No Improvement in Melanoma-Specific Survival With Completion Dissection vs Observation for Sentinel-Node Metastasis

Get Permission

Mark B. Faries, MD

Mark B. Faries, MD

THE PHASE III MSLT- II TRIAL showed that completion dissection was not associated with improved melanoma-specific overall survival vs observation in patients with sentinel-node metastasis, although a benefit was observed in regional disease control. The findings were reported in The New England Journal of Medicine by Mark B. Faries, MD, of The Angeles Clinic and Research Institute, Los Angeles, and colleagues.1 Sentinel-node biopsy has been associated with improved melanoma-specific survival in patients with node-positive intermediate-thickness melanoma. 

Study Details 

IN THE international trial, 1,934 patients with sentinel-node metastases detected by standard pathologic assessment or reverse-transcriptase polymerase chain reaction (RT-PCR) from 63 sites were randomized between December 2004 and March 2014 to receive immediate completion lymph node dissection (dissection group; n = 967) or nodal observation with ultrasonography (observation group; n = 967). The primary endpoint was melanoma-specific survival. For survival data, comparisons between the two groups were performed using means of the log-rank test in univariate testing and Cox regression for adjusted analysis. 

Immediate completion lymph node dissection increased the rate of regional disease control and provided prognostic information but did not increase melanoma-specific survival among patients with melanoma and sentinel-node metastases.
— Mark B. Faries, MD, and colleagues

Baseline characteristics were similar for the per-protocol dissection vs observation groups. Of patients randomized to immediate dissection, 824 received assigned treatment (140 of the 143 excluded from the per-protocol analysis declined assigned treatment). Of those randomized to observation, 931 received assigned treatment. 

Patients in the observation group underwent clinical examination every 4 months during the first 2 years, every 6 months during years 3 to 5, and annually thereafter, with nodal ultrasonography of the sentinel-node basin performed at each visit for the first 5 years. Follow-up in the dissection group was according to the same schedule, without protocol-specified nodal ultrasonography. 

Melanoma-Specific Survival 

AT THE THIRD INTERIM ANALYSIS, the data and safety monitoring board determined that detection of a significant melanoma-specific survival difference was unlikely and recommended that the current primary endpoint data be released. Intention-to-treat and per-protocol analyses of outcomes yielded similar results. The results of the per-protocol analysis are reported in the body of The New England Journal of Medicine article, as they were considered by the authors to be the most clinically relevant data. 

In the per-protocol analysis, there was no significant difference between the dissection group and observation group in melanoma-specific survival. At a median follow-up of 43 months, the mean (± standard error) 3-year rate of melanoma-specific survival was 86 ± 1.3% vs 86 ± 1.2% (P = .42). No difference was observed after adjustment for other prognostic factors (adjusted hazard ratio [HR] = 1.08; P = .42). There was no significant difference in melanoma-specific survival according to whether sentinel node metastasis was identified by RT-PCR (P = .35) or by pathologic assessment (P = .47). 

Disease-Free Survival 

AT 3 YEARS, the disease-free survival rate was increased in the dissection group (68% ± 1.7% vs 63% ± 1.7%; P =.05), reflecting an improved rate of 3-year disease control in the regional nodes (92% ± 1.0% vs 77% ± 1.5%; adjusted HR = 0.31; P < .001). The authors stated, however, that the results of this secondary outcome analyses must be viewed with caution, given the lack of significance for the primary endpoint. 

No difference in distant metastasis–free survival was observed (adjusted HR = 1.10; P = .31). In the dissection group, nonsentinel-node metastases were identified on pathologic assessment in 11.5%, with rates increasing to 17.9% at 3 years and 19.9% at 5 years. In the observation group, ultrasonography or physical examination identified nonsentinel-node metastases in 22.9% at 3 years (P = .02) and 26.1% at 5 years (P = .005). 

Prognostic Factors 

ON MULTIVARIATE ANALYSIS that included only patients with sentinel-node metastases identified on pathologic analysis, significant prognostic factors for poorer melanoma-specific survival included increasing Breslow thickness and ulceration in both the dissection and observation groups. In the dissection group, nonsentinel-node metastasis was an independent prognostic factor for melanoma-specific survival (HR for death = 1.78; P = .005). The number of involved sentinel nodes was not a prognostic factor in either group. 


ADVERSE EVENTS were more common in the dissection group. At the most recent follow-up in April 2016, lymphedema had occurred in 24.1% of the dissection group vs 6.3% of the observation group (P < .001). Among all patients, lymphedema was mild in 64%, moderate in 33%, and severe in 3%. 

The investigators concluded: “Immediate completion lymph-node dissection increased the rate of regional disease control and provided prognostic information but did not increase melanoma-specific survival among patients with melanoma and sentinel-node metastases.” ■

DISCLOSURE: The study was funded by the National Cancer Institute and others. For full disclosures of the study authors, visit 


1. Faries MB, Thompson JF, Cochran AJ, et al: Completion dissection or observation for sentinel-node metastasis in melanoma. N Engl J Med 376:2211-2222, 2017