| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Department of Surgery, Division of Surgical Oncology, University of Louisville, 315 East Broadway, Suite 312, Louisville, Kentucky 40202
Correspondence: Address correspondence and reprint requests to: Anees B. Chagpar, MD, MSc; E-mail: anees.chagpar{at}nortonhealthcare.org.
Although none of us has a crystal ball, the ability to predict outcomes on the basis of clinicopathologic factors is critical in guiding treatment decisions. Such predictive models, often statistically based, are finding widespread acceptance and utility in medicine, particularly in breast cancer management. Models have been used to predict outcomes with various types of chemotherapy1 and to predict the presence of nonsentinel lymph nodes in breast cancer patients with a positive sentinel lymph node.2 In this issue of Annals of Surgical Oncology, Rivers et al.3 further this fine work by discussing the features associated with having four or more positive axillary lymph nodes in breast cancer patients with at least one positive sentinel node.
The ability to predict extensive nodal disease is important, but for some readers its value may be questioned. Why should we try to predict which patients will have four or more positive lymph nodes when the finding of a positive sentinel lymph node generally mandates an axillary node dissection (whereupon the exact number of positive lymph nodes will be known)? The authors discuss the issue that historically there has not been found to be a survival detriment associated with waiting and removing clinically relevant axillary disease at a later time4; however, it could be argued that the concept of leaving positive lymph nodes until they are clinically troublesome has not been widely accepted in clinical practice. The introduction of sentinel lymph node biopsy in breast cancer has been embraced as a staging procedure that can spare patients who are found to be node negative the morbidity of an axillary node dissection, but it has not gained favor as a treatment alone. Although it is known that in roughly a third to half of all patients, the sentinel node may be the only node harboring metastatic disease,5 there has been no reliable technique of determining when the sentinel node will be the only positive node, although statistical models have been developed to try to address this issue.2
The finding of four or more positive lymph nodes, however, has a significant effect on adjuvant therapy, particularly with recommendations of the American Society of Clinical Oncology6 and the American Society of Therapeutic Radiology and Oncology7 for postmastectomy radiotherapy (PMRT) in patients at high risk of chest wall recurrence. Therefore, to be able to predict whether PMRT will be required may affect decision-making. Some may argue, however, that the utility of such prediction models may be limited because adjuvant radiation therapy is often planned after definitive surgery, which would generally include axillary lymph node dissection in sentinel node-positive patients.
One potential application of such a prediction model would be to assist patients’ decision-making regarding their breast cancer surgery. Potentially, patients who would otherwise opt for mastectomy to try to avoid the adjuvant radiation therapy mandated with breast conservation may choose to have a partial mastectomy if PMRT is inevitable. It is therefore interesting that Rivers et al.3 found tumor size and lymphovascular invasion to be correlated with having four or more positive lymph nodes on bivariate analysis. This information alone, however, cannot distinguish those who will have four or more positive nodes from those who will not in the Rivers model. For example, even patients with tumors >2 cm with lymphovascular invasion will have a predicted probability of four or more positive nodes ranging from 5.5% to 90.8%, depending on extranodal extension and number of positive sentinel nodes. Unfortunately, this information from a sentinel node biopsy is frequently not available before surgery, because nodal evaluation is often performed at the time of the definitive operation.
Rivers’ work, however, may significantly benefit intraoperative surgical decision-making. In patients who have opted for immediate reconstruction, surgeons may decide to delay reconstruction if preoperative and intraoperative factors (including sentinel lymph node biopsy data) suggest that PMRT will be indicated. Many radiation oncologists prefer to radiate a flat chest wall rather than a reconstructed breast because the steep slopes associated with reconstruction (particularly if implant based) cause poor matching of tangential radiation fields and uneven dose distribution to the chest wall.8 In addition, many plastic surgeons would prefer to delay reconstruction if radiation is required, because radiating a reconstructed breast may result in a higher complication rate, with flap shrinkage, capsular contracture, and increased reoperation rates.9 Kronowitz et al.10 have described the technique of delayed immediate reconstruction, in which a tissue expander is placed and immediate reconstruction is delayed until the final pathology results. Being able to predict during surgery which patients will require PMRT may obviate the need to delay the immediate reconstruction in controversial cases.
There are currently no recommendations regarding PMRT for the group of patients with tumors <5 cm and one to three positive lymph nodes. The North American Intergroup trial designed to answer the question of whether postmastectomy radiation was of benefit in this group of patients closed because of poor accrual. A European study evaluating this question is being planned.11 However, the recent release of the 20-year follow-up data from the British Columbia trial demonstrating a survival benefit in all node-positive patients (including those with one to three positive nodes)12 may prompt some radiation oncologists to consider the use of PMRT in all node-positive patients, thus obviating the need for a prediction rule of four or more positive nodes. This practice, however, has not been ratified in a revision of the American Society of Clinical Oncology/American Society of Therapeutic Radiology and Oncology consensus statements.
Knowing the factors that predispose to having extensive nodal disease is an important step forward in advancing our knowledge of breast cancer and its management. The work of Rivers et al.3 in this issue of the Annals furthers our understanding of the clinicopathologic features that predict the presence of four or more positive lymph nodes on axillary dissection and may therefore predict the need for postmastectomy radiotherapy. Such work is the foundation for the generation of predictive models that can be incorporated into clinical practice. Although not a crystal ball, these models may prove to be simple, clinically relevant guides to treatment planning and improved patient decision making.
Received for publication October 20, 2005. Accepted for publication October 26, 2005.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
