Background and Aims

Automated systems for anti-nuclear antibody (ANA) detection by indirect immunofluorescence analysis (IFA) provide an estimate of the fluorescence intensity (FI). The aim was to evaluate whether variability in FI estimations could be reduced by using a standard preparation.

Methods

Two standards [IUIS/ASC ANA#4 (large speckled; 1/8 dilution) and DFS Megapool] and three serum samples were distributed to 34 laboratories using automated IFA (ZenitPro;n=2, Zenit G-sight;n=5, EUROPattern;n=10, NOVAView;n=14, Image Navigator;n=3). Samples were analysed in duplicate in 5 runs. Sample1 was also distributed by the Belgian External Quality Assessment scheme (EQA). Instrument-specific FI and/or estimated titers were documented. Besides, relative fluorescence or light units (RLU) and titer ratios were calculated based on the two standards.

Results

EQA results revealed an inter-laboratory FI variability that tended to be lower for automated compared to manual IFA analysis.

With almost every automated IFA system, outlier results were observed.

Imprecision results revealed pattern-dependent coefficient of variations (CV%), with lowest CV% for DFS Megapool for all substrates.

The use of the DFS Megapool as a reference material to calculate RLUs and titer ratios reduced inter-laboratory variation (e.g. sample1: CV% without standard: 12.1-16.9% versus CV% with standard: 1.4-8.9%;p<0.05) and aligned results between instruments from the same manufacturer and between instruments from different manufacturers. Such effect was not observed when the IUIS/ASC ANA#4 was used as reference material.

Conclusions

Automated IFA analysis allows to align FI measurements, if a common protocol is applied, a total process iQC program is implemented and a suitable standard is available. The DFS Megapool presented the best reproducibility.