| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
EDITORIALS |
From the Division of Surgical Oncology, Department of Surgery, University of Louisville, James Graham Brown Cancer Center, Louisville, Kentucky.
Correspondence: Address correspondence to: Kelly M. McMasters, MD, PhD, Division of Surgical Oncology, Department of Surgery, University of Louisville, James Graham Brown Cancer Center, 315 E. Broadway, Suite 308, Louisville, KY 40202; Fax: 502-629-3379; E-mail: kelly.mcmasters{at}nortonhealthcare.org
There are three kinds of explanation in science: explanations which throw a light upon, or give a hint at a matter; explanations which do not explain anything; and explanations which obscure everything. –Friedrich Von Schlegel (1772–1829)All truth passes through three stages. First, it is ridiculed. Second, it is violently opposed. Third, it is accepted as being self-evident. –Arthur Schopenhauer (1788–1860)
In science there is never any error so gross that it won’t one day, from some perspective, appear prophetic. –Jean Rostand (1894–1977)
The work of John F. Thompson et al.1 from the Sydney Melanoma Unit is always thought provoking. It, however, has provoked this meandering and wide-ranging essay, seasoned with a smattering of unrestrained speculation as well as some reflections on the pursuit of scientific truth.
Thompson et al.1 examined the fate of their patients who were clinically node negative from the eras of elective lymph node dissection (ELND; 1983–1992) and sentinel lymph node biopsy (SNB; 1992–2000). Only patients with melanomas 
1.5-mm thick were evaluated. In comparing the two groups, two major endpoints were evaluated: nodal metastasis and survival. Nodal metastasis was detected more frequently by SNB compared with ELND (18.0% vs. 14.6%, respectively; P = .005). This was despite significant differences in age, tumor thickness, and ulceration that would favor a decreased incidence of nodal metastasis in the SNB group. From these findings, the authors conclude that SNB is significantly more accurate than ELND in identifying early nodal metastases. As they discuss, this would be expected given the more intensive pathologic scrutiny of the SLN compared with that of a complete lymphadenectomy specimen.
Two other noteworthy findings were reported based on multivariate analyses of factors predicting nodal metastasis and survival. First, increasing age was significantly associated with both a decreased risk of nodal metastasis and an increased risk of mortality from melanoma. We have noted this paradox previously, and have suggested that, compared with older patients, a more competent immune system in younger patients might allow eradication of minimal nodal disease more frequently.2,3 Alternatively, older patients may have a greater propensity for distant metastasis without concomitant nodal metastasis, perhaps because of differences in the lymphatic system such as lymphatic channel atrophy4 or reduced lymphatic vessel density.5,6 Maybe melanomas with more abundant lymphatic drainage spread more commonly via the lymphatics. A similar idea has been proposed to explain why palpable breast cancers have a much higher incidence of nodal metastasis, size for size, compared with nonpalpable tumors: perhaps they are just closer to the rich subdermal lymphatic network.7 Although we ordinarily consider that melanoma and breast cancer invade the lymphatic and vascular spaces, allowing for cancer dissemination, perhaps lymph node metastasis relies on the ability of the cancer to initiate formation of new lymphatic channels—tumor lymphangiogenesis.8–10 This is obviously analogous to tumor angiogenesis, or the formation of new blood vessels. Thus, perhaps we should stand the dogma on its head: instead of cancer invading the lymphatic vessels, maybe the lymphatics are invading the cancer! Because these neolymphatic vessels do not form out of nowhere (i.e., they develop from existing lymphatic vessels), maybe the propensity of the tumor to express potent lymphangiogenic signals, coupled with proximity to a rich lymphatic network, is what determines the likelihood of lymphatic metastasis.
A recent intriguing study using a mouse animal model of melanoma found that tumor growth resulted in formation of new lymphatics associated with vascular endothelial growth factor-C (VEGF-C) expression. These new lymphatic vessels were functional, but displayed a retrograde drainage pattern, suggesting possible dysfunction of the intraluminal valves. Perhaps this can explain some cases of satellite and in-transit metastasis, false-negative SNB findings, as well as the potential for regional nodal metastasis.11 While we are theorizing: maybe ulcerated melanomas are associated with greater tumor lymphangiogenesis (and angiogenesis), which would explain why the ulcerated phenotype predicts a greater propensity for nodal (and systemic) metastasis. Maybe the fluid that sometimes exudes from the ulcerated melanomas is lymphatic fluid. A lot of "maybes," but among these baseless speculations lay testable hypotheses for future studies.
The Sydney Melanoma Unit study also found that head and neck primary melanoma location was associated with a decreased probability of nodal metastasis. We have also noted this, along with the fact that false-negative SNB results are somewhat more common for head and neck melanomas.12 Thompson et al. offer the suggestion that head and neck melanomas are less likely to develop metastatic disease in the regional nodes as the first site of disease compared with melanomas of the trunk or extremity. In fact, head and neck melanomas have a greater chance of satellite and in-transit metastasis compared with other sites. Whether this has anything to do with lymphatic density or lymphangiogenesis remains to be seen. Again, these possibilities can and should be addressed in future analyses.
Despite a greater incidence of positive nodes, no difference in survival was noted between the ELND and SNB groups, even after adjusting for the differences in the length of follow-up. Several possibilities come to mind. First, maybe SNB (with concomitant therapeutic lymphadenectomy for patients who were node positive) really does not improve survival. This certainly is plausible, and is the subject of an ongoing randomized study, the Multicenter Selective Lymphadenectomy Trial (MSLT), directed by Morton.13 Second, perhaps the absence of a survival difference reflects the fact that the two groups in the study were not evenly balanced in terms of major prognostic factors. Although the SNB group was older and had more positive nodes, they also had fewer ulcerated tumors and a lower mean Breslow thickness. Maybe these prognostic factors balanced themselves out. Third, perhaps SNB detects, in some cases, small foci of micrometastatic disease that do not have clinical significance, or at least not the same prognostic significance as greater tumor burden detected by routine histology of ELND specimens. This argument has certainly been made for SNB staging of breast cancer, and is starting to appear in melanoma.14,15 Whereas we have known for a long time that the number of positive nodes is an important predictor of prognosis, the prognosis for patients with a single positive sentinel node can vary widely, depending on the amount of disease or patterns of metastasis within the node. This, too, is an area ripe for further investigation. Nonetheless, the nonrandomized nature of the present study from the Sydney Melanoma Unit precludes a definitive answer to the question of whether SNB improves survival.
So what good is SNB if it does not improve survival? I agree with Thompson et al. that we must await the results of the MSLT study to determine whether SNB can be recommended as a therapeutic procedure. If SNB is associated with an improvement in survival at all, it probably will be modest. Yet, I do not believe that we need to wait to recommend SNB as a diagnostic procedure. SNB should not be considered a therapeutic procedure; it is a diagnostic test to determine the status of the regional lymph nodes. As with any other staging test, SNB is not 100% accurate, but it is far better than clinical staging to detect nodal metastasis, and far better than ELND in terms of both sensitivity and morbidity. We do not require any other cancer staging tests to improve survival; why should we raise the bar for SNB? Because it is a minimally invasive surgical procedure? I have not heard anyone suggest that positron emission tomography (PET) scan must improve survival for melanoma, but given the cost, the same logic applies.
The presence of nodal metastasis is such a powerful prognostic factor, even considering the caveats discussed above, that it is difficult to envision reverting to clinical staging of regional nodes, even if SNB provides no survival advantage. If nothing else, SNB provides excellent regional disease control; recurrence in a nodal basin after therapeutic lymphadenectomy for a microscopically positive sentinel node is rare.16 This is in contrast to therapeutic lymphadenectomy performed for palpable nodal disease, when extracapsular extension, and invasion and encasement of neurovascular structures can create a nightmare for regional disease control. Those of us who are called on to confront this problem can assure those of you who do not that loss of regional disease control is catastrophic. SNB has virtually eliminated this problem. Remember, if we revert to wide local excision alone as standard treatment for patients with clinically node-negative melanomas, we will not prevent the need for regional lymphadenectomy. Many patients will still need to have a lymph node dissection–just, on average, 2 to 3 years later. Although I fully expect that, in the future, molecular profiling of the primary melanoma will supplant SNB for determining both prognosis and response to therapy, that day has not yet arrived. For the time being, we should strive to learn as much as possible about the mechanisms by which melanoma spreads via hematogenous and lymphatic routes, and hope to find molecular pathways to exploit using novel targeted therapies.
In science, it is always instructive to reflect on what we now know, or think we know, versus what we formerly knew, or thought we knew. In past decades, this essay might well have been a strong defense of 5-cm excision margins, adjuvant hyperthermic isolated limb perfusion, or elective lymph node dissection. It is hard to know the truth in advance of the evidence. That does not keep some of us from trying, but a little bit of data can often straighten that out. That is why we are indebted to the present study from the Sydney Melanoma Unit, and others like it, which add to our knowledge, induce new hypotheses, and ultimately answer some of the questions.
Received for publication July 11, 2004. Accepted for publication July 22, 2004.
REFERENCES
This article has been cited by other articles:
 |
|
 |
|
  C. Caraco, U. Marone, E. Celentano, G. Botti, and N. Mozzillo Impact of False-Negative Sentinel Lymph Node Biopsy on Survival in Patients with Cutaneous Melanoma Ann. Surg. Oncol., September 1, 2007; 14(9): 2662 - 2667. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Mocellin, A. Ambrosi, M. C. Montesco, M. Foletto, G. Zavagno, D. Nitti, M. Lise, and C. R. Rossi Support Vector Machine Learning Model for the Prediction of Sentinel Node Status in Patients With Cutaneous Melanoma Ann. Surg. Oncol., August 1, 2006; 13(8): 1113 - 1122. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
