–48). After stem cell transplantation for myeloma, K/L ratio was normalised in complete remission (0.55–1.18), but normal or outside the range for stable disease with residual monoclonal protein (0.23–2.9).

Conclusions Because of analytical imprecision, repeat testing of borderline samples is recommended to verify clinical classification. The use of FLC measurements alone cannot differentiate some patient groups with monoclonal gammopathy from healthy subjects.


G Koerbin, K Marshall, Northside Pathology, QHPS, Brisbane QLD

Introduction Six Sigma is the latest Quality Management strategy embraced by the industrial community. Six Sigma is about process improvement with the goal of attaining a process performance of 99.9997%. The sigma metric (capability) of an assay may be calculated from the imprecision (SD or CV) and inaccuracy (bias) observed and the allowable limits of error {Sigma = (allowable error-bias)/precision}. Current industrial standards consider a sigma metric of 6 or greater to be of world class and less than 3 to be unacceptable.

Methods We calculated the sigma metric for 5 months of internal quality control data for 41 of our routine automated analytes using the allowable limits of error as proffered by the RCPA/AACB quality assurance program. Process improvement or decline was reviewed in conjunction with reagent, QC and calibrator lot number change, QC rule failure and preventative or other maintenance requirements.

Results Initial results showed 10 assays to have sigma <3 (unacceptable performance). The performance of 5 of these analytes improved with changes in QC lot number and quarterly analyser preventative maintenance. Five analytes consistently under performed. The underperforming assays also showed a greater number of QC rule failures compared with the 18 analytes performing at a sigma of 6 or greater.

Discussion Careful consideration should be given to defining the allowable limits of error if Six Sigma Quality Management is to aid in identifying the performance of assays as unacceptable or world class. For laboratories to make improvements in assay performance the options sought are generally to change the assay or instrument, both of which are expensive and time consuming. Alternatively, Six Sigma used in conjunction with the optimisation of QC use and QC rules, should improve quality management by minimising wrong or bad results and repeat testing whilst improving the laboratory’s ability to monitor troublesome assays.


MJ Whiting*, M Pejakovic*, JA Leach* and A Keech#,

*Central FIELD Laboratory, SouthPath, Flinders Medical Centre, S.A. & #NH&MRC Clinical Trials Centre, University of Sydney, NSW 2006.

Introduction Patients with type 2 diabetes mellitus are at high risk of vascular disease and show many changes in their plasma lipid and lipoprotein profile. Fibrates may be important for the treatment of diabetic dyslipidaemia by altering triglyceride and HDL metabolism through effects on apolipoprotein gene expression. We have examined the effect of administration of fenofibrate (FF) on plasma lipids and various apolipoproteins in diabetic patients.

Methods 100 patients (age range 50–75 y) with type 2 diabetes mellitus were selected from registrants in Australia and New Zealand for the FIELD Study. Fasting plasma was collected prior to randomization to FF (200 mg/d) or placebo, and again after 4 months. Total cholesterol, triglycerides, HDL- & LDL-cholesterol were analyzed enzymatically, while apolipoproteins A1, A2, B, C3 & Lp(a) were measured by immunoturbidimetry. FF in plasma was determined by HPLC.

Results: FF was present in the plasma of 48 patients (mean 8.2; range 1.4–26.7 mg/L). Mean lipid (mmol/L) and apoprotein (g/L) levels were:

An external file that holds a picture, illustration, etc.
Object name is cbr24_3ps01f1.jpg

Conclusions: The improved plasma lipid and lipoprotein heart disease risk profile produced by FF in diabetic patients is consistent with its proposed mode of action via peroxisome proliferator-activated nuclear receptors, which influence the expression of apoprotein genes. In particular, inhibition of the production of apoC3 should increase lipoprotein lipase activity, thus explaining the large reduction in plasma triglycerides. The FIELD Study of 9,795 diabetic patients will determine whether these changes in the lipoprotein profile are translated into clinical benefits in terms of a reduction in cardiovascular events by FF.


C Mamotte1, L Langan1 and S Vasikaran2.

1Departments of Clinical Immunology and Biochemical Genetics, and 2Clinical Core Pathology and Biochemistry, Division of Laboratory Medicine, Royal Perth Hospital, Perth, WA, 6000.

Introduction Gilbert syndrome, a benign cause of mild unconjugated bilirubinaemia, is due to an extra TA dinucleotide in the TA repeat motif (7xTA compared to 6xTA) in the promoter for UDP-glucuronosyl transferase 1A1, a hepatic enzyme responsible for bilirubin conjugation. We evaluated a method based on denaturing HPLC (dHPLC) for this polymorphism.

Methods The genomic region containing the TA repeats was PCR amplified and PCR products from patient samples were then mixed (1:1) with a) a reference Gilberts amplicon (G) and b) a reference normal amplicon (N). The mixtures were then heated to 95oC, slowly cooled to 60oC (35 mins) to promote heteroduplex formation, and 3mL injected onto a dHPLC column at 50oC, with elution of DNA monitored at 260 nm.

Results As expected, results for 7xTA homozygotes showed one main peak when mixed with the reference G amplicon, indicating the presence of 7xTA homoduplexes only, and two well resolved peaks when mixed with the N reference amplicon, consistent with the presence of heteroduplexes and homoduplexes. Conversely, results for 6xTA homozygotes showed a single peak when mixed with N, and two peaks when mixed with G reference amplicons. For 7/6xTA heterozygotes, there were two peaks for samples mixed with either with G or N. Injection of patient samples twice, one mixed G and the other with N, overcomes the potential for false positives/negatives inherent to use of either G or N alone, as in the original description of the method, should patients have other polymorphisms. Detection of heteroduplexes was apparent from 40–55oC, and was relatively insensitive to large differences in amplicon concentrations for the patient samples.

Conclusion dHPLC analysis is a simple, and robust method for the molecular diagnosis of Gilbert syndrome in routine practice, and far more rapid and economical than alternative approaches such as sequencing.


A Pocathikorn1,3,4 , RR Taylor 2,4, C Mamotte1

1Departments of Clinical Immunology and Biochemical Genetics, and of 2Cardiology, Royal Perth Hospital; and Schools of 3Surgery/Pathology and 3Medicine/Pharmacology, University of Western Australia. Perth WA, 6000

Introduction Real-time PCR has greatly facilitated measurement of mRNA expresssion. In this study we evaluated the influence of numerous factors on quantitation of mRNA using genes involved in lipid metabolism as a model.

Methods Total RNA was extracted from human peripheral blood cells using a guanidium-isothiocyanate/phenol-chloroform method, reverse transcribed to cDNA, and the preparation quantified by real-time PCR on the Roche Lightcycler for LDL receptor (LDLR) and LDL receptor-related protein (LRP) cDNAs, as well as for the cDNAs for the reference or house-keeping genes, b-actin and GAPDH. The assays were characterised in terms of cDNA stability, linearity and reproducibility.

Results Our results showed that accuracy and sensitivity in mRNA quantitation by RT-PCR, requires consideration of: a) common, but avoidable errors in spectrophotometric analysis of extracted RNA b) rigour in assessment of RNA integrity c) the choice reverse transcription enzymes (eg SuperscriptII compared to AMV and Omniscript) d) interferences from reverse-transcriptase reaction components on quantitative PCR d) gene specific differences in cDNA stability (LRP cDNA was very unstable) and e) as expected, thoughtful design of PCR primers to avoid amplification of related genes and contaminating DNA. In terms of reproducibility, the within-run coefficients of variation (CV) were 13, 14, 9 and 8%; and between-run CVs, 20, 14, 16 and 12% for LDLR, LRP, b-actin and GAPDH, respectively. Linearity of response spanned over 7 orders of magnitude for all cDNAs. The within-subject CVs, based on repeated analyses on 10 male subjects, were between 24–30% for the LDR or LRP relative to the reference genes.

Conclusion Real-time PCR is a valuable tool for measurement of mRNA expression, and was used to measure the expression of numerous genes pertinent to cholesterol metabolism. However, the technique, while highly sensitive and specific, demands consideration of numerous factors to obtain accurate results.


P Giannopoulos, G Jones, EM Lim

Chemical Pathology, St Vincent's Hospital, Sydney, NSW.

Introduction Recently the Abbott AxSYM hCG assay has been shown to be affected by interference from heterophile antibodies (1). We have introduced a rapid, cost-effective method which has identified two cases of this type of interference in this assay.

Methods Samples with hCG results between 10 and 1000 U/L are diluted using Abbott Total hCG diluent. For results between 10 and 100 U/L a manual 1 in 2 dilution is performed. For samples between 100 and 1000 U/L a 1 in 10 auto-dilution is performed. If the result of the diluted sample is not consistent with the initial result, the possibility of interference is considered. The rationale for this protocol is that addition of extra diluent into the reaction increases the amount of animal serum present, which may assist in blocking interference from heterophile antibodies.

Results We have used this protocol routinely for 14 months and identified two patients with false positive hCG results. The results of the undiluted samples were 87 and 83 U/L for the two patients but in each case hCG was undetectable after dilution in the assay-specific diluent. In both cases urine was negative for hCG and alternative serum methods were also negative. This is a false positive detection rate of 0.07%.

Discussion An Abbott product information letter for hCG suggests testing possible false positive results in another method or with a urine sample. We believe our protocol is an easier, cheaper and faster method of identifying false positive results.

Conclusion As the AxSYM is widely used for hCG measurements we believe that this protocol may be of benefit to other users of the AxSYM hCG assay to reduce the possibility of issuing false positive results.

1. Cole L et al. Clin Chem 1999;45:313–314.


P Graham, B Martin, M Roser, G Jones

Department of Chemical Pathology, St Vincent's Hospital, Sydney NSW.

Aim To evaluate samples collected in the new Becton-Dickinson Plasma Separator Tubes (PST II) and Serum Separator Tubes (SST II) for comparability with results from plain tubes and stability on storage in the gel tubes for up to seven days.

Methods PST, PST II, SST, SST II and a plain tube were collected from five volunteers. Analytes were measured in all tubes and results compared those obtained from the plain tubes. The PST II and SST II were stored in a glass front fridge at 4°C and re-tested after 1, 3 and 7 days. Separate centrifuged PST II and SST II tubes were stored for up to 3 days and re-analysed for certain analytes before and after re-centrifugation of the tube. Routine chemistry tests were performed on an Hitachi 917; basic endocrinology on an Abbott AxSYM; vitamin B12, folate and insulin on a Beckman-Coulter Access; osmolality on an Advanced Osmometer and ionized calcium and pH on a Radiometer ABL 725.

Results Compared to non-gel serum tubes, LD was consistently higher and potassium and phosphate were lower in the plasma tubes. Unexpectedly ALT was slightly lower and folate and prolactin were higher in the heparin tubes. Regarding stability, most analytes were stable for 7 days when stored on the gel. Storage guidelines were developed for the less stable analytes (glucose, bicarbonate, phosphate, bilirubin, LD, sodium, potassium and phosphate). For glucose, LD and phosphate re-centrifugation of the primary tube was not recommended.

Summary Most tested analytes were acceptable in the PST II and SST II tubes on the day of collection although the expected plasma /serum differences were seen. Further evaluation of plasma folate, prolactin and ALT is required. Stability of tested analytes varied between 1 and 7 days under routine storage conditions.


LM Boscato, Chemical Pathology, St Vincent's Hospital, Sydney, NSW.

Introduction The detection of B2 transferrin (B2T) in discharges from the ear or nose is a useful tool in the investigation of suspected CSF leakage. False elevation of B2T can arise from bacterial sialidase activity in the sample (1). Failure to detect a spurious result could lead to inappropriate patient management. While sialidase contamination can be verified by the detection of B2T produced following incubation of serum transferrin with the sample, this procedure is time consuming and labour intensive. BVBlue® is a point-of-care enzyme activity test for detection of bacterial sialidase activity in vaginal swabs. The aim of this study was to assess the potential of the BVBlue® test for the confirmation of sialidase activity in fluids thought to contain CSF.

Methods BVBlue® involves incubation of the specimen with a chromogenic substrate followed by addition of colour developer. The colour can be assessed visually or read at 590nm. Six known sialidase negative samples and two positive samples were tested. Sample volume and incubation time was investigated.

Results The test clearly discriminated between positive and negative samples. Increasing the sample volume and incubation time enhanced the test response. OD at 590nm was superior to visual assessment for identification of samples with lower levels of activity.

Conclusion BVBlue® appears to be a viable alternative to the current technique for detecting sialidase activity. Test sensitivity can be easily increased if necessary. Ease of use and speed make the BVBlue® test suitable for routine screening of specimens.

Reference (1) Boscato LM, Edwards GA, Bayoun R and Brady L. Spurious elevation of B2 transferrin in bacterially contaminated specimens. Clin Biochem Revs 1994;15:135.


T Yen and DR Sullivan, Depts. of Clinical Biochemistry at the Pacific Laboratory Medicine Services (PaLMS) Royal North Shore Hospital and Royal Prince Alfred Hospital, NSW.

Introduction HMG-CoA reductase inhibitors (statins) are thought to affect bone metabolism. We were interested in PTH, vitamin D and calcium metabolism in patients on statins because of their putative anti-resorptive action on bone.

Methods Samples were from 213 sequential patients who attended the RPAH lipid clinic between 2000–2002. In addition to serum biochemistry, intact PTH, 25D and 1,25D were measured.

Results Vitamin D deficiency was identified in 23/104 males and 29/109 females (pop prevalence 24%). There was no season sampling bias (26% winter, and 22.9% summer). Twelve males and 20 females with vitamin D deficiency had 2o hyperparathyroidism, and 7 males and 8 females were hypocalcaemic. In patients on non-statin treatment, 30% of males and 20% of females had 25D < 40nmol/L. In contrast, males taking statins showed a lower incidence of vitamin D deficiency (2/31 and 6/31 patients, winter and summer) with a trend to higher 1,25D levels, whereas statin-treated females a higher incidence of vitamin D deficiency than non-statin counterparts (12/29 and 5/20), higher mean PTH (p < 0.02) and lower 1,25D. During this period, 3 patients had CK >1000 U/mL, and myalgia was reported by 17/111 patients taking statins (15%). Mean serum CK was higher in patients on statins, and in females there was significant correlation between CK and those reporting muscle complaints.

Discussion Vitamin D deficiency was common in patients treated for hyperlipidaemias. Some notable sex-based differences in vitamin D, PTH and calcium levels were observed particularly in relation to statin treatment. The contribution of vitamin D deficiency to musculoskeletal symptoms attributed to statins remains under investigation.


MP Metz1, L Nesci2, F Ku2

1 Clinical Biochemistry, IMVS, Adelaide, SA, Australia and

2 Core Laboratory, WCH, North Adelaide, SA.

Introduction Bilirubin in blood is a complex analyte with many methods used for its detection in the clinical setting. It has great clinical significance in the management of babies both well and unwell. Depending upon the setting, values greater than 50 μmol/L can trigger medical intervention. At the WCH, bilirubin concentrations are measured 8400 times yearly in plasma from neonates with the Beckman Coulter Jendrasik-Grof modification of the diazonium salt method using dichromatic detection. The Radiometer ABL measures absorbance across 128 wavelengths in whole blood. We aimed to compare them.

Methods From babies in the WCH, 118 specimens were collected, split, and analysed with the CX5 and ABL by laboratory staff. 61 were from capillary collections and 57 from indwelling vascular catheters. Values using the CX5 method ranged from 6 μmol/L to 305 μmol/L in capillary collections and from 30 μmol/L to 279 μmol/L in catheter specimens.

Results By the Student's t-test, the ABL and CX5 results were significantly different, p<0.001. Passing-Bablok analysis of CX5 vs. ABL results revealed slopes and intercepts of 0.91 and −8, 0.87 and −8, and 0.93 and −3 for combined, capillary, and catheter specimens respectively. The Pearson correlation coefficients (r) were 0.97, 0.98, and 0.98 respectively. The slopes of the capillary and catheter plots were significantly different (p < 0.001). Bland-Altman analysis was undertaken.

Conclusion The concentrations of bilirubin as measured by the CX5 and ABL were significantly different but correlated well. The difference between catheter and capillary specimens suggest that the matrix and variably conjugated bilirubins are important in this comparison. Results from these two instruments could be used at the bedside with an awareness of their differences improving their usefulness.


E Whitham, Chemical Pathology, Women's and Children's Hospital, Adelaide, SA.

Introduction The Women's and Children's Hospital is a reference centre for the measurement of sweat tests in South Australia. For over 14 years a 2 hour sweat collection time had been used. Because of the recommendation from the National Committee for Clinical Laboratory Standards (NCCLS) for a 30 minute collection time we have changed to this after investigations showed no significant difference in results in sweat collections from volunteers. This poster compares the results of 643 sweat tests with 30 minute collections with those obtained with 2 hour collections.

Methods Chloride and sodium results for each of 11 age groups were compared. Chloride and sodium results in different weight categories were also compared. Because of a change of sweat weights over the 14 year period, comparisons were also made with those 2 hour collections (1999/2000) which had weights that were comparable with the 30min collections.

Results The variation in sweat chloride and sodium concentration with age was confirmed. The 30min results were significantly different from the overall 2 hr results for chloride (p=<0.01) but not sodium (p=0.42) and not significantly different from the 1999/2000 results for either chloride (p= 0.24)or sodium (p=0.69). For both groups the sodium and chloride results decreased as weight of sweat increased although this was more noticeable in the 2 hour results.

Conclusions The difference seen in the chloride values is probably in part due to a change of method but the similarity with the 1999/2000 results where the weights are similar indicates that weight also plays a role. Although the majority of the earlier 2 hr results had a sweat formation rate less than the recommended 1g/m2/min, decreases are also seen with weight seen with the 30min collections which do have sweat formation rates greater than 1g/m2/min. This could be due to a higher percentage of sweat present as water vapour in lower weights, which is lost as the protective sheeting is removed, leading to a concentration effect.


A Read* and M Arumanayagam#,*Eastern Health Pathology, Box Hill Hospital, and PathLab, Vic. #Centre for Complementary Medicine Research, UWS, NSW.

Introduction Currently, as part of a research project, pregnant women, attending their first antenatal clinic (8–10 weeks gestation) at the Box Hill Hospital, are being surveyed, to assess their thyroid status. The first 100 results have been used to derive a reference interval for TSH and FT4 at this stage of pregnancy.

Methods TSH and FT4 were measured on the Roche Elecsys. Thyroid peroxidase antibodies were measured on the DPC Immulite 2000. Non parametric statistics were used to derive the reference intervals.


TSH mIU/LFT4 pmol/L
Laboratory reference interval0.27–4.212.0–22.0
All results<0.01–4.189.6–19.0
Reference interval from all results (n=96)0.04–2.510.3–18.2
Reference interval if raised TPOAb and TSH less than 0.1 plus raised /borderline FT3 excluded (n=83)0.24–2.3110.4–17.8


A reduced TSH in the first trimester is consistent with the findings of other investigators. We also found a reduced range for FT4, whereas most investigators have found FT4 to be the same as females or slightly raised. Thyroid peroxidase antibodies have been reported to be positive in 5–15% of non-pregnant women3. We found them to be positive in 10%. Most reports have not looked at individual patient results. Five women with TSH levels less than 0.1 mIU/L had FT3 ranging from 5.2 to 8.1 pmol/L (non-pregnant reference interval 2.8–7.1pmol/L). We do not have sufficient follow up to show whether these women should be classified as borderline T3 hyperthyroidism. All had quite normal FT4 levels.


JA Burgess, S Daniells and F San Gil

Illawarra Diabetes Centre and Clinical Chemistry Department, Illawarra Pathology, Crown St, Wollongong, NSW, 2500

Introduction Diabetics have a higher risk of developing chronic foot ulcers than non-diabetics. Poor nutrition is a frequently cited factor in wound-healing failure generally, but there is very little information regarding the nutritional status of diabetics with foot ulcers. This pilot study assessed indices of nutritional status relevant to wound healing in diabetics.

Methods Subjects (n=20) recruited from a foot ulcer clinic for diabetics had at least one chronic foot ulcer. Anthropometric data, weighed food records and fasting blood samples were collected. Nutritional and biochemical parameters were measured.

Results Subjects had a median age of 68 y and BMI of 35.7. Dietary data showed that all but one subject had adequate dietary intake of all nutrients. Glycaemic control varied greatly (HbA1c range 5.4 to 11.4%), as did the presence of diabetic complications (albumin/creatinine ratio ranged from 0.6 to 143.3 mg/mmol Cr). Blood levels were very low despite all subjects having a dietary intake well in excess of the RDI.

Discussion This study compared dietary intake with blood levels of the same micronutrients. Blood levels of a number of micronutrients, most importantly Vitamin C, an important micronutrient for wound healing, were very low despite all subjects having a dietary intake well in excess of the RDI. The findings suggest that diabetics have a greater dietary requirement compared to non-diabetics. This is important when assessing poor wound healing in this group. These findings also highlight the fact that RDI recommendations, that meet the nutritional needs of a healthy adult population, may be inadequate for people with diabetes and/or chronic foot ulcers.


W Ferguson, J Tate, P Hickman.

QHPS, Princess Alexandra Hospital, Woolloongabba QLD 4102.

Introduction The Roche Diagnostics Integra 400 Plus HbA1c whole blood assay was tested for precision, dilution linearity, recovery, accuracy and traceability, sample stability and effect of, sedimentation, anticoagulants and interferences. Test values were compared using the BioRad Variant II (V2) and Primus CLC-330 (CLC-330) assays.

Methods Accuracy and traceability to the National Glycohemoglobin Standardization Program (NGSP) were tested using value assigned whole blood samples. Precision, linearity, interferences, differences in anticoagulant type, sedimentation effect on results and stability of HbA1c over 0–28 days at different temperatures were tested. Integra sample values were compared with V2 values (n=60).

Results Precision: between-run CVs were 4.0%, 2.5% and 4.2% at HbA1c concentrations, 5.6%, 7.6% and 11.3%, respectively. Linearity: Hb 50–208 μmol/L and HbA1c 1.2–11.9 μmol/L. Recovery: 98.6%–100.6% in a range of 5.7%–12.5% HbA1c. Accuracy and Traceability: by linear regression analysis of Integra and CLC-330 HbA1c values (n=19); slope, y-intercept and correlation coefficient (R2) were 0.94, 0.71 and 0.991, respectively. The mean bias was 0.21% HbA1c. Sample Stability: EDTA samples were stable for 2 days at 22 °C, 7 days at 4 °C and 1 month at −70 °C. Effect of Sedimentation: Twelve results were comparable if repeated within 1 h. Interferences: HbC, E, F (>10%), G, S and homozygous Hb variants interfered. Anticoagulants: K3-EDTA, Li Heparin, Na-Citrate and Na-Fluoride/K-Oxalate gave comparable HbA1c results. Method Comparison: by Deming regression analysis of Integra and V2 HbA1c values; slope, y-intercept and R2 were 1.08, −0.19 and 0.976, respectively. The mean bias was 0.41% HbA1c.

Conclusion Analytical performance of the Integra 400 Plus whole blood HbA1c assay was generally acceptable. Accuracy was within the NGSP ± 1% HbA1c specifications. However, in our hands, between run imprecision was >3%.

Acknowledgement We wish to thank Ian Goodall for value assignment of patient samples.


K Lee1, B Barr1, J Chang1, S Bazeley2, S Klingberg2, J Tate1.

1Departments of Chemical Pathology, Princess Alexandra and 2 Royal Brisbane Hospitals, Brisbane, QLD.

Introduction The soluble transferrin receptor (sTfR) assay is used to evaluate marrow erythropoietic activity and iron status. sTfR levels are elevated in iron deficiency anaemia in the presence of inflammation and are normal in the anaemia of chronic disease. Some anomalous results may occur using a commercial immunoturbidimetric (ITA) assay.

Methods Precision, linearity, and sample stability were evaluated using the Orion Diagnostica IdeA sTfR assay (Espoo, Finland) and Beckman Coulter IMMAGETM analyser (Brea, CA), and values compared with the Dade Behring immunonephelometric (INA) assay (Marburg, Germany) and ProSpecTM analyser (N=50 fresh-frozen samples).

Results Between-run imprecision (CV): 6.7%, 4.8%, 7.9%, and 3.7% at sTfR concentrations of 0.9, 1.5, 2.6, and 6.4 mg/L, respectively. Dilutional linearity: recovery was 102–142% for two samples diluted up to 8-fold. Stability of samples at 2–8 °C and −20 °C was generally satisfactory with comparable sTfR values for 9 of 10 tested serum and lithium-heparin plasma samples. However, sTfR concentration for one serum sample decreased by 11–19% after 10 days' storage despite no change for the corresponding plasma. Method comparison: by Deming regression analysis, ProSpec sTfR = 0.85*IMMAGE sTfR + 0.06, correlation coefficient (R2) 0.854, mean bias −0.31 mg/L (95% C.I. −1.18 to 0.57). Some discrepant results were observed between methods. Possible explanations are differences in calibrations (no international standard available), differences in assay format (ITA versus INA), lot-to-lot antibody variation that was observed for Orion Diagnostica reagent (average −10% difference between two lots), or unidentified interferences.

Conclusions A user-defined sTfR assay adapted to the IMMAGE analyser gave acceptable precision but showed lot-to-lot variation between kits which may affect the clinical interpretation of results.

Some samples showed instability on storage or non-linearity on dilution and the reasons for this are unclear.


K Sanders, J Tate, J Chang, PE Hickman.

Chemical Pathology Department, QHPS, Princess Alexandra Hospital, Brisbane, QLD 4102

Introduction Testing for tau (β2)-transferrin in nasal, ear, and head and neck wound fluids is used to identify cerebrospinal fluid (CSF) leakage. However, results may be misinterpreted as positive if salivary sample is used for testing.

Methods Agarose-gel electrophoresis (SPE Paragon, Beckman, US), isoelectric focusing (IEF) at pH 4.0–6.5 (Pharmacia Biotech AB, Sweden), and immunofixation (IFE) with anti-human transferrin antiserum (Dako AS, Denmark; Beckman Coulter, US), were used to distinguish brain-specific transferrin (tau-, β2-Trf) that lacks neuraminic acid, from the tetrasialylated serum Trf (β1-Trf). Nasal, throat, and saliva samples were compared with CSF and serum for the presence of tau Trf.

Results In two patients with recurrent sinusitis (Case 1), and meningitis post craniotomy (Case 2), IFE of fluids gave only one Trf band that had similar mobility to β2-Trf, the β1-Trf band being absent. IFE of saliva gave a similar pattern consisting of one band with mobility cathodal to β1-Trf. Addition of normal saliva to CSF pool and incubation overnight resulted in altered electrophoretic mobility of the CSF with only one band visible in the zone between β1- and β2-Trf positions. In Cases 1 and 2, IFE suggested the samples were most probably salivary. By IEF the main Trf species identified in serum were tetrasialo Trf; in CSF, tetrasialo Trf and asialo Trf; in saliva, monosialo- and asialo Trf; in saliva + CSF, a mixture of disialo-, monosialo-, and asialo Trf. In Case 2, IEF identified monosialo-, asialo- and other Trf bands cathodal to asialo Trf that were possibly degradation products. The presence of coincident CSF in the throat fluid could not be determined.

Conclusion Endo-β-N-acetyl glucosaminidase HS enzyme, present in buccal epithelial cells in saliva, cleaves N-linked glycan side chains on the transferrin molecule resulting in a loss of sialic acid residues and electrophoretic patterns showing asialo Trf. To avoid false positive tau transferrin, throat fluids should not be used to test for CSF leakage.


J Tate1, W Ferguson1, L Callaghan1, S Klingberg2, K Marshall3, G Koerbin3. QHPS, Princess Alexandra Hospital1, Royal Brisbane Hospital2, and The Prince Charles Hospital3, Brisbane, QLD.

Introduction Benzethonium chloride (BTCl) and pyrogallol-red molybdate methods are routinely used to measure total protein in urine. However, values for test samples may vary between different methods. We investigated possible causes for the lack of method harmonisation.

Methods Precision, linearity, reactivity to purified proteins, and TUP sample method comparisons (n=46) were evaluated using in-house and Roche Diagnostics (Indianapolis, US) BTCl methods, and a pyrogallol-red molybdate assay from Roche Diagnostics (Mannheim, Germany).

Results The analytical performance of four TUP methods.

An external file that holds a picture, illustration, etc.
Object name is cbr24_3ps01f2.jpg

Generally Fara BTCl values compared with Dimension BTCl and Integra pyrogallol TUP values whereas Modular values were lower compared with other tested methods. Changes to the Modular TUP assay parameters and reagent composition led to increased protein values.

Conclusions The lack of harmonisation of total urine protein values between methods may in part be due to differing reactivity of the various proteins present in urine, assay format and BTCl reagent composition.


B Russell1, B Jones2, J Tate3, G Koerbin4, JM Potter4, PE Hickman3. QHPS, Redlands1, Logan2, Princess Alexandra3, and The Prince Charles Hospitals4, QLD.

Introduction Cardiac troponin T (TnT) is used to detect myocardial damage, with levels of TnT ≥0.03 μg/L shown to be associated with adverse cardiac outcomes in the setting of cardiac ischaemia. Method harmonisation, interferences, and lack of concordance between troponin I (TnI) and T values may affect clinical interpretation.

Methods Roche Diagnostics CARDIAC® reader (CR) and Elecsys 2010 (E2010) using 3rd-generation TnT reagent (Mannheim, Germany).

Results Following on from initial observations that a reaction line was visible on the strip in the absence of detectable TnT on the CR and an increased number of CR low level values (<0.10 μg/L), samples were quantitated on the E2010. Coincidentally, there was a change of the manufacturer's limit of sensitivity from 0.05 to 0.03 mg/L. We observed a higher than expected incidence of low level values (20%, lot no. 19141). Ongoing monitoring showed a decreasing incidence (5%) with later CR lot numbers. The discordance between CR and E2010 TnT at low level values was also variable during this time, from 27% to 9%. Other analytical issues included anomalous TnT values that were observed in two patients having CR TnT testing for the investigation of chest pain. CR TnT was detected in three of four consecutive samples in one patient (0.15, <0.01, 0.21, and 0.14 μg/L), but was <0.01 μg/L by the E2010. In another patient an unidentified interference prevented the control from reacting in the CR method. Comparative TnI values for some patient samples indicated a discrepancy between TnT and TnI methods.

Conclusions Monitoring the frequency of low level CR TNT may be an indicator of potential reagent problems. CR and E2010 TNT values are generally harmonised. Differences between values may be due to the greater imprecision of the CR assay or to matrix-interaction differences between the two assays. Generally there were few discrepancies between TNT and TNI methods for clinical agreement.


V Miranda, G Koerbin, W Ferguson1, QHPS, Northside Pathology and Princess Alexandra Hospital1, Brisbane QLD.

Introduction Diabetes mellitis is a serious universal health problem and its prevalence has increased consistently during the last decade. It is established that good glycaemic control can delay the onset and slow the progression of secondary complications. Glycohaemoglobin (HbA1c) is considered the most reliable indicator of glycaemic control since it reflects the average concentration of blood glucose over the circulating life span of haemoglobin in red the blood cells, (~120 days).

Method We evaluated the Dade Behring HbA1c turbidimetric inhibition (TINIA) assay available for the Dimension RxL following modified NCCLS guidelines using 61 samples that had been standardised with CLC-330 HbA1c values from the HbA1c Australian Reference Laboratory. An additional 45 problematic samples from patients with haemoglobinopathies, high urea, lipid and bilirubin values were assayed. The Dade assay measures both HbA1c and haemoglobin in the sample calculating the percentage of total haemoglobin that is glycated, the final result being standardised to the Diabetes Control and Complications Trial (DCCT). We then compared the results obtained to those obtained by the Biorad Variant II and the Roche Integra Immunoassay.

Results Total precision of the assay was observed as 2.89% and 2.00% at concentrations of 5.73% and 9.75% (n=77, n=97). Deming regression analysis showed a correlation of Dade = 0.94Variant + 0.24, r2=0.97, n=61 and Dade = 0.81Integra + 0.88, r2=0.99, n=61. No appreciable differences in concentration with the problematic samples was observed between the Dade and Variant assays.

Discussion The Dade assay performed well giving good precision and accuracy. The National Glycohaemoglobin Standardisation Programme (Little et al., Clin Chem 2001, 47:11, 1985–1992) recommends a total imprecision of <3.0% for the methods for determination of HbA1c for the Primary and Secondary Reference Laboratories. The CV obtained by the Dade method was well within this recommended range. The Dade assay is a simple one step assay for the determination of HbA1c showing excellent performance characteristics.


V Miranda, L Chan, G Koerbin

Chemical Pathology, Northside Pathology, QHPS, Brisbane QLD.

Introduction Mycophenolic Acid (MPA), the active immunosuppressant form of the pro drug mycophenolate mofetil (MMF) is widely used for the prophylaxis of rejection in solid organ transplant patients. The immunosuppressive effect of MMF, when used in conjunction with other immunosuppressive drugs is well established. There is general agreement that monitoring levels of plasma MPA is required for optimising its immunosuppressive efficacy.

Aim The two major methods available for the determination of plasma concentrations of MPA are HPLC and enzyme immunoassay (EMIT). We evaluated the Dade Behring EMIT 2000 MPA assay on the Dimension RxL and compared the results to those obtained from an HPLC method developed in house.

Results HPLC showed total precision of 1.1–4.9% (range 1.0 –12.5mg/L). The total precision for EMIT was 6.7–12.4% (range 1.2 – 12.7mg/L). Limit of detection for the EMIT assay was 0.28mg/L and 0.25mg/L for the HPLC. Serial dilutions of pooled patient plasma showed good linearity for both assays. There was excellent correlation between the two methods when QC data with assay values 1.0, 7.5 and 12.0mg/L were used. Passing-Bablok regression of heart and lung transplant patients showed EMIT = 1.35HPLC + 0.14, r2 = 0.98; n = 11.

Discussion It is known that the results obtained by immunoassay tend to be 7–35% higher than those obtained by HPLC. Though this positive bias is attributed to several factors including type and time of transplant and type of primary immunosuppressive drugs used, it is mainly due to the cross reactivity of the acyl glucuronide metabolite of MPA. The clinical significance of this metabolite is not clearly understood.

Conclusion The EMIT MPA assay on the Dimension analyser offers a simple and accurate method for the measurement of MPA in plasma. This assay offers a fast and cost effective method suitable for a routine analytical laboratory.


B McWhinney, P Hickman, S Edwards. Department of Chemical Pathology, QHPS, Princess Alexandra Hospital, Brisbane QLD 4102

Introduction Phaeochromocytomas are tumours arising from chromaffin cells and are characterized by excessive production of catecholamines and their metabolites. The diagnosis of these tumours is based upon the measurement of catecholamines and their metabolites in plasma and urine. However, none of these commonly used tests reliably confirms or excludes the presence of a tumour. The recent development of a new test involving the measurement of plasma free Metadrenalines offers more promise than these other commonly used tests.

Methods The samples were extracted as per the method of Eisenhofer1 with the following modifications. All wash steps were allowed to flow through the extraction columns under gravity, only minimal vacuum was applied to start the flow. An extra 0.5mL methanol wash was added as a final wash step. The samples were then eluted as per the Eisenhofer method. Different electrode settings, in particular the first electrode on the High sensivitity cell produced a significant reduction in the size of the paracetamol peak measured by the second electrode.

Results The reference ranges for plasma metadrenalines were that of Eisenhofer1. The assay for plasma free normetadrenaline has a between-run CV of 5.1% at 357 pmol/L (RR <700) and for metadrenaline between-run CV is 3.9% at 154 pmol/L (RR<400). The detection limit of the assay was 5 pmol/L.

Conclusion The modified assay developed, has overcome the paracetamol issue in the Eisenhofer method thus producing cleaner chromatograms and significantly less number of interfering peaks. A number of phaeochromocytomas have subsequently been confirmed in patients.

1. Lenders JWM, Eisenhofer G, Armando I, Keiser HR, Goldstein DS, Kopin IJ. Determination of plasma metanephrines by liquid chromatography with electrochemical detection. Clinical Chemistry. 1993;39:97–103.


S Edwards, B McWhinney.

Department of Chemical Pathology, QHPS, Princess Alexandra Hospital, Brisbane QLD 4102

Introduction The HIV anti-protease inhibitors (PI) have contributed substantially to the improvement in quality of life of many HIV-infected patients in the last decade. Inhibition of the HIV protease enzyme inhibits viral replication. The PI are usually administered in combination with two other antiviral agents, usually selected from the nucleoside reverse transcriptase inhibitors or non- nucleoside reverse transcriptase inhibitors (NNRTI). In addition, PI may be given in combination. This is of particular significance because the PI tend to inhibit each others metabolism. More complex drug interactions may occur when a NNRTI such as Efavirenz is added, which may increase PI metabolism.

Methods The assay is a simple and rapid organic extraction of alkanized plasma with diethyl ether. The ether layer is dried under nitrogen. The precipitate is dissolved in acetonitrile and then washed with hexane to remove lyphophillic interferences. The acetonitrile is then dried under nitrogen and reconstituted in mobile phase. The analytes are then separated from other components in the sample by gradient reverse phase HPLC using a combination of 15mmol/L phosphate buffer and acetonitrile. The peaks are detected at 215nm using a photodiode array detector.

Results Analytical recovery for all the analytes was in excess of 95%. The assay was linear to 10 mg/L for all analytes. Plasma spiked with 2mg/L of each drug produced inter-run imprecision CV's of less than 5%. The detection limit for each analyte was 50μg/L.

Conclusion The assay developed is a significant improvement on current methods and allows the simultaneous TDM of eight drugs commonly used in the treatment of HIV.


B McWhinney, P Hickman, C Davies

Department of Chemical Pathology, QHPS, Princess Alexandra Hospital, Brisbane QLD 4102

Introduction 6-Mercaptopurine (6MP) and its prodrug Azathioprine (AZA) are immunosuppressive drugs used to treat numerous conditions, including childhood acute lymphoblastic leukaemia, organ transplant rejection and inflammatory bowel disease. The 6MP and AZA are themselves inactive, but are converted to their active metabolites, 6-thioguanine nucleotide (6TG) and 6-Methyl Mercaptopurine (6MMP). Currently, most physicians measure treatment efficacy by an improvement in clinical symptoms and quality of life, and monitor toxicity by full blood counts.

Methods The assay is a simple and rapid deproteinization of whole blood by perchloric acid in the presence of dithiothreitol. The nucleotides are then hydrolyzed to their bases by heating the sample for 60 min at 100°C. The 6TG, 6MP and 6MMP are separated from other components in the sample by reverse phase HPLC using a 75mmol/L phosphate buffer/methanol gradient and the peaks detected using a photodiode array detector. The chromatograms are then processed at three different wavelengths.

Results Analytical recovery for the three analytes was in excess of 95%. The assay was linear to 40μmol/L for both 6TG and 6MP and 150μmol/L for 6MMP. Results were expressed as nmol/g of Hb. The detection limit for each analyte was 25 nmol/L.

Conclusion Whole blood was the sample of choice as this provides significant convenience, allowing EDTA blood samples to be transported frozen to the referral lab. The assay developed is a significant improvement on current methods and allows the simultaneous measurement of the 6MP and its metabolites. Photodiode array detection also increases the sensitivity and selectivity of the assay by monitoring and processing the analytes at each of their maximum wavelengths.


G Dimeski, W Ferguson, J Tate. QHPS, Princess Alexandra Hospital, Woolloongabba QLD 4102

Introduction The ACE assay is used to diagnose active pulmonary sarcoidosis and to monitor the effectiveness of treatment of the disease. An in-house ACE method was adapted to the Roche Modular P (Modular) and tested for precision, linearity, interference from haemolysis, and reagent stability. Test values were compared using an in-house 3-(2-furylacryloyl)-L-phenylalanylglycylglycine (FAPGG) assay on the Cobas Fara (Fara).

Method Modular method: a rate assay monitored at 340 nm over 10 min at 37 °C. ACE activity was determined using a commercial standard with a value assigned against the Fara assay from the molar absorptivity of FAPGG. Precision was tested using commercial and serum pools of low, medium and high ACE activity. A patient sample (158 U/L) was diluted in saline and tested at 11 dilutions. The on-board reagent stability was tested over a 30-day period. Modular sample values were compared to Fara values (n= 50).

Results Precision: Within run (n=5–10) CVs were 14.5%, 7.2%, 2.9%, 1.4%, and 1.7% at ACE activity of 6, 35, 68, 125 and 155 U/L, respectively. Between run (n=20) CVs were 11.4%, 9.9% and 6.4% at 35, 65, and 148 U/L, respectively. Linearity: by linear regression analysis of expected versus measured ACE activity (4–158 U/L), slope, y-intercept and correlation coefficient (R2) were 0.99, 0.79, and 0.999, respectively. The assay was unaffected by haemolysis (Haemolytic Index: up to 583). On-board reagent stability was at least 30 days. Method Comparison: by Deming regression analysis of Modular vs Fara ACE activity values, slope, y-intercept, R2, and mean difference were 0.87, 8, 0.956, and −1.5 U/L [95% C.I. −4.4 to 1.5], respectively. Poorer imprecision of the Fara assay contributed to some method discrepancies.

Conclusion We successfully adapted an in-house ACE assay to the Modular P analyser with values generally comparable between methods. Assay performance was better than for the current in-house Fara assay with improved precision and linearity.


Z Wang, Z Bostanian, M McKenna, E Whitters, JD Lei, T Ito, P Hsia, N Panosian-Sahakian, A El Shami. Diagnostic Products Corp., Los Angeles, CA, USA.

Circulating levels of androstenedione (AND), a major precursor of testosterone and estrone, are elevated in patients with abnormal hair growth (hirsutism) and virilization. We have developed competitive chemiluminescent enzyme immunoassays for the determination of AND levels in serum and plasma on the IMMULITE and IMMULITE 2000 automated analyzers. The assays use rabbit anti-AND polyclonal antibody on the solid phase. Alkaline phosphatase-labeled AND is coincubated with sample for 60 minutes. After four washes, enzyme activity is determined by adding chemiluminescent substrate. The detection limit of the assays was approximately 0.3 ng/mL, with a working range of up to 10 ng/mL. The assays exhibited low crossreactivities with 26 structurally related steroids. Dilutional parallelism experiments demonstrated a recovery of 91 to 114% throughout the assays' working range. Total assay CVs for five samples with values ranging from 1.19 to 8.9 ng/mL, tested over 20 runs, were 5.7 to 13.4% by IMMULITE, and 3.6 to 5.8% by IMMULITE 2000. Analyte recovery experiments obtained mean recoveries of 91 to 98% by IMMULITE and 87 to 104% by IMMULITE 2000. Comparison of these assays to a radioimmunoassay developed by DSL, Inc. (Webster, TX) on 34 patients yielded regression equations of IMMULITE = 1.14 (DSL) – 0.38 ng/mL (r = 0.85) and IMMULITE 2000 = 1.19 (DSL) – 0.45 ng/mL (r = 0.84). Patient means were 1.45, 1.45, and 1.62 ng/mL for the IMMULITE, IMMULITE 2000 and DSL assays, respectively. These data suggest that the AND assays developed for use on the IMMULITE and IMMULITE 2000 systems are suitable for routine determination of circulating AND levels, as an aid in the assessment of physiological/pathological changes of sex hormone metabolism.


Z Bostanian, E Gudipati, J Wen, E Whitters, JD Lei, D Sustarsic, E Unver, A S El Shami

Diagnostic Products Corp., Los Angeles, CA, USA.

D-dimer, a breakdown product of fibrin clots, is produced by normal clotting mechanisms. Some reports indicate that D-dimer tests are useful as predictors of both lower-limb deep-vein thrombosis and pulmonary embolism.

We have developed an immunometric assay for the measurement of D-dimer for the IMMULITE 2000 automated immunoassay analyzer. The assay utilizes a monoclonal murine anti-D-dimer antibody coated on a polystyrene bead. The bead is coincubated with serum for 30 minutes then incubated with alkaline phosphatase-conjugated murine anti-D-dimer antibody for an additional 30 minutes. The D-dimer assay has a working range of up to 2000 ng/mL. Dilutional parallelism studies yielded recoveries of 104 to 109% across the assay's working range. The assay demonstrated an analytical sensitivity of 0.5 ng/mL. Total assay CVs for samples ranging from 31 to 327 ng/mL were 6.9 to 16%. Comparison of the IMMULITE 2000 assay with the Dimertest Gold EIA (Agen Biomedical Limited, Acacia Ridge, Australia) on 77 patient samples yielded a correlation coefficient of 0.93, with IMMULITE 2000 = 0.24 (Dimertest Gold) – 12.6 ng/mL; means were 76.9 and 373 ng/mL for IMMULITE 2000 and Dimertest Gold, respectively. Analysis of 112 random normal and disease-state samples, using a cutoff value of 14 ng/mL for IMMULITE 2000 and 100 ng/mL for Dimertest Gold, indicated agreement of 97.4%; sensitivity and specificity were 100 and 94.4%, respectively. Investigation of 96 apparently healthy individuals aged 40 through 69 by the IMMULITE 2000 assay yielded a mean of 6.0 ng/mL and a 95th percentile of 16.9 ng/mL.

These results indicate that the IMMULITE 2000 D-Dimer assay is suitable for use as an aid in the assessment of the degradation products of fibrin clots.


E Whitters, JD Lei, N Figueroa, R Laroya, N Panosian-Sahakian, A S El Shami. Diagnostic Products Corp., Los Angeles, CA.

Gastrin, a key gastrointestinal hormone that serves to stimulate gastric acid secretion, is found in the circulation in three principal forms: G-17, G-34 and G-14. Monitoring the different forms of gastrin is useful for helping to identify Zollinger-Ellison tumors (gastrinomas) which are typically associated with elevated gastrin levels, gastric acid hypersecretion and peptic ulcer disease. We have developed a rapid, chemiluminescent enzyme immunometric assay for the determination of serum gastrin levels using the IMMULITE 2000 automated analyzer. The assay utilizes a monoclonal anti-gastrin antibody on the solid phase that is coincubated with patient serum and a pair of enzyme-labeled monoclonal and polyclonal detection antibodies in a 60-minute assay. The assay was determined to be specific for gastrin, with no crossreactivity with other structurally related hormonal substances and no apparent heterophile interference. The assay is calibrated to G-17 with reactivities for G-34 and G-14 being 18% and 10%, respectively. Dilutional parallelism experiments yielded mean recoveries of 98 to 111% on samples spanning the assay's working range (up to 1000 pg/ml). Total assay CVs were 11.5, 7.0, and 6.5% for samples with values of 17, 224 and 846 pg/mL, respectively. The assay had a detection limit of 20 pg/ml and exhibited no high-dose hook effect at gastrin levels as high as 120,000 pg/mL. Comparison of the IMMULITE 2000 Gastrin assay to the ICN Gastrin assay (n= 87) yielded a correlation coefficient of 0.95. A study performed on 20 apparently healthy fasting individuals demonstrated a mean of 45 pg/mL and a central 95% range of 20 to 84 pg/mL. Moreover, 15-minute postprandial test-meal stimulation of this population shifted the mean to 66 pg/mL and the central 95% range to 34 to 163 pg/mL.

We conclude that IMMULITE 2000 Gastrin is a sensitive immunometric assay for the determination of serum gastrin levels as an aid in the assessment of gastrointestinal hormonal function.


M Elmlinger 1, W Kühnel 2, A Durham3, 1University Hospital, Pediatric Endocrinology, Tübingen, Germany, 2DPC Biermann, Bad Nauheim, GERMANY, 3DPC., LA, CA, USA.

Current IFCC and NCCLS guidelines for the treatment of reference values accommodate discrete cofactors, like sex, but not continuous covariates, like age, where binning (the premature imposition of arbitrary age brackets) is clearly undesirable. The Harris and Boyd textbook (1995) illustrates global parametric techniques, due largely to Royston's group, which work for analytes like PSA with simple trajectories. Insulin-like growth factor I (IGF-1) and IGF binding protein 3 (IGFBP-3), however, exhibit far greater complexity. Moreover, artifacts of data reduction are certainly one major source of discrepancies among published reference ranges (Ranke et al, Horm Res 2000; 54:60). "Local regression" algorithms, constitute a promising alternative technique: Loader's LocFit -available for S-Plus and open source R and as a stand-alone program - is currently the most advanced implementation.

We applied LocFit to IGF-I and IGFBP-3 reference values and their ratios obtained in a study of ~1500 subjects in good health and of normal height, including children and adults (1wk-88y) from the Tübingen region, using automated chemiluminescent assays (IMMULITE, DPC). For each parameter, the age-related mean and variance were estimated globally, on a grid with 0.5y resolution, from log-transformed values restricted to 0.5m-88y and 0.5m-50y, respectively. In the sequential modeling of mean and variance functions, spans of 35% and 55% and degrees of 0–3/adaptive and 0, respectively, proved optimal for this data, as judged by regression diagnostics and visual inspection. The data to be presented will show the resulting graphs with +2, 0, −2 and −3 SD contours for IGF-I values from subjects up to 20y (n=793, both sexes combined). We conclude that LocFit provides a powerful framework for age-related reference range analysis, offering a suitable balance between flexibility and control.


Z Bostanian, G Hall, E Whitters, JD Lei, K Pregger, D Sustarsic, E Unver, A S El Shami. DPC, Los Angeles, CA.

PAPP-A is a placenta-derived glycoprotein, which, during pregnancy is produced in high concentrations by the trophoblast and released into the maternal circulation. Its functional significance is unclear but some studies suggest that reduced PAPP-A concentrations in pregnancy may be associated with chromosomal abnormalities in the fetus. Consequently, it might be used, in conjunction with serum free beta-HCG, for screening at 10 weeks of pregnancy.

Assays for the measurement of PAPP-A on the IMMULITE and IMMULITE 2000 analyzers have now been developed. Both assays utilize a monoclonal murine anti-PAPP-A antibody coated onto a polystyrene bead which is coincubated with sample and alkaline phosphatase-conjugated murine monoclonal anti-PAPP-A. Dilutional parallelism studies obtained recoveries of 92 to 104% across the working range of the assays (up to 10 mIU/ml) which also demonstrated an analytical sensitivity of better than 0.05 mIU/mL. Total assay CVs for both IMMULITE and IMMULITE 2000 were 4.9 to 7.8% for samples with values ranging from 0.42 to 4.92 mIU/mL. Comparison of the IMMULITE assay with the Kryptor PAPP-A assay (Brahms, Berlin, Germany) on 146 patient samples yielded a correlation coefficient of 0.97, with IMMULITE = 0.97 (Kryptor) + 0.52 mIU/mL; means were 3.57 and 3.15 mIU/mL for IMMULITE and Kryptor, respectively. Comparison of the IMMULITE 2000 assay with the Kryptor assay on 152 patient samples yielded a correlation coefficient of 0.99, with IMMULITE 2000 = 0.99 (Kryptor) + 0.498 mIU/mL; means were 3.82 and 3.36 mIU/mL for IMMULITE 2000 and Kryptor, respectively.

These data indicate that the IMMULITE and IMMULITE 2000 PAPP-A assays are suitable for the routine assessment of circulating PAPP-A in maternal subjects.


T Li, T Chuang, S Tse, D Hovanec-Burns, A El Shami DPC, Los Angeles, CA, USA.

Traditional RAST (mRAST) allergy blood test utilizes a single calibrator with 2 overnight incubations with allergens immobilized onto paper discs. Samples are classified as positive or negative and results are reported in mRAST class scores. Second-generation assays are quantitative enzyme immunoassays performed on batch analyzers with a working range of 0.35 to 100 kU/L and require 2 incubations taking up to 3 hours to complete.

In this study, we report the development of a third-generation quantitative chemiluminescent enzyme immunoassay for determination of allergen-specific IgE in human serum, which is performed on the continuous random access IMMULITE® 2000 system. The extremely low nonspecific signal afforded by liquid allergens and the centrifugal washing of the universal bead solid phase permits extension of the assay's low end, for a working range of 0.1 to 100 kU/L. Calibration of the assay is traceable to the WHO 2nd IRP 75/502. Sample and conjugate incubation times are reduced to 30 minutes and concordance, clinical sensitivity and specificity relative to the original mRAST were 90%, 88% and 92%, respectively (n = 812). The total number of allergens evaluated was 66, representing 9 different allergen categories. This method was highly specific for IgE with no significant crossreactivity to other subclasses. There was no interference with high concentrations of total IgE, anticoagulants, bilirubin, hemoglobin and triglyceride. Benefits of this method include an extended working range, lower detection limit, improved precision at the low end, dramatic reduction in allergen and sample setup time and a substantially reduced time-to-first-result of 65 minutes. The availability of the third-generation allergen-specific IgE assay on the IMMULITE® 2000 system allows for sensitive, accurate and reproducible allergy blood test results and facilitates the continued consolidation of laboratory workstations in immunoassay laboratories.


M Goodwin, M Harrop and P Colman. Departments of Biochemistry and Diabetes & Endocrinology, Royal Melbourne Hospital, Grattan St, Parkville, VIC 3050.

Measurement of salivary cortisol at 2300h has recently been suggested as a sensitive and specific screening test for the diagnosis of Cushing's syndrome but it does provide a number of challenges. Firstly, cortisol in saliva is unbound and thus in low concentration and secondly collection of saliva samples is not routinely used. However, cortisol stability potentially allows sampling of saliva within the patient's home with samples transported to the laboratory the next day. Our objective was to develop a method suitable for small numbers of patient samples which could be quickly and easily measured for cortisol using our routine assay, the IMMULITE 2000 cortisol assay. The required limit of detection was less than 3.6nmol/L (upper limit of 2300h reference range)(1). Results were compared to a commercially available RIA, (Coat-a-Count Cortisol, DPC) using Tunn's modification (2). 45 people without pituitary or adrenal problems collected saliva into plain "Salivettes", (Sarstedt Australia Pty Ltd., SA). Subjects were asked to collect saliva at 0800h and 2300h and to store samples at 4°C until transported to the laboratory. Centrifuged samples were stored at −20°C until assayed. Salivary cortisol was extracted using equal volumes of ethyl acetate and saliva. The ethyl acetate extract was reconstituted with DPC cortisol diluent to attain a 10-fold concentration and maintain a suitable matrix for analysis on the IMMULITE 2000. The ranges for salivary cortisol are shown below.

IMMULITE 20004.6–24.6 nmol/L (n = 35)<6.2 nmol/L (n = 43)
RIA3.4–23.6nmol/L (n= 32)0.15–4.1 nmol/L (n = 40)
Correlation (r)0.893420.84598

Modification of the currently used assay to measure salivary cortisol provides comparable results to the commercially available salivary cortisol assay.

  1. Raff et al., JCEM 83(8): 2681.
  2. Tunn et al, Clin Chem. 1992;38:1491–94.


LG Sampson, R Rappel, G Ward, D Jessop, P Hickman, L Price Dept. Chemical Pathology. QHPS-PAH, Brisbane, QLD.

Introduction Two different renin assays were compared for ease of use and correlation of patient results. The assays examined were an activity based plasma renin activity (PRA) assay and a mass based direct renin method using the Nichols Institute immunoassay analyser, the Advantage®.

Methods PRA was measured using the The Diasorin Clinical Assays™ GammaCoat™ [I125] Plasma Renin Activity Radioimmunoassay Kit and direct renin by Nichols Institute immunochemiluminescent (ICMA) Direct Renin method on the Advantage®. Sample, time and technical requirements of each method were examined. Ninety patient samples, including those from several fludrocortisone suppression tests (FST) and renal vein samplings (RVS), were assayed by both methods. Results were assessed by linear regression analysis.

Results The PRA assay requires 1mL EDTA plasma compared to 200μl for direct renin. Total incubation time for the PRA assay is 4.5 hours and includes several technically demanding intermediary steps. This compares with the direct renin incubation time of 40 minutes and availability of results within 1 hour of placing a specimen on the analyser. An on board dilution system allows a 1:10 dilution of specimens with high renin levels. The correlation coefficient (r2) was 0.9404 for specimens with PRA values <170 ng/mL/hr and 0.8 for those with PRA values between 0.15 and 10 ng/mL/hr. No discordant direct renin and PRA results were revealed for specimens from FSTs and RVRs, thereby confirming the clinical interpretation of these challenge tests.

Conclusions The Nichols Institute (ICMA) direct renin method on the Nichols Advantage® is easy to use, requires lower sample volume, and has a shorter turn around time than the PRA assay. The direct renin values correlated well with PRA results in a variety of clinical situations. Comparison of data for plasma pools suggests a possible conversion equation of: Direct renin (mU/L) = PRA (ng/mL/hr) x 10.


LG Sampson, R Rappel, G Ward, N Avsenev, P Hickman, L Price Dept. Chemical Pathology. QHPS-PAH, Brisbane, QLD.

Introduction The plasma renin activity (PRA) assay estimates renin activity while the direct renin assay measures concentration of the renin molecule. Precision of the Diasorin Clinical Assays™ GammaCoat™ [I125] Plasma Renin Activity Radioimmunoassay (RIA) Kit was compared with the Nichols Institute immunochemiluminescent (ICMA) direct renin method on the Advantage®.

Methods Biorad Lyphochek® Trilevel Hypertension (HT) Markers Controls were analysed for PRA on routine PRA runs over several months. Nichols Advantage Direct Renin Chemiluminescent Controls were assayed for direct renin once a day for 16 days. Laboratory prepared plasma pools of 4 different renin values were assayed by the Advantage® twice a day for 16 days and on routine PRA runs.

Results The PRA values of the Biorad HT controls ranged in value from 2 to 25 ng/mL/hr with coefficients of variation (CV) 18% to 14% respectively (n=92). Direct renin concentrations of the Nichols QC were between 7.8 and 194 mU/L with CV's of 10% to 3% (n=16). The mean direct renin levels of the plasma pools were 8, 32, 51 and 530 mU/L with CV's at 13, 5, 4 and 4% respectively (n=30). PRA values for these plasma pools were 0.58, 3.3, 6.6 and >50 ng/mL/hr with CV's ranging from 24% (n=10) for Pool 1 to 5% for Pool 3 (n=5).

Conclusions Precision of the Nichols Institute (ICMA) direct renin method on the Advantage® is acceptable over a wide range of direct renin concentrations. On-board dilution of 1:10 is available for very high renin levels.


LG Sampson, G Ward, R Rappel, B Sipinkoski, P Hickman, L Price, Dept. Chemical Pathology. QHPS-PAH, Brisbane, QLD.

Introduction Suggested changes (1) to improve precision and sensitivity of the plasma renin activity (PRA) method procedure were investigated These changes included extension of the incubation time at 37ºC for patient samples with low PRA levels and the elimination of the blank tube (4ºC angiotensin I generation tube).

Methods Plasma pools with low PRA values were incubated at 37ºC for both 90 minutes and 18 hours at pH 6.0 in the presence of ACE inhibitors to generate angiotensin I. Extended incubation at 4ºC was not performed. No blank value was included in the calculation of the PRA for the 18 hour incubation. The precision of the two incubation times was compared. The effect of elimination of the blank tube on PRA values for specimens with various PRA levels after a 90 minute incubation was examined.

Results The coefficient of variation (CV) of the extended incubation time of a plasma pool with a PRA value of 0.8 ng/mL/hr was 16% compared to 25% for a 90 minute incubation. The blank value of most specimens was low (<2ng/mL/hr) and bore little relationship to the final PRA level. The blank tube does influence the final result for specimens with low PRA levels by determining whether the PRA is/is not detectable.

Conclusion Extension of the angiotensin I generation incubation time improved analytical precision for specimens with low PRA values. The 90 minute generation time would still be required for specimens with mid to high levels of PRA. Hence, if an extended incubation time were introduced to a routine laboratory, the PRA procedure would become more labour intensive and turn around times for results would increase. Elimination of the blank tube would be feasible only if the extended generation time was introduced for specimens with low PRA values.

1 Jean E. Sealy. Plasma renin activity and plasma prorenin assays. Clin Chem 1991;37:1811–1819


LG Sampson, G Ward, R Rappel, J McArthur, P Hickman, L Price Dept. Chemical Pathology. QHPS-PAH, Brisbane, QLD.

Introduction Source of quality control material and laboratory work practices were assessed for their affect on the observed precision of the plasma renin activity (PRA) assay.

Methods The Diasorin Clinical Assays™ GammaCoat™ [I125] Plasma Renin Activity Radioimmunoassay (RIA) Kit was used to measure PRA. Commercial quality controls (QC) assessed were the Diasorin QC supplied with the PRA kit and the Biorad Lyphochek® Trilevel Hypertension (HT) Markers Control. Plasma pools were prepared by thawing frozen patient EDTA plasma of known PRA values, pooling and re-freezing in aliquots until use. QC were run at the beginning and the end of the RIA stage of the PRA assay to assess effect of QC positioning on precision performance.

Results The mean of a single lot of Diasorin QC varied between 5.5 and 7.0ng/mL/hr between kit batches. For a single kit batch, the coefficient of variation (CV) for this QC ranged from 10–13%. Between kit batches, the CV was 18%. Mean values for the Biorad HT controls ranged from 2 to 25ng/mL/hr with CVs 18 to 14% respectively. Similar precision was found for plasma pools while a Low Plasma Pool with a PRA value <1ng/mL/hr had a CV of 25–30%. Plasma pools were unstable after 4 months. PRA values of QC placed at the beginning of the RIA assay were higher than those at the end. Differences between start & end values ranged from 0.3 to 1.7 ng/mL/hr. Precision of the angiotensin I generation at 37ºC and 4ºC were compared. At three different PRA values, the CV's of the 37ºC angiotensin I generation step were lower than those at 4ºC (9% vs16%).

Conclusion Poor precision was observed over a wide range of PRA values. This imprecision was not affected by source of QC material. The behaviour of the PRA assay is represented by the laboratory practice of placing QC's at both the beginning and end of the RIA assay. The PRA assay procedure is divided into three separate steps, each step contributing to the total imprecision of the method. The poor precision of the 4ºC angiotensin I generation step is a major contributor to the overall precision of the PRA assay.


RJ Czajko, PA Shevenan, RX Davey. Dept of Biochemistry, Royal Melbourne Hospital, Parkville, VIC.

Introduction A number of recent advances in Cyclosporin (CsA) methodology have allowed quantitative analysis on automated general chemistry analysers. We evaluated the performance of the Microgenics CEDIA® Plus CsA (CEDIA) assay on an Olympus AU2700.

Methodology The following investigations were undertaken:

  1. A between run precision study using 4 levels of QC material and one patient sample.
  2. An accuracy study using samples from the CsA International Proficiency Testing (IPT) Program (n = 12). Results were compared with the mean of submitted results for the HPLC method group.
  3. A patient correlation study (n=69) comparing our Abbott Flx CsA Monoclonal (Mono) assay with the CEDIA® Plus assay


  1. Between run precision (n = 15) CEDIA (Mono)
    An external file that holds a picture, illustration, etc.
Object name is cbr24_3ps01f3.jpg
  2. Accuracy study (IPT samples n = 12)
    Regression analysis: CEDIA = 1.05 HPLC - 23 r2 = 0.982
  3. Patient correlation (n = 69)
    Regression analysis: CEDIA = 0.90 Mono - 16 r2 = 0.972

Conclusion Overall, the performance of the Microgenics CEDIA® Plus assay on our Olympus AU2700 was equivalent to that of the Abbott Flx monoclonal assay.


M. Lovrincevic, C. Appleton. Biochemistry Department, Queensland Medical Laboratory, West End. QLD.

Introduction Accurate quantitation of kappa and lambda free light chains (FLC) is important in the diagnosis and monitoring of patients with B-cell tumours.

Methods 128 serum samples were assayed on a Hitahi 917 using the Freelite kit (The Binding Site) to quantitate both kappa free and lambda free LC, and their ratio. Urine specimens from the same patients were investigated by electrophoresis / IFE techniques (SPIFE, Helena Laboratories).

Results 91% (20/22 patients) identified as κFLC's by IFE, were positive for serum κFLC (κ/λ ratio elevated) and 93% (14/15 patients) identified as λFLC's by IFE were positive for serum lFLC (κ/λ ratio decreased). In 55 patients with known serum monoclonal paraproteins but without urine FLC (by IFE), 18% (10/55) were positive for serum κFLC (κ/λ ratio elevated) and 7% (4/55) were positive for serum λFLC (κ/λ ratio decreased). In all cases, the nature of the FLC corresponded with the light chain of the monoclonal paraprotein. In 36 patients not known to have an immunoglobulinopathy and without abnormality by IFE, 6% (2/36) demonstrated mildly raised κFLC (κ/λ ratio elevated).

Discussion The Freelite kit assay appears to be a sensitive and specific automated immunoassay for the detection and quantitation of FLC's in serum and may prove to be clinically valuable where there is difficulty in identifying and quantitating low levels of FLC in urine by conventional IFE techniques.


CG Clark, JP Galligan.

Chemical Pathology, QHPS, Royal Brisbane Hospitals Campus, QLD 4029.

Introduction Pseudohypernatraemia is a phenomenon which is not always recognised by laboratory and clinical staff. Indeed, a recent study which clearly documented the difference in plasma sodium, measured by two ion specific electrode (ISE) methods, made no reference to the role of total protein concentration.

We sought to demonstrate the relationship between total protein concentration levels and the difference between the ISE methods used to estimate plasma sodium in our laboratory and in the Intensive Care Unit.

Method In the Intensive Care Unit, plasma sodium was measured by direct ISE on the ABL725 blood gas analyser (Radiometer Pacific). In the laboratory, sodium was measured by indirect ISE on a DDP Modular system (Roche Diagnostics). 158 sets of results were produced from 86 patients in the Intensive Care Unit. This data was subjected to regression analysis and t-test. Less than 15 minutes elapsed between blood sampling and analysis.

Results There was a highly significant negative correlation (P <0.001) between the differences in sodium levels (DDP-ABL) and the total protein concentration. When results from normal and low protein groups were compared, the difference in sodium results from the two methods was highly significant (P <0.001).

Conclusion Hypoproteinaemia in acutely ill patients is common and clinical staff may be confused and misled by plasma sodium results produced in the laboratory. A recent review showed a calculated theoretical difference of 4.5 mmol/L at 40 g/L of total protein. Staff need to be educated to the fact that analysers in the laboratory normally measure sodium concentration whereas results produced on gas analysers measure sodium activity. Sodium activity is considered to give the physiologically correct value.


CMF Yin, SK Ong, S Saw and S Sethi. Department of Laboratory Medicine, National University Hospital, Singapore.

Bio-Rad Laboratories (USA) recently introduced a new small laboratory high performance liquid chromatography (HPLC) instrument, D-10, for the analysis of glycated haemoglobin (GHB). We examined the performance of the D-10 and compared it to our current Variant II (Bio-Rad Laboratories) method. Using routinely requested primary tube EDTA samples from our outpatient clinics the correlation obtained: D-10 = 0.92(Variant II)+0.55, r=0.99, n=121. The bias plot showed good correlation at values below 10%, however, 0.5 – 1.0% lower results were seen above 10% on D-10. Using haemolysate (prepared with 1.5mL haemolysing reagent and 5mL whole blood sample), as utilized in outpatient clinics, the correlation obtained with both samples types run on D-10: Haemolysate = 0.94(primary tube)+0.5, r=0.997, n=20. No bias was observed at the different percentages tested. Precision was evaluated using two levels of controls and two pools of patient blood. Within day imprecision for the GHB range of 5.4 to 10.6% (n=5) obtained a coefficient of variation (CV) of < 1.0% Between-day imprecision (n=6) obtained a CV <1.5%. Total imprecision was <1.5% (n=11). Linearity studies produced y = 1.007x+0.31, r=0.996, with a recovery of 99 – 108%.

In conclusion, this small instrument showed excellent precision, well within GHB performance criteria of <3.0%. Correlation and linearity were good; the accuracy obtained by this HPLC system lending itself to standardization and traceability to NGSP. No notable difference was seen when comparing the results of a sample collected in a capillary and added to haemolysing buffer or a whole blood EDTA primary tube, indicating comparative performance regardless of sample type. It can run up to 10 samples per rack in 36 minutes offering same session outpatient results, results reported in a trace format alerting laboratorians to the presence of abnormal haemoglobins, and more economic than our current small laboratory instruments for glycated haemoglobin.


HA Morris1, PH Anderson1,2 and BK May2, 1Hanson Institute, Frome Rd, Adelaide, SA, Australia, 5000. 2School of Molecular and Biomedical Science, University of Adelaide, North Tce, Adelaide, SA 5000.

Introduction Critical to an understanding of the control of 1,25-dihydroxyvitamin D activity is a molecular appreciation of the regulation of the expression of three genes, 25-hydroxyvitamin D-1α hydroxylase (CYP27B1), 25-hydroxyvitamin D-24-hydroxylase (CYP24) and vitamin D receptor (VDR). We now report the sensitivity, reproducibility and accuracy of a real-time reverse transcriptase-polymerase chain reaction protocol (Taqman™) for the quantification of mRNA levels for these genes in total RNA extracted from soft and calcified tissues.

Methods Total RNA was isolated from homogenised kidney and bone tissues. 1μg of total RNA was reverse transcribed with MMLV-RTase. The sequence of the primers and fluorogenic probes, were designed for CYP27B1, CYP24 and VDR using Primer Express v1.5. PCR reactions were performed in duplicate for each cDNA sample (ABI PRISM 7700 Sequence Detection System).

Results The sensitivity of the protocol was at least 150 copies of mRNA per reaction. Reproducibility, expressed as the coefficient of variation, ranged between 14 and 30% at the level of approximately 104 copies of mRNA per reaction. Accuracy was estimated at greater that 95% for each of the mRNA species.

Discussion This protocol, utilising fluorogenic probes, allows for the comparison of mRNA levels expressed in low abundance between different tissues and animal treatment protocols in extracted total RNA samples from soft and calcified tissues.


GI Raines, M Cole-Sinclair#, HG Schneider. Clinical Biochemistry and Haematology# Units, The Alfred, Commercial Road, Melbourne, Victoria, 3004

Introduction Patients with light chain amyloidosis frequently receive chemotherapy to limit their light chain production. Currently there is no reliable or sensitive tumour marker for monitoring the required dose or effectiveness of their chemotherapy. We have investigated the usefulness of an immunochemical measurement of serum and urine free light chains in these patients.

Methods Five patients undergoing treatment with Melphalan and Prednisolone for amyloidosis were examined. Where possible, both serum and urine were analysed. The kappa/lambda ratio was measured using FreeLite® (The Binding Site, Birmingham) on the Beckman Coulter IMMAGE® and samples were examined by electrophoresis on agarose gels prepared in house.

Results Each patient showed a disturbance in the kappa/lambda ratio in either or both the serum and urine. Visual inspection of the serum electrophoretogram showed hypogammaglobulinaemia without an obvious paraprotein, although in three cases monoclonal light chains (free or intact) were detected at initial presentation and confirmed by immunofixation.


Measurement of free light chains seems to offer increased sensitivity in these patients compared to conventional methods. Drawbacks of the procedure are frequent retesting of the light chains at different dilutions and limited reagent stability after reconstitution. Further data are required to ascertain whether levels of free light chains change in response to treatment and can lead to better outcomes in this patient group.


K Namdarian#*, A Wootton# and M Dobos* #Pathology Department, Royal Melbourne Hospital and *Laboratory Medicine, School of Medical Sciences, RMIT University, VIC.

Introduction Recently, software to optimise Roche LightCycler primer and fluorescent probe design was released. We attempted to design a robust, routine method for PiS and PiZ Alpha-1-Antitrypsin (AAT) mutation detection using this software.

Methods The Roche Applied Science LightCycler Probe Design Software (v1.0) was used to design primer and probe sets. Selection of candidate sets was based on the battery score (taking into account melting temperature and match and mismatch stability) and lack of cross-complementarity. The best performing reagent sets were then synthesised and tested.


Results Ten fold higher fluorescence measurements and improved allele melting curves were observed with the S assay compared to previous PCR-based methods. In contrast, the Z assay performed less well than the existing methods. Asymmetric primer concentrations were found to maximise fluorescence peak detection in optimised amplification reactions.

Discussion The Roche software was useful for assay design. However, experimental testing of several alternative oligonucleotide sets remains necessary to select the best performing reagents under practical assay conditions.

References Aslanidis C et al Clin Chem 1999; 45: 1872–5.

Ortiz-Pallardo M et al J Mol Med 2000; 78: 212–6.


M Lovrincevic, M. Blakey. Biochem. Dept. Queensland Medical Laboratory, West End. QLD.

Introduction Two landmark studies (DCCT, UKPDS) demonstrated the risk of developing long-term complications of diabetes is closely related to the degree of glycaemic control. Thus, glycohaemoglobin assays must reliably detect a clinically relevant change (via acceptable imprecision) in control. This study investigates the suitability of the recently released IFCC-standardised HbA1c method by Roche, using the Cobas Integra 800 high-throughput system for the routine assay of HbA1c on whole blood samples.

Methods Split sample comparisons were performed using the new Roche immunoassay on a Cobas Integra 800 and the Bio-Rad Variant II HPLC system. Both assays are IFCC-standardised.

Results 308 samples were assayed in parallel on the Integra 800 across three Variant IIs (r2: 0.9621, slope: 1.00, intercept: −.020). Within-run imprecision of patient pools did not exceed 1.0% CV at a mean of 10.9% HbA1c using the Roche assay. Between-run imprecision (CV) of the Roche assay, over 5 days, did not exceed 1.7% CV over a range of 5.4% – 11.0% HbA1c.

Discussion This study shows excellent and clinically desirable between-run coefficients of variations that are significantly lower than 5% recommended by IFCC/AACC, and less than 3% which is suggested as being desirable (1).


1. Colman P., et al. Glycohaemoglobin: a crucial measurement in modern diabetes management. Progress towards standardisation and improved precision of measurement. Med J Aust. 1997; 167: 96.


MW Kemp,1,2 S Klingberg,3 L Lloyd,3 P Marr,4 L Vonthethoff,4 Y Wang2, GAC Murrell,2 DF Murrell.1 1Dept of Dermatology, 2Orthopaedic Research Institute & 4Anatomical Pathology, St George Hospital Campus, UNSW, NSW; 3Dept of Chemical Pathology, QHPS-RBHC, QLD.

Introduction Epidermolysis Bullosa Simplex (EBS) is a blistering disease inherited in an autosomal dominant fashion due to mutations in either the Keratin 5 (K5) or Keratin 14 (K14) gene. Here we report a strikingly mild form of EBS-WC caused by a novel 15-basepair (bp) deletion in 2B-rod domain of K5. The proband was the only member of a 3-generation non-consanguineous family to have EBS.

Methods Genomic DNA isolated from whole blood was used to amplify all exons of K5 and K14 using the polymerase chain reaction (PCR). Bi-directional sequencing of the PCR product was done using ABI PRISM BigDye Terminator cycle sequencing.

Results A novel 15bp deletion in exon 5 of K5 was identified. This deletion results in the loss of 5 amino acids (arginine, asparagine, lysine, leucine & alanine) in positions 429 – 433 inclusive. The normal reading frame of K5 is maintained. This deletion was not found in unaffected relatives or 50 unrelated normal controls.

Discussion This is the first report of a small deletion in K5 causing EBS-WC. As the normal alpha helical conformation of the 2B-rod domain naturally accommodates a 3-residue deletion, it is hypothesised that the effect of this 5-residue deletion may be less disruptive than anticipated as it may be partially tolerated by the alpha helix secondary structure. In addition, the conservation of the K5 reading frame probably helps maintain a relatively normal protein secondary structure. Hence the mild phenotypic expression.

Conclusion This novel 15bp deletion in K5 gives further insight into the complicated disease mechanisms in EBS genotype / phenotype correlations.


N Trisnowati,1,3,4 S Klingberg,2 L Lloyd,2 Y Wang,3 GAC Murrell,3 DF Murrell.4 1Dept of Dermatology, Gadjah Mada Univeristy, Yogyakarta, Indonesia; 2Dept of Chemical Pathology, QHPS-RBHC, QLD; 3Orthopaedic Research Institute & 4Dept of Dermatology, St George Hospital, UNSW, NSW.

Introduction The majority of Epidermolysis Bullosa Simplex (EBS) cases are inherited in an autosomal dominant fashion with a single mutation occurring in either the Keratin 5 (K5) or Keratin 14 (K14) genes. Here we report the occurrence of two mutations, K5-E168D and K14-A413T, in an EBS family with variable phenotypic expression.

Methods Genomic DNA, isolated from whole blood, was collected from 3 generations of a non-consanguineous family affected by EBS. Polymerase chain reaction (PCR) was used to amplify all exons of K5 and K14 genes. Direct sequencing of the PCR products was done using ABI PRISM BigDye Terminator cycle sequencing. Results: DNA sequencing showed 2 mutations, K5-E168D and K14-A413T in the proband and her affected daughter. The two mutations were confirmed by NruI and Cac8I digests, respectively. Family studies showed the K14-A413T in the proband's unaffected mother but neither mutation in three other unaffected relatives nor in 50 unrelated normal controls. This suggests that both mutations are more likely to be pathogenic but possibly co-dominant recessive mutations rather than normal polymorphisms.

Discussion Although the two affected individuals have the same genotype, their phenotypic expression is different. The effect of K14-A413T mutation, previously reported in EBS-Koebner (a mild form of EBS), may be enhanced by the novel K5-E168D to cause severe EBS seen in the proband. However, the mild disease seen in the daughter cannot be explained. Variable gene expression or penetrance may explain these phenotypic differences.


S. Ratnaike, P Thurlow, D Macgregor, V Gurtler, P Martinello and J McLeod* Division of Laboratory Medicine, A&RMC & *Douglas Hocking Research Institute, Geelong Hospital, VIC.

We describe a patient with an isolated high serum ferritin.

Case History

Mrs. EB born 1955 was investigated in April 1998.

Her investigations were normal save for a high serum ferritin of 1416μg/L (Reference Range 15–220μg/L). This remained high over the next 3 years, even when she became a blood donor and her serum iron and transferrin saturation dropped below the reference range. She had no stainable iron on liver biopsy. The common mutations of the HFE gene leading to hereditary haemochromatosis were absent.

A literature search lead to a diagnosis of HHCS cataract syndrome. Family studies: Her brother and both her children had cataracts and raised serum ferritin. Her sister had no cataracts and a normal serum ferritin.

Gene Studies: DNA from Mrs EB, her brother and sister were sequenced. She and her brother were found to have a mutation (C to A at position 36 of the IRE of the L ferritin gene) earlier described to cause HHCS.1

Discussion HHCS is an autosomal dominant disease initially described in 1995.2

L ferritin production is unregulated as the mutation in the IRE prevents its interaction with the iron regulatory protein (IRP). The cataracts are due to deposit of L ferritin. Aside from this the disease seems to cause no untoward illness. There is no increase of stainable iron in the liver and no liver damage has been described. Its recognition as a cause of high serum ferritin would prevent unnecessary biopsy of the liver.


1 Mumford AD et al. Blood 1998; 91:367–8

2 Girelli D et al. Br J Haem 1995; 90:931–34


R MacIsaac*1,2, C Tsalamandris1, S Panagiotopoulos1, T Smith1, D Lowe1, S Ratnaike3, M Jenkins3, M Thomas4, G Jerums1

1Endocrinology, 2Department of Medicine, 3Division of Laboratory Medicine, Austin Health, Vic & 4Diabetes Complication Centre, Baker Institute, VIC.

Serum cystatin C levels and calculation of the Modification of Diet in Renal Disease (MDRD-6) formula (1) have been proposed as good markers of GFR. The aim of this study was to compare several markers of renal function in diabetic patients with a DTPA measurement of GFR (using the Brochner-Mortensen correction) that ranged from hyperfiltration to renal impairment. In a cross-sectional study of 221 consecutive clinic patients who had a DTPA-GFR estimation we measured serum cystatin C and creatinine and calculated MDRD-GFR, Cockcroft-Gault (CG)-GFR, and Creatinine Clearance (CC). The correlation coefficients of various markers of renal function with DTPA-GFRs ranging from 9 to 168 ml/min/1.73m2 together with their Receiver Operator Characteristic (ROC) curve analysis (area under the curve %) and Positive Predictive Value (PPV %) for low and high DTPA-GFR cut offs are shown.

An external file that holds a picture, illustration, etc.
Object name is cbr24_3ps01f4.jpg


Although CG-GFR, CC and serum creatinine proved to be good markers of DTPA-GFR, both serum cystatin C and calculation of the MDRD-GFR formula had the best correlation, ROC characteristics and PPV for predicting DTPA-GFR. In conclusion, cystatin C appears to be the best test for predicting a DTPA-GFR < 60 ml/min/1.73m2 whilst the MDRD-6 formula appears to be the best test for predicting a DTPA-GFR > 130 ml/min/1.73m2.

(1) Levey et al. Ann Intern Med 1999;139:461–470.


RJ Czajko, PA Shevenan, RX Davey.

Dept of Biochemistry, Royal Melbourne Hospital, Parkville, VIC.

Introduction The Olympus AU2700 is a fully automated multi-channel analyser with the ability to pre-dilute samples prior to analysis. Our investigation looked at using the 'on-board' diluting facility to carry out the pre-analytical lysis of whole blood samples. The Microgenics CEDIA® Plus Cyclosporin assay was used for this study.

Methodology Our investigations compared sample pre-treatment carried out automatically (Auto) by the analyser with manual (Manual) sample pre-treatment. They included:

  1. Between run precision (4 level QC material and one patient sample).
  2. Patient correlation (n=72).
  3. Sample correlation using the Cyclosporin International Proficiency Testing (IPT) Program samples (n=12).


  1. Between run precision (n=15): Auto(Manual)
    An external file that holds a picture, illustration, etc.
Object name is cbr24_3ps01f5.jpg
  2. Patient correlation (n=72)
    Regression analysis: Manual = 0.937 Auto - 2.0 r2 = 0.983
  3. Sample correlation IPT (n=12)
    Regression analysis: Manual = 1.02 Auto - 3.1 r2 = 0.998


The 'on-board' diluting facility of the Olympus AU2700 can be utilised for the pre-analytical lysis of whole blood samples, for the Microgenics CEDIAâ Plus Cyclosporin assay. Our procedure showed improved precision and greatly improved test efficiency when compared to manually pre-diluting the samples.


R Rappel1, L Sampson1, G Ward1, PE Hickman1, L Price1, D Jessop1, J McArthur1, D Nicholas2 QHPS, Chemical Pathology, Princess Alexandra Hospital, Brisbane1, Immuno Diagnostics, St Peters, NSW2

Introduction Studies have demonstrated that the 7–84 peptide fragment of Parathormone (PTH) (84 amino acid polypeptide) can interfere with some assays for intact PTH. This fragment is present in greater amounts in patients with chronic renal failure. It is claimed the recently released Nichols Advantage Bio Intact PTH assay is not subject to this interference. In this study we perform an analytical evaluation of the intact and 'Bio PTH' assays of the Nichols Advantage analyser and compare these assays with the DPC Immulite intact PTH assay.

Methods Precision was tested using proprietary controls and frozen patient plasma pools run daily over 30 days. The methods were compared using frozen plasma patient specimens (n=40), including specimens from 23 patients with chronic renal failure, and 3 with known primary hyperparathyroidism.

Results For the Advantage Intact PTH method at levels of 0.9, 8.0, 33.9 pmol/L the interassay CV was 9, 6, and 6% respectively. For the Advantage 'Bio Intact PTH' method at levels of 0.7, 4.9, 19.3 pmol/L, the interassay CV was 13, 2, and 3% respectively. For the Immulite intact PTH method at levels of 5.2 and 38.3 pmol/L, the interassay CV was 5%.

Method comparison by Deming regression analysis:

Advantage intact PTH=1.03 Immulite intact PTH+2.8 (r2=0.95, n=40) Advantage Bio intact PTH= 0.58 Immulite intact PTH+0.03 (r2 =0.78, n=39)

Advantage intact PTH= 2.43 Advantage Bio Intact PTH-1.02 (r2=0.95, n=40)

Discussion Between-assay precision was satisfactory across all three PTH methods. A high correlation was observed between the established intact PTH assays (Nichols and DPC), and between the Nichols 'Bio Intact' and Intact PTH assays, and a lower degree of correlation between the Nichols 'Bio Intact' and DPC Intact PTH.


V Wood and R David, Gippsland Pathology Service, Traralgon, VIC.

Introduction Roche Diagnostics supplies two assays for Troponin T, one of which is labelled as Troponin T Stat with a 9 minute incubation and the other Troponin T with an 18 minute incubation. The decision to select the 9 minute incubation may seem obvious considering the clinical requirement for urgent analysis of this test. However the choice is complicated by the fact that the Modular E170 will only perform the 18 minute assay while the Elecys 1010 and 2010 are capable of running both the 9 or 18 minute assays. Hence the dilemma in multi platform laboratory practices which have to decide between offering a faster turnaround times and consistency in applications across its practice.

Methods Both Roche Troponin assays measure Troponin by a sandwich method, with an electrochemiluminescent marker. The comparison studies were performed on patient samples, some samples were analysed retrospectively others as they arrived for analysis.


Troponin TWithin run Precision
9 minutes18 minutes9 minutes18 minutes

r = 0.997 y = 0.02 + 0.84x

Discussion Our results demonstrate good precision, and correlation between the two kits. This allows for flexible usage of these kits interchangeably between different analytical platforms and also to meet the turnaround time expectations.


Q Lam, G Raines and HG Schneider. Biochemistry Unit, Alfred Pathology Service, Melbourne, VIC.

Introduction Over the past few years, troponin measurement has become increasingly fundamental to clinical decision-making. Lack of a universally recognised standard has resulted in heterogenous assays employing unique antibodies each measuring different troponin forms. Validation of such assays must be based on clinical outcomes. This however does not address assay interferences and despite established protocols designed to minimise these in our teaching hospital laboratory, we continue to see clinically-discordant troponin values which are not confirmed when re-analysed using different assays. In response to false positive results from our current assay (Abbott AxSYM cTnI), we evaluated the DPC Immulite 2000 TnI assay as a means of checking suspect results.

Methods 100 plasma samples, representing a wide range of troponin values, were measured in parallel. Results were interpreted quantitatively and qualitatively based on pre-determined "cut-off" values (AxSYM: 0.6–1.9 ng/ml = significantly detectable, >2.0 ng/ml = AMI likely, Immulite 2000: 0.5 –0.9 ng/ml = significantly detectable, > 1.0 ng/ml = AMI likely). Where possible, discrepant results were resolved using clinical evidence and troponin measurements by a third assay.

Results and Discussion Results were discordant in 33/100. In 6 samples, one assay diagnosed AMI while the other indicated no myocardial damage. 4 were false positive AxSYM and 2 false positive Immulite results. In 27 specimens, one assay showed "significant detectable" troponin while the other did not. Overall, we conclude that based on current cut-offs, the Immulite 2000 assay did not offer sufficient discriminating evidence against suspicious AxSYM troponins.


L Lloyd, D Travers and S Klingberg. Chemical Pathology, QHPS Royal Brisbane Hospital, Brisbane, QLD.

Introduction Tau Transferrin (TTf) is the asialyated form of transferrin found in cerebral spinal fluid (CSF) and inner ear perilymph, but not in blood, nasal or ear secretions. Detection of TTf is used as a simple and sensitive method for the detection of CSF leakage. We have adapted the immunofixation protocols of the Sebia Hydrasys system to rapidly detect TTf.

Methods A CSF pool was concentrated 20-fold and diluted either in normal saline or dilute serum. TTf detection limits for the Sebia IF Hydragels was assessed using the 'BJP' and 'IF' protocols with Dako anti-human transferrin antiserum. The gels were stained using the Sebia Acid Violet protocol. CSF protein quantitation was performed on the Roche Modular with the U/CSF method and transferrin quantitation was performed using the Beckman Immage.

Results Protein and transferrin levels in the CSF concentrate were 2430 and 200 mg/L, respectively. TTf was detectable at total transferrin levels of approximately 25 mg/L and 13 mg/L, respectively for IF and BJP protocols. Diluted serum spiked with concentrated CSF had similar detection limits of TTf for both protocols despite higher total protein loads.

Discussion The BJP protocol is capable of detecting TTf at very low levels of transferrin due to increased sample application time and antiserum incubation time as compared to the IF protocol. Although Sebia markets a specific application for TTf using peroxidase conjugated antiserum, specialised equipment is needed to process the gel, which significantly increases the run time. Our adapted BJP protocol turn-around time is less than 2 hours.

Conclusion The Sebia Hydrasys system using IF Hydragel with BJP protocol allows sensitive and rapid detection of TTf in fluid samples in less than 2 hours.


M Charan and R Bais, Clinical Biocemistry, PaLMS, RNSH, St Leonards, NSW, 2065.

Introduction In January, 2002, responsibility for the analytical performance of glucose meters within our hospital was transferred to Clinical Biochemistry. To monitor and improve performance an external proficiency program was introduced for these instruments.

Methods The manufacturer trained 70 Ward staff to operate the instruments (Accu-chek Advantage, Roche Diagnostics) and to perform basic QC and QA. After the initial training, the external proficiency component of the QA product, Solutions™ (Roche Diagnostics, Australia) was introduced. Two glucose samples of various concentration were distributed per month. Ward staff analysed these samples and returned the results to the central laboratory. The results were then entered into the Solutions™ Data Manager and submitted to the RCPA-QAP Pty Ltd for analyses.

Results Allowable limits of performance, data analyses and report formats from the RCAP-QAP were the same as for the general chemistry program for glucose. Every four months, a cumulative report was returned to the Wards with an explanatory letter. Over the past two years, the rate of results being returned averaged 75%. Less than 20% of participating wards have not had at least one result outside allowable limits. Since commencing this program, the p value for all results, a measure of results outside allowable limits, has not changed. The CV for all results ranged from 6% to 16% and have also not changed over the survey.

Discussion The results show there was significant variation in performance between locations. Furthermore, an external proficiency program alone, has not been sufficient to achieve an improvement in acceptable performance. The next phase will be to work with individual sites to achieve acceptable performance.

Acknowledgements This work was partially sponsored by Roche Diagnostics, Australia.


R Rappel1, L Sampson1, G Ward1, PE Hickman1, L Price1, D Jessop1, J McArthur1, D Nicholas2 QHPS, Chemical Pathology, Princess Alexandra Hospital, Brisbane1, Immuno Diagnostics, St Peters, NSW2.

Introduction: The Nichols Advantage System was evaluated to investigate its performance for Adrenocorticotrophic Hormone (ACTH) and Insulin-like Growth Factor-I (IGF-I) assays. Currently these assays are performed in this laboratory on the DPC Immulite analyser (ACTH), and Bioclone radioisotopic assay (IGF-I) involving an acid / alcohol extraction pre-treatment.

Methods Imprecision was tested using Nichols controls once daily, and in-house serum pools twice daily, over 30 days. The Advantage methods were compared with the Immulite assay (ACTH), and RIA assay (IGF-I) using frozen patient specimens (n=45,70) respectively.

Results Imprecision (Interassay CV): ACTH, 4.8, 1.6 and 2.0% at levels of 13, 84, and 390 ng/L (Advantage); 7% and 3% at levels of 35 and 412 ng/L (Immulite); IGF-I, 6% (Advantage) at levels of 7,32 and 68 nmol/L vs. 14, 12, and 10% at levels of 10, 34, and 94nmol/L (Bioclone RIA). Method comparison by Deming regression analysis of Advantage versus Immulite or RIA : Advantage ACTH = 0.84 Immulite + 31.1 (r2=0.96, n=70) and Advantage IGF-I = 1.14 Bioclone - 3.2 (r2 = 0.91, n=45)

Conclusions Imprecision was <5% CV (ACTH) and <10% (IGF-I) for the Advantage assays as compared to <8% (Immulite ACTH) and between 10 and 15% for the manual Bioclone method at similar levels Imprecision for the RIA (Bioclone) assay was greater, probably because the interassay CV was measured over time using multiple reagent kit lots. Correlation was good between the Advantage and the Immulite for ACTH, and acceptable for the Advantage IGF-1 versus Bioclone RIA.

Benefits of using the Advantage system are: savings due to a reduction in operator time, longer reagent shelf life compared with RIA reagents, reduction in need for specialized equipment and more rapid assay turnaround time.


R Rappel, J McArthur, L Sampson, G Ward, PE Hickman, J Potter, L Price, R Mason, B McWhinney, QHPS, Princess Alexandra Hospital, Woolloongabba, QLD.

Introduction Spironolactone and Canrenone have been shown to interfere positively or negatively with the Digoxin assay of various immunoassay analysers, (Steimer et al1). Additionally, most of these analysers displayed varying degrees of positive interference in Digoxin-free serum when either of these compounds was present. In this study we assess the effect of these compounds on the Advia Centaur Digoxin assay, as information on this method has not been published to date.

Methods Stock solutions of Spironolactone and Canrenone were made up in ethanol to the following concentrations: 2800 mg/L, 700 mg/L, 175 mg/L, 43.8 mg/L and 10.9 mg/L. These solutions were then spiked into three separate human serum pools to give final concentrations of 25.2 mg/L, 6.3 mg/L, 1.58 mg/L, 0.39 mg/L, and 0.10 mg/L within each pool. The three serum pools contained Digoxin at concentrations of 0, 2.7 and 2.0μg/L.

Results When Spironolactone was added at 5 different concentrations to drug-free serum to give final levels of 0–25 mg/L, the apparent Digoxin concentration was less than 0.04μg/L, similarly with Canrenone the apparent Digoxin was <0.02μg/L. When Spironolactone was added to the pooled serum with a Digoxin concentration of 2.7μg/L, no interference was observed. Canrenone, as well, showed no interference to the Digoxin result when added in 5 different levels to the serum pool with a Digoxin concentration of 2.0μg/L.

Discussion The Centaur method is shown here not to be subject to positive or negative interference by Spironolactone or Canrenone at the levels usually seen in a clinical laboratory. Furthermore, the presence of Spironolactone and Canrenone do not produce false positive results in Digoxin-free serum with this method.

Reference: 1. Steimer W, Müller C, Eber B. Digoxin Assays: Frequent, Substantial, and Potentially Dangerous Interference by Spironolactone, Canrenone, and Other Steroids. Clin Chem 2002; 48:3: 507–516


F Gray, A Dear, K Byron, L Paiman. Dept. of Molecular Biology, Gribbles Pathology, Clayton, VIC. 3168

Introduction Warfarin is the most frequently prescribed anticoagulant in Australia. However, treatment with this oral anticoagulant is problematic, with the risk of serious haemorrhage being the most potentially dangerous side effect. Cytochrome P450 CYP2C9 is the principal enzyme that metabolises warfarin. Two allelic variants of this gene (CYP2C9*2 and CYP2C9*3) have been well documented. Individuals with these variants have a reduced capacity to metabolise warfarin. Our study aimed to assess the incidence of CYP2C9 allelic variants in an Australian patient population undergoing routine laboratory warfarin dosing and to correlate these findings with dose requirements.

Methods 120 de-identified patients, undergoing routine INR blood testing for Warfarin dosing, were selected for this study. Patients were classified as being either 'low-dose' (Warfarin 1.5 mg. or less) or 'normal-dose' (Warfarin 2.5 mg. or more). Genotypes were determined by standard real-time PCR methods.

Results The frequency of *2 and *3 alleles was found to be similar to that previously documented. In the 'low-dose' group 21/60 (35%) patients were found to be heterozygous for the *2 allele, while 19/60 (32%) patients were heterozygous and 5/60 (8%) homozygous for the *3 allele. In the 'normal-dose' group, 10/60 (17%) were heterozygous for the *2 allele, and only 4/60 (7%) were heterozygous for the *3 allele.

Discussion Our results suggest that CYP2C9 genotyping should be included in the initial assessment of patients commencing warfarin. This measure should reduce clinical difficulties associated with the induction of therapy.


R Scarff1, A Musk1, P Sheehan1, J Hill2, K Smith1.1Department of Core Clinical Pathology and Biochemistry, Royal Perth Hospital, WA, 6000.2National Women's Hospital, Claude Rd Auckland, NZ.

Introduction The concentration of β-Trace Protein (βTP) is approximately 30 times higher in cerebrospinal fluid (CSF) than in serum. It is not present in lacrimal fluid or nasal secretions. Thus, βTP is potentially an ideal marker for the detection of CSF in samples from patients suspected of CSF liquorrhoea. A fast and reliable screening test can aid in the diagnosis of this condition and action can be taken to prevent bacterial meningitis, a life threatening disease. We present a comparison of a βTP and β2-Transferrin (β2T) assay.

Methods 10 nasal secretion samples were from patients presenting with possible CSF leakage. Another 17 samples were from patients who had had a caesarean section performed under spinal or combined spinal-epidural anaesthesia with a fluid leak from the epidural site. βTP was measured by the N Latex βTP assay on the Dade Behring Nephelometer II. The presence of β2T was identified by immunoblotting.

Results The βTP concentration ranged from <0.05 to 34.8 mg/L. There was concordance in 25 of the 27 samples. The 2 discordant results, both with βTP of 6 mg/L initially had no detectable β2T, but with repeat immunoblotting of transferrin showed clearly detectable β2T.

Conclusion The nephelometric assay of βTP appears to be a useful alternative to the detection of β2T for the identification of CSF liquorrhoea. There is good agreement in this study between the βTP and β2T results. The measurement of βTP is rapid and precise whereas the β2T assay is subjective, labour intensive, time consuming and operator dependent.


*KA Sikaris, AR McNeil, G Wilcox. *Melbourne Pathology, Collingwood and Dorevitch Pathology, Heidelberg VIC.

Introduction The standard lactose tolerance test (LTT) uses a 50g oral lactose load and plasma glucose measurements between 0 and 120 minutes. Many laboratories use 1964 criteria which state that a normal response is a glucose rise > 1.0 mmol/L.

Method We reviewed 473 LTTs performed at Dorevitch Pathology over ten years 1993–2002.

Results Seventy percent (332/473) of the LTTs were performed on women, 50% aged 20–40 years. LTT was the only test requested in 36% of patients. The most common associated tests were for coeliac disease (34%). The median maximal rise in plasma glucose was 1.6 mmol/L with the peak value at 15, 30 or 45 minutes in 83% of cases. The glucose increase failed to exceed 1.0 mmol/L in 34% (162/473). There were no differences in the sex or age distributions of lactose tolerant and intolerant patients. The reference interval for the maximal LTT glucose rise was 1.0–2.7 mmol/L (Bhattacharya technique). Eight subjects had findings consistent with diabetes (fasting glucose > 6.9 mmol/L). Two of these patients had abnormal LTT responses. The five patients with Asian surnames in this series had abnormal LTT results.

Discussion Most LTTs were performed on young women, possibly reflecting the prevalence of suggestive symptoms in this group. Our findings support the current diagnostic cut-off of 1.1 mmol/L although correlation with longitudinal studies and other diagnostic techniques would be of value.


R Greaves1, RW Hunt2, J Frederiksen1 1Complex Biochemistry, Women's and Children's Health (WCH), 2Department of Neonatology, WCH, VIC.

Introduction Steroid hormone measurement is often complex due to the similarity in structure of the hormones. The C18 steroids, which include oestrone (E1), oestradiol (E2) and oestriol (E3) are no exception differing by the addition of one hydroxyl group between each step. This similarity makes assay specificity difficult for many immunoassays and often relies on the relatively small amounts of the alternative oestrogens present. Here we describe two oestradiol assays that demonstrate between method concordant results in one group of patients whilst discordant levels in a neonatal study population.

Method Two groups of patients were analysed for oestradiol by both RIA (Diasorin) and the Immulite/Immulite 2000 (DPC).

Group 1a - 88 paediatric and adult patients originally selected for the method comparison of the E2 RIA compared to the Immulite.

Group 1b - 67 paediatric and adult patients selected for the method comparison between the Immulite and the Immulite 2000.

Group 2 - 99 premature neonates participating in a study were assayed both on the Immulite 2000 and by RIA.

Results Using the Passing-Bablok Agreement plot, the method comparisons for each group demonstrated the following:

Group 1a - Immulite = (RIA x 1.28) + 0.6, r=0.954.

Group 1b - Immulite 2000 = (Immulite x 0.93) −37.7, r=0.991

Group 2 - Immulite 2000 = (RIA x 1.31) + 114.1, r=0.86, ie data from preterm infants born less than 30 weeks gestation. Blood was collected at delivery and then weekly for the first six weeks of life.

Discussion The discordant neonatal results may be associated with cross-reactivity from closely related steroids such as E3. The Diasorin RIA assay appears to offer improved specificity over the Immulites. In a specialised laboratory quantitation by GCMS may be performed to confirm the specificity of the working method.

This work is supported by Biomediq DPC & NHMRC.


GRD Jones Chemical Pathology, St Vincent's Hospital, Darlinghurst, NSW, 2010.

Introduction: Westgard multi-rules are commonly used to provide optimal error detection in assay systems. The Westgard website ( provides graphs showing the power of error detection (PED) for various combinations of rules and numbers of QC samples (N). These graphs are known as Power Function Charts.

Hypothesis: The change in bias in an assay which standard Westgard multi-rules can detect with 90% certainty is described in the website as about 2.0 times the SD of the assay. The 10x and 4.1s rules (cross-run rules) are important in this claimed error detection. If only 2 samples are used in each QC run, these rules require data from previous QC runs. However if a change in assay bias has existed during the time of those previous QC runs, other rules (eg 1.3s and 2.2s) may have already detected the shift. My hypothesis is that excluding the contribution of the cross-run rules, when the error would have been detected on a previous QC run, will change the bias which can be reliably detected.

Methods Power Function Charts were produced using an Excel spreadsheet with QC results simulated using a random number generator with a normal distribution. Changes in bias were modelled by adding various constants to the output. QC rules were evaluated by the frequency with which they were triggered at various changes in bias. Standard Westgard rules with N=2 were evaluated for bias detection.

Results When these previously-detected shifts are removed from the data the assay bias which can be detected with 90% certainty is increased from about 2.0 to about 3.3 times the assay SD. The 10x and 4.1s rules add little to the overall PED with 2 QC samples per run.

Conclusion The power function charts on the Westgard website under-estimate the change in bias which can be detected with 90% certainty when the rules include results from previous QC runs.


GRD Jones Chemical Pathology, St Vincent's Hospital, Darlinghurst, NSW, 2010.

Introduction: Westgard multi-rules are commonly incorporated into commercial QC software packages. Using published data ( it is possible to predict the power of error detection (PED) of the rules in use. This data assumes that the actual standard deviations of the method (ASD) is inserted in the QC package. On occasions laboratories use an SD for QC limits (QCSD) larger than the ASD. In this poster I evaluate the effect of using QCSD larger than the ASD on the PED of Westgard multi-rules using Power Function Charts.

Methods: Power Function Charts were produced using an Excel spreadsheet with QC results simulated using a random number generator. Changes in bias were modelled by adding various constants to the output. QC rules were evaluated by the frequency with which they were triggered at various changes in bias. Individual and combined rules were evaluated using N=2.

Results: The effect of setting QCSD>ASD affects the individual components of multi-rules in different ways. There is no change in the shift which can be reliably detected (expressed as a multiple of the ASD) with the 10x rule as QCSD/ASD increases. However the bias which can be reliably detected with the 4.1s, 2.2s and 1.3s rules increase at different rates as QCSD/ASD increases. The effect of changing QCSD on multi-rules appears initially to be small, however with increasing QCSD/ASD, the error detection becomes more dependent on rules which require more than one QC run, potentially delaying error detection.

Conclusion: The determination of the PED of Westgard rules using published data depends on setting the QCSD equal to the ASD. If this criteria is not met the error which can be reliably detected must be determined from other sources.


J Joseph1, EM Lim1, E Rossi1, B De Boer 2, W D Reed3, GP Jeffrey4. Clinical Biochemistry 1 and Anatomical Pathology 2 PathCentre; Hollywood Specialist Centre3, and Department of Medicine, University of Western Australia 4, all at Nedlands, WA, 6009.

Introduction The compound heterozygous [CY/HD] HFE genotype is considered a risk factor for iron loading, however the proportion of CY/HD subjects who develop hepatic iron loading is unknown.

Methods HFE genotyping was performed on 246 patients referred to the Hepatology Clinic of a tertiary hospital for further assessment of elevated serum iron indices. Liver biopsy including quantitative liver iron was performed on 19 CY/HD patients whose liver histology was also re-staged for iron staining, fibrosis and steatosis by a single pathologist.

Results Of the 246 patients genotyped for HFE, 66% were C282Y homozygous, 9% were CY/HD and the remainder were either C282Y heterozygous or wild-type for C282Y. Of the 19 CY/HD patients, mild iron overload, defined as an hepatic iron concentration between 30 and 100 mmol/kg dry weight, was present in 18 patients. One patient had a hepatic iron concentration in the reference range. None had an hepatic iron index which exceeded the published cut-off of 1.9 for hereditary haemochromatosis. The median and range of the hepatic iron concentrations were 58 [23–96] mmol/kg dry weight. Additional risk factors predisposing to hepatic iron loading were recorded in 18/19 CY/HD patients.

Conclusions Although additional risk factors were present in our CY/HD patients, liver biopsy iron levels were in the range defined as mild iron overload.


C Macaulay, J Joseph, E Rossi. Clinical Biochemistry, PathCentre, Nedlands, WA, 6009.

Introduction The NHMRC reduced the level of concern for blood lead in all Australians to 10mg/dL in 1993, however there is particular concern for pre-school age children who are at greatest risk of intellectual impairment.

Methods Pre-school children aged up to 6 years were sampled for blood lead levels in two community based surveys. Lead levels were determined in 1993 in 120 children of mean age 3.3 y from an established coastal suburb and in 2003 in 86 children of mean age 3.7 y from a more recently developed suburb located 18 km inland. For both surveys, blood lead was analysed by furnace atomic absorption spectrometry [Varian SpectrAA300] calibrated with porcine blood based standards.

Results Blood lead levels declined from a mean [SD] of 7.6 [3.2] in 1993 to 3.1 [1.8] mg/dL in 2003. The number of children with blood lead levels of 10 mg/dL and above were 25/120 [21%] in 1993 and 1/86 [1%] in 2003. Results from external quality assurance schemes in both 1993 and 2003 were close to the target result and there was no bias. Possible reasons for the decline in lead levels include:

  1. Lead in petrol was progressively phased out during the 10 year gap between studies and sales ceased in January 2000.
  2. The proportion of housing built pre-1970 and possibly containing lead based paint was much higher in the coastal suburb compared to the inland suburb which was developed post 1980.

Conclusion Although the surveys were conducted in different suburbs, these results are very encouraging, and probably reflect a real decline in mean lead levels in pre-school children


AR McNeil and P Panaya. Dorevitch Pathology, Heidelberg, VIC 3084

Introduction Hyponatraemia is common in people with Guillain-Barré syndrome and has been attributed to pseudohyponatraemia caused by treatment with intravenous gamma globulin (IVGG).

Methods We tested this hypothesis by adding IVGG (Intragam-P from the Australian Red Cross Blood Service, Victoria) to serum in vitro in proportions similar to those one would expect if a dose of 0.4g/kg was given to a person with plasma volume 40mls/kg. We also reviewed the association between hyponatraemia and high serum protein concentrations in our laboratory's database.

Results The addition of IVGG produced the same decrease in serum sodium concentration as the addition of equal volumes of water. IVGG produced a small decrease in protein concentration and an increase in osmolality consistent with the protein (60g/L) and maltose (100g/L) concentrations of this solution. We reviewed 374495 patient episodes from our laboratory database that had concurrent serum sodium and protein measurements. There were 130 results from 82 people with protein concentrations ≥100g/L, nine of whom (11%) had serum sodium <130mmol/L. None of these patients had serum osmolality measurements to investigate the possibility of pseudohyponatraemia. There was no correlation between serum sodium and protein concentrations.

Discussion Hyponatraemia is common in patients with marked increases in protein concentration although this may not be caused by pseudohyponatraemia. IVGG does not cause increased protein concentration or pseudohyponatraemia when added to serum in vitro. Hyponatraemia in people given IVGG is probably caused by dilution or maltose-induced fluid shifts.


J Beilby1, E Rossi1, J Hung2, M Knuiman3, M Divitini3

Clinical Biochemistry1, PathCentre, Departments of Medicine2 and Public Health3, University of Western Australia2 WA.

Introduction There is clear evidence that low folate levels increase the risk of colorectal cancer but the possible role of folate in prostate cancer has not been examined. The aim of this study was to determine if low folate levels in men who participated in a 1969 community survey predicted an increased risk of prostate cancer.

Methods This study was a prospective cohort study over 29 years in Busselton, Western Australia of 962 men aged 20 to 90 years, who were alive more than 3 years after their participation in the 1969 Busselton Health survey. Folate was measured by an automated microbiological method. The adjusted hazard ratio for death from prostate cancer according to baseline levels of serum and red blood cell folate was determined.

Results A total of 31 deaths from prostate cancer were recorded over a 29 year period in the Busselton cohort. After adjustment for age, smoking, alcohol and body mass index, Cox proportional hazards regression analysis found an independent association between serum folate concentrations in 1969 and the risk of subsequent death from prostate cancer. When analysed by quartile, men with the lowest serum folate levels, less than 3.0 μg/L, had an adjusted hazard ratio of 4.79 [1.56, 14.43] for subsequent death from prostate cancer. The overall adjusted hazard ratio for increased serum folate levels across all quartiles gave a reduced risk of 0.64 [0.42, 0.95]. A similar trend was seen with red blood cell folate, but the reduced risk did not achieve significance 0.78 [0.57, 1.07].

Conclusions These findings provide preliminary evidence that folate may have a role in the development of prostate cancer.


H Fong and M Black. Biochemistry Unit, Alfred Pathology Service, Alfred Hospital, Melbourne VIC.

Introduction Gel separator tubes provide a closed system that allows for collection, transport, processing and sampling of specimens. The absorption of specific drugs (antiepileptic and antidepressants) by gels in separator tubes has been reported. As a result manufacturers developed new gel tubes to avoid the problem. We noticed incomplete separation in patients with a large amount of paraprotein. We subsequently examined the effects of paraproteins on separation in BD Vacutainer SST II tube.

Methods The objective of this study was to identify the effect of paraproteins on the gel barrier formation. BD Vacutainer SST II tubes were evaluated. Samples from seven patients with known paraproteinemia were collected in the tubes. The tubes were centrifuged in a Megalfuge 1.0R, Heraeus Instrument at 3000g at 15°C for 15 minutes.

Results and Discussion Incomplete gel barrier formation was observed with specimens from patients with grossly elevated paraproteins, that were identified as IgG(K), IgG(L) or IgM(K). The extent of gel barrier formation ranged from viscous gel layer to incomplete formation of gel barrier. The poor barrier formation may mislead some to believe the tube is faulty. We conclude gel barrier formation can be adversely affected by the impact of density and viscosity of the plasma/serum.

As serum/plasma gel separator tubes are generally used as the primary sample tube we recommend specimens of known or suspected paraproteinemia should be collected in tubes without gel.


C Chin, G Arscott, T Nguyen and J Beilby. Clinical Biochemistry, PathCentre, Nedlands, WA 6009

Introduction Homozygosity of the C282Y HFE mutation accounts for approximately 89% of patients with Hereditary Haemochromatosis (HH). Two other mutations, H63D and S65C, usually accompanied by C282Y compound heterozygosity, are responsible for a mild form of HH. The LightCycler (Roche) using fluorescent hybridization probes enables the real-time detection of both the C282Y(G845A) and H63D(C187G) mutations. The S65C (A193T) genotype, if present, will be detected along with the H63D mutation.

Methods HFE genotyping for the C282Y and H63D mutations were done simultaneously on the LightCycler. The multiplexed reaction contained two sets of primers and LC-Red 640 labelled probes. C282Y and H63D mutations were determined by melting curve analysis.

Results Melting curve temperatures (Tm) differentiated all 4 alleles of C282Y and H63D. In this assay, the 'T' allele of the S65C mutation overlapped the 'A' mutant allele of the C282Y mutation as the difference in Tm was only 1.1oC. This caused C282Y heterozygous subjects to be misclassified as homozygous wild type. Approximately 2% of samples requested for HFE genotype carry the S65C mutation and 20% of these are heterozygous for the C282Y mutation.

Conclusion Multiplexing with fluorescent hybridisation probes provided rapid and sensitive analysis of the HFE gene mutations. However, samples that are heterozygous for the C282Y mutation and carry the 'T' allele of S65C mutation may be genotyped as wild type C282Y. This demonstrates that multiplexed genotyping using probes labelled with the same fluorescent dye has the potential to produce errors.


N Al Hafid, V Poulos. Department of Clinical Biochemistry, RPAH, Camperdown NSW 2050

Introduction Porphyrias are generally a group of inborn errors of haem metabolism which may present with neurological symptoms and/or as skin photosensitivity. Screening of patients by conventional means (fluorimetry, ion exchange chromatography and HPLC methods for porphyrins, precursors and enzyme assays) can usually identify patients with overt disease. However, with these methods the detection of carriers with the disease who are not currently exhibiting symptoms, can be difficult if not impossible. DNA based genetic studies of porphyria patients are available, however these are generally research based studies being both time consuming and relatively expensive to perform.

Method Starting with patients with Acute Intermittent Porphyria (AIP), we are following the method described by A. De Siervi et al. (1999). The method is based on long chain PCR amplification of the entire HMB synthase gene (abnormal in AIP) from genomic DNA in two large overlapping fragments. Following purification of the PCR products sequencing is carried out to detect mutations. After establishing the mutation for the affected patient with AIP, we will then employ denaturing HPLC of the appropriate PCR fragments to screen other family members.

Results In preliminary work we have obtained primers for long chain amplification of genomic DNA of the HMB synthase gene in the form of two fragments 4.5kb and 5.5kb products. We have also obtained 13 primers to be used for sequencing the full 15 exons on the long chain PCR fragments.

Conclusions Although we have only preliminary results, theoretical considerations suggest that long chain PCR followed by sequencing and denaturing HPLC may be useful in establishing a screening regime for patients with AIP and perhaps for screening all genetic forms porphyria.

Articles from The Clinical Biochemist Reviews are provided here courtesy of The Australian Association of Clinical Biochemists