Standard surgery for melanoma with sentinel-node metastases does not extend survival

The surgical management of the regional lymph nodes in melanoma has long been an area of controversy. Sentinel lymph node biopsy has been widely adopted as an alternative to elective lymph node dissection — radical lymphadenectomy for clinically normal lymph nodes without prior knowledge of the pathological status of the nodes — and nodal observation with radical lymphadenectomy at the time of biopsy-proven nodal recurrence, known as therapeutic lymphadenectomy. 

As originally conceived, sentinel lymph node biopsy was intended to obtain pathological staging of the regional nodal basin, with observation of the sentinel node–negative basin but completion lymphadenectomy whenever positive sentinel nodes are encountered.

This approach of sentinel lymph node biopsy followed by completion lymphadenectomy has been shown in a prospective randomized trial — the first Multicenter Selective Lymphadenectomy Trial (MSLT-I) — to be associated with improved melanoma-specific survival for sentinel node–positive patients compared with patients randomly assigned to nodal observation but who subsequently developed nodal recurrence and underwent therapeutic lymphadenectomy.

Often, however, pathological analysis of the completion lymphadenectomy specimen performed after a positive sentinel lymph node biopsy fails to reveal any additional tumor-containing lymph nodes, prompting many to wonder whether completion lymphadenectomy was necessary in all patients with a positive sentinel lymph node biopsy.

Now we have the results of two prospective randomized clinical trials that tested this question: a relatively small study conducted in Germany referred to as the DeCOG-SLT trial, and the second Multicenter Selective Lymphadenectomy Trial, known as MSLT-II.

These two clinical trials involved 2,407 sentinel node-positive patients randomly assigned to either completion lymphadenectomy or nodal observation with serial ultrasonography and physical examination of the affected regional nodal basin.

As expected, few patients randomly assigned to completion lymphadenectomy had further melanoma detected in the lymph node dissection specimen (18% in DeCOG-SLT and 11.5% in MSLT-II).  Median follow-up in both trials was relatively short (35 months in DeCOG-SLT and 43 months in MSLT-II), but the results of both trials are very similar. Neither trial showed any sign that completion lymphadenectomy is associated with improvement in melanoma-specific survival compared with nodal observation and therapeutic lymphadenectomy only for those patients who develop isolated nodal basin recurrence.

In an editorial accompanying the MSLT-II publication, renowned melanoma surgeon Daniel Coit, MD, stated: “If this aggregate of data is insufficient to extinguish the enthusiasm for immediate completion lymph-node dissection, then it is unclear what more is required.” So, is completion lymph node dissection outdated thinking with no role in modern practice, or does it remain an important aspect of the personalized care for patients with melanoma?

Let’s first answer Coit’s rhetorical question regarding what additional data would still inform decision-making regarding completion lymph node dissection. First and foremost is additional follow-up data over a longer period of time. In MSLT-I, the survival advantage associated with immediate lymphadenectomy in node-positive patients didn’t even begin to emerge for 2 years after surgery — and that was with the tumor-containing sentinel nodes left in the patient! It stands to reason that, once the tumor-containing sentinel nodes are removed, any benefit of removing nonsentinel nodes by completion lymphadenectomy would not become apparent for many months or even years. Additional data on the applicability of the trials results across all cohorts of sentinel node–positive patients also would be important before completely abandoning completion lymphadenectomy.

Another way to think about role of completion lymphadenectomy in contemporary melanoma management is to ask about the primary goals of nodal surgery. If the goal of surgery is to perform the fewest lymphadenectomies on patients, then sentinel lymph node biopsy followed by nodal observation in all cases would make sense. But if the goal is to minimize morbidity due to nodal disease and its treatment, there may be subsets of sentinel node–positive patients who have a very high risk for early nodal recurrence and, hence, might be better served by immediate lymphadenectomy rather than risk uncontrolled recurrence or at least a higher morbidity salvage surgery. Moreover, if the goal is to minimize overall morbidity and mortality due to melanoma, then the recognized prognostic information obtained by knowing the status of nonsentinel nodes combined with the more selective application of adjuvant systemic — and regional — therapy might represent the best strategy for at least some subsets of sentinel-node positive patients.

Finally, if the goal is to deliver patient-centered care that allows the patient to actively participate in treatment decision-making whenever two therapeutic strategies have similar anticipated survival rates, then it must be acknowledged that — for appropriately informed and selected patients — a proactive approach that includes completion lymph node dissection offers peace of mind, better staging information, excellent regional disease control and acceptable morbidity in the large majority of patients.

Completion lymph node dissection is not obsolete, at least not yet, and the best management of the sentinel node–positive patient will continue to involve a balanced discussion with patients about the risks and benefits of this procedure — now informed by two randomized trials and a high degree of confidence that there is no early impact of completion lymph node dissection on melanoma-specific mortality. 

The likelihood is that many — perhaps even most — patients will be well served by a strategy that includes a thoroughly performed sentinel lymph node biopsy and careful nodal observation with subsequent therapeutic lymph node dissection in small percentage of patients who experience isolated nodal basin recurrence. Patients who undergo nodal observation after a positive sentinel lymph node biopsy also are appropriate for future adjuvant therapy trials — they have been uniformly excluded to date — and, indeed, nodal basin recurrence might represent an early endpoint for assessing therapeutic efficacy, much like pathological complete response has become in neoadjuvant therapy. The ideal strategy for management of the individual sentinel node-positive patient is still a goal we can only aspire to achieve, not a matter defined even by “practice-changing” clinical trials.

References:

Coit D. N Engl J Med. 2017;doi:10.1056/NEJMe1704290.

Faries MB, et al. N Engl J Med. 2017;doi:10.1056/NEJMoa1613210.

Leiter U, et al. Lancet Oncol. 2016;doi:10.1016/S1470-2045(16)00141-8.

Morton DL, et al. N Engl J Med. 2014;doi:10.1056/NEJMoa1310460.

Disclosure: Sondak reports advisory board roles with Bristol-Myers Squibb, Genentech/Roche, Merck, Novartis and Provectus, as well as data safety monitoring board roles with Array, Bristol-Myers Squibb, Novartis, Pfizer and Polynoma.