Comparative evaluation of six cytological grading systems in breast carcinoma

BACKGROUND: Cytological grading is a useful tool for selection of therapy and prognosis in breast carcinoma. Despite having many cytological grading systems, there is still no agreement among pathologists to accept one of them as a gold standard. AIM: This study was undertaken to evaluate six such three-tier cytological grading systems to determine which system corresponds best to histological grading done by Nottingham modification of Scarff Bloom Richardson (SBR)'s method. MATERIALS AND METHODS: In a double-blind study, preoperative cytological grades obtained by six systems on fine-needle aspiration cytology (FNAC) smears were compared by testing concordance, association and correlation with histological grade derived postoperatively by the SBR's method in 57 patients of breast carcinoma. Bivariate correlation studies and multiple linear regressions were done to assess the significance of the different cytological parameters to predict final cytological grades. RESULTS: Robinson's system demonstrated the best correlation (ρ = 0.799; P = 0.000 and τ = 0.765; P = 0.000), maximum percent agreement (77.19%) and a substantial kappa value of agreement (κ = 0.62) with the SBR's grading system. All the six cytological grading systems correlated with histological grading strongly and positively. In multiple regression analysis, all of the cytological parameters of Robinson's system except cell size and nucleoli had significance in predicting the final cytological grade. CONCLUSIONS: Robinson's grading system is simple, more objective and reproducible, and demonstrated the best concordance with histological grading. So, Robinson's system should be used routinely for breast carcinoma aspirates.