Relationshipof Presenceof ProgesteroneReceptorsto Prognosisin Early BreastCancer'1

the Centre. Pseudo inflammatory, bilateral, or multiple cancers were excluded from the study. The stage of the disease was classified according to the TNM (tumors-nodes-metastasis) system (19). Treatment consisted of either modified partial mastectomy (1 6) with ablation of the axillary nodes or partial mastectomy with low axillary dissection (Levels 1 and 2 of Berg), followed by postsurgical radiotherapy. The patients were examined systematically 1 month after surgery, then every 3 months over a period of 2 years, and finally once a year. Chest X-ray examinations were performed at 3-month intervals during the first 2 years. Every year, liver scmntigraphy,mammography, and thermography were done for each patient. Among the 105 patients, 9 had adjuvant chemotherapy. 5Fluorouracil (500 mg), methotrexate (15 mg), and cyclophos phamide (500 mg) were given weekly over a period of 3 months and then every 2 weeks for the following 18 months. Vincristine sulfate (1 mg) was given in the first 4 infusions and thereafter in every fourth infusion. Pathology. Tumors were graded according to the classifi cation of Bloom eta!. (1), and local recurrences were confirmed by biopsy. PGR and Estrogen Receptor Determinations. Determina tions were performed as described previously (18). Tumors were considered receptor positive (PGR positive or ER positive) if they contained 10 fmol of specific binding sites per mg cytosolic protein. Statistical Analysis. Graphs were derived from the life table analysis. Data were analyzed separately on each day. The comparison between groups was by the log rank test, and the x2testwasusedtoassess thestatistical significance ofdiffer ences(9, 10, 17).