| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Hartford Hospital and the University of Connecticut School of Medicine, Hartford, Connecticut.
Correspondence: Address correspondence to: Kenneth A. Kern, MD, FACS, FSSO, Hartford Hospital and the University of Connecticut School of Medicine, 85 Seymour St., Suite 1011, Hartford, CT 06106; Fax: 860-541-2299; E-mail: Kennet2521{at}aol.com
The search for the lowest false-negative rate in sentinel node biopsy for breast cancer is an undertaking worthy of all surgical oncologists. This search is particularly vital, however, for those surgeons who continue to use the original method of dye-only, peritumoral (PT) injections as described by Giuliano1 because PT injections must overcome the poor migration characteristics related to the relative paucity of lymphatic channels within the breast parenchyma.2–5 False-negative rates reported after PT injections range as high as 11%, with approximately 5% as an average false-negative rate.6 Whether a consistently low false-negative rate using dye-only, PT injections can ever be achieved by a diverse group of surgeons remains to be seen. This search for the lowest false-negative rate using PT injections may be loosely compared with the 12th century Crusaders’ search for the mythological Holy Grail. Word of the relic’s existence prompted heroic search efforts that in the end came to naught because the Holy Grail existed in the imagination of its seekers, and not in reality.
In PT sentinel node biopsy, the search for the Holy Grail, the lowest false-negative rate, may take two forms: either continued adjustments and fine-tuning of the original PT methodology or abandoning the original PT method altogether and seeking a new route to the Holy Grail. I have chosen the latter method and have abandoned PT injections in favor of the subareolar (SA) approach to breast lymphatic mapping. Using combined radiocolloid and blue dye SA injections, I have reported a 98.8% identification (ID) rate, a 0% false-negative rate, and a 96% concordance rate,5 findings supported by the results of other studies.7,8
Nos et al.9 have chosen to persist on the original route to the Holy Grail by advocating a dye-only, PT injection with added modifications to decrease the false-negative rate. The false-negative rate for PT injections before the compensating mechanisms they describe is 11.1%, a finding in line with other dye-only, PT approaches.6,10 Nonetheless, this false-negative rate is clearly an underestimate because the ID rate of sentinel nodes was only 85.5%. This is significantly below the ID rate of the subareolar method of over 98%.5 Using the method of Nos et al. resulted in over 14% of patients never having had the chance to undergo a sentinel node biopsy. Unfortunately, the authors never addressed methods to improve this first step in the sentinel node process, finding the sentinel node in the first place.
To decrease this false-negative rate among patients with an identified sentinel node, Nos et al. retrospectively applied three major compensating mechanisms. First, they checked other nonblue nodes, removed during the completion axillary dissection, for blue staining after fixation in an acid-alcohol-formalin (AFA) fixative (called the "pathological color quality assessment"). They labeled any nodes with delayed blue staining "sentinel nodes" (in this article I will call these latent sentinel nodes). Second, they subjected all latent sentinel nodes to 24 step-sections per node. Last, they defined a retrospectively identified "positive" sentinel node as any latent sentinel node with two or more immunohistochemical-positive cells (IHC [+]), even without the presence of positive hematoxylin and eosin (H&E) cells.
By using these sequential postbiopsy maneuvers, Nos et al. were able to state, after the fact, that some latent sentinel nodes were actually positive because they had at least two IHC (+) cells. These relabeled latent sentinel nodes were used to calculate a lower false-negative rate for this procedure of 2.2%. Although this is an interesting technical achievement, I do not believe this method of identifying latent sentinel nodes, or of recalculating the false-negative rate, is valid or applicable to sentinel node biopsy performed by American surgeons, for several reasons.
First, American pathologists uniformly use 10% formalin fixation and not AFA fixation. Our hospital’s pathologists, who process well over 300 breast cancer cases yearly, were completely unfamiliar with the AFA formulation. When I checked the wax blocks from several cases with extremely blue nodes seen in the operating room, no blue dye was noted, and the pathologists had never noted the latent appearance of blue dye in step-sections of hundreds of sentinel node biopsies. From the outset, this type of latent discovery of blue dye in axillary nodes does not appear to be applicable when using the standard 10% formalin fixation techniques.
Second, American pathologists cut at most four to eight step-sections and do not subject sentinel nodes to the extensive step-sectioning technique described by Nos et al. Extensive step-sectioning of any nodes, sentinel or nonsentinel, may result in finding IHC (+) cells in as high as 20% of cases and often in small tumors not usually associated with positive lymph nodes.11 In my opinion, the clinical significance of IHC (+) cells discovered in the 15th cut-down on an axillary node remains to be seen and does not fit the original definition of a sentinel node. Moreover, Nos et al.’s redefinition of a sentinel node as any node with two or more IHC (+) cells does not meet the criteria established by the American Joint Commission of Cancer Staging (AJCC). The American definition of a positive sentinel node, as codified by the AJCC, requires a cluster of H&E positive cells of 200 microns or greater. Cells that are solely IHC (+), in the absence of H&E positive cells, do not meet the definition of a positive sentinel node.
The maneuvers used by Nos et al. are an attempt to improve the sensitivity of sentinel node biopsy in the setting of poor migration characteristics of PT injections. Although the sensitivity has increased, the specificity has not. That is, these investigators have identified cells deep within nodes that may not have clinical meaning and do not constitute "positive" sentinel nodes by American definitions of metastatic nodal disease. They have redefined a "positive" sentinel node and then constructed a false-negative rate built upon this redefinition. I do not believe this is a valid approach to defining a false-negative rate in sentinel node biopsy.
American surgeons have achieved real improvements in the false-negative rate of sentinel node biopsy by overcoming the problem of poor breast migration of dyes and tracers. They have done so by using combined dye-isotope methods, often injecting one or more tracers in a site separate from the breast parenchyma, such as the dermis or subareolar lymphatic plexus. The false-negative rate using these combined, two-site methods has decreased to 0–2%. This rate is defined using the standard criteria of 4 to 8 step-sections with a positive node requiring at least a 200-micron cluster of H&E positive cells.
We applaud Nos et al. for their arduous trek toward reaching the Holy Grail in dye-only, PT lymphatic mapping. In particular, their use of confirmatory dissections in all their cases is to be acknowledged as a truly scientific and dedicated approach to improving the technique of sentinel node biopsy. However, their chosen route to the Holy Grail is circuitous and certainly not for the faint-hearted because it requires a special AFA fixative, multiple step-sections of suspected sentinel nodes, and a redefinition of a positive sentinel node as one containing two IHC (+) cells. This route is too long and difficult for me. Instead, I will continue to advocate a shorter, simpler, and more direct route to the Holy Grail by avoiding PT injections and relying upon SA injections using combined dye-isotope mapping agents.
Received for publication April 5, 2003. Accepted for publication April 21, 2003.
REFERENCES
| ||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
