Achieving interlaboratory agreement of enzyme activity measurements represents one of the most important standardisation efforts in Laboratory Medicine.¹ The fact that the determinations of some enzymes (i.e. creatine kinase (CK), lactate dehydrogenase (LD), aspartate aminotransferase (AST), alanine aminotransferase (ALT), γ-glutamyltransferase (GGT), alkaline phosphatase (ALP), amylase (AMY), and pancreatic lipase) are among the most frequently ordered tests in clinical laboratories emphasises the importance of standardised measurement results. These enzymatic determinations are important biochemical parameters for the diagnosis and monitoring of diseases of liver, pancreas, skeletal muscle, bone, etc.²

Despite the efforts towards the standardisation of enzymatic assays, variability of the results among laboratories is still observed. In 2002, the Institute for Reference Materials and Measurements (IRMM) of the European Union (EU) surveyed 950 global laboratories in an International Measurement Evaluation Program (IMEP) for the measurement of two commonly measured enzymes in human serum (GGT and AMY).³ As an example, results for GGT showed biases of −60% to +30% at a serum catalytic activity of 35 U/L, a value close to the upper reference limit (URL). It is easy to argue that, on the basis of this large variation, many of the study participating laboratories would have misclassified patients at this critical decision level.

The catalytic activity of an enzyme is a property measured by the catalysed rate of reaction, produced in a specific assay system, and is not an amount of substance. If the components of the reaction system (e.g. pH and buffer, temperature, presence of activators and inhibitors, substrate nature and concentration) are changed, the magnitude of the measured activity will also change. Therefore, the numerical results of catalytic activity measurements depend entirely on the experimental conditions under which the measurements are made, i.e. are method-dependent. Consequently, two procedures that measure the catalytic activity of the same enzyme but under different analytical conditions may produce different results for a given sample thus preventing the use of the same reference interval or decision limit.

In the standardisation of enzyme assays, therefore, a reference measurement procedure, which defines the conditions under which a given enzyme activity is measured, occupies the same role as that of primary reference materials for nonenzyme analytes of known mass or substance concentration.⁴ One way to improve the comparability of results can therefore be through the widespread use of standardised analytical methods (“method globalisation”).⁵ From the 1970s national and international expert panels have carried out fundamental research work to determine the optimised conditions for measurement of the catalytic activity of several enzymes in human serum. Recommended measurement procedures were defined, which gave optimum reaction conditions, such as substrate concentration and reaction temperature. However, it is now clear that the approach of “method globalisation” for enzyme standardisation has its limitations and the goal of a single, universal method to measure the catalytic concentration of a given enzyme cannot be achieved in clinical practice. For instance, although in 1998 Lorentz published on behalf of the IFCC a recommended method for the measurement of AMY catalytic activity,⁶ more than 20 different substrates are still currently employed and large differences are evident for the results among commercial assays.^3,7

To minimise these method differences among laboratories, results are often expressed as multiples of the URL, particularly in multicentre clinical trials. Recent studies have demonstrated however that the expression of enzyme results as URL multiples increases the scatter of results due to both inter-method differences and the different laboratory reference limits. Hence, this mode of result expression should be discouraged.^8,9

The aim of the efforts to standardise enzyme measurements is to achieve comparable test results for human serum samples independent of the test kits and instruments used, and also independently of the laboratory where the measurement is performed. To achieve this goal one approach is needed for a reliable transfer of the measurement values from a higher-order method to methods which are routinely used in the laboratories. Such a “reference system” approach is based on the concepts of metrological traceability and a hierarchy of measurement procedures.¹⁰ The reference measurement procedure, which defines the conditions under which a given enzyme activity is measured, occupies the highest level of this enzyme reference system.¹¹ In addition to the reference measurement procedure the system also requires reference materials for the intermediate transfer of values from the reference procedure to the routine laboratory assays (Figure 1).¹² Once the reference material is certified, this material and the manufacturer’s standing procedure can be used by industry to assign values to commercial calibrators. Routine clinical laboratories that use commercial methods and validated calibrators to measure patient samples in turn obtain values traceable to the reference measurement procedure and independent of the particular method or instrument (“result globalisation”).¹³

Figure 1 The reference system for enzyme measurement (Adapted from reference 12).

The applicability of the enzyme reference system concept is possible only if the reference materials used to transfer trueness to the field methods are commutable and if the reference measurement procedure and corresponding lower - order routine methods have identical, or at least very similar, specificities for the measured enzyme.

Rej has defined commutability in enzymology as “the ability of an enzyme material to show interassay activity changes similar to those of the same enzyme in human serum”.¹⁴ To assure complete traceability, the reference material must be commutable by the reference measurement procedure and each of the commercial methods for which this common reference material is to be used as the master calibrator. Calibration of commercial methods with non-commutable reference materials can cause worse, rather than improved, agreement of results among methods for clinical samples.¹⁵ Commutability of reference materials can be affected by many factors including the source of material, purification procedure, matrix of the solution, lyophilisation, addition of stabilisers or other additives.¹² If commutable reference materials suitable for the direct calibration of field methods are lacking, a panel of native human sera, with values certified by the reference measurement procedure and acting as a secondary reference material, represents the only possible alternative for establishing traceability to the reference system. Calibration of the commercial system must be in accordance with correlation results obtained using the value-assigned reference samples (Figure 2). Benefits of this approach are: 1) trueness of the measurement result is directly traceable to the reference procedure; 2) new calibrator values can be assigned whenever the reagent formulation/lot changes or there are new calibrator batches. By using the validated calibration material that has been stored under ideal conditions, the manufacturer is able to assign values to the new batches of calibrators and reagents; and 3) if “matrix” effects are present in commercial calibrator and control materials, trueness is not affected because traceability of measurement results directly derives from the patient sample comparison with values assigned to these samples by the reference procedure.

Figure 2 Suggested practical approach for establishing traceability of enzyme routine methods if the available reference material is not commutable (symbolised by the traffic lights).

Apart from the need for commutability of reference materials, the practical implementation of a reference system in enzymology also requires that the commercial methods have similar analytical specificities toward different molecular forms of the specific enzyme when compared to the reference procedure. For example, assays for the measurement of aminotransferases that do not incorporate pyridoxal phosphate to activate the apoenzyme will produce different results to the IFCC reference procedure. AMY is another example that illustrates the different selectivity of some assays for pancreatic and salivary isoenzymes. In general, the idea of relating methods to each other through the use of the enzyme reference system does not imply there is a simple correlation between methods that employ different measurement principles.⁷

Ten years ago the IFCC created a working group to promote the interlaboratory comparability of enzyme results through the establishment of reference measurement systems in clinical enzymology.^16,17 The first objective of the group was to select suitable primary reference measurement procedures to underpin accuracy of the system and traceability of measurement results. IFCC reference procedures for the measurement of CK, LD, ALT, AST, GGT, and AMY are now available and a measurement procedure for ALP is in preparation.^18–23 The measurement conditions were described in the form of detailed standard operating procedures (SOPs), validated, and tested for transferability in a network of reference laboratories.

Under the patronage of the IFCC and IRMM, a worldwide group of laboratories was selected to provide the necessary skill and equipment for performing measurements following the above-mentioned SOPs using in-house prepared reagents. In order to achieve low levels of measurement uncertainty, careful control of all metrological aspects related to gravimetry, volumetry, and photometry was necessary. The main objective of this international group of laboratories was to demonstrate acceptable comparability of results and an ability to carry out the analytical work according to the specified analytical requirements for reference procedures. Several practice exercises have shown good agreement between individual laboratories with results within the specified narrow limits of uncertainty. As an example, Figure 3 shows the results obtained by seven of these laboratories, including laboratories from research institutes, hospitals, and manufacturers, for AST on two samples distributed in 2005 during the specific External Quality Assessment (EQA) exercise organised by the IFCC (“RELA trials”).²⁴

Figure 3 Results obtained by laboratory members of the IFCC enzyme network for aspartate aminotransferase (AST) measured in two samples (A and B) distributed during an External Quality Assessment exercise organised by IFCC (“RELA 2005”). Results (more ...)

A parallel objective for the IFCC working group was to select appropriate reference materials and to certify the catalytic activity of enzymes in these materials.²⁵ Accurate value assignment of these materials is crucial when transferring trueness from the higher–level reference procedures to routine clinical measurements. Initially, the IFCC network and IRMM cooperated to certify the reference materials for GGT, LD, ALT, and CK, previously prepared by the Community Bureau of Reference (BCR) of the EU, using the IFCC reference measurement procedures.²⁶ The results obtained from different reference laboratories agreed within very narrow limits, so that the values assigned to the four reference materials carried a very low uncertainty (from 1.39% for LD to 3.96% for CK) (Table 1).²⁷ More recently, a candidate reference material for AST has been identified.²⁸ As a result a lyophilised monoenzyme material for AST has since been prepared and a certification campaign, involving 12 laboratories from the IFCC network, is currently ongoing. The commutability of these materials remains a key issue and the main criterion required for the transfer of trueness to routine enzyme methods. So far commutability has only been shown for a restricted number of methods^29–32 and additional materials that behave in a similar manner to human samples are needed to ensure result comparability for certain methodologies, e.g. procedures using dry chemistry.³³

Table 1 Characteristics of the enzyme reference materials certified by the IFCC enzyme laboratory network in cooperation with the Institute for Reference Materials and Measurements (IRMM).

In the EU the implementation of result traceability in Laboratory Medicine to available reference measurement procedures and reference materials of a higher order is mandatory by law.³⁴ The implementation of a plan to introduce traceable assays is complicated however, and the suggested steps must be recognised as being valid by all those involved in measuring enzymes.

In a recent study involving 70 laboratories from three European countries, enzyme assays from six major manufacturers were assessed for traceability to IFCC reference systems through a commutable serum-based material targeted with ALT, AST, CK, GGT, LD, and AMY reference methods.⁷ Results from commercial methods were assessed by a system using a maximum allowable error derived from the desirable analytical performance that is based on the biological variation model.

Of these enzyme measurements, CK and ALT results were relatively good. For CK more than 95% of laboratories using Abbott, Beckman and Roche systems would be expected to show traceability within the biologically derived limits; for ALT more than 95% of laboratories using systems from Dade, Olympus, Ortho and Roche, 94% using Abbott, and 89% using Beckman would also comply. For AST and GGT only Dade and Olympus would fully comply.⁷ LD and AMY measurements have major drawbacks. LD results for Ortho Clinical Diagnostics systems in Italy and The Netherlands were approximately 2.4-times higher than the established target value. This is the result of using a method with different analytical specificity for LD and obtaining results that are not traceable to the internationally accepted reference system. By contrast, in Germany this same company apparently uses a completely different analytical approach, with results closer to the target. Such ‘‘country-based’’ approaches impede “result globalisation”. In some cases, diagnostic companies may still prefer to produce non-IFCC traceable enzyme methods, e.g. LD methods based on pyruvate to lactate, and so permit laboratories to choose between the different marketed assays, of which some are clearly not traceable to the reference system. For AMY many commercial assays are not yet traceable to the IFCC reference measurement procedure.²³ The result is a significant negative bias for Abbott, Dade, and Ortho AMY assays. The use of different substrates by these manufacturers (and the consequent differences in assay specificity) may in part explain the observed biases. Abbott and Dade assays use the chloro-p-nitrophenyl-α-D-maltotrioside (CNP-G3) as substrate, Ortho the dye amylopectine, and Roche the same substrate that is used in the IFCC reference measurement procedure, i.e. ethylidene-p-nitrophenylmaltoheptaose. As a result only laboratories using Roche assays would be expected to measure AMY within the limits of the biological variation model.⁷

Collectively, these observations suggest that a number of routine analytical systems for serum enzymes are still biased when compared to the reference methods. Despite recommendations for traceability to IFCC reference measurement procedures in clinical enzymology, not all manufacturers in the European market follow this recommendation.

Correct implementation by the manufacturers of the reference system concept, required in the pertinent ISO standard, should address most of the above-mentioned problems.¹³ It is also the responsibility of our profession to verify the accuracy and comparability of the commercially available enzyme methods by performing validation studies of the instruments and methods used in daily practice. This should be a major task for the EQA scheme organisers through the use of commutable materials with values assigned by IFCC reference measurement procedures.³⁵ Discussion with the manufacturers is needed when results are not comparable.

Overall, it appears that method bias can be reduced by better calibration to the internationally accepted reference systems. In addition, commercial assays using methodological principles that differ in analytical specificity when compared with the internationally recommended reference systems should be replaced by analytical procedures in which the traceability of calibration to the corresponding IFCC reference measurement procedure has been experimentally proven.

In conclusion, the time has come to provide comparable, standardised numerical enzyme results for all laboratories in the world. The reference system approach can provide the clinical laboratory and medical community with a universal means of creating and ensuring comparability without requiring disruptive changes to existing working methods or to an individual laboratory’s preference for an analytical system.