Comparison Between Two Immunohistochemistry Systems: A Paired-Design Quality Assessment Study

Objective  To compare the staining quality between a novel domestic automatic high-throughput immunohistochemistry (IHC) system and an established imported automated IHC system.

Methods  Using a paired-design, 94 consecutive tissue sections from identical paraffin blocks were stained with identical primary antibodies (same manufacturer, clone, and batch; total 54 antibodies including Ki-67, ALK, HER2, TTF-1, CK5/6, CEA, and p53, et al.). Staining patterns covered nucleus, cytoplasm and cell membrane. The domestic system (Baisheng BMark-02 automatic immunohistochemical system) was compared against the imported system (Roche BenchMark ULTRA automatic immunohistochemical system) across seven quality parameters: overall good-to-excellent rate, staining uniformity, background clarity, staining intensity, localization accuracy, edge effect, and non-specific staining.

Results  In overall quality assessment, the overall pass rates (staining quality meeting diagnostic standards) were comparable between systems (domestic: 94.7% vs. imported: 97.9%, P = 0.25), However, three-tier quality grading (excellent/good/need-to-improve) revealed significant distribution differences (domestic: excellent 21.3%/good 73.4% vs. imported: excellent 59.6%/good 38.3%; P < 0.001). While the domestic system matched the imported system in staining intensity (97.9% for both) and absence of edge effects (89.4% vs. 96.8%; P>0.05), it showed statistically inferior performance in staining uniformity (70.2% vs. 86.2%), background clarity (61.7% vs. 78.7%), localization accuracy (70.2% vs. 92.6%), and non-specific staining (47.9% vs. 86.2%) (all P < 0.05). The receiver operating characteristic (ROC) curve analysis using imported system as gold standard domonstrated good concordance of these two systems: staining uniformity area under the curve (AUC)=0.818, 95%(confidence interval, CI) 0.693~0.943; P < 0.001; background clarity (AUC=0.797, 95%CI 0.688~0.905; P < 0.001); localization accuracy (AUC=0.879, 95%CI 0.803~0.955; P = 0.001); and non-specific staining (AUC=0.778, 95%CI 0.679~0.877; P = 0.001).

Conclusion  The domestic IHC system achieves comparable overall pass rates and matches the imported IHC system in staining intensity and absence of edge effect, demonstrating capability to deliver accurate and stable results for clinical utility for pathological diagnosis. Further optimization is warranted for staining uniformity, background clarity, localization accuracy, and non-specific staining reduction.