Histological grading and prognosis in breast cancer a study of 1409 cases of which 359 have been followed for 15 years

Abstract

Since there is great variation in the behaviour of carcinoma of the breast, even in patients with tumours of comparable clinical extent, we have attempted to recognise different degrees of malignancy in this disease from the histological appearance of the growth. The tumours of 1409 cases of breast cancer were divided into three grades of malignancy depending upon simple histological criteria and a good correlation with prognosis based upon 5, 10- and 15-year survival rates obtained. The number of survivors with tumours of low grade malignancy was between two and three times greater than those with tumours of high malignancy. A more accurate guide to prognosis can be obtained from histological data by considering the grade of the tumour with the presence or absence of metastases in the axillary lymph nodes. Thus the 5-year survival rate varied from 86 per cent for low grade cases with the axilla free to 19 per cent for high grade cases with the axilla involved. The corresponding figures at 10 years were 61 and 9 per cent, and at 15 years 49 and 7 per cent. Grading reflects the potential malignancy of the tumour and indicates which cases are more likely to have occult distant metastases at the time of treatment. However, as metastases appear to be common in all three grades of tumour when the patient first seeks advice, the real value of histological grading is in providing a guide to the speed with which such metastases become active, produce symptoms and cause death. The technique of histological grading has been described in detail and possible difficulties discussed. Although some variation in histological architecture is common in breast cancer a definite individual pattern can usually be recognised in any particular tumour, and in most instances there is little difficulty in deciding into which grade it should be placed. Furthermore, the histological appearance of metastases usually closely resembles that of the parent tumour. In a group of 397 patients the grade of malignancy in axillary lymph node metastases was identical with that of the breast tumour in 82 per cent of cases. Reference is made to the possible value of grading in determining the most suitable treatment for a particular type of case depending upon the clinical extent of the tumour and its histological type. The use of grading to assist in the evaluation of endocrine therapy is also mentioned. Brief reference is made to the question of "host resistance" as a possible factor in those patients whose clinical outcome does not conform to the predicted prognosis. Factors which may influence the histological grade of breast cancer are mentioned. So far, only the rare association of pregnancy and lactation has been found to have any effect, in which the majority of tumours are of high grade malignancy with a correspondingly poor prognosis. We wish to thank Professor R. W. Scarff for introducing us to his method of histological grading of breast cancer, Professor B. W. Windeyer for encouragement, Dr. P. Strickland for helpful criticism and Dr. J. W. Boag for some statistical advice. Our gratitude is due to the surgeons of the Middlesex Hospital and War-time Sector Units, and to Professor Windeyer and Miss Margaret Snelling of the Meyerstein Institute of Radiotherapy for allowing us to study cases under their care. It is a pleasure to acknowledge the assistance given by the pathologists of the many hospitals referring cases for post-operative irradiation, who have allowed us to study the histology of their surgical specimens. We wish to thank Mr. T. E. Cowan of the Records Department, and Miss J. Chambers of the Follow-up Department of the Hospital for tracing the patients. © 1957, The British Empire Cancer Campaign for Research. All rights reserved.