PMCCPMCCPMCC

Search tips
Search criteria 

Advanced

 
Logo of clinbiorevLink to Publisher's site
 
Clin Biochem Rev. 2008 November; 29(3): S120–S148.
PMCID: PMC2605415

Proceedings of the Australasian Association of Clinical Biochemists’ 45th Annual Scientific Conference

Clin Biochem Rev. 2008 November; 29(3): S120.

S1 The Role of the Profession, Laboratories and Laboratory Personnel in the Evolution of RCPA Chemical Pathology Quality Assurance Programs

The Royal College of Pathologists of Australasia (RCPA) reintroduced Australian produced quality assurance (QA) programs in 1982. These programs were the product of close collaboration between the RCPA and AACB and their respective Fellows and Members with expertise in design and delivery of QA programs. This expertise had its foundation in QA theory and programs of the day. However, the decision to produce “Australian QA programs” was driven by a desire for change and a need for improvement in the information provided by QA programs. The professional bodies were seeking summary information on analytical performance to assist them to focus on problem areas. In addition, the initial reports were designed to provide each participating laboratory with data which comprehensively compared their performance with their peers. Allowable Limits of Performance were also introduced to assess whether a laboratory’s results were acceptable for patient care.

Introduction of the Australian programs generated considerable debate, not always favourable, within laboratories and the profession. This debate was facilitated by the AACB introducing a one-day QC Update in 1984 as part of their annual scientific meeting. These meetings have allowed discussion of the strengths and weaknesses of the programs and provided an annual stimulus for introducing new features or new programs. Both the AACB and RCPA also facilitated Industry Displays at their annual meetings and a variety of other opportunities for debate.

Many individuals from RCPA, AACB and participating laboratories have also provided substantial support in developing the programs. David Rothfield was one of these individuals. He was both a supporter and constructive critic of all aspects of the range of challenges, data management and reporting and the summary statistics. His views of the programs as a laboratory user and subsequently as a tool for assessing a laboratory’s analytical performance has been invaluable.

This presentation will review key events in the development, design and reporting of the RCPA Chemical Pathology QA Programs.

Clin Biochem Rev. 2008 November; 29(3): S120.

S2 Traceability, Reference Systems and Result Comparability

The purpose of standardisation of measurements in Laboratory Medicine is to achieve closer comparability of results obtained using routine analytical systems. In order to achieve standardisation, an approach is required that provides reliable transfer of the measurement values from the highest hierarchical level to methods which are routinely used. Such a structure is presented by the reference measurement system (RMS), based on the concepts of metrological traceability and a hierarchy of analytical measurement procedures. Since the development of metrologically sound RMS is a complicated and expensive process, it is clear that the objective of improving standardisation will only be achieved if the problems are dealt with through international cooperation. This was the reason for the creation of the Joint Committee on Traceability in Laboratory Medicine (JCTLM), which has made available a list of higher order reference materials and methods and also the reference laboratories that are able to deliver a reference measurement service.1 Once a RMS is implemented, clinical validation of the correctly calibrated routine methods should take place. In specific cases, in order to maintain the value of clinical experience, correlation of measurement results obtained with the new traceable calibration to results of measurements obtained with the previous not standardised calibration should be established. Other important requirements concerning the implementation of a metrologically-correct approach for result standardisation are the definition of the clinically allowable error of measurements and the post-market surveillance of the performance of diagnostic products through the organisation of appropriate EQAS. The applicability of the true value concept in EQAS requires, however, the availability of control materials with target values assigned by reference methods and that these materials behave exactly as human patient specimens.

References

1. Joint Committee on Traceability in Laboratory Medicine. [(Accessed September 2008)]; http://www.bipm.org/en/committees/jc/jctlm/jctlm-db.
Clin Biochem Rev. 2008 November; 29(3): S120–S121.

S3 Obesity and the Toxic Environment: Is It Really Toxic?

The epidemic of obesity throughout the world has been blamed on numerous factors in the genetic, epigenetic, and environmental realms. Environmental alterations include the ubiquity of fast food restaurants, urban sprawl, and lack of outside activity for fear of crime. Although euphemistically called the “toxic environment”, in fact we are all exposed to environmental toxins in our environment that foment obesity through various endocrine mechanisms.

The oestrogen DES is a known endocrine disruptor causing cancer when administered in parts per million, but causes obesity in animals in concentrations of parts per billion. Similarly, the environmental oestrogens bisphenol-A (from plastics production), and genistein (a phytoestrogen) stimulate adipogenesis in animals. Phthalates are used in plastics production, but also activate PPARg to promote adipogenesis. Organochlorines, such as atrazine, are used as insecticides, but also increase adipogenesis by acting as an oestrogen and by decreasing serum T3 and resting metabolic rate. The class of organotins, used as fungicides, also inhibit 11bHSD2 to increase blood pressure, and are ligands for RXR and PPARg to induce adipocyte differentiation. Although epidemiologic studies are thus far only correlative, these “persistent organic pollutants” or POPs, are very worrisome, because they are ubiquitous, and may work antenatally to alter foetal programming.

We are also exposed to toxins at the grocery store. Thanks to its abundance, sweetness, and low price, high-fructose corn syrup (HFCS) has become the most common sweetener used in processed foods. It’s not that HFCS is biologically more ominous than sucrose (table sugar); it’s that its low cost has made it available to everyone, especially low socioeconomic groups. Average daily fructose consumption from all its sources (HFCS, sucrose, juice) has doubled over the past 30 years. There is abundant correlative evidence that the growing dependence on fructose in the Western diet is fueling the obesity and T2DM epidemics. In addition, mechanistic models in both animals and humans demonstrate that high-fructose diets lead to increased energy intake, decreased resting energy expenditure, excess fat deposition, and insulin resistance.

Hepatic metabolism of fructose differs significantly from that of glucose. Fructose enters the liver without insulin regulation. Fructose phosphorylation causes hepatic phosphate depletion, which increases uric acid production, contributing to hypertension. Fructose enters the glycolytic pathway without regulation. This leads to an excess accumulation of citrate outside the mitochondria, which then undergoes de novo lipogenesis and is reassembled into free fatty acids, which promote insulin resistance, very low density lipoproteins (which promote atherosclerosis and serve as a substrate for obesity), and triglycerides (some of which precipitate in the liver, and cause non-alcoholic steatohepatitis). Indeed, fructose alone causes all of the manifestations of the Metabolic Syndrome. Fructose also does not suppress secretion of ghrelin from the stomach, levels of which correlate with perceived hunger. Lastly, fructose is seven times more likely than glucose to form advanced glycosylation endproducts (AGEs), which promote atherosclerosis. In sum, fructose consumption has metabolic and hormonal consequences that facilitate development of obesity and the Metabolic Syndrome.

The “toxic environment” is not just a euphemism; and we must work to alert and educate the public and politicians as to the environmental dangers ahead.

Clin Biochem Rev. 2008 November; 29(3): S121.

S4 Geriatric Nutrition: Losing Weight is Bad

As our population ages ageing-associated problems assume greater significance. Despite the rising prevalence of obesity, a surprising number of older people are under-nourished, particularly those in nursing homes and other institutions. The body mass index (BMI) range associated with maximum life expectancy increases with increasing age and is in the order of 22–23 to 27–28 kg/m2 for older people. Weight loss, particularly if involuntary and more than 5%, and low body weight (BMI <22 kg/m2) are both associated with increased morbidity and mortality in older people. We have found that community dwelling older people in Adelaide who are under-nourished or at risk of under-nutrition are twice as likely to need hospital admission as their well-nourished peers. Yet weight loss is common in older people; on average body weight is lost after age 65 years with a preferential gain of fat and loss of muscle. The latter has particularly adverse effects. The causes of under-nutrition and weight loss in older people are complex and multiple, but are caused, at least in part, by the physiological decline in appetite and food intake that accompanies normal ageing, the so-called “anorexia of ageing”. Possible causes include increased activity of the satiety hormones cholecystokinin and altered gastrointestinal function.

Various strategies have been tested to prevent and treat under-nutrition in older people. Because weight loss by older people is often disproportionately of skeletal muscle, anabolic treatments may be beneficial. We have administered oral testosterone undecanoate (160 mg/day in men, 40 mg/day in women) and a nutritional supplement drink (475 kcal/day), alone and combined, for one year to 49 undernourished people, and demonstrated a significant reduction in hospitalisations (p = 0.03) and days in hospital (p = 0.04) in the combined treatment vs no treatment group. Larger follow-up studies of such combined treatments are needed.

Clin Biochem Rev. 2008 November; 29(3): S121.

S5 New Biochemical Markers for Pre-Eclampsia

Risk profiling for hypertension in pregnancy involves obstetric, maternal and placental factors. There are high-risk profiles that relate to advanced renal disease and organ transplantation as well as some forms of chronic hypertension and glomerular disease. Many other risk factors for hypertension relate to obstetric factors such as multiple gestation, prior obstetric history and molar pregnancy. The role of epidemiologically discovered risk factors such as body mass index, family history, parity, inter-pregnancy interval and partner status are more difficult to use as markers of individual risk.

Assessment of placental function is both a new and an old idea and current research has identified the cellular biology of the placenta in more detail. The immunological and cytokine functions of the placenta clearly contribute to maternal adaptation to pregnancy, including blood pressure regulation. The role of recently defined angiogenic factors and hypoxia regulated molecules look to be a major development in our capacity to define placental function in early pregnancy, or at least, preceding the onset of clinical symptoms and signs. The interaction of these molecular systems with placental cell processes such as apoptosis, and villous formation and invasion of the endometrium, may explain early placental failure in pre-eclampsia. The likely value of a predictive test is most challenging in the women with no prior obstetric history.

A composite “scoring” of maternal factors, placental function tests, maternal immunological adaptability and possibly utero-placenta blood flow tests will be more likely to predict an individual case of pre-eclampsia than any test alone. However, the promise of reliable blood tests of stable molecules reflecting the various clinical aspects of “toxaemia” is fast being realised.

Clin Biochem Rev. 2008 November; 29(3): S121.

S6 The Pleiotropic Effects of Vitamin D: Its Importance in Bone and Systemic Health

Vitamin D is a fat-soluble vitamin with a steroid-like structure similar to cholesterol. It has a specific ligand, the vitamin D receptor that is prolific, found in many tissues and through which it mediates its endocrine and paracrine action. The active vitamin D metabolite, 1,25D is central to the regulation of vitamin D physiology. Circulating as a hormone, it regulates calcium homeostasis. At the cellular level, it acts by either suppressing or up-regulating gene transcription, altering cell signalling and cellular differentiation.

It is therefore notwithstanding that vitamin D deficiency can result in an array of medical disorders such as abnormal bone metabolism (osteoporosis and osteomalacia), cardiovascular disease, diabetes, cellular dedifferentiation (oncogenesis) and immune derangement (auto-immune disorders such as lupus, rheumatoid arthritis and multiple sclerosis). The most compelling evidence for its pleiotropic actions, however, comes from reports linking vitamin D deficiency to higher risks of death from all causes, by mechanism as yet unknown.

One of the major disputes in vitamin D research is the lack of universality in defining an optimal serum 25D level required for “good health”. The definition of vitamin D deficiency (defined as serum 25D of less than 12.5 nmol/L) and insufficiency (defined as serum 25D 30–50 nmol/L) has largely been derived from metabolic bone studies. Healthy individuals who receive adequate sunlight exposure usually have serum 25D greater than 90–100 nmol/L. This review will focus on some of the clinical evidence linking low serum 25D levels to a number of vitamin-D sensitive diseases and outline the importance of maintaining high serum 25D for disease prevention.

Clin Biochem Rev. 2008 November; 29(3): S121–S122.

S7 Interpretation of Bone Markers

Introduction

Bone markers are of great value to clinicians both for diagnostic and therapeutic purposes. In the investigation of metabolic bone disease, particularly osteoporosis, it is essential to know not only the bone density (generally measured by dual energy x-ray absorptiometry and expressed in g/cm2) but also the direction of change in bone density. Bone resorption and formation are closely linked and commonly called “bone turnover” though high turnover is generally driven by bone resorption, at least in women. Markers of bone formation and resorption can therefore both be used to represent bone turnover and their clinical value is increased by the fact that bone turnover is a risk factor for fracture in its own right.

Methods

We are reporting results from 184 treatment periods in 179 postmenopausal women in whom the change in bone mineral density (BMD) at five sites was measured over a mean period of 18 months (10–36). Serum osteocalcin was measured in the Immulite 2000 and CTX on the Roche E170.

Results

There was a strong positive correlation between the two markers (r = 0.50, p<0.001), and inverse correlations between each marker and the changes in bone density. The latter were most significant when ΔBMD was expressed as the mean change at proximal forearm, spine and hip. The r-value for each marker with this mean value was 0.25 (p<0.001) but when the two markers were combined, it became 0.29 (p<0.001). At CTX up to 400 ng/L, most patients were gaining bone; at higher values most patients were losing bone (positive predictive value (PPV) 71%; negative predictive value (NPV) 72%). The corresponding breakpoint for osteocalcin was 4 μg/L (PPV 62%; NPV 67%). There was agreement between the two markers in 132 subjects. If both were high, 69% were losing bone; if both were low, 82% were gaining bone. However, there was poor quantitative agreement between the marker levels and the rates of bone loss or gain.

Conclusions

Both markers are clinically useful indicators of the direction of change in bone density. CTX has rather more predictive value than osteocalcin but the combination of the two is more useful than either alone. However, there are many false positives, particularly with CTX; further refinement is needed.

Clin Biochem Rev. 2008 November; 29(3): S122.

S8 The Laboratory and Bone Assessment: It’s all about the Sample

Biochemical markers of bone turnover (formation and resorption) are measured by either immunoassay or HPLC techniques in serum or urine. Bone formation markers include by-products of type 1 collagen synthesis such as procollagen type I N propeptide (PINP) and procollagen type I C propeptide (PICP) measured in serum, or products of osteoblasts e.g. osteocalcin (serum and urine) and bone alkaline phosphatase (serum). Bone resorption is assessed by either collagen breakdown products such as C-terminal crosslinking telopeptide of type I collagen (CTX) (serum, urine), N-terminal crosslinking telopeptide of type I collagen (NTX) (serum, urine) and deoxypyridinoline (DPD) (urine, serum), or osteoclastic activity e.g. tartrate resistant acid phosphatase (TRAP5b) (serum).

Bone markers have in general a large degree of preanalytical variation due to both technical and biological components. Biological components, some of which can be controlled, include: age, sex, season, diet, exercise, pregnancy and medical conditions (fractures, hyperparathyroidism, hyperthyroidism, steroid therapy). Analytical aspects can be confounded by the analyte stability or assay specificity e.g. osteocalcin (unstable and a number of circulating forms) and DPD (photolysis). Further problems include those inherent with collection of urine samples such as time of collection, completeness and use of analyte creatinine ratios. In clinical research studies many of these components can be controlled. Research has demonstrated measurement of bone markers to be useful both in the assessment and diagnosis of metabolic bone disease. In routine practice this has been difficult due to the large day to day variation of bone marker levels. Bone turnover markers currently cannot be used for diagnosis of osteoporosis. Bone marker levels after drug therapy reflect the mechanism of action of the drug relative to the mechanism of marker production. For example, PTH causes bone formation and both formation and resorption markers increase with PTH therapy, whereas bisphosphonates result in decreased production of markers. Bone markers can demonstrate the effectiveness of drug therapy more quickly than measurement of bone mineral density. There are no position statements for the use of biochemical bone turnover markers and clinicians are often unsure of which sample matrix and markers to use. Early studies on bone markers required the use of both serum and urine to assess bone turnover, particularly urine for bone resorption markers. Formation and resorption markers are now available in serum and it is likely that future bone turnover or “bone function” panels will be predominantly serum based.

Clin Biochem Rev. 2008 November; 29(3): S122.

S9 Vitamin D Deficiency and Depletion: How Much Do We Need for Optimal Health?

The vitamin D system plays a primary role in the maintenance of calcium, and phosphate homeostasis. Recently it has been identified as exerting a wider range of biological activities including regulation of cellular differentiation and proliferation and therefore it has implications for cancer, the immune system, and reproduction. The endocrine action of vitamin D acts through the renal metabolism of vitamin D, producing 1,25 dihydroxyvitamin D (1,25D) for the circulation. The major organ responding to plasma 1,25D is the small intestine where it controls intestinal absorption of calcium and phosphate. 1,25D is also synthesised in a wide range of tissues including bone cells where it is believed to exert an autocrine or paracrine activity. Vitamin D insufficiency in the elderly has clearly demonstrated an increased risk of hip fracture. A major question for the clinical laboratory is what is the vitamin D requirement for a healthy skeleton? Measurement of vitamin D status in apparently healthy individuals merely assesses their level of sunlight exposure and does not indicate vitamin D requirements. Current evidence from a wide range of sources including clinical data indicate that levels of serum 25-hydroxyvitamin D (25D) below 20 nmol/L are clearly deficient and have significant adverse effects on bone and calcium homeostasis. However evidence is accumulating to suggest that 25D levels between 20 and 80 nmol/L indicate a depleted vitamin D status increasing the risk of fracture as well as clinical data demonstrating increased risk of colon cancer. These findings suggest that the lower level of the desirable limit for vitamin D should be significantly higher than currently used by many laboratories.

Clin Biochem Rev. 2008 November; 29(3): S122.

S10 Vitamin B12: Total vs Active

Clinical diagnosis of vitamin B12 deficiency is not straight forward due to lack of a diagnostic gold standard and the carrier protein effect on the serum B12 concentration. Vitamin B12 binds to two binding proteins: Haptocorrin (approx 80%) and transcobalamin (approx 20%). Holotranscobalamin (HoloTC) is biologically active which is taken up by the cells to form cofactor for metabolic reactions. We examined the clinical usefulness of the new HoloTC assay (“Active B12”).

Active B12 was measured on the Abbott AxSYM routinely on all patients with a total vitamin B12 (measured on Abbott Architect) ≤200 pmol/L. 8,858 data pairs of Active B12 and total B12 results were obtained May-Nov 2007. Results were divided into four quartiles using cut-off values of 140 pmol/L for the total vitamin B12 and 23 pmol/L for the Active B12. Patients were also grouped based on their Active B12 concentrations to assess the rate of macrocytosis in each group.

Review of the discordance data and individual cases showed that the Active B12 assay was generally better than the total B12 assay for assessing B12 status, especially in pregnant women and in people with hypothyroidism or iron deficiency. The rate of macrocytosis was increased when ‘Active B12’ was below around 20–23 pmol/L.

In conclusion, the Active B12 assay on the Abbott AxSYM may provide a better estimate of the true vitamin B12 status. It has the potential to replace total vitamin B12 assay to become the first line test to screen for vitamin B12 deficiency. This potential is currently limited as the automated Active B12 assay is only available on Abbott AxSYM which requires greater resources to operate than the total B12 assay.

Clin Biochem Rev. 2008 November; 29(3): S123.

S11 Economics of Point-of-Care (POC) Testing for B-Natriuretic Peptide (BNP) and Cardiac Markers

Delay in diagnosis and treatment of cardiac patients presenting in the Emergency Department (ED) with acute symptoms is associated with poor patient outcomes, hence the timely and accurate diagnosis of congestive heart failure (CHF) and acute coronary syndromes (ACS) in the ED now require the 24/7 availability of real-time, rapid (30 minutes or less) testing for BNP and cardiac markers. In addition to the traditional role of BNP as a marker for CHF, recent studies at our institution and others indicate rapid BNP measurements are also useful in evaluation of our ED ACS patients as candidates for possible emergency Percutaneous Cardiac Interventions (PCI) or Coronary Artery Bypass Graft (CABG) interventions. We have observed that creation of a special cardiac “Decision Unit” in our ED together with the rapid availability of laboratory evidence of CHF or ACS due to rapid POC testing for BNP and cardiac markers has greatly facilitated the evidence-based diagnosis and management of our cardiac patients. Tangible benefits to our hospitals of improved clinical decision making for our cardiac patients both in our ED and CCU are well beyond the cost of performing POC cardiac testing; these benefits are clearly in three areas: clinical outcomes improvement, operational improvements, and economic benefits. To date, benefits observed include a reduction in as much as 15 h ED Length of Stay (LOS)/wait time for CHF patients and ACS patients (especially those with non-ST segment elevation acute myocardial infarction patients) to a consistent ED LOS of no more than 8 h. In addition, our inpatient LOS for CCU CHF patients has been dramatically reduced from 5.2 days to 3.2, a potential savings of over $1,000 per day per patient. With our improved cardiac clinical decision making protocols, we have also reduced the amount of time required in the ED to rule out significant cardiac disease. Indeed, since implementation of POC BNP and cardiac marker testing, we have evaluated approximately 4,200 chest pain patients annually (7 percent of our 60,000 total ED encounters per year) with no reported inappropriate discharge to home events.

Clin Biochem Rev. 2008 November; 29(3): S123.

S12 Judicious Use of Cardiac Markers in the Emergency Set-Up

The heart is a non-stop pump supplying nutrition and oxygen to the cells of the body through systemic coronary circulations. Coronary circulation has its unique networks. Any derangement in its own supply would cause cellular anoxia, trigger cellular leakage and jeopardise its pumping function. Acute coronary syndrome (ACS) is diagnosed by elevated cardiac markers, ECG changes and acute chest pain. Although currently there is a strong suggestion to use only the 99th percentile changes in serial cardiac Troponin in the diagnosis of ACS, judicious use of multiple cardiac markers (CK-MB, myoglobin and c-Troponin) could be useful in the early diagnosis of ACS so that timely institution of either thrombolytic or interventional therapy could be done to save this pump.

Any increase in the intraventricular pressure or volume or both stimulate cardiac myocytes which release B-type natriuretic peptide (BNP) detectable immediately in the circulation.

It is quite probable that acute ischaemic events alter the heart pump and bring forth a change in intraventricular pressure and volume thus releasing BNP in the circulation. Therefore, a serial assay of BNP along with CK-MB, myoglobin and c-Troponin could be used not only to diagnose ACS but also to stratify risk associated with it, and further prognosis following therapy.

Clin Biochem Rev. 2008 November; 29(3): S123.

S13 Micronutrients and Vascular Health - Diet or Supplements?

A range of dietary factors may influence vascular health, either directly by altering the function of cells within the arterial wall or indirectly by altering vascular risk factors. Among the active dietary components, substantial evidence supports an important role for nutritional antioxidants. Despite a considerable body of experimental and epidemiological data which supports the importance of dietary modification, clinical trials demonstrating beneficial effects of dietary change are limited.

Antioxidant micronutrients such as ascorbate or tocopherols can be shown to improve endothelial function and increase the resistance of lipoproteins to oxidation when given as individual oral supplements. However, longer term clinical trials in general fail to demonstrate any improvement in clinical endpoints. Several potential reasons exist for this discrepancy, including inadequate understanding of the metabolism and interactions of micronutrients, use of inappropriate pharmacological doses in clinical trials, inadequate duration of exposure and inclusion of antioxidant-replete subjects. In addition, chain-breaking antioxidants may exert pro-oxidant effects in certain conditions, and when administered as pharmacological supplements may adversely affect the availability of other dietary components.

Possible approaches to minimising risks and maximising benefits include conducting studies in defined groups of antioxidant deplete subjects using mixtures of nutrients as supplements or in the form of nutrient-rich foods, or trials which aim to achieve a change in dietary pattern. Dietary interventions which focus on increasing the intake of leafy green vegetables or citrus fruits can be shown to increase concentrations of a range of relevant micronutrients and also to produce improvements in physiological parameters in at-risk populations. Such studies indicate that ascorbate and beta-cryptoxanthin as useful biomarkers of fruit and vegetable intake at both an individual and population level, although published data are difficult to interpret due to lack of assay standardisation.

Clin Biochem Rev. 2008 November; 29(3): S123–S124.

S14 HBA1C Reporting

Measurement of haemoglobin A1c (HbA1c) is used extensively in individuals with diabetes mellitus to monitor long-term glycaemic control and adjust therapy. Large prospective randomised clinical trials, most notably the Diabetes Control and Complications Trial (DCCT) and the United Kingdom Prospective Diabetes Study (UKPDS), documented that HbA1c is a measure of the risk of developing microvascular complications in type 1 and type 2 diabetes, respectively. Based on these data, many diabetes organisations around the world recommend a HbA1c value of 6.5 to 7% as the target for glycaemic control. The HbA1c assay is thus an integral component of diabetes management.

Glycated haemoglobin (GHb) consists of haemoglobin A1a (HbA1a), HbA1b and HbA1c. Over 30 different methods are commercially available to measure GHb. These factors have led to considerable variation among results reported by different laboratories. In the USA and many other countries, the NGSP (previously known as the National Glycohemoglobin Standardization Program) has significantly reduced interlaboratory variation using a standardisation process based on the DCCT reference method. HbA1c values are reported using the same numbers as those in the DCCT and UKPDS. More recently, the IFCC developed a reference method for HbA1c using mass spectrometry. HbA1c values by the IFCC method are significantly lower (~1.5 to 1.9% across the relevant HbA1c range) than NGSP/DCCT results. These findings have generated considerable debate as to how HbA1c should be reported. Should evidence based values from the DCCT/UKPDS continue, should the IFCC “scientifically accurate” values be adopted or should we change to another reporting system (e.g. mean blood glucose)?

Clin Biochem Rev. 2008 November; 29(3): S124.

S15 Mean Blood Glucose

When physiological control of blood glucose levels has been lost in diabetes mellitus, blood glucose levels vary widely following food and exercise. The worse the glucose control and higher the glucose levels, the higher level of glycated HbA1c. HbA1c levels have been clearly shown to predict complications.

HbA1c is linearly related to glucose levels, regardless of whether the glucose levels are fasting, random or following a 75 g glucose load. HbA1c levels are similarly directly related to an individuals mean (or average) glucose level, regardless of whether the estimate is the average of a patients few fingerstick readings per day or an average of a few hundred glucose readings taken over several days using continuous interstitial glucose monitoring.

Mean blood glucose is now a commonly used term that is poorly defined. Consider that nightly glucose levels (when we rest and don’t eat much) are usually lower than daily glucose levels (when we usually eat and exercise). Therefore average glucose estimates performed during daylight hours, which do not include nighttime measurements, will be higher than 24-h averages. HbA1c glycation however is expected to be a 24-h process.

There are two issues causing concern. Firstly, if HbA1c and mean blood glucose are linearly related, can they be interchanged so that HbA1c can be reliably reported in A1c derived average glucose (ADAG) equivalents? While expressing in alternative units may violate theoretical principles of metrology, this could have the practical advantage of reporting HbA1c to patients in the familiar glucose equivalent values and units. The second issue of concern is that if HbA1c and mean blood glucose are directly interchangeable, shouldn’t mean blood glucose also be as predictive of risk as HbA1c? Isn’t it the high mean blood glucose that leads to complications rather than the high HbA1c?

Clin Biochem Rev. 2008 November; 29(3): S124.

S16 Urine Albumin: Standardisation and Reporting Issues.

Urine albumin is well established as an early marker for nephropathy in diabetic subjects. More recently it has also been identified as an important prognostic marker in hypertension and cardiovascular disease. Despite the literature supporting clinical utility there is evidence that clinicians do not often request urine albumin, particularly in their non-diabetic patients. Furthermore, when they do make the request, they find interpretation confusing and adherence to guidelines is poor. Currently there is no reference system for the measurement of urine albumin: in particular the measurand is not defined, there is no reference material and although the manufacturers of the majority of methods in routine laboratory use the same calibrator source, it is a serum based material and individual preparations are not standardised. Despite this, the analytical variation represents a small proportion of the total error in urine albumin measurement. It is the pre-analytical phase which contributes the major variation. Albumin excretion rate (AER) and albumin creatinine ratios (ACR) vary substantially with collection type and patient preparation. Daytime collections yield significantly higher results than nightime collections. Exercise, stress, posture, infection, hydration and blood contamination can contribute to bias in both AER and ACR. Creatinine variation will contribute significantly to variation in ACR but reference ranges are not age/gender/race specific. Decision points for urine albumin, AER and ACR in common use vary widely and none has been clinically validated for the particular method. The IFCC has a urine albumin working group that is collaborating with other professional groups to work towards standardisation; in the meantime individual laboratories can play a crucial role in educating health care providers.

Clin Biochem Rev. 2008 November; 29(3): S124.

S17 Loinc and Observation Coding Systems – What Do I Need this for?

The Logical Observation Identifier Names and Codes (LOINC) is a system for identifying pathology tests and other reportable data, primarily for the purpose of reporting results. Individual laboratories often use different nomenclature conventions for reporting relatively simple pathology tests, making consolidation of results from different providers problematic. By using a specific numeric identifier the odds of correctly identifying a specific test are significantly improved.

Other important coding systems include the IFCC NPU (Nomenclature, Properties & Units) and SNOMED-CT (Systematized Nomenclature of Medicine - Clinical Terms; now maintained by the IHTSDO - International Health Terminology Standards Development Organisation). In Australia, it is expected that SNOMED-CT will gradually take over the roles of specifying orderable tests and reportable results.

Such systems will become increasingly prevalent as Australia and the world moves closer to Electronic Health Records (EHRs). Observation coding systems appears to be a relatively simple concept in understanding and execution, however in practice a number of difficulties arise. These include, but are not limited to, possible ambiguities in reporting units, difficulty in standardising codes from the universe of possible codes for a single test, and difficulties modifying and implementing codes within laboratory information systems (LISs).

The AACB LOINC Working Party was created to foster an improved understanding of LOINC in the Chemical Pathology community, recommend suitable standard codes for LOINC implementation, to help the profession stay abreast of developing health informatics standards, and provide critical feedback on the utility of developing standards.

In conclusion, in order to get maximal utility from EHRs and electronic ordering and reporting, laboratories will need to stay abreast of developing informatics standards and continue to update their LISs to implement the required functionality.

Clin Biochem Rev. 2008 November; 29(3): S124–S125.

S18 Nomenclature, Properties and Units: The IFCC Philosophy

The scientific foundation, management and conduct of laboratory medicine must be supported by computing providing communication solutions to augment the internal flow of data within laboratories and for communication with requesting practitioners. This electronic communication should use concepts that represent examinations and testing conducted in laboratories and concepts used by health care providers. The IFCC and IUPAC in a joint project over 40 years have developed a coding scheme for laboratory medicine based on metrological principles and progress in modern classification science. The IFCC/IUPAC Joint Committee on Nomenclature, Properties and Units (C-NPU) coding scheme is used nationally in Denmark and Sweden. It is based on principles including definitions of the property, the component or biological material and the system or units such that it carries the result unambiguously to the requesting medical practitioner. The syntax of the coding system has been developed and endorsed by the CEN Technique 251 on medical informatics, while the semantic content is based on widely recognised schemes such as those developed by WHO and other international organisations. The underlying electronic structure of the coding scheme provides a technical platform for handling the information in a laboratory information system (LIS). The database is accessible and available for all affiliated National Societies from the IFCC homepage (Scientific Division). Access to an international scientifically documented coding scheme reduces the difficulties and costs required to develop a local LIS. As well the ready availability of a proven, consistent and global coding scheme simplifies quality management in its widest context including electronic health care reporting. Such a coding scheme eliminates the ambiguity of laboratory test requests and reports and thus contributes to patient safety and laboratory integrity.

Clin Biochem Rev. 2008 November; 29(3): S125.

S19 Myeloma – New Tests and Treatments

The management of multiple myeloma is changing rapidly. Improved understanding of the pathophysiology of myeloma has come with new cytogenetic and molecular techniques. New prognostic systems based on clinical and biologic features, such as cytogenetic abnormalities and gene microarray data, have been developed. New guidelines defining response to therapy include tests such as the free light chain assay, bone marrow immunohistochemistry and flow cytometry. Novel agents such as lenalidomide, thalidomide, and bortezomib have been designed to interrupt myeloma growth and survival pathways, and have entered into clinical usage in Australia. Multiple other new agents are also under study or in development. Numerous clinical trials have commenced with these novel agents, alone or in conjunction with established modalities such as conventional cytotoxic agents and stem cell transplantation. These newer treatments have increased the anti-tumour response rates in this disease and have provided options for patients whose disease has become resistant to conventional therapy. Accurate assessment of therapeutic response is becoming more important in this setting. A major challenge will be to define the best use of these new agents and combinations in order to significantly impact the natural history of myeloma and improve overall survival.

Clin Biochem Rev. 2008 November; 29(3): S125.

S20 Update on Prostate Cancer Testing

The identification and management of prostate cancer has been revolutionised over the last twenty years largely due to the advent of the prostate specific antigen (PSA) blood test. Surprisingly this test is not a diagnostic test for prostate cancer and not even specific for the prostate gland as its name would suggest. However, because prostate cancer is a largely subclinical illness in its early course, the impact of this test that reliably predicts the risk of the most common cancer of men has been profound.

Medical professionals and layperson alike are usually surprised to hear that the typical range (95% confidence interval) for cholesterol results in Australia is 3.8–7.8 mmol/L. It is logical to understand that people on the high side of average (>5.5 mmol/L) have higher than average risk of cardiovascular disease. It surprises most people to understand that if the typical range (95% confidence interval) of PSA results in Australia is 0.3–4.0 μg/L, a man with a PSA level above the average has a higher than average risk of prostate cancer. Furthermore, because the PSA distribution is skewed, the median PSA value above which risk is increased is about 1.0 to 1.5 μg/L. Similar interpretations can be made for ‘free to total PSA ratio’ where the median value is 25%, and the rate of rise of PSA (doubling time) which has a median value of five years and risk in both situations is increased at values lower than these medians. These various approaches to PSA interpretation provide the powerful prediction of risk we are looking for, but do not make the diagnosis. While subsequent biopsy and Gleason score are considered the gold standard of diagnosis, these tests are not infallible and new diagnostic tests are constantly been sought and proposed.

Clin Biochem Rev. 2008 November; 29(3): S125.

S21 Update on Faecal Occult Blood Testing

Bowel cancer is the second most common lethal cancer in Australia killing approximately 90 people each week. Faecal occult blood testing can reduce deaths from bowel cancer 15–40% by detecting bleeding polyps or early cancers before they spread. Australia has moved beyond testing small numbers of people who are concerned or symptomatic to nationwide, government-funded testing of all people aged 55 or 65 years. This will soon be expanded to include 50 year olds. Since September 2006 more than one million people have been invited to participate in the program. Subjects collect two rice-grain sized faecal specimens at home and mail them to the central laboratory in tubes containing transport medium. Testing is done with an automated immunochemical method with a positive result reported if blood is detected in either specimen. Both specimens must be negative for a negative result. To date the participation rate has been approximately 40%, the positivity rate 5% and the frequency of collection errors 4%. The growth of the screening program has focused attention on other areas besides faecal occult blood testing such as the quality and availability of colonoscopy services around Australia.

Clin Biochem Rev. 2008 November; 29(3): S125–S126.

S22 Key Incident Monitoring and Management Systems (KIMMS) Workshop

The Key Incident Monitoring & Management Systems (KIMMS) project was initiated by the Quality Assurance Scientific and Education Committee (QASEC) of the Royal College of Pathologists of Australasia. RCPA Quality Assurance Programs P/L (RCPA QAP) acquired the project from QASEC in 2006.

Pathology in Australia has been a leader in the introduction of accreditation and quality assurance for pathology laboratories to demonstrate competence and continual improvement.

Recent studies in many countries have shown that the majority of adverse patient incidents occur in the non-analytical phase of the test-request-report cycle. In order to minimise the risk of errors and incidents in pathology, the pre- and post-analytical phase of testing need to be measured and monitored. KIMMS aims to monitor these pre- and post-analytical phases of the laboratory quality system thus extending measurement of quality to the entire quality system.

The KIMMS 2006 and 2007 Pilots identified a need to improve and clarify key definitions and to develop a uniform national data set, to enable standardised reporting. The Pilot Programs also recommended that the KIMMS program seek formal recognition as a designated Quality Assurance activity, with legally protected privilege, so that participants could openly report incidents without fear of exposure to litigation. The KIMMS Program is pleased to announce that it has now been officially granted Qualified Privilege status by the Australian Government, and similar recognition by the NZ Government is underway.

The aim of this workshop is to further develop this national data set to enable participants to measure and monitor pathology incidents in a standardised manner. This data set now needs to be finalised, prior to work commencing on software development to extract KIMMS data from Laboratory Information Systems (LIS).

The KIMMS Workshop will include a review of:

  • Key definitions
  • Pre-analytical categories
  • Post-analytical categories
  • Incident identification & investigation categories

The workshop will allow participants the opportunity to be involved in developing the national data set. Feedback from these workshops will be utilised to finalise the national data set.

The data generated by KIMMS has the potential to drive systematic improvement of pre- and post-analytical quality in pathology.

Clin Biochem Rev. 2008 November; 29(3): S127.

O1 The Challenges of Lean

Introduction

Internationally Lean Laboratories out-perform conventional laboratories in productivity, profitability and customer satisfaction.1 Many industries have embraced Lean, but despite the promise of greater efficiency many implementations have failed suggesting the approach to organisational change may be a critical factor to success. We describe results achieved by incorporating Lean with effective change management skills.

Methods

Frontline staff were empowered to drive Lean with a ‘learn by doing’ philosophy. Central Laboratory staff applied the principles to integrate the Haematology, Biochemistry and Endocrinology departments. Staff participated in the define and measure phases which included brainstorming sessions, big picture maps, workflow charts, spaghetti maps, cycle times, rework, rate of specimens arriving and capacity versus demand of analysers. Communication strategies were developed to integrate the change with an understanding of why and how.

Results

The Lean Team analysed the ‘current state’ and planned the ‘future state’ by re-designing the laboratory based on the consolidation of work processes. This change impacted on 100 FTE processing 15,000–20,000 specimens per 24 hours. Work-flow was streamlined into groups based on the expertise required, high / low volume throughput and consolidation of analyser platforms. Work processing steps decreased from 22 to 12 by consolidating 70% of high volume work.

Conclusions

A version of ‘Lean’ has been implemented with a bottom up approach, empowering staff to lead the changes with detailed planning, effective communication, motivation and teamwork. Identifying change agents was critical to success because they were able to seek the opportunities and influence others to follow. Textbook ‘Lean’ is easy, but getting people to change is the challenge of Lean.

References

1. [(Accessed January 21 2008)];The Dark Report. www.darkreport.com.
Clin Biochem Rev. 2008 November; 29(3): S127.

O2 Using UM Data to Improve Laboratory Processes Across Pathology Queensland

Introduction

Section 5.6.2 of AS 4633 (ISO 15189:2003) states laboratories must estimate measurement uncertainty (MU) where relevant and possible. Currently there is progressive replacement of all general chemistry and immunoassay analysers in Chemical Pathology at all 33 Pathology Queensland laboratories. The implementation of UM was begun with analytes processed on these instruments.

Methods

A UM data sheet was initially developed for each analyte. UM is calculated for each analyte at each site after an initial period of 2–3 months running on the new assay (unless there are fewer than 30 data points) and then recalculated every six months. A review sheet lists analytes falling outside acceptable limits, likely causes and what action is to be taken.

Results

The state-wide data is then reviewed in two ways –

  1. UM is compared across all sites for each analyte. The goal is to ultimately demonstrate a standardised UM for each analyte regardless of the testing site.
  2. A spreadsheet of analytes falling outside limits is created which can be sorted by site, analyte, control, instrument or comment. This allows trends to be more easily identified. It is this review that highlighted the need for standardised IQC management across the State. The first initiative undertaken was the introduction of state-wide comment codes to encourage appropriate management and annotation of QC results in the LIS.

Conclusions

Calculation of UM is designed to aid the correct interpretation of results. Standardised UM across the state will simplify this process for clinicians using Pathology Queensland laboratories. In addition, review of this data provides an extra tool to monitor, troubleshoot and ultimately improve laboratory processes to achieve the best and most consistent assay precision possible.

Clin Biochem Rev. 2008 November; 29(3): S127.

O3 Evaluation of Analytical Specificity in Method Comparison Studies – The Specificity Plot

Introduction

Method comparison studies typically evaluate precision and between-method proportional and absolute bias. Combining these factors in a single experiment and display using a variation of the difference plot may allow assessment of differences in assay specificity.

Methods

Example data was generated using a Roche Modular <P> analyser and an Ortho-Clinical-Diagnostics Vitros 350 analyser. Samples were obtained from unused serum submitted from GPs for routine analysis. In a single run 24 samples were analysed in duplicate for sodium, creatinine and LFTs on each analyser. Within-run precision was determined on each analyser from the paired measurements. Actual scatter around the line of best fit was compared with scatter predicted from the measured analytical variation using statistical and graphical techniques.

Results

For sodium and protein the between-method scatter of results was within the expected values for the combined analytical precision. For creatinine, ALP, AST, albumin, GGT, LDH and ALT the scatter of results was greater than predicted. This is consistent with different specificities of the two methods for these analytes. Although non-specificity was statistically evident for some samples in the majority of analytes tested, only a single ALP result demonstrated a difference which was of clear clinical significance (33 U/L different at average level of 105 U/L).

Conclusions

A simple graphical analysis has shown differences in analytical specificity between two analysers for several common tests. The approach can be applied to any method comparison data however the study design shown here minimises the effect of precision on the analysis. The study should include the type of samples which are usually analysed in the laboratory. Knowledge of differing analytical specificities may be important when changing methods or combining results from different laboratories.

Clin Biochem Rev. 2008 November; 29(3): S127–S128.

O4 A Damsel in Distress

Introduction

Steroid cell tumours of the ovary are extremely rare, accounting for less than 0.1% of all ovarian neoplasms. These tumours which contain steroid hormone producing cells e.g. lutein cells, leydig cells, and adrenocortical cells, can be divided into 3 groups (a) stromal luteomas, (b) leydig cell tumours and (c) steroid cell tumours, not otherwise specified (NOS). The mean age of presentation for NOS patients is 43 years. This paper details the presentation of a 12 year old girl where the final diagnosis was a steroid cell tumour NOS.

Case Report

A twelve year old girl presented with secondary amenorrhoea accompanied by some virilising features with the suspicion of congenital adrenal hyperplasia. Thyroid function tests, LH, FSH, AFP, hCG and PRL were all within normal limits. E2 result was >15800 pmol/L (RI mid cycle peak 550–1650 pmol/L), progesterone 14.5 nmol/L (RI 2.5–12.0), 17-OH-progesterone 103 nmol/L (RI luteal phase 0.8–12.0), testosterone 49.5 nmol/L (RI <3.2), SHBG >200 nmol/ L (RI 20–110), CA125 422 U/ml (RI <35), cortisol 758 nmol/L (RI 130–650) and inhibin B 111 ng/L (RI 100–250 normal menstruating female).

A pelvic ultrasound was then performed which showed an ovarian mass. Surgery was performed and her right ovary was found to contain a uniform solid grey tan tumour weighing approx 328 g. Histological examination showed strong staining for alpha inhibin (indicating a sex chord stromal tumour), no Reinke crystals seen (Leydig Cell tumour unlikely) and the capsule was intact. Histology results were consistent with a steroid cell tumour (NOS). Subsequent post operative testing has revealed a return to normal of her hormone levels and normal menstruation.

Clin Biochem Rev. 2008 November; 29(3): S128.

O5 Suppression of Bone Resorption Markers by Calcium and Vitamin D Separately and in Combination

Introduction

The combination of vitamin D and calcium has been shown to be more effective than either supplement alone in preventing fractures in people aged 50 years and older. The following study aims to compare the effects of calcium or vitamin D separately and in combination on serum bone resorption markers.

Methods

Nineteen healthy ambulatory women aged 59–74 years were randomised to receive either 1000 IU vitamin D3 daily for 7 weeks or 1000 mg CaCO3 daily for 1 week. Subsequently 1000 mg CaCO3 daily for 1 week or 1000 IU vitamin D3 daily for 7 weeks were added to their regimen respectively so that both groups crossed over to the combination treatment. Fasting blood samples were collected for biochemistry, PTH and crosslaps at baseline, after either of the supplements alone and after the combination period. An inhouse questionnaire was used to estimate the habitual dietary calcium intake.

Results

The mean dietary calcium intake was 778 mg/day (444–1487). Ionised calcium rose significantly from baseline (mean 1.20 mmol/L) after calcium and vitamin D combination (mean 1.26) compared to either supplement alone (mean 1.20 and 1.24 after calcium or vitamin D respectively; p<0.005). Serum PTH fell significantly after calcium alone (mean 1.26; p = 2.8x10–6), vitamin D alone (mean 1.24; p = 0.0003) and after the combination (mean 1.26; p = 4.13x10–6) compared to baseline (mean 6.4 pmol/L). Serum beta crosslaps fell significantly from baseline (mean 456 ng/L) only when calcium and vitamin D were used in combination (mean 275; p = 0.003).

Conclusions

The use of calcium with vitamin D has a more significant effect on suppression of bone resorption markers which may lead to more effective prevention of osteoporotic bone loss. This effect appears to be caused by the greater suppression of PTH with the combination therapy than either supplement alone.

Clin Biochem Rev. 2008 November; 29(3): S128.

O6 Antidiabetogenic Activity of Morus Rubra Leaf Extract in Streptozotocin-Induced Diabetic Rats: A Herbal Remedy

Introduction

Researchers all over the world are exploring herbal supplements to control diabetes and its complications. In our study, we have assessed antihyperglycaemic, antihyperlipidaemic and antioxidant activity of aqueous extract of Morus rubra leaves in streptozotocin (STZ)-induced diabetic rats.

Methods

Rats were divided into four groups (n = 5) i.e. normal control, diabetic control, diabetic + extract (400 mg/kg) and diabetic + glibenclamide (600 μg/mg). After 21 days of treatment, blood samples were taken from overnight fasted animals. Fasting blood glucose, glycosylated haemoglobin (GHb), insulin, C-peptide levels and lipid parameters were measured. Antioxidant enzymes like superoxide dismutase (SOD) and catalase, reduced glutathione (GSH) and lipid peroxidation (LPO) were also measured.

Results

Oral administration of extract to diabetic rats showed significant (p<0.001) reduction in blood glucose when compared with the diabetic control group. Levels of GHb, insulin and C-peptide were found to be near normal following treatment with extract. A significant (p<0.001) decrease in triglycerides, total cholesterol and low-density lipoprotein and increase (p<0.01) in high-density lipoprotein was observed in extract-treated rats as compared to diabetic control rats. Activity of antioxidant enzymes and GSH level were significantly (p<0.001) enhanced, while LPO was significantly (p<0.001) suppressed in extract-treated group compared to diabetic control. The effect of extract was comparable to that of glibenclamide.

Conclusions

It is evident from the data that aqueous leaf extract of Morus rubra leads to glycaemic and lipidaemic control. The effect could be due to insulinogenic action of the extract as revealed by insulin and C-peptide analysis. The study also demonstrates its free radical scavenging activity and antioxidant response, hence the extract may be protective against diabetic complications.

Clin Biochem Rev. 2008 November; 29(3): S128.

O7 Vitamin A and E Methods: Does A Common Calibrator Improve the between Laboratory Agreement of Results?

Introduction

The determination of Vitamin A and E levels relies heavily on the performance and fit for purpose design of the analytical techniques employed. Since the RCPA-QAP Vitamins Program began in 2000, the RCPA-QAP have observed a wide spread of data returned by its participants. The aim of this study was to determine whether the use of a variety of calibrators had a significant effect on the distribution within an external quality assurance program.

Methods

Five laboratories enrolled in the RCPA-QAP Vitamins Program were selected for participation in this study. Each laboratory performed Vitamin A and E analysis by isocratic HPLC separation with spectrophotometric detection. A common calibrator was distributed to each of the participating laboratories. This calibrator was analysed in parallel with RCPA-QAP samples throughout the first half of 2006 (Cycle 14), and the results were submitted to the RCPA-QAP for statistical analysis.

Results

The RCPA-QAP results submitted were compared to the target RCPA-QAP results for both Vitamin A and E. The results were then recalculated based on the common calibrator results submitted. Calculation of the dispersion of results for the raw data compared to the recalculated data, demonstrated a decrease in between laboratory imprecision at the higher analytical levels for both Vitamin A and E. Comparison of the raw data compared to the recalculated data demonstrated an improvement in accuracy compared to the target values with a corresponding decrease in the absolute number of results outside the allowable limits of performance.

Conclusions

The use of a common calibrator improved the spread of results seen between participants of the RCPA-QAP Vitamins Program. However, this standardisation experiment does not account for all the dispersion seen and other factors including vitamin recovery and chromatographic separation need to be considered. A pilot of a common calibrator by all participants enrolled in the RCPA-QAP Vitamins Program will be considered by the AACB VWP.

Clin Biochem Rev. 2008 November; 29(3): S129.

O8 Diagnostic Accuracy of Serum Caeruloplasmin in Wilson Disease - Determination of Sensitivity and Specificity by Receiver Operating Characteristic Curve Analysis Among ATP7B Genotyped Subjects

Introduction

A serum caeruloplasmin concentration below 0.20 g/L is conventionally considered as one of the major diagnostic criteria for Wilson disease. However, this decision threshold has not been fully validated for its diagnostic characteristics. In this study, we evaluate various decision thresholds of serum caeruloplasmin concentration in the diagnosis of Wilson disease based on genotype-verified Wilson disease patients, carriers and normal individuals.

Methods

Serum caeruloplasmin concentration was measured by a nephelometric method in 57 Wilson disease patients and 71 family members (49 heterozygotes and 22 wild-type homozygotes), a validation group of 25 subjects clinically suspected of Wilson disease and 690 normal individuals. Receiver operating characteristic analysis was performed using Analyse-it software. The genotypes were confirmed by direct DNA sequencing of ATP7B gene.

Results

Serum caeruloplasmin concentrations below 0.20, 0.14 and 0.10 g/L showed positive predictive values of 48.3%, 100% and 100% respectively and negative predictive values of 98.7%, 97.1% and 91.9% respectively. In the validation group, a serum caeruloplasmin threshold of 0.14 g/L rendered 100% sensitivity and specificity. Forty of 690 healthy subjects had serum caeruloplasmin concentration below 0.20 g/L. However, these 40 individuals had normal genotypes by DNA sequencing. Additionally, none of the 40 individuals had caeruloplasmin concentrations below 0.14 g/L.

Conclusions

The diagnostic accuracy for Wilson disease using a serum caeruloplasmin concentration of 0.14 g/L as the local decision threshold was better than that using a threshold of 0.20 g/L. We suggest that laboratories providing caeruloplasmin assays determine decision thresholds based on local populations.

Clin Biochem Rev. 2008 November; 29(3): S129.

O9 Simple and Rapid Screening Test for Drugs of Abuse in Meconium

Introduction

The determination of foetal exposure to drugs of abuse (DoA) has clinical values in the future management of the newborn. These DoA and their metabolites enter the foetal circulation during the exposure and pass into the amniotic fluid via urine and other body secretions. A proportion of these DoA is deposited in the meconium when the foetus ingests the amniotic fluid. Therefore, meconium provides a clinical sample which has a longer history of drug exposure than urine. We reviewed a number of methods and found them to be tedious and labour intensive. A simple and rapid method of screening for selected DoA was investigated so that we could readily adapt to our existing testing platform.

Methods

Meconium samples were homogenised by adding equal weight of deionised H2O and vortexing. Samples chosen for this investigation had been proven negative for selected DoA by solvent partitioning and analysis on a Hewlett Packard 6890 gas chromatograph and 5972 mass spectrometer in SIM mode. Samples of blank homogenised meconium (hMec) were spiked with standards at the following concentrations: nordiazepam 200 ng, morphine 300 ng, benzoylecgonine 150 ng, phenobarbital 600 ng per 1 g of hMec. Samples were extracted with an ethanol:methanol mixture, then vortexed followed by ultracentrifugation before analysis by Enyme Multiplied Immunoassay Technique (EMIT).

The analytes and concentrations (nordiazepam 200 ng/mL, benzoylecgonine 150 ng/mL, morphine 300 ng/mL and phenobarbital 600 ng/mL) at which we used to determine if a significance difference existed between negative and positive spiked sample populations was primarily based on the Australian Standard AS4308. Further investigation may prove a lowering of screening cut-off concentrations is possible.

Results

A minimum of 14 negative and 10 positive samples for each of the four drug classes were analysed and statistically evaluated. Absorbance values of the spiked samples were significantly higher than those of the blank samples for each of the spiked analytes at a probability of 0.05 for cut off values previously defined.

Conclusions

This is a rapid, cost effective and simple screening test that can be readily adapted for routine laboratory screening of meconium for these selected DoA.

Clin Biochem Rev. 2008 November; 29(3): S129.

O10 Can Urinary Free Metanephrines Substitute for Free Catecholamines or Total Metanephrines for Testing for Phaeochromocytoma?

Introduction

The measurement of free metanephrines in plasma has been proposed as the best test for the biochemical screening for phaeochromocytoma (PHAEO). The aim of this work was to investigate the diagnostic utility of urinary free metanephrine excretion instead of total (free + conjugated) metanephrines or free catecholamines in 24-hour urine collections.

Methods

Biogenic amines in urine with and without acid hydrolysis were extracted by solid-phase extraction on Bond Elut Plexa from 12 patients with known PHAEO and 23 controls. Metanephrines and catecholamines were measured by tandem mass spectrometry (LC-MSMS) with deuterated internal standards and multiple reaction monitoring over 6 min of 15 ion transitions for 5 biogenic amines. Quantitation was performed using peak areas and Analyst software.

Results

Urinary free metanephrines accounted for a mean of 15–20% of the total, with no significant difference between patients with and without PHAEO, and between normetanephrine (NMET) and metanephrine (MET). Using an upper limit of normal excretion (ULN) of <600 nmol/d for free NMET, both free NMET and total NMET were on average 20 x ULN in the PHAEO patients compared to 8 x ULN for free noradrenalin. An ULN of <250 nmol/d for urine excretion of free MET produced the same classification of patients to total MET and adrenalin.

Conclusions

Free metanephrines in urine appear to be similar to total metanephrines for discriminating between patients with and without PHAEO. Their measurement avoids urine acid hydrolysis, and can be performed simultaneously with catecholamines with a single LC-MSMS method.

Clin Biochem Rev. 2008 November; 29(3): S129–S130.

O11 Urinary Total Metanephrines Measured by HPLC and Tandem Mass Spectrometry (LCMSMS)

Introduction

For the biochemical diagnosis of phaeochromocytoma, expert opinion favours measurement of metanephrines in urine or plasma over catecholamines. The aim of this study was to use a LC-MSMS assay, developed in our laboratory for biogenic amines, to quantitate total urine metanephrines and to compare its performance with our routine method of HPLC coupled to electrochemical detection (ECD).

Methods

Conjugated metanephrines in urine were acid-hydrolysed at 90oC for 25 min prior to the addition of free deuterated metanephrines as internal standards. After neutralisation, sample clean-up with good recovery was achieved on weak cation-exchange cartridges and the metanephrines eluted with acidic acetonitrile. LC-MSMS was carried out on a Waters Atlantis T3 column (150 x 2.1 mm) coupled to an AB 3200 Q-Trap with an electrospray source and a run time per sample of 5 min.

Results

The LC-MSMS assay (y) showed good linearity from 0.1 to 5 μmol/L, high sensitivity with S/N of 10 at 0.01 μmol/L, and between-run precision with CV between 4 and 8% (n = 8) with two commercial QC materials. Correlation with HPLC-ECD (x) was very good (r = 0.98), as was agreement in urine samples from 94 patients for metanephrine (y = 0.96x – 0.04; Passing-Bablok regression) and normetanephrine (y = 0.89x + 0.02) concentrations. Both methods correctly classified 18 patients with known phaeochromocytoma as having excessive urinary excretion of total metanephrines (> 4.3 μmol/day).

Conclusions

LC-MSMS provides a faster, reliable alternative to conventional HPLCECD with lower costs of labour and consumables for the assay of total metanephrines in urine.

Clin Biochem Rev. 2008 November; 29(3): S130.

O12 Cardiac Troponin Increases among Marathon Runners in the Perth Marathon (The Trim Study)

Introduction

Research from the 2002 Boston marathon found that troponin increased in 68% of runners, with 11% at levels typically diagnostic for acute myocardial infarction. We aimed to replicate and extend this study in the Perth marathon.

Methods

Entrants in the 2007 Perth marathon were recruited 1–2 days before the race. Clinical assessments and biochemical investigations were performed pre- and post-race, and a range of demographic and training history data collected. Simple descriptive statistics were performed to describe the prevalence of troponin increases (measured on a Centaur, Siemens Diagnostics) and runner characteristics.

Results

Ninety two (27%) of the 346 runners were enrolled, and 88 (96%) completed the marathon. Sixty five (71%) were male, the mean age was 43.1 ± 9.8 yrs (range 25–64) and the mean BMI was 24.1 kg/m2. Twenty eight (31.8%) of those who completed the race had a raised troponin I (>0.1 μg/L), the highest being 1.43 μg/L. Using logistic regression, the strongest predictor for developing an elevated troponin I was a decrease in weight with an odds ratio (OR) of 0.56 (95% confidence interval [CI] 0.33–0.96). The post-race eGFR was also associated with a raised troponin I (OR 0.96, 95%CI 0.92–0.99) while pre-race eGFR, age, sex, BMI, training factors, marathon experience and race time were not.

Conclusions

Troponin increases were relatively common among marathon finishers. The strongest predictors were weight decrease and post-race eGFR. The pathophysiology of raised troponin in marathon runners is unclear but unlikely to represent classic myocardial infarction. Clinicians need to be aware of this in the event of an athlete presenting to an emergency department post marathon.

Clin Biochem Rev. 2008 November; 29(3): S131.

P3 Study of ANTI-CCP, Rheumatoid Factor and Chronic Disease

Introduction

Rheumatoid Arthritis (RA) is an autoimmune disease that causes joint inflammation and damage. Rheumatoid factor (RF) is the common serological marker which is used for the diagnosis of rheumatoid arthritis, but it is non-specific and can also be positive in other autoimmune disease and inflammatory disease. For example, patients with Hepatitis C may also develop arthritis, and the RF may be positive and therefore unhelpful for distinguishing the cause of the arthritis. Antibody to citrullinated proteins is present in patients with RA and appears to be a more specific indicator of the disease. The aim of this study was to evaluate the Abbott AxSYM method for anti-CCP (anti-cyclic citrullinated peptide antibody), particularly in patients with inflammatory disease.

Methods

To investigate the extent of possible interference in patients with inflammatory disease, anti-CCP (Abbott) and RF (Abbott) was run on patients positive for hepatitis B (core antigen and surface antigen) and hepatitis C (antibody).

Results

No patients we studied with hepatitis had a positive value for anti-CCP, despite many having borderline elevated RF. (Prior to the study, there had been a patient with Hepatitis C with a slightly raised anti-CCP.) Correlation with randomly selected RF results using a Bland Altman scatter plot indicated that there was a good correlation between many RF results and anti-CCP results. However, there was also a group with poor correlation, with raised RF and only borderline raised anti-CCP. This may indicate that RF can be raised in other chronic diseases, whereas anti-CCP is more specific for RA.

Conclusions

The AxSYM method for anti-CCP appears to be more specific for RA than RF in patients with chronic inflammatory disease such as Hepatitis B and C.

Clin Biochem Rev. 2008 November; 29(3): S131.

P4 Identification of a Novel Mutation in a Patient with Variegate Porphyria

Introduction

Variegate porphyria is an autosomal dominant disorder that results from mutations in protoporphyrinogen oxidase (PPOX), an essential enzyme in the haem biosynthetic pathway. Symptoms may manifest as neurological dysfunction and/or cutaneous photosensitivity.

Symptoms may be precipitated by numerous environmental factors including certain drugs in asymptomatic patients. Although the usual biochemical markers for variegate porphyria including raised urine PBG excretion, raised faecal coproporphyrin, faecal protoporphyrin, and raised faecal coproporphyrin III/I isomer ratio are suitable for diagnosis of symptomatic patients they may not identify carriers of the mutation. This, coupled with the fact that variegate porphyria has a post-pubescent onset, has indicated a need to begin mutation analysis for familial cascade testing for asymptomatic individuals including juvenile family members. This study examines one novel mutation identified in the PPOX gene.

Methods

Primers were designed for all 13 exons of the PPOX gene including the intron/exon boundaries, promoter, and poly-A tail regions. PCR products were analysed using Sanger sequencing. The DNA sequences and significance of the mutations were then analysed in silico.

Results

In one individual, the DNA sequence showed a novel mutation identified in exon 11 at c.1132delG. This caused a frameshift that added 23 incorrect downstream amino acids before a new stop codon was introduced. This was confirmed using the forward and reverse sequence.

Conclusions

The primers designed and the methods employed were able to identify a novel mutation in the PPOX gene with strong evidence of pathogenicity in a patient with variegate porphyria.

Clin Biochem Rev. 2008 November; 29(3): S131.

P5 Relationships from the Laboratory Database

Introduction

Our aim was to determine the relationships between some analytes using the statistical tools of moving averages and medians.

Methods

Moving averages/medians of between 50 and 500 analytical results were used to study hundreds of samples from the Austin Pathology database.

We studied the relationships of the thyroid function tests (TFT); the oral glucose tolerance test (OGTT) and glycated haemoglobin (GHb); serum potassium and creatinine.

Results

  1. TFT (6474 samples): showed the expected relationships between free thyroxine (FT4), free triiodothyronine (FT3) and thyrotropin (TSH). Older women had higher TSH at the same FT4 levels (12–15 pmol/L) than younger.
  2. OGTT and GHb (828 outpatients and 927 inpatients): There was a continuous relationship between both fasting and 2-hour plasma glucose and GHb from normal to abnormal levels. Both fasting and 2-hour glucose increased with age to plateau at about 60 years. The inpatients had higher GHb at lower levels of plasma glucose, perhaps because the glucose had been corrected in hospital.
  3. Serum potassium and creatinine (24326 samples): Overall, potassium rose with increasing creatinine levels. However, between 175 and 250 μmol/L creatinine there was a dip and rise in potassium, which in the spinal unit patients was seen between 140–200 μmol/L.

Conclusions

Our findings support the expected observations for thyroid function and glycaemia and show the continuous relationship between the analytes from normal to abnormal range. The relationship between creatinine and potassium requires further study. The study demonstrates the usefulness of moving averages/medians to mine the laboratory data base.

Clin Biochem Rev. 2008 November; 29(3): S131–S132.

P6 Comparison between Serum Adenosine Deaminase (ADA) & CA-125 in Ovarian Cancer

Introduction

Epithelial ovarian cancer demonstrates high mortality due to diagnosis at an advanced stage. In the search for the biomarker for early diagnosis and a target for therapy, the issue of whether adenosine deaminase, shown to be expressed on variety of human cancers, malignant condition and many other infections, is expressed in ovarian cancer and if expressed then to which extent. In addition, this enzyme can serve as a biomarker in diagnosis and prognosis of ovarian cancer. Appropriate therapeutic measures can improve the life expectancy of patients with ovarian cancer.

Methods

The study was conducted among 185 subjects comprised of 50 normal, 50 benign and 85 ovarian cancer. The study groups were all female and the age group of study population was over 35. Adenosine deaminase and CA-125 were estimated using ELISA and enzymatic methods in all cases and control.

Results

The sensitivity of ADA as tumor marker was found to be 54% and the specificity was 88% in malignant ovarian cancer. There was a positive correlation between ADA and histology of ovarian cancer. Both CA-125 and ADA levels are low in normal and benign groups and CA-125 is more specific marker than ADA as there is no significant difference of CA-125 serum level between normal and benign group.

Conclusions

ADA can serve as another serum biomarker in addition to CA-125 and may be useful in diagnosing ovarian cancer along with other clinical trials. Also it is easy to perform in small laboratories set-up, simplicity of technique, rapid turn around time and cost effective as compared to CA-125 estimation in developing countries like Nepal.

Clin Biochem Rev. 2008 November; 29(3): S132.

P7 Low Plasma Magnesium Results

Introduction

We have recently observed differences in plasma magnesium results depending on the method of analysis used. In the 24 regional PathWest laboratories, magnesium is assayed on the Ortho Clinical Diagnostics Vitros 250 dry chemistry system and in the Nedlands laboratory Roche reagents are used on a Hitachi 917. Low plasma magnesium levels (<0.5 mmol/L) are clinically significant, as they are associated with hypocalcaemia and hypokalaemia.

Methods

Plasma samples for magnesium analysis received by either the Regional or Nedlands laboratories were assayed on both the Roche Hitachi 917 and Vitros 250 systems. Statistical comparisons were done using Biostats and Microsoft Excel software.

Results

One hundred samples were analysed for magnesium levels by both methods; 51 samples in the range 0.7–1.8 mmol/L and 49 in the range 0.2–0.8 mmol/L. Passing and Bablok regression analysis of all results gave a slope of 1.20 and an intercept of −0.25. In the 0.2–0.6 mmol/L range, magnesium results from the Vitros 250 analysers were on average 0.14 mmol/L lower than the Hitachi 917. The difference between the results ranged from 0.02–0.29 mmol/L. In the range 0.7–1.1 mmol/L results from the Vitro 250 were on average 0.05 mmol/ L lower than the Hitachi 917. The RCPA QAP General Serum Chemistry program covers magnesium in the range of 0.53 to 1.8 mmol/L.

Conclusions

The Vitros 250 system produces lower plasma magnesium results than the Hitachi 917, particularly for patient results <0.6 mmol/L. This discrepancy is not observed in the external quality control program, as it does not monitor the assay performance adequately at this level. The frequency of low magnesium results from our remote laboratories appeared to be more common in certain areas, with 71% of results coming from 5 of the 24 regional laboratories. We plan to do further work in this area.

Clin Biochem Rev. 2008 November; 29(3): S132.

P8 Assessment Of Cystatin C as A Renal Marker in Oncology Patients

Introduction

The determination of glomerular filtration rate (GFR) is crucial for monitoring the use of nephrotoxic chemotherapy in oncology patients. The aim of our study was to investigate whether plasma cystatin C and derived equations are sufficiently accurate in predicting GFR to replace the use of radionuclide GFR and creatinine based estimated GFR (eGFR).

Methods

Creatinine measured by the Jaffe reaction and a HPLC method, cystatin C levels, and 99mTc-DTPA GFR were determined in oncology patients. Calculations of bias, correlation coefficients and percentage estimates between creatinine based equations, cystatin C equations and measured GFR were compared.

Results

Preliminary results on 33 patients showed equations based on cystatin C and creatinine HPLC performed better than equations based on Jaffe creatinine method. The 186 and 175 MDRD formula produced statistically significant underestimations of GFR (mean difference −24.0 mL/min/1.73m2 and −28.5 mL/min/1.73m2 respectively, p<0.05). Cystatin C equations did not show any statistically significant bias and the proportion of results within 30% of true GFR by all cystatin C equations were higher (p<0.05) than the 175 MDRD equation. ROC curves analysis with a cut-off GFR of <60 mL/min/1.73m2 showed cystatin C equations and equations by the HPLC method (both AUC = 1.00) performed better than equations estimated by the Jaffe method (AUC = 0.947).

Conclusions

Cystatin C based equations and HPLC creatinine based equations performs better than routinely used MDRD equations in the prediction of GFR in oncology patients. A larger population is needed to validate the use of cystatin C in this population.

Clin Biochem Rev. 2008 November; 29(3): S132.

P9 In Vivo Total Antioxidant Capacity of Supplemented β-Carotene and Vitamin A in Asbestos Exposed Former Workers

Introduction

Carotenoids, vitamin A and E have been used as antioxidant supplements aiming to remove free radicals which may damage sensitive structures and triggering diseases such as cancer. Several studies have not shown these beneficial effects. A meta-analysis of 68 interventional trials recently showed β-carotene, vitamin A and E increased mortality risk by 7, 16 and 4% respectively. In this study, we evaluated the association between in vivo total antioxidant capacity (TAC) and antioxidants, specially β-carotene and vitamin A supplemented to people previously exposed to asbestos.

Methods

A sample of 313 former male Wittenoom workers exposed to blue asbestos was selected from the Vitamin A Program, essentially a cancer prevention program in asbestos-related diseases which supplemented participants with β-carotene or vitamin A over 11.7 years on average with blood collected for analysis annually. β-carotene, vitamin A and E were measured by a HPLC method. Albumin and bilirubin were measured by an autoanalyser. TAC was measured by the antioxidant inhibition of oxygen radicals (AIOR) method.

Results

Plasma β-carotene was inversely associated with TAC in both the univariate and multivariate analysis (−0.070, 95% CI −0.100 to −0.040, p<0.001). Vitamin A was positively related to TAC in the univariate analysis but statistical significance was lost after adjustment for albumin. Vitamin E and albumin were positively associated with TAC (0.007, 95% CI 0.000 to 0.010, p<0.001) and 0.076, 95% CI 0.060 to 0.090, p<0.001) respectively. Bilirubin was not related to TAC.

Conclusions

Vitamin supplements given as antioxidants did not necessarily contribute positively to TAC. Assessment of in vivo TAC ascertain the contribution of vitamin supplements as antioxidants. The result may help to explain the increased mortality risk with vitamin supplements in several studies but further investigation are required to confirm these findings.

Clin Biochem Rev. 2008 November; 29(3): S133.

P10 Total Antioxidant Capacity as Predictor of Mesothelioma and Lung Cancer in People Exposed to Blue Asbestos in Wittenoom Western Australia

Introduction

Increased rates of death from asbestos-related diseases have been reported in 7000 people exposed to blue asbestos at Wittenoom in Western Australia. Previous studies have shown that plasma vitamin E levels are inversely associated with lung function, while plasma retinol levels have been associated with a lower incidence of mesothelioma and lung cancer after 5 and 15 years of follow up. Total antioxidant capacity (TAC) has been suggested as an additional predictor for the efficacy of antioxidants in asbestos-related diseases.

Methods

Plasma TAC were compared in 63 cases of mesothelioma, 50 lung cancer and 200 controls from a cohort of former male Wittenoom workers. The follow up for mesothelioma, lung cancer and control cases were 11.7 years on average. TAC was measured by the antioxidant inhibition of oxygen radicals (AIOR) method.

Results

After multivariant adjustment for age and smoking, average plasma TAC were not different among cases and controls (p>0.05). Plasma TAC levels were substantially the same over time for mesothelioma (−0.005 TAC units/ year, p = 0.71) and lung cancer (0.003 TAC units/year, p = 0.88). There was a significant increase of TAC levels over time in controls (0.011 TAC units/year, p = 0.05). A decrease in TAC has been observed during the last 4 years (0.124 mM Trolox equivalents, p = 0.09) before developing mesothelioma but not in lung cancer.

Conclusions

Average plasma TAC were similar among the three groups except for a decrease in TAC observed in the 4 years prior to the diagnosis of mesothelioma. The results suggest that an increase in oxidative stress may be associated with the onset of mesothelioma. Confounding factors such as lifestyle, diets, supplements and oxidative stress contribute to TAC which were not accounted for in this study. Further research is required to confirm these findings.

Clin Biochem Rev. 2008 November; 29(3): S133.

P11 The Relationship of Red Cell Size to ‘Active B12’ Levels

Introduction

Low levels of vitamin B12 are known to be a cause of macrocytosis. Attempts to define the optimal level of vitamin B12 based on changes in red cell morphology have been variable in their success. We examined the relationship of the new holotranscobalamin assay (‘Active B12’) to red cell morphology.

Methods

Active B12 was measured on the Abbott AxSYM routinely on all patients with a total vitamin B12 level (measured on Abbott Architect) ≤ 200 pmol/L. We performed 18,263 Active B12 analyses between May 2007 and March 2008, where we also had a full blood examination requested and performed (Beckman Coulter).

Results

The table below shows that the average MCV was increased only in patients with a very low Active B12 (<5 pmol/L). However, the prevalence of both microcytosis (<84 fL) and macrocytosis (<97 fL to <100 fL depending on age) increased at active B12 levels below 20 pmol/L.

Table

Red cell morphology in people with different levels of Active B12.

Active B12nMCV (fL)MicrocytosisMacrocytosisBoth
<5.05597.911%35%45%
5.0 9.927990.511%13%23%
10.0 14.975089.713%9%22%
15.0 19.9155090.010%8%17%
20.0 24.9234890.27%6%13%
25.0 29.9287490.37%6%13%
30.0 34.9283090.07%6%13%
35.0+757790.47%7%13%

Conclusions

MCV is an insensitive marker of B12 deficiency as the rates of both microcytosis and macrocytosis are apparently increased in people with vitamin B12 deficiency. This is probably due to the well know coexistence of iron deficiency in B12 deficient patients.

Clin Biochem Rev. 2008 November; 29(3): S133–S134.

P12 Defining Cut-Offs for Total Protein, Globulin and Age that Optimise Detection of Paraprotein

Introduction

Routine liver function tests (LFT) often have an unexpected finding of an elevated protein level which may suggest the presence of a paraprotein. In this study we examined cut-off values of total protein, globulin and age to predict when a serum protein electrophoresis might reveal a paraprotein.

Methods

Serum specimens with varying concentrations of total protein (>80 g/L) and globulin (≥ 35 g/L) but with normal CRP and FBE were randomly selected for protein electrophoresis on the Sebia Capillary system. The specimens with abnormal bands were immunofixed to identify monoclonal proteins. Results were evaluated using receiver operated characteristic (ROC) curves.

Results

Fifteen (13%) monoclonal protein bands were detected in the 117 specimens. The majority of bands were IgG (73%) with the remainder being IgM (20%) and IgA (7%). Optimal cut off values obtained from ROC curve analysis were: total protein value 84 g/L (sensitivity 53% and specificity 54%); globulin value 38 g/L (sensitivity 60% and specificity 50 %); and age ≥ 50 years (sensitivity 80% and specificity 34%). Combination of these three optimal cut-offs gave a positive predictive value of 30% and a negative predictive value of 95%.

Conclusions

Total protein and globulin values in LFT are useful for case finding of individuals with possible early stage of multiple myeloma. The optimal cut-off values were 84 g/L for total protein and 38 g/L for globulin in patients who are ≥50 years and without laboratory evidence of inflammation.

Clin Biochem Rev. 2008 November; 29(3): S134.

P13 A High Sensitivity Thyroglobulin Assay to Monitor Thyroid Cancer is Essential – Is Beckman Best?

Introduction

A major role of thyroglobulin assays is to detect and monitor recurrent thyroid cancer following complete thyroidectomy. Early detection of recurrence requires an assay of the highest possible sensitivity. Ideally such an assay can detect thyroglobulin during TSH suppression as well as during stimulation (thyroxine withdrawal or after recombinant TSH known as thyrogen). Beckman-Coulter claim to produce a routine assay with the best sensitivity. Where assay sensitivity is critical to its clinical interpretation it is important for laboratories to determine the functional sensitivity (FS; lowest level at which CV is 20% or better) in their own setting.

Methods

We assessed the FS of the Beckman Access thyroglobulin assay using post thyroidectomy sera with thyroglobulin levels of 0 to 2 μg/L by serial measurements over several weeks with multiple calibrations and reagent lots.

Results

An external file that holds a picture, illustration, etc.
Object name is CBR29_S2_pgS115_G1.jpg

The FS was 0.1 μg/L. This is well below the normal thyroglobulin reference range and superior to the claimed sensitivity of other routine methods on the Australian market. When testing for recurrent thyroid cancer in post thyroidectomy patients the highest achievable sensitivity must always be sought. We also note that FS of the Beckman assay is well below the lowest value seen in the RCPA-QAP programme introduced in 2008 which is about 4 μg/L.

Conclusions

The Beckman Access assay has a functional sensitivity of 0.1 μg/L which clearly exceeds the claimed sensitivity of the other main assays used in Australia. The EQA programme introduced by the RCPA-QAP in 2008 should be adjusted in future to evaluate performance closer to the all-important functional sensitivity of the available thyroglobulin assays. Functional sensitivity is an especially important parameter for any assay where any detectable level of the analyte has immediate diagnostic significance and laboratories should confirm their functional sensitivity by appropriate studies on their own analytical platform.

Clin Biochem Rev. 2008 November; 29(3): S134.

P14 A Case of Seriously Low Cholesterol

Introduction

There is a general understanding that the lower one’s cholesterol the better. However there are some specific conditions in which a low cholesterol is due to a serious underlying metabolic disorder. Recognition and treatment of such conditions in childhood is necessary to prevent mental retardation or even premature death.

Clinical presentation

A 7yr old girl in the Clinical Genetics Clinic was noted to have mental and growth retardation, neuromuscular impairment, evidence of malabsorption, congenital malformations including syndactyly. There was a family history of cleft palate. The constellation of clinical signs suggested the possibility of Smith-Lemli-Opitz (SLO) syndrome.

Results

Cholesterol was 1.7 (RR 2.3–5.4 mmol/L) and 7-dehydrocholesterol 111 (RR <10 μmol/L). The results confirm SLO (deficiency of 7-dehydrocholesterol reductase) and the child was treated with cholesterol supplements with rapid improvement in mental function.

Discussion

Apart from malignancy, sepsis and malabsorption there are two conditions that may cause seriously low cholesterol: abetalipoproteinaemia (chol <1.25) and SLO (chol 0.2–2.6). Relatively common causes of syndactyly include Down’s syndrome and hereditary syndactyly. Rare causes include: Apert syndrome, Carpenter syndrome, Cornelia de Lange syndrome, Pfeiffer syndrome, Smith- Lemli-Opitz syndrome, and using the medication hydantoin during pregnancy (foetal hydantoin effect).

Conclusions

Cholesterol is seldom tested in young children but where specific clinical signs such as described above are present it is essential for the astute physician to consider SLO and measure both serum cholesterol and 7- dehydrocholesterol. Early diagnosis can have a major impact on the long term prognosis.

Clin Biochem Rev. 2008 November; 29(3): S134–S135.

P15 Congenital Disorders of Glycosylation – Are We Missing Them?

Introduction

Many proteins in the human body carry post-translational modifications, the most common being glycosylation. Addition of carbohydrates to proteins requires complex machinery and demands coordinated expression of hundreds of genes. Because of the important biological functions of these glycoproteins, congenital disorders of glycosylation (CDG) may result in severe and clinically diverse multi-system involvement.

Patient

A baby was transferred to newborn services with a range of dysmorphic features including low set ears, inverted nipples and short proximal limbs. Additionally the baby had hypertrophied ventricles, severe pulmonary hypertension and respiratory distress. The mother had three previous miscarriages. Investigations included blood gas, serum electrolytes, serum transferrin isoforms by isoelectric focusing (IEF) and immunofixation, urine metabolic screen, leucocytes, enzymes, genetic studies.

Results

High levels of diasialo-and asialo-transferrin in serum. Markedly reduced level of peripheral leucocyte phosphomannomutase (PMM) activity. Compound heterozygote for the D148N and V231M mutations.

Discussion

CDG are a rapidly expanding group of metabolic syndromes with wide symptomatology and severity. All stem from deficient N-glycosylation of proteins. Main features of CDG include psychomotor retardation, ataxia, seizures, retinopathy, liver fibrosis, coagulopathies, failure to thrive, and dysmorphic features. When CDG is suspected serum should be analysed for deficient transferrin glycosylation (isoforms) either by IEF or mass spectroscopy. However some defects do not manifest as abnormal transferrin isoforms. Definitive diagnosis in all cases is made by specific leucocyte enzyme studies and genetic testing.

Conclusion

Diagnosis of CDG requires a good understanding of the diverse presentations of this condition and appreciation of the respective roles of characterising transferrin isoforms, leucocyte enzyme studies and genetic testing.

Clin Biochem Rev. 2008 November; 29(3): S135.

P16 Malnutrition Screening in a Tertiary Hospital Setting - Are We Doing Too Much or Too Little?

Introduction

Our hospital admission policy includes a nursing nutritional assessment tool (‘MUST’) which involves measurement of height and weight and answering several simple questions about recent unplanned weight loss or acute illness over the preceding 5 days. Admission screening also involves basic haematological and biochemical laboratory tests which may provide valuable clues to nutritional status. The objective of this audit was to assess the consistency of implementation of the MUST and the incidence of haematological and biochemical abnormalities that may indicate protein-calorie or micronutrient deficiencies.

Methods

Over a period of 8 weeks 100 new unselected admissions aged 18–100 yrs (mean 68 yrs) were enlisted in the audit. All procedures including the MUST, FBE and film morphology and biochemical tests for protein calorie malnutrition and micronutrient deficiencies were carried out within 48 hrs.

Results

The low cost MUST clinical assessment tool was seldom performed (2%) in spite of it being a ‘compulsory’ component of the nursing admission procedure. Haematological abnormalities were common with anaemia 50%, thrombocytopenia 12%, lymphocytopenia 33%. Blood film morphology demonstrated 17% polychromasia, 17% elongated cells, 7% stomatocytes, 6% macrocytes, 5% hypersegmented neutrophils, 5% crenated cells and 3% microcytes. Toxic granulation was seen in 19%. Only a single case of acanthocytes with occasional udder cells was seen. Screening for protein calorie malnutrition was positive in the majority of patients; hypoalbuminaemia 73% and low transthyretin in 55%. Most patients suffered micronutrient deficiencies; vitamin D 86% <75 μmol/L and 67% <50. Vitamin C deficiency was seen in 71%, zinc 32%, magnesium 19% but red cell folate was deficient in only 3%.

Conclusions

The incidence of malnutrition in tertiary hospitals is so high that many cases of micronutrient deficiency will probably be undetected if screening is not performed. The cost benefit analysis of screening is less clear and whether identification of defects leads to improved outcomes during the acute admission or long term correction, requires further study. The use of clinical screening tools such as MUST are poorly complied with and unlikely to add value to direct biochemical screening.

Clin Biochem Rev. 2008 November; 29(3): S135.

P17 Analytical Evaluation of the EZ CHEM (Nova Statsensor Creatinine Hospital Meter) Creatinine Point of Care Analyser

Introduction

The analyser is intended for quantitative measurement of creatinine in capillary, venous and arterial whole blood in the point-of-care setting but not recommended for neonates. The required sample volume is 1.2 μL and the analysis time is 30 seconds. The blood is drawn onto the strip by capillary action. The analyser also estimates (optional) GFR value using MDRD algorithm taking into account age, gender and race. The analyser is factory calibrated, can store 1200 results and 1000 user passwords, and can be configured for patient gender, age and race, Physician and Patient ID.

Methods

It is an amperometric sensor using a cascade of enzyme reactions. The evaluation included precision studies, patient comparison and limited interference testing. Precision studies were performed using fresh pooled lithium heparin whole blood samples at 3 different levels for consecutive analysis and 2 levels for between-day. The patient comparison was performed on capillary and venous samples (Greiner Lithium Heparin Tubes #456083) at the same time versus the Beckman DxC800 cup method. Interference studies were performed for the following compounds: bilirubin, haemolysis (HI), and creatine.

Results

Evaluation was performed using strip lot no 4107255129 (expiry September 2008). In total 25 male and 35 female volunteers were used with an age range 14–84 years with creatinine range (capillary samples on the EZ CHEM analyser) of 28-611 μmol/L. The precision data results are a) consecutive: L1 mean-67, SD-6.7, CV-10.1; L2 mean-121, SD-8.2, CV-6.8; L3 mean-257, SD-21.5, CV-8.4, and b) between day: L1 mean-128, DS-8.3, CV-6.5 and L3 mean-412, SD-23.4, CV-5.7. The Passing-Bablok patient correction details are: capillary y = 0.922x – 3.2; whole blood lithium heparin y = 0.810x + 6.1; and lithium heparin plasma y = 0.651 + 2.04. The manufacturer does not recommend plasma as a sample type. No interference was observed with HI>500 mg/dL, bilirubin up to 463 μmol/L or creatine up to 200 μmol/L.

Conclusion

Based on our experience during this study and the findings presented, the analyser is easy to use.

Clin Biochem Rev. 2008 November; 29(3): S135–S136.

P18 Evaluation of the Beckman Intrinsic Factor Antibody

Introduction

Intrinsic factor antibody (IFAb) prevents the binding of vitamin B12 to intrinsic factor, and is present in 50% of patients with pernicious anaemia. The Beckman assay enables laboratories to rapidly analyse IFAb as reflex in samples with low vitamin B12 results.

Methods

The IFAb assay is a competitive immunoenzymatic with ready to load reagent. Results are reported as negative equivocal or positive based on cut-off limits. The evaluation included precision studies, patient comparison, different sample comparison and haemolysis interference testing at 2 IFAb levels. Precision studies were performed using the Beckman IFAb controls. Serum (Greiner Serum Tube #456078) samples were compared with the Siemens Solid Phase IF bAb 57Co radioassay (#KISPZ). The cut-off limits for this method are recalculated with each run.

Results

The precision data is as follows: a) within-run L1 mean-1.05AU/mL, SD- 0.023, CV% 1.07; L2 mean-2.19, SD-0.052, CV-2.22 and b) between day L1 mean-1.04, SD-0.067, CV-6.53, and L2 mean-2.19, SD-0.077, CV-3.52. In total 135 serum samples at random were analysed for the patient comparison with vitamin B12 range of 29 to >1110 pmol/L. Only two discrepant results were found and investigated. In one, IgM paraprotein was possibly causing false positive with the Siemens method, and the second sample was query false negative on the Beckman method for which no clear reason could be found. Both serum and lithium heparin plasma are suitable sample types. No interference was observed with HI >500 mg/dL at the 2 IFAb levels.

Conclusions

The assay provides very comparable results by an automated means of analysing samples without the use of radioactive compounds. It also plus provides rapid results for prompt clinical decisions.

Clin Biochem Rev. 2008 November; 29(3): S136.

P19 A Comparison between IDMS Traceable and Non-IDMS Traceable Creatinine Slides on the Vitros Fusion 5,1 Analyser

Introduction

Standardisation of serum creatinine measurements is expected to improve the detection, diagnosis and treatment of chronic kidney disease by reducing interlaboratory bias. To achieve this aim Ortho-Clinical Diagnostics has released a creatinine calibration that is traceable to the internationally accepted reference method, isotope dilution mass spectrometry (IDMS).

Methods

Plasma samples were obtained from 123 patients (40 of whom were neonates) with creatinine values ranging between 9 and 605 μmol/L. In addition, 27 urine samples were obtained with creatinine values ranging between 1.8 and 24.2 mmol/L. These samples were analysed initially using the non-IDMS traceable creatinine slides and subsequently on the IDMS-traceable creatinine slides using a Vitros Fusion 5,1 analyser.

Results

Regression analysis of the plasma comparison showed an overall correlation of R2=1.00 with a slope of 0.89. Whilst regression analysis of the urine comparison showed an overall correlation of R2 = 1.00 with a slope of 0.97.

Conclusions

The IDMS-traceable creatinine slides were shown to reflect the manufacturer’s performance specifications compared with the non-IDMS traceable slides. However the difference in plasma creatinine measurements between the non-IDMS and IDMS traceable slides is significant. This is especially relevant at low creatinine levels, such as those seen in neonatal samples. Based on these findings plasma reference intervals will need to be reviewed, not only for creatinine itself but for calculations involving creatinine such as the Cockcroft-Gault and Schwartz creatinine clearance calculations. The difference in urinary creatinine measurements was not significant and as such urinary reference intervals need not be reviewed or altered based on the findings of this comparison.

Clin Biochem Rev. 2008 November; 29(3): S136.

P20 Evaluation of the Cyclosporine Assay on the Abbott Architect CI8200® Analyser

Introduction

Cyclosporine is a commonly used immunosuppressive drug. Monitoring of cyclosporine levels is important due to the tight therapeutic range and variable bioavailability. We evaluated the performance of the new Architect Cyclosporine chemiluminescent microparticle immunoassay on the Abbott Architect ci8200 and compared it to the Cyclosporine MEIA assay on the Abbott AxSYM and the Microgenics CEDIA Cyclosporine PLUS assay on the Architect ci8200.

Methods

Whole blood EDTA samples (n = 101) representing a wide range of cyclosporine concentrations (10–2686 μg/L) were analysed on both the Abbott AXSYM and the Abbott Architect ci8200 analysers. Specimens were from patients with a variety of transplants including lung (n = 46), kidney (n = 6), heart (n = 26) and bone marrow (n = 22). We evaluated the intra- and inter-assay precision, and linearity of the cyclosporine assay on the Abbott Architect ci8200. Results were compared with the established cyclosporine MEIA on the AxSYM and the CEDIA Microgenics on the Architect ci8200.

Results

The analytical evaluation of cyclosporine on the Abbott Architect ci8200 showed good precision, with intra-assay CVs of 9.0%, 6.5%, 4.1% and 6.8% for cyclosporine values of 82, 180, 414 and 925 μg/L respectively. Inter-assay CVs were 9.4%, 8.9%, 8.8% and 3.1% for values of 80, 333, 923, 1413 μg/ L. The assay showed acceptable linearity between 150 and 1500 μg/L. The Architect results correlated well with the AxSYM (n = 100, Passing Bablok correlation y = 0.9832x – 3.8918), Microgenics (n = 98, Passing Bablok correlation y = 1.0277x – 5.5201) and external quality assurance LC-MS specimens. Transplant type showed little effect on the correlation.

Conclusions

These results demonstrate that the Abbott Architect Cyclosporine assay is a precise, reproducible method for determining cyclosporine concentrations in whole blood specimens.

Clin Biochem Rev. 2008 November; 29(3): S136.

P21 Evaluation of the ADVIA Centaur® Cyclosporin Immunoassay Compared with Tandem Mass Spectrometry (LCMSMS)

Introduction

Cyclosporin (CsA) immunoassays are commonly used in routine laboratories but may be affected by metabolite cross reactivity. The introduction of 2 hour monitoring (C2) has meant that two calibration curves are often required to cover the concentration range. We evaluated the ADVIA Centaur® CsA immunoassay that uses a single calibration curve (20–2000 μg/L).

Methods

CsA concentrations were measured using the ADVIA Centaur® and LCMSMS in samples collected from heart (HTX, n = 25), lung (LTX, n = 26) and bone marrow (BMT, n=14) transplant patients. Imprecision was determined using Biorad Whole Blood and Elevated Immunosuppressant controls. David Holt QAP samples with weighed-in values were used to determine accuracy. Linearity was assessed with dilution in CsA-free whole blood and Centaur Multi Diluent 12. Functional sensitivity was determined by measuring a patient sample (37 μg/L) over 10 runs.

Results

Between-run precision was 5.6% at 100 μg/L, 3.0% at 209 μg/L, 4.5% at 828 μg/L and 11% at 1635 μg/L. Correlation data for all transplant types was good (y = 1.01x – 7, n = 65, r2 = 0.97) however Centaur results were generally higher than LCMSMS at values below 300 μg/L (y= 1.09x + 6, n = 28, r2 = 0.95). On average Centaur results for BMT were 8% higher than LCMSMS (range: −12 to +31%), 26% higher for HTX (−7.5 to +47%) and 1.5% lower in LTX (−26 to +36%). Centaur results showed no bias compared to spiked QAP samples with average bias 0% (range: −10 to +15%, n = 7). Metabolite crossreactivity may contribute to over-reading in some patient samples. Dilution studies were acceptable for both diluents and precision at 37 μg/L is 17.5% CV.

Conclusions

The precision, accuracy and correlation data were acceptable. The single calibration curve improves workflow and reduces assay costs.

Clin Biochem Rev. 2008 November; 29(3): S136–S137.

P22 Six Month Laboratory Experience with Unity Real Time (URTTM) – New Features Improving QC Monitoring and Reporting

Introduction

Unity Real Time (URTTM) is a Bio-Rad Quality Control (QC) data management software package. It is a tool designed to meet current and future clinical laboratory quality requirements for data access, analysis, review, management and storage where all these functions assist in reporting reliable results. Features include new chart options, bench and supervisor reviews, audit trails and worldwide report peer review.

Implementation

The implementation was a seamless transition from QC OncallTM to URTTM. Decisions had to be made about the connectivity options where SoftConnectTM was the first choice however DirectConnectTM proved to be the better choice for our laboratory.

Features/Benefits

The following features have improved QC monitoring and reporting: Levey- Jennings options where our laboratory QC chart may be graphed against the world wide peer group data allowing real time monitoring of the laboratory QC with international QC data. Colour has facilitated QC review. Direct access to comments or actions at the time of review enhances troubleshooting. Automated data download and bench review access ensures QC review by operator. The supervisor review may be performed remotely giving the supervisor a snapshot of the QCs and allowing opportunity for the supervisor to action or comment on the QC performance. Audit trails record all parameter changes in the product. Electronic archiving of data and comments reduces printing requirements. The use a single QC product streamlines training and documentation in the laboratory. Security of laboratory number permission and administrator permission ensures correct work practice.

Conclusions

URTTM has improved QC monitoring and reporting however we plan to learn and utilise more of the features.

Clin Biochem Rev. 2008 November; 29(3): S137.

P23 Evaluation of a Monoclonal Faecal Elastase Assay for the Diagnosis and Monitoring of Paediatric Patients by Comparison with Traditional Enzyme Markers

Introduction

Exocrine pancreatic insufficiency results in a decrease in secretion of pancreatic enzymes that aid digestion. Faecal elastase measurement has been appraised as the “new gold standard” for non-invasive pancreatic function testing that can replace most traditional non-invasive tests. The aim of this study was to assess the clinical utility of faecal elastase measurement compared to our traditional enzyme markers, tryptic activity and chymotrypsin, for the diagnosis and monitoring of patients with pancreatic insufficiency.

Methods

The monoclonal ELISA Pancreatic Elastase-1 Stool Test (ScheBo®Advantages) was performed on patients with suspected or confirmed pancreatic insufficiency. Intra-assay imprecision (n = 10) was 14.1% and 6.9% for deficient and sufficient patients. Inter-assay (n = 10) was demonstrated to be 5% at 200 μg/g. Faecal elastase measurements were compared to the traditional in-house tryptic activity and/or chymotrypsin assays.

Results

Of the 49 faecal elastase measurements performed, 39 patients were replete in elastase, one had borderline deficiency and 9 were classified as deficient. Of the 12 patients who had chymotrypsin analysis performed, ten results were concordant with the elastase result and two results (17%) were discordant. Of the 48 patients who had tryptic activity analysis performed, 34 results were concordant with elastase and 14 results (29%) were discordant. Investigation of the discordant results identified 10 patients with degradation of the trypsin enzyme, three patients on porcine pancreatic enzyme replacement therapy, one liquid faecal sample, and two unidentified causes.

Conclusions

The elastase assay demonstrates good imprecision. It cannot be used accurately on liquid samples nor to monitor replacement therapy. Tryptic activity can be used to monitor pancreatic enzyme replacement therapy. We recommend tryptic activity as a first line test with a follow up of elastase measurement when clinically indicated. The cumbersome and imprecise chymotrypsin assay will be discontinued.

Clin Biochem Rev. 2008 November; 29(3): S137.

P24 Determination of Serum Copper by Zeeman Graphite Furnace Atomic Absorption Spectrometry

Introduction

The Zeeman graphite furnace atomic absorption spectrometer (GFAAS) was adopted to replace the previous colorimetric analysis, with the requirement that the analysis should be capable of producing quality results from samples as small as 20 μL, and with copper levels down to below 1 μmol/L. In the course of developing the method, little information could be found in the literature since most labs appear to use flame AAS.

Methods

The copper method was established on a Varian AA240Z Zeeman GFAAS with a Programmable Sample Dispenser accessory and utilising a Varian UltrAA lamp. 100 μL of each standard, control and patient sample was diluted with 100 μL of 0.37 M trichloroacetic acid, vortexed and centrifuged prior to analysis. The sample supernatants were then diluted 1 in 12 in a solution of 0.1% nitric acid and 1% Triton X-100, and then analysed with atomisation at 2300°C in the coated graphite tube.

Results

The method was linear in the studied concentration range of 0.1 to 40 μmol/L The inter-assay CV was 3.5% at 16 μmol/L (n = 20). Results obtained for quality assurance samples have demonstrated substantial improvement in performance compared to the previous colorimetric method.

Conclusions

This study demonstrated that Zeeman GFAAS can be effectively used for the analysis of copper in very small paediatric serum samples. The method is linear and precise and does not suffer from the false positive results occasionally observed in the colorimetric analysis of copper. The initial protein precipitation step can be scaled down to as low as 20 μL of serum without loss of result quality.

Clin Biochem Rev. 2008 November; 29(3): S137–S138.

P25 Significance of C-Reactive Protein in the Diagnosis of Neonatal Infection

Introduction

The diagnosis of neonatal bacterial infection remains one of the greatest and most tantalising challenges to neonatologists. Empirical treatment should not be delayed because of the high mortality. C-reactive protein (CRP), an acute phase protein, increases in inflammatory disorders and tissue injury is consider as a useful test in diagnosis neonatal infections. The study was conducted at the Saint Paul Hospital between January 2006, and April 2008 to determine whether, in the presence of proved or presumed bacterial infection, the sensitivity of serum C-reactive protein (CRP) response could be enhanced by serial rather than single determinations in a developing country.

Methods

The CRP responses of 117 suspected infection neonates were assessed. CRP levels were measured initially and again after (immunoturbidimetry). Assessments also included a blood culture, complete blood cell count, and chest radiograph and culture of spinal fluid when appropriate. CRP responses were correlated with: (1) positive blood or cerebrospinal fluid cultures; (2) negative blood culture-definite infection (necrotising enterocolitis, pneumonia, subcutaneous abscess); (3) negative blood culture- possible non-infection.

Results

Blood cultures had been done in only 14.3% cases, and CSF culture in 9.4% cases due to difficulty in facilities and techniques in a developing country. 17.7% blood cultures were positive (Klebsiella pneumoniae, Enterobacte cloacae, streptococcus). CRP levels were elevated in 100% of the positive blood culture group; in the negative blood culture-definite infection group, CRP levels were abnormal in 91.7%. In the negative culture- possible non-infection group, CRP was not elevated, suggesting high specificity; Serial determinations of CRP resulted in enhanced sensitivity in the positive blood culture group. Initial determinations by themselves were inadequate sensitive. However, single CRP level determination and inadequate time of sampling was popular in this study.

Conclusions

In the developing country like Viet Nam, where blood cultures are difficult to realise, the determination of CRP - a simple, fast test could be useful in detecting neonate infections. The data also indicate the necessity for serial determinations of CRP at adequate time for optimal sensitivity.

Clin Biochem Rev. 2008 November; 29(3): S138.

P26 Assessment of Minicap Capillary Electrophoresis System

Introduction

We examined the Minicap Capillary Electrophoresis System from Dade Behring for three weeks. The Minicap has two capillaries, a bar code reader, can handle 22 serum protein electrophoresis samples per hour and was easy to operate. As the Minicap is regulated to operate at 37 oC, it is able to run specimens containing cryoglobulins without damaging the capillary. The Minicap is not able to run urine protein electrophoresis or provide immunofixation of serum samples at this time.

Methods

The method used was the proprietary Minicap method using Minicap buffer. The QC used was serum of a patient with a monoclonal band of 24 g/L. We chose 135 specimens, 70 were from patients with monoclonal bands in their serum and 20 samples with small abnormalities in the gamma region. These included specimens with oligoclonal IgG, free light chain bands, and small monoclonal bands <3 g/L. We compared the size of the monoclonal bands (from 1 to 91 g/L) to the sizes of the monoclonal bands on our usual Beckman MDQ Capillary Electrophoresis instrument.

Results

From the Minicap electropherograms, we identified all 20 samples which had abnormalities in the gamma region using the Beckman MDQ. These samples then required immunofixation to identify the abnormality. Results for the QC ranged between 24.0 g/L and 24.8 g/L; n = 9, (mean ± SD) 24.3 ± 0.23 g/L. Comparing the size of the 70 monoclonal bands on the Minicap to the Beckman MDQ, the correlation was R2 = 0.98, with y = 1.03X + 0.02.

Conclusions

The Minicap is a robust two capillary system which is suitable for analysing samples for serum protein electrophoresis in a medium sized protein laboratory. The usefulness of the system could be improved if it analysed urines for urine protein electrophoresis.

Clin Biochem Rev. 2008 November; 29(3): S138.

P27 Critical Differences: Are Doctors Estimates better than Laboratory Calculations?

Introduction

The Critical Difference (CD) is the smallest difference between two results from the same patient which is likely to be real, i.e. not caused by random changes due to analytical or biological variation. While doctors rarely have knowledge of CD calculations they frequently make decisions on the differences in laboratory results. This study compares CD values derived from laboratory studies with clinician estimates.

Methods

Staff attending Medical Grand Rounds were provided with scenarios seeking their estimates of CD for sodium, haemoglobin, calcium and ALT. Responses were recorded with responder keypads and compared with calculated CD values.

Results

Results were obtained from 16 staff specialists, 13 registrars, 15 medical students and 9 non-medical personnel. For serum sodium and whole blood haemoglobin the responses of the doctors were very closely aligned with theoretical values with over 75% estimating changes within ±1 mmol/L and ±4 g/L respectively. By contrast for serum ALT and total calcium, where the changes were increases away from normal, the estimate CDs were much smaller than predicted by theory.

Discussion

Either medical staff were unaware of the possible variability of some results or current CD theory is inadequate. The scenario formats included starting concentrations as well as incremental changes. Doctors may have considered the meaning of the absolute values as well as the change in values. The absolute values have been shown to affect the CD1 and when this factor is included in the CD calculation doctors estimates aligned more closely with predicted values.

Conclusions

For some analytes doctors showed good knowledge of predicted CDs. For other analytes CDs were underestimated however this may reflect limitations in current theory rather medical misinterpretation.

References

1. Jones GRD. Clin Biochem Rev. 2007;28:S15.
Clin Biochem Rev. 2008 November; 29(3): S138.

P28 Evaluation of the Accu-Chek® Performa, Medisense Optium® and Medisense Optium® Point-of-Care Blood Glucose Meters Against a Laboratory Comparative Method

Introduction

Following recent IFCC recommendations blood glucose meters should report results in plasma equivalent units, instead of whole blood capillary levels. We evaluated the accuracy of the recalibration of the Accu-Chek® Performa strip, one of three blood glucose strips currently available to Australian hospitals that according to the manufacturer meet this new recommendation.

Methods

We analysed 121 lithium heparinised whole blood samples collected from inpatients in our hospital. Each of the three strips, Accu-Chek® Performa (strip lot: 320040), Medisense Optium® 5 second (strip lot: 4230) and Medisense Optium® point of care 20 second (strip lot: 53684) were sampled in random order. Following blood glucose strip sampling, aliquots were centrifuged within 15 minutes to obtain plasma used to measure glucose concentration by our routine laboratory glucose method (Roche Modular).

Results

Overall bias from the laboratory results was calculated using a Bland & Altman analysis. A bias of −0.30 (95% CI: −0.37 to −0.24) mmol/L for the Performa, 0.44 (95% CI: 0.23 to 0.64) mmol/L for the Optium 20 sec and −0.71 (95% CI: −0.89 to −0.52) for the Optium 5 sec was measured. Passing and Bablock regression analysis for agreement of each glucose meter versus the laboratory glucose method yielded results as: Accu-chek Performa y = −0.043 + 0.95x, r2 = 0.99; Medisense Optium (20 sec) y = 1.81 + 0.73x, r2 = 0.90; and the Medisense Optium (5 sec) was y = 0.57 + 0.76x, r2 = 0.93.

Compliance to the ISO 15197.2003 (agreement to within 5% of the reference value using hypo and hyperglycaemic criteria) showed the Accu-Chek Performa had 98.3% compliance with 1.7% outside acceptable limit. For the Medisense Optium meter 20 sec strip 83.5% compliance 16.5% outside, and the 5 sec strip 79.8% compliance 20.2% outside.

Conclusions

Results indicate the Accu-Chek® Performa compared closest to the laboratory method of the three types of blood glucose strips tested. The observed differences between the two Medisense Optium glucose strips and the difference versus the comparative laboratory method were unable to be explained.

Clin Biochem Rev. 2008 November; 29(3): S139.

P29 QAP Median Values: Majority Rules, but the Majority Isn’t Always Correct

Introduction

In the first cycle of the 2008 RCPA-AACB QAP General Serum Chemistry and Special Drugs Program, the reported Vancomycin values varied by more than three fold at the midpoint and upper end of the concentration range, with the absolute differences up to eleven times the allowable limit of error. Method group differences predominantly accounted for this disparity. Significant negative bias compared to QAP median values prompted our method comparison study with Vancomycin spiked pool sera.

Methods

Pooled sera obtained from community patients with no clinical history of Vancomycin administration or recent infection was confirmed as Vancomycin free by analysis on the Siemens Dimension RxL Max. Different concentrations of Vancomycin were prepared from 1 g vial Vancomycin (Mayne Pharma Pty Ltd) by serial dilutions with the drug free sera. Aliquots were analysed on the Siemens Dimension RxL Max and Roche Modular. Analyses were undertaken on the same day by each method, within 24 hours of sample preparation.

Results

The Dimension RxL method, which had an apparent proportional negative bias of 20% for QAP samples demonstrated <1.0% positive bias for spiked sera. The Modular method, which had no significant bias for QAP samples demonstrated 20% positive bias for spiked sera.

Conclusions

Similar differences between method groups were observed for QAP materials and spiked pooled sera. Whereas the QAP program evaluated the Dimension RXL method with negative bias, it is more likely that the median lab had significant positive bias.

The RCPA-AACB QAP program provides external proficiency testing for the majority of Australian laboratories. Assessment of laboratory performance by this QAP program is important for NATA accreditation. Therefore, it is appropriate to provide therapeutic drug target values instead of median laboratory values for the General Serum Chemistry and Special Drugs Programs.

Clin Biochem Rev. 2008 November; 29(3): S139.

P30 Stability Studies of Fifty Five Common Analytes in Serum and Plasma

Introduction

Spurious test results have been known to be caused by inappropriate storage conditions. We assessed the stability of 55 common analytes in serum and plasma stored at 4°C over a 6 day period.

Methods

Fasting blood samples were collected into 7 × 4.5 mL sample tubes containing lithium heparin with a gel separator, 7 × 4.5 mL tubes containing lithium heparin without a gel separator and 7 × 9.5 mL tubes containing no additives but with a gel separator. All samples were centrifuged and one from each group analysed within 30 minutes. The other samples were stored at 4°C. Samples were aliquotted after 24, 48, 72, 96 120 and 148 hours, re-centrifuged and analysed. All analyses were performed on the Abbott Architect Ci8200. Clinically significant changes were assessed using the significant change limit (SCL).

Results

No result exceeded the SCL for the lithium heparin plasma with gel separator. Potassium and ferritin results exceeded this limit at day 3 and 4 respectively with the serum samples. For the plasma sample containing the gel separator, LD, glucose, phosphate, potassium and ferritin exceeded the SCL after days 1, 2, 2, 3 and 5 respectively.

Conclusions

Many analytes show stability in both serum and plasma when stored at 4°C over a long period of time. Knowledge of that stability is required if “add on” requests are to be undertaken.

Clin Biochem Rev. 2008 November; 29(3): S139.

P31 The Effect of Time and Storage Conditions on the Analysis of 49 Commonly Performed Analytes

Introduction

A common cause of erroneous results is prolonged contact of the plasma with red blood cells. We assessed the effect of time and storage condition on 49 commonly performed analytes over a 24 hour period.

Methods

Non-fasting blood samples were collected into 16 × 4.5 mL sample tubes containing lithium heparin. The samples were divided into 3 groups, the first group containing six samples were centrifuged with one sample analysed within 30 minutes. The other 5 samples were stored at room temperature (RT). The other two groups each containing 5 samples remained un-centrifuged with one group stored at RT the other at 4°C. Samples from the un-centrifuged group were centrifuged at 2, 4, 8, 16 and 24 hours post collection and analysed in conjunction with a sample from the centrifuged group. All analyses were performed on the Abbott Architect Ci8200. Clinically significant changes were assessed using the significant change limit (SCL)

Results

Potassium and LD at 4 hours exceeded the SCL for the un-centrifuged samples stored at 4°C and LD at 24 hours exceeded the SCL for the centrifuged samples stored at RT. No result exceeded the SCL for the un-centrifuged sample stored at RT.

Conclusions

Many analytes show stability over a 24 hour period when collected into lithium heparin and do best when stored at room temperature in an un-centrifuged state. Knowledge of the stability of these analytes is of use to regional centres that refer pathology samples to central locations.

Clin Biochem Rev. 2008 November; 29(3): S139–S140.

P32 The Effect of Sample Type, Time and Storage Condition on the Estimation of Glucose

Introduction

Prolonged contact of the plasma with red blood cells is known to cause spurious results when measuring glucose concentration. We assessed the effect of sample type, time and storage condition and glucose analysis over 24 hours.

Methods

Non-fasting samples were collected into 18 × 4.5 mL tubes containing lithium heparin (LiH) and 12 containing fluoride oxalate. The LiH samples were divided into 3 groups, the first containing 6 samples were centrifuged with one sample analysed within 30 minutes. The other 5 samples were stored at room temperature (RT). The other two groups each containing 5 samples remained un-centrifuged with one group stored at RT the other at 4°C. The fluoride oxalate samples were divided into 2 groups, the first containing 6 samples were centrifuged with one sample analysed within 30 minutes. The other 5 samples were stored at RT. The other group containing 5 samples remained un-centrifuged and was stored at RT. Samples from the un-centrifuged group were centrifuged at 2, 4, 8, 16 and 24 hours post collection and analysed in conjunction with a sample from the centrifuged group. All analyses were performed on the Abbott Architect Ci8200. Clinically significant changes were assessed using the significant change limit (SCL).

Results

No significant changes were observed in the fluoride oxalate groups over the 24 hour period. Significant changes were seen after 2 hours in the non centrifuged LiH group stored at RT and after 8 hours in the other two LiH groups.

Conclusions

Where there is a delay in the transportation of samples for glucose estimation of more than 2 hours then samples should be collected into tubes containing fluoride oxalate.

Clin Biochem Rev. 2008 November; 29(3): S140.

P33 Comparison of 3 Bio-Rad Variant Columns for the Determination Of HBA1C

Introduction

Bio-Rad (Hercules CA) recently introduced a new cation exchange resin for the Variant II instrument, providing improved separation for labile and carbamylated fractions and better reporting of the variant haemoglobins. We have compared the results from patients and NGSP standards with the total Glycated Hb kits and the old and new resin kits.

Methods

The Variant II Haemoglobin A1c or the total Glycated Haemoglobin Instruments were used for all samples. The reagent kits and columns were used according to manufacturers’ instructions.

Results

The results obtained with the new resin were all highly correlated.(see Table).

MethodNCorrellation coef cient, Slope and intercept
Old resin370.97, 1.13, −1.24
Boronate af nity Resin260.98, 1.24, −0.97
NGSP vs new resin300.98, 0.99,0.15
NGSP vs Boronate af nity300.91, 0.85, 0.78
NGSP vs old resin100.99, 1.0, 0.32

Bias plots between the old and new resins were closer but there were significant differences between the values for the boronate affinity resins (See Figure). In addition the differences between the NGSP standards and the new resin were less variable than the values from the boronate affinity resin.

An external file that holds a picture, illustration, etc.
Object name is CBR29_S2_pgS115_G2.jpg

Conclusions

The new resin is a significant improvement because it does not suffer from the carbamylated, labile and HbF problems seen with the old resin and would not significantly change patient results from previous studies. Curiously boronate chromatography which also does not suffer from these same interferences showed greater variability both with the patient results and when NGSP standards were analysed.

Clin Biochem Rev. 2008 November; 29(3): S140.

P34 Bilirubin Measurement in CSF on the Abbott Architect

Introduction

With the gold standard of bilirubin measurement in cerebrospinal fluid (CSF), scanning spectrophotometry, unavailable due to the specialised equipment and training it requires, it was necessary to develop an automated method to support the Neurosurgery Unit protocol for subarachnoid haemorrhage (SAH) diagnosis at the Royal Hobart Hospital.

Methods

The assay was created by altering the total bilirubin assay on the Abbott Architect. The serum method configuration was modified to achieve greater sensitivity by increasing the sample volume, delaying the last read time and calibrating the assay using onboard dilutions of the low calibrator. Internal validation involved a linear regression of expected and measured values of diluted QC material and imprecision studies. A comparison between the reported median values and measured results was performed using external quality assurance material for external validation. The effect of haemolysis was determined by spiking pooled CSF with various volumes of a prepared haemolysate. Non-SAH patient samples were analysed to determine normal CSF bilirubin concentrations.

Results

The linear regression and external comparison yielded R2 values of 0.9999 and 0.9638, respectively. The lower limit of quantitation (LOQ) was determined to be 0.2 μmol/L and the assay linear range 0.2–5.8 μmol/L. The %CV was found to be 7.5% at concentrations of 0.4 μmol/L. Due to the negative interference caused by haemolysis, samples with haemoglobin concentrations >0.30 g/L (haemolysis index of 1+) are considered unsuitable for analysis. The analysis of patients without suspected SAH established a reference range of <0.4 μmol/L, and thus a cut-off of 0.4 μmol/L.

Conclusions

The changes implemented to produce this assay proved successful. The method was found to have acceptable imprecision and accuracy, a dynamic range relevant to the clinical application and workable parameters for routine use.

Clin Biochem Rev. 2008 November; 29(3): S140–S141.

P35 Lead Contamination Signalled by Bird Deaths in Esperance WA

Introduction

Three months after the residents of Esperance reported the unexplained deaths of some 5000 birds, lead toxicity was implicated and exports of lead carbonate were suspended. The Health Department opened a community clinic offering free determination of blood lead levels for concerned residents.

Methods

EDTA blood was collected from 2706 residents (21% of the population) and lead analysed by furnace atomic absorption spectrometry. The current level of concern is a whole blood lead of 10 μg/dL, although the intelligence quotient of children begins to decline at lead levels below this level.

Results

Pre-school children aged under five years comprised 318 of those sampled: 2.5% (eight children) had a blood lead levels ≥10 μg/dL and 26% (83 children) had levels of ≥5 μg/dL. Of comparable children resident in the port city of Fremantle, none were ≥10 μg/dL and 4% were ≥5 μg/dL. Of adults aged 20 years and over, 2.0% (37 adults) had blood lead levels ≥10 μg/dL and 14.4% (265 adults) had levels ≥5 μg/dL. Mean blood lead levels varied with age; children under five years and adults over 60 years had the highest means of 3.5 μg/dL and 3.6 μg/dL respectively, the 10–19 year age group recorded the lowest mean of 1.8 μg/dL.

Conclusions

The residents of Esperance were exposed to lead dust, an unforeseen hazard of port handling procedures. Exposure of the respiratory tract to lead dust results in the absorption of up to 30–40% of inhaled lead. Blood lead levels exceeded the level of concern (≥10 μg/dL) in eight pre-school children and 37 adults.

Clin Biochem Rev. 2008 November; 29(3): S141.

P36 Plasma Homocysteine Measurement with an Automated Enzymatic Cycling Assay

Introduction

The amino acid and vascular disease risk factor homocysteine is measured routinely by immunoassay or by HPLC with fluorescence or mass spectrometric detection. The aim of this study was to evaluate a new enzymatic assay for homocysteine, which is cycled between cystathionine synthase and lyase to generate pyruvate. In the presence of lactate dehydrogenase, the rate of conversion of NADH to NAD is proportional to the homocysteine concentration.

Methods

Liquid-stable reagents with recombinant enzymes were manufactured by Catch Inc., Bothell, WA, USA and purchased in kit form. Assay parameters were provided for the Roche Modular P and included a 2-point calibration and plasma volume of 16 μL. 63 routine plasma samples covering the range 4–60 μmol/L were assayed in duplicate and results compared to a tandem mass spectrometric (MSMS) assay.

Results

The enzymatic assay was specific for L-homocysteine and was linear over the concentration range 5–50 μmol/L. Between-day imprecision (n = 9) of two commercial QC materials at low (14.6 μmol/L) and high (37.6 μmol/L) concentrations was 4.5% and 1.6% respectively, with a CV of 1.9% for assay of patient duplicates. Passing-Bablok regression of the enzymatic assay (y) against the MSMS assay (x) gave y = 1.09x – 1.2, with a correlation coefficient of 0.96. Bland-Altman analysis showed a mean bias of 0.3 μmol/L between the two assays, which both used calibrators traceable to SRM 1955. Two outliers were detected.

Conclusions

The automated enzymatic assay for plasma homocysteine gives comparable results to tandem mass spectrometry, and is quick and convenient. This assay could be a useful alternative method to immunoassay or chromatography, depending on reagent costs.

Clin Biochem Rev. 2008 November; 29(3): S141.

P39 A Survey of Therapeutic Drug Monitoring in Australia and New Zealand – Preliminary Results

Introduction

Measurements of therapeutic drug concentrations are performed in many laboratories, however the provision of a complete Therapeutic Drug Monitoring (TDM) service requires additional factors such as pre- and post analytical advice. It is unknown to what extent Australian and New Zealand laboratories provide these services.

Methods

A survey of therapeutic drug monitoring (TDM) practices in Australia and New Zealand was undertaken in late 2007 with 57 responses received. Of these, 42% described their institution as a hospital, 11% as a hospital pathology provider and 47% as a pathology provider. Areas surveyed included assays provided, dosing regimen advice provided and monitoring advice provided.

Results

The responses indicated good support for therapeutic drug measurements in terms of the assays provided. However the level of supporting advice with respect to both dosing and monitoring was less frequent. For example, with aminoglycosides, less than 50% of those respondents who provide advice on dosing and monitoring, use pharmacokinetic principles to provide individualised advice (tobramycin: 42%, gentamicin: 28% and vancomycin: 24%).

Conclusions

The provision of high quality TDM services requires more than issuing drug assay results. The survey has highlighted the need to define the appropriate level of practice in this area.

Clin Biochem Rev. 2008 November; 29(3): S141.

P40 Ionised Calcium – Validation of Reference Interval

Introduction

Many laboratory reference intervals, like QML’s ionised calcium (IC) range, were determined decades ago using instrumentation and data analysis tools of the time. This study aims to validate the incumbent range in light of new instrumentation, improved data analysis and in the context of changing patient demographics.

Methods/Equipment

  • IL GEM Premier 3000 analyser, iQM enabled reagent cartridges
  • Analyse-It Version 1.69 (2003), Analyse-It Software Ltd.

All ionised calcium analyses were performed on anaerobically sealed serum (SST) samples (during 2006 and 2007). Collections followed standard protocols and analysis was performed using a GEM Premier 3000 analyser. Quality of results was monitored by the instrument’s on board iQM software and verified by other findings. A data extract producing 21,137 ionised calcium results was undertaken in April 2008, with other parameters used to identify normals as well as multiple patient episodes which could potentially bias subsequent calculations. The final data set was used to calculate both parametric and non-parametric reference intervals using Analyse-It software.

Results

QML’s current IC reference interval of 1.20–1.30 mmol/L (pH 7.4), although somewhat higher than quoted by “Tietz”, is comparable to that used by many Australian laboratories. To confirm this range, abnormals and duplicate patients were firstly deleted from the database. Since the distribution of the remaining 534 patients was essentially Gaussian, parametric and non-parametric range determinations were expected to produce similar values. This was confirmed with values of 1.18–1.31 and 1.19–1.32 mmol/L respectively.

Conclusions

The current IC range was essentially confirmed although both parametric and non-parametric calculations implied that it may be a little tight. To avoid the potential disruption of a formal change to the reported range, the “extended range” is factored into relevant commenting protocols.

Clin Biochem Rev. 2008 November; 29(3): S141–S142.

P41 Are Routine Serum Albumin Assays Suitable for Haemodialysis Patients?

Introduction

Serum albumin is an important marker of life expectancy for patients on haemodialysis. There is conflicting information on how to measure albumin in these patients. The National Kidney Foundation (USA) recommends using BCG because a uraemic toxin (CMPF, 3-carboxy-4-methyl-5-propyl-2-furanpropionic acid) can cause falsely low results using BCP (competition for albumin binding sites). BCG, however, can give falsely high results due to its reaction with certain globulins. Because nephrologists have concerns on this issue, which can confuse audits between dialysis centres, we investigated the magnitude of this problem.

Methods

Pre and post dialysis samples of serum (S) and plasma (P) (n=88) were assayed on Roche Modular using standard Hitachi assays (BCP, BCG, immunoturbidimetry). IFCC protein standard (CRM 470) and albumin standards (Sigma cat.no. A-9511, 25, 37.5 and 50 g/L) were used to ensure traceability. All samples (both S+P) were assayed in duplicate by each method.

Results

Mean results (SD) are summarised below using immunoturbidimetry (IT) as a “gold standard”.

TP (g/L)BCP (g/L)

DialysisSPSP
PRE69.9 (5.1)73.8 (5.0)35.6 (3.6)35.8 (3.5)
POST75.4 (6.8)79.4 (6.9)37.8 (4.2)37.9 (4.2)

IT (g/L)BCG (g/L)

DialysisSPSP
PRE36.1 (3.7)35.6 (3.7)39.8 (3.3)39.0 (3.3)
POST38.8 (4.3)38.1 (4.3)42.6 (3.9)41.8 (3.9)

Conclusions

Albumin (but not TP) results were similar for serum and plasma. For most patients, BCP measures albumin more accurately than BCG, which reads 4.0 g/L and 3.7 g/L higher than BCP and the “reference” method (on average). Five patients (5.7%) had unique results (BCP 3-> 6 g/L lower than the “reference” method and 20% lower than BCG). We conclude that both dye-binding methods have limitations for measuring serum albumin in haemodialysis patients. By using a combination of BCP and BCG, it may be possible to identify patients where CMPF occupies albumin binding sites. Higher CMPF concentrations have been reported in haemodialysis rather than CAPD patients, and are associated with anaemia and neurological impairment.

Clin Biochem Rev. 2008 November; 29(3): S142.

P42 Assessment of the Effect of Macroprolactin on 6 Different Prolactin Assays Using a Sample Containing Predominantly Big Big Prolactin

Introduction

Macroprolactin can lead to unnecessary investigations, misinterpretation of results and misdiagnosis of patients with symptoms of prolactin excess. Prolactin is a hormone secreted by the anterior pituitary and though its main action is in the control of lactation, serum levels are useful in the investigation of hypogonadism, fertility and sexual function. Prolactin is found in serum in a number of forms. Biologically active prolactin is secreted in the monomeric form, molecular mass 23 kDa. Big prolactin, which may have some biological activity, is a complex, possibly a dimer or associated with a receptor, has a molecular mass 50–60 kDa. Big big prolactin, which is considered biologically inactive, is complexed with immunoglobulin G and has a molecular mass of 150–170 kDa. Commercial immunoassays for prolactin differ in reactivity with the macromolecular complex and the purpose of this study was to show what differences might be expected from the most common methods available on mainstream analysers in use today.

Methods

A serum sample containing predominantly big big prolactin collected from a single patient previously investigated for hyperprolactinaemia was divided and sent to a number of laboratories chosen to represent a wide cross-section of methods. All of the participating laboratories used a PEG precipitation protocol. The analysers in the survey comprised an Abbott Architect i2000 SR, Beckman DXi, Roche E170, Bayer Centaur, Vitros ECi and an Immulite 2000.

Results

The sample distributed was from a male and was elevated in all but the Bayer Centaur assay. The various assays returned prolactin results between 299 and 1073 U/L. All laboratories reported a significant decrease in recovery following PEG precipitation with all but one laboratory using a specific cutoff value. An audit of all the prolactins run in our laboratory between 1st January 2008 and the 21st May 2008 shows 184 were measured, of these 67 were from males and 117 from females. Of these 107 were elevated with 8 macroprolactin screens requested over this period, one on a sample with a normal level of prolactin. Of the eight screens, 4 were negative, 2 were equivocal and 2 detected the presence of macroprolactin, one of these was the sample with the normal level of prolactin. Although the numbers are small the figures are consistent with literature reports that 23–25% of elevated prolactins are due to macroprolactin.

Conclusions

The data show that all methods react with macroprolactin and that the level is no guide to the presence of macroprolactin. Until this assay improves we should run a macroprolactin screen on every prolactin request.

Clin Biochem Rev. 2008 November; 29(3): S142.

P43 Circulating Troponin I-IGG Complex (Macrotroponin) as a Cause Of Raised Plasma Troponin I

Introduction

Troponin (either as I or T measurement) is the most important analyte in the investigation of cases of suspected acute myocardial infarction (AMI). In all immunoassays interference by a variety of mechanisms – most often involving heterophile antibodies – can be seen. We have identified a number of patients, where we detected the presence of circulating Troponin I (TnI) complexed with IgG. We developed a protocol to determine the cause of the interference and confirm the presence of TnI-IgG complex.

Methods

Samples with persistently elevated or suspect TnI results were treated with Scantibodies Heterophile Blocking Tubes (HBT). Samples that did not show a drop in the apparent TnI level were then passed through a Protein G affinity cartridge. The unbound fraction was concentrated and analysed for TnI, IgG and albumin. The bound fraction was eluted with a low pH buffer and the same analytes measured. Samples from patients after AMI with comparable TnI level were processed as controls using the identical protocol.

Results

We identified 6 patients who showed continuously elevated TnI and whose samples were positive in multiple TnI assays. The TnI values did not decrease after preincubation with HBT. Protein G extracted IgG from both patient and control samples. The flow through of control samples was comparable to the pre-treated level but markedly reduced in patient samples. After elution of the IgG and pH adjustment the TnI immunoactivity in the patient samples but not in controls was recovered.

Conclusions

Immobilised Protein G may be used to investigate a persistently elevated TnI that is inconsistent with the clinical history or is not corrected by treatment with heterophile antibody blocking tubes. Early identification of the spurious result may reduce the incidence of unwarranted patient testing and further investigations, and shorten lengthy hospital admissions.

Clin Biochem Rev. 2008 November; 29(3): S143.

P44 IGG4 and Antigen Excess in the Investigation of Autoimmune Pancreatitis

Introduction

IgG subclasses have been measured for many years in the investigation of immune deficiencies that contribute to chronic infections. Recently an association between an elevation in IgG4 and autoimmune pancreatitis (“Japanese pancreatitis”) has been recognised. Protocols have now been published that include IgG4 measurement as one criterion for distinguishing between autoimmune and other forms of pancreatitis. While investigating one such case, we noted a large discrepancy between the measured IgG level and the sum of the four IgG subclasses. Suspecting an antigen excess problem, we repeated the analysis of IgG4 at a tenfold dilution and found that the original result was underestimated by a factor of five.

Methods

IgG subclass measurements were performed on a Beckman IMMAGE using The Binding Site reagents supplied by Beckman Australia. The sample was diluted tenfold using Beckman diluent for the repeat of the IgG4. IgG was measured on an Abbott Architect 8200sr with reagents supplied by Abbott Australia.

Discussion

Antigen excess in the measurement of one of the IgG subclasses would most likely be encountered when a patient with monoclonal gammopathy was tested. In the case of autoimmune pancreatitis, the subclasses are all polyclonal in nature.

Although the measurement if IgG4 is only one of three or four criteria used in the diagnosis of autoimmune pancreatitis, the reporting of a false negative result could be confusing or cause difficulties in reaching a diagnosis and perhaps lead to unnecessary procedures being undertaken.

Conclusions

There needs to be an awareness of the potential for antigen excess causing underestimation of IgG subclasses and that this population is significantly different to what is usually encountered in the investigation of immune deficiencies. Comparing the sum of individual subclasses to the total IgG should alert the analyst to the problem in most cases.

Clin Biochem Rev. 2008 November; 29(3): S143.

P45 AACB CSF Working Party Investigation of CSF Total Protein Result Variation.

Introduction

CSF total protein results from the RCPA QAP show distinct assay groups. The CSF Working Party conducted a survey of patient CSF samples to determine the cause of this variation in the assay results.

Methods

Three laboratories distributed 12 unfrozen patient CSF samples. The following analysers were used: Roche Modular PPE- Lab A; Abbott Laboratories Architect ci8200- Lab B; Beckman Coulter DXC 800- Lab C; Ortho-Clinical Diagnostics Vitros Fusion 5.1- Lab D and Siemens Dimension RXL Max- Lab E.

Results

The mean of Lab A and B results were used and a ratio was determined between this mean and the results of the other 3 labs. The ratio for the RCPA CSF QAP samples for Lab C = 1.09; Lab D = 1.37; Lab E = 1.19. The ratio for the patient CSF samples for Lab C = 1.09; Lab D = 1.37; Lab E = 1.14. Patient CSF samples 3, 8 and 9 were haemolysed. Lab E was not able to assay four of the patient samples.

Conclusions

Patient CSF total protein results showed the same pattern of variability as seen in the RCPA CSF QAP samples. Since a ratio was used we conclude that there is no matrix effect contributing to the variation seen in the RCPA QAP reports. Thus the groups of results can be contributed to calibration or method variation and more standardisation of assays is required to improve comparability between laboratories.

Clin Biochem Rev. 2008 November; 29(3): S143.

P46 Observations on the Impact of Sam Newman’s Prostate Cancer Publicity on Pathology Testing in Men

Introduction

Celebrity endorsement of medical testing and medical procedures is known to have an impact on usage. Pathology laboratories often brace themselves following media exposure concerning a pathology test. This study examines the effect of a television personality’s announcement of prostate cancer and his promotion of the PSA test.

Methods

The media announcements of Sam Newman’s prostate cancer were made in the local and national press on March 6 and 7, 2008. During that period both he and other celebrities encouraged men to talk to their GP about having a PSA blood test. We examined our PSA request rates for the period following that announcement and sought to control for seasonal variations in test requesting by comparing PSA request rates with total cholesterol test rates in both men and women.

Results

For the three weeks following the announcement, the daily PSA requests rate increased by 49%. A closer analysis shows that there was also an increased rate of cholesterol testing in men during the same period (28%), but only a mild increase in the rate of cholesterol testing in women during the same period (3%).

Conclusions

There was a sudden, large and sustained increase in PSA request rates following the media announcements. Many men also had cholesterol tests ordered. The impact of a celebrity health message may extend beyond the particular illness concerned.

Clin Biochem Rev. 2008 November; 29(3): S143.

P47 Likely Matrix Effect with RCPA-QAP Samples for Ultrasensitive PSA on the Immulite 2500 Analyser

Introduction

PSA measurements are an important tool in the diagnosis, prognosis and monitoring of prostatic carcinoma. Since changing from total PSA (tPSA) to ultrasensitive PSA (sPS) method on the I2500, under-recovery was noted for both the Endocrine and the Tumour Marker program. No method subgroup exists in the RCPA-QAP program for sPS, so we were unable to determine if we have any method peers. We sought to identify the nature of the apparent discrepancy.

Methods

RCPA-AQAP samples which had been tested for Endocrine Cycle 29 and Tumour Markers Cycle 31 were retested on the Immulite 2500 using tPSA and sPS. In addition, samples from previous cycles and 40 patient samples were retested with both methods. We performed dilution studies using the manufacturer’s analyser-specific diluent with QAP and patient material to test for linearity.

Results

Test results were analysed with the Passing Bablok regression test. Patient samples showed the expected good correlation between methods (slope 0.979, intercept −3.31). The QAP samples showed a consistent negative bias with sPS (Endocrine slope 0.746, intercept 0.3596, Tumour Markers slope 0.595, intercept −2.4). The difference was independent of PSA concentration. Good linearity was seen for QAP and patient material by dilution. Three non-Immulite RCPA-QAP respondents in the low recovery subgroup were identified as Ortho Diagnostics users.

Conclusions

A significant negative bias, presumably a matrix effect, was identified for QAP samples, the effect was independent of PSA concentration. We suggest that a new subgroup classification for high sensitive PSA assays is required for the RCPA-QAP program. In the meantime, users of sensitive PSA methods need to interpret QAP data with caution. The nature and extent of the matrix effect requires further clarification.

Clin Biochem Rev. 2008 November; 29(3): S144.

P48 Evaluation of the Abaxis Piccolo Xpress Point of Care Instrument

Introduction

The Piccolo Xpress is a portable blood analyser designed for use at the point of care. 100 μl of whole blood, serum or plasma is added to the single use plastic disc and placed in the analyser where centrifugal and capillary forces are used to mix the reagents and sample in the disc. The Piccolo Xpress calculates the results from the absorbance data and reports the results. Results are available after approximately 12 minutes of the reagent disc being placed in the instrument.

Methods

We evaluated the accuracy, precision and linearity of the following panels on the Piccolo Xpress; Metlyte 8 Panel (sodium, potassium, chloride, bicarbonate, creatinine, glucose, CK, and urea), Liver Panel Plus (total protein, albumin, ALT, ALP, AST, GGT, total bilirubin and amylase) and Lipid Panel Plus (total cholesterol, HDL, triglycerides, ALT, AST and total glucose). Accuracy was evaluated by performing 50 patient comparisons with the local laboratory. Between day and within day precision was calculated using Abaxis controls. Linearity was evaluated using linearly related samples provided by the RCPA Chemical Pathology Quality Assurance Program

Results

Samples tested in this evaluation showed closely related results between the Piccolo Xpress and laboratory. Precision of the Piccolo Xpress was also acceptable with CVs ranging from 1% to 8% for within day precision and 1% to 12% for between day precision for all analytes. Linearity results were generally good across the majority of analytes.

Conclusions

The Piccolo Xpress displayed high-quality analytical performance in this study. The instrument would prove useful where point of care testing for the above analytes is indicated.

Clin Biochem Rev. 2008 November; 29(3): S144.

P49 The Effect of Digibind® on the Abbott CI8200 Digoxin Assay

Introduction

DIGIBIND® (GlaxoSmithKline, Parma, Italy) is a digoxin fab antibody therapy designed to bind to excess digoxin in overdosed patients. The antibody therapy is known to interfere with the digoxin immunoassays, though the specific effect has not been quantified for the Abbott Ci8200. The aim of this study was to characterise the effect of DIGIBIND® on the Abbott Ci8200 and to measure the free digoxin levels in patients administered DIGIBIND®.

Methods

DIGIBIND® [9.5 g/L] was diluted in human serum albumin, concentration range 0.3–9.5 mg/L, to test its interference of the digoxin assays. Four digoxin serum pools (0, 1, 2, 6 nmol/L) were tested with increasing concentrations of DIGIBIND® (0.05–4 mg/L). Digoxin levels were measured using the Abbott Ci8200 and compared to the Abbott TDx following ultrafiltration (Amico Ultra, Millipore). Assays were calibrated according to manufacturers’ specifications. Blood samples were collected from four patients post DIGIBIND® therapy, and digoxin concentrations were measured.

Results

DIGIBIND® interference of the Ci8200 digoxin assay was displayed as a linear relationship of digoxin concentration to DIGIBIND® concentration. Ultrafiltration of serum pools and patients’ samples removed digoxin bound to DIGIBIND® and allowed free digoxin to be measured on both digoxin assays. DIGIBIND® concentrations of 0.8–4 mg/L in all serum pooled samples caused interference in the Abbott Ci8200 assay. The patients’ samples showed little or no detectable free digoxin post administration of DIGIBIND®.

Conclusions

We recommend that patients’ samples post antibody therapy are unsuitable for digoxin measurement. Furthermore, the assessment of the digoxin antidote DIGIBIND® must be determined using clinical signs.

Clin Biochem Rev. 2008 November; 29(3): S144.

P50 Association of -2548 G/A Polymorphism in the Leptin Gene with Pre-Eclampsia/Pregnancy Induced Hypertension (PIH)

Introduction

The OB gene which codes for leptin consists of three exons and two introns and spans approximately ~18 kb of genomic DNA. The promoter region contains several putative binding sites for transcription factors. Cytosine enhancer binding protein (C/EBP) and peroxisome proliferator-activated receptor-γ of the OB gene promoter are up regulated in pre-eclampsia (PE)/pregnancy induced hypertension (PIH). This study was carried out to investigate possible associations of -2548 G/A polymorphism in leptin gene promoter and the serum leptin and soluble leptin receptor (SLR) concentrations of patients with PE/PIH.

Methods

Non-fasting venous blood samples were obtained from women with PE/PIH (n = 62) and with normal pregnancies (63). Plasma leptin and SLR were determined using immunoassays. Polymerase chain reaction followed by RFLP with CfoI was used to determine the -2548 G/A polymorphism. Banding patterns were visualised on agarose gel and the polymorphism was confirmed by direct sequencing. In addition, the -2548 G/A polymorphism was also studied in 60 unrelated healthy controls.

Results

Relative risk (RR) for PE/PIH in the presence of A allele (AA and AG genotypes) against carriers of GG genotype was significantly higher (RR = 1.67, 95%CI 1.28 – 2.18, p<0.0001). Two way ANOVA indicates a significant effect of PE/PIH (p<0.0001) but not the genotype for leptin, leptin/SLR, BMI and leptin/BMI. For SLR both PE/PIH (p<0.0001) and genotype (p<0.05) has a significant effect.

Conclusions

The leptin system, in particular the leptin genotype, could therefore be a contributory factor in the development of PE/PIH.

Clin Biochem Rev. 2008 November; 29(3): S144–S145.

P51 Reference Range Study of Soluble Transferrin Receptor (sTfR) And sTfR/Log Ferritin Ratio on Local Population

Introduction

Serum ferritin is sensitive in diagnosing iron deficiency anaemia but being an acute-phase protein, it is a poor measurement of iron status in patients with other diseases involving the acute-phase response. Serum Transferrin Receptor (sTfR), not affected by acute-phase reactions, is valuable in such situations. This study seeks to establish the reference ranges for sTfR and sTfR/log ferritin ratio (sTfR-F index) in our local population.

Methods

Serum samples from 85 males (20–60years) and 85 females (25–45years) with normal haematological indices were analysed for sTfR and ferritin on ROCHE COBAS INTEGRA400 Tina-quant sTfR assay and SIEMENS ADVIA Centaur respectively. Precision and linearity studies were performed on Tina-quant sTfR assay.

Results

Reference ranges (2.5th to 97.5th percentile) of sTfR and sTfR-F index: males 2.0–4.8 mg/L and 0.7–2.4; females 2.0–5.9 mg/L and 1.1–4.9. Linearity for sTfR: y = x + 3.50, r = 0.998, recovery 97.2–99.8%. Precision at sTfR concentrations 4.1 and 15.5 mg/L has intra-assay CVs of 1.4 and 3.1% and inter-assay CVs of 3.1 and 3.0%.

Conclusions

Local male population has a sTfR reference range similar to manufacturer’s range (males: 2.5–5.0 mg/L, females: 1.9–4.4 mg/L), females are slightly higher. Local sTfR-F index reference ranges for both males and females are also higher compared to study done by Thomas, et al. Clin Chem 2002;48:1066-76 (males: 0.2–3.7, females: 0.6–3.8). These reference ranges would aid in the assessment of iron status in patients with diseases accompanied by acutephase response in a local context.

Clin Biochem Rev. 2008 November; 29(3): S145.

P52 Biochemical Assessment of Twin Pregnancies in First Trimester Screening (FTS) for Down’s Syndrome

Introduction

Combined FTS involves the measurement of Nuchal Translucency by ultrasound and Free βHCG and Pregnancy Associated Plasma Protein-A (PAPP-A) using immunoassay. This data is corrected for gestation to produce Multiple of Medians (MoMs) using medians established in specific populations. The aims of this study was to a) compare gestation specific medians in populations of twin and singleton pregnancies and b) determine how these medians compared to twin specific medians determined in other populations.

Methods

During the period 17/11/05 to 30/04/08 166 FTS requests from twin pregnancies and 30,607 from singleton pregnancies were measured at 10–14 weeks gestation. Serum levels of Free βHCG and PAPP-A were measured using a Siemens Immulite 2000. Gestation medians were calculated and compared at each week in both twin and singleton pregnancies.

Results

Requests from twin pregnancies constituted 0.6% of all FTS requests received. Serum levels of Free βHCG and PAPP-A were significantly higher (p <0.01) in twin pregnancies compared to singleton pregnancies. Median Free βHCG levels were 87.2, 60.6, 67.7, 53.9 in twin and 52.5, 44.70, 37.9, 32.1 IU/L in singleton pregnancies at 10, 11, 12, and 13 weeks respectively. Median PAPP-A levels were 2.58, 2.61, 4.16, 7.32 in twin and 0.85, 1.30, 1.89, 2.77 IU/L in singleton pregnancies at 10, 11, 12, and 13 weeks respectively.

Conclusions

Serum levels of biochemical markers used in FTS are two-fold higher in a small local population of twin pregnancies compared to singleton pregnancies. This finding is broadly in agreement with larger studies using mixed populations. Larger numbers would be required to establish population specific regression data in twin pregnancies.

Clin Biochem Rev. 2008 November; 29(3): S145.

P53 Measurement of TSH Receptor Antibodies on the Roche E170 and the Brahms Trakhuman Radio Receptor Assay

Introduction

The objective of this study is to compare the measurement of TSH receptor antibody (TRAB) between two different assays. The Roche E170 uses a human thyroid stimulating monoclonal antibody with electrochemiluminescence immunoassay (ECLIA). The ECLIA method is referenced to the BRAHMS radio receptor assay (RRA).

Methods

Patient samples (n = 85) were assayed by Roche E170 and BRAHMS RRA. Intra-assay precision was evaluated at three TRAB levels. Calibration and internal QC samples were performed daily for both assay types over one working week.

Results

For all patient samples (n = 85) linear regression was E170 = 1.24RRA − 0.32 (r2 = 0.91). At TRAB levels between 1 and 10 IU/L linear regression of patient data (n = 75) yielded E170 = 1.14RRA − 0.02 (r2 = 0.8665). E170 intra-assay respective CVs were 11.1%, 5% and 1% at TRAB levels of 1.35 IU/L, 3.28 IU/L and 10.4 IU/L. Inter-assay precision for the E170 was 4% and 3.8% CV at TRAB of 3.5 IU/L and 15.3 IU/L respectively. Different QC material was used for inter-assay precision on the BRAHMS respectively giving CVs of 7.8% and 4.5% at TRAB levels of 2.7 IU/L and 20.6 IU/L.

Conclusions

This study demonstrates that these two methods for determination of TSH receptor antibody in human serum correlate well. Intra and inter-assay precision on the E170 were comparable to the Brahms RRA. There was clinical concordance between the two methods for all patient samples.

Clin Biochem Rev. 2008 November; 29(3): S145.

P54 Extractable Negative Interferent of 25 Hydroxy Vitamin D Immunoassay

Introduction

Erroneous laboratory results attributable to interferents can lead to adverse effects on patient care. The most common interferent in immunoassays is animal antibodies. However, atypical proteins and autoantibodies can also interfere. Usually these interferents lead to a falsely elevated or a false positive result, while false negative or falsely low results due to interferents are extremely rare. We report a case of an extractable negative interferent in vitamin D enzyme immunoassay.

Methods

Four consecutive serum samples collected over a period of 4 months from a 72 year old man were assayed using the IDS 25OHD enzyme immunoassay (EIA) kits and an automatic plate reader (Best 2000). Aliquots of these samples were read on MS/MS for vitamin D3, re-assayed with EIA following treatment with heterophile antibody blocking agents (HBA) and following NaOH and acetonitrile extraction. Apparently normal serum samples showing high and low levels of 25OHD were treated identical to the patient’s serum sample and assayed on the same plate for comparison.

Results

EIA results for all serum specimens were undetectably low with very high optical densities. Examination of plates showed a black precipitate which did not improve following the treatment with HBA. However, extraction with NaOH and acetonitrile eliminated the black precipitate which allowed the EIA method to read results consistent with the MS/MS method. Rheumatoid factor was in the reference range.

Conclusions

An apparent negative interferent other than a heterophile antibody in the patient serum was successfully eliminated by extraction. This interferent may be high levels of D binding protein or some other protein that strongly binds 25OH vitamin D such as heat shock proteins.

Clin Biochem Rev. 2008 November; 29(3): S145–S146.

P55 Age Related Increase in Cea Levels

Introduction

CEA has been used to monitor lung and bowel cancer for over 30 years. The earliest studies have described an increase in CEA values with age and subsequent studies have found that the rise is also seen in non-smoking elderly patients without confounding illnesses. This study defines CEA reference intervals by an indirect technique using a large private pathology database.

Methods

57,026 CEA value were obtained using the Abbott Architect CEA method between March 2005 and October 2007. Using the Bhattacharya technique of separating underlying Gaussian distributions from a mixed population we were able to define the dominant underlying population across each decade of age.

Results

The CEA distribution was log Gaussian in both sexes and although the male interval was slightly higher than the female interval at most ages, the difference was considered to be minor compared with the overall impact of aging. CEA level rose almost linearly with age (see Table).

Table

Reference intervals defined by the Bhattacharya technique for Abbott Architect CEA.

Age18–3030–4040–5050–6060–7070–8080+
N7382,2866,37611,77313,90113,3358,617
Ref Int0.4–4.00.4–5.00.4–6.00.5–7.00.5–8.00.6–9.00.7–10.0

Conclusions

CEA reference intervals appear to increase with age. Like most age related reference intervals their application will result in improved specificity of the test, but will probably decrease sensitivity. While age related reference intervals are not likely to improve the poor suitability of the CEA for bowel cancer screening, the age related increase should be considered in prognostic studies.

Clin Biochem Rev. 2008 November; 29(3): S146.

P56 Performance of the New Bio-Rad HBA1C Nu Variant II Reagent with Comparison to the Previous HBA1C Variant II Reagent

Introduction

Bio-Rad introduced new reagents for the Variant II HbA1c method with claims of increased cartridge injections, greater stability of reconstituted calibrators and interchangeable buffers within cartridge resin lots. The aim of this study was to compare the accuracy and precision of the two kits.

Methods

Routine samples submitted for HbA1c testing were analysed using the previous Variant II reagent, and then using the new Variant II HbA1c NU reagent. Both the intra- and inter-assay precision was evaluated. QC was run 8 times daily for 1 month.

Results

The intra-assay respective CV at HbA1c levels of 5.1% and 12.6% were 0.99% and 0.87%. Inter-assay precision was 2.2% and 1.5% CV at HbA1c levels of 5.6% and 9.8% respectively. Using the same QC material, the old reagent means were unchanged with inter-assay CV of 2.1% and 1.8% respectively. Correlation of patient data (n = 134) yielded r2 = 0.996 (slope = 1.034, intercept = −0.25).

Conclusions

These results demonstrate that the new Variant HbA1c kit is a precise, reproducible method for determining levels of HbA1c. NU kits also show better inter-assay precision at higher levels. Cartridge life is doubled from 500 to 1000 injections.

Clin Biochem Rev. 2008 November; 29(3): S146.

P57 Comparison of Bio-Rad and Roche Cobas Integra HBA1C Assays

Introduction

Bio-Rad and Roche have respectively recently introduced improved HPLC and immunoassay HbA1c assays. It has been demonstrated that both of these assays are unaffected by variant haemoglobin S. These studies did not address the problem of carbamylated haemoglobin (CHb). This study discusses the effect of CHb on the Bio-Rad Variant II, Variant II Turbo assay and the Roche Cobas Integra 800 assays for HbA1c.

Methods

Patient samples (149) were run on both the Bio-Rad Variant II and Roche Integra HbA1c assays. Samples (n = 54) with CHb from the Bio-RadVariant Turbo were reanalysed on the Bio-Rad Variant II.

Results

Linaer regression analysis of patient data (n = 149) was Roche Integra = 0.90 Variant II + 0.35 (r2 0.96). Carbamylated Hb did not interfere in Roche Integra Immunoassay. Artefactual results were observed in all 54 samples containing CHb when assayed by the Bio-Rad Variant II Turbo assay. The Bio-Rad Variant II assay suffered CHb interference in four of the 54 samples.

Conclusions

Our results demonstrate that the Roche Integra Immunoassay and the Bio-Rad Variant II HPLC assays for HbA1c performs better than the Bio-Rad Variant Turbo for patients who have CHb. This observation is consistent with recent RCPA QAP data using a sample spiked with CHb.

Clin Biochem Rev. 2008 November; 29(3): S147.

P58 A Correlation between the Radiometer AQT90 Flex Poc System and the Abbott Architect I2000SR Immunoassay System for Troponin I

Introduction

The Radiometer AQT90 FLEX is a new random-access point-of-care immunoassay analyser intended for the quantitative determination of cardiac, coagulation and sepsis markers in blood. The Radiometer Troponin I assay utilises a sandwich assay with all the specific reagents provided in a dry stable form inside a cup. The capture antibodies are pre-immobilised to the surface of the cup. The tracer antibodies are labelled with europium chelate and bind to the top of the capture antibodies.

Methods

Whole blood lithium heparin samples from 179 patients were analysed for Troponin I on the AQT90. These samples where then centrifuged, and the plasma analysed for Troponin I on the Abbott Architect iSR2000. The samples were analysed on average within 100 minutes of each other. Linearity and precision studies were also performed to evaluate assay performance.

Results

AQT 90 Troponin I correlated well with the Architect Troponin I (r = 0.85; p = <0.0001) with a slope of 0.23. The assay gave a between run precision of 10.8 % at 0.024 μg/L, 4.6% at 0.220 μg/L and 3.5% at 0.956 μg/L. A sample diluted linearly from 4.1 μg/L to 0.016 μg/L. Two patients identified as false positives and one patient was borderline false negative.

Conclusions

The AQT90 Troponin I assay shows a high degree of correlation with the Abbott Architect Troponin I assay. The quality of results, the system’s ease of use, and ability to perform testing on whole blood, make it a potential clinical asset for Emergency Department based point-of-care testing.

Clin Biochem Rev. 2008 November; 29(3): S147.

P59 Assay of Direct Renin by Automated Chemiluminescence Immunoassay

Introduction

Renin measurement is used to evaluate the renin-angiotensin system in disease states, in particular for primary and secondary hyperaldosteronism. An evaluation of the Liaison automated chemiluminescence immunoassay (CLIA) for Direct Renin (DR) was undertaken to improve turnaround time and convenience compared with a plasma renin activity (PRA) radioimmunoassay (RIA).

Methods

Plasma samples (n = 109) were assayed using the Diasorin PRA RIA kit and Diasorin Liaison CLIA automated DR assay. Precision data were derived under routine operating conditions. Assay turnaround was monitored using laboratory information system data.

Results

DR = (12.9 x PRA) + 1.4 (all results); DR = (19.5 x PRA) − 0.1 (n = 64, PRA <1.0 ng/mL/hr). (Passing-Bablok comparison). Day-to-day precision of new and old assays was as follows:

AssayRef IntervalQCMeanSDCV %
Direct Renin (uIU/mL)erect 7–50Low QC1.80.844.4
supine 4–40Kit QC 126.52.17.9
Kit QC 2104.85.85.5
PRA (ng/mL/hr)erect 1.3–4.0QC 10.80.112.5
supine 0.1–2.4Pool 14.60.510.8
Pool 215.92.113.2

The quoted functional sensitivity of approximately 2.0 uIU/mL for the DR assay (CV between-assay 20%) was not achieved (see table).

Conclusions

The automated DR assay is much easier to perform than manual RIA, with a sample turnaround of under two hours. Performance is adequate but not as good as PRA at low renin levels found in patients with primary hyperaldosteronism. Manufacturers of DR assays should provide quality control material to validate performance at the lower limit of their recommended reference interval. After pathologist review, the clinical decisions reached using either assay were the same for all but one patient.

Clin Biochem Rev. 2008 November; 29(3): S147.

P60 Measurement of Serum Testosterone by Isotope Dilution-Liquid Chromatographytandem Mass Spectrometry (ID-MSMS)

Introduction

All immunoassays for serum testosterone in women give unpredictable false-high results for some samples. The aim of this study was to examine the use of ID-MSMS as an alternative method to immunoassay and to establish the analytical sensitivity of our instrument.

Methods

Serum (200 μL male or 500 μL female) and di-deuterated testosterone as internal standard (IS) were co-extracted with methyl-tert-butyl-ether. After evaporation, steroids were dissolved in 50% methanol and subjected to HPLC with MSMS detection on an AB 3200 Q-Trap in positive electrospray mode. The intensity of ion transitions m/z 289»109,97 for testosterone and 291»99 for IS were recorded over 3 min for peak area quantitation. Immunoassays were performed on a Roche E170 analyzer (males) or a Siemens Advia Centaur (females).

Results

For sera from 22 males with a concentration range 1–42 nmol/L, there was a high correlation (r = 0.99) and good agreement between immunoassay (y) and ID-MSMS (x) (y = 1.11x − 1.5; mean bias 0.6 nmol/L). For 94 female sera with testosterone <3 nmol/L, ID-MSMS produced concentrations less than immunoassay (y = 1.88x − 0.29; mean bias 0.5 nmol/L) and the correlation was not as good (r = 0.73). Inter-day precision for a serum testosterone level of 0.6 nmol/L gave a CV of 4.8% if assayed in duplicate and 11.2% in singlicate.

Conclusions

Our ID-MSMS method is easy to perform and gives serum testosterone concentrations less than immunoassay, presumably because of its very high specificity. Analytical sensitivity was adequate to measure serum testosterone in women with a limit of quantitation around 0.5 nmol/L.

Clin Biochem Rev. 2008 November; 29(3): S147.

P61 Comparison of a Complete Commercial Kitset Clinrep® with a Validated Extraction Method for the HPLC Determination of Serum Vitamins A & E

Introduction

Fat soluble vitamins are routinely measured to assess the vitamin status of patients suspected or known to have a malabsorption syndrome or possible nutritional deficiency. Recipe® offers ClinRep® a newly modified complete HPLC kit for measurement of vitamin A (retinol) and vitamin E (α-tocopherol). We compared the performance of ClinRep® with our validated solvent extraction HPLC method.

Methods

ClinRep® printed instructions were followed which required precipitation and stabilisation by simple sequential addition of two reagents followed by direct HPLC injection. These reagents are supplied by the manufacturer and were used without modification. ClinRep® supplies a single δ-tocopherol internal standard as a part of the precipitation reagent. Our solvent extraction method is more involved and follows the widely recognised technique of sample precipitation, extraction into hexane, evaporation and subsequent reconstitution. A single calibrator (from Chromsystems®) is used. Our extraction method uses dual retinyl acetate and α-tocopheryl acetate internal standards. Controls were supplied by Chromsystems® and Recipe® ClinChek®. Both HPLC methods use time programmed wavelength changes at 325nm and 292nm.

Results

Regression between ClinRep® precipitation method and our extraction method is 1.81 (retinol) + 0.1 (r2 = 0.92) and 1.26 (tocopherol) − 0.7 (r2 = 0.87) showing significant non-unity slope especially for retinol. Overall between-batch imprecision for ClinRep® is 9.6% for retinol levels ranging from 1.9 to 4.6 μmol/L and 9.2% for tocopherol levels ranging from 19.8 to 56.3 μmol/L (12 specimens over 5 batches). The corresponding imprecision for our extraction method is 3.9% and 5.8% respectively. The values for trilevel Recipe® ClinChek® controls as estimated by ClinRep® were higher than the stated targets with both extraction and precipitation methods.

Conclusions

The ClinRep® precipitation method is rapid and easy to perform and does not block the fluidic lines. Co-extracted bilirubin is resolved from retinol. Carotenoids are not reliably extracted. The ClinRep® method is less precise and there is a bias between the methods due to differences between calibrators especially for retinol.

Clin Biochem Rev. 2008 November; 29(3): S147–S148.

P62 The Ability of Different GFR Estimating Equations to Predict Creatinine Clearance in Subjects with Varying BMI and Height

Introduction

It is not always clear which estimating GFR equation should be used for drug dosing calculations. The modification of diet in renal disease (MDRD) equation is suggested by the Australasian Creatinine Consensus Working Group for use in non-critical-dose drugs, emphasising the need to correct for body surface area (BSA) in people of abnormal body sizes (uncorrected MDRD). The Australian Medicines Handbook cautions against using the MDRD equation, advocating the Cockcroft-Gault formula or measured creatinine clearance in all situations. What matters for drug dosing is which equation best predicts the patient’s actual creatinine clearance.

Methods

Estimating GFR equations were calculated for 643 patients and compared to measured creatinine clearance in subjects with varying body mass index (BMI) and height using linear regression.

Results

Plasma creatinine increases slightly with BMI with a greater increase in 24h urine creatinine excretion resulting in a small increase in measured creatinine clearance. This rise with BMI is reflected in the Cockcroft-Gault formula using actual body weight and the uncorrected MDRD formula. The Cockcroft-Gault formula using lean body weight and the MDRD formula expressed per 1.73m2 in contrast give a decreasing estimate of creatinine clearance as BMI rises. Plasma creatinine does not change as height increases whereas 24h urine creatinine excretion increases by 44–47%. All equations tested predicted the rising creatinine clearance with height.

Conclusions

The equations which predict creatinine clearance from plasma creatinine behave differently in subjects whose increased BSA is due to increased height, increased BMI or pregnancy. Equations that do not take these factors into account are unsuitable for drug dosing calculations.

Clin Biochem Rev. 2008 November; 29(3): S148.

P63 Performance Characteristics of an Automated EIA For 25-Hydroxyvitamin D

Introduction

The number of requests for 25-hydroxyvitamin D (25-OH D) has been growing rapidly. We investigated the performance of the IDS-EIA assay on the robotic BEST 2000 ELISA plate pipettor (Abacus ALS). Reproducibility, linearity, dilution of high end values, across-plate drift and the characteristics of singleton performance were assessed.

Methods

Kit-supplied and in-house quality control were evaluated over 80 batches. Samples with 25-OH D values >120 nmol/L were serially diluted in either a low value sample, or with the recommended sample diluent (PBS+Alb). Singleton performance was assessed by comparing quality control data, and by performing 304 unknowns twice in separate singleton batches. Across-plate drift was assessed using in-house QC material distributed evenly through several batches. We participated in 3 cycles of the Vitamin D External Quality Assessment Scheme (DEQAS).

Results

Interassay CVs were 3.5% (duplicate runs, n = 80) and 4.3% (singleton runs, n = 8) at 28.9 nmol/l, 4.3% and 4.8% at 98.0 nmol/L and 5.6% and 6.2% at 73.0 nmol/L. Recoveries for dilutions in human serum were 82–104% (mean 96%), while dilutions in PBS+Alb were 104–149% (125%). The CV for unknowns repeated in separate singleton batches was 5.9%, ±6.5 (mean, SD), with 10.7% of CVs in the range 11–20% and 1.8% were >20%. Within assay unknown CVs were 3.3% ±2.7 (n = 149). Mean across-plate variation was 4.0%. Altman-Bland analysis of DEQAS data indicated a bias of +2.50 nmol/L compared to ALTM.

Conclusions

The assay was sufficiently precise and accurate for clinical application. Diluting samples in the recommended buffer gave unpredictable recoveries, suggesting that the IDS-EIA is matrix sensitive. Further improvements in assay performance are required to permit singleton analyses.


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