Development and commutability assessment of serum-based candidate reference materials for homocysteine and their application in external quality assessment.

Background

Elevated serum homocysteine (HCY) is associated with cardiovascular and other systemic risks. This study aimed to develop serum-based secondary candidate reference materials (Can-RMs) for metrological traceability and validating their dual utility as commutability-verified materials for external quality assessment (EQA).

Methods

First, 41 fresh serum samples were analyzed by isotope dilution liquid chromatography-tandem mass spectrometry reference method and four routine detection systems to evaluate baseline comparability. Second, the commutability of Can-RMs, EQA materials spiked with purified reference standards (Sta-EQA materials), and in-house prepared EQA materials with serum matrix (Ser-EQA materials) was assessed via Deming regression and bias analysis. Third, the commutable Can-RMs were deployed in a 62-laboratory EQA program; performance (CV, bias, total error) were evaluated per biological variation-based minimum analytical quality specifications (6.23%, 12.94%, and 23.21%).

Results

Routine systems exhibited positive proportional biases (0.5%-13.1%) versus the reference method. The Can-RMs (L1-L5) and Ser-EQA materials were commutable across all systems (with only Can-RM L1 for the Leadman system was inconclusive under IFCC criteria). All Sta-EQA materials were non-commutable under both approaches. In the EQA program, the average laboratory pass rates for TE, bias, and CV were 91%, 87%, and 98%, respectively, with the Roche and Strong Biotechnologies systems achieving 100% pass rate.

Conclusions

In-house prepared serum-based Can-RMs and Ser-EQA materials exhibit excellent commutability, supporting reliable standardization assessment and accurate trueness validation for clinical procedures. A major limitation is that all reagents were tested on a single Roche Cobas c701 analyzer; future multi-instrument trials are required to verify generalizability.