Alanine aminotransferase (ALT) accuracy determination is a critical performance evaluation that verifies how closely the results of an ALT measurement system agree with assigned reference values, ensuring reliable liver function testing. This technical article explains the analytical principles, sample requirements, testing items, evaluation methods, and practical applications of ALT accuracy assessment. It is intended for IVD manufacturers, clinical laboratory professionals, and quality managers who must demonstrate metrological traceability and regulatory compliance.
Technical Overview
Alanine aminotransferase (ALT, EC 2.6.1.2), formerly known as serum glutamic pyruvic transaminase (SGPT), is a cytosolic enzyme found at high concentrations in hepatocytes and is one of the most widely used biomarkers of hepatocellular injury. In quantitative clinical chemistry, accuracy is defined as the closeness of agreement between measured values and true values, and it combines two components: trueness (absence of systematic bias) and precision. Accuracy determination for an ALT assay therefore consists of an experimentally based evaluation of the bias between the routine measurement procedure under test and an accepted reference, typically the IFCC reference measurement procedure at 37 degrees Celsius or a certified reference material with an assigned value.
The ALT reaction couples transamination with a pyridine nucleotide indicator reaction. ALT catalyzes the conversion of L-alanine and 2-oxoglutarate to pyruvate and L-glutamate; lactate dehydrogenase then reduces pyruvate to lactate while oxidizing NADH to NAD+. The rate of decrease in NADH absorbance, monitored spectrophotometrically at the main wavelength of 340 nm (molar absorptivity approximately 6.3 L·mmol⁻¹·cm⁻¹), is directly proportional to the catalytic activity concentration of ALT expressed in U/L. Because ALT results guide clinical decisions at relatively low decision limits, even modest systematic bias can cause misclassification of patients, which is why documented accuracy is a mandatory performance characteristic under ISO 15189, the EU IVDR 2017/746, and equivalent regulatory frameworks.
The metrological traceability chain for ALT is established through ISO 17511 and ISO 18153: manufacturer calibrators are assigned values by the IFCC reference measurement procedure, and routine systems are calibrated against those materials. Accuracy determination verifies that this chain functions correctly at the level of the end user's results.
Testing Samples
Human Serum Samples
Fresh human serum is the primary matrix for method-comparison-based accuracy studies. Specimens are collected by standard venipuncture into plain or separator tubes, allowed to clot, and centrifuged under controlled conditions. Samples must be free of hemolysis, icterus, and lipemia where possible, since these interfere with the photometric measurement, and the activity range of the collected set should span the measuring interval of the assay, from low normal to markedly elevated concentrations.
Human Plasma Samples
Lithium heparin plasma is acceptable for systems validated for plasma. Citrate and EDTA plasma are generally unsuitable because anticoagulant dilution and cation chelation alter the measured activity concentration. When both matrices are claimed, accuracy must be verified separately for each sample type.
Certified Reference Materials and Calibrators
Certified reference materials (CRMs) and manufacturer calibrators with values assigned by the IFCC reference measurement procedure are used for direct bias assessment. These materials should be commutable, meaning they behave like native patient samples in the routine system, and the assigned value must be accompanied by its measurement uncertainty.
Quality Control and Proficiency Testing Materials
Frozen or lyophilized quality control materials and external quality assessment (proficiency testing) specimens provide an ongoing, longitudinal check of accuracy. They are measured repeatedly over time to detect drift, lot-to-lot reagent differences, and calibration instability after the initial validation study.
Testing Items
An ALT accuracy determination protocol typically evaluates the following items:
- ALT catalytic activity concentration (U/L) across the medically relevant concentration range, including values near decision limits of approximately 40 to 50 U/L.
- Absolute bias and percentage bias between the routine method and the comparative reference values at defined medical decision levels.
- Regression parameters from method comparison, including slope, intercept, correlation coefficient, and their 95% confidence intervals.
- Recovery rates from spiked or gravimetrically prepared samples, typically targeting 95% to 105%.
- Total error of the measurement system, compared against the allowable total error (TEa), which for ALT is commonly specified in the range of ±15% to ±20% depending on the applicable biological variation or regulatory quality goals.
- Verification of bias from certified reference materials, expressed relative to the certified value and its expanded uncertainty.
Testing Methods
Accuracy determination follows a structured, statistics-driven approach based on internationally recognized standards, principally CLSI EP09 (measurement procedure comparison and bias estimation using patient samples), CLSI EP15 (user verification of precision and estimation of bias), and ISO 18153 (metrological traceability of catalytic concentration values for enzymes).
Reference Method Comparison
A minimum of 40 fresh patient specimens, and preferably 100 or more, covering the full measuring interval are measured by both the routine ALT system and the IFCC reference measurement procedure. Duplicate measurements are performed over at least three to five separate days to distribute matrix and calibration effects. The IFCC procedure specifies defined reagent composition, including 500 mmol/L L-alanine and 15 mmol/L 2-oxoglutarate in Tris buffer at pH 7.15 and 37 degrees Celsius, a pre-incubation step to consume endogenous pyruvate, and calculation of activity from the NADH decrease at 340 nm.
Statistical Analysis
Paired results are analyzed using regression techniques appropriate to the error structure, such as Deming or Passing-Bablok regression, supported by Bland-Altman difference plots. Bias and its confidence interval are estimated at each medical decision level and compared with predefined acceptance criteria. If the upper confidence limit of bias remains within the allowable limits, accuracy is accepted.
Recovery Exstudy and Reference Material Verification
In recovery experiments, patient pools are supplemented with known amounts of purified ALT and the measured increment is compared with the expected increment. In parallel, certified reference materials are measured in replicate on the routine system, and the mean result is evaluated against the certified value plus or minus its expanded uncertainty. Both approaches provide an independent estimate of trueness that complements the patient sample comparison.
Acceptance Criteria
Typical acceptance criteria include percentage bias within ±10% at clinically relevant levels, recovery between 95% and 105%, and demonstrated total error (bias plus two standard deviations) not exceeding the specified allowable total error of ±15% to ±20%. Criteria must be defined prospectively in the validation protocol and justified against clinical requirements and biological variation data.
Applications
ALT accuracy determination supports multiple stakeholders across the in vitro diagnostics and clinical laboratory value chain:
- IVD manufacturers performing product development, design verification, regulatory submissions under the EU IVDR, and lot release testing of ALT reagent kits and calibrators.
- Clinical laboratories verifying manufacturer accuracy claims before implementing new ALT assays, as required by ISO 15189 and clinical laboratory accreditation programs.
- External quality assessment and proficiency testing providers evaluating inter-laboratory agreement and detecting systematic errors.
- Reference measurement laboratories assigning target values to calibrators and control materials through implementation of the IFCC reference procedure.
- Post-market surveillance activities, including comparability studies after reagent reformulation, analyzer replacement, or calibration lot changes.
- Epidemiological and pharmaceutical research, where harmonized ALT values across sites and study periods are essential for data pooling and safety monitoring.
FAQ
Q: What is the difference between accuracy and precision in ALT testing?
A: Accuracy describes the closeness of a measured ALT value to the true value and combines trueness with precision. Precision reflects reproducibility under repeat conditions, while accuracy determination specifically quantifies systematic bias against a reference measurement procedure or certified reference material.
Q: Which reference method is used as the comparator?
A: The IFCC reference measurement procedure for ALT at 37 degrees Celsius is the internationally recognized highest-level comparator. It is an optimized UV kinetic method based on NADH oxidation measured at 340 nm and provides the anchor for metrological traceability under ISO 18153.
Q: How many patient samples are required for a valid accuracy study?
A: CLSI EP09 recommends a minimum of 40 specimens, with 100 or more providing stronger statistical power. The samples must cover the analytical measuring interval and include concentrations near clinical decision limits.
Q: What are the typical acceptance criteria?
A: Common criteria are percentage bias within ±10% at medical decision levels, recovery of 95% to 105% for spiked samples, and total error not exceeding an allowable total error generally set between ±15% and ±20% for ALT.
Q: Can hemolyzed or lipemic samples be used?
A: No. Hemolysis releases intracellular enzymes and adds optical interference, while lipemia and marked icterus disturb the photometric measurement. Such specimens should be excluded or flagged according to the manufacturer's interference specifications.
Q: How often should accuracy be re-verified?
A: Accuracy should be verified at initial method implementation, after any major change such as a new reagent or calibrator lot from a different formulation, after instrument service, and at intervals defined by the laboratory's quality management system, supplemented by routine quality control monitoring.
