Proceedings of the Australasian Association of Clinical Biochemists’ 47th Annual Scientific Conference

. 2009 Nov;30(4):S7–S47.

Proceedings of the Australasian Association of Clinical Biochemists’ 47th Annual Scientific Conference

PMCID: PMC2791774

CBR. 2009 Nov;30(4):S7.

S1 Experiences with Heart and Brain Biomarkers

The first substance in the blood measured to reflect damage to the heart was the enzyme ALT in the 1950s. Subsequently, the enzyme creatine kinase and its iso-enzyme CK-MB were found to be more specific. With the advent of immunoassays and then the ability to develop monoclonal antibodies early in the 1980s, alternate approaches were possible. The development of the monoclonal antibody, Conan-MB and a rapid assay for CK-MB will be described, as will the development and validation of the assay of Troponin-I for assessing cardiac damage.

The use of various cardiac biomarkers evolved empirically. We found that an unbiased discovery approach (gene expression) predicted the abundance and specificity of the various cardiac biomarkers. We then utilised gene expression chips (Affymetrix) to identify potential markers of brain injury. The experiments were performed with mice in order to get undegraded RNA and then human homologues identified. We identified 17 candidate genes based on abundance and a molecular size of ≥70 kDa, which should be small enough to cross the blood-brain barrier. The specificity of the proteins coded by the genes was confirmed by Western blot and antibody and immunoassay development begun.

The most abundant of the brain genes codes for visinin like protein-1 (VILIP-1). The peak values of VILIP-1 in blood following stroke occur 24–72 hours following onset. Analysis of VILIP-1 in cerebrospinal fluid of patients with dementia of the Alzheimer type (DAT) showed elevation above controls and a strong correlation with Tau protein. We are currently looking at VILIP-1 and some of the other candidate markers in longitudinal samples from relatives of patients with DAT and also in suspected traumatic brain injury.

CBR. 2009 Nov;30(4):S7.

S2 Therapeutic Drug Monitoring: Overview and Update

L Broussard

Therapeutic Drug Monitoring/Management (TDM) has several definitions but typically is considered to be measurement of drug concentrations in biological fluids, usually blood/serum, at defined periods for the purpose of optimising an individual’s dosage to achieve maximum efficacy and minimum adverse reactions. The value of TDM for drugs exhibiting a narrow therapeutic window and a demonstrable dosage/clinical effect was proven as early as the 1950s for monitoring of anticonvulsants. As pharmacokinetic knowledge increased and appropriate analytical tools (e.g. gas chromatography, high performance liquid chromatography, immunoassay, mass spectrometry, tandem mass spectrometry) became available TDM continued to develop and evolve. The scope of TDM has expanded to include monitoring of antiarrhythmics, immunosuppressants, antidepressants, antimicrobials, antiretrovirals, antineoplastics, antiasthmatics, antifungals, and analgesic drugs in addition to anticonvulsants. Of these classes of drugs monitoring of immunosuppressive drugs has become perhaps the most rapidly expanding area.

Practical consideration for those drugs qualifying for TDM includes multiple factors such as patient status (pharmacogenetics, physiological and/or pathological conditions, age, diet, compliance), drug and dosage properties (route of administration, dosage regimen, pharmacokinetics, formulation, drug interactions), preanalytical factors (specimen type, tube, collection time, handling, storage), analyte to measure (prodrug, drug, metabolite, free versus bound, biomarker, surrogate), methodology (availability, specificity, sensitivity, cost-effectiveness, comparative results), establishment of therapeutic ranges (trough, peak, single sample, limited sampling strategy, area under the curve (AUC), population pharmacokinetics), interpretation of results (software, guidelines, decision algorithms), and regulatory/reimbursement issues.

CBR. 2009 Nov;30(4):S7.

S3 Pathologists Overseas

JH Ladenson ^1,²

Pathologists Overseas was founded by Dr. Heinz Hoenecke in 1991 with the mission to help improve pathology and clinical laboratory services in developing countries through the efforts of volunteer pathologists, technologists and laboratory scientists. The experience of Pathologists Overseas in laboratory medicine (clinical pathology) will be reviewed, particularly the countries of Eritrea and Bhutan. The rationale and use of coordinated improvement of laboratory testing and training in clinical management for chronic diseases, such as diabetes, will be emphasised.

The most common laboratory problems observed in our experience over the last 15 years are effective communication, instrument service and training, reagent inventory and supply, and quality assurance.

A new quality assurance training and monitoring program is being developed as a collaborative program by Pathologists Overseas, Inc., Australasian Association of Clinical Biochemists, Royal College of Pathologists of Australasia, Beckman-Coulter Foundation, the Division of Laboratory and Genomic Medicine at Washington University in St. Louis, and the Department of Laboratory Medicine, University of Washington in Seattle.

The quality assurance program will consist of a site visit of local facilities to determine need and scope, followed by three months of analysis of QA samples followed by an on-site training program for which the manual has been prepared. Continuing education modules on DVD are in preparation and the three pilot countries will be Eritrea, Uganda, and Bhutan. Eritrea is now analysing the QA samples; a site visit team will visit Uganda in October-November 2009 and a date for a site visit to Bhutan in early 2010 will be set soon.

CBR. 2009 Nov;30(4):S7–S8.

S4 PoCT in Australia – Implementation and Quality Management, Where are we up to?

R Tirimacco ¹

Anyone can purchase and use a point of care pathology testing device in Australia. There are currently no constraints on the use of such devices. In 2005–06 Department of Health and Ageing (DoHA) on behalf of the Australian Health Ministers' Advisory Council (AHMAC) commissioned the Phillips Fox Report to assess extent of public health risks associated with non-Medicare pathology services including PoCT.

Although the report stated that available evidence did not support unacceptable risk associated with non-Medicare pathology services it did suggest with respect to PoCT that appropriate quality processes should be developed.

The Point of Care Testing (PoCT) in General Practice Trial conducted in 2005–2007 was an Australian Government funded multi-centre, cluster randomised controlled trial to determine the safety, clinical effectiveness and satisfaction of PoCT in General Practice. An interim set of PoCT standards were developed for the trial by the PoCT Implementation Subcommittee of the Quality Use of Pathology Committee of the Pathology Section of DoHA, after consultation with key representatives from relevant scientific and general practice groups. These standards underpinned all quality related activities in the Trial and provided the framework for the Accreditation Program. The trial provided an opportunity to evaluate the effectiveness and relevance of the Standards for use in the general practice environment.

Results of the trial released by the Health Minister in February 2009 reported that the trial framework appears to be acceptable to all stakeholders and provides evidence that non laboratory operators can successfully perform PoCT benefiting the management of chronic disease patients in primary care. These interim standards may be a suitable starting point in defining a quality framework that all PoCT in Australia should be operated within. DoHA is currently consulting with stakeholders on all aspects of the trial results including suitability of interim standards as an appropriate quality framework for PoCT in Australia.

CBR. 2009 Nov;30(4):S8.

S5 Advances in PoCT Technology and Range of Available Tests

C Martin ¹

Introduction

Point of Care Testing (PoCT) has endured varied fortunes during the development of clinical medicine. From being the principal means of testing in medieval times and earlier, point of care testing all but became extinct in hospitals during the latter half of the 20th century, as it fell into the shadow of the mighty automated clinical laboratory. Point of care has undergone a renaissance in recent years, as technologies once reserved for large automated clinical analysers, have become smaller, simpler, more reliable, and more affordable.

Main

Up until the last 15 years point of care testing was largely limited to ward based glucose testing and the use of remote blood gas analysers. The traditional imprecision inherent in strip based glucose meters and the reliability issues of remotely placed lab analysers tended to limit their use to patient maintenance rather than primary care, and to use close to an established laboratory. New emergent technologies combined with increasing cost pressures on health services have made primary diagnosis and treatment by PoCT devices not only possible, but a practical alternative to traditional laboratory services, particularly in the rural and home care setting. In combination with the development of clinical pathway models for patient treatment, this represents a linked paradigm shift in both testing and treatment technologies.

Conclusions

The introduction of new technologies combined with increasing pressures on TAT and cost containment may well signal a new return to prominence for Point of Care Testing in primary patient care which may forever change pathology testing models as we know them.

CBR. 2009 Nov;30(4):S8.

S6 BNP Testing in the Emergency Departmentpanacea or a Waste of Money?

H Schneider ¹

The advent of the natriuretic peptides in the diagnosis and management of heart failure has caused considerable excitement in the Cardiology and the Chemical Pathology community. These markers promise to be of very high diagnostic accuracy and to be powerful predictors of outcome. Subsequent to the introduction into clinical practise they have become widely available increasingly making clinicians aware of issues with the natriuretic peptides. These include that values are not only elevated in heart failure, but also in other conditions and that there are marked age differences as well as changes associated with obesity. The most common clinically used natriuretic peptide assays are for B-type natriuretic peptide (BNP) and for the N-terminal fragment of proBNP (NTproBNP). A number of papers compare the clinical characteristics of the assays and dependent on the diagnostic company one or the other is claimed to be superior. Newer studies have shown that the majority of the natriuretic peptide immunoactivity in the plasma of patients with heart failure is actually a larger precursor and so NTproBNP and BNP might be detecting the same molecule. Future assays for proBNP might improve further the diagnostic potential of the natriuretic peptides.

A Swiss randomised controlled trial used BNP in the diagnosis of patients presenting with shortness of breath to the emergency department and found that the measurement of BNP could reduce hospital admission and shorten hospital length of stay of these patients. Based on the data of this landmark trial and the excellent diagnostic performance of the natriuretic Peptides, the BNP assays gained PBS approval in Australia and are reimbursed at a relatively high price. We have recently published a large randomised controlled trial using BNP in the Emergency Department in two Victorian hospitals. We investigated over 600 patients presenting with severe shortness of breath to two emergency departments and were randomised to receive a BNP test or not. In our trial we did neither see a decrease in hospital admission rates nor a shortening of length of stay. Compared with the Swiss trial the length of stay was significantly shorter although the patient population was somewhat older. We further investigated the impact of BNP on the accuracy of diagnosis in patients presenting with shortness of breath to the Emergency Department. Interestingly, in a randomised controlled setting BNP did not increase the overall accuracy of the diagnosis of heart failure. This is despite excellent diagnostic separation of patients with heart failure from patients without heart failure by the BNP test. This might be due to the fact that while the test has a similar accuracy as the clinical assessment, in patients with a high degree of clinical uncertainty the test does not perform as well.

In summary, from our clinical studies we cannot recommend to use the BNP test in all patients presenting to the Emergency Department with shortness of breath. There might still be sub-groups that will benefit from diagnostic testing and the test has a high degree of diagnostic accuracy and predictive power. Future studies should focus on specific subgroups and patients presenting with a lesser degree of shortness of breath.

CBR. 2009 Nov;30(4):S8–S9.

S7 Analytical and Clinical Performance of Cardiac Troponin Assays

J Tate ¹

Introduction

Global Task Force guidelines for diagnosis of myocardial infarction (MI) and risk stratification of acute coronary syndromes recommend the detection of a rise and/or fall of cardiac troponin (cTn) with at least one value above the 99th percentile of a reference population distribution and a corresponding imprecision of ≤10% coefficient of variation (CV) at this cutoff. Development of troponin assays with improved low-level imprecision is required for conformance to these guidelines.

Methods

Analytical performance of current and future troponin assays for imprecision, limit of detection, 99th percentile concentration, and interferences according to manufacturers’ information, peer-reviewed literature and local laboratory results are described. Assessment of: 1) quarterly local hospital troponin positivity rates; 2) delta change cTnI values (Beckman Coulter AccuTnI) compared with peak levels; and 3) clinical classification of cTnI values between 0.04 μg/L (99th centile) and 0.15 μg/L is discussed.

Results and Discussion

Of 1,711 patient episodes, 39.9% gave positive AccuTnI values and 208 of the 682 patient episodes were between 0.04 and 0.075 μg/L. 79 of these episodes had estimated glomerular filtration rate of <60 mL/min/1.73m2. Using statistical modelling which incorporates assay imprecision profiling and a delta change value calculation for ≥2 cTnI episodes per patient (Z score 2.58), no significant change value was observed for cTnI concentrations between 0.04 and 0.075 μg/L. According to diagnostic manufacturers’ 2008 data, only four of 19 troponin assays met the guideline imprecision requirement of ≤10% CV at the 99th centile.¹ Next generation high sensitivity troponin assays will likely achieve this goal. Their introduction however necessitates that local laboratories collaborate with Emergency Departments to determine optimal clinical decision algorithms that incorporate troponin.

References

1.International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) http://www.ifcc.org/PDF/IFCC_Troponin_Web_Page_Table_of_Assays_Oct_2008.pdf.

CBR. 2009 Nov;30(4):S9.

S8 Markers of Alcohol Tolerance

JB Whitfield ¹

The concept of tolerance to alcohol’s effects is a comparative one in that some people are more tolerant while others are more sensitive. Definitions of tolerance can include short- and long-term outcomes: resistance to intoxication after drinking, presence or absence of adverse reactions to alcohol such as flushing, ability to use alcohol without becoming dependent, and the ability to drink heavily for many years without progressing to liver cirrhosis or other forms of alcohol-induced organ damage. Measures of these processes range from subjective self-reports or physiological measures of intoxication, through quantitative biochemical or haematological markers, to the presence or absence of diagnosed disease. As variation in susceptibility has a genetic basis, genotype at one or many loci can potentially be used to predict susceptibility to alcohol’s effects. This presentation will be an overview of biomarkers and genetic markers in the assessment or prediction of tolerance.

Some aspects of the immediate consequences are well understood; the alcohol flush reaction is explainable through variation in aldehyde dehydrogenase (ALDH2) and alcohol dehydrogenase (ADH1B). Other reactions to alcohol are more common but largely unexplained, and although variation in the degree of intoxication has a genetic basis the genes and polymorphisms responsible are not known. Sensitivity to intoxication is important both in relation to driving impairment and because it affects the probability of excessive drinking and dependence. Effects of alcohol on the liver can be detected in many hazardous drinkers by measurement of GGT and ALT, and these can be used as markers of fatty liver and/or fibrosis to identify genetic effects on susceptibility to alcoholic liver disease.

Study of variation in the effects of alcohol employs a wide range of techniques; understanding the causes of variation requires many approaches but biomarkers and genetic markers are important tools.

CBR. 2009 Nov;30(4):S9.

S9 Severe Glucocorticoid Resistance in a Neonate Presenting with Hypoglycaemia – A Case Presentation

SK McMahon ¹, CJ Pretorius ², JPJ Ungerer ², NJ Salmon ², LS Conwell ¹, JA Batch ¹

Introduction

Resistance to glucocorticoids (OMIM +138040) is a rare inherited disorder of the glucocorticoid receptor α gene (NR3C1), characterised by reduced sensitivity to cortisol signalling and activation of the hypothalamic-pituitary-adrenal (HPA) axis with increased cortisol production in an attempt to compensate for the reduced biological effect at the tissue level. The clinical presentation is variable and patients may present with fatigue or signs of mineralocorticoid or androgen excess. The mutations reported in the GR gene are clustered in the glucocorticoid binding domain and are mostly point mutations with one case of a small deletion in the intron-exon boundary. Most of the reported cases were heterozygotes with a few homozygous cases also reported. Case reports of glucocorticoid resistance without a demonstrable mutation of the glucocorticoid receptor have appeared with a similar phenotype. We report a case with features of hyperandrogenism, increased mineralocorticoid levels and a severe deficiency in the glucocorticoid tissue effects that presented with neonatal hypoglycaemia.

Methods

The clinical diagnosis was confirmed by:

The in vitro response of peripheral blood mononuclear cells to dexamethasone.
The binding of [6, 7 3H] dexamethasone to intact EBV-immortalised lymphocytes.
Sequencing of the glucocorticoid receptor α gene (NR3C1).

Results

The index patient demonstrated no in vitro response to dexamethasone or to the [6, 7 3H] dexamethasone binding assay and the parents and siblings demonstrated an intermediate response. DNA sequencing revealed a 2 base pair (TG) deletion at nucleotide position 2318 that is predicted to cause a frame shift mutation in exon 9 of the GRα gene. Both parents and both siblings were heterozygous for the deletion.

Conclusions

The homozygous mutation in this infant has resulted in a phenotype consistent with complete generalised glucocorticoid resistance. This case is unusual as the presentation included a severe deficiency in glucocorticoid effects with hypoglycaemia.

CBR. 2009 Nov;30(4):S9.

S10 Primary Hyperaldosteronism

M Stowasser ¹

At Greenslopes Hospital Hypertension Unit (GHHU), adopting the policy in 1991 to screen all (and not just hypokalaemic or resistant) hypertensives by aldosterone/renin ratio (ARR) testing led to a tenfold increase in detection rate of primary aldosteronism (PAL) (only 22% hypokalaemic) with a prevalence of 9.5% among referred normokalaemic hypertensives, and a fourfold increase in removal rate of aldosterone-producing adenomas (APAs). PAL is now regarded as the commonest potentially curable and specifically treatable form of hypertension. The Greenslopes/Princess Alexandra Hospital Hypertension Unit series stands at >1400 patients diagnosed (confirmed by fludrocortisone suppression testing [FST]) and >300 APAs removed with cure of hypertension in 50–60% (remainder improved). Reliable detection requires that medications known to alter the ratio are withdrawn before ARR measurement (or their effects taken into account), and reliable methods (such as FST) used to confirm PAL. Because CT frequently misses APAs yet demonstrates non-functioning nodules, adrenal venous sampling (AVS) is the only dependable way to differentiate APA (surgically correctable) from bilateral adrenal hyperplasia (usually treated with aldosterone antagonist medications). For the glucocorticoid-remediable familial form of PAL (familial hyperaldosteronism type I, FH-I), genetic testing for the causative "hybrid" 11beta-hydroxylase/aldosterone synthase gene has greatly facilitated detection. Identification of mutations causing a more common familial variety (FH-II) described by the GHHU in 1991 should further aid detection of PAL. Linkage of FH-II at chromosome 7p22 has been demonstrated in three Australian, one South American and two Italian families. Follow-up of the Framingham cohort has revealed the ARR to be an independent predictor of hypertension development among normotensive participants. Importantly, the ratio showed significant heritability, and a possible role for chromosome 7p21-22, already implicated in FH-II. Clinically indistinguishable from apparently non-familial PAL, FHII could be represented in a substantial proportion of the PAL population, and its further study is therefore a high priority. Sequencing at 7p22 is underway.

CBR. 2009 Nov;30(4):S10.

S11 Insulin Resistance: Recent Advances

P Williams ^1,², T Yen ^1,², S Twigg ^1,²

The condition of prediabetes, defined as impaired glucose tolerance or impaired fasting glucose, is common and recent guidelines have been developed to help in its clinical management.¹ The situation where an isolated increase in blood glucose level (spike) on the 75 g OGTT occurs, with a normal 2h OGTT reading is also quite common and yet its clinical significance is unclear. We examined a population undergoing a 75 g OGTT where 0, 1 and 2h glucose and insulin values were determined. In 17,543 patients, the prevalence of 1h spikers varied from 37% for levels above 8.5 mmol/L, to 14.5% for plasma glucose above 10.0 mmol/L. A spike at 1h greater than 8.5 mmol/L had predictive value to determine those who would progress to IGT. This prediction appeared to be enhanced by the inclusion of 1 and 2 h serum insulin values.

Methods to determine insulin resistance are imprecise. The Gold standards, the euglycaemic insulin clamp and the insulin tolerance test, are unsuitable for general use. Oral glucose tolerance tests have several pitfalls, requiring special dietary preparation and have inherent variability between individuals and in the population. This prompted the ADA to use fasting glucose alone and more recently to a new call for HbA1c to be used as the criteria for diagnosis of diabetes, but this does not allow pre-diabetes to be diagnosed. HOMA and QIKI are ratios relating fasting glucose and Insulin levels as a surrogate for the glucose clamp but are dependent upon insulin values that are not internationally standardised and the results are not universally interchangeable. The insulin values may differ by as much as 50–100% between assay platforms. The ADA and the AACC have recently called for and are defining insulin standards to correct this situation.

References

1.Twigg SM, et al. Med J Aust. 2007;186:461–5. doi: 10.5694/j.1326-5377.2007.tb00998.x. [DOI] [PubMed] [Google Scholar]

CBR. 2009 Nov;30(4):S10.

S12 Re-Aligning Allowable Limits of Performance

K Sikaris ¹

The RCPA/AACB Australasian Quality Assurance Program (AQAP) in Chemical Pathology allows laboratories to assess their performance against their peers. This fulfills the ISO15189 requirement (5.6.4) for interlaboratory comparison which is a relative approach to quality requirements. There is another ISO15189 requirement (5.5.2) that requires examination procedures to be suitable for intended use which is an absolute approach for quality requirements.

Most tests are used for either monitoring (comparing a patient against themselves) or diagnosis (comparing a patient against others). Monitoring is a naturally more difficult task. It can be shown that when comparing a patient against themselves, systematic analytical bias cancels itself out so only imprecision is important. However when comparing against others, bias against reference limits can be an issue therefore both bias and imprecision are important - i.e. total error.

The current AQAP Allowable Limits of Performance were derived by expert consensus and largely based on what can be achieved by Australasian peers. There has been a general focus for EQA organisers to set analytical goals based on what is required clinically. The favoured approach is one based on biological variability. Nevertheless, even when accepting the biological variability approach, should the goals be designed for clinical monitoring (imprecision) or diagnosis (total error)? Furthermore, should they be minimal requirements, optimal requirements or requirements that vary based on current peer capability?

If we decide that total analytical error requirements are required and that they may vary based on current peer capability, we would come full circle to accepting the current peer-relative requirements. However if we decide that there should be an absolute minimum standard for analysis, and this could be defined as the minimum standard for the simplest task of diagnosis, do any or all of our analyses fit such a minimal standard?

CBR. 2009 Nov;30(4):S10.

S13 Quality Issues for Diagnostic Reagent Manufacturers

G Koumantakis ¹

Most laboratories when using a product such as a reagent are probably unaware that the accompanying package insert represent a formal performance undertaking by the manufacturer about this product. Although we tend to take such documentation for granted it is worth noting that any claims made within the package insert are only made based on extensive internal and external investigations by the manufacturer of the product.

The complexity of manufacturing an in vitro diagnostic (IVD) product is significant. The manufacturer must achieve optimum productivity and regulatory compliance with solutions that address manufacturing, regulatory and business needs simultaneously.

To achieve this a manufacturer requires to have well established internal systems that are capable of addressing any issues. Such systems are composed of formal procedures, internal and external quality controls, use of reference material and laboratories and of course very experienced clinical and analytical scientists to ensure the final product meets the required specifications.

The manufacturing of IVD reagents has many complex steps that must be continuously monitored starting with the quality of the raw material right down to the amount of time the reagent can stay on a clinical analyser during the testing process and its mode of disposal. Although assay characteristics such as sensitivity and specificity form a major part of this process it must be appreciated that the true validation of the product will be once it is used by thousands of different laboratories on millions of patients. An open communication channel between the clinical laboratory and the manufacturer allows the final manufacturing step to be completed. Without such an interaction no product could have reached the quality that is currently available to our clinical laboratories around the world.

CBR. 2009 Nov;30(4):S10–S11.

S14 Insight into Transfusion QAP

CE Hughes ¹, NR Herrmann ¹

Introduction

The Transfusion QAP commenced operating in 1969 with one survey per year and 102 participants. This was to become known as the A Program and was the signature program that tested all elements of basic blood banking. In 1983, the program was permanently co-located with the Australian Red Cross Blood Service in Brisbane. In 1991, the B Program was commenced followed by the F Program in 2001 for the assessment of foeto-maternal haemorrhage by Kleihauer and Flow methods. In 2006, the I Program was introduced which gave supervisors an external QAP aimed at assessing and improving individual performance.

New Developments

This year the A Program was replaced by the General Transfusion (GT) Program. This is a web based program containing four modules: General Compatibility, Basic Compatibility, Phenotyping, and Anti-D Titre.

Materials and Scenarios

Transfusion QAPs are unique in that all surveys are prepared from fresh donor blood. Preparation of the survey commences two weeks prior to despatch with the sourcing of appropriate units from the ARCBS Processing Department. Each survey is based on a scenario pre-determined by the Transfusion QAP Advisory Committee. The scenario may be complex giving blood bankers the opportunity to tackle problems not commonly seen in routine practice. Surveys may simulate HDNB, transfusion reaction investigation, bone marrow transplant and massive transfusion.

Conclusions

The General Transfusion Program aims at not only assessing laboratory performance but has a strong educational bias by providing blood bankers with surveys that are outside the normal routine.

CBR. 2009 Nov;30(4):S11.

S15 Insight into the Cytopathology QAP

J Finnimore ¹

The Cytopathology QAP was originally developed in Sydney in 1985 and moved to Brisbane in 1992. The Program has grown considerably since then, now annually issuing 20 pap smears per gynaecological module (2) and 4 non-gynaecological (sputum, body fluids and fine-needle aspirates) cases per module (2). In addition we issue a virtual non-gynaecological case with each glass slide survey. The logistics of offering a glass slide program to over 200 laboratories worldwide are considerable. To prepare a fine needle aspiration specimen suitable for inclusion in the Program (e.g. breast carcinoma) requires a pathologist to procure an appropriate fresh surgical specimen and perform over 100 needle passes into the tissue with a fine gauge needle to prepare the 150 slides required. All slides must be stained and cover-slipped and screened to ensure there is adequate material for assessment and diagnosis. The Program requires a number of these specimens from a variety of body sites, reflecting the routine work of the laboratories enrolled in the Program and providing cases of interest for educational benefit. Cases to be issued to participants must then be selected by the Advisory Committee members. Due to the subjective nature of the discipline this is often a lively multi-header microscope session.

Each gynaecological slide is a unique specimen and the manufacture of cases is not possible. The Program therefore relies entirely on donations for establishing its gynaecological slide library. Every slide lent to the Program must be re-screened to ensure it is a true representation of the diagnostic response. As well, all slides must be re-screened at least every two years to ensure retention of staining quality. The Cytopathology QAP is the only Program which requires return of all its material. Slides may be broken in transit or lost in the mail or returned very late in which case they are not available to re-issue in a subsequent survey. Slides may also be held in quarantine until fees are paid; this adds a significant cost to the operation of the Program.

These many and varied factors make delivery of a QA Program in Cytopathology challenging and rewarding.

CBR. 2009 Nov;30(4):S11.

S16 QAP –Strategic Directions

P Stewart ¹

RCPA QAP is 21 years old this year. The company now generates revenue of over $11 M per annum across all disciplines of pathology. Growth continues, based on increased overseas participation and development of new programmes for the Australasian markets. There are currently over 55 staff working for the company at 10 sites. All programmes are accredited to the relevant standards. There is an independent governing Board that sets company policy.

The Board believes that the company’s traditional market, Australasia now gives little opportunity for growth except in genetics and pre- and post-analytical QA. There is increasing emphasis on development of overseas markets. Penetration has been strong in Hong Kong and particularly in Malaysia. Opportunities also seem likely in India. Company strength is enhanced by the ability to offer programs in all disciplines and in particular Anatomical Pathology. The AACB has contributed to promotional activities internationally and this will remain a high priority for the company. Diversification into education and training are also being explored as business opportunities.

There are now a series of common report formats designed so that cross discipline understanding of QA performance is feasible.

It is less likely in the future that programs will be co-located with pathology laboratories or hospitals. Chemistry was the first program to stand alone. Likewise, Program Chairs are less likely to be located in the same city. The company continues to focus on costs especially in information technology.

It is also seen as high priority for the company to support professional development and career building in Quality and to that end is offering a number of scholarships.

CBR. 2009 Nov;30(4):S11.

S17 Point of Care Testing Working Party

R Tirimacco ¹, R Bais ¹, B Glastonbury ¹, J Gill ¹, B Heffernan ¹, G Koumantakis ¹, C Martin ¹, M Shephard ¹, A St John ¹, V Hauke ¹, L Watkinson ¹, R White ¹

The AACB PoCT Working Party continues to work at increasing its profile as an expert body in all facets of PoCT for both laboratory and non-laboratory operators. The group is well placed to achieve this as it represents all of the major stakeholders in PoCT including private and public pathology laboratories, PoCT Specialists, instrument companies, EQA, NATA and general practice.

Two major PoCT events held in Sydney this year were a Scientific Education Seminar (SES) and a QC/QA workshop. The SES was titled “Importance of Getting it Right”. Speakers had the challenge of highlighting what could go wrong if the wrong instrument is selected, wrong sample is used and if no QC/QA is performed.

This year’s SES gave industry the opportunity to demonstrate PoCT instruments as part of the program. Participants were rotated through industry stands where they were given a demonstration on how equipment works and given opportunity to ask questions. This proved to be a rewarding experience for both registrants and industry.

The QC/QA workshop expertly chaired by Andrew St John commenced with three industry presentations bringing the audience up to date on quality aspects of common PoCT equipment currently in use. The main challenge of the afternoon was to find answers to common questions constantly asked such as: should QC be run, how often should it be run, should QA be run, how often should it be run, who should run QC/QA samples and should the frequency of QC/QA be dependent on the complexity of the PoCT equipment. Participants were asked to fill out a QC/QA survey. Results of the survey will be collated and used to form a PoCT QC/QA policy which will be presented to AACB executive for endorsement.

Development of an AACB QC/QA policy will be a major achievement when accomplished representing the expert opinions of AACB members including industry currently involved in PoCT.

CBR. 2009 Nov;30(4):S11–S12.

S18 Vitamin D Uncovered

GH Beastall ¹

Introduction

The role of vitamin D in calcium homeostasis and metabolic bone disease has been established for many years. The measurement of serum 25-hydroxyvitamin D (25OHD) is recognised as a marker of the nutritional status of vitamin D. Recently, there has been an explosion of interest in the potential role of vitamin D as a protective agent against a wide range of chronic disorders associated with northern climates that see relatively little sunshine.

Methods

The scientific literature may be reviewed in three areas. Firstly, there is evidence suggesting an association between global latitude and the prevalence of chronic diseases. Secondly, there is evidence to suggest that serum 25OHD, serving as a marker of sunshine exposure, is relatively deficient in populations where these chronic diseases are prevalent. Thirdly, vitamin D replacement studies are under way to see if clinical outcomes can be altered.

Results

There is evidence that a relative deficiency of serum 25OHD correlates with the prevalence of chronic disease. But is it cause or effect? Early evidence from vitamin D replacement studies suggest that in some cases raising the serum 25OHD concentration is associated with improved clinical outcomes. However, to achieve this effect the serum 25OHD concentration needs to be higher than that associated with mitigation of the effects of vitamin D deficiency in metabolic bone disease. To complicate matters the inter-laboratory variability in measuring 25OHD is substantial.

Conclusions

There is huge interest in the hypothesis that relative vitamin D deficiency is causative in common chronic diseases. Vitamin D replacement studies and studies on the mechanism of action of active vitamin D will help to clarify the position. In the mean time clinical chemists need to better harmonise the measurement of serum 25OHD.

CBR. 2009 Nov;30(4):S13.

O1 Comparability Assessment of Three Point of Care Analysers for Troponin I Against A Central Laboratory Platform

A Coleman ¹

Introduction

The AACB Point of Care Testing (POCT) Implementation Guide makes a number of recommendations regarding the selection of analysers, including result comparability with local laboratory results. In a large laboratory network, including rural, regional, metropolitan and central laboratories comparability is a key factor. This study assessed the comparability of three POCT devices against a central laboratory platform.

Methods

Lithium heparin patient samples were selected from patients with detectable levels of troponin I by the Centaur TNI Ultra method. These samples were then either analysed neat, or diluted using samples from patients without detectable levels, for troponin I on the following analysers; Siemens Advia Centaur, Immuno Mitsubishi Pathfast, Siemens Stratus CS, and Abbott POC iSTAT.

Results

Results were assessed at the 99^th centile as quoted by the specific manufacturer. 78% of the patient samples (31/40) were defined as positive using the Centaur 99^th centile. Of these the Pathfast provided the greatest sensitivity at 77% (24/31), the Stratus the next at 55% (17/31) and the iSTAT the least at 45% (14/31). None of the POCT analysers gave a ‘false positive’ result. This gave equal sensitivities and positive predictive values of 100% for each analyser, with negative predictive values and efficiencies of 56% and 83% respectively for the Pathfast, 39% and 65% for the Stratus, and 35% and 58% for the iSTAT.

Conclusions

On the basis of the sensitivity of each POCT device, the Pathfast provided the greatest results comparability with the Centaur Ultra TNI method. Laboratories should also look at a range of other parameters when assessing POCT devices as recommended by the AACB POCT Implementation Guidelines.

CBR. 2009 Nov;30(4):S13.

O2 Carbamazepine Interference In Urine Free Cortisol By UPLC

T Harrower ¹, J Creces ¹, G Ward ¹

Introduction

Urinary free cortisol (UFC) measurement is used in the diagnosis of Cushing’s syndrome and techniques commonly used include immunoassay and HPLC. Recently observed was a peak that co-eluted with cortisol by HPLC in a patient with Cushing’s syndrome. Investigation discovered the interfering peak was caused by carbamazepine, and in this paper we discuss how this interference was removed.

Methods

UFC measurement was by Waters UPLC. The samples are extracted using SEP PAK Vac 3cc (500mg) C18 cartridges then passed through a Waters Acquity UPLC BEH 1.7μm, 2.1x10mm C18 column. The mobile phase consisted of 2 solutions; purified 0.2 micron degassed water (A) and 30 % tetrahydrofuran in water (B). A non linear gradient was run using starting conditions of 50% (A) and 50% (B). The sample was injected and the gradient started using a ratio of 30% A and 70% B increasing to 100% B over 3.5 minutes. The gradient was then returned to initial conditions for re-equilibration. A Photodiode Array is used to measure at a wavelength of 247 nm. The initial UFC result pre-transphenoidal surgery was 3171 nmol/ L (RI <150). Post-surgery the UFC result remained elevated at 330 nmol/L due to carbamazepine interference.

Due to the co-eluting peaks the method in use had to be altered to eliminate the interference. The extraction cartridges, C18 column and detector all remained the same. The mobile phase was adjusted to isocratic conditions using 40% (A) and 60% (B), thus altering peak separation and extending the overall run time to 12 minutes. Using this modified method the patient’s true UFC result returned to the normal reference interval with a level of 30 nmol/L.

Results and Conclusions

An existing UFC method was provided to Waters so they could evaluate suitability with UPLC chemistry. However, carbamazepine was not considered or evaluated for the Waters method. The altered gradient successfully eliminated the carbamazepine interference in the cortisol peak, and has now been included as a modified method to use on any patients taking carbamazepine. Carbamazepine interference has been reported in one other HPLC method for UFC.

CBR. 2009 Nov;30(4):S13.

O3 Negative Interference in an Insulin Assay

M McLaren ¹, T Yen ², L Price ¹, G Ward ¹

Introduction

Insulin assays in current use are two-site immunometric assays. Clinical applications include the assessment of insulin resistance and the investigation of hypoglycaemia. Heterophilic antibody interference in two-site immunometric assays is commonly observed as an artefactually elevated analyte level. We report a patient who had negative interference in a two-site immunometric insulin assay.

Methods

Insulin and C-peptide were respectively assayed by the Abbott AxSYM and the Siemens Immulite 2000. Scantibody tubes were used according to the manufacturer’s instructions. Polyethylene glycol (PEG) was added to an equal volume of patient sample, mixed, centrifuged and the supernatant re-assayed for hormones.

Results

A 58 year old male patient was assessed for insulin resistance by oral 75g glucose tolerance test. Insulin levels in all patient samples were undetectable despite the presence of C-peptide. Insulin was detected in patient samples after pre-treatment with Scantibody Tubes but not detected after PEG precipitation (Table).

Table.

Analyte levels before and after Scantibody Tube or PEG Treatment.

Sample	Glucose	C-peptide	Insulin (mU/L)
	mmol/L	nmol/L	Untreated	PEG	Scantibody Tube
0 hour	4.6	0.5	<1	<1	10
1 hour	5.6	2.3	<1	<1	27
2 hour	2.5	0.6	<1	<1	4

Open in a new tab

Negative interference due to heterophilic antibodies is rarely detected or reported. In this patient PEG was ineffective in eliminating interference. The incidence of positive heterophilic antibody interference is approximately 1 in 3000. We have not observed negative interference previously in the AxSYM insulin assay and conservatively the incidence of negative interference is 1 in 20,000.

Conclusions

Negative interference in an insulin assay has the potential to confound the differential diagnosis of hypoglycaemia. Heterophilic antibody interference should be considered whenever discordant insulin and C-peptide results are observed.

CBR. 2009 Nov;30(4):S14.

O4 Comparison of Two Immunoassays and Two LC-MS/MS Methods for 25-Hydroxyvitamin D

D Fredline ¹, V Bell ¹, S Howells ¹, K Young ¹, L Rogers ¹, B McWhinney ², J Galligan ², D Kanowski ¹, L Price ¹, G Ward ¹

Introduction

Assays for measurement of 25-hydroxyvitamin D require both accuracy and precision to ensure correct classification of vitamin D status. Current immunoassay formulations are often modified to meet the clinical and analytical expectations. This paper compares results from two immunoassays and two LC-MS/MS methods.

Methods

Fifty patient samples were assayed for 25-hydroxyvitamin D by DiaSorin Liaison and Roche E170 immunoassays and two independent LC-MS/MS methods (Pathology Queensland (PQTMS); Roche Diagnostics Referral (RDRTMS)). Inter-assay and intra-assay precision for immunoassays was respectively assessed by patient duplicates on two separate analytical runs; and 20 replicates of patient pools on a single analytical run.

Results

Table.

Correlation data for assay comparison by linear regression.

X- Axis	Y-Axis	Slope	Intercept	r²
E170	PQTMS	1.21	−5.1	0.71
E170	RDRTMS	1.05	−5	0.69
RDRTMS	PQTMS	0.78	+5.6	0.8
Liaison	E170	1.16	−23	0.81
PQTMS	Liaison	0.74	+33.4	0.55
RDRTMS	Liaison	0.64	+28.4	0.53

Open in a new tab

The lowest level measurable by the Roche E170 was approximately 25 nmol/L with the inter- and intra-assay precision 6% and 5% respectively. The DiaSorin Liaison had inter-and intra-assay precision of 20% and 14% at 18 nmol/L and 25 nmol/L respectively. For the samples tested the Roche E170 while precise appeared unable to give results <25 nmol/L. The DiaSorin Liaison had interassay CVs on patient duplicates up to 15–20% from 18 nmol/L to 100 nmol/ L. There was no strong correlation between any of the methods assessed.

Conclusions

This data further supports the urgent need for an international reference method and standards for 25-hydroxyvitamin D.

CBR. 2009 Nov;30(4):S14.

O5 The Body Mass Index (BMI) in the First Trimester Screening (FTS) Population

R Sinnadurai ¹, JM Morris ¹, V Tasevski ¹

Introduction

Increasing BMI of the population is an important public health problem. Biochemical markers used in FTS are affected by maternal weight. Maternal BMI is associated with adverse pregnancy outcomes. We assessed changes in BMI over time and compared the BMI of pregnant women in the first trimester to an age matched female population.

Methods

Pregnant maternal weights were collected for six years as part of routine screening. Weights were converted to BMI with the formula weight(kg)/ height(m)² using an average female height of 1.68 metres. BMI data was compared to an age matched non-pregnant population using the National Health survey 2004–05. Individuals were characterised as underweight (<20), normal (20–25), overweight (25–30) and obese (>30) and data expressed as percentage of total.

Results

The percentage of normal, overweight and obese individuals was little changed over time whilst the percentage of underweight individuals increased. In the pregnant population 33% of individuals were either overweight or obese. By comparison, the percentage of either overweight or obese individuals in the age matched non-pregnant population was higher still at 42%. This difference reflects a larger number of overweight individuals in the non-pregnant group compared to the pregnant population, whilst the percentage of obese individuals was similar.

Conclusions

Maternal weight correction of the biochemical component of FTS is important, as a large percentage of women’s final risk would be affected by weight alone. Whilst our pregnant population would be considered heavy it is still however less heavy compared to an age matched non pregnant population. Future monitoring of maternal weight is required as further increases may reasonably be expected.

CBR. 2009 Nov;30(4):S14.

O6 Evaluation of the Roche Diagnostics High Sensitivity Troponin T Assay

R Bais ¹

Introduction

Roche Diagnostics have developed a high sensitivity troponin T (TnTh-s) assay with a limit of detection of 0.003 μg/L compared to the current assay lower limit of 0.01 μg/L. A guideline specified cut-off of 0.03 μg/L (CV≤10%) is the current level used to detect myocyte damage in acute coronary syndrome (ACS) patients. This is 0.013 μg/L for TnT-hs. We have compared the TnT-hs assay with the current TnT assay, determined imprecision, functional sensitivity and estimated clinical implications if the 0.013 μg/L cutoff is adopted.

Methods

Imprecision was determined on two levels of QC over 14 days. To measure functional sensitivity, patient samples were assayed 10 times and the mean and SD calculated for each sample. The mean was then plotted against the %CV, and the data fitted to an exponential curve. The equation was used to determine the 10% CV and the CV at the 0.013 μg/L cutoff. 58 samples (580 assays) were assayed with two different calibrations.

Results

Imprecision (CV) for TnT-hs was 4.7% and 3.2% at 0.03 μg/L and 2.42 μg/L, respectively compared with 11.7% and 3.0 % at 0.016 μg/L and 2.6 μg/L, respectively for TnT. Comparison data in the TnT range 0.0032 to 0.13 μg/L gave an equation of TnT-hs = 0.96TnT + 0.01. The exponential curve fitted to mean vs CV data gave a 10% CV of 0.0033 μg/L and a 3.7% CV at 0.013 μg/L. If the cutoff is lowered from 0.03 μg/L to 0.013 μg/L, the new TnT-hs will report quantitative values in 42% of results previously reported as <0.01 μg/L and an increase of 26% of results being positive.

Conclusions

The TnT-hs assay meets the criteria of CV ≤10% at the 99^th percentile of the reference interval. However, its introduction into clinical practice will result in a significant increase in the number of quantitative troponin results.

CBR. 2009 Nov;30(4):S14–S15.

O7 Emerging Biochemical Risk Markers for Coronary Artery Disease: A Case-Control Study

HV Singh ¹, S Bhandari ², N Singh ³, S Singh ³, A Raizada ²

Introduction

To provide the best treatment for coronary heart disease (CHD) patients, one must go beyond LDL cholesterol to detect factors contributing to CHD risk or existing CHD. These emerging risk factors appear to identify individuals at increased CHD risk and the assessment of these can modulate clinical judgment when making therapeutic decisions. This study focus on five markers lipoprotein(a) (Lp(a)), high sensitive c-reactive protein (hs-CRP), total antioxidant status (TAS), total homocysteine (tHcy) and fibrinogen.

Methods

269 angiographically proven coronary artery disease (CAD) cases were selected for the present study. 127 healthy controls were selected from hospital staff by screening and it was ensured that they are not suffering from hyperlipidaemias, dyslipidaemia or diabetes. Lp(a), hs-CRP, TAS was estimated using Hitachi 912 analyzer. tHcy levels were estimated by Bio- Rad CODA analyzer and fibrinogen levels were estimated by Clauss method using Sysmax Autoanalyzer.

Results

Emerging markers Lp(a), hs-CRP, homocysteine and fibrinogen levels were significantly increased in CAD cases. TAS level was significantly decreased in the CAD group. The tHcy level was significantly increased (p<0.001) in the present study. Oxidative indices like TAS was found decrease or reduced significantly (p<0.001) and it is confirmed by the elevated lipid peroxidation product MDA in both groups.

Conclusions

The correlative and logistic regression analysis shows that these markers are more closely associated with atherosclerosis than traditional risk markers. The screening with these emerging risk markers is helpful in risk assessment of CHD. This additional diagnostic information can assist in the prevention or early diagnosis and treatment of atherosclerotic CHD.

CBR. 2009 Nov;30(4):S15.

O8 Addressing the Workforce Crisis

G Koerbin ^1,², R Smith ³, B Lidbury ²

Introduction

ACT Pathology provides services to Canberra and the surrounding region. Service provision comes under pressure from increasing workload, turnaround time requirements, resignations and retirements and limited ability to attract qualified experienced staff for supervisory and base grade positions. Michael Legg and Associates stated that workforce shortages in pathology are not new and there are difficulties in attracting recruits because suitable education is not available. Also, the cost of medical scientist training is high and not attractive to universities and there is some misleading advice about the appropriateness of certain courses as training for work in pathology. Some courses have an insufficient vocational component and there are too few work placements opportunities. The report also noted that industry need and course content have not always been aligned. ACT Pathology, the University of Canberra (UC) and Siemens Healthcare Diagnostics are collectively attempting to address some of these factors.

Methods

Discussion between senior staff of ACT Pathology, UC and Siemens identified the need to modify the current Biomedical Science course focusing on industry need. ACT Pathology and UC staff have rewritten the curriculum incorporating a work placement program using ACT Pathology and Siemens facilities.

Results

The updated course commenced semester 1 2009. ACT Pathology staff provided lectures and practical supervision. The work placement program is commencing in semester 2. Revamped units in clinical chemistry, haematology, transfusion and histology will commence in 2010.

Conclusions

With laboratory staff interacting with academic staff to provide curriculum development, delivery of lectures and practical sessions and with the direct involvement of the diagnostics industry, it is hoped that graduates will become more workforce ready enhancing their employability and retention in the workforce.

CBR. 2009 Nov;30(4):S15.

O9 Is the Aldosterone DPC Immunoassay Suitable for Fludrocortisone Suppression Testing?

PJ Taylor ¹, DP Cooper ², RD Gordon ¹, D Cowley ¹, M Stowasser ¹

Introduction

We have recently reported a LC-MS/MS method¹ for plasma aldosterone (PA). Radioimmunoassays (RIAs) may have lower precision and specificity, but have been clinically useful in diagnosis of primary aldosteronism (PAL). We used the LC-MS/MS assay to examine whether the RIA (DPC Coat-a- Count) might have led to either over- or under-diagnosis of PAL.

Methods

A total of 19 patients undergoing fludrocortisone suppression testing (FST) were studied. FST involved sampling at 0800h (overnight recumbency) and 1000h (2h upright) basally and after 3 and 4 days of fludrocortisone and oral salt loading. Day 4 upright PA of >165 pmol/L was considered diagnostic of PAL. PA concentrations were determined by LC-MS/MS and RIA and these results compared.

Results

Results by LC-MS/MS for 16 patients with PAL diagnosed by RIA supported the diagnosis in each. In 3 patients undergoing post-op FST following unilateral adrenalectomy for aldosterone-producing adenoma, both methods confirmed cure. Deming regression analysis comparing results by the two methods gave the equation LC-MS/MS = 0.80*RIA + 43 (n=94). Data was stratified into 3 groups: (A) <165 pmol/L (n=11), (B) 165-400 pmol/L (n=43) and (C) >400 pmol/L (n=40). Bland Altman analysis was performed on each group. For groups (A) and (B) the range of differences between methods was ±50%. For group (C) the mean bias was substantial (−200 pmol/L), and the range of differences was ±80%.

Conclusions

These findings validate the clinical utility of RIA in diagnosis of PAL, but raise questions about acceptable accuracy of measurement, particularly with high PA concentrations.

References

1.Taylor PJ, et al. Clin Chem. 2009;55:1155–62. doi: 10.1373/clinchem.2008.116004. [DOI] [PubMed] [Google Scholar]

CBR. 2009 Nov;30(4):S15–S16.

O10 Cockcroft and Gault - Revision with Australian Data

JY Wu ¹, DM Cowley ¹, GRD Jones ²

Introduction

It is recommended in Australia to assess renal function for drug dosing decisions using the Cockcroft and Gault (C&G) formula, with or without estimation of ideal body weight (IBW) from height. The C&G formula used data from 249 Canadian men and 10 women. Using Australian data we assess the performance of the original C&G formula and compare it with alternatives.

Methods

Data was generated from 396 women and 321 men who had creatinine clearance (CC) measured at Mater Pathology from 1998-2008. Creatinine was measured with a pre-IDMS aligned Vitros method (Ortho-Clinical Diagnostics). Age, sex, weight and height were recorded. Formulae were derived from the data using the C&G methodology and using an estimate of lean body weight (LBW) and results from these and other formulae compared to measured CC.

Results

Australian men have a higher creatinine output (CrOP) than in the original study, giving the “Mater C&G formula” improved performance. LBW and body surface area (BSA) showed best correlation with CrOP however these produced only marginal improvements. The use of IBW led to marked underestimation of CC. The use of MDRD “uncorrected for BSA” showed good performance with an expected small negative bias.

Conclusions

C&G can be improved using local data with small further improvements using LBW. “Uncorrected” MDRD (Modification of Diet in Renal Disease) also provides a good estimate of CC. Although this study used pre-IDMS aligned creatinine assays this affects all formulae similarly. The reference standard for this study was measured CC which is known to be unreliable. Traceability to measured glomerular filtration rate (GFR) using standardised creatinine assays is preferred. Thus an “uncorrected” MDRD may provide the most robust measure of absolute GFR.

CBR. 2009 Nov;30(4):S16.

O11 Not Even a Little Bit Pregnant - Ca Colon as a Cause of False Positive Pregnancy Test

B Dayanath ¹, I Simpson ¹, ZX Lu ^1,², JCG Doery ^1,³

Introduction

Pregnancy testing is performed in women prior to any procedure which could harm a developing embryo. False positive pregnancy tests have the potential to cause delay in treatment and great consternation for the patient. Patient: A 38 yr old female with advanced colorectal carcinoma (Duke’s stage 4) had serum beta HCG tested prior to receiving chemotherapy. HCG was 74 IU/L (RR<10) which was interpreted as unambiguously positive. Therefore chemotherapy was delayed.

Investigations

LH and FSH were normal. Ultrasound showed no evidence pregnancy. The initial HCG result was confirmed on alternate analysers and repeated one week later when it had risen to 84. Elimination of heterophile antibodies by use of Scantibodies and urine HCG confirmed the presence of free HCG. Immunohistochemical staining of the original tumour revealed patchy production of HCG in about 5% of the tumour population.

Discussion

An HCG of 74 IU/L is normally considered unambiguously positive and the absence of heterophile antibodies or peri/post-menopausal source of pituitary HCG appeared to further increase the possibility pregnancy. Demonstration of HCG producing tumor tissue confirmed the likely origin of the circulating HCG.

Conclusions

Laboratories must be aware of the common sources of false positive pregnancy tests including heterophile antibodies, ‘macro’ HCG, pituitary HCG in peri/post menopausal women and tumours. Production of HCG by germ cell tumours and teratomas is well recognised but the production by gasto-intestinal cancers is much less well recognised. Recent reports suggest non-trophoblastic tumours producing HCG are associated with a poor prognosis but a technique to stimulate anti-HCG antibodies may provide therapeutic promise.

CBR. 2009 Nov;30(4):S16.

O12 High Dose Hook Effect of BHCG Assay in Six Analytical Platforms

HA Al-Mahdili ¹, GRD Jones ¹

Introduction

High dose hook effect is a well-known phenomenon of two-site immunoassays including those for hCG. We studied this effect in hCG assays using six different analytical platforms and a sample with hCG concentration of 3,600,000 IU/L.

Methods

Dilutions of the high hCG sample were analysed on the following analytical systems: Advia Centaur, Immulite 2000, Dimension RxL (Siemens Diagnostics Solutions, Australia), Unicel DxI 800 (Beckman Coulter, Australia), Roche E170 (Roche Diagnostics Australia) and Architect (Abbott Diagnostics, Australia). The measured concentrations and corresponding assay signals were obtained for each method. Performance was compared with manufacturer’s claims.

Results

Three of the tested platforms demonstrated a clear high dose hook effect in the assay signal which produced falsely low results with no instrument flag in the neat sample (Advia Centaur, Immulite 2000 and Unicel DxI 800). The other methods showed no hook effect in the signal up to the highest level tested and no falsely low results. The hCG concentration up to which no false results were seen was equal to or higher than the manufacturers claims in all cases. The hook effect was more common in single step assays than those with a washing step prior to addition of the second antibody.

Conclusions

Our results indicate that hook effect may occur for some hCG assays, although at higher concentrations than those indicated in manufacturer’s product information. Assay design plays a role in its occurrence. Laboratories should be aware of the assay limitations in this regard.

CBR. 2009 Nov;30(4):S16.

O13 Serum Testosterone Measurement by Isotope-Dilution Liquid Chromatographytandem Mass Spectrometry (LC-MS/MS)

BR Cooke ¹, K Hoad ¹, SD Vasikaran ¹

Introduction

The use of LC-MS/MS for the analysis of serum testosterone provides positive identification of compound to low concentrations, without the cross-reactivity problems with other compounds of the steroid pathway that is inherent in immunoassay methods that results in poor specificity at low testosterone levels.

Methods

100 μL serum and 10 μL di-deuterated testosterone as internal standard (IS) were combined and extracted with ethylacetate/hexane (3:2, v/v). Organic phase was reduced to dryness under nitrogen gas and the residue redissolved in 100 μL 50% methanol. Chromatographic separation was by a C18 UPLC column with a gradient mobile phase consisting of 0.1% formic acid and methanol at a flow rate of 0.2 ml/min. The run time was 8 min, with testosterone eluting at 4.3 min. Testosterone was quantitated on a Waters Micromass Quattro Premier XE mass spectrometer in positive electrospray mode, Multiple Reaction Monitoring recorded the ion transitions from m/z 289.3 – m/z 108.8 for testosterone and m/z 291.3 – m/z 110.9 for IS.

Results

Testosterone calibrators showed a linear response from 0.1 nmol/L to 60 nmol/ L. Calculated intra-assay imprecision was 8.4%, 1.8% and 3.5%, while interassay imprecision was 12.6%, 6.5% and 5.0% over the concentration range 0.5, 3.6 and 9.2 nmol/L, respectively. Functional sensitivity was determined as 0.3 nmol/L. AACB WA QC subcommittee testosterone survey of patient samples (pools) were assayed by our method and were found to be in good agreement with data measured by tandem mass spectrometry from the Mayo Clinic, USA.

Conclusions

This validated tandem mass spectrometry assay offers a highly specific, accurate method of measuring extracted serum testosterone, especially useful for use in women and children.

CBR. 2009 Nov;30(4):S16–S17.

O14 Within-Method Bias Assessment Using QAP Data

GRD Jones ^1,², JP Gill ²

Introduction

The design of the RCPA Chemical Pathology QAP programs allows assessment of precision and bias. As the material is not native patient samples, assessment of between-method bias may be compromised by “matrix” effects. However within a manufacturer’s reagent group bias can be validly assessed. We examine bias within manufacturer method groups using group biological variation and other criteria.

Methods

Data from cycle 79 of the General Serum Chemistry and Therapeutic Drugs Program (Aug – Dec 2008) was extracted from the QAP database. For each analyte examined bias was assessed within each manufacturer’s group at the center of reference interval, using individual laboratories’ linear regression data. Laboratories were excluded if fewer than 16 results were returned or if the coefficient of variation (CV) for the analyte was outside the majority group of participating laboratories. Reference intervals were taken from the RCPA Manual.

Results

For some analytes the vast majority of laboratories within most method groups met the minimal or desirable standards for bias compared to the centre of the method group (e.g. iron, potassium, phosphate, transferrin, urate, urea). On analytical grounds most laboratories sharing the same methods may adopt common reference intervals. For some tests all or nearly all method groups were unable to meet the criteria (e.g. bicarbonate, chloride, calcium, creatinine, magnesium, sodium). This may indicate that laboratories need to reduce bias; or that wider common reference intervals or laboratory-specific intervals may be required.

Conclusions

QAP results can be used to assess within-method bias. Review of a number of selected analytes indicates the potential for shared reference intervals for some, and the need for bias reduction or other approaches for others.

CBR. 2009 Nov;30(4):S17.

O15 Should We Use HbA1c of 6.0% or 5.6% for Screening of Diabetes?

ZX Lu ¹, KA Sikaris ¹

Introduction

HbA_1c of >6.0% for screening and of ≥6.5% for diagnosis of diabetes has been recommended recently by a panel of experts although HbA_1c has traditionally been used only for monitoring treatment of diabetes. We aimed to evaluate the effectiveness of an HbA_1c cut-off of 6.0% for screening of diabetes using the WHO criteria for oral glucose tolerance test (OGTT).

Methods

Data from patients (n=2,494) whom had an HbA_1c (Bio-Rad Variant II Turbo analyser) and a 75 g OGTT performed at the same time was extracted from our pathology database. Patients were grouped based on their HbA_1c levels and their glycaemic status was classified as (WHO criteria) normal: fasting plasma glucose (FPG) ≤6.0 and 2-hour glucose (GLU-2h): ≤7.7; impaired fasting glucose (IFG): FPG 6.1–6.9; impaired glucose tolerance (IGT): GLU- 2h 7.8–11.0; diabetes: FPG ≥7.0 and/or GLU-2h ≥ 11.1 (all units in mmol/L).

Results

Of the 2,494 individuals, 34.6% were diabetic and 23.0% were normoglycaemic. In the diabetic group (n=864), the 2.5th percentile HbA_1c was 5.6%. Nineteen diabetic patients had HbA_1c below 5.6% and a further 100 had HbA_1c between 5.6 and 6.0%.

Table.

Distribution of glycaemic status by WHO criteria in each group of

	HbA1_c	HbA1_c	HbA1_c	HbA1_c
	≤5.5%	5.6–6.0 %	6.1–6.4 %	≥6.5%
Number	457	732	522	763
Normoglycaemia	63.5%	46.7%	23.6%	4.3%
IFG and/or IGT	22.3%	39.6%	48.9%	18.9%
Diabetes	4.2%	13.7%	27.6%	76.8%

Open in a new tab

Conclusions

A high proportion (13.7%) of patients with HbA_1c between 5.6-6.0% were diabetic. Using HbA_1c >6.0% for screening of diabetes, 119 (13.8%) diabetic were missed. We suggest using an HbA_1c at 2.5th percentile (≥5.6%) of the diabetic group as a cut-off for screening of diabetes as this gives 97.5% of sensitivity.

CBR. 2009 Nov;30(4):S17.

O16 Albumin Concentration in Stored Urine Samples: Losses and Gains

K Kania ¹, L Tay ¹, EA Byrnes ², JP Beilby ^1,², SAR Webb ³, KJ Strong ¹

Introduction

Protein degradation is a known concern for researchers who use stored urine samples. This study quantified changes to albumin during storage of urine collected from a cohort of critically ill patients.

Methods

Seventy-three urine samples were collected on the day of admission to ICU. The concentration of albumin was measured once by nephelometry, then again after storage for 12 months at −20°C. The CV of the nephelometry assay is stable at <5%. A different set of 40 samples was distributed into two separate aliquots, which were stored concurrently for 12 months at each of −20°C and −80°C. The two samples stored at different temperatures were then visually analysed after separation by SDS-PAGE.

Results

Sample degradation varied considerably between samples, ranging from 80% gain to 70% loss of nephelometric albumin concentration during storage (mean 3.3% loss, SD 23.4%). Loss was dependent on sample pH (regression, p<0.01, greatest losses at pH <5) but not initial albumin concentration. Eleven out of 40 samples analysed by SDS-PAGE had reduced albumin after storage at −20°C. Two of those samples acquired a new protein band approximately 45–50 kDa, likely to be a cleaved fragment of albumin.

Conclusions

The measured concentration of albumin can change over time in frozen urine samples. Conformational changes during degradation may expose internal epitopes leading to apparent gains in albumin concentration. Sample degradation might be prevented by adjusting pH prior to freezing.

CBR. 2009 Nov;30(4):S17.

O17 Diagnosis of CLCN7 Gene Related Osteopetrosis Using Serum Enzymes – A Case Report

SDC Thomas ¹, K Pearson ¹, G Hinds ¹, P Coates ¹

Introduction

Osteopetrosis is a metabolic bone disorder where osteoclast mediated bone resorption is impaired. The CLCN7 gene related spectrum of osteopetrosis includes infantile malignant recessive (ARO), intermediate autosomal (IA) and autosomal dominant type II (ADO II). The diagnosis usually relies on typical radiographic changes. Definitive diagnosis is by sequencing exons and adjacent splice sites. The proportion of cases caused by de novo mutations is unknown. We present the case of two affected sisters.

Case

The index case is a 4 year old girl who presented with a hip fracture following a fall. Radiography was suspicious of osteopetrosis. Following her biochemical analysis her asymptomatic mother and sister (6 years) were investigated. A maternal female cousin and her two adult daughters are affected and symptomatic. Skeletal surveys of the sister were suggestive of osteopetrosis, while that of the mother was normal.

Biochemistry

Both sisters had normal electrolytes, renal and liver function tests. Both sisters had markedly elevated total CK with a prominent CK-BB fraction and a significantly elevated total acid phosphatase (>95% tartrate resistant isoenzyme TRAP). ALP were within the reference interval with a normal isoenzyme pattern. The above tests in the mother were all normal.

Conclusions

Both siblings had demonstrated fractures and phenotypic characteristics of ADO II, i.e. elevated CKBB and TRAP. Obligate carriers cannot be identified using isoenzyme patterns. TRAP and CKBB can be used as specific diagnostic tests to screen high risk individuals who have not had radiographic assessment and for early diagnosis.¹

References

1.Waguespack SG, et al. J Clin Endocrinol Metab. 2002;87:2212–7. doi: 10.1210/jcem.87.5.8497. [DOI] [PubMed] [Google Scholar]

CBR. 2009 Nov;30(4):S19.

P1 Seasonal Prevalence of Suboptimal Vitamin D in First Trimester Pregnancy

EM Lim ^1,², S Abbs ², S Brown ³, N Hadlow ³

Introduction

Vitamin D deficiency in pregnancy has adverse consequences, such as neonatal hypocalcaemic convulsions, rickets in infancy and reduced bone mass in childhood. Furthermore, maternal vitamin D deficiency may predispose to preeclampsia. Maternal vitamin deficiency is a recognised public health problem in certain migrant populations. A Melbourne antenatal clinic reported 80% of dark skinned and veiled pregnant women are vitamin D deficient. We report the prevalence of vitamin D deficiency in pregnant women receiving routine antenatal care in Western Australia.

Methods

Vitamin D was measured on serum obtained from 485 consecutive women attending Western Diagnostic Pathology for first trimester screening during February 2008 (summer) and 496 women during August 2008 (winter). Vitamin D was analysed by chemiluminescent immunoassay (Diasorin Liaison).

Results

The results are shown by seasonal group.

	Summer	Winter
Number	485	496
Mean age (y)	31	31
Av gestational age (weeks)	11.3 ±1.1	11.2 ±1.0
Vit D <25 nmol/L	5 (1.0%)	51 (10.3%)
Vit D 25–50 nmol/L	66 (13.6%)	222 (44.8%)
Vit D 50–80 nmol/L	265 (54.6%)	208 (53.4%)
Vit D >80 nmol/L	148 (31%)	15 (3%)

Open in a new tab

The mean vitamin D levels in summer and winter were statistically different (p<0.001). There was no correlation between maternal age and vitamin D level.

Conclusions

During the first trimester of pregnancy, the prevalence of vitamin D insufficiency (<80 nmol/L) was 69% in summer, increasing to 97% in winter. Moderately severe deficiency (<25 nmol/L) was seen in 1% during summer, increasing to 10% in winter. These findings have potentially important adverse consequences for their offspring.

CBR. 2009 Nov;30(4):S19.

P2 Seasonal Effect on the Laboratory Prevalence of Vitamin D Deficiency Across Australia

K Sikaris ¹, D Kanowski ², G Ward ², Z Lu ¹

Introduction

Serum vitamin D (VitD) levels are mainly determined by sunshine exposure. Changes in levels may occur due to the duration of sunshine across different seasons, as well as other sunshine factors such as latitude. Both these factors would affect the distribution of VitD levels across Australia. We sought to examine the differences in VitD levels, using the same method, in patient samples across Australia for seasonal effects.

Methods

VitD was measured both at Melbourne Pathology (MP) and Sullivan Nicolaides Pathology (SNP) using the Diasorin Liaison analysers between 1/1/2008 and 22/8/2008. SNP acts as a reference laboratory for the Sonic Health Network and receives samples not only from Brisbane but also Hobart, Perth and Canberra. MP receives samples mainly from the city of Melbourne. Typical sunshine hours for each city were obtained from the Australian Bureau of Meteorology Website. VitD deficiency is defined as <50 nmol/L.

Results

There was a strong correlation (r² = 0.81) between the prevalence of VitD deficiency and sunshine hours.

Table.

Prevalence of Vitamin D deficiency in patient samples across Australia.

City	Summer

	N	VitD deficiency	Sun Hrs
Brisbane	2,163	21.6%	7.5
Canberra	3,699	22.1%	10.0
Hobart	1,415	30.5%	8.0
Melbourne	20,444	20.1%	10.0
Perth	2,127	17.3%	11.0

City	Winter

	N	VitD deficiency	Sun Hrs
Brisbane	3,427	32.3%	7.0
Canberra	4,892	45.5%	6.0
Hobart	3,201	54.8%	5.5
Melbourne	27,776	44.8%	4.0
Perth	3,197	35.4%	6.5

Open in a new tab

Conclusions

Laboratory prevalence of VitD deficiency is high in Australian cities and most of the variation from city to city and season to season is directly attributable to the number of sunshine hours.

CBR. 2009 Nov;30(4):S20.

P3 Deriving Sonic Network Reference Intervals for Pregnancy

K Sikaris ¹, Z Lu ¹, D Kanowski ², L Price ², R Flatman ³, G Caldwell ³, N Taylor ³, T Yen ³, S Sacks ⁴, M Metz ⁵, R Hanlon ⁶, J Andriolo ⁷

Introduction

As part of the Sonic Reference Interval Project, reference intervals for different stages of pregnancy need to be defined and validated. This would allow the network to move from site specified intervals based on either manufacturer suggestions, literature or locally defined reference limits to a common set of validated reference intervals.

Methods

Following a survey and literature review of existing reference intervals, U&E (n = 8,750) and LFT (n = 9,970) data was extracted from the Apollo Laboratory Information System at Melbourne Pathology and Sullivan Nicolaides Pathology for Bhattacharya analysis. As gestational age was also recorded, partitioning into trimester specific intervals could be investigated. Both laboratories use Roche Modular platforms for routine chemistry.

Results

Table.

Reference intervals for U&Es and liver function tests in pregnancy by trimester (Trim).

Analyte	1^st Trim	2^nd Trim	3^rd Trim	Not Pregnant
Sodium mmol/L	134–142			135–145
Potassium mmol/L	3.4–4.8			3.5–5.5
Chloride mmol/L	95–108			95–110
Bicarbonate mmol/L	18–28			20–32
Urea mmol/L	1.5–5.5			2.5–7.0
Creatinine μmol/L	30–70			45–85
Bilirubin μmol/L	2–15	2–10		3–15
Total Protein g/L	63–80	60–75	55–72	64–81
Albumin g/L	35–48	32–43	28–40	37–48
ALP IU/L	30–100		60–300	20–105
GGT IU/L	5–30		3–30	5–35
ALT IU/L	5–30			5–30
AST IU/L	8–30		8–35	10–35
LD IU/L	100–200		100–200	120–250

Open in a new tab

Conclusions

All reference intervals for pregnancy were lower than in a non-pregnant young woman with the exception of ALP which rose dramatically at the start of the third trimester.

CBR. 2009 Nov;30(4):S20.

P4 Deriving Sonic Network Reference Intervals for Children

K Sikaris ¹, Z Lu ¹, D Kanowski ², L Price ², R Flatman ³, G Caldwell ³, N Taylor ³, T Yen ³, S Sacks ⁴, M Metz ⁵, R Hanlon ⁶, J Andriolo ⁷

Introduction

As part of the Sonic Reference Interval Project, reference intervals for different stages of childhood need to be defined and validated. This would allow our network to move from site specified intervals based on either manufacturers suggestions, literature defined limits and locally defined reference intervals to a common set of validated reference intervals.

Methods

Following a survey and literature review of existing reference intervals, U&E and LFT data were extracted from the Laboratory Information System at Melbourne Pathology and Sullivan Nicolaides Pathology for trend analysis. Both laboratories use Roche Modular platforms for routine chemistry. Age and gender specific reference intervals were investigated.

Results

Table.

U&E and LFT reference intervals in childhood.

Analyte	6 months	5 years	10 years	15 years
Sodium mmol/L	132–145
Potassium mmol/L	3.6–5.8		3.5–5.5
Chloride mmol/L	95–110
Bicarbonate mmol/L	17–29		21–31
Urea mmol/L	1.5–5.5	2.5–6.5	2.5–6.0
Creatinine μmol/L	20–35	20–40	20–65	40–75
Bilirubin μmol/L	2–10	2–12	3–15
Total Protein g/L	50–70	63–79	65–80	65–81
Albumin g/L	35–48	37–48	38–49
ALP IU/L	120–350	120–300	100–400	50–200
GGT IU/L	5–60	5–20	5–30
ALT IU/L	5–35		5–30
AST IU/L	20–65	15–45	15–40	10–35
LD IU/L	150–450	150–350	120–300	120–275

Open in a new tab

Conclusions

The results imply that paediatric reference intervals change at varying stages of childhood depending on the physiological changes affecting each analyte. Defining the age for making a reference interval transition in childhood is just as important as the determining the limits of the reference interval for the age interval selected.

CBR. 2009 Nov;30(4):S21.

P5 Highly Specific Reference Intervals for CA125 II Assays

K Sikaris ¹, Z Lu ¹, S Greco ¹, D Kanowski ², M Freemantle ², T Yen ³, A Liu ³, G Caldwell ³

Introduction

The Centocor CA125II assay was revised in 1993 to improve the analytical sensitivity of the original Centocor CA125 assay. Despite most studies demonstrating differences between CA125II and CA125 results, many manufacturers maintain the same ‘reference interval’ of <35 IU/L defined as 99% specificity for the original assay. We sought to investigate the reference intervals for CA125 in women and men using current assays.

Methods

Data was extracted for Abbott AxSym CA125 (n = 26,337) and Abbott Architect CA125II (n = 34,602) from Melbourne Pathology (MP), Abbott Architect CA125II (n = 51,910) from Sullivan Nicolaides Pathology (SNP) and Roche E170 CA125II (n = 33,561) from Douglas Hanly Moir Pathology (DHM). About 5% of all CA125 requests were on men. The Bhattacharya procedure was used to define the dominant underlying Gaussian population which is assumed to represent the many unaffected patients having CA125 tests in private pathology.

Results

The distribution of CA125 results is log normal, consistent with what has been previously described for most tumour markers.

Table.

99% Reference Intervals for CA125 using a variety of assays (units: IU/L).

CA125 method	Women 30–39y	Women 60–69y
MP AxSym CA125	5–26	5–21
MP Architect CA125II	3–43	3–34
SNP Architect CA125II	1–38	3–32
DHM E170 CA125II	3–45	2–36

CA125 method	Women 80+y	Men 18–59y

MP AxSym CA125	5–24	5–20
MP Architect CA125II	3–40	4–27
SNP Architect CA125II	3–42	4–28
DHM E170 CA125II	3–40	2–39

Open in a new tab

Conclusions

Upper reference limits for both the CA125II assays appeared seem to be considerably higher than 35 IU/L upper limit intended for the original CA125 assay. Furthermore reference intervals are lower in post menopausal women and men but rise in the elderly. Although CA125 should not be used for screening, improved definition of limits may reduce false alarms and improve the understanding of CA125 elevation in benign disease.

CBR. 2009 Nov;30(4):S21.

P6 Globulin Levels Expected in CRP Defined Inflammation

K Sikaris ¹, Z Lu ¹, K Dahanayaka ¹

Introduction

Inflammation is associated with increased production of acute phase proteins and immunoglobulins, together leading to an increase in serum globulins. In this study we attempt to define the degree of globulin increase expected with various degrees of inflammation defined by increasing C-reactive protein (CRP) levels.

Methods

Data extraction was 1/1/2009 – 15/3/2009 when there were 31,336 LFTs with CRP requested during the same episode. Globulins were calculated from the difference between the serum total protein and serum albumin (modified BCG) measurements. All tests were done using a Roche Modular.

Results

Table.

Relationship between CRP and 95% confidence limits for serum globulins.

CRP Level mg/L	Number of pairs	Serum Globulins (95% Limits) g/L
< 2	12,808	22 – 38
3 – 7	6,934	22 – 40
8 – 15	3,340	22 – 42
16 – 25	1,817	20 – 43
26 – 50	2,019	19 – 45
51 – 150	2,972	18 – 46
> 150	1,446	17 – 44

Open in a new tab

Conclusions

With worsening inflammation, as indicated by elevated CRP, serum globulins can be expected to rise up to a level of 46 g/L. However when serum globulins levels are greater than 46 g/L, alternative causes of the globulin rise should be considered regardless of the degree of inflammation present.

CBR. 2009 Nov;30(4):S21.

P7 Detection of Accidental Glimepiride Ingestion Causing Factitious Hypoglycaemia

A Konpa ¹, RO Fullinfaw ¹

Introduction

An increase in the prescription of oral hypoglycaemic agents reflects the increasing prevalence of type II diabetes mellitus in our community. Elderly patients in whom confusion regarding medications can be significant may overdose or administer incorrect medication leading to adverse outcomes. Thus screening for the presence of sulfonylureas is recommended in the assessment of patients without diabetes experiencing hypoglycaemia. Glimepiride is a commonly used sulfonylurea in Australia but is not included in commonly available serum sulfonylurea screens. We report a case of hypoglycaemia due to accidental glimepiride ingestion confirmed by detection of this agent in serum using high performance liquid chromatography.

Methods

High performance liquid chromatography coupled with photodiode array detection was undertaken on the patient’s plasma sample taken four hours after presentation. Spectral matching was then used to verify that the resultant peak conformed to the known glimepiride spectrum.

Results

A peak corresponding to glimepiride was detected in the patient’s plasma. The computerised spectral matching score was rated at 972 out of a perfect score of 1000. It was concluded that the patient’s sample did contain glimepiride at an approximate concentration of 0.11 mg/L.

Conclusions

Misdiagnosis of the cause of hypoglycaemia can result in adverse outcomes. Clinicians rely upon accurate detection of sulfonylureas given that insulinoma and insulin secretagogue induced hypoglycaemia may have identical biochemistry. However, many techniques for the detection of first-generation sulfonylureas do not detect second-generation sulfonylureas, including glimepiride. Further development of such assays such would assist clinicians faced with this diagnostic dilemma in the future.

CBR. 2009 Nov;30(4):S22.

P8 OGTT Results Using Fasting Glucose Cut-Off of 5.5 Compared to 6.0 MMOL/L

ZX Lu ¹, KA Sikaris ¹

Introduction

Impaired fasting plasma glucose (IFG) is currently defined differently in different countries. It is defined as fasting plasma glucose (FPG) 5.6–6.9 mmol/L in the United States and 6.1–6.9 mmol/L in Australia. Patients with IFG are usually followed up with a 75 g oral glucose tolerance test (OGTT) to further clarify their glycaemic status in both countries. We evaluated the impact of different lower limit of IFG (5.6 vs 6.1 mmol/L) on the rate diabetes defined by the 2-hour glucose post-OGTT (GLU-2h).

Methods

Data from adult patients (n = 52,134) aged (mean ± SD) 50.9 ± 17.1 years whom had a 75g-OGTT performed over 2003–2009 was extracted from our pathology database. GLU-2h status was classified (WHO GLU-2h criteria) as normal: GLU-2h ≤7.7, impaired glucose tolerance (IGT): GLU-2h 7.8–11.0, diabetes: GLU-2h ≥11.1 mmol/L.

Results

Of the 52,134 individuals, 12.5% were diabetic and 64.8% were normal by WHO GLU-2h criteria for OGTT. The distribution of GLU-2h in people with different FPG was:

	Fasting plasma glucose (mmol/L)
	≤ 5.5	5.6–6.0	6.1–6.9	≥ 7.0

n	29,996	8,369	9,256	492
GLU-2h status
Normal	82.5%	60.7%	38.5%	10.9%
IGT	15.8%	30.2%	37.7%	22.0%
Diabetes	1.7%	9.0%	23.8%	67.1%

Open in a new tab

Conclusions

Using the American Diabetic Association definition of IFG (FPG 5.6–6.9 mmol/L) for screening of diabetes, only 1.7% of patients who are diabetic by WHO GLU-2h criteria would have been missed. Using the WHO definition of IFG (FPG 6.1–6.9 mmol/L), over five times more diabetics by GLU-2h would be missed. Furthermore, the incidence of IGT is similar in patients with FPG between 5.6–6.0 mmol/L as with FPG between 6.1–6.9 mmol/L (30% vs 37%, respectively). We believe that significant portion of patients with a FPG between 5.6 and 6.0 mmol/L have glucose intolerance and should not be ignored.

CBR. 2009 Nov;30(4):S22.

P9 Critical Difference Calculations: Use of Multiple Samples

GRD Jones ¹

Introduction

The commonly used calculation for Critical Differences (CD) is based on the significance of differences between two singleton measurements. The use of multiple measurements is sometimes recommended e.g. baseline lipid assessment prior to starting lipid lowering therapy. A formula for CD with multiple measurements is described together with a calculator for multiple measurements before and/or after an event or intervention.

Methods

The uncertainty ascribed to a pathology result (± 2 x CVtot, the combination of within-subject and analytical variation) is the interval within which a patient’s homeostatic set point for that analyte lies with a specified certainty, usually 95%. When multiple measurements are taken, this interval becomes smaller by a factor of 1/√n, where n is the number of measurements, provided the variability is Gaussian in nature. A formula was developed and modelled in Microsoft Excel to determine the probability of a true change from before to after an event with variable sample numbers.

Results

Increasing the number of samples before and/or after a clinical event leads to a fall in the CD by approximately 29% (for 2 measurements before and 2 after), 42% (3 and 3) and 50% (4 and 4) when compared to 2 singleton measurements. An unequal number (e.g. 3 before and 1 after) dose not provide the same reduction in CD as equal numbers. A spreadsheet calculator is described for performing these calculations on patient results.

Conclusions

Repeat testing before and/or after a clinical event improves the precision of the estimates of the true patient values, allowing detection of smaller changes with greater certainty. These calculations are facilitated by provision of calculators.

CBR. 2009 Nov;30(4):S22.

P10 Alpha-1-Antitrypsin Deficiency – 2 Year Review in Transplantation Assessment

S Sykes ¹, I Valentine ¹, S Klingberg ¹, J Tate ¹

Introduction

Alpha-1-antitrypsin (AAT) is a polymorphic serine protease inhibitor encoded by the 12.2 kb gene on chromosome 14. There are now at least 75 AAT variants with some alleles (e.g. Z, S, I, F) associated with genetically low levels and early onset liver or lung disease.

Methods

AAT was measured by automated immunochemistry. Pi (protease inhibitor) phenotyping was performed without immunofixation by polyacrylamide gel isoelectric focusing (IEF) (pH range 4.2 to 4.9). AAT requests were analysed by IEF when the AAT was <1.5 g/L. Specific requests for AAT typing were analysed by IEF irrespective of the AAT value and genotyped after review of the Pi patterns. Genotyping or sequencing of the AAT gene was also performed in samples negative for the common deficiency alleles. Patient records for AAT requests for years 2006 and 2007 were analysed with Microsoft Access database software.

Results

Of the 5415 requests, 876 had one or more abnormal (non M) Pi type. 666 were MZ or MS heterozygotes. Three Pi types with IM alleles and AAT values above 1.5g/L would have been missed by our current screening algorithm. Rare allelic variants, nulls and deletions were also found including an AAT/AAT*E132K genotype (non-HGVS nomenclature) reported as previously undescribed. In three instances there were discrepant pre- and post-transplantation Pi types with two normal MM phenotypes pre-transplantation receiving a donor organ heterozygous for the Z deficiency allele.

Conclusions

AAT measurement and Pi typing are important in transplantation assessment. Although AAT is an acute phase reactant our screening algorithm detected the common alleles associated with significant disease. This study also highlights problems of assessing AAT status prior to transplantation particularly where only genotyping is used.

CBR. 2009 Nov;30(4):S22–S23.

P11 Coagulation Factor Level Most Important Variable in Triaging Hepatitis B Induced Acute Liver Failure Patients for Liver Transplantation

N Gupta ¹, S Sharma ¹, A Mukhopadhyay ¹

Introduction

In acute liver failure, variables predicting survival of patients are based on age, interval between onset of jaundice and encephalopathy, coagulation factor level, metabolic parameters such as arterial blood pH, blood gas level, and lactic acid levels. The exceptions are liver failure due to paracetamol poisoning or Wilson’s disease, where the metabolic derangements are much higher. In the Indian subcontinent the major cause of acute liver failure is hepatitis B and we investigated the major predictors of survival in a cohort of patients with viral hepatitis induced acute liver failure, to assist in triaging these patients for liver transplantation.

Methods

Subjects - 40 patients admitted in gastroenterology intensive care unit following diagnosis of acute liver failure were followed for one month. Coagulation factor V, factor VIII, arterial blood gas and lactic acid levels were measured on day 1 and day 3 of admission.

Results

The analysis showed that patients who survived had factor V level on day 3 of diagnosis in the range of 10%–25% and those who died had factor V level below 10% (p<0.004), with other predictors of survivals being non-significant.

Conclusions

Thus factor V level on day 3 of diagnosis is the most important variable in predicting survival in patients of viral hepatitis B induced acute liver failure, thus help in triage for the need of orthotopic liver transplantation which is the only definitive curative therapy in these patients.

CBR. 2009 Nov;30(4):S23.

P12 Laboratory Measurement of Serum IGG and Implications for Therapeutic Supplementation

D Gillis ¹, J Tate ², S Jovanovich ³, R Wong ¹

Introduction

Guidelines for use of therapeutic IgG depend on accurate IgG concentration to ensure the concentration at which supplementation is given is independent of IgG method or the laboratory performing the assay. Performance of IgG assays was evaluated using the data from the RCPA Immunology Specific Proteins QAP.

Methods

Mean IgG concentration and coefficient of variation (CV) were determined for 52 patient specimens measured by between 93 to101 laboratories in QAP Cycles 21 to 25 (2007 until April 2009). CV data were combined in a variance function plot to assess total error for IgG for all methods (n = 19–20).

Results

Mean tested IgG concentration was 3.55 to 21.89 g/L with between-laboratory variation of 5.18 to 17.36 %CV. Forty of the 52 specimens gave CVs below 8%. From the profile of CV versus IgG concentration, total error was constant across the IgG range of 5 to 22 g/L and increased at lower concentration. Of the 12 samples with a CV above 8%, removal of gross outlier measurements reduced the between-laboratory variation to below 8% in 11 of 12 specimens and was 9.6% at 3.65 g/L IgG for specimen 23–06. Outliers were method-independent.

Conclusions

According to intra- and inter-individual biological variation measurements, the desirable total error (imprecision and bias) for serum IgG is 8%. Ten years after the initiation of protein standardisation and traceability of IgG results to the reference material CRM-470, in Australia there seems to be good harmonisation of IgG methods. IgG levels throughout Australia can be used with confidence in therapeutic IgG guidelines.

CBR. 2009 Nov;30(4):S23.

P13 Cryofibrinogen Measurement and its Clinical Utility

J Tate ¹, D Gillis ², N Avsenev ¹, L Callaghan ¹, B Gluchowska ¹, S Klingberg ¹, I Valentine ¹, S Sykes ¹

Introduction

Cryofibrinogen is a cryoprotein that precipitates out in plasma that has been refrigerated at 2–8°C. The clinical significance of cryofibrinogen greater than 0.5 g/L versus less than 0.5 g/L was evaluated according to clinical outcomes.

Methods

Whole plasma (WP), supernatant minus the precipitate (SN) and precipitate resolubilised in 0.15 M saline were reheated to 37°C and electrophoresed on Sebia Hydrasys (Sebia, France) 9IFE gels followed by protein fixation, immunofixation with anti-fibrinogen and staining. Fibrinogen in WP and SN was measured indirectly by difference in total protein and albumin (Globulins-WP minus Globulins-SN) or directly by immunological fibrinogen (FIBI) immunoassay using an IMMAGE nephelometer (Beckman Coulter, CA) (FIBI-WP minus FIBI-SN).

Results

Of 1,597 cryoprotein screens performed in Pathology Queensland during 2007 and 2008, 213 (13.3%) were positive for cryofibrinogen and 60 for cryoglobulins (3.8%). Cryofibrinogen concentration was 0.5 to 5.9 g/L in 44 (20.7%). Of these 44 patients, fourteen had a history of leucocytoclastic vasculitis. However seven of 44 randomly selected patients with cryofibrinogen less than 0.5 g/L also had a history of leucocytoclastic vasculitis. There was no other specific disease association in patients with the cryofibrinogen greater than 0.5 g/L, who had a wide variety of clinical diagnoses.

Conclusions

Measurement of cryofibrinogen has a limited role in clinical diagnostics. Although there was an association of small vessel vasculitis with a cryofibrinogen concentration greater than 0.5 g/L, this association is not significant in clinical practice. Whether or not this is due to analytical limitations of the laboratory measurements is unclear.

CBR. 2009 Nov;30(4):S23.

P14 Protocol for Bias Assessment and Commutability Testing Using Urine Albumin Measurement

J Tate ¹, W Ferguson (Chair) ¹, K Jones ¹, M Freemantle ¹, J Oostenbroek ¹, J Gill ², G Jones ²

Introduction

External Quality Assessment materials can suffer from non-commutability which can cause difficulties distinguishing between a true bias for an analytical system and an underlying matrix problem. We developed a protocol to assess bias and commutability for urine albumin measurement.

Methods

The study evaluated assay precision (imprecision profile) and bias (Bland-Altman plot) for Cobas Integra 800, Vitros 5 1/FS, Beckman DxC800 and Cobas 6000 urine albumin assays and tested commutability for three pair wise sample comparisons using linear regression and 95% prediction interval analysis. Albumin materials included 80 patient urine samples, six RCPA Urine Chemistry QAP (2008), CRM-470 plasma proteins reference material diluted to 40, 100 and 400 mg/L in water, saline and urine albumin pool (<1 mg/L), BioRad Lyphochek and Liquichek controls.

Results

Total result variability (within-run, between-run, between-method) for 80 samples ranged from 24.5% at 2–15 mg/L albumin (n = 9 samples), 5.3% at 15–80 mg/L (n = 18), to 4.5% at 80–1006 mg/L (n = 53). Between-method bias was generally within 10% limits except at around 10 mg/L albumin where the difference was 35%. Dilutions of CRM-470 were approximately 12% low by all methods. RCPA QAP and Liquichek samples were commutable by all methods whereas Lyphochek was not commutable by DxC800.

Conclusions

Currently accuracy of urine albumin methods cannot be validated without use of a suitable matrix-matched reference material. QAP samples showed commutability to native urine samples and can be used as an alternative to assess methodology issues. The study indicates that the methods tested may not be able to be used interchangeably within the normal range.

CBR. 2009 Nov;30(4):S24.

P15 Albumin in Urine is Cleaved by Endogenous Proteases

K Kania ¹, EA Byrnes ², JP Beilby ^1,², SAR Webb ³, KJ Strong ¹

Introduction

Albumin in stored urine loses reactivity with assays over time; losses are greatest in samples with low pH conditions. The aim of this study was to investigate the role of endogenous urinary proteases in albumin degradation.

Methods

Two normoalbuminuric and five macroalbuminuric urine samples were collected from healthy volunteers and critically ill patients respectively. Urine samples were adjusted to pH <2.5 with HCl and incubated for up to 48 hours at room temperature. Albumin fragmentation was observed using the biuret assay for total protein, SDS-PAGE, nephelometry and the HPLC assay for urinary albumin. A pepsin digest of albumin was used as a comparison.

Results

Within minutes after pH adjustment, both endogenous albumin and exogenous albumin added to urine was extensively fragmented. The speed of degradation was dependent on pH, time and temperature. The fragments produced were consistent with those produced during digestion with pepsin. The HPLC assay was most dramatically affected, with near complete loss of albumin-sized material within one hour of low pH incubation. Reactivity of the sample with nephelometry antiserum initially declined then increased, which may reflect epitopes usually internal to the albumin molecule being exposed during digestion. Urinary proteolytic activity was completely inhibited by pepstatin.

Conclusions

Urine contains proteases capable of degrading albumin. HPLC may considerably underestimate albumin present in a stored sample while immunoassays such as nephelometry can give inaccurate results. These findings have implications for studies using stored samples.

CBR. 2009 Nov;30(4):S24.

P16 Comparison of Albumin Methods

H Martin ¹, B Sonza ¹

Introduction

Two methods for albumin are in common use in the wider laboratory community. Bromocresol purple (BCP) is used on Dimension RXL analysers in our satellite laboratories and bromocresol green (BCG) in used on Advia 1600/2400 analysers in our main laboratories While BCP is generally accepted to be highly specific for human albumin, it is well known that BCG suffers from non-specific binding to alpha and beta globulins leading to over-estimated albumin in sick people. Since in our organisation it is possible that consecutive sera for an individual patient be measured by different albumin methods we aimed to quantify the extent of any difference.

Methods

Current samples from patients identified by routine result validation scrutiny as having had albumin measured by both BCP and BCG at some time were selected and analysed by both methods. A subset of these was also examined by electrophoresis

Results

42 sera were analysed by BCP and BCG; all had unacceptable differences as defined by the Royal College of Pathologists of Australasia Quality Assurance Programs’ allowable limits of performance for albumin. Correlation between BCP and BCG was poor, y = 1.26x, r² = 0.65, mean difference was 7.4 g/L. For the subset of 19 samples also run on electrophoresis, correlation with albumin by BCP was excellent, y = 0.98x, r² = 0.95, mean difference -1 g/L while correlation with BCG was poor, y = 1.22x, r² = 0.75, mean difference 8.2 g/L.

Conclusions

BCP albumin is more specific than BCG. Interchanging albumin results from BCP and BCG methods may result in clinically unacceptable differences. In order to provide uniformity within our organisation we are actively seeking a BCP albumin method for our Advia platforms.

CBR. 2009 Nov;30(4):S24.

P17 Hba1c and Microalbuminuria in the Monitoring of Treatment of Diabetes Mellitus

VS Hoang ¹, AT Vu ¹, MC Vu ¹

Introduction

The determinations of HbA_1C in blood and microalbuminuria are important in the monitoring of treatment of diabetes mellitus. During the period 2006–2008, we studied the relationship between glucose, HbA_1C, microalbumin and the clinical symptoms of 124 diabetes patients treated in the Trang An Hospital, to evaluate the clinical utility of these tests and provide related information to practitioners in less favourable conditions.

Methods

The investigation population comprised of 124 non-insulin-dependent diabetes mellitus patients (71 males, 53 females) aged 27–85 years. HbA_1C and microalbumin were determined using immunoassay on the DCA 2000+. Glucose, insulin and creatinine in blood, protein in urine, and a biopsy of the kidney were also performed to assess for renal disease.

Results

Abnormal values of glucose, HbA_1C and microalbumin were found in 92.4%, 63.7% and 51% of patients, respectively. The patients with abnormal HbA_1C had glucoses from 7.1 to 30.0 mmol/L. Mean test values of the patient population: glucose: 9.8 mmol/L, HbA_1C: 8.1%, microalbumin: 76.50 mg/L. However, the creatinine in blood and urine were all normal. 10% of patients had some disturbances of renal function to a mild degree, without renal failure (biopsy result: high mesangial proliferation of the glomerulus cell; 8.9% with mild proteinuria).

Conclusions

There were close relationships between glucose and HbA_1C, and between microalbumin and renal function. However, no relationship between microalbumin and glucose and HbA_1C was identified. HbA_1C and microalbumin are good predictors of diabetic nephropathy. We have proposed a diabetes control scheme using glucose and HbA_1C data.

CBR. 2009 Nov;30(4):S24–S25.

P18 A Local and National Database for Variants for Breast Cancer

V Hyland ¹, J MacMillan ², R Cotton ³

Introduction

BRCA1 and BRCA2 gene analyses are performed to detect variants in the normal DNA sequences that predispose an individual to breast and/or ovarian cancer. Variants of the normal gene sequence are screened for and reported using nomenclature recommended by the Human Genome Variation Society. The reporting of a variant of clinical significance allows predictive gene testing in other family members. Variants of unknown clinical significance may not used for predictive gene testing.

Methods

A local Leiden Open Variation Database (LOVD) is used to collect data from 192 breast cancer BRCA1 and BRCA2 gene analyses. Data is stored on a personal computer that is not connected to any network. Continual updating of the local LOVD database is achieved using a modification of existing laboratory worksheets to produce tab delineated data sets. The data is curated in house. Other data include a de-identified patient number, molecular techniques used, variant nucleotide and protein check, SIFT values, PolyPhen values, Align GVGD values, splicing values, UTR values, variant detection in mRNA, publication and database web links in the laboratory fields with individual and pedigree specific cancer data in the clinical fields. At present no functional assays are performed.

Results

192 BRCA1 and BRCA2 gene analyses were performed from 2007 to date. Fifteen variants of clinical significance were detected in BRCA1. Twenty three variants of clinical significance were detected in BRCA2. Three BRCA1 variants of unknown clinical significance co-occurred with a BRCA1 clinically significant variant. Five BRCA2 variants of unknown clinical significance cooccurred with a BRCA2 clinically significant variant. Six BRCA2 variants of unknown clinical significance co-occurred with a BRCA1 clinically significant variant. One BRCA1 variant of unknown clinical significance co-occurred with a clinically significant BRCA2 variant.

Conclusions

The local LOVD is used as a record to determine the specific type, frequency of any variant and the co-occurrence of variants of unknown clinical significance with a variant of clinical significance for the BRCA1 and BRCA2 genes. The data will be available to upload to the Australian Human Variome database and then to international databases. The aggregation of similar data from other laboratories in Australia will allow the clinical significance of the unknown clinical BRCA1 and BRCA2 variants to be revised.

CBR. 2009 Nov;30(4):S25.

P19 Biomarkers in Genetic and Geographic Epidemiology

JB Whitfield ¹

Introduction

Many of the laboratory tests used for diagnosis and monitoring of disease have predictive significance as risk factors, or are able to detect pre-clinical disease. They are therefore useful for research into a wide range of conditions, and allow inclusion of people who have not reached the stage of clinical disease. The recent burst of publications on gene variation in common diseases comprises both case-control studies and quantitative studies on risk factors. Applications of biomarkers in genetic epidemiology will be presented, together with early results on regional variation in risk factors across Australia.

Methods

The Genetic Epidemiology group at QIMR has accumulated biomarker data from multiple twin and family studies on adults and adolescents, since 1979. This includes lipids, glucose and insulin, hepatic and renal function tests, and markers of iron status. This complements an extensive dataset of demographic characteristics and a substantial resource of genome-wide association markers covering some 20,000 participants.

Results

Earlier studies concentrated on variation in liver function tests associated with alcohol use or obesity, and on the presence of genetic effects. Recent studies have located genes and variants which affect lipids, uric acid and markers of iron status. For transferrin and cholinesterase, variation close to the structural gene affects the plasma concentration or activity. The data have also allowed us to define regional variation in blood lead concentration across Australia, in addition to the genetic variation we previously discovered.

Conclusions

The concept of biomarkers offers opportunities which overlap with diagnostic testing, but also embrace research areas such as epidemiology, genetics, drug safety assessment and of course the development of new markers of disease. What was clinical chemistry has been re-branded as laboratory medicine or in-vitro diagnostics, but it is also a powerful tool for health research.

CBR. 2009 Nov;30(4):S25.

P20 The Effects of Increased Water Drinking on 24 Hour Urine Volumes - Endemic Psychogenic Polydipsia

AR McNeil ¹, S Pereira ¹, Y Walker ¹, D Mitchell ¹, QT Lam ¹

Introduction

The increasing sales of bottled water and popularity of drinking water while exercising or working suggest that people are drinking more now than they did in the past. The aim of this study was to see whether the volumes of 24 hour urine specimens submitted to a large pathology laboratory for diagnostic testing have changed over the last 15 years.

Methods

The average volumes of 24 hour urine specimens for men and women aged 19–99 years were examined excluding those with volumes less than 500 mL or greater than 10,000 mL. The analysis included one group of 110,835 specimens tested predominantly for protein excretion 1993–2006. The second group of 65,055 specimens were tested for creatinine clearance 1993–2008.

Results

The average volumes of urine protein collections increased by 14% in women (1660 to 1900 mL) and 9% in men (1880 to 2040 mL). The same trend was seen with urine collected for creatinine clearance measurements where the volumes increased 11% in women (1818 to 2015 mL) and 11% in men (2010 to 2229 mL). The average creatinine concentrations fell by 14% in women and 20% in men in the same specimens indicating that people were passing larger volumes of more dilute urine. The changes were similar each year and there was no suggestion that a plateau had been reached.

Conclusions

Adults collecting 24 hour urine specimens for diagnostic purposes appear to have been drinking on average 15 mL more per day each year since 1993. Average 24 hour urine volumes have increased by approximately 10% or 200 mL over the last 15 years.

CBR. 2009 Nov;30(4):S25.

P21 Classification of Proton Magnetic Resonance in Vivo Spectra from the Normal Human Brain by Means of Pattern Recognition Methods

M Sokol ¹, A Polnik ¹, M Wicher ², T Banasik ², E Jamroz ³, J Paprocka ³

Introduction

Metabonomics based on pattern recognition methods proved to be an effective tool for analysis of high resolution MR spectra and has recently gain popularity in the field of in vivo MRS. In this work PCA and PLS-DA were chosen for evaluation of metabolic heterogeneity of the normal brain.

Methods

The studied group consisted of 28 volunteers. The ¹H MRS spectra were acquired from thalamus, pons, cerebellum, frontal and occipital white matter and hippocampal areas using 1.5T scanner. PCA and PLS-DA were applied to the metabolite levels determined with LCModel and to the unresolved spectra. Pre-processing of the raw data included eddy current correction, zero-filling, 3Hz exponential line broadening, Fourier transformation, frequency alignment, residual water signal removal and baseline correction.

Results

As reveals from the PCA analysis the spectra acquired from the hippocampal and cerebellar regions are separated from the remaining localizations along the component characterised by a positive contribution of GPC+PCh, Ins, Cr and Glu+Gln and a negative contribution of NAA and NAA+NAAG. The PLS-DA analysis shows that cerebellum is characterised by the highest Cr content, thalamus has high NAA, and the concentrations of GPC+PCh, Ins and NAA+NAAG are high in pons, whereas the spectral region corresponding to NAA+NAAG contributes to the differentiation between the frontal and occipital localisations. Significant separation between the spectra acquired from cerebellum, pons and thalamus was also obtained from the analysis of the unresolved spectra.

Conclusions

PCA and PLS-DA provide an alternative way of MRS data structure evaluation.

CBR. 2009 Nov;30(4):S26.

P22 Illicit Drug Exposure During Pregnancies with Gastroschisis in Queensland Australia

C Gurnsey ¹, L Johnson ¹, C Warnholtz ¹, T Donovan ²

Introduction

Gastoschisis is a life-threatening foetal defect of the abdominal wall through which loops of intestine, and at times stomach protrude. The birth incidence of gastroschisis increased >4.5 fold in Queensland (1988 to 2006) from 1.23 to 5.65 per 10,000 births. Meconium analysis provides a longer term view of pregnancy drug exposures from 20 weeks gestation to delivery compared with maternal urine testing and eliminates the biases of maternal self reporting. This case-control study aims to provide preliminary results of an investigation into whether prenatal illicit drug exposure as measured by meconium analysis, has a significant association with gastroschisis.

Methods

A structured questionnaire, maternal urine (14–18 weeks) and neonatal meconium were collected. Maternal urine and neonate meconium were analysed for amphetamine type substances, opiates, cannabinoids, cocaine and benzodiazepines using immunoassay and gas chromatography/mass spectrometry.

Results

Of 34 case and 30 control meconium analyses the following results – drug and (frequency) were observed:

CASES: Morphine (6), Ephedrine (5), Methylenedioxymethylamphetamine (1), Methylamphetamine (1) and Cannabinoids (6).

CONTROLS: Morphine/Codeine (1/2), Ephedrine (1).

When drugs used for the management of the mother’s delivery (morphine, codeine and ephedrine) were excluded, significant associations for increased risk of gastroschisis were present. Illicit drug exposure by meconium analysis occurred in 20.6% of cases and none in controls (p = 0.009). Cannabis exposure explained almost all of this association with detection in 17.6% of cases and none in controls (p = 0.018). Amphetamine exposure was insignificant.

Conclusions

This study has identified that illicit drug exposure during pregnancy is associated with gastroschisis. Highest risks were with cannabis.

CBR. 2009 Nov;30(4):S26.

P23 Benign Transient Hyperphosphatasaemia at the Women’s & Children’s Hospital

M Freeman ¹, F Ku ¹, G Hinds ², MP Metz ¹

Introduction

Benign transient hyperphosphatasaemia (BTH) is a condition displaying a unique isoform of alkaline phosphatase (ALP) whose existence has been known for a long time, yet continues to mystify. We present a retrospective descriptive study of BTH in a children’s hospital during a period of 36 months.

Methods

All alkaline phosphatase results determined in our laboratory greater than 500 IU/L from a 36 month period were reviewed in regard to the presence of BTH. A diagnosis of BTH was made either by wheat germ lectin precipitation or neuraminidase digestion of serum followed by electrophoresis.

Results

Of 50,181 ALP results, 1246 were greater than 500 IU/L. 25 individuals had BTH. There was a clustering of cases in October to December and April to June. The mean and median values (IU/L) at presentation for BTH were 2596 and 2557 compared to 1770 and 1198 for non-BTH ALP. Time from presentation to decrease in ALP value to within the reference interval was 89 and 62 (mean and median) days with a range of 28 to 263 days.

Conclusions

BTH is predominantly a curiosity at this time. It is common and not known to be associated with any pathologic process. The seasonal clustering suggests the possibility of an infectious aetiology, although a number of non-infectious seasonal disorders are known.

Many clinicians are not familiar with BTH. By recognising this diagnosis and informing clinicians, unnecessary anxiety, investigations and follow-up can be avoided.¹

References

1.Carroll AJ, Coakley JC. J Paediatr Child Health. 2001;37:359–62. doi: 10.1046/j.1440-1754.2001.00686.x. [DOI] [PubMed] [Google Scholar]

CBR. 2009 Nov;30(4):S26.

P24 Naproxen Induced Hyperbilirubinaemia

G Dimeski ¹, B Jones ¹, G Marshall ¹, P Slater ², J Ungerer ¹

Introduction

A 39 year old female ingested 25 g of naproxen in a suicide attempt ~12 hours prior to presenting to the emergency department. Except for the unexplainably high total bilirubin (TB), 186 μmol/L on Beckman DxC600 (6 μmol/L on Roche Modular), the biochemistry results were normal. A sample taken 45 minutes later had TB of 206 μmol/L. There was no liver disease or other causes that may explain this significant elevation of TB to such an extent. The Beckman and the Roche TB methods are both Jendrassik and Grof diazo methods. It has been documented the major naproxen metabolite, Odesmethylnaproxen (ODMN) interferes with diazo methods on the Siemens Advia 1600 (superceded method) and Dade Behring RXL-Max analysers.¹

Methods

Normal serum was spiked with Naproxen (#N5160) (range 11.5–2500 mg/L) and ODMN (#UC305) (range 25–2500 mg/L). The samples were analysed on the Beckman DxC800 analyser, and interference was detected with ODMN only. It forms a reddish end product on addition of reagent B. The estimated ratio of formation of ODMN from naproxen is 100:1 in therapeutic doses when ingested in isolation.²

Results

From the 25 g of naproxen the patient took, 250 mg/L ODMN would have been formed, and in the spiked sample this resulted in 283 μmol/L of TB. Taking into consideration the half-life of naproxen (12–17 hours), the stated dose and time taken by the patient, the artefactual increase in the TB suggests the patient would have taken at least 25 g.

Conclusions

ODMN interferes with the TB method by forming a measurable coloured compound with reagent B compounds on the DxC analysers.

References

1.Al Riyami N, et al. Clin Biochem. 2009;42:129–31. doi: 10.1016/j.clinbiochem.2008.09.119. [DOI] [PubMed] [Google Scholar]
2.Vree TB, et al. Br J Clin Pharmacol. 1993;35:467–72. doi: 10.1111/j.1365-2125.1993.tb04171.x. [DOI] [PMC free article] [PubMed] [Google Scholar]

CBR. 2009 Nov;30(4):S26.

P25 Vitamin B12 Deficiency in Bahrain

A Madan ¹, N Das ², N Turaif ²

Introduction

Vitamin B12 deficiency is one of the most common vitamin deficiencies especially in pure vegetarian people. However, it is not common to find such findings in non-vegetarian people. Therefore, we chose a population that expected to have balanced efficient non-vegetarian diet to study this phenomenon. Bahrain has a good population sample for this study.

Methods

Vitamin B12 assay was performed in Salmaniya Medical Centre laboratory in 2008 where 4190 samples were analysed.

Results

Of 4190 samples, 207 had a B12 less than 132 pmol/L and considered deficient. Of these 10 patients (all Bahrainis) were repeated and found to be deficient. Of the 197 patients 173 are Bahrainis, 9 Indians, and 7 Pakistanis. Members from Ethiopia, Syria, Nepal, Jordan, Iran, and Sri Lanka are also in the deficient list. The diet pattern is non-vegetarian. 85% of them are women. This indicates that non-vegetarian diet does not prevent the people from vitamin B12 deficiency.

Conclusions

A diet rich in vitamin B12, based on red meat, chicken and fish, may not enough to prevent vitamin B12 deficiency. It appears that there is a hidden factor(s) amongst Bahrain population that lead to vitamin B12 deficiency other than dietary factors.

CBR. 2009 Nov;30(4):S27.

P26 An Unexpected Relationship Between Paracetamol and Bilirubin

M Rozkin ¹, C Cocotsi ¹, D Murphy ¹, Q Lam ¹

Introduction

Bilirubin interference of chemical assays for paracetamol has been reported previously. Here, we present our experience and recommendations for paracetamol determination in icteric samples on the Roche Integra.

Subject and Methods

A 40 year old previously healthy woman was referred to the emergency department with a 2 day history of painless jaundice and deranged liver function tests including a total (predominantly conjugated) bilirubin of 436 μmol/L and an ALT of 4486 IU/L. Ultrasound imaging did not demonstrate a structural cause for the cholestasis and further laboratory investigations excluded viral and autoimmune causes. Her medical history did not implicate any toxins or drugs. Paracetamol was measured by enzymatic assay (Roche Integra) and immunoassay (Beckman Coulter DxI) on icteric samples from our patient (both neat and diluted) and samples from other jaundiced patients.

Results

Serum paracetamol was found to be 95 μmol/L and an N-acetyl infusion commenced. Unusually, her serum paracetamol levels continued to rise over the next few days, peaking at 370 μmol/L. This paralleled a rise in total bilirubin, which peaked on the same occasion at 954 μmol/L. Serum was sent for paracetamol measurement by immunoassay which failed to detect the drug therefore confirming an interference.

Conclusions

Very high total bilirubin levels are typically not seen in paracetamol overdose but where the aetiology of fulminant hepatic failure is not clear, paracetamol measurement is usually requested. It is important that laboratories are aware of the possible interference of bilirubin in paracetamol analysis and that such interference is not seen in all enzymatic assays. We determined that for the Roche Integra, the icteric index could be used to identify samples where paracetamol measurement by immunoassay would be required.

CBR. 2009 Nov;30(4):S27.

P27 Utilising UPLC for the High Throughput Analysis of Amino Acids in a Clinical Setting

RJ Swenson ¹, BC McWhinney ¹, BM Walker ¹, B Sipinkoski ¹, AJ Wilce ¹

Introduction

Current physiological amino acid analysis techniques have a number of limitations including poor sensitivity, peak resolution, large solvent waste generation, long run times, and a large number of chromatographic interferences. The quantum leap in the resolving power of Ultra Performance Liquid Chromatography (UPLC) is utilised to enable the accurate quantitation of 42 amino acids and related compounds with a run time of 45 minutes.

Methods

Samples (25 μL) are deproteinised using 5-Sulfosalicylic Acid (containing Internal Standard Norvaline) and the supernatant is derivatised with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC). The derivatised sample is injected onto a UPLC system and the amino acids are separated using reverse-phase chromatography and detected at 260nm.

Results

The intra-run precision performed over a 1 day period and the inter-run precision performed over a 3 month period by 4 analysts was <2.5% and <5.0% respectively. Correlation of the Waters MassTrak AAA kit to our ninhydrin amino acid system gave coefficients (r²) ranging from 0.88–1.00. Over 95% of the samples with interference in the ninhydrin system did not exhibit any interfering peaks or baseline disturbances. Advantages over the ninhydrin system that we have found in our laboratory include:

Reduced sample volume, 25 μL versus 200 μL
Decreased runtime, 45 minutes versus 165 minutes
Increased sensitivity, 2 μmol/L versus 20 μmol/L
Decreased solvent waste generation
Urgent sample TAT of 3 hours and routine sample TAT of 24 hours

Conclusions

The Waters MassTrak AAA kit has proven to be a robust, reliable, sensitive, and reproducible method and is therefore suitable for use in a high throughput clinical setting.

CBR. 2009 Nov;30(4):S27.

P28 Utilising UPLC to Develop a Less Invasive Method for Monitoring Amino Acids in Maple Syrup Urine Disease Patients

RJ Swenson ¹, BC McWhinney ¹

Introduction

Current monitoring regimens for Maple Syrup Urine Disease (MSUD) patients have a number of limitations. Blood spot analysis using newborn screening mass spectrometry procedures do not quantitate the individual branched-chain amino acids (leucine, isoleucine and alloisoleucine) rather they provide a total leucine result. Plasma analysis can quantitate the individual branched-chain amino acids but requires regular phlebotomy. Utilising blood spot analysis on Ultra Performance Liquid Chromatography (UPLC) overcomes the need for regular venipuncture and facilitates timely and close patient monitoring.

Methods

Amino acids are extracted from the 3 mm blood spot using methanol (containing Internal Standard Norvaline). The supernatant is dried down, reconstituted using borate buffer and acetonitrile then derivatised with 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC). The derivatised sample is injected onto a UPLC system and the amino acids are separated using reverse-phase chromatography and detected at 260nm.

Results

A comparison of whole blood amino acids in samples with those from blood spots showed >90% recovery in the blood spot samples. Blood spot branched-chain amino acids levels were lower than plasma samples due to a dilution effect, however the levels mirrored those of plasma.

Advantages over current procedures include:

Less invasive to the patient, particularly to the paediatric population
Critical branched-chain amino acid levels are individually quantitated
Significant improvement for patients with specimen collection at home and sent via mail.

Conclusions

The UPLC method has been validated for the determination of amino acids utilising blood spots thus providing the ability to individually quantitate the branched-chain amino acids.

CBR. 2009 Nov;30(4):S27–S28.

P29 Familial Hypocalcaemia with Absence of Casr Mutation

CW Lam ¹

Introduction

The extracellular calcium-sensing receptor (CASR) is a plasma membrane G protein-coupled receptor that is expressed in the parathyroid gland and the kidney. Activating CASR mutations result in a syndrome of increased responsiveness of the parathyroid gland and kidney to extracellular calcium concentration - autosomal dominant hypocalcaemia (ADH; OMIM#601198). In our studied family, the youngest affected family member is 3 years old, and the oldest affected family member is 82 years old. The proband came to our notice after a routine blood analysis performed for an annual health check. In all ADH individuals, serum intact PTH levels and serum magnesium concentrations are normal.

Methods

Genomic DNA was extracted from peripheral blood samples with a QIAamp blood kit (Qiagen, Hilden, Germany). All the exons, flanking introns, and promoter, and UTR regions of the CASR gene were PCR amplified. PCR products were purified by Microspin S-300 HR columns (GE, Uppsala, Sweden), and both strands were sequenced using the amplification primers as sequencing primers and BigDyeDeoxy terminator cycle sequencing reagents (PE Biosystems, Foster City, CA). Products of sequencing reactions were purified by Centri-Sep spin columns (Princeton Separations, Adelphia, NJ). Purified sequencing fragments were separated by capillary electrophoresis and detected via laser-induced fluorescence on an ABI Prism 3100 genetic analyser (PE Biosystems).

Results

By bidirectional sequencing of the PCR products of all the exons, intronic flanking region, promotors, and UTR regions, no mutations of the CASR gene were identified.

Conclusions

In this study, we have identified a Chinese family with ADH but without detectable CASR mutation by mutational analysis. The ADH in this family is likely to be caused by a new ADH locus.

CBR. 2009 Nov;30(4):S28.

P30 Statistical Analysis of PSA Results from the Siemens Centaur

A Coleman ¹, H Martin ¹, K Sikaris ²

Introduction

Recent changes to Australia’s Medicare Schedule require laboratories to know the method specific, age related median and 97.5% for their total prostate-specific antigen (PSA) method. The changes have been introduced in order to optimise the use of free to total PSA ratio as a tool in the early diagnosis of prostate cancer. Local data was not available so we used database extraction and Bhattacharya analysis to provide this.

Methods

Gribbles Pathology data extraction included PSA results performed at the main laboratories in both Adelaide and Melbourne from November and December 2007. Data was extensively trimmed by using LabWizard comment codes to exclude results from patients being monitored for prostate disease. The remaining 30,858 points were analysed by non-parametric and Bhattacharya procedures.

Results

Non-parametric
Age	n	Median	97.5%
18 – 29	137	0.5	2.2
30 – 39	886	0.6	1.9
40 – 49	4,372	0.6	2.3
50 – 59	9,124	0.8	3.9
60 – 69	8,968	1.1	5.5
70 – 79	5,456	1.4	7.0
80+	1,903	1.6	7.8

Open in a new tab

Bhattacharya
Age	n	2.5%	50%	97.5%
18 – 29	137	0.25	0.7	2.1
30 – 39	886	0.30	0.8	2.2
40 – 49	4,372	0.30	0.9	2.5
50 – 59	9,124	0.30	1.0	3.5
60 – 69	8,968	0.35	1.2	4.5
70 – 79	5,456	0.35	1.4	6.0
80+	1,903	0.35	1.5	7.5

Open in a new tab

Conclusions

The close agreement between the age related values derived by non-parametric and the Bhattacharya analysis provides some confidence that the raw data was appropriately trimmed. Incorporation of this information into PSA billing will allow us to fulfil Medicare billing requirements for free to total PSA ratio and should facilitate rational use of this test in diagnosing prostate cancer at an early stage.

CBR. 2009 Nov;30(4):S28.

P31 Dynamic Endocrine Tests – Whose Responsibility?

JCG Doery ^1,², B Dayanath ¹, ZX Lu ^1,³

Introduction

Dynamic endocrine tests are important diagnostic procedures with both clinical and laboratory components. The clinical components include reviewing clinical indications for the test, appointment, advice on preparation required, current medication that may require temporary withdrawal, test counselling re risks and consent for procedure, calculation of drug dosage especially in children, administration of drugs or procedure e.g. exercise, Synacthen, oversight of specimen collection times, critical transport requirements, registration and processing, recording of clinical notes, test details and specimen relative times in the laboratory records, interpretive comments and discussion with requesting clinician where indicated. The laboratory components include standard analytical procedures, with special attention to avoid and check for collection time errors and specimen transposition, because of multiple samples for the same test on the same patient.

Methods

We audited the current requesting pattern for dynamic endocrine tests at Southern Health and evaluated the impact of developing a pathology based service on test volumes, efficiency of specimen collection and handling and capacity to provide interpretive reporting.

Outpatient tests included: growth hormone exercise, growth hormone and dexamethasone suppression, Synacthen, HCG, LHRH and CRH stimulation, lactose and xylose absorption tests.

Adrenal vein sampling and bilateral inferior petrosal sinus sampling were performed in conjunction with interventional radiologists. Prolonged fasting tests, water deprivation test and arginine-glucagon test were performed with endocrine registrar as a Day Admission. Total tests >160/annum.

Results

Tests are performed efficiently by staff who perform sufficient procedures to maintain expertise. Many Day Admissions are avoided. Error rates markedly reduced.

Interpretation greatly facilitated by correct patient preparation and clear clinical history captured in the LIS. Clinicians have increased requesting as the service is simple to access and reliably and efficiently performed.

Conclusions

Performance of dynamic endocrine tests by our Department has increased the clinical profile of Chemical Pathology, increased test volumes, improved the quality of the service and reduced overall cost.

CBR. 2009 Nov;30(4):S29.

P32 Evaluation of the Point of Care HBA1C IN2IT

K Lee ¹, K Tan ¹, M Conte ², P Williams ¹

Introduction

The Variant Boronate Affinity (VBA) method for HbA_1c is withdrawn from the market and interference in HbA_1c determinations using the Variant II NU (V II nu) method will still need to be addressed. We have evaluated the Bio- Rad In2it as a suitable method for dealing with this interference.

Methods

The V II nu, the VBA and the Bio-Rad In2it were used for all samples in this study. The reagent kits, columns and cartridges were used according to the manufacturer’s instructions. The In2it consumables were subsidised by Bio-Rad.

Results

All methods gave results that were highly correlated.

graphic file with name CBR_30(4)_S29_g001A.jpg

graphic file with name CBR_30(4)_S29_g001B.jpg

The difference and bias plots showed a positive bias in HbA_1c between the VII nu and the In2it of 0.23% and a negative bias of −1.35 % for the VII nu vs VBA and −1.12%. for the VBA and In2it. The 95% confidence intervals for concordance between the V II nu, the In2it and VBA were 0.96 to 0.99 (bias correction factor of 0.996) and 0.74 to 0.91 (bias correction factor of 0.88) respectively: between the VBA and the In2it it was 0.83 to 0.94 (bias correction of 0.91).

Conclusions

The In2it is a very capable replacement for the VBA. It provides HbA_1c values that are in excellent agreement with the V II Nu method and does not have significant bias. It provides a reasonable and simple solution for the determination of HbA_1c when interference is present. It is unaffected by HbF and HbJ and the absence of bias provides excellent agreement between methods that will not cause patient therapy to change.

CBR. 2009 Nov;30(4):S29.

P33 HPLC Separation of Macroprolactin

L Leong ¹, K Hoad ¹, J Abu Bakar ¹, P Sheehan ¹, S Vasikaran ¹

Introduction

All prolactin immunoassays commonly used in routine laboratories show cross-reactivity with macroprolactin, an IgG-complexed prolactin with limited bioactivity. A simple assay with polyethylene glycol (PEG) 6000 to precipitate macroprolactin can screen for its presence, while lengthier analysis and separation by gel filtration chromatography is often used for confirmation in specialised laboratories. We have separated macroprolactin easily and quickly by size exclusion high performance liquid chromatography (HPLC) using readily available instrumentation.

Methods

Patient plasma was diluted with the mobile phase (Tris/NaCl buffer, pH 7.2) and filtered before injecting onto a Zorbax GF-250 column (Agilent). The sample was eluted for 15 min with a flow rate of 1 mL/min. The first 7 mL were discarded and subsequent eluent collected (6 drops/fraction) using an automated fraction collector. Thirty four fractions were assayed for prolactin by immunoassay on an Immulite 2000.

Results

Macroprolactin eluted first, well separated from monomeric prolactin while other prolactin variants of intermediate sizes (big prolactin), eluted in-between. Elution profiles of molecular weight markers and IgG confirmed the sizes of the different prolactin species. Macroprolactin, monomeric prolactin and where present, intermediate size prolactin variants, showed consistent profiles.

Conclusions

Size exclusion HPLC offers a rapid and reliable method for the separation of macroprolactin from monomeric and big prolactin. It is more informative than PEG precipitation. It also demonstrates the heterogeneity present in the prolactin profile of many patients, particularly those with PEG screening results within the 40–60 % recovery zone.

CBR. 2009 Nov;30(4):S29–S30.

P34 Evaluation of the Access hTSH Assay as a Hypersensitive (Third Generation) and Fast (Second Generation) hTSH Assay for Clinical use

IA Bittar ¹, QT Lam ¹

Introduction

Third generation thyroid stimulating hormone (TSH) assays (functional sensitivity 0.01–0.02 mIU/L) facilitate the discrimination of hyperthyroid patients from those euthyroid with a subnormal TSH. Their performance is usually compared to older second generation assays where functional sensitivity approaches 0.1 mIU/L. Some newer “second generation” TSH assays achieve much better functional sensitivity and even approach those of third generation assays. The Beckman Coulter Access hTSH assay can be run as either a third generation or a faster (20 vs. 45 mins) second generation assay. We evaluated its performance in both modes for clinical use.

Methods

Analytical and functional sensitivity was determined for both assay modes on the Beckman Coulter DxI800 using patient sample pools and quality control material. Correlation and clinical agreement was assessed using 32 patient samples.

Results

Analytical (0.003 and <0.01 mIU/L) and functional (0.015 and 0.030 mIU/L) sensitivities for high sensitivity and fast hTSH assays respectively were consistent with manufacturer quoted figures. Ten patient samples with levels above 0.05 mIU/L showed good agreement. Among 22 samples with levels <0.05 mIU/L correlation was poor, although the largest absolute difference between paired results was 0.03 mIU/L. In three samples, high sensitivity TSH showed appropriate suppression (judged by corresponding fT4 and fT3 values) while fast TSH was detectable. However, these fast TSH results were still very low (0.04 mIU/L) and it is arguable whether this difference would adversely affect clinical interpretation.

Conclusions

The Access fast hTSH assay has a functional sensitivity approaching the high sensitivity hTSH assay but poorer (although not necessarily unsatisfactory) performance with low level samples. Laboratories should discuss with their clinicians whether the shorter analysis time and lower sample volume are worth the diminished sensitivity.

CBR. 2009 Nov;30(4):S30.

P35 Common TSH Cut-Offs – Should They be Applied?

J Calleja ¹, J Ryan ¹

Introduction

UK Clinical Practice Guidelines for use with thyroid function tests have recently been published. Guidelines focusing on thyroid dysfunction during pregnancy and postpartum have also been written. Both have put forward clinical management cut-offs or decision limits, in relation to thyroid stimulating hormone (TSH) results. Because inter-method bias is noted on External Quality Assessment Schemes, which may suffer from sample commutability issues, the Victorian AACB QC Sub-Committee investigated this for TSH assays using patient serum pools. TSH reference intervals used across methods were also examined.

Methods

Six fresh serum pools were prepared with target TSH levels of ~ 0.01, 1.0, 1.5, 5.0, 12 and 20 mIU/L. Two sets of samples were distributed to 15 labs in Victoria and Tasmania. Samples were thawed and analysed on separate days of analysis and assayed in duplicate, within two weeks of receipt.

Results

Results were collected from 15 laboratories involving six different instrument groups, including Roche Modular E170 (n = 2), Abbott Architect (n = 3), Bayer Centaur (n = 3), Beckman Unicel DXi (n = 3), Vitros ECi (n = 2) and Immulite 2000 (n = 2). Good agreement between laboratories using the same methods was noted, but inter-method bias is evident. Analysis of bias by Bland-Altman analysis for one sample pool (median TSH = 5.11 mIU/L) showed a 95% confidence interval of 3.88 to 5.6 mIU/L. The Modular E170 and Vitros ECI groups recovered values notably above their respective upper reference limit, whilst the Architect and Immulite groups recovered below. The Beckman and Bayer groups gave mixed results depending on what upper reference limits were employed.

Conclusions

Accordingly given the variability of recoveries noted and reference intervals employed, the application of common clinical decision cut-offs and/or reference intervals may be misleading.

CBR. 2009 Nov;30(4):S30.

P36 Relationship Between Plasma Free Cortisol/Cortisone and Salivary Cortisol/Cortisone Levels in Healthy Subjects

BC McWhinney ¹, SE Briscoe ¹, JP Galligan ¹

Introduction

11β-hydroxysteroid dehydrogenase (11β-HSD) is the enzyme which regulates mineralocorticoid-like action of glucocorticoid by the oxidation of active cortisol to inactive cortisone. The blood concentration of cortisol/cortisone has been shown to reflect the conversion of cortisol to cortisone thus providing an index of the systemic activity of 11β-HSD. In this study, we investigated the relationship between plasma free and salivary cortisol/cortisone levels.

Methods

Blood and saliva specimens were obtained from 24 healthy subjects between 0800 and 1000 h. Total, plasma free and salivary cortisol and cortisone levels were measured using a Waters Premier Tandem Mass Spectrometer coupled to an Acquity UPLC^TM (LC-MS/MS).

Results

The median plasma free cortisol and salivary cortisol levels for the group were 2.5 and 2.6 nmol/L with 5–95% confidence intervals of 1.2–6.2 and 1.1–7.7 nmol/L. There was not a significant difference between the analyses using the paired t-test (t = −1.301, p = 0.207). There was also no significant difference when the results were expressed as a percentage of plasma total cortisol. The median plasma free cortisone and salivary cortisone levels for the group were 3.6 and 19.9 nmol/L with 5–95% confidence intervals of 2.5–6.3 and 10.3–34.9 nmol/L. There was an obvious significant difference between the analyses with saliva levels being ~5x the concentration of plasma free levels.

Conclusions

Salivary cortisol levels correlate with plasma free cortisol levels when measured by LC-MS/MS. Given the relatively high levels of cortisone in saliva compared to plasma free levels, further studies are required to assess if these levels are the result of higher activity of Type 2 11β-HSD in tissues secreting saliva.

CBR. 2009 Nov;30(4):S30.

P37 Plasma Free Metadrenaline Determination Using LC-MS/MS

BC McWhinney ¹, B Sipinkoski ¹

Introduction

A number of studies have demonstrated the higher diagnostic efficacy of plasma free metadrenalines (PFM) in the diagnosis of patients with phaeochromocytoma. Traditionally the PFM assays were performed with HPLC and electrochemical detection due to its specificity and sensitivity. However the assays are labour-intensive and time consuming with long run times. A LC-MS/MS method with a novel extraction procedure has been validated that uses small sample volumes, is non-labour-intensive and has a shorter run time compared to published methods.

Methods

A Waters Premier Tandem Mass Spectrometer coupled to an Acquity UPLC™ was used for all analyses. MS control and data acquisition was performed using Masslynx v4.1 software. 200 μL of plasma was mixed with deuterated Internal Standard and extracted using Waters OASIS MCX Mixed mode SPE columns. The extracts are separated on a HILIC column using Ammonium Formate/Acetonitrile gradient. The appropriate multiple reaction monitoring (MRM) is then carried out for each analyte

Results

Excellent linear calibration curves were obtained for all analytes over the range 0.05 to 80.00 nmol/L (r² >0.995). Two levels of commercial controls used for the intra-day precision study (n = 10) gave <5%CV. Inter-day precision studies (n = 20) gave <7% CV. Comparison of patient sample results to a HPLC ECD method gave good agreement r² >0.950.

Conclusions

This method uses 2–5-fold less sample and has a shorter run time than other LC-MS/MS methods. The OASIS MCX SPE sample cleanup produces an extract that is chromatographically cleaner and has significantly less regions of ion suppression leading to improved sensitivity and precision.

CBR. 2009 Nov;30(4):S31.

P38 Analysis of 25-Hydroxyvitamin D in Serum using Semi-Automated Solid Phase Extraction and UPLC-MS/MS

LJ Calton ¹, BJ Molloy ¹, B Keevil ², DP Cooper ¹, S Wilson ³

Introduction

The demand for serum 25-hydroxyvitamin D (25OHD) analysis has increased significantly. 25OHD concentration is accepted as the clinical indicator for determining vitamin D status and is important for monitoring supplementation therapy which is available in two forms; vitamin D2 and vitamin D3. This method describes a semi-automated sample pre-treatment protocol in combination with UPLC-MS/MS for simultaneous analysis of 25OHD2 and 25OHD3.

Methods

For the initial study, twenty anonymised patient samples were analysed using an established manual liquid/liquid solvent extraction LC-MS/MS method at UHSM. Commercial calibrator and QC material (Chromsystems), primary samples and DEQAS proficiency testing samples were placed on a robotic liquid-handling system and identified by bar code to be tracked throughout the extraction procedure. The entire Oasis^® μElution SPE procedure is automated except for an off-line centrifugation step following protein precipitation. Eluted analytes were directly separated using an ACQUITY UPLC^® and detected by multiple reaction monitoring mass spectrometry (Waters ACQUITY^® TQD).

Results

The assay was linear over the range 6.25–250 nmol/L (r² >0.997). Inter- and intra-assay imprecision at three levels was less than 10% CV (n = 5, over 5 days). Calculated 25OHD concentrations of twenty patient samples were in good agreement with measurements from UHSM, using Passing-Bablok analysis Waters = 1.01(UHSM) − 3.6.

Conclusions

A routine semi-automated solid phase extraction UPLC-MS/MS method for the analysis of 25OHD2 and 25OHD3 in serum has been developed. The assay has acceptable performance characteristics (linearity, sensitivity, precision and accuracy) allowing the analysis of up to 192 samples per working day. This semi-automated procedure significantly reduces the manual operation steps, eliminates solvent evaporation and reconstitution to help minimise inter-operator variability.

CBR. 2009 Nov;30(4):S31.

P39 Quantification of 17-Hydroxyprogesterone from Dried Blood Spots using Liquid Chromatography Tandem Mass Spectrometry

C Rossi ¹, HA Brown ², LJ Calton ², SD Gillingwater ², S Wilson ³

Introduction

Congenital adrenal hyperplasia (CAH) is caused by inherited defects in steroid biosynthesis, in particular 21-hydroxylase deficiency, the classical form of CAH. Hormonal imbalances are reflected in decreased levels of aldosterone and cortisol and excessive secretion of 17-hydroxyprogesterone (17-OHP) and androstenedione. Diagnosis of CAH is based on the quantification of 17-OHP, usually by immunoassay. Compared with other neonatal screening tests, the specificity of screening for 21-CAH by immunoassays is low, characterised by high false-positive results. This is due to cross-reactivity with steroids other than 17-OHP, especially in pre-term neonates and in critically ill newborns. We have developed a simple method to determine the levels of 17-OHP in a dried blood spot (DBS) using ultra performance liquid chromatography tandem mass spectrometry (UPLC-MS/MS).

Methods

Patient and control DBS samples equalling 6 μL whole blood were extracted in acetone/acetonitrile. Extract was analysed by UPLC tandem quadrupole mass spectrometry (ACQUITY UPLC^® Xevo TQ MS) using multiple reaction monitoring acquisition. 17-OHP was quantified by calibration against DBS standards prepared at Glasgow Royal Infirmary from washed red blood cells (9–1270 nmol/L). Method precision and accuracy was ensured using CDC proficiency testing DBS material measuring 75, 150 and 300 nmol/L.

Results

The assay was shown to be linear over each analyte concentration range with all correlation coefficients (r² >0.997). Inter and intra-day imprecision of the assay was <6.5%CV and measured values were ±10 % of target across the analytical range.

Conclusions

A method for quantification of 17-OHP in DBS over the range of the clinical assay with good linearity, sensitivity and precision has been developed.

CBR. 2009 Nov;30(4):S31.

P42 Evaluation of the Diasorin Liaison Ostase Bone Alkaline Phosphatase Assay

A Kadiric ¹, H Martin ¹

Introduction

The DiaSorin Liaison assay for bone specific alkaline phosphatase (BAP) has been developed as an aid in management of Paget’s disease and osteoporosis. We sought to evaluate the assay’s analytical performance and to compare its utility with our routine alkaline phosphatase isoenzyme method.

Methods

Samples were sourced from those presented for routine alkaline phosphatase iosenzyme analysis supplemented by additional samples with isolated high ALP identified from routine biochemistry. The test method, DiaSorin BAP Ostase (310970), is a single step delayed addition sandwich chemiluminescent immunoassay. The routine method, Sebia Hydragel ISO-PAL (4132), is an electrophoresis method which separates alkaline phosphatase isoenzymes on the basis of charge and interaction with wheat germ lectin.

Results

A total of 66 patient samples were compared. Correlation between bone ALP activity determined by electrophoresis (eALP) and BAP mass was excellent, r² = 0.94. Correlation between total ALP measured by Advia 2400 (tALP) and BAP was good, r² = 0.78 and was better when pairs were selected on the basis of either raised BAP, r² = 0.81 or raised eALP percent, r² = 0.94. 24 of 27 samples (89%) with an eALP >50% had raised BAP; 100% could be achieved with minor changes to the manufacturer-defined cut-offs. The 42 samples with elevated BAP showed a median percent eALP of 60 (range 4–91). Nine of these had percentage eALP <40. Reproducibility for the BAP method was 6.7% at normal and 5.2% high QC levels.

Conclusions

BAP Ostase assay has acceptable precision and accuracy for routine use where bone specific ALP is clinically required. However electrophoresis remains preferred to answer the more general question of tissue source for (elevated) total ALP.

CBR. 2009 Nov;30(4):S31–S32.

P43 Procalcitonin Immunoassay on the Roche cobas e 411

JS Moshides ¹, M Moriatis ¹, VM Ellis ¹

Introduction

Procalcitonin (PCT), a prohormone, increases in blood during bacterial infection and is a marker for severe sepsis and septic shock. The Roche Elecsys Brahms PCT assay is five times more sensitive than the DiaSorin Liaison Brahms assay having a functional sensitivity of 0.06 ng/mL compared to the Liaison’s 0.30 ng/mL (detection limits are 0.02 and 0.04 ng/mL respectively). A higher sensitivity may allow the earlier detection of systemic infections. The Roche PCT assay time is 18 minutes. Our objective was to correlate the Roche PCT assay on the cobas e 411analyser with the Liaison analyser method.

Methods

Correlation was made by comparing results of patient specimens assayed on the e411 and the Liaison analysers. Statistical analyses for linear regression, correlation coefficient and Altman-Bland bias plots were made. Imprecision was calculated from between-day results of low level and high level commercial quality control materials.

Results

The mean bias between the e411and Liaison from 111 patient comparisons over a range of 0.03 to 98.58 was 0.01 ng/mL which is less than the detection limits of both assays. The linear regression equation was y = 0.9973x + 0.0825 and r was 0.9956. CVs on the e411 were 3.3 and 3.7% for PCT values of 0.49 and 9.68 ng/mL. CVs on the Liaison were 11.0 and 5.4% for values of 1.17 and 39.19 ng/mL.

Conclusions

The results indicate that PCT on the cobas e411 correlates well to the Liaison method and is much more precise especially near the clinical cut-off level of 0.50 ng/mL. Other advantages are five-fold greater sensitivity and acceptable assay time.

CBR. 2009 Nov;30(4):S32.

P44 Validation of DSL ELISA Assay for Anti Mullerian Hormone

L Krebs ¹, M Freemantle ¹, G Ward ¹, D Kanowski ¹, K Sikaris ²

Introduction

Anti-Mullerian Hormone (AMH) is expressed by granulosa cells of the ovary in the reproductive age and controls the formation of primary follicles by inhibiting excessive follicular recruitment by FSH. As levels fall during menopausal, reduction in ovarian reserve and high levels may be seen with ovarian hyperstimulation (or polycystic ovary syndrome), it is becoming an increasingly popular investigation in infertility testing. We compared two commercially available AMH kits.

Methods

We compared the results from 60 requests for AMH received mainly from IVF doctors. The two methods used were the Immunotech ELISA method and the DSL ELISA method which are both distributed by Beckman Coulter.

Results

There was a strong correlation (r² = 0.938) between the methods with a linear regression of DSL = 0.743 Immunotech - 1. The correlation between two separately analysed Immunotech results was r² = 0.90 with a linear regression of Immunotech 1 = 0.85 Immunotech 2 + 1.8. This reflected a poorer imprecision with batch to batch variability by the Immunotech method. The precision of the DSL method was slightly better CV = 6.6 % at 12 pmol/L and CV = 5.2% at 55 pmol/L.

Conclusions

We decided to use the DSL method for the following reasons (i) ease of use (ii) slightly better imprecision and (iii) the possibility that the older Immunotech method may be discontinued earlier. Nevertheless, the consistently lower results did require an adjustment of the clinical decision points from <14 to <10 pmol/L for reduced ovarian reserve and from >30 to >20 pmol/L for ovarian hyperstimulation. These new limits give a 92% concordance in classification for the samples tested.

CBR. 2009 Nov;30(4):S32.

P45 A Sertoli-Leydig Tumour in a Female: A Case Study

B Teis ¹, L Price ¹, G Ward ¹

Introduction

Androgen secreting neoplasms (ASN) account for 0.6 to 2.1% of female patients investigated for clinical features of hyperandrogenism. Approximately 50% of ASN are ovarian and Sertoli-Leydig tumours are a rare form of sex cord stromal tumours which represent less than 0.5% of all ovarian tumours. These tumours are characterised by testicular structures that produce androgens which are mostly unilateral and confined to the ovaries. In this paper we report a patient who had an ovarian ASN characterised as a Sertoli-Leydig cell tumour.

Case Report

A 53-year-old woman who two years previously had testosterone levels within the reference interval presented with hirsutism, hoarse voice and increased libido. On initial testing testosterone was 10.4 nmol/L (RI <3.2), TSH 3.2 mU/L, prolactin 189 mIU/L (RI <500), cortisol 650 nmol/L (RI 160–650), ACTH 14 ng/L (RI 9–51) and 17 OH-progesterone 14.6 nmol/L (RI 0.3–3.6). Cushing’s syndrome was excluded as serum cortisol suppressed from baseline 240 nmol/L to <35 nmol/L following administration of dexamethasone (RI <50). Oophorectomy and hysterectomy were performed and with macroscopic examination showing a nodule on the left ovary. Microscopically the nodule consisted of immature Sertoli cells with small ovoid nuclei and scant cytoplasm consistent with an ovarian Sertoli-Leydig cell tumour (intermediate grade). At follow-up testing, six weeks post-operatively the testosterone and 17-OH-progesterone levels were respectively 0.5 nmol/L and 0.9 nmol/L and within the reference interval.

Conclusions

The initial presentation of hirsutism is a common clinical presentation for women. This case has the usual presentation for Sertoli-Leydig tumours, with the patient first noting changes in secondary sexual characteristics. The clinical features and testosterone >7 nmol/L were suggestive of an ASN but only histology can confirm the tumour type.

CBR. 2009 Nov;30(4):S32.

P46 Evaluation of an Ultrasensitive Method for the Determination of Serum Thyroglobulin

CCG Wood ¹, P Ward ¹

Introduction

The major use of measuring the level of serum thyroglobulin (Tg) is in the management of differentiated thyroid carcinoma. After total thyroid ablation for papillary or follicular thyroid carcinoma, Tg should not be detectable, and its subsequent appearance signifies the presence of persistent or recurrent disease. The functional sensitivity of Thyroglobulin assays has improved 10-fold since their introduction as radioimmunoassays thirty years ago. Here we report on the evaluation of a commercially available Tg ELISA with further improvements in functional sensitivity.

Methods

RSR’s USTG ELISA is a sandwich assay in which Tg in test sera is captured by high affinity Tg antibody coated ELISA plate wells. This immobilised Tg is then detected by the addition of a second Tg antibody conjugated to horseradish peroxidase.

Results

There was a reasonable correlation between the Siemens Immulite 2000 Tg assay and the USTG ELISA (r = 0.87). Samples were diluted in parallel with the standard curve. Analytical sensitivity of the assay is 0.03 μg/L. Investigations of functional sensitivity revealed a CV of 8.8% at a dose of 0.08 μg/L suggesting a functional sensitivity of approximately 0.05 μg/L.

Conclusions

Royal North Shore Hospital has long been a specialist referral centre for Thyroid Cancer. Each month there is a formal review of those patients presenting in the previous month. Results generated by the USTG ELISA on patients with Tg of <2.0 μg/L as determined on the Siemens Immulite 2000 are now included in this formal monthly review. The RSR USTG ELISA represents a great clinical tool for the management of patients with recurrent thyroid tissue, and may alter current clinical intervention points.

CBR. 2009 Nov;30(4):S33.

P47 Simultaneous Measurment of Aldosterone and Cortisol by Semi-Automated HPLC-Tandem Mass Spectrometry

PJ Taylor ¹, RD Gordon ¹, M Stowasser ¹

Introduction

Aldosterone and cortisol are measured during fludrocortisone suppression testing and adrenal vein sampling. The aim of this study was to develop an HPLC-tandem mass spectrometric method for the simultaneous measurement of aldosterone and cortisol in human plasma, with potential application to the study of adrenal disorders.

Methods

Samples were prepared and analysed using a Symbiosis™ HPLC on-line sample preparation system coupled to a mass spectrometer. Plasma standards, controls and patient samples (200 μL) were pretreated with precipitation reagent containing d₇-aldosterone and d₄-cortisol (200 μL). The samples were mixed and centrifuged. Preconditioned Hysphere C18 HD cartridges (7 μm, Spark Holland) were loaded with the supernatant (250 μL). The cartridges were washed with 10% acetonitrile in 0.1% ammonium hydroxide (1 mL), 10% acetonitrile in 0.1% formic acid (1 mL) and 10% acetonitrile in water (1 mL). Chromatography was performed on a Sunfire C18 analytical column (50x3.0mm, 3 μm, Waters) under isocratic conditions at a flow rate of 0.3 mL/min. The mobile phase consisted of 35% acetonitrile/water. Mass spectrometric detection was by selected reaction monitoring (aldosterone m/z358.9 → 330.9; d₇-aldosterone m/z365.9 → 337.9; cortisol m/z361.0 → 331.1; d₄-cortisol m/z365.0 → 335.1).

Results

The assay had an analytical range of 69 to 1387 pmol/L and 69 to 1379 nmol/L for aldosterone and cortisol respectively (r² >0.996, n = 8). Inter-day accuracy and imprecision for quality control samples was 99.4 to 106% and <16 %, respectively (n = 10). The lower limit of quantification was 69 pmol/L for aldosterone and 69 nmol/L for cortisol.

Conclusions

This semi-automated, simultaneous method is highly accurate within analytical ranges that would be suitable for fludrocortisone suppression testing used to definitively confirm or exclude primary aldosteronism, and, with sample dilution, for adrenal vein sampling studies used to differentiate unilateral from bilateral forms.

CBR. 2009 Nov;30(4):S33.

P48 Evaluation of Two External Quality Control Schemes for Aldosterone

PJ Taylor ¹, DP Cooper ², RD Gordon ¹, M Stowasser ¹

Introduction

We tested our recently reported liquid chromatography-tandem mass spectrometry (LC-MS/MS) method for aldosterone against a series of standard issue quality control samples from the the Royal College of Pathologists of Australasia (RCPA) and German Society for Clinical Chemistry and Laboratory Medicine (DGKL) quality assurance schemes.¹

Methods

Samples obtained from the RCPA (n = 6) and DGKL (n = 5) were handled according to the respective recommended protocols and analysed in singlicate by LC-MS/MS. The results obtained by LC-MS/MS were compared with target concentrations assigned by each scheme.

Results

The range of measured aldosterone results for RCPA and DGKL samples were 221 to 1718 pmol/L and 209 to 1390 pmol/L, respectively. Linear regression analysis of the measured versus assigned target concentrations revealed the following equations: LC-MS/MS = 1.08*RCPA Target + 112 (r² = 0.997) and LC-MS/MS = 0.974*DGKL Target + 11 (r² = 0.996). The mean percentage bias (range) between LC-MS/MS and target concentration was 29.1% (11.3 to 67.5%) for RCPA and −1.8% (−7.8 to 4.5%), for DGKL.

Conclusions

Whereas a systematic bias was observed between LC-MS/MS results and the RCPA target values, excellent agreement was found for DGKL. This discrepancy could be attributable to the methodology used to assign target values: immunoassay for RCPA and gas chromatography for DGKL. Although the study sample size was small, these findings raise an argument for using reference chromatographic methods to assign target values for aldosterone external quality control schemes.

References

1.Taylor PJ, et al. Clin Chem. 2009;55:1155–62. doi: 10.1373/clinchem.2008.116004. [DOI] [PubMed] [Google Scholar]

CBR. 2009 Nov;30(4):S33.

P49 Vitamin D Analysis with the Roche E170

P Dunn ¹, MP Metz ¹

Introduction

The Vitamin D assay available for use with the Roche E170 is a random access immunoanalyser ready Vitamin D assay. The method is a competitive chemiluminescent assay. We have reported more than 25,000 results in routine use since August 2007. In addition to standard QC and QAP practice, on a day to day basis, Vitamin D results are all reviewed by a senior staff member. Results higher than the Bone Health Australia reference interval (50–140 nmol/L) published as MJA(2005) 182:281 are particularly reviewed. We report three artefactually increased results discovered this year.

Methods

The analytical range is from 10 to 250 nmol/L. For values greater than 250 nmol/L, the recommended diluent is low Vitamin D pooled human serum. When results higher than the reference interval are obtained, the requesting doctor is notified and asked about the clinical setting including Vitamin D supplementation. In the absence of clinical history supporting high Vitamin D values, the sample was analysed by LC-MSMS.

Results

Three outliers with values of 564 nmol/L, 345 nmol/L, and 487 nmol/L were discovered. All results repeated on reanalysis. Analysis of the serum by LC-MSMS revealed values of 154 nmol/L, 125 nmol/L, and 104 nmol/L.

Conclusions

We do not know the cause for these aberrant results. Interferences are well known in immunoassays. Interpretation must always consider any of a number of causes for interference in a result. This random access automated immunoassay for Vitamin D is a step forward in evaluating Vitamin D status but there are more steps left to make. We would like to thank Dr Malcolm Whiting for LC-MSMS analysis of the outliers.

CBR. 2009 Nov;30(4):S33–S34.

P50 Serum Testosterone in Females by Liquid Chromatography Tandem Mass Spectrometry

J Joseph ¹, EM Lim ¹, M Tanner ¹, S Mahathavan ¹

Introduction

Biochemical assessment of female hyperandrogenism depends on measurement of testosterone and SHBG. Automated testosterone immunoassays are non-specific nor sensitive enough in the female range.

Methods

A liquid chromatography tandem mass spectrometry (LCMSMS) assay for the measurement of testosterone in females was developed. Reference interval data was obtained on 94 post partum women not on oral contraceptive preparations and was compared to automated immunoassay (DPC Immulite).

Results

Testosterone was extracted from 200 μL of serum by simple solvent extraction using methyl tetra-butyl ether with deuterated (D2) testosterone as Internal Standard and standards spiked into charcoal stripped bovine serum. Following evaporation to dryness, samples were reconstituted with methanol/water, injected onto the LCMSMS and precursor/product ions for Testosterone and D2-Testosterone were monitored at 289.1/96.8 m/z and 291.1/98.8 m/z respectively. The CVs at 0.62, 1.34 and 6.58 nmol/L were 13.6%, 11.0% and 5.9% respectively. The assay was controlled with commercial quality control sera. The mean recovery over the range 2.6 to 10.1 nmol/L was 95.4%. The range of LCMSMS reference interval data was 0.18–2.0 nmol/L, the 2.5^th and 97.5^th percentiles were 0.3 and 1.9 nmol/L respectively. A female testosterone LCMSMS reference interval of < 2.0 nmol/L is proposed. The range for the DPC Immulite was 0.2–3.6 nmol/L, with 2.5^th and 97.5^th percentiles of 0.2 and 2.6 nmol/L respectively and the reference interval is <3.2 nmol/L.

Conclusions

The reference interval obtained is significantly lower than that of automated immunoassay and agrees with other published data. The method is rapid and cost effective for laboratories already equipped with this technology.

CBR. 2009 Nov;30(4):S34.

P51 Haemolysis and Plasma Polypeptide Hormone Sample Stability: Interactions with Pre-Analytical Time and Temperature

JH Livesey ¹

Introduction

Haemolysis is well known to interfere with the analysis of many plasma constituents, but there are few systematic studies of whether the pre-analytical variables of specimen temperature and time affect the degree of interference. These may be particularly relevant where enzyme activity (proteolysis) is suspected, thus we have examined the joint effects of haemolysis, temperature and time on the measured concentrations of nine polypeptides in plasma.

Methods

Human EDTA plasma (n = 8) containing up to 10% of haemolysed blood was incubated at 0º or 22º for up to 24 hours. ACTH, C-peptide, C-telopeptide, insulin and PTH were measured using the Elecsys 2010, and FSH, LH, prolactin and thyroglobulin using the Access2.

Results

Slight haemolysis (0.35 g/L Hb) resulted in a mean (±sem) 16.9±2.9% loss of ACTH after one hour at 22º and a 32.3±4.0% loss after 4 hours (both p<0.001). At 0º the comparable losses were 5.0±1.3% and 6.3±2.0% (both p<0.02). For insulin, the respective losses were 7.5±2.0% and 18.8±1.9% at 22º and 5.3±1.4% and 5.8±1.4% at 0º (p<0.01).

Moderate haemolysis (3.5 g/L Hb) resulted in a 13.6±2.1% loss of C-peptide after 6 hours at 22º or a 2.2±0.4% loss at 0º (p<0.001). For C-telopeptide the losses after 24 hours were 11.0±2.1% (p<0.01) and 4.4±1.7% (p<0.05) respectively. PTH loss was 9.8±0.9% (p<0.001) independent of time or temperature. FSH, LH, prolactin and thyroglobulin were not significantly affected.

Conclusions

The duration and temperature of exposure to haemolysis can be important pre-analytical variables for polypeptide analytes. ACTH and insulin are particularly susceptible. Ideally samples for these tests should be handled cold, and any with even slight visually detectable haemolysis must be flagged or rejected.

CBR. 2009 Nov;30(4):S34.

P52 Alpha-Fetoprotein – A Possible Marker of Regeneration of the Human Newborn Intestine

J Wojtulewicz ¹, J Coakley ¹

Introduction

Plasma α-fetoprotein (AFP) concentrations are very high at birth and decline steadily to reach adult levels by about eight months of age. Any increase in AFP levels in infancy is usually attributed to liver pathology, in particular hepatoblastoma which causes extremely high plasma concentrations of AFP. Disordered hepatic function due to biliary obstruction and total parenteral nutrition cause more modest rises. As endodermal cells produce AFP, we postulated that regeneration of gut following extensive surgical resection in infancy, might also result in post-natal rises in AFP.

Methods

We conducted measurements of AFP in four infants (gestations between 24 and 41 weeks) who had extensive bowel resections and prolonged treatment with total parenteral nutrition (TPN). Another infant with a bowel motility disorder who only underwent colostomy/ileostomy but received prolonged TPN acted as a control.

Results

The AFP results, with the number of days post first surgery in brackets, were as follows. Infant 1: 3000 kIU/L (94 days), 22,400 kIU/L (119 days); infant 2: 242,300 kIU/L (0 days), 338,700 kIU/L (6 days); infant 3: 3600 kIU/L (164 days), 4000 kIU/L (176 days); infant 4: 32,800 kIU/L (190 days); infant 5 (control): 23 kIU/L (167 days).

Conclusions

The results from the bowel resection babies show either an unexpected rise in AFP or AFP levels much higher than expected for age (corrected for gestation), whereas the control result is low. We contend that rises in AFP in this age period may not only be caused by pathology, they may also indicate regeneration of intestinal mucosal cells. Our findings have not been previously described and warrant further detailed studies.

CBR. 2009 Nov;30(4):S34.

P53 Stability of Vitamin C When Stored at −70 Degrees Centigrade

J Burns ¹, S Fotinos ², G Whittaker ², J Coakley ¹

Introduction

Vitamin C (L-ascorbic acid) in plasma is readily oxidised and must be stored, protected from light, at −70°C. Even so, it may degrade with prolonged storage. Our aim was to assess the stability of ascorbic acid when stored for 65 days at −70°C.

Methods

Heparinised plasma from six adult participants was placed into four separate tubes and protected from light. These tubes were then stored at −70°C, covered in foil. Three days after collection, the plasma from one tube from each volunteer was assayed for vitamin C by HPLC, with electrochemical detection. The CV for this assay is 5.7% at a concentration of 46 μmol/L and 3.8% at 88 μmol/L. The other stored plasmas were then assayed for vitamin C after 17, 32 and 65 days of storage respectively.

Results

For five out of the six subjects, plasma vitamin C did not decline between day 0 and day 32. (The day 32 sample was lost for the remaining subject). The results for the day 65 specimens decreased in all subjects. The percentage decrease over the mean of the previous results in each of the six subjects was as follows: 14%, 24%, 17%, 9%, 17%, and 11%.

Conclusions

Ascorbic acid in the plasma is stable when stored, protected from light, at − 70°C for 32 days. Thereafter some degradation occurs, so samples for vitamin C estimation should not be stored for more than 32 days at −70°C.

CBR. 2009 Nov;30(4):S34–S35.

P54 Improved Extraction and HPLC Assay of Fatsoluble Vitamins in Serum

LA Johnson ¹, SE Briscoe ¹, C Lam ¹, BC McWhinney ¹

Introduction

Serum levels of the fat-soluble vitamins (A, E, β-carotene & carotenoids) are decreased in malabsorption states which may require supplementation. Elevated Vitamin A and retinyl palmitate indicate toxic excess. A recent AACB Working Party survey found wide variation in assay techniques and results. One procedure allows their simultaneous assessment by HPLC but has a complex extraction step. We found high recoveries by direct isopropanol extraction, and modified the HPLC procedure to improve separation and shorten run time.

Methods

The extract from 200 μL of serum plus 1 mL of isopropanol (containing α-tocopherol acetate internal standard) is dried under N₂ at 37°C, and redissolved in 200 μL of mobile phase (25% ethanol, 75% acetonitrile). Analytes separated on a Waters™ Atlantis® C18 3μm 4.6 x 150 mm column are quantitated with a PDA timed-wavelength detector and two-point calibration (Chromsystems Calibrators for Vitamins A, E & β-carotene, in-house retinyl palmitate standard (Sigma R-3375).

Results

Direct extraction, the 3μm column, and timed-wavelength detection give baseline resolution of vitamins A, E, internal standard and retinyl palmitate with retention times (RT) 2.8, 7.4, 7.9 and 17.1 min, at wavelengths 330, 292, 292 and 330 nm, inter-assay CVs 5.9, 4.6, -, and 9.9 %, and recoveries 108, 103, -, and 97 % respectively. Reprocessing the data at 454 nm shows β-carotene (18.4 min RT, 16% CV, 101% R) separated from other carotenioids which are quantitated relative to the area of the β-carotene peak.

Conclusions

Our simple extraction and chromatographic modifications facilitate the simultaneous assessment of all the fat soluble vitamins of clinical interest with improved sensitivity, recovery and reproducibility.

CBR. 2009 Nov;30(4):S35.

P55 How Does Renal Function and Dosage Affect Serum Digoxin Levels?

ZX Lu ¹, KA Sikaris ¹

Introduction

Digoxin, a second line agent for the treatment of heart failure, is primarily eliminated as an unchanged drug through the kidneys. Thus, renal impairment prolongs its elimination half-time and increases toxicity risk. As most people with heart failure are elderly and often have chronic renal failure, we examined how renal function and different dosage affects serum digoxin concentration (SDC). Optimal SDC has recently been decreased from below 2.6 to below 1.3 nmol/L.

Methods

We reviewed 8 years of clinical laboratory data that included approximately 42,000 patients who had SDC and creatinine test at the same time. Of these, reliable information on dosage were provided for 1,519 patients (aged 81.4±9.7 years). Data are grouped according to dosage and renal function (estimated Glomerular Filtration Rate, eGFR).

Results

Median SDC (nmol/L) and proportion of patients with SDC ≥1.3 nmol/L in each group based on dosage and eGFR:

eGFR^*	Digoxin dosage
	1PG^**	2PG	3PG	4PG
≥90	0.6 (3.8%)	0.8 (25.0%)	1.1 (20.0%)	1.2 (47.4%)
61–90	0.8 (24.3%)	1.0 (34.0%)	1.4 (53.3%)	1.3 (52.4%)
31–60	0.9 (26.3%)	1.3 (52.5%)	1.5 (75.8%)	1.6 (69.0%)
≤30	1.2 (47.9%)	2.0 (87.2%)	2.3 (88.9%)	2.1 (71.4%)

Open in a new tab

Data are median (% patients with SDC ≥1.3 nmol/L);

^*

unit: mL/min/1.73m²;

^**

PG =Digoxin 62.5μg

Conclusions

Higher dosage and decrease eGFR result synergistically in higher SDC. In people with eGFR >90, dosage up to 4 PG/day (250 μg) is unlikely to result in SDC in the undesirable range. However dosage >2 PG/day in people with eGFR 61–90 or dosage >1 PG/day in people with eGFR 31–60 are likely to result in toxic SDC. In people with eGFR ≤30, any dosage >1 PG could almost certainly result in SDC in the toxic range.

CBR. 2009 Nov;30(4):S35.

P56 High Performance Liquid Chromatography Determination of Various Classes of β-Lactam Antibiotics for Monitoring Purposes in a Clinical Setting

SE Briscoe ¹, BC McWhinney ¹, T Hillister ¹

Introduction

Recently there has been an increasing demand from clinicians for measurement of β-lactam antibiotics in real-time, particularly from the Intensive Care Units. The use of serum concentration monitoring may aid in the determination of appropriate dosing and may allow the potential for favourable clinical outcomes while minimising the risks of treatment-related toxicity. We have developed a high performance liquid chromatography (HPLC) method for the determination in plasma of cephalosporins, penicillins and carbapenems.

Methods

The HPLC system consisted of a Waters 2695 Separation Module, and a Waters 996 PDA. A Waters XBridge C18 2.5 μm 4.6x30mm column was used for all chromatographic separations. Peak homogeneity and identification was aided by utilising peak purity algorithms and 3D spectral library matching within the chromatography software. 200 μL of plasma was combined with 100 μL of internal standard and 600 μL of acetonitrile. The mixture was vortexed and centrifuged. The supernatant was combined with 600 μL of chloroform, vortexed and centrifuged. A 10 μL aliquot of the upper aqueous phase was injected.

Results

Calibration curves were linear over the concentration range of 1 to 500 mg/L with correlation coefficients (r²) ≥0.998. Within-run precision (n = 10) across three levels was ≤3.1% CV. Inter-run precision (n = 10) was ≤6.9% CV and the limit of quantitation was 1 mg/L.

Conclusions

This method is simple, robust and provides the clinical laboratory with an effective means of determining the concentration of a number of frequently prescribed antibiotics in an efficient and timely manner. In addition, through utilisation of PDA spectral library matching and peak purity criteria, the confidence in correct peak identification, quantification and homogeneity is significantly increased.

CBR. 2009 Nov;30(4):S36.

P57 Should a Common Digoxin Toxicity Cut-Off be Applied?

J Calleja ¹, ZX Lu ¹, Q Lam ¹, D Rae ¹

Introduction

Several studies have shown that serum digoxin concentrations (SDC) ≥1.3 nmol/L are associated with increased mortality without additional clinical benefit. In response, many laboratories decreased their upper therapeutic limit (UTL) for SDC from 2.6 to 1.2 nmol/L. Having a uniform UTL requires agreement of results between laboratories but it is apparent from the RCPA Quality Assurance Program (QAP) that significant inter-method bias exists. The Victorian-Tasmanian AACB QC Subcommittee investigated whether this finding was also true in patient samples.

Methods

Three serum pools with different SDC were prepared. Two sets of these pools were transported frozen to 11 laboratories in Victoria and analysed on two separate occasions in duplicate (i.e. 4 data points from each lab). The eight digoxin methods tested were: Roche Modular E170 and P-Unit, Abbott Architect and AxSYM, Bayer Centaur, Beckman Unicel, and Olympus AU640/AU2700 using Microgenics DRI-reagent.

Results

There was good agreement between laboratories using the same methods. Summary of the serum digoxin results (n = 44) for each serum pool from 11 laboratories in Victoria are below:

	Pool-1	Pool-2	Pool-3
Mean (nmol/L)	0.6	1.4	2.5
SD	0.1	0.2	0.2
Inter-lab CV (%)	18.3	11.5	8.2
Min (nmol/L)	0.4	1.1	2.1
Max (nmol/L)	0.8	1.7	2.9

Open in a new tab

Conclusions

Marked inter-method bias similar to that noted in QAP was also observed in patient pools suggesting that the adoption of a common UTL across different methods may be misleading. Patients being monitored using an assay with a high bias are at risk of under-treatment whereas those monitored using an assay with low bias are at increased risk of toxicity. Standardisation of all digoxin methods by the manufacturers is required for better patient care.

CBR. 2009 Nov;30(4):S36.

P58 Development and Implementation of an Algorithm for Determining Future Gentamicin Dosage

H Robins ¹, JM Potter ^1,², M Woodward ¹, P Collingnon ^1,², K Daveson ³, G Koerbin ¹, PE Hickman ^1,²

Introduction

The control of gentamicin dosage to maintain therapeutic levels is poorly undertaken. It is believed that this is not only a local issue but one that is poorly undertaken nationally. We developed an algorithm that enables the laboratory to predict whether the gentamicin dosage is therapeutic or not and recommend any alteration to that dosage.

Methods

An algorithm was developed using last dose (mg), time and date of last dose, patient weight (kg), time of collection (hours post dose) and the gentamicin concentration (mg/L). Two calculation models have been developed. The first is a single level calculation using the gentamicin concentration at 6–8 hours post dose. This is used if the patient is non-pregnant, >15 years of age, has an eGFR >30 mL/min and is on once a day IV dosing. The second model is a 2 level calculation using the gentamicin concentrations at 30 minutes and 6–8 hours post dose. This model is used for patients with altered pharmacokinetics including those admitted to ICU, and patients with shock, burns, obesity, quadriplegia, cystic fibrosis or ascites.

Results

The uptake of this value added service was slow to commence however, it has now been accepted by most requesting physicians. Incomplete data is still being received by the laboratory preventing the provision of this service for all relevant gentamicin requests.

Conclusions

The efficacy of this service is still under evaluation. This algorithm is not useful for patients with renal impairment (eGFR <30 mL/min) or patients with infectious endocarditis.

CBR. 2009 Nov;30(4):S36.

P59 A Simple and Robust Method for the Simultaneous Determination of Whole Blood Cyclosporin a, Tacrolimus, Sirolimus and Everolimus Using LC-MS/MS. An Australian Perspective

BC McWhinney ¹, SE Briscoe ¹, ST Bildsten ¹, RJ Bahrdt ¹

Introduction

LC-MS/MS allows accurate, precise quantification of parent drug in the presence of metabolites and other interferences that impair the performance of other technologies. A strong demand for multi-analyte methods has arisen due to the increasing use of combination therapy and the need to streamline laboratory workflows.

Methods

A Waters Premier Tandem Mass Spectrometer coupled to an Acquity UPLC™ was used for all analyses. Whole blood (50 μL) was mixed with 0.1 M zinc sulphate and acetonitrile containing internal standards (4 μg/L d4-everolimus, 2 μg/L Ascomycin, and 25 μg/L d12-cyclosporin A), vortex and centrifuged. The supernatant was injected onto the LC-MS/MS system and a ballistic gradient is then introduced and the eluent flow is switched to the Premier Tandem Mass Spectrometer. The appropriate MRMs are then carried out for each analyte.

Results

Excellent linear calibration curves were obtained for all analytes using the Chromsystems calibrators (r²>0.995). Four levels of Chromsystem controls used for the intra-day precision study (n = 10) gave <5%CV. Inter-day precision studies (n = 30) gave <7% CV. Comparison of patient sample results to single analyte methods give good agreement. Tacrolimus (LC-MS/MS) r² = 0.995, Cyclosporin (HPLC-UV) r² = 0.992, Sirolimus (HPLC-UV) r² = 0.962 and Everolimus (TDx) r² = 0.865. External proficiency program samples were also assayed and all 4 analytes had very good agreement with target values (r²>0.992). The on-line sample cleanup allows the analytes to elute away from significant ion suppression regions.

Conclusions

This method is simple, robust and allows a clinical laboratory to assay multi-immunosuppressants within one run and provide same day results.

CBR. 2009 Nov;30(4):S36–S37.

P60 One Versus Two Blood Samples for Determination of Tobramicin AUC in Paediatric Cystic Fibrosis Patients

M Barras ^1,², H Alraman ¹, C Kirkpatrick ², M Harris ¹, C Dakin ¹, S Suresh ¹, D Urquhart ¹, M Pilbeam ¹, R Norris ^1,^2,³

Introduction

Intravenous tobramycin is a mainstay in the treatment of Pseudomonas aeruginosa in children with cystic fibrosis (CF) during pulmonary exacerbation. Ideally, doses should be adjusted to achieve an area under the time-concentration curve (AUC) of 70 to 100 mg/L.hr. Currently a minimum of two blood concentrations are required to estimate AUC. This adds to morbidity as blood collection is by finger-prick. This study provides preliminary data to evaluate the minimum blood samples required to achieve an accurate AUC using TCIWorks^®, a public domain Bayesian pharmacokinetic modelling program.

Methods

Retrospective data from CF patients on once daily intravenous tobramycin were used. Each patient had two serum tobramycin concentrations taken, on two separate occasions. Data from the first occasion were entered into TCIWorks^® to establish an individual pharmacokinetic model. Data from the second occasion were used to estimate AUC’s using one concentration versus using two concentrations. Patients included children (n = 14) where accurate dosing details had been obtained, and children (n = 16) where sampling/dosing times were obtained from medication charts.

Results

The correlation between AUCs for one sample and two samples was high (r² = 0.84, p <0.001) and bias (0.72 mg/L.hr (95 % CI = −10.1 – 8.7 mg/L.hr)) low.

Conclusions

If TCIWorks^® is used, these data suggest only one blood sample may be required to calculate an AUC once an individual patient model is established. A larger prospective study is required to confirm this finding.

CBR. 2009 Nov;30(4):S37.

P61 Fabrication of Flexible Re-Usable Conductive Silicone Rubber Electrodes for Pilocarpine Iontophoresis for Sweat Testing

JCG Doery ^1,², A Chao ¹, ZX Lu ^1,^2,³, F Meacco ⁴

Introduction

Cystic fibrosis is an autosomal recessive condition in which sweat chloride testing is still the definitive test for diagnosis especially when two abnormal genes are not apparent from routine genetic analysis. Testing consists of two distinct phases; collection of sweat after pilocarpine iontophoresis and then sweat chloride analysis by one of several techniques. Apart from the Macroduct sweat collection system no other commercial systems are available to Australian Laboratories for use. Most laboratories are using in-house fabrications built from various combinations of rigid metal electrodes and gauze or filter paper to hold the pilocarpine electrolyte. Rigid metal electrodes are inconvenient for small children where a soft flexible electrode is more desirable to get sufficient surface area exposed to electrolyte solution. Based on an in-house prototype produced some years ago by Pathology Queensland we have used commercially available materials to produce a flexible electrode suitable for pilocarpine iontophoresis in both adults and children.

Methods

The electrode base is conductive silicone rubber (Product code: RE3) obtained from Bio Electronics Pty Ltd, Niddrie, Victoria, 3042 sold for use as a ‘TENS’ electrode for cutaneous electrical stimulation. Other materials including rubber glue and non conductive rubber strips are basic items found in a typical handyman’s toolbox! Using the conductive rubber base (35x40 mm) a 3 mm wide and 3 mm high edging was then attached to the base creating a chamber into which pilocarpine soaked gauze can be placed. Electrodes were attached to the standard Gibson and Cooke electronic control unit.

Results

The electrodes required care and some dexterity to construct but within the capability of any laboratory technical officer. Electrodes were easily attached with Velcro type strips and easily washed and re-used many times. The fit on curved surfaces was superior to, and electrical conductivity equivalent to, our previous metal electrodes.

Conclusions

Flexible skin electrodes for pilocarpine electrophoresis can be conveniently constructed in-house from commercially available conductive silicone rubber. They are durable and more convenient to use than rigid flat metal electrodes especially in small children.

CBR. 2009 Nov;30(4):S37.

P62 An Automated Procedure for Performing High Volume Disaccharidase Testing Using Ultrasonication

M Baxter ¹, T Neville ¹, L Price ¹, R Flatman ¹, G Ward ¹

Introduction

Measurement of the duodenal disaccharidases (lactase, sucrase and maltase) can aid in the diagnosis of malabsorption disorders. To determine the disaccharidase activity, duodenal biopsy samples are homogenised, incubated with specific carbohydrate substrates (lactose, sucrose and maltose), and the reaction product of glucose then measured. A continual increase in requests for disaccharidase testing at SNP has necessitated the development of a more automated method of analysis. Traditionally biopsy samples were manually homogenised using a glass homogenising vessel with a Teflon tipped pestle. Problems with this methodology included:

Laborious and time consuming
Variability between samples and operators in attaining a consistent level of fine tissue homogenisation
Safety – breakage of glass vessels during process an infrequent but hazardous event.

Methods

Biopsy samples are loaded into 10 mL plastic tubes containing 0.7 mL of KCl buffer. Ultrasonication is achieved using a Sonics Vibra Cell Ultrasonicator, set at 65% amplitude for 15 seconds. Ultrasonication is carried out in a separate closed room, with the use of ear protection in addition to usual Personal Protective Equipment. Homogenates are centrifuged and loaded onto a Roche 912 analyser for measurement of disaccharidase activities using a modified method derived from IMVS Adelaide.

Results

Ultrasonication consistently resulted in a finer regular homogenate than manual techniques. Variation between disaccharidase activity for multiple biopsies from a single patient makes statistical comparison of disaccharidase methods challenging. For example the Deming equation for Lactase was Sonication = 0.78 (±0.26) Homogenisation. Similar slopes with a wide 95% confidence interval were observed for Sucrase and Maltase. The clinical utility of enzyme methods was unaffected by the introduction of ultrasonication.

Conclusions

The use of ultrasonication for homogenising dissacharidase duodenal biopsy samples has increased reliability and safety, while reducing testing time and relieving scientists from performing laborious manual procedures.

CBR. 2009 Nov;30(4):S37–S38.

P63 Evaluation of the Quanta Litetm H-Ttg/Dgp Screen Elisa Kit

J Bartle ¹, P Vervaart ¹

Introduction

We evaluated the QUANTA Lite^TM h-tTG/DGP Screen (Inova Diagnostics) to determine if it could be used to distinguish patient samples requiring referral for coeliac serology testing from normal patient samples.

Methods

Serum samples requiring coeliac serology testing were sub-aliquotted and stored at −20°C until analysis. 32 patient samples, 13 RCPA QAP samples and 3 controls were tested in accordance with the ELISA kit instructions. Duplicate sub-aliquots of patient samples were referred to another pathology laboratory for anti-tissue transglutaminase (IgA) and anti-deamidated gliadin peptide (IgA) testing. Results obtained by the referral laboratory and by RCPA QAP report were compared with results from the QUANTA Lite^TM kit.

Results

The screen results were 30 negative and 15 positive tests. This was consistent with external results in all but 2 cases. Both exceptions were negative by the screen; one gave a borderline positive result on referral (re-testing in 3–6 months was recommended). The other was a QAP sample positive for gliadin (IgA), which the RCPA suggest is likely to represent a false positive result. From our data, the QUANTA Lite^TM kit is calculated to have a sensitivity of 100% and a specificity of 93%.

Conclusions

It is anticipated that the screening kit could be used to save money by eliminating the referral of normal samples for coeliac serology testing (82% of samples referred by our laboratory test negative). Samples screening positive would still be referred for testing. We have accepted the QUANTA Lite^TM h-tTG/DGP IgA/IgG Screen kit for routine use in our laboratory. Negative screen results are reported using the RCPA recommended comment.

CBR. 2009 Nov;30(4):S38.

P64 Evaluation of Abbott Next Generation Calcium for Renal Dialysis Patients

D Armbruster ¹, L Legendre ²

Introduction

Study compared Abbott Next Generation Calcium (Next Gen Ca), on-market Abbott Calcium (Ca), and Randox Ca (R-Ca) assays for imprecision, method comparison, and suitability for renal dialysis patients.

Methods

The University of Virginia Medical Laboratories (UVA) evaluated Next Gen Ca assay on the ARCHITECT ci8200 analyser for a large population of renal dialysis patients. Sporadic falsely increased Ca results for dialysis patients prompted specific evaluation of factors that might affect Ca values. Calcium assays were evaluated for 5-day precision at two control levels. Method comparison examined Next Gen Ca, Ca, and R-Ca reagents for 199 serum and plasma samples, 103 from renal patients (pre and post dialysis) including 25 geriatric patients. Other samples consisted of 25 pediatric and 32 geriatric samples, 22 samples spiked with high concentrations of magnesium (common potential interferent), and 17 normal samples. All samples were analysed using the three reagents simultaneously.

Results

Next Gen Ca, % CV = 1.0 at 2.18 mmol/L, % CV = 1.4 at 2.83 mmol/L (5 day precision). Next Gen Ca vs. Ca: y = 0.97x + 0.1 mmol/L, r = 0.994. Next Gen Ca vs. R-Ca: y = 1.02x + 0.08 mmol/L, r = 0.998. R-Ca vs. Ca: y = 1.00x + 0.15 mmol/L, r = 0.995.

Conclusions

Abbott Next Gen Ca demonstrated good precision, lacked interference from high magnesium concentrations (critical in renal patient populations), and serum and plasma samples (including those from dialysis patients, pediatric patients and geriatric patients) correlated well with other methods. The large kit size meets the workflow needs of large volume laboratories such as UVA, allowing for less operator intervention.

CBR. 2009 Nov;30(4):S38.

P65 Evaluation of Salivary Amylase Measurement

KE Song ¹, JR Park ¹, YK Kim ¹, SJ Lee ², KI Ha ³

Introduction

Saliva is increasingly being used as a specimen for systemic disease, oral health as well as an index for psychological stress. Authors evaluated the measurement of salivary amylase activities collected by Salivette (Sarstedt, Germany) as a biological marker of stress.

Methods

Forty-six healthy subjects participated in this study. For reproducibility and recovery studies, we used diluted samples from pooled saliva. For the stress experiment, saliva was collected between 9:00 to 11:00 a.m. once before and three times after to 5-min stressor consisting of 20 negative affective pictures. Amylase was measured by Dimension RxL Max (Dade Behring Inc., USA).

Results

Intra-run CVs for three levels were 1.59%, 1.63%, and 1.07% respectively. Dosing Salivettes with 2 mL, 1.5 mL, and 1 mL yielded sample recovery 85.5±2.4%, 82.4±1.5%, 72.2±3.1%, respectively and amylase recovery 78.9±10.9%, 74.1±13.7%, 37.3±26.9%, respectively. Salivary amylase was significantly decreased from baseline 64.2±7.1 U/mL to 54.0±7.0, 53.7±7.8, 50.0±7.2 over time with the negative affect score and state-anxiety.

Conclusions

Measurement of salivary amylase using Salivette had good CVs. More than 1.5 mL of saliva would be needed to have more than 70% recovery of Salivette. For the stress experiment, the baseline level of salivary amylase and its reactivity depend on individual trait of anxiety as well as state of reactivity to affective stress.

CBR. 2009 Nov;30(4):S38.

P66 Effect of Haemolysis on Zinc Levels in Edta Plasma

M Baxter ¹, P Moore ¹, G Ward ¹, D Kanowski ¹, R Flatman ¹

Introduction

We recently introduced EDTA trace element collection tubes for plasma metals (aluminium, copper, selenium, zinc, strontium, manganese). It was observed that some collections particularly short collections from children were slightly haemolysed which would routinely result in a recollection for zinc. This is because zinc is present in red cells at around 20 to 30 times the concentration of zinc in plasma. Other elements analysed in plasma exist in relatively the same concentration within red cells and are not affected by haemolysis. A reduction in recollections due to contamination was one of the main reasons for the introduction of EDTA Tubes. This study investigated the effect of gross haemolysis on the level of zinc.

Method

A patient sample collected into an EDTA trace element tube was centrifuged and the plasma was analysed for zinc. 1000 μL aliquot of plasma was removed from the tube and mixed with 10 μL and red cells from the original tube. This subsequent tube was vortexed to haemolyse the cells. This sample was analysed for zinc also.

Results

The baseline zinc in plasma was 11.2 μmol/L. The zinc concentration increased to 13.1 μmol/L in the haemolysis sample. The haemoglobin in the haemolysed sample was measured at 4 g/L and is considered gross haemolysis.

Conclusions

The concentration of zinc in red cells is approximately 200 μmol/L and therefore the addition of 1% red cells should have increased the plasma zinc concentration by about 2 μmol/L which is almost exactly what was observed. This manufactured sample represented a substantial amount of haemolysis and therefore any increase in plasma zinc attributed to ‘micro-haemolysis’ would not contribute significantly in the measurement of plasma zinc.

CBR. 2009 Nov;30(4):S38–S39.

P67 Implementation of Becton Dickinson Edta Trace Element Collection Tubes for Heavy Metals

P Moore ¹, M Baxter ¹, D Kanowski ¹, G Ward ¹, R Flatman ¹

Introduction

The number of trace metal tests performed in the laboratory is constantly increasing. Tests for zinc, aluminium and chromium are easily contaminated by instrument sample probes, incorrect specimen handling and incorrect tube collection. We investigated a single collection tube for metals to avoid contamination. The proposal was to collect samples for serum and whole blood metals into Becton Dickinson (BD) trace element collection tubes with potassium EDTA anticoagulant. This collection would require a switch from serum to plasma analysis for elements aluminium, cobalt, copper, manganese, nickel, selenium, strontium, and zinc. The outcome of this simple investigation to reduce the number of collection tube errors proved to be a significant exercise in LEAN engineering for the ICP-MS section.

Methods

Several patients presenting for zinc or manganese collections had samples collected into both plain (no additive) and EDTA trace element tubes. Specimens were analysed for the serum profile on the Agilent 7500CE ICPMS to ascertain if a difference existed between the level of elements in serum and plasma.

Results

The results showed no significant difference between serum and plasma. The within run variation for the assay was greater than the observed difference between the plasma and serum data.

Conclusions

From the data it was concluded that serum and plasma from the trace element tubes were equivalent for these elements allowing the collection of a single tube of blood for any trace metal analysis performed in our laboratory. Having a dedicated tube for the metals significantly saved processing time through specimen reception as well as an almost complete reduction in samples contaminated by processing through other analysers or departments. It also greatly reduced the number of recollections of serum zinc due to SST rubber stopper contamination.

CBR. 2009 Nov;30(4):S39.

P68 Stability of Trace Metals in Edta Trace Element Tubes: A 24 Hour Study

M Baxter ¹, P Moore ¹, D Kanowski ¹, T Neville ¹, G Ward ¹, T Yen ², W Marshall ¹

Introduction

Previous studies in the literature on trace element tube type and the time between collection and separation of plasma from red blood cells for zinc analysis were performed on heparin tubes. These indicated an increase in plasma zinc of up to 6.3% in 1 hour and 40.7% in 24 hours for samples at room temperature prior to separation. There is no published data on the stability of plasma trace elements in EDTA trace element tubes. This study was performed to evaluate the EDTA tubes.

Methods

One staff member was bled and two Becton Dickinson (BD) EDTA plastic trace element tubes were collected. One of the tubes was stored at room temperature, and the other was stored at 4°C. An aliquot was removed immediately, centrifuged and the plasma separated into an Abbott TDX centrifuge tube. The samples were gently mixed twice by inversion and further aliquots removed, centrifuged and plasma aliquotted into TDX centrifuge tubes at 1 hour, 2 hours and 24 hours. All of the plasma aliquots were stored at 4°C until analysis on the same run. The samples were diluted and analysed on an Agilent 7500ce ICPMS for aluminium, manganese, nickel, cobalt, zinc, copper, selenium and strontium.

Results

The results show a 0.3% increase in plasma zinc at 4°C, and 1.5% at room temperature over 24 hours. Plasma copper increased 3.3% at 4°C and 4.7% at room temperature over 24 hours. There was no significant increase in any of the other elements analysed.

Conclusions

The BD EDTA trace element tube is a suitable tube for the collection of samples for plasma trace metal analysis. The unspun EDTA samples can be kept at 4°C for 24 hours unspun without a significant rise in the analytes measured.

CBR. 2009 Nov;30(4):S39.

P69 Zinc Stability in Becton Dickinson Trace Element Tubes: A 72 Hour Study

M Baxter ¹, P Moore ¹, D Kanowski ¹, T Neville ¹, G Ward ¹, T Yen ²

Introduction

Previous studies in the literature on trace element tube type and the time between collection and separation of plasma from red blood cells for zinc analysis were performed on heparin tubes. These indicated an increase in plasma zinc of up to 6.3% in 1 hour and 40.7% in 24 hours for samples at room temperature (RT) prior to separation. There is no published data on the stability of plasma zinc in EDTA trace element tubes. This study was performed to evaluate the EDTA tubes.

Methods

Blood from four volunteers was collected into the following Becton Dickinson (BD) trace element tube types: plastic EDTA; plain glass; and sodium heparin glass. Blood was stored at 4°C and RT. Blood aliquots were removed from each tube type at time-points post collection as follows: 0 h; 2 h; 6.5 h; 24 h; 48 h and 72 h. Aliquots were centrifuged and plasma/serum stored at 4°C until zinc analysis on the same run by an Agilent 7500ce ICPMS.

Results

Table.

Percent increase in plasma zinc at 4°C and RT with different collection tubes and varying times between collection and separation.

Time	EDTA		Sodium Heparin		Plain Glass

Hrs	4°C	RT	4°C	RT	4°C	RT
2	0.9	3.2	2.1	9.4	2.6	9.0
24	4.0	9.8	12.5	50.9	8.3	21.9
48	7.1	13.2	14.2	54.5	12.2	34.8
72	5.8	19.2	18.1	68.3	16.3	52.0

Open in a new tab

Conclusions

The BD EDTA trace element collection tube is suitable for plasma zinc analysis. Unspun EDTA samples can be kept at 4°C until analysis, but samples should be spun and aliquotted if temperature cannot be maintained at 4°C between collection and sample analysis. Plain glass or sodium heparin collection tubes should be centrifuged and samples aliquotted within two hours of collection with blood optimally stored at 4°C.

CBR. 2009 Nov;30(4):S39–S40.

P70 Evaluation of Aliquotting Plasma Samples into Becton Dickinson Trace Element Tubes

M Baxter ¹, P Moore ¹, D Kanowski ¹, T Neville ¹, G Ward ¹, T Yen ²

Introduction

Pre-analytical contamination of samples is a common problem for trace element analysis. To prevent this we have chosen to aliquot samples into trace element collection tubes to ensure samples are forwarded direct to the inductively coupled plasma mass spectrometry (ICPMS) area. Becton Dickinson (BD) trace element sample tubes were evaluated to determine which was most suitable to contain plasma sample aliquots for trace metal analysis.

Methods

BD Trace Element Tubes evaluated were: EDTA; Plain Glass and Plain Plastic. One ml of an EDTA plasma pool was added to duplicates of each trace element tube type. The original pool and each of the aliquots were tested for Al, Mn, Co, Ni, Cu, Zn, Se and Sr on the same analytical run on an Agilent 7500ce ICPMS.

Results

The results showed no significant increase above the level in the original pool for all elements for either the EDTA trace element tube or the plain glass trace element tube. The plain plastic trace element tube showed significant elevations for both aluminium and manganese.

Conclusions

Plain glass and EDTA trace element tubes are suitable containers for holding plasma trace element sample aliquots. The EDTA tube is preferred as this eliminates the need for other trace element tube types to be kept at collection centres. Confusion between different tube types held in collection centres is a risk that can lead to recollections when the incorrect sample is collected for whole blood trace elements. The plain plastic trace element tube is unacceptable and probably due to contamination from clot activators. ICPMS is a multielement technique and receiving a plain plastic trace element tube containing a plasma aliquot may lead to incorrect results reported for plasma aluminium and manganese. We consider that the EDTA trace element tube is the safest option for aliquot storage for reliable plasma trace element testing.

CBR. 2009 Nov;30(4):S40.

P71 Laboratory Automation and the Pvt Rsd Pro

CV Bell ¹, DR Fredline ¹, GJ Ward ¹, RJ Flatman ¹, C Ison ¹, P Norman ¹

Introduction

The PVT RSD Pro sorting system is a computer-controlled stand-alone device for sorting bar-coded sample tubes. Our laboratory recently introduced two PVTs to accommodate for the significant rise in pathology requests and urgent need for further automation of sample processing for automated chemistry (Roche Modular) and immunoassay (Abbott Architect). Implementation of the PVT has resulted in changes in staff utilisation.

Methods

Prior to PVT implementation, samples were sorted by specimen reception area (SRA), placed through the Pathfinder then redirected to the Modular or Architect work benches. Specimens were further logged, sorted and manually stored. After PVT installation, SRA send samples directly to PVT which sorts Modular and Architect samples. The PVT also redirects, caps and stores samples.

Results

Implementation of the PVT RSD Pro has assisted in automation within the laboratory through removing majority of the sorting and storing of samples. While the PVT sorts majority of our work, some specimens still require manual handling. Samples which are not in standard SST tubes or tipped into cups cannot be processed by the PVT. Shared specimens between Modular and Architect have reduced collection tubes by 50%.

Conclusions

Pathfinder aliquot tubes are still required to avoid HCG and infectious serology carryover however a third Pathfinder no long needs to be purchased. There has been a 30% reduction in capping and 50% reduction in the number of staff required to process samples. Biochemistry and Endocrinology sorting benches have been integrated into one working bench and lab assistants now perform some of the roles previously done by scientists.

CBR. 2009 Nov;30(4):S40.

P72 Precision Profile for Urine Sodium Analysis: Integra vs Modular

D Masters ¹, J Calleja ¹, ZX Lu ¹, KA Sikaris ¹

Introduction

Urine sodium (UNa) concentration is helpful for investigation of the causes of hyponatraemia (UNa ≤20 mmol/L: gut loss; UNa >20 mmol/L: renal loss). Thus, assay accuracy and precision at ~20 mmol/L is important. Many assays, as indicated in the RCPA Quality Assurance Program however, have bias and poor precision at low UNa concentrations. In addition, UNa concentration of commercially available urine quality control materials are usually too high to monitor precision around this cut-off. Accordingly, we examined the bias and precision profile of UNa using the Roche Integra-800 and Modular-P. The analytical sensitivity claimed by Roche is 20 and 10 mmol/L, respectively, for Integra 800 and Modular-P.

Methods

An aqueous standard of 500 mmol/L NaCl was gravimetrically prepared. From this, eight working solutions with sodium concentrations of 10, 15, 20, 25, 30, 50, 150 and 250 mmol/L were prepared. The samples were tested on both a Integra-800 and Modular-P analysers.

Results

Table.

Precision profile of urine sodium using Integra and Modular.

	Urine sodium concentration (mmol/L)
	10	15	20	25
Integra	N/A	N/A	22 (5.9%)	27 (5.4%)
Modular	7(25%)	13 (16%)	17 (12%)	23 (7.5%)

	Urine sodium concentration (mmol/L)
	30	50	100	150
Integra	32 (3.3%)	52 (3.9%)	101 (4.7%)	150 (3.0%)
Modular	29 (7.9%)	50 (3.4%)	100 (1.8%)	148 (1.2%)

Open in a new tab

Data are mean (CV%), CV = coef cient variation. n = 7

Conclusions

Integra gives higher values than Modular-P for UNa up to 50 mmol/L (p <0.05) but aligned well with Modular-P for UNa ≥100 mmol/L. At UNa cut-off of 20 mmol/L, Integra has a positive bias and values varying up to 24 mmol/L (1.56 x CV) could be obtained. Modular has a negative bias, thus any value >21 mmol/L (1.56 x CV) is significantly different from the cut-off.

CBR. 2009 Nov;30(4):S40.

P73 Interference in the Measurement Of The Plasma Lipaemic and Icteric Indices on the Beckman Coulter Dxc Chemistry Analysers

AI Bransden ¹, AE Coriat ²

Introduction

A number of patient lithium-heparin plasma samples analysed for general chemistry on the Beckman DXC800 and DXC600i analysers in Pathology QLD laboratories were found to give spuriously raised results for the lipaemic and icteric indices. The plasma haemolytic, icteric and lipaemic indices are measured spectrophotometrically on the Beckman DXC analysers after addition of a diluent to an aliquot of sample. Investigation of the samples was undertaken to determine the possible causes.

Methods

Available lithium-heparin plasma and serum samples from four patients were examined after reaction with the diluent used in the performance of the indices on the Beckman DXC analysers. Electrophoresis was performed on available samples of plasma and serum from four of the patients showing the raised indices.

Results

Of the four patients, one was found to have an IgG paraprotein and three were found to have an IgM paraprotein. All of the patients’ lithium-heparin plasma samples developed an obvious turbidity when mixed with diluent indicating that the paraproteins may have precipitated on reaction with the diluent. Serum samples available on two of the patients did not develop turbidity when mixed with diluent.

Conclusions

IgG or IgM paraproteins found in some patient lithium-heparin plasma samples may react with the diluent used for the plasma indices on the Beckman DXC analysers to produce turbidity that interferes in the spectrophotometric measurement of the plasma lipaemic and icteric indices. Serum samples from the same patients may not react in the same way.

CBR. 2009 Nov;30(4):S40–S41.

P74 Contaminated Heparin in Blood Collection Tubes – Rapid Screening for Possible Effects on Test Results

GRD Jones ¹, R Wood ²

Introduction

In January 2008 it became known that there was contamination of therapeutic heparin leading to patient deaths in the USA. The contaminant was identified as over-sulfated chondroitin sulphate (OSCS), a heparin-like molecule not detected by the usual heparin purity checks. The TGA in Australia raised the question of possible contamination in heparin-containing blood tubes which could affect pathology results and sought assistance from the RCPA in assessing the risk of clinical errors.

Methods

Following the identification of OCSC-contaminated tubes by a manufacturer, a plan was rapidly formed to assess possible analytical interferences. Using volunteer laboratories, 5 pairs of matched pooled serum samples in contaminated and uncontaminated heparin tubes were distributed and measured for 54 analytes on up to 7 analytical systems. A number of further contaminated tubes and capillary collection devices were also tested with smaller studies. A screening criterion of an average difference less than the commonly used reporting interval, or less than ¼ of the RCPA QAP Allowable Limits of Performance was used.

Results

A total of 2730 measurements were made in the initial study. The screening criteria identified four analyte/analyser combinations with potential interference. For three of these additional testing was possible and interference was not confirmed. The findings of the study were reviewed at the TGA and by manufacturers and international regulatory bodies.

Conclusions

A rapid screen of multiple analyte/analyser combinations did not reveal any significant interferences. This data provided reassurance to the TGA and other agencies and professional bodies that the tubes were unlikely to cause clinical errors. The authors gratefully acknowledge the participating laboratories and scientists.

CBR. 2009 Nov;30(4):S41.

P85 Innovative Web-Based Training and Competency Program for Poct Device Operators in the ‘Qaams’ Program

BC Mazzachi ¹, MDS Shephard ¹

Introduction

The Quality Assurance for Aboriginal and Torres Strait Islander Medical Services (QAAMS) Program is a national point-of-care testing (PoCT) program for diabetes management which is operating in over 100 Aboriginal medical services across Australia. Aboriginal Health Professionals perform PoCT for Haemoglobin A_1c (HbA_1c) and urine albumin:creatinine ratio (ACR) on the Siemens DCA. A major on-going challenge for the program, given its large participant base, is the training and competency of device operators. Annual, regional and onsite training is available but, in a recent innovation, a web-based training and competency assessment option has been developed and implemented.

Methods

Through a password protected access on the main QAAMS website (www.qaams.org.au), participants can view a live web-streamed video presentation of training. Training comprises a series of short video clips which cover both the theory and practice of PoCT. Competency assessment can also be undertaken on-line, initially by downloading and answering a set of written competency questions which must be marked by a member of the QAAMS scientific team. A practical test using Quality Control and Quality Assurance samples must also be completed and sent to the QAAMS scientific team for accuracy verification. A competency certificate is then issued to the successful operator.

Results

24 operators from 12 services, comprising 16 new operators and 8 operators undertaking competency update, have completed on-line training.

Conclusions

The implementation of web-based training has provided QAAMS participants with a flexible, state-of-the-art training option which is available whenever required. This fills the void for services that may have otherwise had to wait for the next regional or annual workshop to receive training.

CBR. 2009 Nov;30(4):S41.

P86 Evaluation of Four Hospital Glucose Meters

PA Simpson ¹, R Tirimacco ¹, PA Tideman ¹

Introduction

Glucose meters are used routinely in hospitals wards and important clinical decisions are made from the results generated by these instruments. In this study we evaluated the accuracy and precision of the Abbott Medisense Optium (glucose oxidase), Roche Accu-Chek Inform (glucose dehydrogenase), Hemocue Glucose 201+ (glucose dehydrogenase) and Nova StatStrip (modified glucose oxidase) glucose meters.

Methods

Accuracy was evaluated by hospital nurses performing a finger stick capillary test on the glucose meter and taking a venous sample to be sent to the laboratory for testing (Roche Modular – hexokinase). Accuracy was assessed using the ISO 15197 standard, which sets minimum acceptable accuracy for glucose meters. Precision was evaluated by nurses performing quality control testing on each of the days that the instrument was in routine use. Quality control material supplied by the manufacturers was used for precision analysis.

Results

The Nova StatStrip and Roche Accu-Chek inform meters met the ISO 15197 standard with 96.9% and 95.9% of results falling within the allowable limits respectively. The Abbott Medisense Optium and Hemocue Glucose 201+ did not reach the standard with only 89.5% and 92.6% of results falling within the allowable limits respectively. The Abbott Medisense Optium, Roche Accu-Chek Inform, Hemocue Glucose 201+ and Nova StatStrip glucose meters produced precision levels of 6.2%, 6.6%, 10.2% and 4.0% respectively for the level 1 control and 4.8%, 5.1%, 4.3% and 2.6% respectively for the level 2 control.

Conclusions

The glucose meters evaluated in this study had varying levels of performance. The Nova StatStrip and Roche Accu-Chek inform displayed the best performance and were the only two to meet the accuracy requirements set by the ISO 15197 standard.

CBR. 2009 Nov;30(4):S41.

P87 Evaluation of the Hemocue White Blood Cell Point of Care Instrument

PA Simpson ¹, R Tirimacco ¹, C Chooi ², PA Tideman ¹

Introduction

The Hemocue White Blood Cell (WBC) System is a point of care instrument for the quantitative determination of WBC count in capillary or venous whole blood. Approximately 10μL of blood is drawn into a test cuvette where a haemolysing agent lyses the red cells and a staining agent stains WBCs. The number of WBCs is calculated by image analysis via a small camera within the instrument and a result is displayed within 3 minutes. The instrument has a measuring range of 0.3–30.0x10⁹/L.

Methods

Accuracy was evaluated by testing K-EDTA venous whole blood samples on both the Hemocue WBC System and hospital laboratory analyser (Beckman Coulter LH750). Precision was evaluated by performing replicate tests on two different blood samples.

Results

Samples tested in this study (n = 178) showed a high level of correlation between the Hemocue WBC System and the laboratory analyser (y = 1.0513x - 0.6654, r² = 0.985, range = 0.3–29.4x10⁹/L). Precision analysis of the Hemocue System gave CV levels of 4.8% for blood sample 1 (mean = 5.0, SD = 0.24, n = 35) and 2.2% for blood sample 2 (mean = 13.1, SD = 0.29, n = 35).

Conclusions

The Hemocue WBC System displayed excellent analytical performance in this evaluation proving it is suitable for use at the point of care. Further investigation into correlations with the laboratory using capillary samples would be recommended as this would be the most common blood type used at the point of care.

CBR. 2009 Nov;30(4):S42.

P88 Assessment of the Nova Statstrip^TM Glucose Meter

G Dimeski ¹, V Tilley ¹, B Jones ¹, N Salmon ¹, Z Lenz ¹

Introduction

Millions of glucose analyses are annually performed on glucose meters around the world. These are often used in conjunction with laboratory analyser results for patient decisions.

Methods

We evaluated patient samples with the Nova StatStrip glucose meter against lithium heparin plasma as the standard sample type on the Beckman DxC800 glucose oxidase method with different sample types: 1) capillary (finger prick), 2) venous whole blood collected in Rapidlyte blood gas syringes, l3) lithium heparin whole blood and lithium heparin plasma and serum collected in Greiner tubes. Precision and limited interferences studies were also performed. The analyser is factory calibrated, can store 1200 results and 1000 user passwords, can be configured for patient gender, age, race, physician and patient ID. The analyser comes with battery charger/downloader, and uses a modified glucose oxidase enzyme system. Evaluation was performed using strip lot no 0308273249.

Results

Precision results for the three levels were: consecutive run CV% 1.2–3.4% (n = 10), and between day CV% 2.9–8.6% (n = 12). No interferences were observed with maltose, galactose, lactose, ascorbic acid, Intragam, and 7.5% w/v Icodextrin. No glucose results were produced with paracetamol >400 mg/L.

Patient comparison	n	r²	Range mmol/L
Capillary	53	0.92	4.7-20.0
Venous blood gas syringe	53	0.96	3.3-20.8
Li heparin whole blood	53	0.97	3.3-22.2
Li heparin plasma	59	0.99	1.4-30.0
Serum	49	0.98	3.4-21.9

Open in a new tab

Patient comparison	Passing Bablok Regression		% difference from Beckman
	Slope	Intercept
Capillary	0.941	0.814	9.2
Venous blood gas syringe	0.996	0.11	1.4
Li heparin whole blood	0.952	−2.2	2.2
Li heparin plasma	0.952	−0.033	−4.7
Serum	1.016	−0.364	−4.8

Open in a new tab

Conclusions

The meter demonstrated acceptable imprecision and inaccuracy, it is relatively free from common interferences, and it is very easy to use.

CBR. 2009 Nov;30(4):S42.

P89 Analytical Assessment of the Il Gem Premier 4000 Analyser

G Dimeski ¹, V Tilley ¹, B Jones ¹

Introduction

The IL Gem Premier 4000 blood gas analyser was evaluated for imprecision, sample comparison on different samples, and for bicarbonate interference with the Cl⁻. The analytes tested were pH, pCO₂, pO₂, THb, O₂Hb, COHb, MetHb, HHb, iCa⁺⁺, Na⁺, K⁺, Cl⁻ and glucose.

Methods

The comparative systems were Siemens Rapidlab 1265, Beckman DxC800 and Sysmex XE-5000 analysers. Samples were obtained from within the hospital patients and selected to cover the widest possible measuring range. Pooled serum was spiked with NaHCO₃ for the Cl⁻ interference.

Results

Acceptable precision was obtained for all the analytes at the different concentration levels (consecutive-run CV range 0.2–7.0 %; between-run CV range 0.3–6.8 %). No outliers were excluded in the sample comparison analysis and there were no aberrant unexplained results. Except for the Cl⁻, good correlations were obtained for the other analytes with the different samples and analytical systems with r² values of 0.92 to 1.00. The Cl⁻ bias was investigated and confirmed to be due to non specificity of the Cl⁻ sensor, bicarbonate interference as per published procedure.¹ New Cl⁻ sensor will be released in the very near future and hopefully overcome such non selectivity.

Conclusions

From the data obtained in this study, the analyser showed good analytical performance for analytes except for the Cl⁻. The analyser is easy to use for all level operators, and has a single cartridge all the components necessary to operate the instrument. 1. Dimeski G, et al. Clin Chem 2004;50:1106–7.

CBR. 2009 Nov;30(4):S42.

P90 Evaluation of the Siemens Rapidlab 1265 Blood Gas Analyser Neonatal Bilirubin Method for Adult Samples

G Dimeski ¹, T Jarman ¹, B Meggitt ¹, D Vega ¹, B Jones ¹

Introduction

Bilirubin is used to assess severity of jaundice, especially in neonates. This study was to verify the transferability of the results from the Rapidlab 1265 spectrophotometric method on whole blood compared with the plasma Beckman DxC800 diazo method in adult patients for the emergency department (ED).

Methods

Unhaemolysed whole blood samples are passed through the CO-oximetry module where 47 absorbance measurements are made (473–672 nm). Bilirubin is determined via patented sets of equations which also correct for other pigments that may be present in the sample. The evaluation included precision, whole blood patient comparison (plasma alone is not suitable) and limited interference studies: haemolysis (haemoglobin) and lipaemia.

Results

The linear regression data was (n=103, range 37–767 μmol/L) y=1.19x-56, r² 0.965. The lowest detectable limit is 34 μmol/L. Neither haemolysis (<450 mg/dL) nor lipaemia (triglycerides <19.9 mmol/L) caused interference. The precision was: consecutive run CV range 2.2–2.7%, and the between-run CV range 2.3–6.0%. The 1265 bilirubin concentrations were lower at the low end and higher at >300 μmol/L than the DxC800. Such differences were not observed in Siemens’ own evaluation using neonatal samples (Rapidlab 1265 = 1.16x Cobas Mirror Plus -9.6, r² 0.995). The observed differences may be due to a number of factors: matrix (neonatal samples contain high levels of HbF and higher unconjugated bilirubin compared with adult), the calibrator (synthetic versus human based or assignment value), and the Rapidlab 1265 algorithm may have been derived for neonatal samples only. This needs further investigation.

Conclusions

The bilirubin availability will add useful clinical information for prompt and improved decisions in our adult only institution in the many patients (e.g. liver transplant that present with elevated bilirubin) presenting for the first time in the ED.

CBR. 2009 Nov;30(4):S43.

P91 Evaluation of Hemocue, Accu-Chek and Optium Blood Glucose Meters for Point-Of-care Use in Paediatric and Adult Hospital Patients

JV Warner ¹, JY Wu ¹, N Buckingham ¹, DSA McLeod ¹, AC Carter ¹

Introduction

Bedside blood glucose measurement is widely practised to investigate suspected hyper- or hypoglycaemia, to monitor diabetes and to adjust glucose-lowering medication in hospitalised patients. A number of point-of-care glucose meters are available, but for large hospitals there are practical and economic advantages to using the same type of meter in all settings. Hospitalised populations include patients with a wide range of haematocrits and supraphysiological levels of galactose or maltose (Icodextrin metabolite) which may interfere in some glucose meter methods. This investigation sought to identify a meter which was accurate, precise and free from interferences as well as being suitable for use across a wide spectrum of age groups and diseases.

Methods

Lithium-heparinised whole blood was analysed on the HemoCue Glucose 201, Accu-Chek Performa, using maltose-insenstitive strips (Roche), and Optium (Abbott) glucose meters and compared with plasma glucose measured on the Vitros 5.1FS analyser (Ortho-Clinical Diagnostics).

Results

The Accu-Chek was the most accurate at all glucose levels. The meters’ within-run imprecision varied from 4.1–5.8% at 2.1 mmol/L glucose, and 2.5–3.4% at 14 mmol/L glucose. Between-run imprecision varied between 2.1 and 6.8% with the HemoCue performing best. All meters measured to 1.1 mmol/L with acceptable precision (CV<14%). All were affected by variations in haematocrit in the range 0.2–0.7L/L. Significant interference at clinically relevant concentrations of galactose and maltose was demonstrated with the AccuChek meter.

Conclusions

All meters are sufficiently accurate and precise to be useful in the hospital setting but they do not perform equally. Choosing a single meter involves a trade-off between accuracy, precision and the risk of adverse outcomes from interferences.

CBR. 2009 Nov;30(4):S43.

P92 Is the Incidence of Macroprolactinaemia Increased in Patients on Antipsychotic Medication

KS Lee ¹, JP Galligan ¹, AE Vinning ¹, S Bazeley ¹

Introduction

Hyperprolactinaemia is a well documented adverse effect of some antipsychotic medications. Macroprolactin is a high molecular-mass IgG-prolactin complex detected by many immunoassay analysers currently in use. Macroprolactin has reduced bioactivity; hyperprolactinaemia resulting from increased concentrations of macroprolactin may result in misdiagnosis and mismanagement of patients. In this study we attempted to determine the contribution of macroprolactin to the incidence of hyperprolactinaemia associated with sustained antipsychotic drug medication, particularly risperidone.

Methods

Plasma specimens from hospital inpatients with prolactin levels >500 mIU/L were used in the study. Prolactin was measured on the Beckman Coulter DxI 800 immunoassay analyser, pre- and post-treatment with 25% polyethylene glycol (PEG). Gel filtration chromatography using Sephadex^TM G-100 was used to confirm the PEG results.

Results

The median recovery of mono-prolactin post-PEG treatment in the non-antipsychotic drug-treated group was 94.5% with a 5–95% confidence interval of 83–108%. In the general hospital group, 25 of 234 patients (11%) had a significant level of macroprolactin. In the antipsychotic drug-treated group, 17 of 275 patients (6.2%) had a significant level of macroprolactin.

Conclusions

Although the incidence of macroprolactin in our drug-treated group was higher than that quoted in an earlier report for a group of pregnant women with hyperprolactinaemia (6.2% vs. 2.9%), it was however, lower than the rate found in our general hospital group (11%). It was also lower than the incidence (9.7%) reported in a similar general hospital setting. These figures indicate that hyperprolactinaemia resulting from therapy with some antipsychotic drugs is not associated with an increase in the prevalence of macroprolactinaemia.

CBR. 2009 Nov;30(4):S43.

P93 Evaluation of the Adams A1C HA-8160 HPLC Analyser for Glycated Haemoglobin in the Laboratory and a Point of Care Setting

GS Streitberg ¹, PR Edwards ¹

Introduction

Measurement of glycated haemoglobin (βN1-deoxyfructosyl-hemoglobin) is an important marker of glycaemic control in diabetes. The Biochemistry Unit provides a glycated haemoglobin (HbA_1c) analytical service in the diabetes clinics conducted by Southern Health. Current instrumentation requires the Biochemistry Unit to maintain two reference intervals due to the different analytical principles and technology employed in the laboratory and the clinics.

Methods

The Adams A1c HA-8160 HPLC analyser was evaluated using patient samples and quality control material and compared with the laboratory TOSOH 2.2plus and the Siemens DCA2000.

Results

The coefficients of variation (CV) (inter-run) on the Adams A1c HA-8160 for BioRad quality control material were less than 1% at HbA_1c levels of 5.4% and 9.4%. Imprecision studies using patient whole blood were all less than 2% (the desirable maximum imprecision for HbA_1c) between 5.1% and 8.9%.

HbA_1c results (50 patients): Y (Adams) = 1.015 (TOSOH 2.2plus) - 0.05 HbA_1c results (72 patients): Y (Adams) = 1.000 (Siemens DCA2000) + 0.10 Processing of patients in the clinic with one Adams HA-8160 was found to be more effective than using two DCA2000 instruments. The Adams A1c HA-8160 analyser is portable, produces the first result in 2.9 minutes and can use an haemolysate made from a sample of 4 μL of blood obtained from a finger prick or EDTA whole blood.

Conclusions

The analytical characteristics and throughput make the Adams A1c HA-8160 analyser a suitable instrument for use in the Biochemistry Unit laboratory and as a POC instrument in the outpatient diabetes clinics held at the Clayton and Dandenong campuses of Southern Health.

CBR. 2009 Nov;30(4):S44.

P94 An Evaluation of the Bio-Rad D-10 for Glycated Haemoglobin Analysis for Point of Care and the Laboratory

PR Edwards ¹, GS Streitberg ¹

Introduction

Glycated haemoglobin (βN1-deoxyfructosyl-haemoglobin) is an important marker of glycaemic control in diabetes. The Biochemistry Unit provides a glycated haemoglobin (HbA_1c) analytical service in the diabetes clinics conducted by Southern Health. Current instrumentation requires the Biochemistry Unit to maintain two reference intervals due to the different analytical principles and technology employed in the laboratory and the clinics.

Method

The Bio-Rad D-10 HPLC analyser was evaluated using i. patient samples ii. quality control material and compared with the laboratory TOSOH 2.2plus and the Siemens DCA2000.

Results

The inter-run coefficients of variation (CV) on the Bio-Rad D-10 for Bio-Rad Lyphochek Diabetes Bilevel Control material were 1.3% at 5.8% and 1.1% at 9.4%. Imprecision studies using patient whole blood were all less than 2% (the desirable maximum imprecision for HbA_1c) between 4.9% and 9.0%. HbA_1c results (44 patients): Y(D-10) = 1.021(TOSOH 2.2plus) - 0.16 HbA_1c results (41 patients): Y(D-10) = 1.000(Siemens DCA2000) + 0.10 Processing of patients in the clinic with one Bio-Rad D-10 was found to be more effective than using two DCA2000 instruments. The Bio-Rad D-10 analyser is portable, produces a result in 3 minutes and can use haemolysate made from blood obtained from a fingerprick or EDTA whole blood.

Conclusions

The analytical characteristics and throughput make the Bio-Rad D-10 analyser a suitable instrument for use in the Biochemistry Unit laboratory and as a point-of-care instrument in the outpatient diabetes clinics held at the Clayton and Dandenong campuses of Southern Health.

CBR. 2009 Nov;30(4):S44.

P95 Evaluation of the Afinion AS100 Analyser for HBA1C Determination

G Dimeski ¹, P Salm ¹, S Nand ¹, M Petroff ¹, J Ungerer ¹

Introduction

The Afinion™ HbA_1c is a boronate affinity assay. The test cartridge contains all the components for the analysis. The assay requires 1.5 μL of capillary (from finger prick) or venous (EDTA, heparin, citrate or NaF) whole blood samples with analysis time of 3 min.

Methods

The evaluation protocol included precision studies, a patient comparison and limited interference testing: bilirubin, haemolysis (haemoglobin), and lipaemia. Precision studies were performed using Afinion HbA_1c controls (L1 and L2). The patient comparison was performed on capillary and EDTA venous samples. The comparative method was the Bio Rad D-10 HPLC analyser.

Results

Evaluation was performed using cartridge lot no 10133162. The glucose range in the samples was up to 59.5 mmol/L, and the HbA_1c range in the EDTA whole blood samples on the Afinion analyser was 4.8–13.7%. The precision data is as follows: a) consecutive run L1 mean ± SD of 5.6 ± 0.11%, CV 1.9% (n = 10); L2 mean of 9.2 ± 0.05%, CV 0.6% (n = 10) and b) the between day L1 mean of 6.1 ± 0.07%, CV 1.1% and L2 mean 8.2 ± 0.13%, CV 1.6%. Comparison studies using patient samples (n = 43) yielded the following linear regression equations: Afinion capillary (y) vs Bio Rad D-10 EDTA (x) (y = 0.869x + 0.55, r² 0.9907, bias: 0.4 ± 0.4%); and Afinion EDTA blood (y) vs Bio Rad D-10 EDTA blood (x) (y = 0.855 x + 0.60, r² 0.9906, bias: 0.4 ± 0.4%). No interference was observed with haemolysis >500 mg/dL, bilirubin ≤348 μmol/L or triglycerides ≤51.1mmol/L.

Conclusions

The analyser is easy to use, and requires minimal training and maintenance. The data indicates a strong agreement between methods with capillary samples equally suitable for analysis.

CBR. 2009 Nov;30(4):S44.

P96 Influence of Methaemoglobin on the Accuracy of Whole Blood Total Bilirubin Measurements on Co-Oximetry Analysers

PVA Pamidi ¹, M DeAbreu ¹, D Kim ¹, S Mansouri ¹, X Coll ¹

Introduction

Total bilirubin measurements are traditionally performed on plasma or serum samples. Such assays suffer interference from haemolysis or lipid turbidity. Most modern blood gas and Co-Oximetry analysers offer whole blood total bilirubin assays without interference from haemolysis or turbidity within clinically significant levels. However, blood based bilirubin assays have to correct for the significant spectral background of 1000 times more concentrated haemoglobin. Co-Oximetry based total bilirubin measurements are generally optimised for samples containing functional haemoglobin and results correlate well with plasma based analysers. In this study the accuracy of measuring whole blood total bilirubin in the presence of methaemoglobin is compared.

Methods

Co-Oximetry based systems; Radiometer ABL 735, Instrumentation Laboratory GEM Premier and Reichert Unistat (plasma/serum) were compared. Fully oxygenated blood samples were spiked with methaemoglobin (metHb) stock solution and bilirubin samples from two different sources (NIST SRM 916 and pooled clinical samples from Lahey Clinic, Burlington MA) were used.

Results

Sample	tHb, g/dL	MetHb, %
Oxyhemoglobin + Cinical Serum	15	0.5
metHb Control	16	25
metHb +NIST bili	16	25
MetHb + Clinical Serum	16	25
Plasma from MetHb+Clinical Serum	N/A	0
1 month old Cord Blood Sample	13	9

Open in a new tab

	Total Bilirubin, mg/dL
Sample	GEM Premier 4000	ABL 735	Unistat
Oxyhemoglobin + Cinical Serum	25.9	24.6	25.3
metHb Control	2.8	4.1	0.4
metHb +NIST bili	23.7	23.1	13.9
MetHb + Clinical Serum	18.1	18.8	14.7
Plasma from MetHb+ Clinical Serum	14.0	14.4	14.7
1 month old Cord Blood Sample	2.4	4	0

Open in a new tab

The data highlights the impact of methaemoglobin on whole blood bilirubin measurements compared to plasma from the same sample.

Conclusions

Bilirubin measurements on samples containing clinically significant levels of methaemoglobin require appropriate bias corrections or interference flags. Bilirubin measurements on the GEM Premier 4000 will attempt to correct or flag the results for samples containing methaemoglobin.

CBR. 2009 Nov;30(4):S45.

P97 Evaluation of the Gem Premier 4000 Blood Gas Analyser

G Koerbin ¹, H Robins ¹, P Talsma ¹

Introduction

An evaluation of the GEM Premier 4000 (Instrumentation Laboratory), was undertaken prior to routine installation in a number of ward environments at the Canberra Hospital. The only reagents required of this analyser are disposable cartridges, GEM Premier 4000 PAK, that comprise all components necessary for analysis, including quality controls

Methods

Method comparison was undertaken using 45 patient samples with the Gem Premier 4000 being compared with the Siemens/Bayer 865 blood gas analyser. Analysis was undertaken using Passing-Bablok regression. Precision for each analyte determined was expressed as coefficient of variation (CV%).

Results

Between-run imprecision showed good CVs for all parameters. Method comparisons demonstrated close agreement with the Siemens/Bayer 865 blood gas analyser for all analytes. Regression and precision data are shown in the following table.

	Regression analysis

	Slope	y-intercept	r²
pH	1.08	−0.62	0.99
pO2	0.99	2.9	0.99
pCO2	1.04	−1.6	0.99
Sodium	1.00	19	0.93
Potassium	1.00	0.1	0.99
Chloride	0.91	10	0.92
Glucose	1.03	−0.3	0.99
Lactate	0.96	−0.23	0.99
iCa	1.09	−0.11	0.98
Hb	1.00	−2	0.99

Open in a new tab

	Precision data

	Mean	CV%
pH	7.43	1.0
pO2	86.2	4.2
pCO2	37.3	2.0
Sodium	142	0.7
Potassium	5.0	1.7
Chloride	111	0.6
Glucose	7.8	0.9
Lactate	1.2	6.8
iCa	1.10	1.3
Hb	100	1.3

Open in a new tab

Conclusions

The GEM Premier 4000 shows good analytical performance and close analyte agreement when compared with the Siemens/Bayer 865 blood gas analyser.

CBR. 2009 Nov;30(4):S45.

P98 Evaluation of the Roche HS-TNT Assay

G Koerbin ¹, E Miller ¹, PE Hickman ^1,²

Introduction

Determination of cardiac troponin is the current standard for the diagnosis or exclusion of acute myocardial infarction. Roche Diagnostics have recently introduced a high sensitivity troponin T (hs-TnT) assay to its suite of tests available for the Elecsys range of analysers. We evaluated the performance of this assay and compared it with the troponin T STAT assay and the Abbott Architect Troponin I (TNI-II) assay.

Methods

Analytical imprecision, linearity of the hs-TnT assay and method comparison using a cohort of 58 patient samples was undertaken and compared with the troponin T STAT assay and the Abbott Architect Troponin I (TNI-II) assay.

Results

Between-run imprecision showed coefficients of variation (CVs) of 6.5% at 28.7 ng/L (0.029 μg/L) and 4.1% at 2352 ng/L (2.35 μg/L). Method comparisons demonstrated significantly improved low end sensitivity when compared with the troponin T STAT assay and the Abbott Architect Troponin I (TNI-II) assay. The number of patient samples with detectable troponin at the manufacturer quoted limit of detection (5 ng/L) and 10% CV (13 ng/L) for each assay in this small cohort are shown in the following table.

n = 58	Troponin T (Elecsys 1010)	Troponin T (Elecsys 2010)
Limit of Detection	7	13
10% CV	6	6

n = 58	Troponin T hs (Elecsys 2010)	Troponin I (Abbott Architect)
Limit of Detection	58	32
10% CV	38	16

Open in a new tab

Conclusions

The high sensitivity troponin T from Roche diagnostics has improved low end sensitivity when compared with the Troponin T STAT and Abbott Troponin I TNI-II assays with CVs of approximately 10% at near the assay limit of detection. The clinical significance of the increased positivity rate with the hs-TnT assay compared with the previous generation TnT and Architect TnI assays remains to be elucidated.

CBR. 2009 Nov;30(4):S45–S46.

P99 Correlation Study of Myoglobin and Troponin Assays on Radiometer AQT90 and Abbott Architect Analysers

RI King ¹, PM George ¹

Introduction

The diagnosis of acute myocardial infarction has gone through a revolution in the past decade with the introduction of cardiac troponin (TnI). The combination of myoglobin, an early but non-specific marker for myocardial damage, with troponin, a specific and sensitive marker is an established strategy for screening patients presenting with chest pain for myocardial ischaemia. Rapid evaluation of patients with cardiac symptoms is important and can be fulfilled by point-of-care testing. We investigated the correlation between the Radiometer AQT90, a random access point-of-care analyser with the main laboratory Abbott Architect assays for myoglobin and TnI.

Methods

Total imprecision and linearity for both myoglobin and TnI were checked. Correlations between the Abbott Architect and Radiometer AQT90 myoglobin and TnI assays were determined from consecutive patient samples arriving in the laboratory. Using 99^th percentile TnI cut-offs an error grid was constructed and the clinical details of discrepant results reviewed.

Results

Both analytes showed acceptable linearity. CV at the 99^th percentile (0.023 μg/L) for AQT TnI was 8.6%. The correlations (Deming regression) were: myoglobin AQT = 1.026(Architect) + 18.1 μg/L, r = 0.973; range, 25 to 300 μg/L; N = 47, TnI AQT = 0.234(Architect) + 0.025 μg/L, r = 0.703, range, 0.01 to 10 μg/L; N = 308. ROC analysis - Architect cut-off <0.03 μg/L, sensitivity and specificity 97%, AQT <0.023 μg/L, sensitivity 78%, specificity 94%.

Conclusions

The AQT provides myoglobin and TnI assays in an easy to use format suitable for use by non-laboratory staff. There was good agreement for the myoglobin assay but a calibration difference was identified between the AQT and Architect TnI assays. Twelve percent of TnI samples showed clinically significant differences between the two assays.

CBR. 2009 Nov;30(4):S46.

P100 Comparison of the Radiometer AQT 90 Flex Analyser Versus Beckman Coulter DXI 800 for the Measurement of Troponin I

L Mackay ¹, N Crinis ¹, Q Lam ¹, R Scott ¹

Introduction

Troponin I (TnI) is used in assessing patients presenting to the Emergency Department with cardiac symptoms. The performance of the TnI assay on the Radiometer AQT90 Flex analyser was evaluated to determine if it would be a suitable point of care option for the Emergency Department. The AQT90 Flex method was compared to the Beckman Coulter AccuTnI on the Unicel DxI 800.

Methods

Patient samples with detectable TnI levels using the Unicel DxI 800 were selected. EDTA samples taken during the same episode were analysed within four hours of collection on the AQT90 Flex. Assay and precision evaluated using a low positive patient pool and two levels of a commercially available quality control material. Patient samples and QC materials were measured on the same day, over an evaluation period of six months.

Results

Correlation by linear regression (n = 152 paired samples), TnI range: 0.01 – 12.0 μg/L yielded the regression equation DxI = 4.399AQT - 0.126, r² = 0.90. Precision for the patient pool was 11.2% at 0.043 μg/L, and for the QC material was 9.1% and 11.9% for levels of 0.42 μg/L and 1.64 μg/L respectively.

Conclusions

The AQT90 Flex TnI assay demonstrated acceptable precision. However initial data analysis suggests that the regression equation is only applicable to AQT values greater than 0.10 μg/L making the application of a correction factor of limited value in the clinical decision range.

CBR. 2009 Nov;30(4):S46.

P103 Effect of Troponin-I Turnaround Time on Length of Stay (LOS) in the Emergency Department in Acute Coronary Syndrome (ACS) Patients

A Tiberio ¹, F San Gil ², A Lee ^1,³

Introduction

The benefit of point-of-care (POC) troponin on LOS in the Emergency Department (ED) is contentious. We assessed the impact of POC troponin on LOS in a medium-sized regional ED.

Methods

Patients (n = 110) presenting to ED from 5/9/08 to 13/2/09 with chest pain and triaged as intermediate-risk ACS were recruited. Patients were assigned randomly into Laboratory or POC groups. POC troponin was determined using the Abbott Point-Of-Care I-stat and the result made immediately available to clinicians. Troponin I results for the Laboratory group were obtained using current practice from the on-site hospital pathology service.

Results

Time from admission to test availability was 50 min (35–73 min) for the POC group and 135 min (75–215 min) for the Laboratory group. Based on troponin results, patients were reclassified as low or high risk. LOS in the ED for intermediate risk patients was 4.5 hours (4–5hrs) for the POC group (n = 12) and 5.4 hours for the Laboratory group (n = 11). LOS before being admitted to the Coronary Care Unit for patients (n = 6) reclassified as high risk based on the troponin result, was 4 (3–5) hours for the POC group and 6 (4–8) hours for the Laboratory group.

Conclusions

POC troponin testing reduced total time from admission to result by 90 min. LOS in ED fell by 30 minutes for intermediate-risk ACS patients and 2 hours for high-risk patients.

CBR. 2009 Nov;30(4):S46.

P104 Temporary False Positive Troponin-I Results for Two Commonly Used Methods

R McFarlane ¹, P Quigley ¹

Introduction

We defined positive plasma troponin-I results, which become negative on repeating after high-speed centrifugation, as temporary false positives (TFP). We report our experience of relatively high incidences of temporary false positive results in our patient population using two troponin I methods, and our procedures for investigating and resolving them.

Methods

We used the Beckman-Coulter Access AccuTnI method from 2003–2007, and the Abbott Architect stat Troponin-I method from mid-2007 to date. Specimens were collected into Becton-Dickinson PST-II Lithium Heparin Vacutainers.

Results

With the Beckman-Coulter method, we initially found a TFP rate greater than 3% of all specimens. Increasing routine sample centrifugation from 3500 rpm/5 minutes to 5600 rpm/5 minutes had no effect. We adopted a routine practice of re-spinning the first positive specimen from each patient at 10,000 rpm for 10 minutes in an Abbott X-systems centrifuge. Beckman Coulter later changed the reagent formulation from “cTnI” to “aTnI”. This improved the TFP rate to 2%. When we changed to the Abbott method, using the same centrifugation conditions, our TFP rate dropped to 0.35%.

TFPs occurred almost exclusively in septic patients with raised C–reactive protein levels, and were more than 3 times as common in Intensive Care as Emergency Department patients. Our attempts to demonstrate interference using platelet-rich plasma from the same patients gave inconclusive results. Matched serum specimens did not give TFPs, and refrigerated storage of TFP specimens for 4–12 hours eliminated the TFP. We also found TFPs using the Access hCG method.

Conclusions

Troponin methods vary in their sensitivity to TFPs, which appear to be caused by a yet-unidentified labile particulate component of plasma, removable by high-speed centrifugation. We postulate that it prevents efficient washing of the assay microparticles. The possibility of this interference affecting other immunoassay methods, including the more sensitive troponin methods under development, should be considered.

CBR. 2009 Nov;30(4):S46–S47.

P105 Troponin T Point-of-Care Testing in an Emergency Department Using Roche Cobas H232

R Bais ¹, R Day ²

Introduction

The RNS Emergency Department (ED) Chest Pain Acute Coronary Syndrome (ACS) pathway requires an initial Troponin T (TnT) and, if not elevated, a second TnT 6–8 hours after admission. The test results assist in risk stratification of patients with possible ACS, and determination of their appropriate clinical management strategy, including a decision regarding whether to admit or send home.

Methods

Point-of-care testing (POCT) of TnT on the h232 was used in ED testing for the second TnT sample from 315 patients over a 10 month period. The h232 enables a TnT result to be available within 12 minutes of taking the blood sample. These results were compared to the laboratory TnT results measured on a Roche Modular E170 using a blood sample collected at the same time.

Results

The imprecision was 14% at 0.269 μg/L on the h232, 3.9% at 0.345 μg/L and 9.2% at 0.033 μg/L on the E170. The h232 gives TnT results as <0.03 μg/L, 0.03–0.1 μg/L and a numerical value at >0.1 μg/L. The results from the E170 were classified in the same way and the following 3 X 3 result matrix was obtained:

graphic file with name CBR_30(4)_S47_g002.jpg

The 23 results >0.1 μg/L gave a linear relationship with an equation h232 = 1.17 E170 - 0.01 and no significant difference between the pairs (p = 0.083).

Conclusions

There was 100% diagnostic concordance in this study of 315 patients presenting to ED. The TnT result was available in the ED at least 40 min earlier than the result from the laboratory. The level of accuracy of the h232 is sufficient to warrant its safe use in ED.

[b1-cbr_30_4_s1] 1.International Federation of Clinical Chemistry and Laboratory Medicine (IFCC) http://www.ifcc.org/PDF/IFCC_Troponin_Web_Page_Table_of_Assays_Oct_2008.pdf.

PERMALINK

Proceedings of the Australasian Association of Clinical Biochemists’ 47th Annual Scientific Conference

S1 Experiences with Heart and Brain Biomarkers

JH Ladenson

S2 Therapeutic Drug Monitoring: Overview and Update

L Broussard

S3 Pathologists Overseas

JH Ladenson

S4 PoCT in Australia – Implementation and Quality Management, Where are we up to?

R Tirimacco

S5 Advances in PoCT Technology and Range of Available Tests

C Martin

S6 BNP Testing in the Emergency Departmentpanacea or a Waste of Money?

H Schneider

S7 Analytical and Clinical Performance of Cardiac Troponin Assays

J Tate

References

S8 Markers of Alcohol Tolerance

JB Whitfield

S9 Severe Glucocorticoid Resistance in a Neonate Presenting with Hypoglycaemia – A Case Presentation

SK McMahon

CJ Pretorius

JPJ Ungerer

NJ Salmon

LS Conwell

JA Batch

S10 Primary Hyperaldosteronism

M Stowasser

S11 Insulin Resistance: Recent Advances

P Williams

T Yen

S Twigg

References

S12 Re-Aligning Allowable Limits of Performance

K Sikaris

S13 Quality Issues for Diagnostic Reagent Manufacturers

G Koumantakis

S14 Insight into Transfusion QAP

CE Hughes

NR Herrmann

S15 Insight into the Cytopathology QAP

J Finnimore

S16 QAP –Strategic Directions

P Stewart

S17 Point of Care Testing Working Party

R Tirimacco

R Bais

B Glastonbury

J Gill

B Heffernan

G Koumantakis

C Martin

M Shephard

A St John

V Hauke

L Watkinson

R White

S18 Vitamin D Uncovered

GH Beastall

O1 Comparability Assessment of Three Point of Care Analysers for Troponin I Against A Central Laboratory Platform

A Coleman

O2 Carbamazepine Interference In Urine Free Cortisol By UPLC

T Harrower

J Creces

G Ward

O3 Negative Interference in an Insulin Assay

M McLaren

T Yen

L Price

G Ward

Table.

O4 Comparison of Two Immunoassays and Two LC-MS/MS Methods for 25-Hydroxyvitamin D

D Fredline

V Bell

S Howells

K Young

L Rogers

B McWhinney

J Galligan

D Kanowski