Cutaneous melanoma: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up

Stage IIB-IIC disease

Considering the 10-year melanoma-specific survival (MSS) rate in stage IIB-IIC disease, and the high risk of recurrence in these patients,^14,15 two trials investigating adjuvant therapy, with a primary endpoint of recurrence-free survival (RFS), were conducted.

In the KEYNOTE-716 trial, 976 patients (age ≥12 years) with completely resected AJCC8 pathological stage IIB-IIC melanoma received intravenous pembrolizumab 200 mg (2 mg/kg in paediatric patients) or placebo every 3 weeks (q3w) for 17 cycles or until disease recurrence or unacceptable toxicity. The estimated 36-month RFS rate was 76.2% for pembrolizumab and 63.4% for placebo [hazard ratio (HR) 0.62, 95% confidence interval (CI) 0.49-0.79], and the estimated 36-month distant metastasis-free survival (DMFS) rate was 84.4% for pembrolizumab versus 74.7% for placebo (HR 0.59, 95% CI 0.44-0.79).¹⁶ In June 2022, the EMA approved the use of adjuvant pembrolizumab for patients ≥12 years old with stage IIB-IIC melanoma following complete resection.

In the CheckMate 76K trial, patients with AJCC8 pathological stage IIB-IIC melanoma (i.e. similar to KEYNOTE-716 eligibility criteria) were randomised to receive either adjuvant nivolumab or placebo for up to 1 year or until disease recurrence or unacceptable toxicity. At a median follow-up of ~23 months, adjuvant nivolumab improved RFS versus placebo (HR 0.53, 95% CI 0.40-0.71). Higher RFS rates were observed for nivolumab regardless of disease stage or T category.¹⁷ The licensed indication for nivolumab as adjuvant therapy has been expanded to include patients with stage IIB-IIC melanoma.

Currently, no overall survival (OS) data are available from the KEYNOTE-716 and CheckMate 76K trials.

Based on these findings, clinicians can offer adjuvant anti-PD-1 treatment for patients with AJCC8 stage IIB-IIC disease after a detailed discussion with the patient to weigh the pros and cons of treatment benefit versus toxicity and a careful evaluation of clinical factors, including patient age, comorbidities, performance status (PS), reproductive potential, personal/family history of autoimmune disease and compliance in case of immune-related toxicity.

Resectable stage III disease

It is worth noting that entry criteria for most adjuvant trials in this setting were based on the AJCC seventh edition (AJCC7),¹⁸ and that stage IIIA disease defined by AJCC7 includes a higher-risk group of patients than stage IIIA defined by AJCC8, which also incorporates Breslow thickness into stage III disease (5-year MSS rates for AJCC7 versus AJCC8 stage IIIA disease are 78% versus 93%, respectively).¹⁵ Moreover, patients with AJCC7 stage IIIA disease were either excluded (CheckMate 238) or had to have >1 mm tumour metastasis from the SLNB to be included. Therefore, for patients with AJCC8 stage IIIA and <1 mm SLNB tumour burden, in the absence of prospective validation of the benefit of adjuvant therapy in this patient population, adjuvant therapy should not be considered as the standard treatment.¹⁹

Efficacy and safety data from prospective randomised controlled trials evaluating adjuvant treatment with immune checkpoint inhibitors (ICIs; ipilimumab, nivolumab, pembrolizumab) in patients with high-risk resected stage III melanoma are summarised below. It is important to note that the eligibility criteria for all trials except CheckMate 915 included complete resection of all disease, including primary tumour excision with adequate margins and CLND in patients with nodal metastases detected by SLNB and those with clinical evidence of regional disease. Although it is unclear if the recommended adjuvant treatment options have similar efficacy in the absence of CLND following a positive SLNB, the consistent 24-month RFS rates reported for nivolumab in CheckMate 238 and CheckMate 915²⁰ suggest that CLND should not be a factor in the decision to use adjuvant therapy in patients with nodal metastases detected by SLNB.

Anti-cytotoxic T lymphocyte-associated antigen 4

In the EORTC 18071 trial,²¹ long-term adjuvant therapy with the anti-cytotoxic T lymphocyte-associated antigen 4 (CTLA-4) agent ipilimumab resulted in improved 5-year RFS (HR 0.76, 95% CI 0.64-0.89, P < 0.001) and OS rates (HR 0.72, 95.1% CI 0.58-0.88, P = 0.001) compared with placebo. The benefit was also observed for patients with N1b and higher disease stages. However, the treatment schedule (10 mg/kg q3w for four doses and then every 3 months for up to 3 years) was associated with several severe and long-lasting adverse reactions and is no longer used. Given the toxicity profile of anti-CTLA-4 and the duration of therapy investigated, adjuvant therapy with either anti-PD-1 agents or dabrafenib–trametinib for patients with BRAF-mutated melanoma is preferred.

Anti-PD-1

In the CheckMate 238 trial, adjuvant therapy with the anti-PD-1 agent nivolumab has shown a significant RFS benefit²² but no significant OS benefit compared with ipilimumab 10 mg/kg for patients with stage IIIB, IIIC or IV (AJCC7) resected melanoma with no evidence of disease (NED) (see details in ‘Resectable stage III and IV NED’ section below).²³

Adjuvant treatment with the anti-PD-1 agent pembrolizumab was evaluated in patients with AJCC7 stage IIIA (SLN >1 mm), IIIB or IIIC (without ITM) melanoma in the placebo-controlled EORTC 1325/KEYNOTE-054 trial. At a median follow-up of 4.9 years, the 5-year RFS rate was longer in the pembrolizumab group than in the placebo group in the intention-to-treat population (HR 0.61, 95% CI 0.51-0.72) and in those with PD-L1-positive tumours (HR 0.62, 95% CI 0.48-0.79). Moreover, the 5-year DMFS rate was higher in the pembrolizumab group than in the placebo group (HR 0.62, 95% CI 0.52-0.75).²⁴ OS data are not yet available.

These results were validated in the phase III S1404 trial, which compared adjuvant pembrolizumab with either of the two standard-of-care (SoC) options at the time (i.e. high-dose interferon-α or ipilimumab 10 mg/kg). RFS was significantly longer in the pembrolizumab group than in the SoC group (HR 0.76, 99.62% CI 0.59-0.99, log-rank P= 0.002).²⁵

Based on these data, the EMA approved nivolumab and pembrolizumab for use in the adjuvant setting in August and December 2018, respectively.

Resectable stage III and IV NED

In the CheckMate 238 trial,²³ which compared nivolumab with ipilimumab 10 mg/kg in patients with AJCC7 stage IIIB, IIIC or IV resected melanoma, the 5-year RFS rates were 50% in the nivolumab group and 39% in the ipilimumab group (HR 0.72, 95% CI 0.60-0.86). The 5-year RFS rates according to disease stage (IIIB, IIIC, IV M1a-b, IV M1c) were 58%, 43%, 47% and 55% for nivolumab and 48%, 35%, 29% and 49% for ipilimumab. The 5-year DMFS and OS rates were 58% versus 51% (HR 0.79, 95% CI 0.63-0.99) and 76% versus 72% (HR 0.86, 95% CI 0.66-1.12), respectively.²² Moreover, nivolumab treatment was associated with fewer grade 3/4 adverse events (AEs) compared with ipilimumab (14.4% versus 45.9%, respectively).²³

In the IMMUNED phase II trial, 167 patients with stage IV melanoma and NED after surgery or RT were randomised to receive nivolumab 1 mg/kg plus ipilimumab 3 mg/kg q3w for four doses followed by nivolumab 3 mg/kg every 2 weeks (q2w), nivolumab monotherapy (3 mg/kg q2w) or matching placebo for up to 1 year. The HR for RFS for nivolumab–ipilimumab versus placebo was 0.25 (97.5% CI 0.13-0.48, P < 0.0001) and for nivolumab versus placebo was 0.60 (97.5% CI 0.36-1.00, P = 0.024). The HR (95% CI) for RFS according to disease stage (M1a-b and M1c) for nivolumab–ipilimumab versus placebo was 0.29 (0.15-0.55) and 0.18 (0.06-0.57), and for nivolumab versus placebo was 0.50 (0.29-0.86) and 0.87 (0.40-1.90). The HR for OS was significantly in favour of nivolumab–ipilimumab versus placebo (HR 0.41, 95% CI 0.17-0.99, P = 0·040) but not for nivolumab versus placebo (HR 0.75, 95% CI 0.36-1.56, P = 0.44). Rates of grade 3/4 treatment-related AEs (TRAEs) were numerically higher with combination therapy, but types of toxicity were similar to what is already known for these agents.²⁶

CheckMate 915 was a phase III, double-blind trial in resected stage IIIB-IIID or IV melanoma. Patients were randomised to receive nivolumab 240 mg q2w plus ipilimumab 1 mg/kg once every 6 weeks or nivolumab 480 mg every 4 weeks (q4w) for ≤1 year. There was no significant difference between treatment groups reported for median RFS in the all-randomly assigned patient population (HR 0.92, 95% CI 0.77-1.09, P= 0.269) or in patients with PD-L1 tumour expression <1% (HR 0.91, 95% CI 0.73-1.14), or in 2-year RFS rates according to disease stage (IIIB: HR 0.91, 95% CI 0.68-1.21; IIIC: HR 0.92, 95% CI 0.75-1.13; IIID: HR 1.61, 95% CI 0.70-3.67; IV: HR 0.88, 95% CI 0.58-1.32).²⁰

Currently, adjuvant treatment with nivolumab–ipilimumab is included as a treatment option in the National Comprehensive Cancer Network guideline for cutaneous melanoma in patients with stage IV melanoma and NED,²⁷ but this adjuvant treatment combination is not approved by the EMA or Food and Drug Administration (FDA).

Other systemic therapies are currently being evaluated as adjuvant therapy for patients with resected stage IIB-IV melanoma. In an ongoing phase II clinical trial, the addition of messenger RNA-4157 to adjuvant pembrolizumab has shown a 49% risk reduction in recurrence and/or death (HR 0.510, 95% CI 0.288-0.906, two-sided nominal P = 0.019) versus pembrolizumab alone.²⁸ A phase III trial is also ongoing.

Treatment of localised melanoma

• Full depth, WLE of primary tumours with safety margins of 0.5 cm for in situ melanomas, 1 cm for tumours with a thickness ≤2 mm and 2 cm for tumours >2 mm is recommended [III, B].

Treatment of locoregional melanoma

• Patients with pT1b-T4b cN0 cM0 melanoma and a positive SLNB should undergo standard follow-up [III, A] ± systemic therapy according to disease stage (see Figure 1) [I, A].
• CLND is not recommended for patients with a positive SLNB [I, E].
• Patients with a negative SLNB can be offered standard follow-up [III, A], clinical trial participation [V, A] or anti-PD-1 therapy for 12 months (stages IIB-IIC) [I, A].
• Enrolment into a clinical trial wherever possible is preferred [V, A].

Treatment of ITMs

• Patients with resectable ITMs should undergo complete excision with clear margins [IV, B]. These patients can also be evaluated for neoadjuvant nivolumab–ipilimumab [I, A; ESMO-Magnitude of Clinical Benefit Scale (ESMO-MCBS) v1.1 score: A; not EMA or FDA approved] followed by adjuvant therapy based on pathological response and BRAF status, neoadjuvant plus adjuvant pembrolizumab [II, A; not EMA or FDA approved] or adjuvant therapy [I, A].
• Patients with resectable disease but a short disease-free interval (<6 months), unresectable satellite, ITMs or inoperable primary tumours of the limbs without additional metastases should be treated with systemic therapy [III, B].

• Patients with unresectable satellite or ITMs may be treated with systemic therapy with anti-PD-1 based immunotherapy or BRAFi–MEKi, according to BRAF mutation status [I, A].

  • ○

    Local therapy with T-VEC, [I, B; ESMO-MCBS v1.1 score: 3], isolated limb infusion or isolated limb perfusion [IV, C], RT [IV, C], electrochemotherapy [IV, C] or limited palliative excision [IV, C] can also be considered (no impact on OS).

Adjuvant RT

• Adjuvant RT is not routinely recommended [III, D].
• RT can be considered for local tumour control in cases of inadequate resection margins of lentigo maligna [III, B].
• Adjuvant RT to the primary excision site should be considered for patients with desmoplastic or neurotropic melanoma for whom adequate (≥8 mm) pathological resection margins cannot be achieved [IV, C].
• RT could be discussed for patients with an R1 resection (resection with microscopic tumour at the margin) or after resection of bulky LN metastases, especially if further surgical clearance is not feasible [III, C].

Adjuvant systemic therapy in stage IIB-IIC melanoma

• Adjuvant therapy with either pembrolizumab [ESMO-MCBS v1.1 score: A] or nivolumab [ESMO-MCBS v1.1 score: A] for 12 months should be considered for patients with stage IIB-IIC disease; treatment discussions with the patient should include consideration of the RFS benefit but lack of mature OS data [I, A].

Adjuvant systemic therapy in resected stage III and IV NED

• Adjuvant systemic therapy options are anti-PD-1 therapy (nivolumab for resected stage IIIB-IV NED [I, A; ESMO-MCBS v1.1 score: no evaluable benefit] or pembrolizumab for resected stage III [I, A; ESMO-MCBS v1.1 score: A]) or dabrafenib–trametinib for patients with resected stage III BRAF V600E-mutated melanoma [I, A; ESCAT score: I-A].

  • ○

    For anti-PD-1-based therapy, treatment discussions with the patient should consider the DMFS and RFS benefits but lack of mature OS data compared with placebo [I, A].

  • ○

    For dabrafenib–trametinib, these discussions should also consider the DMFS and RFS benefits and potential OS benefit for patients with BRAF V600E-mutated melanoma [I, A].

  • ○

    These treatments should be given within 12 weeks of complete resection [I, A].

• Targeted therapy should not be offered to patients with BRAF V600K-mutated melanoma in light of the potential detrimental effect on OS reported in the COMBI-AD trial [II, D].
• For patients with AJCC8 stage IIIA and <1 mm tumour burden, adjuvant systemic treatment is generally not recommended [I, D].
• The use of adjuvant nivolumab–ipilimumab is not recommended for resected stage III melanoma [I, D; not EMA or FDA approved].

• Patients with resectable stage IV melanoma can be offered systemic therapy [V, A], clinical trial [V, A] or metastasectomy or local ablative therapy [III, B] followed by adjuvant anti-PD-1 therapy [I, A].

  • ○

    The use of adjuvant nivolumab–ipilimumab according to the dosing schedule utilised in the phase II IMMUNED trial may be an option for selected patients with resected stage IV melanoma [II, C; not EMA or FDA approved].

Immunotherapy

The superiority of nivolumab over dacarbazine (DTIC) chemotherapy (ChT) for the first-line treatment of patients with BRAF-wild-type (WT) melanoma was demonstrated in the prospective randomised CheckMate 066 trial, with an HR for death of 0.42 (99.79% CI 0.25-0.73, P < 0.001) and an HR for death or progression of disease of 0.43 (95% CI 0.34-0.56, P < 0.001).⁴⁴ Superiority of PD-1 blockade (nivolumab, pembrolizumab) over ipilimumab was demonstrated in two prospective randomised trials, CheckMate 067 and KEYNOTE-006.^45,46 After a minimum follow-up of 10 years, CheckMate 067 had an HR for death for nivolumab versus ipilimumab of 0.63 (95% CI 0.52-0.76)⁴⁷ and KEYNOTE-006 (patients included in the KEYNOTE-587 extension study only) had an HR for death for pembrolizumab (both dose arms combined) versus ipilimumab of 0.71 (95% CI 0.60-0.85).⁴⁸ Based on these trials, PD-1 blockade is now an SoC option for all patients, regardless of tumour BRAF status, in the first-line setting.

The benefit of adding ipilimumab to nivolumab was also assessed in the CheckMate 067 trial. Treatment with nivolumab–ipilimumab according to this trial’s dosing regimen resulted in numerically higher response rates (RRs) and longer response durations, time to subsequent therapies, patients alive after stopping therapy, progression-free survival (PFS) and OS.⁴⁵ By study design, the two nivolumab-containing arms could not be compared. At a minimum follow-up of 10 years, the median OS was 71.9, 36.9 and 19.9 months in the nivolumab–ipilimumab, nivolumab-only and ipilimumab-only groups, respectively. Median MSS was NR, 49.4 and 21.9 months, respectively, and median duration of response (DoR) was NR (>120 months), 103.2 months and 19.2 months, respectively.⁴⁷ Grade 3-4 AEs were reported in 55.0%, 16.3% and 27.3% of patients in the nivolumab–ipilimumab, nivolumab-only and ipilimumab-only groups, respectively,⁴⁹ with no new safety signals observed in subsequent trial reports.⁵⁰

A phase IIIb/IV trial showed that an alternative dosing schedule of nivolumab–ipilimumab (nivolumab 3 mg/kg plus ipilimumab 1 mg/kg) was associated with reduced toxicity.⁵¹ Therefore, this dosing schedule could be discussed for some frail patients. However, as this trial was designed to evaluate the safety (and not the efficacy) profile of this alternative dosing schedule, its general use cannot be recommended.

More recently, an improvement in PFS has been reported for relatlimab–nivolumab. In the phase II/III, multicentre, double-blind, randomised RELATIVITY-047 trial, relatlimab–nivolumab was administered as a fixed-dose q4w to patients with previously untreated unresectable or metastatic melanoma. A median PFS of 10.1 months was reached with relatlimab–nivolumab versus 4.6 months with nivolumab (HR 0.75, 95% CI 0.62-0.92, P = 0.006 log-rank test). The 12-month PFS rate was 47.7% with relatlimab–nivolumab versus 36% with nivolumab. PFS across key subgroups also favoured relatlimab–nivolumab over nivolumab. Grade 3/4 TRAEs occurred in 18.9% of patients in the relatlimab–nivolumab group versus 9.7% in the nivolumab group.⁵² In an updated analysis at a median follow-up of 19.3 months, the HR for PFS by BICR in patients with a PD-L1 expression of <1% (n = 209) and ≥1% (n = 147) was 0.68 (95% CI 0.53-0.86) and 0.96 (95% CI 0.70-1.31), respectively. The OS HRs in these subsets were 0.78 (95% CI 0.59-1.04) and 0.84 (95% CI 0.57-1.24), respectively.⁵³ Based on these data, relatlimab–nivolumab can be considered as a first-line treatment option. For patients who need to discontinue relatlimab–nivolumab due to toxicity, continuation of anti-PD-1 monotherapy can be discussed. In July 2022, the EMA approved the use of relatlimab–nivolumab for the first-line treatment of adults and adolescents (≥12 years of age) with advanced melanoma and a PD-L1 tumour expression of <1%.

Pembrolizumab–lenvatinib as first-line treatment for patients with unresectable or metastatic melanoma was evaluated in the phase III LEAP-003 trial. Despite a significant improvement in PFS observed for the combination in an early interim analysis, findings from a subsequent analysis showed no OS benefit and a significant increase in toxicity and the trial was subsequently discontinued.⁵⁴

Given this collective evidence, treatment decisions should be tailored based on several parameters, including prior (neo)adjuvant therapy received, timing of recurrence on/after adjuvant therapy, resectability status, suitability to receive ICI therapy and PD-L1 status. For patients eligible to receive ICI therapy, the treatment choice of single-agent PD-1 blockade versus nivolumab–ipilimumab or relatlimab–nivolumab should be individualised to each patient.

Targeted therapy

In case of BRAF-mutated melanoma, additional first-line options are provided by BRAFis and MEKis. Combined BRAFi–MEKi is superior to single-agent BRAFi in terms of RRs, PFS and OS.^55–58 However, findings from a post hoc analysis suggest that encorafenib as monotherapy provides a similar OS benefit to combined BRAFi–MEKi (HR 0.93, 95% CI 0.73-1.18) and so could be an option for patients with contraindications to MEKis, although encorafenib is not EMA or FDA approved as monotherapy.⁵⁹ In addition to improved efficacy, skin-related side-effects and the incidence of squamous-cell carcinomas are reduced with the combination, although MEKis add specific toxicities (e.g. muscle, heart, eyes). Single-agent BRAFis should only be used in case of an absolute contraindication for MEKis.

Treatment selection

First-line treatment decisions between targeted therapies and immunotherapies have been evaluated in several prospective trials, with the aim of defining the best sequencing approach. The phase III DREAMSEQ trial included 265 patients with treatment-naive BRAF V600-mutated metastatic melanoma, stratified by Eastern Cooperative Oncology Group PS 0 or 1 and lactate dehydrogenase (LDH) level; patients with untreated melanoma BMs (MBMs) were excluded. Patients were randomised 1 : 1 to receive either nivolumab–ipilimumab (arm A) or dabrafenib–trametinib (arm B), and at disease progression, patients in arm A received dabrafenib–trametinib (arm C) and patients in arm B received nivolumab–ipilimumab (arm D). The median DoR was significantly longer for arm A than for arm B (NR versus 12.7 months, P < 0.001). The PFS showed a trend in favour of arm A (log-rank P = 0.054). The 2-year OS rate for those starting in arm A was 71.8% and was 51.5% for those starting with arm B (log-rank P = 0.010). It is worth noting that only ~50% of patients who had disease progression in arm A or B were enrolled into arm C or D, respectively, since most died within 6 months of their initial disease progression, many due to MBMs. Also, given the significant difference in terms of 2-year OS rates between arms A and B, the data safety monitoring committee recommended the study be closed to accrual and for patients in arm B to be given the option to switch to arm D without disease progression. Given this, the initial question regarding the best therapeutic sequence was not completely answered. Overall, grade ≥3 toxicity was 60% in arm A and 52% in arm B. Grade 5 TRAEs included two patients in arm A and one in arm C.⁶⁰

In the randomised, three-arm, non-comparative phase II SECOMBIT trial, patients with previously untreated, metastatic BRAF V600-mutated melanoma were randomly assigned to arm A (n = 69; encorafenib–binimetinib until disease progression and then ipilimumab–nivolumab), arm B (n = 71; ipilimumab–nivolumab until disease progression and then encorafenib–binimetinib) or arm C (n = 69; encorafenib–binimetinib for 8 weeks followed by ipilimumab–nivolumab until disease progression and then encorafenib–binimetinib). At a median follow-up of 32.2 months, median OS was NR in any arm and >30 patients were alive in all arms. No new safety signals emerged.⁶¹ The 5-year OS rates were 45% in arm A, 52% in arm B and 57% in arm C.⁶²

The randomised, phase II EBIN trial evaluated a total of 2 years of therapy, either with immunotherapy (nivolumab–ipilimumab for four cycles followed by nivolumab; arm A) or encorafenib–binimetinib for 3 months followed by immunotherapy, as per arm A (arm B) in 271 patients with BRAF-V600E/K unresectable stage III/IV melanoma. At a median follow-up of 21 months, there was no PFS benefit associated with the addition of induction targeted therapy (HR 0.87, 90% CI 0.67-1.12, P = 0.36). However, findings from a prespecified subgroup analysis suggested a PFS benefit for induction targeted therapy among patients with LDH >2× the upper limit of normal (ULN; HR 0.46, 95% CI 0.21-1.03). In a post hoc analysis, patients with liver metastases also benefited from the sequential design (HR 0.48, 95% CI 0.28-0.80).⁶³

Based on these results, first-line nivolumab–ipilimumab is the preferred treatment when this can be safely delivered for the first few months (i.e. when a rapid response is not required due to aggressive/symptomatic disease), with targeted therapies reserved for subsequent treatment lines. The optimal duration of induction targeted therapy and the best targeted therapy combination is currently unknown.

Oligometastatic disease

Oligometastatic disease is difficult to define. It is dependent on the number and localisation of metastatic sites and can be resectable or unresectable, but there is currently no consensus regarding its definition. Treatment options mimic those for patients with resectable stage IV melanoma and include systemic therapy, surgery, local treatment (stereotactic RT or ablative therapy) or a clinical trial (see Figure 4). However, based on results from trials evaluating neoadjuvant systemic therapy, the use of first-line systemic therapy instead of ablative therapy, even in patients with resectable oligometastatic disease, seems preferable and should be discussed. Data from clinical trials investigating this question are required.

BRAF-WT melanoma

For BRAF-WT melanoma, approved second-line options are very limited. Thus, consideration for clinical trials and/or personalised approaches is appropriate. If the first-line treatment was anti-PD-1 monotherapy or if patients had primary refractory disease following anti-PD-1 therapy, ipilimumab and ipilimumab–nivolumab are options based on results from the phase II SWOG S1616 trial.^64–66 In this trial, treatment with ipilimumab–nivolumab was associated with a statistically significant improvement in PFS compared with ipilimumab alone (HR 0.63, 90% CI 0.41-0.97, one-sided P = 0.04).⁶⁶ Nivolumab–relatlimab might also represent an option after failure of single-agent anti-PD-1 therapy; in the phase I/IIa RELATIVITY-020 trial, objective response rates (ORRs) of 12.0% and 9.2% were reported for patients who had progressed after one or more than one prior anti-PD-1-containing regimen, respectively. Corresponding median PFS values were 2.1 and 3.2 months, respectively.⁶⁷

In some cases, such as (i) toxicity to anti-PD-1-based immunotherapy which precludes the use of second-line anti-PD-1-based therapy, (ii) rapidly progressing disease or (iii) high tumour volume with symptomatic disease, clinical trials including bispecifics, T-cell engagers, etc. should be a preferred option. ChT with DTIC, carboplatin–paclitaxel, temozolomide or fotemustine can be discussed. However, none of these ChT regimens provide an OS advantage⁶⁸ and they are associated with a low RR, short PFS and increased toxicity in patients who have progressed on ICI therapy.⁶⁹

TIL therapy is another option which can be manufactured using different techniques. However, one of the limitations of this therapy is the time needed for manufacturing, which is currently 3-6 weeks, making it an option only for a selected group of patients. TIL therapy is currently not EMA approved for use in this setting.

In a phase II trial of TIL therapy, lifileucel (an autologous, centrally manufactured TIL product) demonstrated durable responses in patients with previously treated metastatic melanoma and limited treatment options. In this trial, patients received a non-myeloablative lymphodepletion regimen, a single infusion of lifileucel and up to six doses of high-dose interleukin-2 (IL-2). The ORR was 36%, with two complete responses (CRs) and 22 partial responses. The disease control rate (DCR) was 80% and the median DoR was NR after a median follow-up of 18.7 months. In the subset of patients with primary refractory disease following prior anti-PD-1 therapy, the ORR and DCR were 41% and 81%, respectively. The safety profile was consistent with AEs associated with non-myeloablative lymphodepletion and IL-2.⁷⁰

In an open-label phase III trial, 168 patients (86% with anti-PD-1-refractory disease) with unresectable stage IIIC-IV melanoma (AJCC7 criteria) were randomised 1 : 1 to receive either TILs (manufactured at each trial centre) or ipilimumab. Infusion of TILs was preceded by non-myeloablative, lymphodepleting ChT followed by high-dose IL-2. After a median follow-up of 33.0 months, median PFS was 7.2 months for the TIL group versus 3.1 months for the ipilimumab group (HR 0.50, 95% CI 0.35-0.72, P < 0.001). The ORR was 49% for the TIL group and 21% for the ipilimumab group, with 20% and 7% achieving a CR, respectively. Median OS was 25.8 months for the TIL group and 18.9 months for the ipilimumab group (HR 0.83, 95% CI 0.54-1.27, P = 0.39). Grade ≥3 TRAEs occurred in all TIL-treated patients and 57% of ipilimumab-treated patients.⁷¹ Currently, TIL therapy remains a highly toxic treatment option for selected patients who can tolerate its side-effects, administered within regional reference centres.⁷² Current clinical trial evidence suggests that those who derive most benefit are young patients with stage IV M1a-c melanoma, PS 0, normal LDH and one to three prior treatments.⁷³

In the phase II, LEAP-004 trial, pembrolizumab–lenvatinib was evaluated in patients progressing within 12 weeks of the last dose of an anti-PD-1 inhibitor, given as monotherapy or with other therapies, including CTLA-4 inhibitors. In the overall population, after a median follow-up of 15.3 months, the ORR was 21.4%, median DoR was 8.3 months, median PFS was 4.2 months and median OS was 14.0 months.⁷⁴ Pembrolizumab–lenvatinib is not EMA or FDA approved for use in this setting.

BRAF-mutated melanoma

For BRAF-mutated melanoma, all the options available for BRAF-WT melanoma are still valid, with the addition of combined BRAFi–MEKi therapy if not already used as the immediate prior treatment. BRAFi–MEKi therapy after disease progression with first-line immunotherapy should be offered.

NRAS-mutated melanoma

For NRAS-mutated melanoma, due to the limited efficacy of MEKis, first-line immunotherapy options are identical to those for NRAS-WT melanoma. Binimetinib as a single agent, however, can be considered for patients who do not benefit from prior anti-PD-1 therapy, in accordance with findings from the NEMO trial, but there is no OS benefit and its use as monotherapy is not EMA approved.⁷⁵

General recommendations

• Patients with metastatic melanoma should have metastases (preferably) or the primary tumour screened for the detection of BRAF V600 mutation [IV, A; ESCAT score: I-A].

  • ○

    If no tumour tissue is available, circulating tumour DNA may be an alternative [III, C].

  • Enrolment into a clinical trial wherever possible is preferred [V, A].
• In addition to the treatment options outlined below, palliative resection [IV, C] and/or RT [IV, B] and/or T-VEC [I, C] can be considered for symptomatic extracranial disease.

First-line treatment

• First-line treatment decisions must take into consideration prior neoadjuvant and/or adjuvant therapy received, timing of recurrence on/after adjuvant therapy, resectability status and suitability to receive ICI therapy, as outlined in Figure 4 [V, A]. Primary or secondary resistance must be considered as this is also an eligibility criterion for clinical trials in pretreated patients [V, A].

• Patients with treatment-naive resectable disease can be offered the following:

  • ○

    Stage III:

    • ▪

      Wide excision of the primary tumour [III, B].

    • ▪

      Neoadjuvant nivolumab–ipilimumab [I, A; ESMO-MCBS v1.1 score: A; not EMA or FDA approved] followed by adjuvant therapy based on pathological response and BRAF status.

    • ▪

      Neoadjuvant plus adjuvant pembrolizumab [II, A; not EMA or FDA approved].

    • ▪

      Adjuvant anti-PD-1 therapy [I, A] or dabrafenib–trametinib for BRAF V600E-mutated tumours [I, A].

  • ○

    Stage IV:

    • ▪

      Clinical trial [V, A].

    • ▪

      Metastasectomy or local ablative therapy [III, B] followed by adjuvant anti-PD-1 therapy [I, A].

    • ▪

      Anti-PD-1 therapy alone [V, A].

• For patients with unresectable disease, first-line ipilimumab–nivolumab [ESMO-MCBS v1.1 score: A/4] is a preferred option for all patients regardless of BRAF status when this can be safely delivered for the first few months (i.e. when a rapid response is not required due to aggressive/symptomatic disease) [I, A].

• First-line nivolumab [I, A; ESMO-MCBS v1.1 score: A/4] or pembrolizumab [I, A; ESMO-MCBS v1.1 score: A/4] is also recommended.

  • ○

    Nivolumab–relatlimab can be offered as first-line treatment but EMA approval is only for patients with tumour cell PD-L1 expression <1% [I, B; ESMO-MCBS v1.1 score: 3; EMA approved for PD-L1 expression <1%, FDA approval is regardless of PD-L1 expression].

• If anti-PD-1-based therapy is not available or patients are considered ineligible for its use, BRAFi–MEKi combination therapy (dabrafenib–trametinib [ESMO-MCBS v1.1 score: 5]; vemurafenib–cobimetinib [ESMO-MCBS v1.1 score: A/5]; binimetinib–encorafenib [ESMO-MCBS v1.1 score: A/5]) is also an option in the first line for patients with BRAF-mutated melanoma [I; A; ESCAT score: I-A].

  • ○

    BRAFi–MEKi for 8-12 weeks followed by ipilimumab–nivolumab (as per SECOMBIT arm C or EBIN arm B) is also an option, especially for patients with high LDH levels and/or liver metastases [II, C; ESCAT score: I-A; induction targeted therapy is not EMA or FDA approved].

  • ○

    For patients in whom the decision to treat with targeted therapy has been made, those who cannot receive a MEKi (e.g. due to cardiovascular comorbidities, a recent BM bleeding event, history of retinal detachment or other ophthalmological contraindications) can be offered encorafenib as monotherapy [II, B; not FDA or EMA approved].

• Patients with BRAF-mutated melanoma who have relapsed on or within 6 months of adjuvant BRAFi–MEKi therapy and who have an immediate or absolute contraindication to ICI can be offered the following first-line treatments:

  • ○

    BRAFi/MEKi (if >3 months after stopping adjuvant BRAFi/MEKi) [V, A].

  • ○

    ChT [II, C] (no OS benefit).

• First-line immunotherapy options for patients with NRAS-mutated melanoma are identical to those for patients with NRAS-WT disease [I, A].

Second-line treatment

• Treatment options for the second-line setting depend on the therapy used in the first line and include ipilimumab–nivolumab [II, B], pembrolizumab [I, A; ESMO-MCBS v1.1 score: A/4], nivolumab [II, B], ipilimumab [II, B; ESMO-MCBS v1.1 score: 4] and BRAFi–MEKi combination therapy for patients with BRAF-mutated melanoma [II, B; ESCAT score: I-A]. Only pembrolizumab and ipilimumab monotherapy are EMA or FDA approved for second-line use.
• Nivolumab–relatlimab might also represent an option after failure of single-agent anti-PD-1 therapy [III, B; not EMA or FDA approved as second-line therapy].
• TIL therapy is an aggressive treatment option for selected patients (young, stage IV M1a-c melanoma, PS ≤1, LDH <2× ULN and 1-3 prior treatments) who can tolerate its side-effects [II, B; not EMA or FDA approved].

Subsequent lines of treatment

• Third-line treatment rechallenge with the drug class (BRAFi–MEKi [IV, C] or ICI [IV, B]) not used in the immediate previous line can be considered, if feasible.
• If clinical trials, ICIs or BRAFis/MEKis are not available, ChT may be administered as later-line therapy [IV, C], with modest activity and no impact on OS.
• For patients with NRAS-mutated melanoma, binimetinib as a single agent can be offered to patients who do not benefit from prior anti-PD-1 therapy [III, C].

MBMs

• There are currently no systemic treatment options specifically approved for use in treating MBMs; enrolment into a clinical trial wherever possible is preferred [V, A].
• Patients with MBMs should be evaluated for stereotactic radiosurgery (SRS) [III, B]. Early concurrent SRS may be preferred over late SRS as salvage treatment [IV, C]. Since multiple sessions of SRS can be carried out at different time points of the disease course, close monitoring using MRI is recommended so that SRS can be added when indicated [IV, B].

• Patients with asymptomatic MBMs should preferably be treated upfront with nivolumab–ipilimumab [II, A].

  • ○

    If unsuitable for immunotherapy, patients with asymptomatic MBMs and BRAF V600-mutated melanoma can be offered BRAFi–MEKi [III, B; ESCAT I-A].

• For patients with symptomatic MBMs requiring steroids (<10 mg/day prednisolone or equivalent):

  • ○

    BRAFi–MEKi if BRAF V600-mutated can be offered [III, A; ESCAT I-A].

  • ○

    Nivolumab–ipilimumab (BRAF-mutated or BRAF-WT) can be offered [III, A].

  • ○

    Neurosurgery should be discussed, especially if an accessible, resectable tumour is causing symptoms as this may render the patient asymptomatic and provide a bridging strategy to nivolumab–ipilimumab [IV, C].

• Patients with MBMs and neurological symptoms requiring steroids (>10 mg prednisolone/day or equivalent) for whom local therapy is not an option can be considered for the following:

  • ○

    Clinical trial [V, A].

  • ○

    BRAFi–MEKi (if BRAF V600-mutated) [III, A; ESCAT I-A].

  • ○

    Ipilimumab–nivolumab [III, B].

  • ○

    ChT (if BRAF-WT) [IV, C].

• Patients with leptomeningeal disease can also receive the above therapies. Local treatment (RT [III, B] or intrathecal nivolumab [III, C]) can also be considered.
• Best supportive and palliative care should be discussed and activated for all patients with MBMs [V, A].