Patients with thin melanomas comprise a majority of those encountered in clinical practice [45]. Generally, patients with thin melanoma have a favorable prognosis and low risk of metastasis and melanoma-related death. Primary tumors less than 1 mm in thickness have less than a 5 % risk of metastasis overall [46]. It is recognized, however, that even for those with thin melanoma, a certain percentage of patients may recur.

The routine use of SLNB in thin melanoma is controversial. Determination of which patients are at highest risk for nodal recurrence and possibly subsequent melanoma-related death is important given the number of new melanoma diagnoses yearly in this category. Several potential predictors of SLN metastasis have been identified including but not limited to gender, Breslow thickness, Clark level, age, mitotic rate and tumor infiltrating lymphocytes [47–51]. The mitotic rate has been identified to be an extremely important variable in the outcome of thin melanoma, so much so that it was added to the seventh edition of the AJCC staging system. However, by itself, mitotic rate has not been shown to be a predictor of nodal metastases in the AJCC analysis [4, 52–54]. Ulceration has been associated with higher propensity for nodal metastasis and overall worse outcome. While the presence of ulcerative lesions warrants further investigation, it is an infrequent finding in thin melanoma, affecting less than 3 % of patients [55, 56]. Age is also a risk factor for thin melanoma and several predictive models identify it as prognostic for increased risk of nodal involvement [56, 57]. No appropriate age cut off has been identified of when to offer SLNB and therefore cannot be used alone to pursue the procedure. Interestingly, while prognosis is worse with advancing age, the incidence of nodal metastases is lower. Similarly, while prognosis is better with younger age, the incidence of nodal metastases is higher [39, 58]. Nomograms are clinically helpful in identifying which patients with thin melanoma may be at increased risk for lymph node metastasis and may benefit from SLNB. However, while nomograms may be useful, it should also be recognized that a surgeon selection bias exists in which patients with thin melanoma are considered for SLNB testing and therefore may have an impact on which factors are ultimately included as selection criteria.

There has been considerable debate on the management of thin melanoma (AJCC Stage IA/IB) with respect to SLNB testing. The NCCN guidelines recommend SLNB for melanoma ≥ 1.0 mm in thickness. For melanomas less than 0.76 mm without adverse features, SLNB is not recommended. For lesions 0.76–1.0 mm, discussion of SLNB is at the discretion of the treating physician and patient [16, 59]. Recently, the American Society of Clinical Oncology (ASCO) and Society of Surgical Oncology (SSO) published consensus guidelines with evidence-based recommendations for SLNB in melanoma [60]. Four key recommendations were issued. With regard to thin melanomas, specifically, they reported there to be insufficient evidence to support routine SLNB for melanoma <1 mm in Breslow thickness, but suggest that it may be considered in select high-risk individuals.

Discussion of SLNB for patients with thin melanoma is highly individualized and must take into account the available pathology report, the patient’s level of anxiety and expectations and growing body of emerging published data. Long-term data on outcome for SLNB in thin melanoma requires longer follow up and randomized clinical trials to clarify this issue.

There is agreement and uniformity on the role of SLNB for intermediate thickness melanoma. Our understanding of the importance of SLNB on staging and prognosis and ultimately overall outcome is due to the work of Dr. Morton and colleagues of the MSLT Cooperative Group [32, 61, 62]. A large, multicenter randomized trial was conducted in which patients with clinically localized melanoma were assigned to wide excision and sentinel-node biopsy (biopsy group) or wide excision and postoperative observation of the regional nodal basin (observation group). Patients in the biopsy group underwent immediate lymphadenectomy if micro-metastasis were detected at the time of SLNB. Patients in the observation group underwent delayed lymphadenectomy if nodal recurrences developed. Patients with melanomas 1.2–3.5 mm were selected as the primary study group based on prospective database analysis from the John Wayne Cancer Institute [63]. With a median follow-up of 59.8 months, there was no difference in melanoma-specific survival between groups (86.6 % and 87.1 %; p = 0.58). For patients with a positive sentinel node, the disease-free survival was 53.4 %, compared to 83.2 % if the sentinel node was free of metastases (p < 0.001).

The third interim analysis suggested a survival benefit for patients with intermediate-thickness melanoma who were found to be node-positive and underwent immediate lymphadenectomy (72.3 % vs 52.4 % 5-year survival) [43]. Seventy-eight patients in the observation group developed nodal relapse in the regional basin. The median time to development of nodal recurrence was 1.33 years. Furthermore, the mean number of clinically involved tumor-positive nodes in patients who underwent delayed lymphadenectomy (observation group) was 3.3 compared to 1.4 involved nodes for those who underwent immediate lymphadenectomy (p < 0.001), suggesting that even in the absence of clinically evident disease, micro-metastatic disease continues to grow and if left untreated will eventually develop into significant disease.

Subsequent long-term follow-up data was recently reported [64]. There was no difference in 10-year melanoma-specific survival between the biopsy or observation group. The overall rate of nodal disease was 20.8 %, suggesting that 79.2 % of patients did not derive a benefit from SLNB. On subset analysis, for patients with nodal disease and intermediate-thickness melanoma (1.2–3.5 mm), early treatment following a positive SLNB was associated with improved 10-year distant DFS and melanoma-specific survival.

Others have demonstrated similar findings with the use of SLNB testing for intermediate thickness melanoma. Post hoc analysis of study data from patients enrolled in the Sunbelt Melanoma Trial was performed looking at the use of SLNB in melanoma intermediate thickness (1–2 mm) melanoma [65]. Over 1100 patients were evaluated and divided into two groups: Group A (1.0–1.59 mm) and Group B (1.60–2.0 mm). The SLN was positive in 133 (12 %), including 8.7 % in Group A (66/672) and 19.3 % of Group B (67/348). Patient age, Breslow thickness and lymphovascular invasion were independently predictive of a positive SLN on multivariate analysis. The DFS and OS were significantly better for Group A than Group B. Based on review of the data, the authors were not able to identify or reasonably predict which patients with melanoma between 1 and 2 mm in thickness would be at minimal risk for SLN metastasis and therefore unlikely to benefit from SLNB. The authors recommend SLNB for all patients with intermediate thickness melanoma.

Just as with the role of SLNB in thin lesions, there is lack of unanimity with the use of SLNB in thick (>4 mm) melanoma. The risk of SLN metastasis increases as primary tumor thickness increases, with tumors greater than 4 mm in depth having a risk of metastasis of greater than 30 % [33, 66]. Some clinicians, however, contend that SLNB may not provide the same benefit in patients with thick melanoma as these patients have a high rate of occult systemic disease at the time of presentation [67]. The SLN would not provide useful prognostic information and completion lymph node dissection would not have a significant impact on outcome in the presence of distant disease.

Several series have demonstrated conflicting results with the use of SLNB in thick melanomas. For instance, Jacobs et al. identified no statistical difference in overall survival (OS) in 43 patients with thick melanomas (median thickness 6.4 mm) [68]. Essner and colleagues, similarly, identified no difference in OS but there was a disease free survival (DFS) observed for node-negative patients with thick melanomas who underwent SLNB (N = 135 patients, median thickness 5.9 mm) [69]. Cherpelis also identified no difference in OS between node-negative and node-positive patients as identified by SLNB in thick melanomas (N = 201 patients, median thickness 5.2 mm) [70]. These results have led some clinicians to suggest that the routine use of SLNB in thick melanomas may not have the same clinical significance as in intermediate thickness melanoma.

Other series have demonstrated a benefit to the routine use of SLNB in thick melanoma. Gershenwald and colleagues at MD Anderson Cancer Center identified the status of the SLN (along with ulceration) to be the most significant predictor of DFS and OS in 131 patients with T4N0 melanoma and recommended routine use of SLNB in this patient population [66]. Similarly, Carlson and colleagues from Emory University, identified that 37 of 114 patients (32.5 %) with thick melanomas had a positive SLNB, 18 patients (48.6 %) of which had a single tumor-positive lymph node after lymphadenectomy [71]. The status of the SLN was the strongest independent predictor of OS. The authors recommended routine SLN mapping for those with thick (≥ 4 mm) melanoma. The authors of a study from the University of Michigan reported on one of the largest experiences with the use of SLNB for patients with thick melanoma [72]. Of 227 patients with thick melanomas, 107 (47 %) were found to be SLN-positive. Angiolymphatic invasion, satellitosis and ulceration of the primary tumor were the strongest predictors of a positive-SLN. The SLN status was the most significant predictor of distant DFS (DDFS) and OS in that population. Patients with T4 melanoma who were node-negative had a significantly better DDFS (85 % vs 48 %, p < 0.0001) and OS (80 % vs 47 %, p < 0.0001) compared to those who were found to have metastasis on SLNB. The authors recommended strongly considering SLNB, regardless of Breslow depth, for patients with clinically node-negative T4 melanoma.

The use of SLNB for patients with thick melanoma is generally less controversial than for thin melanoma. While there is some conflicting data regarding the impact and outcome for patients who undergo SLNB with thick melanoma, the standard practice for most high-volume melanoma centers is offer SLNB and treat based on results of the biopsy. No randomized clinical trials are currently underway addressing this issue specifically.

Lymphadenectomy is the standard of care for patients with clinically node-positive or SLN-positive melanoma [4, 16]. For patients who have a positive SLNB, up to 15 % of patients will have occult disease identified in non-SLNs at the time of completion of lymphadenectomy [78, 79]. Perhaps more accurately, data from MSLT-1 has demonstrated that 88 % of patients who have a single tumor-containing sentinel node will have no additional metastasis identified at the time of CLND by hematoxylin and eosin staining [43]. This suggests that the majority of patients derive no benefit in undergoing CLND and, therefore, undergo a potentially unnecessary procedure with concomitant increased risk for complications. While the NCCN recommendation is for patients with a positive SLN to undergo CLND, at least one study demonstrated that up to 50 % of patients forego CLND [80]. Currently, the natural history of patients with a positive-SLN who do not undergo CLND is unknown.

Because a majority of patients have no additional disease identified in non-sentinel lymph nodes (NSLNs), several centers have begun to assess predictive factors that may provide information on who may avoid CLND after a positive SLNB. Sabel and authors from University of Michigan queried their prospective melanoma database and identified 980 patients who underwent SLNB for cutaneous melanoma [81]. A positive SLN was identified in 24 % of patients (232). At CLND, 34 patients (15 %) had one or more positive NSLN. Three or more positive SLNs, male gender, Breslow thickness and extranodal extension were all associated with likelihood of finding additional positive nodes on CLND. The authors at Emory University reviewed their experience of 70 patients with a positive SLN and drainage to a single nodal basin [82]. Nineteen patients (24 %) were found to have NSLNs after CLND. Breslow thickness, ulceration, SLN tumor burden, number of positive SLNs and number of SLNs removed were examined and a predictive model developed to identify positive NSLNs was developed. Neither comparison of the tumor factors examined, nor the predictive model could accurately predict NSLN involvement. Others have been likewise unsuccessful in identification of a group of patients at zero-risk for NSLN metastasis when using algorithms or predictive models [83].

Frankel and colleagues examined whether size and location of metastases within the SLN may help better stratify the likelihood of finding additional positive NSLNs [84]. The presence of a head/neck or lower extremity primary, angiolymphatic invasion, mitosis, Breslow thickness >4 mm, extranodal extension, ≥3 positive SLNs and tumor burden involving >1 % of SLN surface area were significantly associated with finding additional disease on CLND. Location of metastases within the SLN (capsular, subcapsular, or parenchymal) did not correlate with a positive NSLN. The Swiss performed a retrospective analysis of 392 patients and investigated whether SLN tumor load had an effect on NSLN positivity or DFS, possibly sparing some patients from CLND [85]. A total of 114 positive SLNs were identified and at the time of CLND, 22 % were found to have additional positive NSLNs. Of those with SLN micrometastasis, 16.4 % had a positive NSLN identified at CLND (p = 0.09). SLN tumor burden, however, did not correlate with NSLN-positivity. Similarly, the authors were not able to reliably or reproducibly predict NSLN-positivity at the time of CLND.

The lack of uniform results has led some authors to question the utility of CLND for positive SLN patients. Kingham and authors reviewed the Memorial Sloan-Kettering Cancer Center experience for patients who hade a positive SLN and did not undergo CLND [86]. Of 2269 patients who underwent SLNB, 313 (13.7 %) had a positive SLN of which 271 (87 %) underwent CLND and 42 (13 %) did not (no-CLND). Patients in the no-CLND group were older, had a higher percentage of lower extremity melanomas, and a trend toward thicker melanomas. The most common reason (45 %) for not performing CLND was refusal by the patient. The patterns and rates of recurrences were similar between groups, suggesting that possibly CLND may not need to be performed in all melanoma patients with a positive SLN. Bamboat and colleagues recently reported updated information on 4310 patients undergoing SLNB over a 20-year period [87]. A positive SLN was observed in 495 (11 %) of which 328 (66 %) underwent immediate CLND and 167 (34 %) underwent nodal observation. There were no differences in Breslow thickness, Clark level, ulceration or SLN tumor burden between groups. Nodal disease was the site of first recurrence in 15 % of patients in the no-CLND group and 6 % of the CLND group (p = 0.002). There was no difference in local and in-transit recurrence between groups. Systemic recurrences occurred in 8 % of the no-CLND patients compared with 27 % of the CLND patients (p < 0.001). Immediate CLND after a positive-SLNB was associated with fewer initial nodal basin recurrences but no difference in melanoma-specific survival when compared with those who were observed and did not undergo CLND. The authors concluded that these results further validate the ongoing, pending results of the Multicenter Selective Lymphadenectomy Trial II (MSLT-II) .

MSLT-II is a phase III multicenter, randomized trial of SLNB followed by CLND or SLNB followed by observation for node-positive melanoma (Clinicaltrials.gov, NCT000297895). Because most patients with SLN positive disease have no additional nodal involvement, this suggest that nodal metastasis may be limited to only 1 or 2 sentinel nodes and that SLNB may be therapeutic as well as diagnostic. The underlying hypothesis of MSLT-II is that CLND can be avoided in most patients with SLN metastasis. Enrollment of a planned 1925 subjects began in 2005. SLN positive patients are randomized to CLND or ultrasound observation (+ delayed CLND if recurrence is detected) and are followed for 10 years with a primary outcome measure of melanoma-specific survival. The results from this large clinical trial will hopefully provide valuable insight and demonstrate the true impact of CLND for node-positive melanoma.

For metastasis to the inguinal lymph nodes, disagreement exists about the extent of surgical dissection required. Specifically, whether a superficial (inguinal) groin dissection is sufficient or whether a combined superficial and deep (ilioinguinal) LND is necessary. Arguments against performing a deep groin dissection are valid as complication rates after combined superficial and deep groin dissection have been reported to be as high as 50 % [88]. Furthermore, several studies support the argument that ilioinguinal metastatic involvement represents more systemic disease and an aggressive surgical approach or the extent of surgery may not have an impact on outcome [89, 90]. Conversely, others support the routine surgical practice of performing deep pelvic LND as there is no difference in OS for involved versus negative deep pelvic nodes. These supporters maintain that metastatic pelvic nodal disease behaves as stage III disease rather than stage IV disease [91]. A recent survey demonstrated that only 30 % of melanoma surgeons routinely perform ilioinguinal lymph node dissection [92]. These issues demonstrate some of the uncertainty and lack of uniformity of when to perform a deep lymphadenectomy for ilioinguinal disease.

Several issues are central to the discussion of when to perform a deep pelvic lymphadenectomy for metastatic disease to the groin. Preoperative lymphoscintigraphy of the lower extremity for SLNB will at times identify selective drainage to the pelvis. This may represent drainage via separate lymphatic channels or second-echelon lymph node drainage from superficial groin nodes. Kaoutzanis and colleagues reviewed their experience of 82 patients over a 3-year period that underwent SLNB of the groin, pelvis or both [93]. Of the 82 patients, 19 (24 %) had positive SLNs. Eleven patients underwent pelvic SLNB, none of which had a positive node. Pre-operative lymphoscintigraphy identified that for primary tumors located below the knee, pelvic nodes appeared to be second level nodes. For primary tumors located on the thigh/trunk, lymphoscintigraphy identified individual tracks draining directly to the pelvis. The complication rate was higher following SLNB in the pelvis but was not statistically significant when compared with SLNB of the groin alone. Soteldo and authors of the European Institute of Oncology (Milan, Italy) reviewed their experience for patients who underwent pelvic SLNB or developed recurrent pelvic disease after a negative inguinal SLNB [94]. One hundred four patients with stage I/II melanoma with primary tumors of the lower extremity or trunk underwent SLNB and were found to have hot spots both in the superficial (groin) and deep (iliac-obturator) areas during dynamic lymphoscintigraphy. Of the 104 patients, 21 patients (20 %) had a positive SLNB and all underwent superficial and deep inguinal dissection. Three patients who underwent ilioinguinal dissection were found to have positive pelvic lymph nodes. Two patients (2.4 %) who were initially SLN negative developed pelvic recurrence. With a 60-month follow up, the DFS was 69 % for SLN-negative patients and 53 % for SLN-positive patients, which was not significant (p = 0.15). Chu and colleagues from Emory University reviewed a single-surgeon experience of 40 patients with positive inguinal SLNB who underwent 42 complete inguinopelvic lymphadenectomies [95]. The median Breslow thickness was 2.3 mm and 79 % had lower extremity primaries. Five patients (11.9 %) had synchronous pelvic disease. All five cases with pelvic metastases had extremity primaries (4 distal, 1 proximal). Three of the five patients (60 %) had ≥3 total involved inguinal lymph nodes. The inguinal node ratio (ratio of positive to total number inguinal lymph nodes retrieved) was >0.2 in 80 % of cases with pelvic disease compared to 8.6 % of cases without pelvic disease (p = 0.002). The authors noted that more involved inguinal LNs and inguinal ratio >0.2 appear more likely to harbor pelvic disease. The impact of lymph node ratio provides prognostic information, with a lower ratio as a marker for better outcome [96, 97]. Karakousis evaluated the prognostic significance of drainage to pelvic nodes at SLN mapping in 325 patients with melanomas of the lower extremity or buttocks [98]. Drainage to pelvic nodes (DPN) was identified in 23 % of cases and associated with increased Breslow thickness (p = 0.007) and age (p = 0.01) on multivariate analysis. Patients with DPN were not more likely to have a positive SLN. The pelvic recurrence rates were similar in patients with recurrence with DPN compared to those without DPN (39 % in both groups, p = NS). SLN negative patients with DPN showed a shorter time to melanoma recurrence in a multivariable analysis model when considering tumor thickness and ulceration (p = 0.002) but marginally when age was included (p = 0.08).

Management of clinical (palpable) lymph node metastasis to the groin consists of a superficial and deep inguinal lymphadenectomy. In practice, however, some perform only a superficial or inguinal groin dissection and forego performing an ilioinguinal (deep) dissection. Proponents of this approach only perform a combined superficial and deep inguinal lymphadenectomy when multiple positive nodes are involved in the groin or the evidence of pelvic nodal involvement as identified on computed tomography (CT) imaging [99]. The Dutch reported their experience on 169 melanoma patients with palpable groin metastases [100]. Of the 169 patients, 121 underwent combined (superficial and deep) groin dissection and 48 underwent superficial groin dissection (SGD) . Patients were clinically diagnosed by CT, fine-needle aspiration and/or ultrasound (US). In general, patients with palpable nodes underwent CGD. The indication for SGD was surgeon preference. Thirty patients (24.8 %) who underwent CGD for palpable groin metastasis had involved deep pelvic nodes. CGD patients had significantly more patients with large superficial nodes (≥3 cm) than SGD patients (70.8 % vs 50 %, p = 0.002), more harvested superficial lymph nodes (15 nodes vs 8 nodes, p < 0.001) and lower superficial lymph node ratio (11 % vs 20 %, p = 0.0004). There was no difference in morbidity rates between groups, although patients undergoing CGD did have a trend toward more chronic lymphedema. There was no difference in local control rates, DFS or OS between SGD or CGD patients. CGD patients with involved deep lymph nodes (24.8 %) had an estimated 5-year OS of 12 % compared with 40 % without involved deep lymph nodes (p = 0.001). The authors noted that survival and recurrence do not differ between patients with palpable groin metastases who are treated by CGD or SGD. They suggest that patients without iliac nodes on CT may undergo SGD and CGD reserved for patients with multiple positive nodes on SGD or deep nodes evident on CT imaging. Authors from the National Institute of Cancer in Naples, Italy also came to similar conclusions [101]. One hundred thirty-three patients underwent superficial and deep groin dissection for melanoma groin metastasis (84 had clinically positive inguinal nodes at diagnosis, 49 patients had tumor-positive SLN). None of the 133 patients had clinical evidence of involvement of deep lymph nodes at initial staging with CT or US. Of the 49 patients with a positive SLNB, 3 (6.1 %) had evidence of disease in deep nodes and 27/84 (20.3 %) with clinically positive inguinal nodes had positive deep nodes identified. The 5-year DFS and melanoma-specific survival was significantly better for patients with superficial lymph node metastasis than both superficial and deep lymph nodes, respectively (34.9 % vs 19 %, p = 0.001 and 55.6 % vs 33.3 %, p = 0.001). Metastasis to the deep nodes was found to be the strongest predictor of DFS and melanoma-specific survival. The authors commented that a deep groin dissection should be considered for all patients with clinical nodal involvement but may be spared in patients with only a positive sentinel lymph node.