Breast preservation versus mastectomy in early breast cancer - 1991 update of the GBSG 1 - Protocol and prognostic facors

In 1983, the German Breast Cancer Study Group (GBSG), sponsored by the Federal Ministry of Research and Technology, started a prospective multicenter trial on the treatment of early breast cancer (pT1 pN0 M0). This was preceded by a three year reviewing period because of some novelties of medical, judicial and ethical problems in the FRG. University and, in the majority, community hospitals participated, combining all together 69 different institutions. From 11/1983 to 12/1989, 1112 patients were recruited. From 1036 patients, 733 underwent breast preservation (71%) and 303 mastectomy (29%). The randomization rate was only 6%. In 268 patients (26%) the tumor size was ≤10 mm, in 765 patients (74%) 11 to 22 mm. In 129 cases, we subdivided the tumor grading II [3] into IIa and IIb. Moreover, the immunohistochemical detection of the transmembrane proteins EGFR, p-185 and p-148 by oncogene overexpression and c-myc oncogene were undertaken in 425 breast cancers. After tumorectomy (or wide excision) and a lower axillary dissection (at least eight lymph nodes) the breast was irradiated up to 50 Gy in 25 fractions. A boost of 12 Gy was given to the tumor bed. The medial located lymph nodes were also irradiated in case of medially or centrally tumors. Quality control was performed by pathological, radiotherapeutic and methodical reference centers. Significant correlations could be demonstrated between receptor status and tumor grading, patient age and grading, and tumor size and grading. The results emphasize the central role of tumor grading among the prognostic factors. Especially the differentiation of the Bloom and Richardson score II into IIa and IIb seems to play an important role. After a median follow-up of 41 months, the frequency of local recurrences (4.4%), regional recurrences (1%) and distant metastases (4.6%) was exactly the same in both treatment groups. In multivariate analysis, only tumor size and tumor grading had a significant impact on disease-free survival. 23 patients with tumor-involved margins had a higher recurrence rate (DFS 62% versus 85% after five years). Without any impact on DFS were the other conventionally evaluated prognostic factors: age, menopausal status, hormone receptor status, histological tumor type, tumor localisation, degree of differentiation, pleomorphism, mitotic index and degree of dissociation. Among the transmembrane proteins EGFR, p-185, p-148 and c-myc, only the impact of p-185 and EGRF positivity on DSF is significant. The p-185-expression rate was depending on tumor grade (4% in GI- versus 15% in GIII-tumors). In young patients, the p-185-expression was three times higher than in older ones. 5% of the cases with locoregional recurrences and 25% of the tumors with systemic failures were p-185-positive. A partial correlation existed between p-185 positive cases and EGFR expression (p = 0.05). These findings strongly suggest the prognostic significance of c-erbB1 and c-erbB2 oncogene overexpression.