The importance of surgery and accurate axillary staging for survival in breast cancer

Aims. The purpose of this study was to investigate, within the context of t he Danish Breast Cancer Cooperative Group (DBCG) programmes, whether a dedi cated surgical approach had a significant bearing on the outcome of breast cancer treatment. Methods. From 1 January 1980 to 31 December 1990, patients below 70 years o f age with operable breast cancer from Odense University Hospital (n = 743) were compared with those from the rest of Denmark (denoted rest-DR) (n = 1 5,419). All patients were treated according to nationwide DBCG guidelines a nd reported to the DBCG Data Centre. The potential median observation time was 11.2 years (range 6.0-16.9). Patients underwent mastectomy or breast co nserving therapy, and high risk lymph-node positive patients had adjuvant s ystemic therapy with or without radiotherapy. Results. Comparing total patients series, overall survival (OS) was signifi cantly superior in patients from Odense compared with rest-DR (P = 0.02), w ith 10-year OSs of 62% (95% CI: 58-65%) and 56% (55-57%), respectively. In subgroups, the OS of low-risk node negative patients (protocol A) in Odense compared with rest-DR was significantly better (P = 0.02); 10-year OS was 78% (73-84%) versus 72% (70-73%). Among the high-risk pre-menopausal patien ts (protocol B), the OS was significantly better in Odense (P = 0.009); 10- year OS was 67% (60-75%,) versus 53% (51-55%) in rest-DR. Post-menopausal h igh-risk patients (protocol C) did not differ significantly in OS between O dense and rest-DR (P = 0.61). Locoregional control in the Odense series was superior compared with rest-D R. More lymph nodes were recovered and examined from the axilla in the Oden se series than in rest-DR, a median of 10 vs. 6 nodes. In the Odense series , a significantly higher proportion of pre-menopausal patients had positive lymph nodes, predominantly one to three positive nodes, and subsequently a lower proportion of pre-menopausal patients had negative lymph nodes compa red with rest-DR (P = 0.02). indicating a more accurate staging in Odense v s. rest-DR. The survival benefit among the patients from Odense cannot be explained by stage migration alone, but seems to represent a true survival advantage. Ov erall mortality was significantly lower in the Odense series compared with rest-DK. Whether or not this difference could be explained by lower backgro und mortality in the Odense series or was caused by superior treatment is d iscussed. Conclusions. The extent of surgery seems important for locoregional tumour control and acurate axillary lymphnode staging. In combination, these might lead to superior recurrence-free and overall survival, although difference s in background mortality cannot be ignored. Surgery; therefore, might repr esent a risk factor by itself.