INTRODUCTION — Surgical excision is the appropriate treatment for melanoma. Although resection usually controls the primary lesion, melanoma often metastasizes through lymphatic channels to regional lymph nodes. Accurate staging at diagnosis is important to assess the prognosis and determine whether the patient would benefit from adjuvant therapy or is eligible for clinical trials (table 1A-B). (See "Tumor node metastasis (TNM) staging system and other prognostic factors in cutaneous melanoma".)
The physical examination of regional lymph nodes is often inaccurate, since approximately 20 percent of clinically node-negative patients have metastatic involvement, and 20 percent of those with clinically positive nodes are pathologically negative. More definitive information about the status of the regional nodes can be obtained from sentinel lymph node biopsy (SLNB), elective lymph node dissection, or fine needle aspiration [1,2].
The staging of subclinical disease using lymphatic mapping followed by SLNB, the management of patients with a positive SLNB, and the management of clinically apparent regional lymph nodes will be reviewed here. The surgical management of the primary lesion and adjuvant therapy are discussed separately. (See "Initial surgical management of melanoma of the skin and unusual sites" and "Adjuvant therapy for cutaneous melanoma".)
CLINICALLY NEGATIVE REGIONAL LYMPH NODES
Sentinel lymph node biopsy — Lymphatic mapping and sentinel lymph node biopsy (SLNB) is the standard approach for the management of patients with melanoma in whom there is a substantial risk of regional node metastasis. This approach provides important prognostic information and permits the identification of patients with a positive sentinel lymph node who may be candidates for adjuvant therapy. Proper application of this technique requires sufficient experience to identify the sentinel lymph nodes and ensure a low false negative rate.
Rationale — Lymphatic mapping is based upon the concept that sites of cutaneous melanoma have specific patterns of lymphatic spread and that one or more nodes are the first to be involved with metastatic disease within a given lymph node basin (figure 1). If the sentinel lymph nodes are not involved, the entire basin should be free of tumor [3-5].
Knowledge of the regional lymph node status is important for several purposes [6,7]:
●To determine prognosis
●To facilitate regional disease control in patients with regional lymph node involvement with possible improvement in survival 
●To select patients who may benefit from adjuvant therapy
●To select candidates for clinical trials
Tc-99 can be used for both preoperative mapping and for sentinel node identification in the operating room using a hand-held gamma probe [3,9]. The technique for SLNB is discussed elsewhere. (See "Imaging studies in melanoma", section on 'Lymph node evaluation'.)
Lymphatic mapping is particularly useful to identify the lymph node basins draining from primary sites located in areas of ambiguous and possibly multiple nodal basin drainage. These areas include the trunk, head and neck, and distal extremities. Dynamic lymphatic flow studies have shown that ambiguous and multidirectional flows are more common than previously thought [10,11].
While a learning curve exists for the technical aspects of SLNB, the importance of a coordinated and consistent approach, including surgeon, nuclear medicine clinician, and pathologist cannot be overstated.
In contrast to the widespread acceptance of lymphatic mapping and sentinel lymph node biopsy for extremity and truncal melanomas, the role of this technique for melanomas of the head and neck is evolving. The use of this approach has been limited by the complexity of lymphatic drainage patterns and the frequent need to remove sentinel nodes from the parotid gland, thereby risking damage to the facial nerve. That being said, sentinel nodes can be identified in 95 percent of cases with appropriate imaging techniques . (See "Initial surgical management of melanoma of the skin and unusual sites", section on 'Skin of head and neck'.)
Sensitivity and specificity — In the initial feasibility studies, SLNB was followed by completion lymph node dissection in all patients. The sentinel node was successfully identified in 75 to 90 percent of cases, and a negative SLNB with a positive nonsentinel node occurred in 1 to 2 percent [4,5]. In subsequent studies, SLNB was followed by completion lymph node dissection only if the sentinel node contained tumor. In a review of the literature and meta-analysis involving a much larger recent experience, sentinel nodes were successfully identified in 97 to 98 percent of patients, with metastases in the sentinel node being detected at the frequency expected, based upon the incidence of regional disease observed in earlier studies .
The false-negative rate in a meta-analysis that included data on 25,000 patients in 71 studies was 12.5 percent . The false negative rate needs to be distinguished from the negative predictive value (the number of patients with a negative SLNB who do not recur divided by the total number of patients with a negative SLNB), which depends upon the patient population studied.
The test characteristics for SLNB are illustrated by results from the Sunbelt Melanoma Trial . In that trial, 59 patients had a late regional lymphatic recurrence and 486 had a positive SLNB, for an overall false negative rate of 11 percent, and a negative predictive value of 97 percent in that patient population.
Multicenter Selective Lymphadenectomy Trial-I — The landmark Multicenter Selective Lymphadenectomy Trial-I (MSLT-I) is the largest trial to address the role of lymphatic mapping with sentinel lymph node biopsy (SLNB) in determining prognosis and its impact on survival [8,15]. The subsequent Multicenter Selective Lymphadenectomy Trial-II (MSLT-II) provides important information on the management of patients found to have a positive SLNB. (See 'Multicenter Selective Lymphadenectomy Trial II' below.)
The final results of this trial confirmed the role of lymphatic mapping with SLNB as a prognostic tool. Furthermore, the trial demonstrated a significant melanoma-specific survival advantage in those patients with intermediate thickness melanoma with microscopic lymph node involvement who were assigned to lymphatic mapping with SLNB and underwent early regional lymphadenectomy, compared with those managed with wide excision of the primary melanoma followed by observation without lymphatic mapping and SLNB. The conclusions of the MSLT-I are also supported by the larger, retrospective series of 5840 patients in the Melanoma Institute Australia database treated between 1992 and 2008 .
In the MSLT-I, 2001 patients were randomly assigned to lymphatic mapping with SLNB (60 percent) or to observation (40 percent) between 1994 and 2002 .
●Patients randomly assigned to lymphatic mapping with SLNB underwent immediate completion lymphadenectomy if a histologically positive node was identified, while those with a negative SLNB were observed and underwent therapeutic lymphadenectomy if there was clinical evidence of subsequent nodal recurrence.
●Individuals assigned to observation had a therapeutic lymphadenectomy only if there was clinical evidence of nodal recurrence.
The final report of the trial was published in 2014, and was based upon ten year follow-up .
Prognostic significance of the sentinel node — The presence of microscopic involvement of the sentinel lymph node at presentation was a significant predictor of subsequent relapse in patients with intermediate or thick primary melanomas. There were too few events to analyze in those with thin melanomas.
●For patients with intermediate thickness melanoma (1.2 to 3.5 mm), the melanoma-specific survival rate at 10 years was significantly worse in patients with lymph node involvement compared with those who had a negative lymph node biopsy (62 versus 85 percent, HR 3.09, 95% CI 2.12-4.49).
●For patients with thick melanoma (>3.50 mm), the melanoma-specific survival rate at 10 years was significantly worse in patients with lymph node involvement compared with those who had a negative lymph node biopsy (48.0 versus 64.6 percent, HR 1.75, 95% CI 1.07-2.87).
Intermediate thickness melanomas — The primary study group was defined as the 1347 patients with intermediate thickness melanomas (1.2 to 3.5 mm), which was the group thought to be most likely to benefit from lymphatic mapping with SLNB . In this group, lymphatic mapping with SLNB resulted in a significant improvement in melanoma-specific survival in those patients found to have lymph node involvement.
Incidence of nodal metastases — The cumulative incidence of regional lymph node involvement in patients with intermediate thickness melanoma was similar in the two groups :
●Among the 770 patients randomly assigned to lymphatic mapping with SLNB, a positive lymph node was found in 122 of 765 (16.0 percent) who had a sentinel lymph node. An additional 4.8 percent subsequently relapsed in the regional lymph nodes, for an estimated overall incidence of nodal involvement of 20.8 percent.
●Among the 500 patients randomly assigned to observation, 87 (17.4 percent) had a clinical relapse in regional lymph nodes, at a median of 19 months after randomization.
Melanoma-specific survival — For the entire group of patients with intermediate thickness melanoma randomly assigned to lymphatic mapping with SLNB, the difference in melanoma-specific survival was not statistically significant compared with those managed with observation (81.4 versus 78.3 percent; hazard ratio [HR] for death 0.84, 95% CI 0.64-1.09).
However, the 10-year melanoma-specific survival rate was significantly improved in those found to have nodal metastases who underwent SLNB compared with those initially managed with observation followed by treatment when clinical disease was detected (62.1 versus 41.5 percent; HR for death 0.56, 95% CI 0.37-0.84). The 10-year distant disease-free survival rate was also significantly improved in this subset of patients (54.8 versus 35.6 percent, HR 0.62 95% CI 0.42-0.91).
Thick melanomas — The study included 314 patients with thick melanomas (>3.50 mm), of whom 311 were available for per protocol analysis. Lymph node metastases were identified in 32.9 percent of patients with thick melanomas at SLNB and an additional 12 patients who were initially sentinel lymph node biopsy negative subsequently were found to have nodal disease. Thus the estimated incidence of disease at 10 years was 42 percent.
Overall, there was no significant difference between the two treatment approaches in melanoma-specific survival for the entire group of patients with thick melanoma (10 year melanoma specific survival rate 58.9 versus 64.4 percent for the lymphatic mapping with SLNB versus initial observation, respectively, HR 1.12, 95% CI 0.76-1.67).
When the analysis was limited to those patients in whom lymph node disease was present, the 10-year melanoma-specific survival rate was not significantly different in those who underwent SLNB compared with those initially managed with observation (48.0 versus 45.8 percent; HR for death 0.92, 95% CI 0.53-1.60). There also was no significant difference in the distant disease-free survival rate (45.3 versus 43.8 percent, HR 0.96, 95% CI 0.56-1.64).
Thin melanomas — The study included 340 patients with thin melanomas (<1.20 mm). However, there were too few events in this group to permit analysis of outcomes.
Complications of SLNB — SLNB is associated with significantly fewer complications than regional lymphadenectomy. This was illustrated in 2120 patients enrolled on the Sunbelt Melanoma Trial, of whom 444 underwent completion lymph node dissection . At a median follow-up of 16 months, the overall complication rate was significantly lower for SLNB alone (5 versus 23 percent with SLNB plus completion lymphadenectomy).
The decrease in the rate of complications included wound infection (1 versus 7 percent with completion lymphadenectomy), lymphedema (0.7 versus 11.7 percent), hematoma/seroma (2 versus 6 percent) and sensory nerve injury (0.2 versus 1.8 percent) . Differences in complication rates were particularly striking in patients who underwent groin procedures (total complications 8 versus 51 percent, and lymphedema 2 versus 32 percent).
Minimal metastases — Because fewer nodes are resected with SLNB compared with elective lymph node dissection, more careful pathologic evaluation is possible. Serial sectioning increases the likelihood of detecting melanoma metastases, although eight to ten sections may be required .
Even though studies have shown that low volume involvement of the SLN is only rarely associated with tumor involvement of other lymph nodes, follow-up suggests that minimal metastasis in the SLN confers an increased risk of relapse and death relative to patients with no SLN involvement.
The impact of extent of SLN involvement is illustrated by a study of 1009 patients with a positive SLN who underwent completion lymph node dissection over a fifteen-year period at nine European Organisation for Research and Treatment of Cancer (EORTC) melanoma centers . Among the 113 patients with SLN tumor size <0.1 mm, 10 (9 percent) were found to have tumor involvement of non-sentinel lymph nodes. The frequency of involvement of nonsentinel lymph nodes increased progressively with tumor size in the SLN (16 percent for those with 0.1 to 1.0 mm tumor and 25 percent for those with tumor >1.0 mm). On multivariate analysis, lesser tumor involvement was significantly associated with improved melanoma-specific survival.
However, the data are inadequate to distinguish subsets of patients based upon the extent of tumor involvement in the sentinel lymph node, and completion lymph node dissection is indicated for patients with a positive SLNB regardless of the extent of tumor involvement. These issues are being addressed in the Multicenter Selective Lymphadenectomy Trial II (NCT002978950).
However, a small number of SLNBs will be falsely reported as negative even with optimal technique. In the MSLT-I trial, for example, 3.4 percent of patients with a negative SLNB subsequently relapsed in the regional lymph nodes .
A similar rate of relapse was observed in a report from MD Anderson . Ten of 243 patients (4.1 percent) with histologically negative SLNB subsequently developed a nodal recurrence in the previously mapped basin. Reexamination of the original pathologic material using serial sections and/or immunohistochemical staining identified melanoma in the sentinel nodes in eight of these ten patients.
Immunohistochemical staining for the melanoma markers S100, HMB45, or Melan-A/Mart-1 further enhances sensitivity, detecting 1 melanoma cell in 100,000 cells compared with 1 in 10,000 cells with routine hematoxylin and eosin stains (H&E).
Patient selection — SLNB is the standard approach for the management of patients with melanoma in whom there is a substantial risk of regional node metastasis. The likelihood of detecting metastatic deposits in a SLNB increases with the thickness of the primary lesion, providing a rationale for deciding which patients may benefit from this procedure.
The generally accepted approach focuses on tumor thickness and related risk factors:
●SLNB is indicated for melanomas ≥1 mm thick. For lesions 1.01 to 2.0, 2.01 to 4.0, and >4 mm, the risk of regional lymph node metastasis is approximately 12, 28, and 44 percent, respectively .
•Tumors >4 mm thick have a risk of distant metastasis that can vary widely and is dependent on a variety of features including ulceration, primary location, patient age and gender. However, status of the SLNB may provide the most important prognostic information. For example, in one series of 571 patients with primary lesions >4 mm thick, the median overall, disease-specific, and relapse-free survival were significantly longer in biopsy-proven node-negative patients compared with those with positive nodes (53 versus 35 months, 32 versus 14 months) . This information may be important to the patient and clinician and will very likely influence the choice of adjuvant therapy either as standard care or as part of a clinical trial.
●For thin melanomas (ie, <1 mm thick), the overall risk of regional node metastases in patients undergoing SLNB is estimated to be about 5 percent [22,24]. Various prognostic factors have been studied in an effort to identify subgroups in which the risk of nodal disease is high enough to justify SLNB.
A retrospective review from the Sentinel Lymph Node Working Group of 1250 patients with thin melanoma who underwent SLNB provides the most extensive data in this group . Breslow thickness ≥0.75 mm, Clark level ≥IV, and ulceration were associated with lymph node metastases in 6.3, 7.0, and 11.6 percent of cases, respectively, and these factors were all statistically significant on multivariate analysis. In contrast, melanomas <0.75 mm thick had positive SLNBs less than 5 percent of the time regardless of the Clark level or the presence of ulceration. The low risk of a positive SLNB in patients with a primary lesion <0.75 mm and without any other risk factors has also been seen in earlier studies .
Management of a positive SLNB — Two randomized trials comparing immediate completion lymph node dissection with observation followed by lymph node dissection in the event of a regional lymph node recurrence have not demonstrated any improvement or a significant difference in melanoma-specific survival or distant metastasis-free survival [26,27].
Although completion dissection of all involved nodal basins previously was considered the standard treatment approach for patients with a positive sentinel lymph node biopsy, the results of these two randomized trials support observation with careful follow-up that includes serial ultrasonography in this setting.
Multicenter Selective Lymphadenectomy Trial II — The phase III MSLT-II trial included 1934 evaluable patients who had a positive sentinel lymph node biopsy as well as a wide local excision of a primary melanoma that had a Breslow thickness ≥1.20 mm or greater and Clark level III, a Clark level IV or V regardless of Breslow thickness, or a primary tumor with ulceration regardless of thickness or Clark level .
Patients were randomly assigned to either completion lymph node dissection or observation, which importantly, included ultrasound evaluation of appropriate lymph node basins prior to lymphatic mapping and sentinel node biopsy and at each follow-up visit. Completion lymph node dissection was performed if there was evidence of regional lymph node recurrence. Follow-up for all patients was carried out every four months for two years, every six months for years 3 to 5, and then annually. Most patients had low-volume (<1 mm) metastasis deposits in the positive sentinel node.
At a median follow-up of 43 months, key results included the following:
●The per-protocol analysis included 824 patients who underwent immediate completion lymph node dissection and 931 who were managed with observation. Following randomization, 140 people refused surgery. In a separate intention to treat analysis, there was also no difference in results.
●Melanoma-specific survival, the primary endpoint of the trial, was the same for both immediate lymph node dissection and observation (three-year rate 86 versus 86 percent, adjusted HR 1.08, 95% CI 0.88-1.34).
●Disease-free survival at three years was improved in patients managed with immediate lymph node dissection (68 versus 63 percent), reflecting a lower rate of recurrence in regional lymph nodes in patients undergoing immediate lymph node dissection compared with observation (8 versus 23 percent).
●The incidence of lymphedema was higher in patients who underwent immediate lymph node dissection (24.1 versus 6.3 percent).
●Although melanoma-specific survival was worse in patients with thicker primaries, ulceration of the primary tumor, or positive nonsentinel nodes, no subgroups could be identified in whom completion lymphadenectomy provided benefit.
DeCOG-SLT trial — In the multicenter DeCOG-SLT trial, 483 patients, again most with low-volume metastasis in the sentinel node, were randomly assigned to immediate surgery or to observation including ultrasound of the primary site and appropriate lymph node basins, and 473 were included in the intention to treat analysis . At a median follow-up of 35 months, there was no significant difference in the three-year distant metastasis-free survival rate, the primary endpoint of the trial (77 versus 75 percent for the observation versus immediate complete dissection groups, HR 1.03, 95% CI 0.70-1.50). There were significantly fewer regional recurrences with immediate surgery (20 versus 34). Longer follow-up is required to understand the true impact on distant metastasis-free survival.
The staging evaluation of a patient with a positive regional lymph node is discussed separately. (See "Staging work-up and surveillance after treatment of melanoma", section on 'Positive sentinel lymph node'.)
The role of adjuvant immunotherapy following completion lymph node dissection is discussed separately. (See "Adjuvant therapy for cutaneous melanoma".)
Elective lymph node dissection — Elective lymph node dissection was advocated as an approach to the regional lymph nodes prior to the development of SLNB. However, the availability of lymphatic mapping with SLNB and its success in predicting regional lymph node involvement have obviated a possible role for elective lymph node dissection.
Guidelines — Joint guidelines from the American Society of Clinical Oncology (ASCO) and the Society for Surgical Oncology are consistent with the above approach, recommending SLNB for patients with intermediate thickness (1 to 4 mm) melanomas as well as suggesting SLNB for patients with thin melanomas (<1 mm) that have other high risk features [6,7]. Those guidelines also suggest considering SLNB for patients with thick melanoma (>4 mm) for prognostic purposes, to aid in achieving regional disease control, and to identify patients for adjuvant therapy or investigational trials.
CLINICALLY APPARENT REGIONAL LYMPH NODES
Therapeutic lymphadenectomy — Surgical (therapeutic) lymphadenectomy is the preferred treatment for clinically detectable and cytologically (fine needle aspirate) or pathologically proven regional lymph node involvement in patients with melanoma . This approach is associated with long term disease free survival in a subset of patients. Even if patients subsequently develop distant metastases, aggressive regional therapy at presentation can prevent morbidity caused by mass effect from involved nodes or skin breakdown.
Approximately 20 to 40 percent of patients with clinically apparent (macroscopic, N1b or N2b) metastatic involvement of regional nodes are alive at 10 years following therapeutic lymphadenectomy . The number of metastatic lymph nodes is a significant prognostic factor; patients with only one positive node have a better prognosis (40 to 50 percent) than those with more than one positive node . (See "Tumor node metastasis (TNM) staging system and other prognostic factors in cutaneous melanoma".)
Extent of dissection and morbidity — Complete regional lymphadenectomy, rather than partial dissection or sampling, is necessary for melanoma because of the high risk of involvement of other lymph nodes within the basin or extranodal spread.
The morbidity of completion axillary dissection (ie, level I, II, and III nodes) for upper extremity melanomas or truncal melanomas with lymphatic drainage to the axilla includes short- and long-term complications. Common short-term complications include wound infection and breakdown, seroma formation, and shoulder dysfunction after axillary dissection (table 2) [18,30]. Long-term complications include lymphedema and paresthesias.
Complications are more common after inguinal lymph node dissection (for lower extremity lesions or truncal melanoma with inguinal drainage) than after axillary nodal dissection, with lymphedema being a frequent problem [18,30]. Some surgeons electively add a deep ilioinguinal dissection to the superficial inguinal node dissection when the highest superficial node (Cloquet's node ) contains metastatic melanoma, when multiple superficial inguinal nodes contain melanoma, and when imaging suggests deep ilioinguinal metastatic disease.
The role of a deep ilioinguinal dissection is controversial, since it is not clear whether the addition of a more extensive dissection improves survival [31,32]. The addition of the deep ilioinguinal lymphadenectomy significantly increases the risk of subsequent lymphedema . Patients undergoing inguinal node dissection usually benefit from compression support stockings for at least six months postoperatively. Prevention of lymphedema is critical since established lymphedema requires lifelong treatment. (See "Clinical staging and conservative management of peripheral lymphedema".)
INFORMATION FOR PATIENTS — UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or e-mail these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Beyond the Basics topics (see "Patient education: Melanoma treatment; localized melanoma (Beyond the Basics)" and "Patient education: Melanoma treatment; advanced or metastatic melanoma (Beyond the Basics)")
SUMMARY AND RECOMMENDATIONS
●Lymphatic mapping with sentinel lymph node biopsy (SLNB) is the standard clinicopathologic approach to evaluate regional lymph nodes in patients without clinical evidence of lymph node involvement. (See 'Sentinel lymph node biopsy' above.)
●For patients with clinically negative regional lymph nodes, the decision whether or not to perform an SLNB is based upon the likelihood of regional lymph node involvement: (See 'Patient selection' above.)
•For patients with a melanoma 0.75 mm thick or greater, we recommend performing an SLNB. This criterion includes patients with intermediate and thick primary lesions, as well as those with thin melanomas 0.75 to 1.00 mm thick. These patients should have nodal basin ultrasonography prior to lymphatic mapping and sentinel lymphadenectomy.
•For patients with a melanoma less than 0.75 mm thick and one or more factors placing the patient at increased risk for nodal involvement (ulceration of the primary tumor, a mitotic rate ≥1/mm2, or lymphovascular invasion), we favor performing an SLNB. However, the absolute risk of a positive sentinel lymph node is small in this group, and a detailed discussion with the patient is needed to discuss the potential risks and benefits.
•SLNB is not indicated for patients with a melanoma <0.75 mm and no other factors placing the patient at increased risk for nodal involvement.
●For patients with a positive SLNB, we suggest careful clinical observation coupled with ultrasound surveillance of the positive nodal basin rather than immediate completion lymph node dissection (Grade 2B). Completion lymph node dissection is indicated if there is subsequent evidence of regional lymph node recurrence in the absence of distant metastases. (See 'Management of a positive SLNB' above.)
●In patients with microscopic involvement of a single lymph node (N1a disease), we suggest adjuvant immunotherapy with nivolumab (Grade 2B). If there is macroscopic involvement of a lymph node or if there is involvement of more than one node (N1b, N2, or N3 disease), we recommend adjuvant immunotherapy with nivolumab (Grade 1B). (See "Adjuvant therapy for cutaneous melanoma".)
●For patients who present with clinically apparent regional lymph node involvement that is confirmed cytologically (fine needle aspirate) or histologically, we recommend therapeutic regional lymphadenectomy (Grade 1B). Partial dissection or lymph node sampling is not considered an acceptable alternative because of the propensity of melanoma to spread microscopically to other nodes in the basin. (See 'Clinically apparent regional lymph nodes' above.)