Introduction

In the United States, 84% of patients with melanoma present with localized disease, 9% with regional lymph node disease, and 4% with distant metastatic disease. Lymph node metastases constitute the most important prognostic factor in the early stages and their incidence increases with increasing Breslow thickness, presence of ulceration and mitotic index¹. Sentinel lymph node (SLN) biopsy is currently recommended for the diagnosis of lymph node involvement in melanoma patients with a Breslow thickness ≥ 0.80 mm and absence of clinically suspicious lymph nodes¹.

The presence of SLN metastases is an important prognostic factor and has been considered an indication for completion lymph node dissection (LND) of the territory involved to improve staging as the presence of positive non-sentinel lymph nodes (NSLN) is another important prognostic factor, and to reduce regional recurrences². However, in some series only 20% of positive SLN-patients present metastases in the NSLN^3-5.

In 2017, the results of the international MSLTII and DeCOG-SLT studies showed that melanomaspecific survival does not change significantly when SLN-positive patients are regularly observed with ultrasound and undergo therapeutic LND compared with the group with immediate completion LND^6,7.

Because of this possible paradigm shift in the management of SLN-positive melanoma patients and the convenience of its implementation, the aim of this study was to describe the percentages and factors associated with SLN metastasis and NSLN metastases in LNDs of SLN-positive patients.

Material and methods

We conducted a retrospective, descriptive and observational study in a tertiary university-based hospital. The electronic pathology records entered between June 2012 to December 2019 were reviewed, using the term “melanoma and the filter “cutaneous location”. Mucosal and ocular melanomas and those only seen in a medical visit were excluded. The demographic and clinical variables and pathological characteristics of the primary tumor and lymph nodes of 139 patients operated on during the aforementioned period were recorded. Thirteen patients underwent regional LND without SLN biopsy as initial treatment because they had suspicious lymph nodes with positive fine needle aspiration (FNA) biopsy. In 30 cases lymph nodes were not examined due to melanoma thickness < 0.75 mm in 17 cases and melanoma in situ in 13 patients. Sentinel lymph node biopsy was performed in 96 patients (69%) who constitute the study population. Sentinel lymph node biopsy was indicated in all patients with thickness ≥ 0.75 mm and absence of clinically suspicious lymph nodes after 2014, and in those with thickness ≥ 1 mm after 2014. Mean age was 61.7 ± 17.5 years (range: 19-93 years) and 55.2% (n = 53) were men. Fifty-six (40.3%) patients lived in the city of Buenos Aires (CABA), 44 (31.7%) in outskirts and 39 (28%) in inland cities.

Preoperative lymphoscintigraphy was performed in all the cases; the radiotracer was used alone or with 3% patent blue for intraoperative identification, and indocyanine green was used in two occasions. Those patients with positive SLN biopsy underwent immediate completion LND or delayed LND of the territory involved.

Lymph node dissection was performed during the same surgery the biopsy was done in those cases in which the intraoperative examination showed a clear positive diagnosis: when the SLN was macroscopically suspicious (black and hard), it was excised and a touch imprint cytology was done. When the SLN was not macroscopically suspicious, it was submitted for delayed histology with immunohistochemistry. Frozen section examination with cryostat was not performed in any case. Data were incorporated into a database and were later analyzed using SPSS 16 BY SPSS^® statistical software package. Appropriate descriptive statistics and significance tests were determined for each variable according to its scale of measurement and distribution. The significance level established was an alpha error of 0.05.

The study was conducted following the ethical principles of the Declaration of Helsinki and revised in Tokyo. The clinical data were protected so as not to identify to whom they belong and not to be accessible to persons not bound by professional secrecy. All the patients signed an informed consent approved by the Committee on Ethics of the institution.

Results

Melanomas were located in the extremities in 47 (49%) cases, in the trunk in 39 (40.6%), and in the head and neck in 10 (10.4%).

The histological types found were superficial spreading (n = 45; 46.9%), nodular (n = 24; 25%), acral lentiginous (n = 13; 13.6%), epitheloid (n = 4; 4,1%), and others (n = 10; 10,4%). Mean Breslow thickness was 5.08 mm (0.8-50 mm), and ulceration was found in 46 patients (33%). Lymphatic mapping was done in 96 cases with Breslow thickness ≥ 0.75 mm and absence of suspicious regional nodes.

A combination of 3% patent blue and a radiotracer based on 99mTc labelled colloid (Linfofast^®) was used in 68 patients (70.6%), patent blue 3% alone in 15 patients (15.7%), radiotracer alone in 11 patients (11.5%) and indocyanine green in 2 patients (2.2%). The variation in lymphatic mapping methods was due to the availability of materials and preferences of the operating surgeon. The SLN was found in all the cases and was positive in 39 (40.6%). Mean Breslow thickness was 5.01 ± 6.6 mm (1.05-50) in SLN-positive patients, and 4.7 ± 5. 3mm (0.8-28) in SLN-negative patients. The Student’s t-test was not significant for these variables: 1.675 (p 0.122). The comparison of variables between SLNpositive patients versus SLN-negative patients is detailed in Table 1. A cut-off point of Breslow ≥ 3 mm was chosen, following the recommendations of Sinnamon et al⁸.

Table 1 Comparison between sentinel lymph node (SLN) positive patients versus SLN-negative patients 

Lymph node dissection was carried out in the 39 patients with positive SLN: axillary LND in 23 (59%), with a mean number of 11 nodes found, inguinal LND in 10 (25.6%) with a mean number of 19 nodes, and cervical LND in 6 (15.4%) with a mean number of 22 nodes. Non-sentinel lymph nodes were positive in 23 cases (59%). The following variables were compared between NSLN-positive patients versus NSLN-negative patients: Breslow thickness, presence of ulceration, spreading superficial histological type, location in the extremities and male sex; the results are detailed in Table 2.

Table 2 Comparison between non-sentinel lymph node (NSLN) positive patients versus NSLN-negative patients in SLN-positive patients 

Thickness ≥ 3 mm and ulceration occurred simultaneously in 48.7% (19/39) of patients with positive SLN on lymphatic mapping, and in 65.2% (15/23) of those with positive NSLN on LND.

Discussion

Sentinel lymph node biopsy in melanoma has enabled accurate staging and indication of systemic therapy in cases of positive results⁹. The percentage of SLN metastases varies among the different series and is associated with the characteristics of the population analyzed. Although SLN biopsy was initially indicated for patients with intermediate thickness melanoma, different investigators have also reported benefits for patients with thicker melanomas^10-12.

Different guidelines for the management of melanoma do not indicate any upper cutoff value of thickness to perform SLN biopsy in the absence of evidence of regional or distant metastases^1,13. In Argentina, the initial experience with the method was reported in 2002. Distant metastases in the SLN were found in 17.5% of the cases. The predictive factors were ulceration, thickness and number of draining nodal basins, and 30% positive NSLN in the LND¹⁴. In 2015, Schlottmann et al. reported metastases in 21.3% of SLN biopsies and the predictive factors were Breslow thickness and nodular histological type¹⁵.

Until recently, the presence of a positive SLN has been considered a precise indication for completion LND, since knowing the tumor status of NSLN provides prognostic information and reduces regional recurrence². The American Joint Committee on Cancer (AJCC) Cancer Staging Manual considers the number of positive regional nodes for cutaneous melanoma staging, to define prognosis and indicate systemic adjuvant treatment¹⁶. Several investigators have reported low rates of positive NSLN in their respective series. This has led to the development of predictive models, usually linked to factors concerning the primary tumor (thickness, ulceration, mitotic index and location) and tumor burden in LND to reduce morbidity and costs associated with LND^8,17-19. Nevertheless, such morbidity would not be the same in all nodal basins and patients²⁰. Furthermore, the prospective randomized clinical trials previously mentioned demonstrated that observation with systematic follow-up could avoid LND without affecting survival^6,7.

The results of the MSLT-II trial, which compared completion LND versus observation in SLN-positive patients were published in 2017. Among patients who underwent LND, positive NSLN were detected in 11.5%. During follow-up, the rate of regional recurrences in these patients was 17.9% at 3 years and 19.9% at 5 years.

In the observation group, the percentage of patients in whom ultrasonographic or physical examination revealed NSLN involvement increased to 22.9% at 3 years and 26.1% at 5 years. There were no differences in melanomaspecific survival at 3 years between both groups (86 ± 1.3% and 86 ± 1.2%). Lymphedema was more common in the dissection group (24.1% vs. 6.3%). At 5 years, regional recurrence was more common in the observation group. Lymph node dissection did not increase survival in SLN-positive patients but provided valuable information about staging and increased the rate of regional disease control⁶. It should be noted that this study was performed in a population with some precise characteristics: mean Breslow thickness was 2.76 mm in the completion LND group and 2.70 mm in the observation group, and in this latter group, 77.2% had Breslow thickness ≤ 3.5 mm. Another similar study performed in Germany included patients with melanoma and positive SLN assigned to observation (n = 233) or LND (n = 240); 311 (66% of the entire sample) had metastases of 1 mm or less. After a median follow-up of 35 months distant metastasis-free survival at 3 years was 77·0% in the observation group and 74·9% in the LND group which had greater rate of complications. For this reason, the authors did not recommend LND in this group of patients⁷. In Australia, Nijhuis et al. analyzed 483 patients who underwent SLN biopsy for melanoma and 61 had positive SLN (13%). Of these, only two underwent LND, and NSLN were negative in both patients²¹.

Recently, a retrospective study analyzed 166 patients with melanoma and positive SLNs excluded from the MSLT-II for presenting high-risk features (microsatellites, extranodal extension, or > 3 positive SLNs). Of the 166 patients, 114 (69%) underwent observation while LND was done in 52 (31%). Analysis of matched data found differences in the regional recurrence rate (14% vs. 6%) but no differences in melanoma-specific mortality. The authors concluded that, due to the increased risk of distant metastases, LND should be avoided in these patients so as not to delay the initiation of systemic treatment²². In the present series there was a non-significant difference in mean thickness between SLN-negative patients and SLN-positive patients (4.7 mm vs. 5.01 mm, respectively) This occurred because a case with a thickness of 28 mm was included in the first group. In this series of SLN-positive patients, 66% have a thickness of 4 mm. This would explain the greater percentage of positive SLNs (59%) in the LNDs performed.

Sentinel lymph node biopsy in melanoma patients with Breslow thickness > 4 mm has gained attention recently. El Sharouni et al. studied a cohort of 1150 melanoma patients with thickness > 4 mm who underwent SLN biopsy and found a disease-free survival of 48.1% at 5 years when SNL was positive, and 70.5% when SNL was negative, which was similar to the 71.5% found in 1877 patients with thickness ≤ 4 mm with positive SNL, and therefore suggest including these patients in clinical trials of systemic therapies²³.

Hans et al. investigated the prognostic factors in 1235 melanoma patients with thickness > 4 mm. The patients were divided by thickness into 3 groups: > 4 to 6, > 6 to 10, and > 10. On multivariate analysis, SLN status predicted overall survival in all the groups²⁴.

The initial results of nodal observation in SLNpositive patients reported by international publications, such as the study by Nijhuis and the MSLT-II and DeCOGSLT trials seem promising in terms of omitting LND and thus avoiding the associated complications, with no changes in long-term survival. However, this approach could be difficult to implement in other countries, as it requires adherence to strict clinical and ultrasound follow-up, which is not always available. Strict observation, both in terms of frequency and quality of studies, of SLN-positive patients without LND, to detect the highest percentage of early regional recurrence is difficult to achieve in a population treated in a public hospital, with wide geographic dispersion and economic limitations aggravated by the recent health emergency due to the COVID-19 pandemic and access barriers to other expensive systemic treatments. The limitations of our study are its retrospective design, the size of the sample, and the lack of long-term follow-up. Its strength is that it encompasses all the lymphatic territories with a uniform criterion in a local population.

In conclusion, and based on the results found, we can state that ulceration and Breslow thickness > 3 mm are associated with positive SLNs and positive NSLN in the LND, although it has not been possible to establish the joint relationship of both variables due to the lack of multivariate analysis. The presence of a high percentage of positive non-sentinel nodes and the difficulty of strict follow-up suggest that completion lymph node dissection should not be omitted in patients with positive sentinel nodes. Future development and improved accessibility to adjuvant therapies will likely lead to a reformulation of this surgical strategy.