Current and Future Use of mRNA and Peptide Vaccines for Melanoma

Melanoma is the most aggressive form of skin cancer and is a leading cause of death from cutaneous malignancies [1]. The advent of immune checkpoint inhibitors (ICIs) has dramatically transformed outcomes, with up to approximately half of all patients diagnosed with advanced melanoma now achieving long-term survival beyond 10 years [2]. In Stage III melanoma, the use of neoadjuvant (before surgery) and adjuvant (after surgery) ICIs has demonstrated marked benefits for relapse-free survival (RFS) [3,4,5,6]. The remarkable success of ICIs is underpinned by the inherent immunogenicity of melanoma, driven by high tumour mutation burden (TMB), abundant tumour infiltrating lymphocytes (TILs), and high expression of tumour-associated antigens (TAAs). These characteristics continue to propel the development of the next generation of immune-directed therapies, including cancer vaccines, adoptive cellular therapies such as engineered melanoma-specific T-cell receptors (TCRs) and TILs, and therapies targeting specific cytokine receptors such as CCR2, CCR4, IL-6, and IL-15, among others.

Vaccination has long been a cornerstone of infectious disease prevention and typically involves the inoculation of a molecular preparation to elicit durable protective immunity [7]. A similar concept underlies cancer vaccines, whereby tumour-derived antigens are used as inoculants to stimulate an adaptive T-cell response capable of recognising and eliminating tumour cells. Firstly, tumour antigens are taken up and processed by professional antigen-presenting cells (APCs), primarily dendritic cells, and presented on major histocompatibility complex (MHC) II molecules, which activate naïve CD4+ T-cells by binding with TCRs and co-stimulatory receptor activation between CD28 on T-cells and CD80/86 on APCs. This activation induces cytokines, including IL-2, IFN-ꝩ and lymphotoxin, which prime and sustain antigen-specific cytotoxic T-cells and macrophages, as well as promote antibody production via B-cells and memory T-cell formation. Antigen cross-presentation on MHC I molecules by APCs activates effector CD8+ T-cells to recognise tumour cells displaying tumour antigens and to exert cytotoxic effects through perforin and granzyme-mediated apoptosis [8,9,10]. While non-professional APCs can also present antigens via MHC I and engage CD8+ T-cells, the absence of concomitant CD4+ T-cell activation via MHC class II limits broader adaptive immunity, and consequently, the overall immune response tends to be less sustained and is prone to tolerance [9,11,12].

The identification of relevant tumour antigens presented by APCs is central to cancer vaccine development. TAAs are self-antigens that are aberrantly expressed or overexpressed in tumour cells but present on normal cells at lower levels. In contrast, tumour-specific antigens (TSAs, also known as neoantigens) are self-antigens that arise from somatic genetic mutations during tumorigenesis and are exclusively expressed by tumour cells [13,14].

Despite decade-long efforts, early cancer vaccines yielded limited clinical efficacy due to weak immunogenicity and failure to overcome immune tolerance [7,15,16]. However, recent advances, including the advent of ICIs, synthetic long peptide (SLP) vaccine platforms and messenger RNA (mRNA) vaccines, have rekindled significant interest. Early clinical data have shown that these vaccines, particularly when combined with ICIs, can elicit potent antitumour responses as adjuncts in combination therapies.

Historically, peptide vaccines encoding short segments (8–11 amino acids) of tumour antigens elicited limited tumour-specific immune responses. Their efficacy was constrained by MHC I restriction, incomplete T-cell activation, and susceptibility to rapid enzymatic degradation, leading to poor cytotoxic T-cell responses and the emergence of immune tolerance. In contrast, SLPs, typically 11–30 amino acids in length, encoding multiple epitopes, are presented on both MHC I and MHC II molecules by professional and non-professional APCs, enabling the activation of both CD4+ and CD8+ immune responses that promote a more robust and durable antitumour response [11,12,14,15,17].

mRNA vaccine technology, accelerated by the COVID-19 pandemic, has further revolutionised the therapeutic vaccine landscape. Contemporary mRNA vaccine technology enables rapid, scalable and cost-effective production of mRNA sequences that encode antigens which can be delivered safely to patients to induce strong antigen-specific immune responses [18,19,20]. Both SLP and mRNA vaccine platforms allow for the inclusion of multiple TAAs or TSAs within a single vaccine formulation to overcome human leukocyte antigen (HLA) restriction and antigen heterogeneity across sites while enhancing antigen presentation through both MHC I and II pathways [12].

Beyond direct T-cell activation, emerging evidence suggests that vaccines may also reshape the tumour microenvironment (TME) to improve immune recognition of tumour cells and response to immune-directed therapies. The immunosuppressive TME is mediated by the recruitment of suppressive cells, including FoxP3+ regulatory T-cells (Treg), which suppress effector T-cell responses by secretion of immunosuppressive molecules, including TGF-β, IL-8 and IL-10 or direct cell contact [21,22]. Tumour cell surface expression of immunosuppressive ligands, such as programmed death-ligand 1 (PD-L1), also actively inhibits cytotoxic T-cell responses. The expression of cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) on the T-cell surface is another inhibitory mechanism by which the T-cell response is suppressed [22]. Blockade of PD-1/PD-L1 or CTLA-4 using ICIs has demonstrated remarkable antitumour activities in melanoma, and the combination of anti-PD-1 and anti-CTLA-4 ICIs showed improved efficacy compared to single-agent ICI [2]. However, a certain proportion of melanoma patients exhibit primary resistance to ICIs, and a proportion of patients who initially responded to ICIs subsequently develop adaptive resistance. The mechanisms behind primary and adaptive resistance are thought to be associated with the high Treg-to-effector T-cell ratio and the expression of immunosuppressive cytokines in the TME, downregulation of tumour antigen presentation, effector T-cell exhaustion, and development of escape mutation variants in tumour cells [22]. Preclinical investigations suggest vaccines, especially mRNA vaccines, were able to increase the release of proinflammatory cytokines, including TNF-α and IFN-ꝩ, induce CD4+ and CD8+ T-cell infiltration, and promote tumour antigen expression, thereby reshaping the TME from being immunosuppressive to immune sensitive and potentially overcoming primary and adaptive resistance to immune-directed therapies. In addition, combination vaccine and ICI therapies appear to induce a stronger anti-tumoural response than either therapy alone in preclinical models, suggesting a synergistic effect and supporting further investigation of combination therapy approaches [23,24,25,26]. Collectively, SLP and mRNA vaccines represent a new generation of immunotherapies that can potentially propel cancer vaccines into mainstream clinical application.

This review aims to provide an overview of the mechanistic and technological foundations underpinning mRNA and SLP vaccine development in melanoma, as well as the immunological and clinical evidence supporting adjuvant and neoadjuvant immune-directed therapies. It also explores opportunities, benefits and challenges of incorporating mRNA and peptide vaccines into standard treatment paradigms for melanoma.

Many cancer cells express peptides that are released in the bloodstream or presented on the cell surface by MHC molecules. Peptides recognised by the immune system that are capable of eliciting an immune response are considered antigenic, including TAAs and TSAs. TSAs are unique self-antigens that arise as a consequence of either viral oncoproteins or somatic genomic mutations generating neoantigens. These include single-nucleotide variants (SNVs), single-nucleotide base insertions or deletions (INDELs), gene fusions, aberrant RNA splicing, disordered post-translational modification, and integrated viral open reading frames (ORF) [27]. Viral oncoprotein-associated TSAs, such as those found in human papillomavirus-associated cervical cancer and Epstein–Barr virus-associated nasopharyngeal cancer, are created by virally encoded ORFs [28]. Among somatic mutations, SNVs account for most of the nonsynonymous neoantigen peptides. These peptides, generated during proteasomal degradation of aberrant proteins and presented by MHC molecules on the surface of tumour cells, can elicit an antigen-specific immune response [29].

A minority of neoantigens, arising from mutations in oncogenes like PIK3CA, RAS and BRAF, may be shared across patients in multiple cancer types [30]. INDEL mutations generate nonsynonymous ORFs, producing neoantigens that are often highly immunogenic and display greater binding affinity to MHC I [31]. INDEL frequency also reflects the mutagenic aetiology of the tumour; for example, higher numbers of INDELS are observed in cutaneous melanoma compared with mucosal melanoma, reflecting ultraviolet (UV)-induced DNA damage [32]. Gene fusions represent another mechanism for neoantigen generation through chromosomal translocations or inversions and can elicit a potent cytotoxic T-cell response. In patients with melanoma treated with anti-PD-1, responders showed a reduction in the number of fusion neoantigens (2.22 vs. 0.67; p = 0.019), whereas no numerical difference was observed in tumours that were stable (2.29 vs. 1.71; p = 0.32) or progressing (2.53 vs. 2.43; p = 0.94) on treatment [30,33]. This data suggests that immune surveillance may selectively eliminate tumours carrying fusion-derived neoantigens.

Cutaneous melanoma exhibits one of the highest rates of TMB, often exceeding 100 mutations per megabase, largely due to UV exposure [34]. UV radiation leads to impairment in DNA damage repair mechanisms, promoting genomic instability and mutagenesis. Melanoma cells express high levels of TAAs, such as the melanoma antigen gene (MAGE) family antigens and cancer testis antigens such as NY-ESO-1 [35]. Additionally, the high TMB confers a high number of tumour-specific neoantigens that are readily recognised by the immune system. Multiple studies have demonstrated high neoantigen load in melanoma (9 to 16,300 neoantigens per tumour), and this has been shown to correlate with ICI responsiveness [36,37]. Snyder et al. [37] demonstrated that shared tetrapeptide neoepitopes were enriched amongst patients deriving durable benefit from CTLA-4 blockade, supporting the persistence of antigen-driven memory after expanded TILs.

For cancer vaccines to be efficacious, the selection of target antigens is crucial. The ideal antigens should be abundantly expressed on tumour cells, exhibit no or minimal expression in normal tissues, and be sufficiently distinct from host antigens to overcome immunological tolerance and mitigate the risk of autoimmunity. Historically, melanoma vaccines have focused on highly expressed TAAs and embryonic antigens such as ganglioside GM2, glycoprotein-100 (gp-100), preferentially expressed antigen in melanoma (PRAME), and MAGE family antigens, as well as restricted gangliosides including GM2, GD2, GT3 and 9-O-Ac-GD3 [38]. Despite strong preclinical data, vaccines directed against TAAs have historically yielded limited clinical efficacy, likely due to the rapid emergence of immune tolerance and tumour antigen heterogeneity [15,30,39]. Furthermore, the application of this technology has been constrained by the limited accuracy of antigen prediction due to the vast heterogeneity of TMB and distant antigen presentation among tumour types [40].

Recent advances in bioinformatics, next-generation sequencing (NGS), RNA sequencing and proteomics have accelerated the identification of immunogenic neoantigens, paving the way for personalised cancer vaccines [15]. Importantly, research is now expanding beyond MHC class I-restricted neoantigens that induce CD8+ T-cell responses to include MHC class II-binding neoantigens that induce CD4+ T-cell-mediated regulation of antitumour immunity. Moreover, nonspecific mRNA vaccines designed to boost T-cell activation in conjunction with ICI therapy are also showing promise. Recent preclinical data indicate mRNA vaccines can upregulate IFN-ꝩ and augment cytokine-mediated T-cell recruitment and increase PD-L1 expression on tumours, rendering immunologically “cold” tumours more sensitive to concurrent ICIs, leading to improved survival outcomes [24,41].

Beyond antigen selection, the vaccine delivery platform also represents an important consideration. An ideal vaccine should be straightforward and cost-effective to manufacture, possess stability with an extended shelf life, be applicable across diverse clinical settings, and elicit strong immune activation with minimal toxicity. Broadly, vaccine delivery strategies can be categorised as cellular-based, peptide and mRNA technologies. Cellular-based vaccines involve the ex vivo loading of APCs, most commonly dendritic cells, with tumour antigens, followed by reintroduction of these antigen-loaded APCs to the host to stimulate a tumour-specific T-cell response. This process involves the harvesting of APCs via leukapheresis, ex vivo maturation, activation, antigen loading and subset selection, all of which are technically demanding, costly and limited by a short shelf life, restricting use to selected specialist centres and clinical settings [17,42,43].

mRNA transfers genetic information from DNA to ribosomes, where it is synthesised into proteins. The applicability of this technology is expanding, as it can be used for antigen expression, gene editing, and protein therapies against infectious diseases or for cancer treatment [20,44]. The development of mRNA vaccines for tumour prevention dates back to the 1970s, when Langer et al. demonstrated that nucleic acids could be encapsulated and delivered by biocompatible polymers [45]. Advances in chemical modification and purification of mRNA, as well as improvements in drug delivery, such as lipid, polymeric and hybrid nanoparticles, have improved the in vivo viability and delivery of mRNA [46,47,48,49]. The COVID-19 pandemic accelerated this field, demonstrating production scalability and improving immunogenicity, resulting in the success of mRNA vaccines such as the Moderna mRNA-1273 and the Pfizer/BioNTech BNT162b2, with others in early-phase clinical development (Moderna mRNA-4359) [50,51,52]. mRNA vaccines offer several advantages, such as encoding multiple antigens to be presented by APCs across diverse HLA types, reducing the potential for HLA restriction, and broadening T-cell activation [18,53,54]. Unlike DNA-based therapies, mRNAs do not integrate into the host genome, eliminating the risk of insertional mutagenesis [19]. The transient expression and rapid degradation by normal cellular processes reduce the risk of potential autoimmunity [17].

The manufacture of mRNA vaccines begins with the identification and selection of target antigens (Figure 1). In the case of a neoantigen vaccine, nonsynonymous mutations that are tumour-specific are first identified through comparative whole exome and RNA sequencing of tumour tissues and normal tissues and prioritised based on predicted MHC class I and II binding affinity [53]. The selected RNA sequences are incorporated into a synthetic mRNA construct comprising a 5′ cap to add stability, and together with 5′ and 3′ untranslated regions, an open reading frame, and a poly(A) tail to enhance stability and translation efficiency [18,55].

Key challenges with mRNA vaccines include the inherent instability of mRNA, large molecular size, negative charge and susceptibility to enzymatic degradation [47,56]. To improve mRNA vaccine stability and delivery, the mRNA molecules are encapsulated in delivery vehicles, most commonly lipid nanoparticles (LNPs), to prevent enzymatic degradation and improve organ- or tissue-specific delivery [55,57]. LNPs also exhibit additional immunostimulant activity, triggering the release of cytokines such as IL-6, IL-12 and IL-1β and upregulating co-stimulatory toll-like receptor agonists (TLRs), including TLR4, to augment the immunogenicity of the vaccine [58]. Once administered, the mRNA molecules are delivered to the cytoplasm of APCs, where tumour antigens are translated, processed and presented on MHC class I and II molecules.

Peptide vaccine synthesis has also benefitted from significant technological advances, making long-chain peptide synthesis and purification techniques relatively straightforward and cost-effective [59]. SLPs are assembled from amino acids by stepwise deprotection and coupling on a solid support such as resin, with downstream purification using high-performance liquid chromatography and mass spectrometry [59]. These advances enable scalable production of multi-epitope SLP vaccine constructs that can activate diverse T-cell clones to induce a more robust immune response and mitigate immune tolerance. Once administered, the long peptides are internalised, processed and presented by APCs on MHC I and II molecules to elicit both CD4+ and CD8+ immune responses [12]. Peptide vaccines are flexible and precise in antigen design and have high tumour specificity, low risk of autoimmunity, and a good safety profile [60].

To further enhance immunogenicity, adjuvants are incorporated into mRNA or peptide vaccines to activate innate and adaptive immunity. Adjuvants are generally classified into three categories: (1) RNAs with intrinsic adjuvant properties, (2) components of the delivery system, and (3) exogenous immunostimulants [61]. Unmodified exogenous mRNA, such as that contained within mRNA vaccines, has intrinsic adjuvant activity and can activate pattern recognition receptors, including TLR3, TLR7 and TLR8, leading to cytokine release, including TNF-α, IL-6, IL-12, IL-1β and IFNα/β [61]. Exogenous immunostimulants, including arginine-rich protamine peptides, may be added to vaccines to further activate TLR7/8 pathways, promoting B- and T-cell responses and suppressing circulating FoxP3+/CD4+ Treg cells, as demonstrated in an early-phase trial [62,63]. Ongoing research is focused on optimising adjuvant formulations that further improve immunogenicity without causing excessive reactogenicity [64].

Clinical application of mRNA and SLP vaccines in melanoma is increasingly garnering interest, with multiple trials evaluating efficacy, safety and optimal therapeutic settings (Table 1).

The presence of tumour antigens is a prerequisite for effective immune cell activation. While surgical resection of tumours offers clear therapeutic benefit, it simultaneously removes the majority of the tumour antigens, potentially diminishing immune activation, even in the presence of immunotherapy. Administering immunotherapy in the neoadjuvant setting, when the tumour bulk and corresponding draining lymph node basins are in situ, has been shown to induce a stronger and more diverse T-cell response, leading to significantly improved clinical outcomes [85,86].

Several landmark randomised trials, most notably SWOG1801 [6], a Phase II trial of perioperative pembrolizumab versus adjuvant pembrolizumab, and the Phase III NADINA trial of neoadjuvant ipilimumab and nivolumab versus adjuvant nivolumab, have firmly established neoadjuvant ICIs as a new standard of care in melanoma [87,88,89]. In SWOG1801, the 3-year event-free survival (EFS) was 68% in the neoadjuvant pembrolizumab arm, compared to 56% in the adjuvant pembrolizumab cohort (HR 0.67; 95% CI 0.42–0.94) [90]. Similarly, NADINA showed a 2-year EFS of 77.3% in the neoadjuvant ipilimumab and nivolumab arm and 55.7% in the adjuvant nivolumab arm (HR 0.40; 95% CI 0.28–0.57; p < 0.001). For patients who achieved a pathological complete response following neoadjuvant therapy, the 2-year EFS was 95.1% [91].

One of the important outcomes of the NADINA study, building on the findings from the PRADO study [88], was the validation of a response-directed adjuvant therapy strategy. In both studies, patients who achieved a major pathological response to neoadjuvant ICIs, defined as <10% viable tumour at surgical resection, were able to forgo additional adjuvant ICIs following surgery without compromising EFS. This approach supports the personalisation of adjuvant treatment based on tumour response assessment, allowing some patients to avoid unnecessary exposure to adjuvant ICI while maintaining excellent efficacy.

Given the mechanism of cancer vaccines, one could hypothesise that introducing cancer vaccines into the neoadjuvant setting could further enhance immune response and clinical benefit [92]. For instance, an adenovirus-based vaccine encoding tyrosinase-related protein 2, which is widely expressed in melanoma, was tested in preclinical models and found to confer protection against tumour recurrence when given before surgery, while no benefit was observed when given after surgery [93]. However, one of the challenges with neoadjuvant vaccine delivery is the lengthy manufacturing process, which may not be practical when there is urgency to treat the malignancy in the neoadjuvant setting. In KEYNOTE-942, the median time from surgery to the first dose of pembrolizumab was 10 weeks, and the majority (81%) of the patients received the first dose of vaccine with cycle 3 pembrolizumab, approximately 6 weeks after the initial dose of pembrolizumab [67].

The KEYNOTE-942 study recruitment spanned the years of the COVID-19 pandemic, and prioritisation for COVID-19 vaccine manufacturing during the pandemic may have partly impacted the manufacturing capacity of the cancer vaccine. However, dedicated production lines and marked improvements in mRNA and protein production technologies are shortening the vaccine production timelines, potentially overcoming this barrier. One question that remains is the optimal number of vaccine doses required in the neoadjuvant setting for an effective response. A Phase II/III study (NCT06295809↗) is also currently underway to investigate the efficacy and safety of neoadjuvant-adjuvant pembrolizumab and the personalised mRNA V940 vaccine in cutaneous squamous cell carcinoma. In the pembrolizumab/V940 arm, patients receive two doses of pembrolizumab delivered every 6 weeks and two doses of the V940 vaccine every 3 weeks before surgery [94]. A Phase II study (NCT05280314↗) investigating an off-the-shelf peptide vaccine (IO102-IO103) in combination with pembrolizumab as neoadjuvant-adjuvant therapy for resectable melanoma is also currently underway. Beyond assessing efficacy, these studies will provide important proof of logistic feasibility for delivering cancer vaccines in the neoadjuvant context.

Overall, mRNA- or SLP-based cancer vaccines represent promising new frontiers in melanoma therapy, particularly in combination with ICIs. Advances in manufacturing technologies and ongoing clinical research are expected to address current logistical challenges and support their integration into the mainstream melanoma treatment paradigm.

Genomics-led personalised cancer vaccines represent a rapidly evolving frontier in melanoma treatment. mRNA and SLP vaccine technologies offer clinically viable approaches to modulate the tumour immune microenvironment, improve the effectiveness of ICIs, and generate robust antitumour responses to both tumour-associated antigens and neoantigens, with minimal additional systemic toxicity. The success of neoadjuvant ICI therapies in melanoma underscores the potential for vaccines to further improve outcomes in this setting. The main barrier to neoadjuvant implementation is the timely manufacture of a vaccine to be suitable for the neoadjuvant window, an obstacle likely to be overcome through continued optimisation of mRNA and peptide vaccine platforms.

Original draft preparation: J.J.Q. and Y.W.; review and editing: J.J.Q., Y.W. and S.S.; expert content direction and supervision: S.S.; project administration: J.J.Q. All authors have read and agreed to the published version of the manuscript.

No new data were created or analysed in this study.

J.J.Q. and Z.W. have no conflicts of interest. S.S. has advisory roles (all fees to the institution) for AbbVie, AstraZeneca, Bristol Myers Squibb, Merck Sharp & Dohme, Novartis, Roche/Genentech, Janssen, AdvanCell, IO Biotech and Skyline Diagnostics; received research funding (all institutional) from Amgen, AstraZeneca, Merck Sharp & Dohme, Endocyte/Advanced Accelerator Applications, Roche/Genentech, Novartis, Pfizer, and Senhwa Biosciences; and has stock ownership in AdvanCell.

This research received no external funding.