Breast surgery in general, and axillary dissection in particular, have come to be regarded by many as staging procedures that are useful for maintaining local control but do not affect survival. This is a far cry from the Halstedian concept of meticulous radical surgery for the cure of breast cancer that governed surgical thinking in the United States until the 1970s. Acceptance of axillary surgery as a staging, rather than a therapeutic, modality resulted from clinical observations that most women with axillary node metastases treated with surgery alone die of breast cancer, and the demonstration that adjuvant systemic chemotherapy improves survival in women with node-positive breast cancer. The most direct articulation of this new role for surgery came from Bernard Fisher, who stated “breast cancer is a systemic disease involving a complex spectrum of host-tumor interactions and variations in effective local treatment are unlikely to affect survival substantially.” 1 This “systemic disease” hypothesis dominated our thinking about breast cancer management from the 1970s to the 1990s, with a number of beneficial results. Most notable among these were the development of breast-conserving therapy and the recognition that node-negative breast cancer did not always imply a good prognosis for the disease. However, a number of recent clinical observations have led to a resurgence of interest in the potential therapeutic role of aggressive local therapy of breast cancer. As one of these reanalyses, Orr presents a metaanalysis of the impact of axillary dissection on survival in this issue of the Annals of Surgical Oncology. 2 The technique of meta-analysis has become extremely popular, but when evaluating the results of such analyses, it is important to remember that their validity is determined by the quality of the studies included in the analysis and their relevance to current practice. In this regard, there are several important points to consider about the studies cited in Orr’s analysis. Of the 2936 patients, approximately 21% were from the Guy’s I and II trials. These studies compared patients treated by radical mastectomy to those treated by wide local excision and an inadequate dose of breast irradiation. 3 A systematic plan of surgical salvage was not followed for patients who relapsed in the breast or the axilla. If one presumes that untreated metastases in the axilla can serve as the source of additional metastases and result in decreased survival, it is equally biologically plausible that uncontrolled local failure in the breast can act in the same fashion. Thus, it is difficult to conclude much more from these two trials than that inadequate local therapy with inadequate salvage therapy has the potential to decrease survival. This general concept, that inadequate local therapy does have an impact on survival, is consistent with the idea that axillary dissection could affect survival, but it is not proof. The patients for whom axillary dissection often is considered least beneficial are those with small cancers at lowest risk for axillary node metastases, usually T1 lesions, particularly those 1.0 cm or less in size. It is difficult to imagine how the removal of histologically normal nodes would alter survival. For this reason, the authors’ analysis of Stage I patients is of particular interest. However, it is incorrectly stated that the B04 and Guy’s II trials included only Stage I patients. Both studies included T2 and T3a tumors in their eligibility criteria, 3, 4 and these larger tumors would be expected to have a higher risk of nodal metastases than would T1 lesions. The Stage I analysis appears to be an analysis of clinically node-negative breast cancers, and no conclusions can be drawn from it regarding …