Targeting Immune Checkpoint Pathways in Melanoma: Triumphs and Challenges

Abstract

The immune checkpoint inhibitors (ICIs) have revolutionized the treatment of advanced melanoma by significantly increasing survival rates, with the promise of durable disease remission in some patients. Herein we review the role of immune checkpoints in melanoma; the history of melanoma immunotherapy; pivotal clinical trial data for ipilimumab, pembrolizumab, nivolumab and relatlimab; and the current clinical role of each ICI. We discuss the challenges that accompany these triumphs in the treatment of melanoma, including: how to distinguish between responders and nonresponders; how to optimize ICI dosing and combinatorial approaches; and the best practices for monitoring response and managing immune-related toxicities. We offer our perspective on the financial toxicity of ICIs and new developments that could deliver answers to current challenges.

Plain language summary

Immune checkpoint inhibitors are a class of medication called immunotherapy that use the patient’s immune system as a tool to combat cancer. When used in advanced melanoma, they increase the chance of survival. In this article we review the role of such treatments in melanoma, the history of melanoma immunotherapy and important information about the four approved drugs for this type of cancer: ipilimumab, pembrolizumab, nivolumab and relatlimab. We discuss the challenges of treating patients with these medications, how to distinguish patients who might respond from those who might not, how to determine the best way to dose these treatments and the best way to monitor side effects. We also present our perspectives on the cost/benefit of these treatments.

Keywords::

• CLTA-4
• immune checkpoint
• immunotherapy
• ipilimumab
• LAG-3
• melanoma
• nivolumab
• PD-1
• pembrolizumab
• relatlimab

Melanoma incidence continues to increase, and it remains the most lethal skin cancer, responsible for more than 80% of skin cancer-related deaths [1]. Melanoma originates in melanocytes, skin cells responsible for producing melanin upon UV stimulation. Since 1970, it has been classified into four different stages according to the depth of invasion: stages I and II, localized melanomas, correspond to lower risk of metastasis, and surgical removal is the only required treatment [2]; stages III–IV are characterized by regional and distant (metastatic) spreading of tumors, requiring both surgical resection and systemic therapy [3]. The median overall survival after the onset of distant metastases is only 6–9 months, and the 5-year overall survival rate is <5% [4]. Melanoma has the highest mutation rate of all cancer types [2,5,6], and more than half of melanoma cases (∼60%) carry activating mutations in the BRAF oncogene [7–9]. Initial treatment with the combination of a BRAF inhibitor and a MEK inhibitor, such as dabrafenib and trametinib, improves progression-free survival [7]. However, while these molecularly targeted therapies have high initial response rates of 70%, almost all patients acquire drug resistance within 1 year post-treatment [10].

Due to their high mutational characteristics, melanoma tumors can generate immunogenic antigen epitopes that activate the immune system [2,5,6]. Tumor antigens are broadly classified as tumor-associated antigens (TAAs), which are also expressed in low quantities on normal cells, or tumor-specific antigens, which are mutated self-antigens expressed only by tumor cells. In melanoma, a subset of TAAs, the melanoma differentiation antigens, are expressed only by melanocyte-derived cells, both normal and malignant [2,11]. Tumor antigens are presented by antigen-presenting cells (APCs) to T-cell receptors (TCRs) [2,11,12] through the interaction of MHC class I or II, requiring costimulatory signals from CD28 receptors on T cells and cytokines [12]. After antigen priming and activation, lymphocytes are then directed to the tumor site, becoming tumor-infiltrating lymphocytes (TILs) [2,11], and, in the case of cytotoxic CD8⁺ T cells, are activated by MHC class I to become responsible for killing tumor cells [2]. TILs can also include Tregs, natural killer (NK) cells, macrophages, dendritic cells (DCs) and myeloid-derived suppressor cells [11]. Tumor-specific antigens generally have increased immunogenicity due to lack of central tolerance by T cells [11]. In contrast, TAAs typically do not activate the immune system due to central immune tolerance against them [11].

Despite the intrinsic immunogenicity of melanoma, early clinical investigations of cancer vaccines and cytokines failed to demonstrate sufficient efficacy for regulatory approval [2]. Since then, several mechanisms of suppressing or evading antitumor immune responses have been discovered that explain the lack of response to these immunotherapies. Highly mutagenic melanoma tumor cells can evade the immune system by several mechanisms, such as reducing MHC class I or decreasing TAAs and melanoma differentiation antigens on their cell surface, consequently reducing antigen presentation to APCs and removing cytotoxic T-cell stimulation [11]. Melanoma cells can secrete TGF-β, prostaglandin E2 and suppressive cytokines, leading to immune-cell inhibition [2,11]. Fas ligand, whose expression may occur on melanoma cells, is responsible for apoptosis induction in T cells [11]. NK cells may also have their activity decreased through the downregulation of their ligands on tumor cells [11]. However, it was the identification of the overexpression of immune checkpoints, such as PD-L1/2 and CTLA-4 ligands, responsible for immune inactivation [2,11], that was considered one of the largest breakthroughs in cancer immunotherapy, and these checkpoints became novel drug targets.

Treatment with blockading antibodies against the immune checkpoints achieves long-term disease remission in melanoma patients [13], and the combination of two immune checkpoint inhibitors (ICIs) can further increase response rates [14]. ICIs are now classified as preferred first-line therapy for metastatic and unresectable melanoma [15]. Instead of targeting tumor cell growth, their function is to increase the antitumor immune response [12]. In fact, for stage III–V disease, immunotherapy has been shown to significantly improve patient outcomes over the last 30 years [2]. In this regard, this review aims to explore the role of ICIs in melanoma, current clinical challenges and promising research in the field.

History of melanoma immunotherapy

The start of the cancer immunology field is credited to Ehrlich who was one of the first scientists to put forth the theory that the immune system is able to recognize and protect against cancer, while the beginnings of modern cancer immunotherapy is attributed to William Coley, a bone sarcoma surgeon (Table 1). In 1891, Coley performed the first documented successful immunotherapy treatment for cancer when he injected tumors with different mixtures of live, inactivated, Streptococcus pyogenes and Serratia marcescens strains in order to induce infection and recruit the immune system specifically to that part of the body [1]. In 1908, Paul Ehrlich and Elie Metchnikoff received the Nobel Prize in Physiology or Medicine for their studies on mechanisms of immunological defense, which became the foundation for humoral and cellular immunology [2]. Further major developments in understanding immune responses against cancer came in 1957–76, when Prehn and Main showed that tumor cells could generate a strong immune response by expressing tumor-specific antigens [16]. Klein et al. demonstrated that an immune response could also be induced in primary autochthonous tumor hosts [17], and interferon and IL-2 were identified as tumoricidal cytokines [18,19]. Although interferon was first discovered in 1957, it was not approved for the treatment of cancer until 1995 when a trial in metastatic melanoma showed a 42% improvement in the relapse-free interval of high-risk, previously resected melanoma patients [20]. Recombinant IL-2 was first approved for use in renal cell carcinoma, but gained an indication for melanoma treatment in 1998 based on a clinical trial in the metastatic population that showed an objective response in 16% of patients on therapy, with 6% of patients (n = 17) in the trial experiencing a complete response to therapy. In patients with a complete response, the duration of that response ranged from 24 to 106 months, showing that this type of therapy had the potential to induce durable disease remission [21]. Both interferon and IL-2 are immunostimulatory cytokines that increase immune responses against a patient’s cancer. However, patients receiving IL-2 need an exceptionally high level of care and monitoring during treatment and must be admitted to an intensive care unit due to severe adverse effects such as capillary leak syndrome (which can lead to hypotension) and hepatic and renal dysfunction [22].

Table 1. Timeline of immunotherapy developments.

Year	Development	Ref.
c.1891	William Coley described using bacterial inoculation of tumors to incite immune-based treatment of cancer	[23]
1908	Paul Ehrlich and Elie Metchnikoff set the foundation for humoral and cellular immunology	[24]
1957	Interferon discovered by Isaacs and Lindenmann	[20]
1957	Immune surveillance of cancer theory published by L Thomas and F Macfarlane Burnet	[23]
1982	Discovery of T-cell receptor by J Allison, B McIntyre and D Bloch	[23,25]
1995	Interferon approved for use in melanoma	[20]
1996	CTLA-4-blocking antibodies discovered to treat tumors in animal models	[26]
1998	Recombinant IL-2 approved by US FDA for melanoma	[21]
2011	Ipilimumab, anti-CTLA-4, is the first checkpoint inhibitor approved by the FDA for melanoma	[27]
2014	Pembrolizumab and nivolumab, two anti-PD-1 medications, approved by the FDA for melanoma	[28,29]
2015	Talimogene laherprepvec, the first oncolytic viral therapy in the USA, is approved for local treatment of unresectable cutaneous, subcutaneous and nodal lesions in patients with melanoma recurrent after initial surgery	[30]
2022	Relatlimab in combination with nivolumab approved by FDA for melanoma	[31]

An unanswered question remained: why did the immune system need to be told to become more active in order to eradicate cancer cells? In the early 1990s, the works of James Allison and Tasuku Honjo on CTLA-4 and PD-1, respectively, shed light on the molecular mechanisms that act as brakes on the immune system and prevent it from doing its job [32]. These ‘immune checkpoints’ may be found on tumor cells, creating an ‘invisibility cloak’ that allows them to escape detection by shutting down immunosurveillance and immune attack by local T cells. This ability to evade immune destruction is now accepted as the eighth hallmark of cancer [33].

The interaction between tumor cells and the immune system can be divided into three phases: initial immune elimination of tumor cells; equilibrium between immune defense and proliferating tumor cells; and tumor cell immune escape, enabling tumor growth [33,34]. In the last phase, tumor cells upregulate multiple immunosuppression mechanisms, including secretion of tumor growth factors and immune inhibitory cytokines, recruitment of Tregs and myeloid-derived suppressor cells, production of metabolic modulators [12,35], expression of Fas ligand, which induces CD8⁺ T-cell apoptosis [2,11], and expression of immune checkpoint molecules [2]. Although there are many immune checkpoints, only CTLA-4, PD-1 and LAG-3 are proven therapeutic targets in melanoma patients (Figure 1) [2,11].

Figure 1. Therapeutically targeted immune checkpoints.

(A) Immune checkpoint targets involved in naive T-cell antigen priming/activation. (B) Immune checkpoint targets involved in effector T-cell activity.

APC: Antigen-presenting cell; TCR: T-cell receptor.

CTLA-4 (CD152) is constitutively expressed on Tregs and competes with CD28 on T cells for binding to B7-1/2 receptors on APCs [2,12]. Its affinity for CD80 and CD86 (B7 family receptors) is much higher than that of the intrinsically expressed CD28, favoring T-cell inhibition. Concurrently, CTLA-4 can also promote the removal of CD28 receptors and intracellular domains from APCs by transendocytosis, limiting costimulatory signals for T-cell activation and thus inducing immune tolerance [12]. Melanoma cells may also express MHC class II, a molecule typically only present on APCs, promoting the activation of FoxP3⁺CD4⁺ T-helper cells (Tregs) [36]. When present, Tregs inhibit antigen presentation by APCs in the tumor milieu and its draining lymph nodes [37]. In addition, Tregs can also express other immune checkpoint molecules such as PD-1 [35,38], which has also been related to clinical outcomes in patients with MHC class II expression in melanoma cells [36]. Ipilimumab (Yervoy^®), an anti-CTLA-4 antibody, entered clinical trials in 2000 and was approved for use in melanoma in 2011 [23]. Its first phase III clinical trial took place in previously treated patients with advanced melanoma. The overall response rate in patients receiving ipilimumab (3 mg/kg) alone was 10.9%, with a median overall survival of 10.1 months. Of the 15 patients with an objective response, nine (60%) maintained their response for at least 2 years [39]. This durable response expanded on the previously seen data of IL-2, showing the short- and long-term promise of immune checkpoint inhibition, all while decreasing the need for patient intensive care admission, as ipilimumab can be given in the outpatient setting every 3 weeks [40].

PD-1 (PDCD1/CD279) expression starts during CD4⁻CD8⁻ T-cell maturation in the thymus, and it can be further expressed in CD4⁺ or CD8⁺ T cells, NK T cells, monocytes, B cells and DCs [5,38]. Its expression is modulated in normal tissues to avoid damage by the immune system [6] and it is not normally expressed in circulating T cells, but it can be induced by TCR activation or cytokine stimulation (e.g., by IL-2, IL-7, IL-15, IL-21 and TGF-β) [12]. Its ligands are mainly induced by IFN-γ [12] and can be either PD-L1 (B7-H1/CD274; found on fibroblasts, APCs and tumor cells) or PD-L2 (B7-DC/CD273/PDCD1LG2; present on mast cells, DCs and macrophages) [5,12,35]. When bound to PD-1 on T cells, PD-L1 and PD-L2 lead to T-cell anergy [5], reduction of proliferation, inhibition of cytokine release, and apoptosis [38]. This interaction plays an important role in maintaining an adequate balance between T-cell activation and tissue damage against self-antigens, chronic infections and tumors [35,38]. While normal tissues rarely express these inhibitory ligands, tumor cells have the ability to overexpress them, using this mechanism to evade the immune cells and reduce the antitumor response [5,35,38], thereby facilitating tumor growth [35]. The first PD-1 inhibitor approved by the US FDA was pembrolizumab (Keytruda^®) in September 2014, with nivolumab (Opdivo^®) approved in December 2014 [41]. The FDA granted accelerated approval for pembrolizumab in unresectable or metastatic melanoma patients who had progressed on ipilimumab or a BRAF inhibitor, in BRAF-positive disease, based on the data from the KEYNOTE-001 trial [42]. In this study, patients received pembrolizumab 2 mg/kg every 3 weeks, 10 mg/kg every 3 weeks or 10 mg/kg every 2 weeks. Patients receiving pembrolizumab had an overall response rate of 33% and a median overall survival of 23 months [42]. The second PD-1 inhibitor, nivolumab (Opdivo^®), was granted FDA accelerated approval based on the CheckMate 037 trial [43]. This trial focused on the treatment of patients who had unresectable or metastatic disease that progressed after ipilimumab and/or a BRAF inhibitor, in BRAF-positive disease. This trial was notable for comparing nivolumab against the then standard of care for previously treated disease, cytotoxic chemotherapy (carboplatin combined with either dacarbazine or paclitaxel). Patients received nivolumab at 3 mg/kg every 2 weeks and showed an improved objective response rate of 31.7% compared with 10.6% in the chemotherapy group [43,44]. While these efficacy improvements are significant, there still were 68.3% of patients who did not benefit from treatment, and it was not reported whether any of these nonresponders experienced significant adverse events.

LAG-3 presents similar activity to that of CTLA-4 by inhibiting the TCR signaling pathway [45]. The expression of LAG-3 can be upregulated in melanoma tumors, being expressed on activated CD4⁺ and CD8⁺ T cells, Tregs, NK cells, B cells, plasmacytoid DCs and tumor-associated macrophages [45,46]. Investigations of LAG-3 expression on CD8⁺ T cells suggest that LAG-3/MHC II interactions lead to T-cell anergy [47]. LAG-3 binds with higher affinity to stable MHC II/peptide complexes on APCs than CD4, blocking TCR–CD4 binding and consequently impairing T-cell function, while its crosslinking with CD3 diminishes T-cell proliferation [45]. Moreover, LAG-3 expression on Tregs is crucial to their activity. FoxP3⁺LAG-3⁺ Tregs in tumors promote immunosuppression by secretion of IL-10 and TGF-β [45]. T cells can express LAG-3 concomitantly with PD-1 [31,45], suggesting that the blockade of both checkpoints may enhance antitumor effects. Preclinical investigations verified the synergistic enhancement of therapeutic response in melanoma with dual blockade of LAG-3 and PD-1 [48,49], and the clinical efficacy of this combinatorial treatment was initially evaluated in a phase I/IIa study (NCT01968109), which included patients with melanoma. In this study, relatlimab (BMS-986016), a LAG-3 blockading antibody, in combination with nivolumab demonstrated peripheral T-cell activation, evidence of anticancer activity and tolerability [50]. Subsequently, relatlimab in combination with nivolumab (Opdualag™) was granted FDA approval in March 2022 for patients aged 12 years and older with unresectable or metastatic melanoma [51]. This approval was based on the RELATIVITY-047 trial, in which patients received either nivolumab 480 mg monotherapy or relatlimab 160 mg with nivolumab 480 mg in a fixed-dose combination product every 4 weeks. In the monotherapy arm, median progression-free survival was 4.6 months, compared with 10.1 months in the combination therapy arm. This increased disease control was accompanied by an increase in grade 3–4 treatment-related adverse events, which occurred in 9.7% of monotherapy arm patients compared with 18.9% in those receiving combination therapy [31].

Role of ICIs

Currently, four ICIs are approved by the FDA for melanoma (Table 2): ipilimumab (CTLA-4 inhibitor), nivolumab (PD-1 inhibitor), pembrolizumab (PD-1 inhibitor) and relatlimab (LAG-3 inhibitor) [52]. ICIs were the first class of therapy shown to improve survival for patients with melanoma and are considered standard of care. Their current role in therapy begins with stage IIIA disease, which is the lowest risk group for which adjuvant therapy should be considered, as there are no currently FDA-approved adjuvant ICI therapies approved in the stage I/II disease setting.

Table 2. Approved immune checkpoint inhibitors for melanoma.

Drug	Target	Initial approval	Melanoma indication	Ref.
Ipilimumab (Yervoy^®)	CTLA-4	2011	• Unresectable or metastatic melanoma as a single agent in adult and pediatric patients (12 years and older) as a single agent • Unresectable or metastatic melanoma in combination with nivolumab in adults • Adjuvant treatment in cutaneous melanoma with pathological involvement of regional lymph nodes >1 mm that have undergone complete resection, including total lymphadenectomy	[27]
Pembrolizumab (Keytruda^®)	PD-1	2014	• Unresectable or metastatic melanoma • Adjuvant treatment of stage IIB, IIC or III melanoma following complete resection in adult and pediatric patients (12 years and older)	[28]
Nivolumab (Opdivo^®)	PD-1	2014	• BRAF V600 wild-type or mutation-positive unresectable or metastatic melanoma as a single agent in adult patients • Unresectable or metastatic melanoma in combination with ipilimumab in adult patients • Melanoma with lymph node involvement or metastatic disease after complete resection as adjuvant treatment in adult patients	[29]
Relatlimab in combination with nivolumab (Opdualag™)	LAG-3 + PD-1	2022	• Unresectable or metastatic melanoma in adult and pediatric patients (12 years or older)	[51]

In the adjuvant setting, PD-1 inhibitors are the preferred treatment option as they have demonstrated improved efficacy and safety compared with ipilimumab. The CheckMate 238 trial demonstrated that ipilimumab had inferior recurrence-free survival (60.8 vs 70.5%) and greater treatment-related grade 3 or 4 toxicity (45.9 vs 14.4%) compared with nivolumab monotherapy in resected stage III or IV melanoma [44]. Long-term follow-up from this study showed that the superiority of nivolumab versus ipilimumab was maintained for up to 4 years, with recurrence-free survival rates of 52 versus 41%, respectively [53]. Pembrolizumab monotherapy was approved based on the EORTC-1325 study investigating pembrolizumab versus placebo as adjuvant therapy in patients with resected, high-risk stage III melanoma. Pembrolizumab demonstrated improved 1-year recurrence-free survival rates of 75.4% compared with 61% in the total patient population; in patients with PD-L1-positive tumors, recurrence-free survival was 77.1 versus 62.6% [54]. Due to high rates of toxicity, adjuvant ipilimumab monotherapy is not commonly used in clinical settings, except for selected patients with resected stage IV disease and prior anti-PD-1 exposure [15].

Combination therapy with anti-PD-1 and ipilimumab may be preferred in patients with good performance status and is associated with higher rates of response, progression-free survival and overall survival [55]. However, increased frequency and severity of immune-related adverse events (irAEs) is a major limitation. The CheckMate 067 trial compared adjuvant nivolumab alone versus a nivolumab plus ipilimumab combination versus ipilimumab alone in unresectable stage III or IV disease. The findings revealed progression-free survival of 6.9 versus 11.5 versus 2.9 months, respectively, with grade 3 or 4 toxicity occurring in 16.3 versus 55 versus 27.3% [56]. The phase II IMMUNED trial investigating adjuvant nivolumab plus ipilimumab versus nivolumab monotherapy versus placebo in stage IV disease following surgery or radiation demonstrated 2-year recurrence-free survival rates of 70, 52 and 42%, respectively [57]. Treatment-related grade 3–4 adverse events were reported in 71% in the combination arm versus 27% on nivolumab monotherapy, with treatment discontinuation occurring in 62 versus 13%, respectively [57].

The PD-1 inhibitors are the preferred first- and second-line regimens in metastatic or unresectable disease and can be used after a patient progresses on ipilimumab therapy. Findings from the CheckMate 037 phase III trial demonstrated that nivolumab monotherapy after failure on ipilimumab was a viable option, with an objective response rate of 32 versus 10.6% when compared with chemotherapy (carboplatin with dacarbazine or paclitaxel) [43]. A subsequent trial, CheckMate 064, attempted to determine whether sequential administration of nivolumab followed by ipilimumab, or the reverse sequence, could improve safety without compromising efficacy in patients with stage III/IV disease who received one or fewer lines of previous systemic therapy. The primary end point was treatment-related grade 3–5 adverse events. The primary outcome was experienced by 50% of patients in the nivolumab followed by ipilimumab arm, compared with 43% in the reverse-sequence arm; after 20 months of follow-up, the median overall survival was not reached versus 14.7 months in each of the arms, respectively [58]. Nivolumab followed by ipilimumab has more toxicity, but appears to be a more clinically beneficial option compared with the reverse sequence. Thus nivolumab monotherapy is a preferred second-line regimen.

Pembrolizumab monotherapy in the second-line setting was approved based on findings from the KEYNOTE-006 phase III trial in patients with advanced melanoma who received one or fewer lines of previous systemic therapy. This trial compared the efficacy of pembrolizumab dosed every 2 weeks versus every 3 weeks versus ipilimumab. Approximately 33% of patients had one previous line of therapy, with about 50% having received a targeted therapy, 40% chemotherapy and 10% immunotherapy. Estimated 1-year overall survival was 74.1 versus 68.4 versus 58.2%, respectively, and the rates of grade 3 and above toxicities were 13.3 versus 10.1 versus 19.9% [59]. Efficacy was similar between the two pembrolizumab groups and was superior to ipilimumab monotherapy.

Both anti-PD-1 monotherapy and anti-PD-1 and ipilimumab combination therapy have been shown to provide durable disease control. However, ipilimumab’s place in therapy in the second-line setting for metastatic disease hinges around patients who experience progression during or shortly after anti-PD-1 monotherapy or in those with central nervous system disease. In both scenarios, the preferred regimen is combination therapy with an anti-PD-1 agent and ipilimumab – and especially so in those with asymptomatic brain metastasis not requiring corticosteroids, due to superior intracranial activity compared with anti-PD-1 monotherapy agents [15].

Cytotoxic therapy can be considered in patients with contraindications to immunotherapy (disease progression on prior therapy, unacceptable toxicity, comorbidities etc.). However, chemotherapy agents have not demonstrated superior outcomes nor improved overall survival in phase III trial settings [15]. Cytotoxic agents that have been used alone or in combination include, but are not limited to: dacarbazine, temozolomide, paclitaxel (traditional or albumin-bound), carboplatin/paclitaxel and cisplatin/vinblastine/dacarbazine. Among these options, combinations of carboplatin and paclitaxel or single-agent temozolomide are preferred [15]. In general, immunotherapy and targeted therapy are preferred for treatment of unresectable or distant metastatic disease. Guidelines suggest that the use of ICIs in combination with targeted therapies (e.g., BRAF and MEK inhibitors) has demonstrated efficacy and overall survival benefit in patients with targetable mutations [15]. However, discussion about targeted therapies is beyond the scope of this article.

Clinical problems/challenges

Redefining the response criteria for immunotherapy

While ICI therapy has advanced many of the patient-focused metrics beyond initial expectations in the cytotoxic therapy era of melanoma treatment, there are critical lingering questions about how to maximize the benefits and minimize the negatives associated with these therapies. One of the first challenges encountered was the slow onset of action compared with prior agents, which could be interpreted as progression of disease during a clinical trial. Early trials adopted a new approach to defining progression, such as requiring confirmation of disease progression by diagnostic imaging ≥4 weeks (KEYNOTE-001); using the traditional Response Evaluation Criteria In Solid Tumors for primary efficacy outcomes, but modified immune-related response criteria for managing treatment (KEYNOTE-006); or waiting until the 12-week mark to rescan (CheckMate 067) [10,42,43]. Hodi et al. reviewed the KEYNOTE-001 data and found that 14% of patients who survived ≥12 weeks had progressive disease via the traditional Response Evaluation Criteria In Solid Tumors (v. 1.1), but nonprogressive disease based on immune-related response criteria [60]. Evidence suggests that most responses to ICI therapy will occur within 6 months [15]. Clinicians have to monitor their patients closely if there is a concern for progressive disease to determine the best course of therapy moving forward.

Optimization of dosing & regimen

While ipilimumab dosing for unresectable disease has stayed constant at 10 mg/kg every 3 weeks, the same cannot be said for our PD-1 targeting agents [27]. Following the initial clinical trials for nivolumab and pembrolizumab, there has been an evolution in both dose and duration of these therapies [15]. Nivolumab was originally approved at 3 mg/kg every 2 weeks until disease progression or unacceptable toxicity, but has switched to flat dosing of either 240 mg every 2 weeks or 480 mg every 4 weeks [15]; duration remains unchanged [29]. Pembrolizumab started at 2 mg/kg every 3 weeks until disease progression or unacceptable toxicity or 24 months, but now also supports a flat dose of 200 mg every 3 weeks or 400 mg every 6 weeks, and no longer carries the duration cap of 24 months [15,28]. In terms of the most appropriate data regarding duration of therapy, outcomes have been mixed: FDA-led pooled analyses reported a decrease in tumor size of 30% or more in 19% of patients in whom PD-1 directed therapy was continued past first progression, while another pooled analysis by Schadendorf and colleagues failed to find a significant impact on clinical outcomes from continued therapy [61,62]. The 2021 National Comprehensive Cancer Network Guidelines for Melanoma suggest prospective randomized trials will be needed to find the optimal duration of therapy [15].

The transition to flat dosing from weight-based dosing was predicated on pharmacokinetic data that examined safety and efficacy of pembrolizumab and nivolumab [63,64] rather than on clinical outcomes data. In contrast, a single-center, retrospective review of 297 patients who had received immunotherapy for diseases such as non-small-cell lung cancer (53%), melanoma (19%) or another malignancy reported results that appear to support weight-based dosing [65]. In the overall cohort, overweight BMI was associated with improved progression-free survival (hazard ratio [HR]: 0.69; 95% CI: 0.51–0.94; p = 0.02) and trended toward improved overall survival (HR: 0.77; 95% CI: 0.57–1.04; p = 0.08). In the overweight population, improved outcomes were limited to patients who received weight-based ICI dosing and not those who received fixed dosing: progression-free survival HR was 0.53 (vs 0.79) and overall survival HR was 0.56 (vs 0.89) for weight-based versus fixed dosing. Patients with BMI <25 tended to have better outcomes with fixed-dose compared with weight-based ICI dosing, while patients with BMI ≥25 tended to have better outcomes with weight-based ICI dosing, although these differences did not achieve statistical significance [65]. Due to limitations in the design of these dosing studies, additional prospective and controlled studies are necessary to fully elucidate the optimal dosing strategy for ICIs in obese and non-obese patients.

Biomarkers for response

Melanoma patients do not always respond to ICIs due to various factors, including differences in expression of PD-1 and PD-L1 between tumors and patients [12] and intrinsic or acquired resistance. Given that ICI therapies are costly and associated with high risk of adverse effects [5,59], there is a critical need to identify biomarkers to better predict responses [66]. In 2016, the FDA approved the first immunohistochemical test (22C3 pharmDx) to evaluate the expression of PD-L1 on tumor cells in melanoma patients as a strategy to identify patients who are likely to respond to anti-PD-1 therapy [5]. KEYNOTE-001 showed that patients have a better chance of responding to pembrolizumab when PD-L1 expression is present in >10% of melanoma cells [5]. Although some studies show that patients with higher expressions of these ligands seem to be more likely to respond to ICI therapy and achieve better clinical outcomes [38,67,68], others show low predictive correlation [69]. The lack of PD-L1 expression also does not exclude the possibility of durable response to treatment with ICIs [68]. Espinoza et al. reported that only 50% of melanoma patients who expressed PD-L1 responded to treatment with anti-PD-1, while approximately 15% who did not express PD-L1 responded to treatment [70]. Disparate results such as these appear to be due to the lack of standardization of PD-L1 expression criteria between samples in different clinical trials, as well as the usage of different immunohistochemical reagents and analytical techniques [35,66,68]. Indeed, McLaughlin et al. showed that although there is heterogeneity in PD-L1 expression between tumors, the different PD-L1 detection antibodies used in each immunohistochemical assay had different binding affinity, specificity, crossreactivity and target epitopes. They also pointed out that companies have been producing their own diagnostic tests and submitting them to the FDA without external peer review and validation, making comparisons between products harder to interpret [69].

In addition to biomarker expression, tumor size and tumor immunological milieu also appear to impact response to ICI. Patients with smaller tumors are more likely to respond to ICIs [42], which are also more effective in tumors with higher infiltration of CD8⁺ T cells [71] or higher mutation rates, suggesting that immunogenicity prior to treatment is a prognostic biomarker [72]. Tumeh et al. demonstrated that high baseline levels of CD8⁺ T cells, PD-1 and PD-L1 at the invasive margin and center of tumors, and increased CD8⁺ T cells and PD-L1 antibodies in tumors after a single dose of pembrolizumab, were directly correlated with better therapy outcomes. No correlation was found between CD4⁺ T-cell levels and previous treatment with ipilimumab with current treatment outcomes [71]. Snyder et al. analyzed tumor DNA from patients treated with CTLA-4 blockading antibodies and found a correlation between an increased mutational load and sustained clinical benefit. However, these mutations were not characterized in the same single gene across patients; instead, multiple mutated genes expressed somatic neoepitopes, which were found to correlate with prolonged benefit from anti-CTLA-4 therapy [72]. A study in 1662 patients with advanced cancers who received at least one dose of an ICI showed that the higher the number of somatic mutations in the tumor, the longer the overall survival when treated with this class of therapy [67]. In the case of ipilimumab, no difference in clinical response was seen in patients with regard to sex, age, tumor history, primary tumor site or the number of therapies administered prior to ICI treatment. However, the load of tumor neoantigens, along with tumor expression of immune checkpoints, showed a significant correlation with therapeutic response [73]. Indeed, patients with higher mutation rates and higher infiltration of T cells in tumors were more likely to benefit from PD-1 and CTLA-4 inhibitors [12]. Nonetheless, some tumors expressing elevated mutations did not respond to therapy, and there is not sufficient evidence to support widespread use of tumor mutation burden as a prognostic biomarker.

Immune-related adverse events

The irAEs are distinct from the toxicities that are characteristic of cytotoxic chemotherapy in terms of their onset, mechanism, type, incidence and severity [74]. While traditional chemotherapy agents are linked to myelosuppression, in general, with drug-class-specific organ toxicities (e.g., anthracyclines and cardiotoxicity; bleomycin and pulmonary toxicity; phosphoramide mustards and hemorrhagic cystitis), irAEs can affect any organ. The onset of irAEs is often delayed compared with traditional cytotoxic treatments: dermatological toxicity (2–3 weeks), colitis (6–7 weeks) and hepatitis (6–7 weeks) occur relatively early, while nephritis, endocrinopathies and pulmonary toxicity occur later in treatment, usually after 9 weeks [74]. While these timelines represent the onset most commonly seen in the literature, a broad range is possible. Patients have been reported to have irAEs shortly following their first dose (e.g., pneumonitis being reported as little as 9 days post-dose) or several weeks after discontinuation of therapy [74,75].

Mechanistically, the exact pathophysiology of irAEs is not fully understood, but they are believed to be largely related to the disruption of immunological homeostasis caused by altering these immune checkpoints [76]. Comparatively, this is a more complex process than the adverse events associated with DNA poisoning that are a hallmark of cytotoxic chemotherapy. Because irAEs are believed to be due to an overactive inflammatory action, as in autoimmune disease, the mainstay of irAE management is immunosuppression, usually via corticosteroids. Treatment may involve topical steroids for low-grade topical toxicities, or more aggressive parenteral corticosteroids ranging from 0.5 to 2 mg/kg/day depending on the grade and location of the irAE [77]. While there is a theoretical concern that treatment of irAEs using immunosuppressive drugs may compromise the efficacy of ICI therapy, it has not been proven. In addition, recent data suggest that patients who experience irAEs have the potential to see increased survival outcomes, provided that the irAEs are managed swiftly and carefully [44,78].

The incidence of reported irAEs ranges across ICI trials from 15 to 90% for any grade, and from 0.5 to 13% for grade 3–4 irAEs [79]. The incidence and severity of irAEs depends on the class of ICI received, whether ICI therapy is used in combination versus monotherapy, and on patient-specific factors [77]. Patient-specific factors include pre-existing autoimmune disorders, though recent data suggest that safe treatment of these patients with ICIs is possible [80]. Age may also play a role, as younger patients have been linked to an increased incidence of severe irAEs, especially higher rates of colitis and hepatitis. In contrast, pneumonitis and myocarditis are more frequent in older patients, who are also at risk for increased death rate and length of hospitalization [81]. Cancer type appears to affect rates of irAEs as well. Grade 3–4 pneumonitis has been reported in 5–7% of patients receiving ICI therapy for non-small-cell lung cancer, but in the melanoma population, pneumonitis has been reported in <2% of patients and is typically grade 1–2 [15,56,79]. The CheckMate 067 trial demonstrated the impact of ICI class and regimen on rates of irAEs. The nivolumab monotherapy arm saw 99.4% of patients reporting any-grade treatment-related adverse events and 43.5% of patients with grade 3–4 toxicities, while patients receiving ipilimumab monotherapy reported 99 and 55.6%, respectively. However, the combination nivolumab and ipilimumab arm saw treatment-related adverse events in 99.7% of patients and grade 3–4 toxicities in 68.7% of patients, showing a clear increase in high-grade toxicity from combination therapy; there was also a smaller, but still evident, increase in the CTLA-4 monotherapy arm versus the PD-1 monotherapy arm [56].

Sequencing ICIs & BRAF-targeted therapies

One lingering question as to the sequencing of therapy in melanoma may be answered shortly. Due to their high response rates and rapid onset of action, BRAF/MEK combination therapy has historically been used as frontline therapy for patients who harbor a targetable BRAF mutation, while ICI are thought to produce a slower, but more durable, response in these same patients [82]. New information presented in November 2021 at an American Society of Clinical Oncology plenary session showed that giving combination ipilimumab and nivolumab prior to dabrafenib and trametinib led to a 20% increase in overall survival compared with the reverse sequence of therapy. Due to this finding, the trial was halted with just 59% of the proposed accrual completed. In terms of toxicity, grade 3 or higher adverse events occurred in 60% of patients who started ICI therapy first and in 52% of patients who received targeted therapy first [83]. The full study results, once published, have the potential to change the current recommendations on the sequence of ICI and molecular-targeted therapies in BRAF mutation-positive melanoma patients.

Cost–effectiveness

One question that remains unanswered is how to approach the issue of cost with these medications. Oh et al. reported an incremental cost–effectiveness ratio of US$454,092 per progression-free quality-adjusted life-year for combination ipilimumab and nivolumab as compared with nivolumab monotherapy [84]. They also found that ipilimumab would require, at minimum, a US$21,555 per dose cost reduction to reach an incremental cost–effectiveness ratio of <US$100,000, the willingness-to-pay threshold used in the study. Ipilimumab would also need a 42% cost reduction before the combination therapy became more cost-effective than nivolumab monotherapy. These data were based on the average wholesale prices for nivolumab and ipilimumab for a 70-kg patient in 2015, which were US$5732 and US$33,162, respectively. Current data using 240 mg flat dosing for nivolumab and 10 mg/kg dosing, assuming 70 kg patient as well, for ipilimumab shows average wholesale prices of US$8297.92 and US$132,429.36, respectively, for a single dose.

New developments

Extensive characterization of the TILs of metastatic melanoma patients has revealed the presence of additional T-cell checkpoint proteins, including costimulatory molecules that have shown promise as therapeutic targets and are under clinical investigation. One such molecule is 4-1BB (CD137), a costimulatory receptor expressed on CD8⁺ tumor-infiltrating T cells which increases T-cell survival [85], induces CD8⁺ T-cell proliferation and enhances T-cell function by amplification of the cytotoxic response [86]. In melanoma, it was found that 4-1BB agonism in combination with ICIs like anti-CTLA-4 or anti-PD-1 inhibitors produces synergistic levels of tumor rejection [87,88]. However, hyper-costimulation of 4-1BB disrupts homeostasis of the immune cells and causes lymphatic anomalies such as lymphadenopathy and splenomegaly [89]. Nevertheless, clinical trials are underway to determine the safety and efficacy of 4-1BB with ICI combinatorial therapy [89,90]. Whether these inhibitory or stimulatory co-receptors promote antitumoral or protumoral immune responses is dependent on the type of interaction and the type of immune cells. For example, although costimulatory, ICOS–ICOSL interaction between melanoma cells and Tregs drives the expansion of Tregs and immune evasion by the tumor cells [91]. ICOS⁺ Tregs isolated from patients with stage IV metastatic melanoma were found to inhibit T-cell-mediated immune responses [91,92]. As a costimulatory molecule, ICOS expression is upregulated on activated T cells; however, sustained ICOS–ICOSL interaction can also lead to suppressive immune responses [93]. In advanced melanoma patients treated with anti-CTLA4, researchers found that elevation of ICOS⁺ CD4⁺ T cells correlated with increased patient survival [94]. Due to this dual nature of ICOS–ICOSL interactions, clinical trials have been designed for both agonism and antagonism of the ICOS receptor [93].

Other molecules, such as OX40, were found to have critical roles in the antitumor immune response against BRAF-mutant melanoma [95]. Although OX40 is not an immune checkpoint molecule, OX40–OX40L agonism has the potential to diminish the effect of the checkpoint inhibition and reduce the resistance to ICI immunotherapies [95]. OX40 is a member of the TNF receptor family, found to be expressed on activated CD4⁺ or CD8⁺ cells in primary melanoma tumors and in their draining lymph nodes [96]. Preclinical studies showed that antitumor immune responses are enhanced through OX40 binding to its ligand OX40L, expressed on APCs [97]. Investigators have leveraged this by designing an anti-OX40 agonistic antibody that binds to OX40, and several clinical trials are currently ongoing to evaluate the efficacy of anti-OX40 in melanoma, both as a monotherapy and in combination with other immunotherapies [98,99].

The gut microbiome has also been related with the efficacy and tolerability of ICIs [6,11,12,35,100]. Bacteroides species have been found to modulate the Th1 cell response through IL-12 production, positively interfering with CTLA-4 blockade response. The severity of intestinal irAEs such as colitis seems to be dependent on the gut microbiota [6]. Faecalibacterium has been connected to a better antitumor response due to a higher interaction and infiltration of T cells [100], while Bifidobacterium has been suggested to act either by directly interacting with the host immune cells or by releasing soluble factors that could enhance systemic DC function [12]. Consequently, antibiotic therapy can potentially play a negative role in treatments with ICIs [11], and the modulation of the host microbiome through oral supplementation could possibly reduce ICI resistance. Several ongoing clinical trials (e.g., NCT05102773, NCT05037825, NCT04734704, NCT04107168, NCT03819296, NCT03817125) are currently investigating the relationship between the microbiome and ICIs.

Conclusion

Since the approval of the first ICI over a decade ago, this class of therapy has revolutionized the treatment of advanced melanoma by significantly increasing the 2-year survival rate from 15 to 35%, with the promise of a durable complete disease remission in some patients [101]. These triumphs also come with new challenges as clinicians and scientists work to understand how to better distinguish between responders and nonresponders; enhance efficacy by optimizing ICI dosing, sequencing and combinatorial approaches; and develop strategies to overcome ICI resistance.

Future perspective

The intense investment of human and capital resources in these endeavors to better understand cancer immunology is expected to significantly advance this field over the next decade and continue to propel the rapid growth of cancer immunotherapeutics. Over the next few years, results are expected from clinical trials of immunotherapies targeting 4-1BB, ICOS and anti-OX40 that will open up new avenues of treatments and combinations for melanoma patients.

Executive summary

• Melanoma is responsible for the majority of skin cancer-related deaths.

• Immune checkpoint inhibitors (ICIs) achieve long-term disease remission and are now the preferred first-line therapy for metastatic and unresectable melanoma.

• Four ICIs are approved by the US FDA for melanoma: ipilimumab (CTLA-4 inhibitor), nivolumab (PD-1 inhibitor), pembrolizumab (PD-1 inhibitor) and relatlimab (LAG-3 inhibitor).

• The PD-1 inhibitors are the preferred first- and second-line regimens and can be used after a patient progresses on ipilimumab therapy.

• Current clinical challenges with ICI therapy include identification of biomarkers to predict efficacy and immune-related adverse events; optimization of dosing strategies; and management of financial toxicity associated with these new therapies.

• New developments in understanding other immune checkpoints, costimulatory receptors and the interplay between the gut microbiome and immune functionality are expected to lead to the development of new immunotherapies for the treatment of melanoma.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.