A at the present time we have 2465 cases of mammary carcinoma and we are in process of reviewing the data from them from a somewhat different point of view than has been done in many clinical studies.
We have taken the hypothesis that many of the clinical staging systems are empirical and that they have come into being by the empirical grouping of different clinical and pathological variables. T h e variables may or may not be related.
In this study 704 consecutive radical mastectomies performed between 1940 and 1958 were reviewed. We have more than 600 clinical, pathological and therapeutic variables quantitated on these cases and we are in process of analyzing the impact of these individual variables on the natural history of mammary cancer.
T h e staff of the cancer hospital in the past rarely has restricted its criteria for operability to those of Haagensen's and has considered that most patients with mammary cancers restricted to the chest wall and the axilla without evidence of remote metastases are candidates for radical mastectomy. This has resulted in the performance of radical mastectomy for some rather far-advanced lesions throughout the years.
Fig. 1 is a plot of the time of recurrence after radical mastectomy and the time of appearance of distant mestastases-almost a straight line plot. At present, 146 cases have developed locally recurrent cancer and only four of them have not manifested distant disease. This log scale is used frequently throughout our analysis.
Growth curves of pulmonary metastasis