Documenting and validating metrological traceability of serum alanine aminotransferase measurements: a priority for medical laboratory community for providing high quality service in hepatology

Testing alanine aminotransferase (ALT) is of major clinical interest being the more liver-specific aminotransferase. It represents the first-level test to detect individuals with hepatocellular damage of any etiology [1]. Recently, a commentary published on Gastroenterology has rekindled the debate about the importance of considering the lack of assay harmonization in interpreting ALT values in hepatology and the risk of using guideline-recommended fixed cut-offs for deciding further patient follow-up [2]. Some years ago, together with other colleagues, I already cautioned the guideline recipients by explaining why the implementation of these recommendations may lead to overdiagnosis and unnecessary further testing and highlighting the need for laboratory expertise when drafting clinical guidelines involving the use of laboratory tests [3]. A series of actions are however urgently expected and no longer postponeable by laboratory community itself to definitively solve the issue of ALT (dis)harmonization.

The ISO 17511:2020 standard describes six model calibration hierarchies for metrological traceability implementation, unanimously considered as a unique tool to improve standardization in laboratory medicine [4]. Specifically, an enzyme measurand such as ALT cannot be described only by kind of quantity, name of enzyme and of system, but also requires the specified measurement procedure and the indicator component of the measured reaction, as a property of the enzyme is measured, i.e., its catalytic concentration. Accordingly, the definition of ALT catalytic concentration (activity) is the rate of conversion of NADH in the 2002 IFCC reference measurement procedure (RMP) for ALT, which defines the reaction conditions [5]. The IFCC RMP for ALT describes in detail aspects related to gravimetry, volumetry, pH, reaction temperature, and photometry that must be controlled in reference laboratories performing it to achieve a level of measurement uncertainty (MU) as lowest as possible [6, 7]. According to ISO 17511:2020, to transfer trueness from higher-order references to commercial calibrators for ALT, in vitro diagnostic (IVD) manufacturers have theoretically two possibilities: (a) directly aligning the IVD medical device (IVD-MD) to the IFCC RMP by a comparison study, or (b) calibrating their internal measurement procedure for IVD-MD calibrator value assignment with a suitable reference material (RM) [4]. The first approach asks for the use of an appropriate panel of native clinical samples, whose values are assigned by the RMP and resulting MU based on the inherent MU characteristics of the standard operating procedure implementing the IFCC RMP in the reference laboratory performing it and from the specific value-assignment protocol employed. IVD-MD should be calibrated in accordance with correlation results obtained using the RMP-assigned values of clinical samples, which therefore act as RM.

Use of higher-order RM represents a possible alternative for assigning traceable values to commercial calibrators. RM must be carefully characterized based on their intended use and compliance with ISO 15194:2009 standard, which lists the characteristics that should be documented for an RM [8]. To use the RM as common calibrator in a calibration hierarchy, its certificate must provide information about the commutability with clinical samples for all commercial measurement procedures with which it may be used as higher-order calibrator for implementing metrological traceability [9]. Unfortunately, at present there are no RM available fully complying with these requirements. The Joint Committee for Traceability in Laboratory Medicine (JCTLM) database is listing one RM for ALT (ERM^®-AD454k/IFCC, European Commission Joint Research Centre), but no definitive information about its commutability is provided as the intended use of the material is limited to control the performance of the IFCC RMP [10]. Although a preliminary commutability study was carried out on a trial batch when the raw starting material was selected, results were correctly considered not enough by the provider to fulfill commutability requirements of ISO 15194:2009 [11]. Therefore, the lack of the robust information about the commutability of this RM has the potential to become a serious problem in order to adopt the strategy proposed above as ISO 17511:2020 option (b), as recently showed also for other enzymes [12]. If this RM would be used to calibrate a commercial IVD-MD, an extended commutability study should be preliminarily performed with this procedure vs. the IFCC RMP to show the RM suitability for this use [13]. Being the RM already released, the responsibility of its commutability assessment is difficult to be assigned [14]. It seems sensible, however, that interested IVD manufacturers together with international standardization bodies urgently prioritize a study assessing the commutability of ERM^®-AD454k/IFCC for using it in implementing ALT IVD-MD traceability. On the other hand, as a JCTLM database gap analysis has highlighted [9], the supply of other suitable ALT RM appears urgently needed.

To ensure an unbroken chain of relationships and enable reporting of measured values that are traceable to the IFCC RMP, problems associated with differences in the measured quantity among the different measurement procedures at the various levels in the calibration hierarchy should be addressed and/or prevented [4]. The concept of a reference measurement system is indeed valid only if the RMP and corresponding IVD-MD have identical, or at least very similar, selectivity for the measured enzyme [15]. Thus, it will not be possible to calibrate IVD-MDs for ALT that do not incorporate pyridoxal-5-phosphate (P-5′-P) using a procedure that does, such as the IFCC RMP, because no constant proportion of pre-formed holoenzyme to reactivatable apoenzyme can be established in patients’ specimens, the ratio of pre-formed holoenzyme to apoenzyme being different among specimens [1]. This is the primary reason why ALT IVD-MDs without P-5′-P activation are often unable to fulfil quality specifications when ALT results are compared to RMP [3]. Unfortunately, almost all IVD manufacturers still market IVD-MDs with or without the addition of P-5′-P and declare that both are traceable to the IFCC RMP. Laboratorians have therefore an active role in definitively replacing IVD-MD not incorporating P-5′-P and supporting the choice of ALT selective assays which permit to report accurate enzyme values in all the pertinent clinical situations. The often raised issue of increased reagent stability (if any) when P-5′-P is not added is a false problem. The choice of analytical methods that are adequate for the clinical application of the measurements should indeed represent the only driving force in making IVD-MD choice by medical laboratories.

As mentioned above, when the IFCC RMP is used to transfer trueness to the IVD-MD calibrator’s value though a comparison study using a unique set of reference samples, IVD-MD should be calibrated in accordance with correlation results obtained using the RMP-assigned values of clinical samples. In case a measurement bias is detected, the manufacturer should decide if the observed bias is significant. An estimate of measurement bias is considered statistically significant if: x _ref − x _IVD-MD>2 u_bias, where x _ref is the ALT value by RMP for a particular human sample, x _IVD-MD is the mean value of n replicates for ALT in the same human sample with the IVD-MD, and u_bias is equal to √(u_ref ² + SD_mean ²). Note that u_ref is the uncertainty associated to the clinical sample provided by the reference laboratory making the RMP determination, and SD_mean is the standard deviation of the mean value of the clinical sample obtained by IVD-MD, calculated by the equation SD/√n, where SD is the standard deviation of the replicate measurement of the sample and n is the number of replicates. It appears sensible to leave the observed calibrator value uncorrected for nonsignificant bias and to correct the observed value for significant bias.

In 2014, the European Federation of Clinical Chemistry and Laboratory Medicine (EFLM) Strategic Conference defined three models for establishing analytical performance specifications (APS) for laboratory tests: model 1, based on the effect of analytical performance on the clinical outcome; model 2, based on components of biological variation (BV) of the measurands; and model 3, based on the state of the art of the measurement [16]. Allocation of ALT to one of these models to derive APS cannot be however straightforward. Although ALT has a clinical role in detection of liver disease, outcome-based data are currently not available to enable APS setting for ALT using model 1. In this situation, as the enzyme demonstrated rather stable catalytic concentrations in healthy individuals, the model based on BV of measurand can be temporarily used to derive APS [17]. Accordingly, the recommended maximum allowable standard MU for serum ALT measurements on clinical samples are 4.65 % (desirable quality) and 6.98 % (minimum quality level), respectively [9, 17, 18]. Commercial calibrators should display MU representing a fraction of this APS allowed for laboratory results to leave enough MU budget usable by individual laboratories to produce clinical results of acceptable quality [19]. Overall, the information that IVD manufacturers should make explicit about the implementation of metrological traceability of their commercial IVD-MD for measuring ALT are the following: (a) an indication of higher-order references used to assign values to calibrators; (b) which calibration hierarchy has been applied, with a detailed description of each step; (c) the combined MU of commercial calibrators, and (d) which acceptable limits for MU of calibrators were applied in the IVD-MD validation [4, 20].

In agreement with ISO 17511:2020, strategies applicable to validation of a calibration traceability claims for an IVD-MD for serum ALT measurements may include: (a) examination of commutable RM intended for use as trueness control material; (b) participation in suitable External Quality Assessment Schemes (EQAS); (c) method comparison study on a set of human serum samples with the IFCC RMP [4]. However, option (a) remains so far only theoretical as an international certified ALT RM with commutability assessed to be used as trueness control material is currently not available. Some surrogate materials, such as pooled serum samples value-assigned using the IFCC RMP, have been used locally to test the quality of ALT measurements, but at the global level there is no possibility to pursue this strategy [21]. Regarding option (b), only value assignment by IFCC RMP to EQAS commutable materials allows an objective evaluation of the performance of ALT measurements through a trueness-based grading of the competency of participating laboratories and of the employed IVD-MD [22]. To this scope, fresh-frozen (−80 °C) off-the-clot pooled sera can be prepared and distributed in dry ice to laboratories participating to EQAS, with ALT target values (and corresponding MU) assigned with the IFCC RMP by a reference laboratory. At a local level, to validate the trueness and to estimate if the MU of ALT measurements using a certain IVD-MD fulfills established APS, a method comparison study with the IFCC RMP can be also performed [21]. Although the availability of RMP logistics is at present still difficult, as reference laboratories performing RMP are rare and the corresponding measurements relatively expensive, a sustainable approach involving professional bodies and EQAS providers (possibly under a consortium structure) as well as a reevaluation of the costs that should be sustained not only by medical laboratories but also by IVD manufacturers is needed to make this approach more feasible. A paradigm shift in the thinking of laboratory professionals and IVD manufacturers is expected to place alongside the traditional strategy of commercial competition and costs the provision of high-quality service in hepatology [20, 22]. The value of a laboratory test must be evaluated according to its influence on medical care, and the cost aspects must be considered in the wider overall context of the impact on the health outcomes and not within the narrow focus of pure laboratory economics.

In conclusion, it is time that all medical laboratory stakeholders (laboratorians, IVD manufacturers, EQAS organizers, RM providers) actively co-operate to implement the ALT standardization in a concerted action following well defined tasks as described in this paper and summarized in Table 1.