Pathologic correlates of survival in 378 lymph node-negative infiltrating ductal breast carcinomas. Mitotic count is the best single predictor
✍ Scribed by Frederic Clayton
- Publisher
- John Wiley and Sons
- Year
- 1991
- Tongue
- English
- Weight
- 815 KB
- Volume
- 68
- Category
- Article
- ISSN
- 0008-543X
No coin nor oath required. For personal study only.
✦ Synopsis
Pathologic features of 378 breast carcinomas without axillary lymph node metastases were correlated with long-term prognosis. Mitotic count, the feature best predicting prognosis, was most useful at lower mitotic count levels than usually studied. Cases with less than 4.5 mitotic figures per ten high-power fields had a 20-year tumor-related survival of 73% versus 48% for those with higher mitotic counts. Low and high mitotic count groups had significantly different prognoses for Tlc, T2, and T3 tumors. Multivariate analysis showed that the best combination of features predicting tumor-related death were high mitotic count (relative risk, X2.81, large tumor size (relative risk, X1.7), lymphatic invasion (relative risk, X1.7), and skin or subjacent muscle or chest wall invasion (relative risk, X2.5). Nuclear grade, Bloom-Richardson grade, modified Scarff-Bloom-Richardson grade, Fisher's grade, and necrosis correlated well with survival but had no significant additional predictive value when adjusted for the mitotic count. Cancer 68:1309-1317,1991.
ECENT STUDIES concluded that adjuvant chemo-R therapy is useful in breast carcinomas without axillary lymph node metastases.'-4 Serious consideration of chemotherapy has been recommended for patients with node-negative breast carcinomas, particularly if an unfavorable p r o g n ~s i s ~. ~ is suspected. Therefore the prognosis must be determined accurately.
Many histologic features were proposed as prognostic correlates for breast carcinomas, but their relative prognostic importance is unclear. Tumors of different histologic types and stages were studied together, limiting the value of the analysis in any specific situation. The reproducibility of some histologic features and subjective grading schemes was widely variable and often p 0 0 r , ~-~ ranging