ST. LOUIS - Non–sentinel node status is an independent prognostic factor for disease-free survival and overall survival in patients with melanoma, results from a large study with long-term follow-up demonstrated.
“This effect is not simply a result of greater numbers of positive nodes,” Dr. Russell E. Brown said at a symposium sponsored by the Society of Surgical Oncology.
The findings support the notion that the sentinel lymph node (SLN) acts as the first line of defense against metastasis in melanoma and that metastasis past the SLN predicts a poorer prognosis.
“Status of sentinel lymph nodes is the most powerful predictor of outcomes for early-stage melanoma,” said Dr. Brown, a surgical oncology fellow at the University of Louisville, Kentucky. “Non–sentinel nodes (NSN) are evaluated at completion of lymph node dissection and are found in 8%-33% of complete lymph node dissection specimens.”
Multiple factors are used to predict NSN metastases, he said, including include age, tumor thickness, number of sentinel lymph nodes, SLN tumor burden, ulceration, male gender, perinodal intralymphatic tumor, and extranodal extension. While single-institution studies have shown a survival disadvantage for NSN-positive patients, Dr. Brown and his associates hypothesized that metastasis beyond the SLN to the non–sentinel nodes is an important predictor of survival.
To test their hypothesis, the researchers performed a post hoc analysis of data from 2,335 patients in the Sunbelt Melanoma Trial, a prospective, randomized study of patients aged 18-70 years at 79 centers in the United States and Canada. The Sunbelt Melanoma Trial was the world’s largest prospective study of melanoma, led by Dr. Kelly McMasters, chair of the department of surgery at the University of Louisville. Patients were treated between 1997 and 2003 and were clinically node negative, with primary tumors at least 1 mm thick. All subjects underwent SLN biopsy and were followed for a median of 68 months.
The researchers used Kaplan-Meier analysis and univariate and multivariate analysis to compare disease-free survival and overall survival in three groups of patients: 1,988 who were SLN negative (group 1), 296 who were SLN-only positive (group 2), and 51 who were SLN and NSN positive (group 3). Dr. Brown reported that the 5-year disease-free survival rates for groups 1, 2, and 3 were 86%, 65%, and 43%, respectively, while the 5-year overall survival rates were 86%, 65%, and 49%, respectively.
Univariate analysis revealed the following significant independent predictors of decreased disease-free survival: SLN positivity (hazard ratio of 2.4), NSN positivity (HR 4.57), increased total number of positive lymph nodes (HR 1.18), increased ratio of positive lymph node count to total lymph node count (HR 1.03), increased Breslow thickness (HR 3.28), presence of ulceration (HR 2.76), Clark level of 4 or greater (HR 1.65), axial primary (HR 1.45), and male gender (1.31).
When the total number of positive lymph nodes and NSN status were evaluated in a multivariate Cox model with other univariate predictors of decreased disease-free survival, NSN status remained statistically significant (HR 3.95), while the total number of positive lymph nodes and lymph node ratio did not (both were an HR of 1.01).
Univariate analysis also revealed the following significant predictors of decreased overall survival: SLN positivity (HR 2.29), NSN positivity (HR 3.6), increased total number of positive lymph nodes (HR 1.16), increased ratio of positive lymph node count to total lymph node count (HR 1.03), increased Breslow thickness (HR 2.96), presence of ulceration (HR 2.92), Clark level of 4 or greater (HR 1.41), axial primary (HR 1.71), male gender (1.68), and increased age (HR 1.24).
“The results of Dr. Brown’s study are important and suggest that there is a biological significance to the sentinel nodes as the first line of defense against metastasis in melanoma,” said Dr. McMasters, a melanoma specialist.
When the total number of positive lymph nodes and NSN status were evaluated in a multivariate Cox model with other univariate predictors of decreased overall survival, NSN status remained statistically significant (HR 2.81) while the total number of positive lymph nodes and lymph node ratio did not (HR of 1.03 and 1.11, respectively).
Dr. Brown noted that lymphovascular invasion, the number of involved nodal basins, the presence of a vertical growth phase, and evidence of regression were not significant predictors of overall survival.
Dr. McMasters has received grant support for the study from Schering Oncology/Biotech and has been on the company’s speakers bureau.
Dr. Brown had no relevant financial disclosures.