Validation of a nuclear grading system for resected stage I–IIIA, high-risk, node-negative invasive breast carcinoma in the N·SAS-BC 01 trial

Abstract

Background: This retrospective observational study validated nuclear grading criteria developed to identify a high-risk group with recurrence rate ≥20–30% and local pathology diagnosis used in a previous multi-institutional randomized N·SAS-BC 01 trial, where the efficacy of adjuvant chemotherapy regimens was evaluated in 733 high-risk node-negative invasive breast cancer patients. Methods: Of 545 patients with long-term follow-up data (median 12.1 years), pathology slides, and local pathology diagnosis, 530 eligible patients were subjected to central pathology review (CPR) for histological type and nuclear grade (NG). Concordance in NGs was compared with local diagnosis. The 10/15-year recurrence-free survival (RFS) and overall survival (OS) rates stratified by NG and histological type were calculated. Results: Local diagnoses were invasive ductal carcinoma (IDC)-NG2, IDC-NG3, invasive lobular carcinoma (ILC), and metaplastic carcinoma (MC) in 158/327/38/7 patients, respectively. The 10/15-year RFS rates were 87.2/82.6% for IDC-NG2 and 81.8/75.0% for IDC-NG3 (p = 0.061), and OS rates were 95.0/92.8% for IDC-NG2 and 90.8/85.7% for IDC-NG3 (p = 0.042). CPR graded 485 locally diagnosed IDCs as IDC-NG1/NG2/NG3/unknown in 98/116/267/4 patients, respectively. No significant difference was found among survival curves for the three NG groups. Although the agreement level between local and CPR diagnoses was low (κ = 0.311), both diagnoses identified a patient group with a 15-year recurrence rate ≥ 20%. The 10/15-year RFS rates were 79.4/63.5% for ILC and 68.6%/unknown for MC. Conclusions: The N·SAS grading system identified a patient group with high-risk node-negative invasive breast cancer, suggesting that local diagnosis was performed efficiently in the N·SAS-BC 01 trial. Trial registration number: UMIN000022571. Date of registration: June 1, 2016.