In the majority of solid tumors, demonstration of regional or distant metastases provides prognostic information that supercedes any primary tumor characteristic. The development of sentinel lymph node technology by Morton et al.1 and others has been an important advance in the management of patients with melanoma and breast carcinoma, with emerging applications in other solid tumors. The current techniques for sentinel lymph node localization using radioactive tracers and vital blue dye, although not universally standardized, are remarkably robust with good results that are widely reproduced. In experienced hands, this low-morbidity procedure has a false-negative rate of approximately 3–4% and an overall accuracy of approximately 99%. The histologic status of the sentinel lymph node is recognized as perhaps the most important primary prognostic factor in patients with early-stage cutaneous melanoma.2 These clinical facts have propelled sentinel lymph node biopsy into the standard of care at virtually all major melanoma centers worldwide.
It is worth considering what is unique about the sentinel lymph node. The anatomic and working definition of the sentinel lymph node is the first draining lymph node(s) in the direct lymphatic drainage pathway for the primary tumor site.3 Therefore, it is the lymph node(s) most likely to receive lymphatic-borne metastases from the primary tumor. Surgeons use anatomic techniques to localize sentinel lymph nodes and harvest them for pathologic analysis. However, the sentinel lymph node also is clearly a biologic entity. The immune-competent cells in the sentinel lymph node are likely the first in the body to encounter tumor-associated antigens. The sentinel lymph nodes also contain lymphocytes that are activated, occasionally in a relatively nonspecific manner, in response to increased lymphatic drainage, debris, and nontumor antigens related to surgical procedures, trauma, and even inoculations.
In the current issue of Cancer, Kahle et al. provided a thoughtful and detailed description of their prospective study of preoperative ultrasound for localization of the sentinel lymph node in patients undergoing lymph node staging for melanoma.4 In this intriguing study, the authors were able to preoperatively identify sonographically the same lymph nodes they later harvested as sentinel lymph nodes. The primary technical challenge to this study was the coregistration of lymphoscintigraphy findings with ultrasound findings. For this purpose, the authors relied principally on the agreement of preoperative skin markings, comparison of sonographically measured sentinel lymph nodes in vivo with directly measured ex vivo lymph node dimensions, and location with reference to neighboring vascular structures. These comparisons demonstrated that the accuracy of ultrasound in detecting sentinel lymph nodes in the inguinal area was 100% and that in the axillary area was 72.5%. The authors concluded that preoperative sentinel lymph node ultrasonography is a potentially useful adjunct to sentinel lymph node biopsy for improving the surgical approach and strategy.
Although the results are interesting and support the authors' conclusions, several study design shortcomings deserve comment. There is, in fact, no common marker that reliably labels a lymph node for both ultrasound and radioguided detection. Short of finding tumor in the sentinel lymph node by ultrasound, or information provided by a complete lymph node dissection with measurement of all nonsentinel lymph nodes in the basin, comparison of skin markings and even lymph node dimensions cannot confirm the lymph nodes were the same precisely. Coregistration of findings is vitally important when one considers that only approximately 15–20% of patients undergoing sentinel lymph node staging will be found to have positive lymph nodes, and 80–85% of these will have only a single tumor-positive lymph node. A slightly less precise but more practical approach would have been to perform an ultrasound-guided fine-needle biopsy of the sonographic sentinel lymph node candidate(s) shortly after intradermal injection of the radiotracer but before surgical harvest to determine whether the aspirate from the lymph node contained radioactive tissue, a finding that strongly supports sentinel lymph node status. It also is necessary to recognize the potential bias of the sentinel lymph node procedure that could have been introduced by a concomitant ultrasound examination performed by the operating surgeon in approximately 33% of cases.
These critiques notwithstanding, perhaps the most important finding in the study by Kahle et al.4 was the finding of an apparently unique sonomorphologic pattern, an asymmetric “cap phenomenon,” that typified sentinel lymph nodes harvested by traditional radiolocalization/blue dye techniques. Not only could the “cap” be identified sonographically much of the time, but this area of the lymph node also corresponded to the point of maximum blue dye staining and radioactivity within most of the sentinel lymph node(s). The “cap phenomenon” was distinguishable sonographically from the more symmetrical appearance of nonsentinel reactive lymph nodes. The “cap” appears to be nonspecific because it was noted in both tumor-positive and tumor-negative sentinel lymph nodes. The “cap” was also a seemingly transient phenomenon, because repeat sonographic evaluation just before lymph node biopsy demonstrated diminished findings. These observations suggest that some form of nonspecific biologic reaction is occurring in the sentinel lymph node. Whether this reaction is due to reactive hyperplasia, lymphatic engorgement, the lymphoscintigraphy tracer, or some other mechanism is unclear.
Although nonspecific, this reaction nonetheless may be clinically useful to help identify the sentinel lymph node. Our own anecdotes and those of other experienced sentinel lymph node surgeons are supportive in this regard. Although the importance of lymphoscintigraphy in identifying the basins of interest is obvious, and radioactive counts and blue dye staining are the primary identifiers of sentinel lymph nodes, we also rely in part on tactile identification of the sentinel lymph node candidates. Through a small incision, we initially use the probe to define the approximate direction and probable location of the sentinel lymph node within the basin, and then use a finger to bluntly dissect and find a palpable lymph node in close proximity to this hot spot. Often, this lymph node is a bit larger or firmer than its neighbors. Once the sentinel lymph node candidate is delivered into the wound sufficiently to interrogate it further, it often is demonstrated to be both blue and radioactive. Frequently, only one pole or end of the lymph node is blue and/or radioactive. It is amazing how consistently this constellation of findings coexists. Therefore, we can believe that the sentinel lymph node might have at least a transiently distinguishable sonomorphologic identity.
What does this mean for today's practicing sentinel lymph node surgeon? First, prospective studies in well defined clinical scenarios comparing with the gold standard (in this case, sentinel lymph node staging) are necessary to ascertain the true sensitivity and specificity of any diagnostic test for lymph node metastases. Our experience with positron emission tomography with 18F-fluordeoxyglucose for lymph node staging in patients with early-stage melanoma illustrates that extrapolation of results from populations with a higher prevalence of detectable disease burdens may overestimate sensitivity.5 Ultrasonography has been shown to be a sensitive (perhaps the most sensitive) noninvasive indicator of small lymph node metastases in the melanoma patient.6 However, the sensitivity and specificity of ultrasound in this application are inferior to histologic analysis of sentinel lymph node tissues. To our knowledge, considerable clinical experience with ultrasound staging of lymph node basins has yet to produce a consensus regarding the clinical role of lymph node sonography in the sentinel lymph node era.
Given the excellent and robust results observed with sentinel lymph node biopsy, it is not clear how most experienced surgeons would find sonographically assisted sentinel lymph node harvest useful. Perhaps in the axillary lymph node basin of an obese patient, one might envision a limited role for the preoperative ultrasound localization of the most likely sentinel lymph node candidate(s) to facilitate a more rapid harvest. Unfortunately, the authors noted the poorest results in this population. However, it is conceivable that intraoperative ultrasound sentinel lymphadenectomy might have some applicability in detecting sentinel lymph nodes in more difficult situations, such as the emerging field of sentinel lymph node staging for thoracic, upper aerodigestive mucosal, or gastrointestinal tumors.
This is not to say that today's treatment algorithms are optimal. Room for improvement exists. A different role for ultrasound sentinel lymphadenectomy might be envisioned. A common clinical scenario in patients with melanoma would be the potential avoidance of a diagnostic sentinel lymph node procedure if the patients can be proven nonsurgically to have tumor-positive sentinel lymph nodes. If ultrasound can identify the sentinel lymph node with enough precision to guide fine needle aspiration biopsy, preferably of the “cap” area, patients determined to have positive lymph nodes might go straight to therapeutic lymphadenectomy. Although surgery is currently the treatment of choice for patients with melanoma lymph node metastases, alternative treatment options exist for other malignancies. For example, in patients with breast carcinoma, selected lymph node-positive patients might receive lymphadenectomy, radiation, and/or chemotherapy without primary sentinel lymph node biopsy.7 Similar treatment algorithms could be envisioned in gynecologic malignancies and tumors of the head and neck, in which elective lymph node irradiation for potential microscopic disease is commonly considered.
Kahle et al. are to be congratulated for their potentially seminal research demonstrating that ultrasound identification of sentinel lymph nodes may be feasible. It now is time for additional definitive studies to verify and refine these results as a prelude to the next era of minimally invasive staging procedures.