Prediction of node-negative breast cancer outcome by histologic grading and S-phase analysis by flow cytometry: an Eastern Cooperative Oncology Group Study (2192).

Histologic evaluation and reporting of invasive breast cancer has effectively used Nottingham combined histologic grade (NCHG). This approach to predict outcome in invasive breast cancer has not been tested in multicenter cooperative trials. Histologic slides from selected breast cancer cases entered on node-negative Eastern Cooperative Oncology Group trials were assigned grades. Two pathologists evaluated cases for NCHG defined from differentiation, mitotic index, and nuclear grade. The study population consisted of separate samples from low- and high-risk strata, where low risk was estrogen receptor positive with a tumor size of less than 3 cm and high risk was estrogen receptor negative or tumor size greater than or equal to 3 cm. The rate of agreement was generally good, with 80% of cases classified the same for mitotic count and 76% of the cases classified the same for combined grade. There were no cases disagreeing from the lowest to the highest of the three categories. The median follow-up is 11.6 years, but for analysis of survival, this was truncated at 5 years. Mitotic index and combined grade as assessed by both pathologists showed significant associations with survival. High combined histologic grade was predictive for response to cyclophosphamide/methotrexate/5-fluorouracil (CMF) with survival differences at 5 years of 30% in the treated high-grade patients over the untreated patients. Overall, it is clear that pathologists can have close agreement in assignment of combined histologic grades, with highly significant prediction in univariate and borderline significance in multivariate analysis in prognostication of time to recurrence as well as survival. Thus, stratification used in these trials was highly prognostic as hoped, leaving a role for histologic grading in these relatively large tumors, more powerful than S-phase analysis in this series. In the subgroups of high-risk patients randomized between CMF and observation, there was a suggestion that the high-combined-grade group was predictive of treatment efficacy. We conclude that a combined histologic grade with defined criteria may be reliably assigned by practiced pathologists using readily available criteria, and that the measure may be of use in prognostication and prediction of therapeutic responsiveness when done in a technically ideal fashion.