| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Department of Surgery, University of Florida College of Medicine, Gainesville, Florida.
Correspondence: Address correspondence to: Edward M. Copeland, III, MD, Department of Surgery, University of Florida College of Medicine, P.O. Box 100286, Gainesville, FL 32610-0286; Fax: 352-338-9809; E-mail: copelem{at}mail.surgery.ufl.edu
In this issue of the Annals of Surgical Oncology, Skinner et al.1 examine the usual multiple variables for survival of breast cancer patients such as age, socioeconomic status, tumor size, nodal status, race, and so on and find the expected correlates. The "meat" of the study, however, is in the observation that patients treated by surgical oncologist (SO) had a significantly lower risk of dying compared with those treated by a nonsurgical oncologist (non-SO) when all patients at all hospitals were considered but with the single exception of patients treated in a National Cancer Institute designated cancer center. Because adjuvant therapy is known to improve both disease-free and absolute survival, the most logical conclusion would be that patients treated by a SO or in a designated cancer center had better access to multidisciplinary treatment. Technical expertise, however, cannot be excluded completely.
In my Presidential Address to the Society of Surgical Oncology in 1999,2 I catalogued the surgical oncology volume done during surgical residency and surgical oncology fellowship for 1997 and compared the surgical oncology volume between SO and non-SO surgeons (as defined by Skinner and colleagues) at the time of 10-year recertification for the American Board of Surgery during the study period between 1995 and 1997. The data reflect a difference in practice patterns between the two groups compared. Survival and local failure rates were not addressed. The average numbers of axillary dissections done by a general surgical resident and surgical oncology fellow during 1997 were 29 and 38, respectively. Sentinel lymph node biopsy had not yet replaced standard axillary dissection for patients with negative lymph nodes. Twenty-nine axillary dissections seemed an appropriate number to adequately train a general surgeon in the procedure if the surgeon continued to do the operation in practice. At the time of 10-year recertification, 50% of non-SO surgeons did fewer than six axillary dissections per year and 50% of SO did fewer than 12. Skinner and colleagues noted from the study of the Los Angeles County Cancer Surveillance database during the same time period that 61% of non-SO surgeons and 16.7% of SO did only one breast cancer operation per year. Since sentinel node biopsy is an accurate predictor of a negative axilla, the number of axillary dissections done during surgical residency and surgical practice, no doubt, has already decreased. If technical expertise does have a prognostic effect on survival and local recurrence, sentinel node biopsy will have a predictable negative impact for patients with positive nodes needing an axillary dissection.
Axillary dissection can be a technically challenging operation to do appropriately. It is also an operation that can be attempted, completed, and billed for with retrieval of only a few lymph nodes and without ever visualizing the axillary vein. The number of lymph nodes removed can have prognostic significance. The Danish Breast Cancer Cooperative Group study3 of 13,951 patients with negative lymph nodes and lesions less than 5 cm in size showed that patients who had 10 or more lymph nodes retrieved from the axilla had an improved survival and disease-free recurrence when compared with patients who had fewer than 10 nodes removed. In the early days of the NSABP-06 trial, the number of lymph nodes removed to define an adequate axillary dissection was determined to be 11. Since this number had to be established, the assumption is that clinical investigators were reporting patients who had many fewer nodes retrieved. No question an undirected axillary sampling procedure is easier than an anatomical axillary dissection (and reimbursement is the same). The question is whether or not an undirected axillary sampling is a safe oncologic procedure?
The addition of radiotherapy to chemotherapy improved disease-free and absolute survival in both the Vancouver4 and Danish trials.5 The median numbers of lymph nodes retrieved in these two studies were 11 and 7, respectively, and 80% or more of the nodes were positive. A hypothesis could be that positive nodes were left behind in the axillae and sterilized by radiotherapy, thereby leading to an improvement in survival that would have been anticipated by an initial adequate anatomical axillary dissection. Skinner and colleagues prejudiced their data against the SO because most of the patients excluded from the study had advanced disease and possibly would have benefited the most from a therapeutic axillary dissection.
Everyone had agreed that appropriate surgical staging of the axilla was important at least for predicting survival and dictating chemotherapy. Now that chemotherapy is offered to almost everyone with invasive breast cancer, there are those who question if the axilla needs to be staged at all (the resurrection of the old theory that breast cancer is always a systemic disease).
Randomized, prospective trials to determine a survival advantage for prophylactic axillary dissection have been difficult to compose and complete. In a meta-analysis on the subject, Orr6 attributed a 5.4% survival advantage to axillary dissection. Let us hope the figure is this low if anatomical axillary dissection disappears from the armamentarium of the general surgeon. Let us also hope that adjuvant therapy covers our surgical flank. If not, then diagnostic tests such as sentinel lymph node biopsy, or even better, a noninvasive one, may be used in the future to identify patients who need an anatomical axillary dissection so that patients can be referred to surgeons who are still technically competent to do the procedure.
Received for publication April 21, 2003. Accepted for publication May 15, 2003.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
