Current surgical management for melanoma

Melanoma is a major malignant cutaneous neoplasm with a high mortality rate. In recent years, the treatment of melanoma has developed dramatically with the invention of new therapeutic agents, including immune checkpoint inhibitors and molecular‐targeted agents. These agents are available as adjuvant therapies for postoperative patients with stage IIB, IIC, and III melanomas. Furthermore, neoadjuvant therapy has been studied in several global clinical trials and has demonstrated promising and favorable clinical efficacy, mainly in patients with palpable regional lymph nodes. A recent large phase III clinical trial investigating early lymph node dissection for sentinel lymph node metastases demonstrated no survival benefits. Based on these data, surgery should be reconsidered as an appropriate treatment modality for melanoma. The need for invasive surgical procedures will be reduced with the invention of effective adjuvant and neoadjuvant therapies and novel clinical trial data on regional lymph node dissection. However, surgery still plays an important role in treating early‐stage melanoma, accurately determining the disease stage, and effective palliative treatment for advanced melanoma. In this article, we focus on surgery for primary tumors, regional lymph nodes, and metastatic sites in an era of remarkably revolutionary drug treatments for melanoma.

[12][13][14][15][16][17][18] Additionally, a recent large phase III clinical trial, the Multicenter Selective Lymphadenectomy Trial II (MSLT-II), investigating the role of early regional completion lymph node dissection (CLND) for sentinel lymph node (SLN) metastasis showed no superior survival benefit of early CLND compared with observation cohort. 19The invention of effective adjuvant and neoadjuvant therapies and novel clinical trial data regarding CLND will lead to less-invasive surgery in melanoma treatment.Although a new standard of care involving less-invasive surgical treatment combined with novel agents will be developed, surgery still plays an important role in treating early-stage melanoma, accurately determining the disease stage and effective palliative treatment for advanced melanoma.

| WIDE LO C AL E XCIS I ON FOR PRIMARY TUMOR
In resectable locoregional melanomas, surgery for primary tumors aims to excise the tumor completely.The side margins of wide local excision, a fundamental surgical procedure, have historically changed in several clinical trials.

| Side margins
In the past, wide-sided margin excision was thought to improve local control and long-term prognosis.Historical surgical approaches for primary tumors have been aggressive, and wide local excision with a side margin of 5 cm is commonly performed. 202][23][24][25][26][27] A first randomized trial reported in 1991 researched the relationship between the different side margins (1 cm vs ≥ 3 cm), local recurrence, and overall survival (OS) in 612 patients with a TT of 0.8-2.0mm.The local recurrence rates during follow-up (median, 7.5 years) were similar between the two groups (1.3% vs 0%).OS showed no statistical significance (8-year OS: 86.9% and 90.3%, P = 0.64). 21,28,29Another randomized trial 22 reported in 2000 analyzed the relationship between different side margins (2 cm vs 5 cm) and local recurrence, relapse-free survival (RFS), and OS rates in 989 patients with a TT of 0.8-2.0mm.During a median follow-up of 11 years, local recurrence rates were similar between the two groups (0.6% vs 1%).No statistical significance was found in RFS and OS (10-year RFS 71% vs 70%, 10-year OS 79% vs 76%, P values unavailable). 22Another randomized trial 23 reported in 2001 compared local recurrence and OS between two different side margins (2 cm vs 4 cm) excised in 468 patients with a TT of 1.0-4.0mm.Following a median follow-up of 10 years, there was no statistical significance in local recurrence and OS between the two groups (local recurrence rates 2.1% vs 2.6%, P value unavailable; 10-year OS 70% and 77%, P = 0.074). 23Furthermore, a randomized multicenter trial 24 reported in 2003 researched local recurrence, disease-free survival (DFS), and OS in the two different margins (2 cm vs 5 cm) in 326 patients with a TT of ≤2 mm.The local recurrence rates were similar between the two groups (0.6% vs 2.4%).No statistical significance was observed in DFS and OS (10year DFS 85% vs 83%, P = 0.83; 10-year OS 87% vs 86%, P = 0.56). 24reover, another randomized multicenter trial 25 reported in 2011 analyzed local recurrence and OS comparing two different margins (2 cm vs 4 cm) in 936 patients with thickness >2 mm.No statistically significant difference between the two groups was observed in the local recurrence rates (4.3% vs 1.9%, P = 0.06).RFS and OS also had no statistical significance (5-year RFS 56% vs 56%, P = 0.69; 5year OS 65% and 65%, P = 0.69). 25Lastly, a randomized multicenter trial 26,27 reported in 2004 and 2016 compared local recurrence, DFS, and OS in the different excisional margins (1 cm vs 3 cm) in 900 patients with a TT >2.0 mm.There was a statistically significant difference in the locoregional recurrence (local recurrence, in-transit metastasis, and regional lymph node metastasis) rate between the two groups (37.0%vs 31.8%,P = 0.05).Statistically significant differences were found in melanoma-specific survival (MSS) (5-year MSS 57.1% vs 63.6%, P = 0.04).However, there was no statistically significant difference in OS (44.1% vs 46.8% P = 0.14). 26,27Based on the results of the randomized trials above (    with those treated with narrower margins (N group).In patients with T1-3 melanoma, the mortality rates in the N and R groups were almost identical (1.36 and 1.28 per 100 person-years).Among patients with T4 melanoma, the N group had a higher mortality rate than the R group (11.44 vs 5.03 per 100 person-years).Multivariate analysis showed that surgical margin was not a risk factor for MSS (P = 0.38) or DFS (P = 0.31). 31Additionally, Lino-Silva et al. 32 analyzed recurrence (local recurrence or distant metastasis) rates and OS between two margins (1-2 cm vs > 2 cm) in 306 patients with acral melanoma at pathologic stages pT3 and pT4.No statistical significance was found in recurrence rates and OS between the two groups (mean follow-up of 31.9 months, recurrence rates 30.0%vs 35.4%, P = 0.32; 5-year OS 69.9% vs 61.4%, P = 0.41). 32Sun et al. 33 compared local and in-transit recurrence-free survival (LITRFS), DFS, OS in the different side margin excisions (1-2 cm vs > 2 cm) in 207 patients with acral melanoma with stage pT3-T4 disease.The side margin was not correlated with patients' LITRFS, RFS, or OS (LTTRF, P = 0.35; DFS, P = 0.08; OS, P = 0.20). 33 for an ongoing trial, the Melanoma Margins Trial II (MelMarT-II; Clini calTr ials.gov number, NCT03860883) compared clinical outcomes between 1-cm margin wide local excision and 2-cm margin wide local excision in patients with a Breslow thickness of >2 mm or 1-2 mm with ulceration (pT2b-pT4b).

| Deep margins
The appropriate determination of deep margins is as crucial as that of side margins for complete removal of the primary tumor.However, there is a lack of studies investigating adequate deep margins. 34,35e NCCN Guidelines do not clearly state recommendations for deep margins. 30A retrospective study by Grotz et al. 35 compared the clinical outcomes of different deep margins between the fascia resection and preserved groups.Fascial resection was associated with a 2.5-fold increased risk of regional nodal recurrence.However, it was not associated with local recurrence or OS, 35 indicating that wide local excision with underlying fascia resection would not improve local control and survival.Additionally, this study had a considerable selection bias, as the TT was thicker in the fascial resection group.In general, the distance from the deepest base of the primary tumor to the fascia varied greatly, depending on the patient's obesity.
Therefore, in real-world practice, the determination of deep margins appears to depend on the discretion of the individual surgeon.
Studies on appropriate deep margins in nail apparatus melanoma are crucial for preserving digit function.[39] The loss of a digit by amputation leads to poor cosmesis and interferes with activities of daily living.Meanwhile, the prognosis in patients with nail apparatus melanoma depends on the T stage at the initial diagnosis, not the degree of radical surgery for the primary tumor. 361][42][43] A recent histologic study also supported that many patients with nail apparatus melanoma with TT of ≤4 mm did not show invasion to the underlying distal phalanx. 44sed on the results from these studies, non-amputative digit preservation surgery may be applicable even in patients with invasive nail apparatus melanoma without compromising their digit cosmesis, digit function, and vital prognosis.An investigator-initiated clinical trial, JCOG1602 (J-NAIL study, clinical trial no.UMIN000029997), to evaluate the safety and efficacy of non-amputative digit preservation surgery with a side margin of 5-10 mm in patients with invasive nail apparatus melanoma is ongoing in Japan (Figure 1). 45

| S LN B I OPSY
The SLNs are the first regional lymph nodes to receive lymphatic flow from a primary tumor.If the SLNs do not metastasize, it is considered that there are no regional metastatic nodes downstream of the SLNs.In this context, SLN biopsy (SLNB) remains an important diagnostic procedure for detecting nodal micrometastases and predicting the prognosis.

| Development of SLNB techniques
To identify the SLN, blue dye injection around the primary tumor was performed in the first advent of the SLNB concept.However, the SLN identification rate with blue dye injection alone was low, approximately 82%. 46Lymphoscintigraphy using radioisotope (RI) injection, such as TA B L E 2 Recommended surgical side margins according to tumor thickness in the National Comprehensive Cancer Network Guidelines.

Recommended side margin (cm)
In situ 0.5-1 technetium tin colloid or technetium phytate, and the intraoperative use of a hand-held gamma probe (RI method) dramatically increased the SLN identification rate.Combined with the RI method, blue dye injection significantly improved SLN identification by up to 99%, 47,48 and has been the standard technique for SLNB.Additionally, hybrid single-photon emission computed tomography/computed tomography (SPECT/CT) can accurately visualize the anatomic position of radioactive SLNs and interval nodes, which is of great value when radioactive SLNs are located in complex anatomical areas such as the head and neck region.
0][51][52][53] ICG binds to albumin in the human body after injection and generates a peak wavelength of 840 nm near-infrared fluorescence when excited with 765 nm light.The intraoperative use of a near-infrared camera enabled dermatologic surgeons to observe ICG as superficial subcutaneous lymphatic flow.The small particle size of ICG (2.1 nm) also led to the smooth flow of ICG in the lymphatics and easy identification of SLNs that were not identified by lymphoscintigraphy because of the poor flow of the RI due to its larger particle size than ICG.The synergistic use of the ICG method with RI method has been proposed in cases of difficult identification of SLN using the RI method alone, such as the shine-through phenomenon due to the proximity between the primary tumor and radioactive SLNs. 54

| Complications of SLNB
[57] A systematic review reported in 2017 57 analyzed 9047 patients in 21 articles, and the overall incidence rate of complications was 11.3%.

| Target patients for SLNB
SLNB is a crucial procedure for the accurate pathological staging of patients with clinical stages I and II.SLNB is generally recommended for patients with a TT of ≥0.8 mm because the probability of harboring positive SLNs in patients with a TT of <0.8 mm without ulceration (T1a disease) is extremely low (<5%). 58Additionally, a recent retrospective study 59   (n = 77) received CLND plus adjuvant therapies. 63Another retrospective study 64 from Japan also demonstrated that the proportion of patients with positive SLN who underwent CLND was significantly lower after the publication of the MSLT-II trial results and approval of novel adjuvant therapies (66.1% [37/56] vs 90.5%

| CLND
As CLND for patients with positive SLN tends to be omitted owing to recent evidence from DeCOG-SLT and MSLT-II, and novel effective adjuvant therapies, the current real-world indication of CLND is considered for patients with clinical nodal disease.However, the true therapeutic value of CLND for clinical nodal disease is still unknown because no trials have compared the prognosis after CLND with that without CLND or alternative nonsurgical treatment modalities in those cohorts, therefore CLND provides regional control and prognostic information.[67][68][69][70][71][72][73]

| Appropriate extent of CLND
The extent of CLND is often modified based on the anatomic location of the clinical nodal disease and lymphatic flow in the regional lymphatic basin.However, the appropriate extent of CLND that safely balances the therapeutic effects remains controversial.
Although many studies have attempted to clarify the appropriate extent of CLND, 74,[83][84][85][86][87] these studies included patients with positive SLN who no longer were absolute candidates for CLND.

| Neck CLND (neck dissection) and parotidectomy
The lymphatic drainage pattern of the head and neck region is complicated, therefore selective neck dissection has become a major treatment of choice to minimize the extent of dissection. 88t depends on the predicted lymphatic flow from the location of the primary tumor and the different extent of dissection, including superficial parotidectomy, 89 has been proposed by O'Brien et al. 90 and the Netherland Cancer Institute 91 as representative guides for selective neck dissection (Table 3 and Figure 2a-d).However, the extents of selective neck dissection of all primary tumors are not covered by those guides (Figure 2c) and there is no definitive evidence that the therapeutic effect of those extents of selective neck dissection is truly not inferior to that of a larger extent of dissection, including radical or modified radical neck dissection.
Additionally, in the head and neck region, unlike head and neck cancer, dissection of superficial lymph nodes, such as preauricular, postauricular, occipital, and superficial cervical nodes, is sometimes required. 92

| Axillary CLND
The axial lymph nodes were divided into three levels (I, II, and III).
Although the recommended level of CLND has been from I to III, [93][94][95] no significant difference was observed in locoregional recurrence

| WILL NEOADJ U VANT THER APIE S LE AD TO LE SS -INVA S IVE SURG ERY ?
As described above, the current standards of care for resectable clinical nodal disease (clinical stage III) are wide local excision and CLND, followed by adjuvant therapy.However, recent attempts, mainly in this clinical stage, have shown that these novel agents can be used as neoadjuvant therapies before surgery.Neoadjuvant therapies have the potential advantages of tumor reduction, resulting in less-invasive excision for primary lesions, omission of CLND, and prolonged survival, but there may be a risk of disease progression during neoadjuvant therapies if ineffective. 981][12][13][14][15][16][17][18] The latest representative phase II study, PRADO trial, 17 using nivolumab plus ipilimumab combination as neoadjuvant therapies, showed a pathologic complete response (pCR) of 49% and nearpathologic complete response of 12%, which led to the omission of CLND after neoadjuvant therapy.The 2-year RFS in the entire cohort was 85% (2-year RFS in patients with pCR + near pCR was 93%).
The other representative phase II trial 18 that used pembrolizumab as neoadjuvant therapy and adjuvant therapy compared with the adjuvant pembrolizumab alone, in patients with resectable stage IIIB to IV melanoma, also demonstrated favorable outcomes.The neoadjuvant-adjuvant cohort (154 patients) revealed pCR of 21% and statistically significant prolonged event-free survival compared to the adjuvant-only cohort (159 patients) (2-year event-free survival 72% vs 49%, P = 0.004).The 2-year OSs of each cohort were >80% and <80%, respectively (P value unavailable).These results suggest a potentially superior therapeutic effect of neoadjuvant therapy compared with adjuvant therapy and the possibility of a personalized surgical approach in combination with immune checkpoint inhibitors such as neoadjuvant therapy in these stages, 17 which may be a new standard of care.

| M E TA S TA S EC TO M Y
In patients with melanoma with resectable distant metastases, such as oligometastasis, the role of metastasectomy is controversial in the era of novel, effective therapeutic agents.0][101][102] A meta-analysis 103 103 Another study reported the role of metastasectomy in combination with immune checkpoint inhibitors. 104In this study, 52 patients with oligometastasis underwent metastasectomy at least 6 months after initiating immune checkpoint inhibitors.The result showed no evidence of disease or non-progressive residual disease, which resulted in a 3-year PFS of 31% and a 5-year MSS of 60%.Stratified by patterns of failure, patients with progression of established tumors had a 3-year PFS of 70%; in contrast, those with new metastases had a 3-year PFS of 6% (P = 0.001).Five-year MSS in patients with the progression of established tumors and those with new metastases after metastasectomy were 93% and 31%, respectively (P = 0.046).
These data indicate that metastasectomy for oligometastasis after immune checkpoint inhibitor treatment may achieve sustained PFS in certain patients.However, we must consider that this study also has a selection bias.

| CON CLUS IONS
Compared with the past standard of care regarding surgery for melanoma, the furture standard of care for surgery will be less invasive, Although CLND has traditionally been performed in patients with positive SLN, the results of recent pivotal phase III randomized trials F I G U R E 1 A male patient with invasive nail apparatus melanoma who was enrolled in JCOG 1602 and received non-amputative digit preservation surgery.(a) Black-colored nail plate with ulcerative nodular lesion and Hutchinson's phenomenon on the left big toenail.Black line indicates incision line of non-amputative digit preservation surgery.(b) After removal of the nail apparatus at the level of the distal phalanx.(c) Histopathology of resected nail apparatus.The black dotted line indicates the proliferative area of melanoma cells.The yellow double-ended arrow indicates tumor thickness (2.2 mm).The blue double-ended arrow indicates the shortest tumor-to-deep margin distance (0.9 mm).(d) Clinical appearance 3 years after coverage of defect by skin grafting.investigating the therapeutic value of immediate CLND after positive SLNB and the advent of novel adjuvant therapies, including immune checkpoint inhibitors and BRAF/MEK inhibitors, have changed this traditional strategy.As a first phase III randomized clinical trial, the Dermatologic Cooperative Oncology Group-SLN Trial (DeCOG-SLT)60 was conducted to investigate the therapeutic role of immediate CLND in patients with a TT of ≥1.0 mm melanoma harboring positive SLN.Comparing the immediate CLND cohort (n = 240) with the observation cohort (n = 233, patients who underwent CLND when regional lymph node metastasis was suspected based on ultrasonographic examination), no significant differences were observed in the distant metastasis-free survival (DMFS) (5-year DMFS 67.6% vs 64.9%, hazard ratio [HR] 1.08, P = 0.87), RFS (HR, 1.01), and OS (HR, 0.99) between the two treatment cohorts, but most patients (n = 311) had small tumor burdens of ≤1.0 mm in the SLNs.Another representative pivotal trial, MSLT-II,19 enrolled a larger number of patients with positive SLN and compared the immediate CLND cohort (n = 824) with the observation cohort (n = 931).There was no significant difference in MSS between the two cohorts (mean 3-year MSS rate 86 ± 1.3% vs 86 ± 1.2%, P = 0.42).The DFS was slightly prolonged in the CLND cohort but insignificant (3-year DFS rate 68 ± 1.7% and 63 ± 1.7%, P = 0.05).Furthermore, the incidence of postoperative lymphedema was higher in the CLND cohort than in the observation cohort (24.1% vs 6.3%).A positive non-SLN status, determined by CLND, was an independent prognostic factor for recurrence (HR 1.78, P = 0.005).Based on the data from the two major clinical trials mentioned above, immediate CLND was not associated with prolonged survival, although immediate CLND may provide patient prognostic information.The advent of effective adjuvant therapies using novel agents, including anti-PD-1 antibodies and BRAF/MEK inhibitors, has drastically changed the real-world management of patients with positive SLN.Simultaneous pivotal phase III clinical trials for adjuvant systemic therapy showed that anti-PD-1 antibody (nivolumab, pembrolizumab) and BRAF/MEK inhibitors are more effective than historical alternatives, such as high-dose interferon-alpha or observation with no adjuvant treatment.The protocols of these pivotal phase III clinical trials, including CheckMate 237, 61 KEYNOTE-054,7 and COMBI-AD,62 investigating the efficacy of nivolumab, pembrolizumab, and dabrafenib plus trametinib combination as adjuvant therapies, all included patients who received adjuvant therapy after CLND, even if they had positive SLNs.However, in real-world practice, physicians often administer adjuvant therapy after skipping CLND in patients with positive SLNs.A recent large retrospective study63 enrolled 1109 patients with positive SLN from 21 institutions in Australia, Europe, and the United States.The results showed that 47% of patients (n = 519) were treated with nodal observation alone, 37% of patients (n = 411) were treated with nodal observation plus adjuvant systemic therapy, 9% of patients (n = 102) were treated with CLND alone, and only 7% of patients

TA B L E 3 F I G U R E 2
Abbreviations: O, occipital node; P, parotid node.
accompanied by the development of novel therapeutic agents for adjuvant or neoadjuvant use and verification of the correlation between surgical procedures, including excision margins for the primary tumor and the extent of CLND, and patients' complications and prognosis.ACK N OWLED G M ENTSThis work was supported by the Japan Agency for Medical Research and Development (grant number JP23ck0106765h0002) and the National Cancer Center Research and Development Fund (grant number 2023-J-3).The institutions and funding sources were not involved in the study design, collection, analysis, and interpretation of data, writing of the report, or decision to submit the article for publication.CO N FLI C T O F I NTE R E S T S TATE M E NT S.K. has no conflicts of interest to disclose.T.I. received institutional research funding from Kakenhi, Maruho, Taiho, Daiichi Sankyo, Torii, and Sun Pharma, and has received honoraria from BMS, MSD, and Ono Pharma.Y.N. receives institutional research funding from Torii and has served as a consultant or/and has received honoraria from Alexion Pharma, Bristol-Myers Squibb (BMS), Kyowa Kirin, Leo Pharma, Maruho, Merck Sharp & Dohme (MSD), Novartis, Ono Pharma, Sanofi, Sun Pharma, Tanabe-Mitsubishi Pharma, and Torii.

Table
Studies evaluating different surgical side margins for wide local excision.
TA B L E 1