Skin cancer surgery on referral

The role of sentinel node in melanoma management

MSLT spells a halt to sentinel node biopsy


In 1892 Dr. Herbert Snow observed that melanoma patients who developed lymph node involvement invariably died of their malignant melanoma. Ever since that time, clinicians and researchers have been trying to find a way to turn this observation into a better outcome for our patients. Snow proposed that all melanoma patients have their regional lymph nodes resected260. However, several subsequent random control trials have failed to demonstrate a survival advantage from elective lymph node dissection (ELND).


80% of the nodes resected through ELND had no melanoma involvement. Patients were being subjected to considerable morbidity for no apparent benefit. ELND has long been abandoned from routine management of malignant melanoma. But if we knew which 20% of patients had the melanoma in their nodes, maybe these melanoma patients could be selected for lymph node dissection.


Sentinel node biopsy described:

100 years later, Dr. Donald Morton described sentinel lymph node biopsy (SNB); a method of identifying patients that have early node involvement such that these patients could then be offered completion lymphadenectomy (CL)261. Morton’s approach promised three advantages for patients: improved survival, enhanced patient information and guidance for future management.


SNB involves the use of lymphoscintography and / or dye to identify the sentinel node (SN). The identifier is injected around the site of the primary melanoma, enabling the material to be carried through the lymph system to the SN. Dissection of the node basin can then identify this SN and it is removed leaving adjacent nodes in place. Involvement of melanoma in the processed SN would then direct management to CL.


Lens262 identified that the likelihood of melanoma involvement in SNs is proportional to the Breslow thickness of the primary tumor. 1% of tumors below Breslow 0.75mm will have SN involvement. 8.3% of patients with Breslow 0.75 to 1.5 mm melanomata will have positive SNs, 22.7% of patients with Breslow 1.5 to 4 mm and 35.5% of more advanced tumors.


The original idea of SNB involved resection of the SN at the time of resection of the primary tumor. In two separate studies, Leong263 and Evans264 have identified that outcomes are not different if the primary tumor excision is undertaken before the SNB. Given the only management of melanoma that improves outcome is prompt wide local excision (WLE), a lengthy delay for tumor resection just to enable simultaneous SNB is not justified.


Multicenter trial of SNB

After developing the concept of SNB, Morton went on to test the efficacy of this technique through a well designed and powered random control trial known as the Multicenter Selective Lymphadenectomy Trial (MSLT – 1). This trial was based at the John Wayne Cancer Institute in Santa Monica California. The multinational centers involved in the trial included Sydney Australia where a large number of patients were recruited for the trial.


MSLT excluded melanomata with Breslow below 1 mm. This has proved an appropriate exclusion given the low incidence of positive SNs in patients with early tumors.


In this study, 1200 patients were randomized to SNB with all those having positive nodes proceeding to CL. There were 800 controls who did not have SNB performed, but had lymphadenectomy if and when nodes became subsequently involved. The survival data from the MSLT – trial has now been presented at the American Society of Clinical Oncology meeting in May 2005. This same data has also been presented to the 6th World Congress on Melanoma in September 2005 and American Society for Dermatologic Surgery Annual Meeting in October 2005.


The five year survival for patients in the SNB group was 87%, versus 86% in the control group (p=0.4). The sentinel node patients gained no survival benefit.



Unfortunately a flawed sub-analysis of this study has created more interest than the core data. Morton showed that SNB positive patients demonstrated a 22% 5 year survival advantage over patients who later developed tumor involved nodes. Such a so-called advantage is hardly surprising. This simply reminds us those patients likely to develop involved lymph nodes fare better than patients who have proven lymph node involvement.


At the end of the day, the one comparison that is valid in the MSLT study is the primary outcome for which the study was designed, the 5 year survival of treatment versus control arms. Their outcomes are the same.


An earlier study by Kretschmer265 has suggested an outcome benefit for SNB patients. This paper is much quoted but flawed. The study compared SNB patients with those who later developed positive clinical nodes and had delayed lymph node dissection (DLND). 5 year survival for DLND patients was 50.2% compared with 62.5% for SNB positive patients.

However, the study did not involve patients who did not have SNB and did not develop clinical nodes. As such, the study simply compared those who might develop metastatic nodal involvement with those that did develop involved nodes. The discussion within this manuscript goes to some length to explain how lead time bias was factored into the analysis. Lead time bias was not the predominant flaw in the analysis, rather it was the failure to compare patients on an “intention to treat” basis.


How much melanoma in the SN?:

In the MSLT trial, any melanoma in the SN meant it was deemed positive and CL was offered. But there is increasing evidence that small amounts of melanoma within the SN may not be of great concern. Vuylesteke266 demonstrated that patients with Breslow thickness less than 2.5 mm who had positive SNs with tumor load of less than 0.3 mm did not have additional nodes involved with melanoma. They had excellent survival prospects and did not need to have CL.


The significance of a small amount of melanoma involvement in a node (or micrometastasis) was further questioned by Giblin et al267. If there are tiny foci of melanoma in SNs then it is difficult to justify further offering the patient CL. At times there will be pockets of melanoma in peripheral sinuses of SNs which, apparently for immunological reasons, do not progress to metastasis. It would be devastating to effect a block dissection and cause morbidity such as lymphoedema only to find predictably uninvolved nodes in the large dissection specimen.


Starz268 data demonstrates any SN involvement of less than 1 mm has no adverse prognostic significance. Spanknebel269 demonstrated that micrometastases identified only on immunochemical staining have no adverse prognostic significance.


We can no longer justifiably advise all our SN positive patients to proceed to CL. Cochran270 has demonstrated that only one third of patients with positive SNs will have melanoma in adjacent nodes. Indeed proceeding to CL may only be appropriate if the original tumor had a large Breslow thickness, there was considerable volume of melanoma in the SN and also dense dendritic cells in the SN. It is difficult to justify advising a patient with an early Breslow tumor to have the SNB procedure if we would not then advise proceeding to CL even when the SN contained melanoma.


Complications of SNB:

The MSLT trial demonstrated that 10.1% of patients who undergo SNB, develop complications.271 Indeed the complication rate rose to 37.2% in those patients who were SNB positive and went on to CL.

SNB complications are such a problem on the head and neck that some centers exclude patients who have primary head and neck tumors. For example, some 25% of these SNs are located within the parotid gland, with dissection to identify and remove the node risking facial nerve damage.272


If SNB provides no survival benefit but is associated with such concerning morbidity figures, there must be some other reason to justify patients undergoing this procedure.


Information for patients:

Mohrle et al273 studied 283 SLN patients finding that these patients gained no benefit in survival time. While these findings supported the MSLT data, this paper also raised again the value of SNB as a means of providing accurate advice for our patients.


Zogakis47 demonstrated that 9% of SNB negative patients developed tumor recurrence. Similarly, Gershenwald274 demonstrated 13% of patients who were SNB negative would develop recurrence somewhere else within three years. Clearly SNB negative patients are surviving poorly compared with patients who did not have SNB and did not subsequently develop nodal involvement.


SNB negative patients can be advised that they are less likely to develop later metastatic disease. But they cannot be advised that metastatic disease has been ruled out. The advice that one’s SN is negative simply means a slightly lower risk, not a removed risk of mortality from the malignancy. Melanoma can spread via the circulatory system. Melanoma patients need careful follow up even if they have a negative SNB.


Melanomata with large Breslow thickness are associated with increased risk of mortality and increased risk of SN positivity. As such, to what extent is that added risk that comes with having a SN positive merely reflect the underlying Breslow thickness?


Roka275 undertook a multivariate analysis of risk factors for survival in malignant melanoma. The analysis was able to separate the predictive value of Breslow thickness from the predictive value of SNB status. Roka found that SNB was not predictive of overall survival. Rather, Breslow thickness was the only independent predictor of overall survival. In an analysis of disease free survival, Breslow thickness, SNB status and ulceration were all predictive. Breslow thickness remains the single most important detail of a melanoma when it comes to providing information and advice to our patients.


In transit melanoma:

Thomas et al101 raised the concern that the SNB procedure may increase our patient’s risk of subsequent in transit and loco-regional metastases. These same investigators published the excellent random control trial of 1 cm versus 3 cm clearance for high risk melanoma in the New England Journal of Medicine276. In this metanalysis101, they demonstrated an incidence of in-transit metastasis in SN positive patients following CL was 4 – 5 times that expected for that mean Breslow thickness. In other words, it is doing synchronous or near-synchronous wide excision and completion lymphadenectomy, while melanoma cells are still in transit, which may cause the increased incidence of in-transit disease.


Others have not confirmed this in transit risk277. Van Poll 278 demonstrated Sydney patients were not more likely to develop in transit disease following SNB. Kretschmer 279 refuted Thomas concluding that SNB and CL avoid later tumor recurrence and prolong disease free survival in melanoma patients. Unfortunately this study has the same “intention to treat” problems that were characteristic of their earlier work265.


Regardless of whether or not SNB leads to an increased risk of in transit and loco-regional disease or not, the vital impact for our patients is whether they will live longer through the SNB procedure. The MSLT study demonstrates they will not. As a sufficient number of authors, independently, have now described an increased incidence of in-transit metastasis, all patients considering this operation should be warned of this possible risk.


Does SNB guide future management?:

We do not have treatments for metastatic melanoma that prolong life. Chemotherapy, radiotherapy and other approaches have not demonstrated improved long term survival for our melanoma patients. We cannot even justify CL in SNB positive patients with tumor load in the SN beneath 1 mm.


This situation might change in the future. Emerging treatments such as B RAF inhibitors280 may provide a future management for patients with early nodal disease. If such treatments are demonstrated to improve patient survival, then we may have a reason to identify early metastatic disease through the SNB procedure.


Abandon SNB?:

Medalie and Ackerman281 consider that SNB now lacks any value to the extent that it should be abandoned. Thomas101 does not suggest the technique be abandoned, but does consider the technique should no longer be used outside of randomized control trials (RCTs). Indeed given the lack of benefit of SNB and the significant morbidity that comes with the procedure, it is now difficult to justify the test outside of RCTs. At best SNB provides some extra advice for patients with melanomata between Breslow 1 and 3 mm in thickness.



SNB promised three things; improved survival, enhanced patient information and guidance for future management. SNB does provide added information for patients who have tumors with Breslow thickness between 1 and 3 mm. In all other respects the SNB procedure has not delivered on its promise.

SNB may have a role in the future if and when an adjuvant melanoma therapy is developed that improves survival in patients with demonstrated early metastatic involvement.

The MSLT results are very helpful. If a pharmaceutical product performed as poorly as SNB as shown in MSLT the product would never achieve FDA approval, but rather it would be abandoned or invited to proceed to phase iii trials. Why should a surgical management be regarded any differently?

It is now difficult to justify ongoing usage of the SNB procedure other than in the context of an RCT to evaluate possible future medical or surgical interventions for melanoma patients. SNB is certainly not a “standard of care” for melanoma patients