Responding to the results of the ACOSOG Z1071 study, Seema A. Khan, MD, Professor of Surgery at Northwestern University Feinberg School of Medicine, Chicago, urged caution in adopting the practice of sentinel lymph node surgery after chemotherapy for some patients with breast cancer at this time.
“It is always a little hard to interpret a registry study, as there is no comparison group, but the strategy that was tested in this study reflects an important question: Do women with known nodal disease prior to neoadjuvant chemotherapy need to have their lymph nodes removed, or could they just have a sentinel biopsy?” she told The ASCO Post.
The latter approach does provoke concern, she maintained. “The reason people worry about this is that we know that response to chemotherapy is not uniform. You could have some sites that respond to treatment and others that don’t. You can have a great response in the breast and less response in the axillae. For example, a patient with several nodes invloved prior to chemotherapy may respond well in the sentinel node, but not in all nodes,” she explained.
Dr. Khan was also concerned that the 12.4% false-negative rate exceeded the study’s prespecified goal of 10%, and that the conclusions regarding the number of sentinel nodes required for accuracy were based on what appears to be an unplanned analysis.
“The factors that were associated with low false-positive rates such as clip placement and number of sentinel nodes removed were not hypothesized a priori. It was a post hoc analysis, and the upper end of the confidence interval was almost 17%. I worry about leaving disease behind in the axilla and not obtaining information that may be valuable in guiding radiotherapy plans. Although we learned from ACOSOG Z0011 that residual axillary disease is rarely followed by recurrence, all the women in that trial subsequently received chemotherapy, whereas in this study they had already had chemotherapy (though some would also receive endocrine therapy),” she said. “My overall take on ACOSOG Z1071 is that there are several important caveats that may be open to misinterpretation in the application of the data.” ■
Disclosure: Dr. Khan reported no potential conflicts of interest.