Evaluation of key analytical performance metrics for reliable lymphocyte subset enumeration using full-spectrum flow...

Full-spectrum flow cytometry (FS-FCM) is increasingly used in clinical laboratories, but practical analytical performance evaluation protocols tailored to FS-FCM-based lymphocyte subset enumeration remain insufficiently defined. To address this gap, this study aimed to evaluate a practical framework proposed for assessing the analytical performance of FS-FCM in TBNK lymphocyte subset enumeration, including both percentages and absolute counts.

Performance evaluation was conducted in accordance with relevant clinical guidelines and industry standards, including CLSI EP15 and EP06. Two protocols for precision and bias verification were evaluated: a "3 × 5" protocol with three replicates per day over five days and a "3 × 10" protocol with three replicates per day over ten days. The performance of the two protocols was compared using analytical performance specifications derived from biological variation (BV). Bias was estimated using two levels of reference materials with assigned values and associated expanded uncertainties provided by the National Center for Clinical Laboratories of China. Single-result agreement was assessed against predefined total error criteria. Linearity was evaluated using weighted regression with segmented assessment across the analytical range. Fluorescence detection threshold and fluorescence parameter linearity were assessed using an approach adapted to the optical characteristics of FS-FCM.

FS-FCM showed acceptable precision. Most coefficients of variation (CVs) were < 5%, whereas CVs for low-abundance subsets were generally <7%. Biases were < 10% for most parameters, although some B-cell and NK-cell measurements in normal-level samples showed biases of 10%-15%. Bias estimates met the minimum BV-derived allowable bias criteria, and most met the desirable criteria, all within the verification intervals. The shorter "3 × 5" protocol yielded performance estimates similar to those obtained with the "3 × 10" protocol and satisfied the predefined BV-derived allowable criteria, indicating that it may serve as an acceptable streamlined alternative. Single-result agreement and segmented linearity assessed by weighted regression also met the predefined criteria. Using the modified approach, fluorescence detection thresholds were < 10 MESF, and the fluorescence parameters showed good linearity (r ≥ 0.9975), meeting the applicable evaluation requirements.

Under the evaluated conditions, FS-FCM demonstrated acceptable analytical performance for the enumeration of the TBNK lymphocyte subset. The proposed evaluation protocols may provide a practical approach for routine local verification and may support future multicenter harmonization of FS-FCM assays.