Audit of Nottingham System Grades Assigned to Breast Cancer Cases in a Teaching Hospital

Abstract

Background: Histologic grade of breast cancer is a Category I prognostic factor. Thus, accurate and reproducible grading must be ensured by periodic auditing and training. Objective: The objective of this study is to audit interobserver agreement in grading and evaluate agreement of grading on core needle biopsies with surgical pathology. Materials and Methods: Slides of 73 cases of invasive carcinomas of the breast were retrieved from archives. These were grouped into 43 cases on which consensus grades assigned by 3 pathologists were compared with initial grades assigned to the cases. The next 24 cases were graded independently by 4 pathologists and compared. In the last 6 cases, consensus grades were assigned to both the initial core needle biopsy (CNB) and the excisional biopsy. Kappa values were then calculated. Results: Kappa value (κ) for the first 43 cases was 0.50 (moderate agreement; p < 0.05). There was only fair agreement (κ = 0.25) between CNB grade and final surgical pathology grade, with 50% being upgraded from Grades 2–3. There was moderate agreement (κ = 0.53, p < 0.05) in the 24 cases on which inter‑rater agreement was tested. Specific rating of both mitotic count and nuclear pleomorphism showed fair agreement (κ = 0.25 and 0.34, respectively) while rating of tubule formation showed moderate agreement (κ = 0.57; p < 0.05). Pairwise kappa agreement ranged from fair to good (0.31–0.63). Conclusion: To ensure reproducibility, greater attention should be paid to accurately assessing nuclear pleomorphism and mitotic activity in particular. Re-grading of excisional breast biopsies after initial CNB is also essential to prevent undergrading.