Oluwaseyi Adeuyan1*
Emily R. Gordon1Divya Kenchappa2Yadriel Bracero2
Ajay Singh2Gerardo Espinoza2
Larisa J. Geskin3Yvonne M. Saenger2*- 1Columbia University Vagelos College of Physicians and Surgeons, New York, NY, United States
- 2Albert Einstein College of Medicine, Bronx, NY, United States
- 3Department of Dermatology, Columbia University Irving Medical Center, New York, NY, United States
The approval of immunotherapy for stage II-IV melanoma has underscored the need for improved immune-based predictive and prognostic biomarkers. For resectable stage II-III patients, adjuvant immunotherapy has proven clinical benefit, yet many patients experience significant adverse events and may not require therapy. In the metastatic setting, single agent immunotherapy cures many patients but, in some cases, more intensive combination therapies against specific molecular targets are required. Therefore, the establishment of additional biomarkers to determine a patient’s disease outcome (i.e., prognostic) or response to treatment (i.e., predictive) is of utmost importance. Multiple methods ranging from gene expression profiling of bulk tissue, to spatial transcriptomics of single cells and artificial intelligence-based image analysis have been utilized to better characterize the immune microenvironment in melanoma to provide novel predictive and prognostic biomarkers. In this review, we will highlight the different techniques currently under investigation for the detection of prognostic and predictive immune biomarkers in melanoma.
Introduction
The yearly incidence of skin cancer in the United States is greater than all other types of cancer combined (O’Neill and Scoggins, 2019; Dzwierzynski, 2021). Among skin cancers, melanoma is the most aggressive. An individual’s lifetime risk of developing melanoma has gone from 1 in 500 in 1935 to 1 in 50 in 2023, partially due to increased awareness and early detection of disease (Volkovova et al., 2012; Rastrelli et al., 2014; Dzwierzynski, 2021). Thus, the need for better prevention and treatment of this disease is increasingly critical.
The development of immunotherapy has served as a pivotal turning point in the treatment of many cancers and melanoma in particular (Hu-Lieskovan et al., 2020). Specifically, the discovery of antibodies directed to immune checkpoint molecules such as programmed cell death protein 1 (PD-1) and cytotoxic T-lymphocyte-associated-protein 4 (CTLA-4), and, more recently, lymphocyte activation gene 3 (LAG-3), have drastically prolonged survival in melanoma (Uhara, 2019; Huuhtanen et al., 2023). Ipilimumab, an anti-CTLA-4 antibody, was FDA approved for the treatment of unresectable or metastatic melanoma in 2011 based on phase 3 data showing superiority to chemotherapy and treatment with a well-studied vaccine against glycoprotein 100 (gp100) and was subsequently found to also show promise in the adjuvant setting (Hodi et al., 2010; Robert et al., 2011; Sarnaik et al., 2011; Sanlorenzo et al., 2014). Anti-PD-1 antibodies, however, have become the gold standard treatment for melanoma based on studies comparing anti-PD-1 to anti-CTLA-4 and to chemotherapy (Topalian et al., 2012). A recent study found that pembrolizumab, an anti-PD-1 antibody, may be used as adjuvant therapy in stage IIB and IIC melanoma, in addition to advanced melanoma and resected stage III disease (Luke et al., 2022). There is now very strong data to support neoadjuvant therapy for stage III melanoma, although integrating with surgical management in small volume disease may not always be straightforward and this is currently an area of investigation (Patel et al., 2023). Combination anti-CTLA-4 and anti-PD-1 is more effective than either alone but at the cost of significant toxicity (Wolchok et al., 2013). Most recently, the combination of anti-PD-1 and anti-LAG-3 gained FDA approval, providing a second immunotherapy option, although prolonged overall survival (OS) with this regimen is not yet proven (Huuhtanen et al., 2023). Emerging immune checkpoint molecules [e.g., adenosine A2A receptor (A2AR), T cell immunoglobulin and mucin domain 3 (TIM3), V-domain Ig suppressor of T cell activation (VISTA)] are being explored in clinical trials with already approved ICIs, offering potentially additional treatment options (Hu-Lieskovan et al., 2020). Tumor-infiltrating lymphocyte (TIL) therapy has also been shown to have high efficacy rates as first-line therapy as well as in the post-PD-1 setting in phase II and phase III trials, but has not yet reached regulatory approval (Rohaan et al., 2022; Monberg et al., 2023).
The notable phase III DREAMseq trial established that immunotherapy should precede targeted therapy for patients with treatment-naive BRAF V600-mutant metastatic melanoma (Atkins et al., 2023). Thus, immunotherapy has become the standard of care for resected stage III melanoma, and most recently for resected stage IIB-C melanoma. Immunotherapy is FDA approved despite the fact that over 75% of patients are cured by surgery alone. Therefore, the need for biomarkers focuses on two areas: selecting patients with advanced disease for combination immunotherapy and selecting patients with early stage disease for single-agent adjuvant immunotherapy.
Although dramatic success has come from the advent of immune checkpoint inhibitors (ICI), these therapies have also been associated with multiple adverse events including elevated liver enzymes, rash, pruritus and fatigue, among many others. Grade 3 or 4 adverse events have also been commonly reported, oftentimes in over 50% of melanoma patients treated in clinical trials with ICI (Wolchok et al., 2010; Robert et al., 2011; Wolchok et al., 2013). Given these significant clinical findings, as well as the expense and inconvenience incurred by treatment, the establishment of biomarkers will be crucial in helping clinicians better weigh the potential response to immunotherapy against the potential risks and adverse effects.
Biomarkers are measurable biological indicators that can be subdivided into two main groups: prognostic and predictive (Nalejska et al., 2014; Rizk et al., 2019). While prognostic biomarkers indicate a patient’s disease outcome regardless of treatment, predictive biomarkers are used to estimate how likely a patient will respond to a given therapy. Thus, prognostic biomarkers may identify high-risk patients that may benefit from multimodal, more aggressive therapy while predictive biomarkers may indicate a patient’s specific treatment outcome. To discover these biomarkers, multiple techniques have been utilized including gene expression profiling of bulk tissue and spatial transcriptomics of single cells, among others. This review summarizes these different techniques from the current standard to newer technological areas of innovation such as multiplexing and artificial intelligence (AI)-based image analysis for the development of prognostic and predictive biomarkers in melanoma and highlights directions for future biomarker development.
Current gold standard
Traditionally, the major clinical and histological prognostic biomarkers of melanoma have included primary tumor (i.e., Breslow) thickness, ulceration, mitotic rate, anatomic site (i.e., acral, cutaneous, mucosal or uveal) and sentinel lymph node (SLN) involvement (Balch et al., 2001; Abbas et al., 2014). These factors have been largely captured in the Tumor, Node and Metastasis (TNM) system by the American Joint Committee on Cancer (AJCC). In the current (eighth) edition of the TNM system, Breslow thickness and ulceration, which comprise the T category, continue to be correlated with survival with thicker and ulcerated tumors more associated with poorer prognosis (Keung and Gershenwald, 2018; In ’t Hout et al., 2012). However, mitotic rate, which remains a strong indicator of prognosis in melanomas of varying thickness, is no longer incorporated into the AJCC melanoma staging system.
Regarding the N category, patients with “clinically occult” nodal metastasis (i.e., patients with regional node metastasis at SLN biopsy but without accompanying clinical or radiographic evidence) have been shown to have better survival than patients with clinically or radiographically evident disease. Thus, SLN biopsies are often performed in these patients as nodal status is an important independent predictor of prognosis (Cascinelli et al., 2000; van Akkooi et al., 2008; Balch et al., 2010; Gershenwald et al., 2017; Keung and Gershenwald, 2018). A positive SLN biopsy historically warranted complete lymph node dissection (CLND). However, two multicenter randomized controlled trials found that there was no significant difference in OS between immediate CLND versus nodal observation in these patients, demonstrating new prognostic implications and roles for adjuvant systemic therapies. Non-nodal regional (e.g., microsatellite, satellite or in-transit) metastases also serve as criteria for the N category and have been associated with worse prognosis (Rao et al., 2002; Van Es et al., 2008; Wilmott et al., 2012; Read et al., 2015). For the M category of the TNM system, which refers to sites of distant metastases, patients with non-visceral (e.g., subcutaneous, cutaneous, nodal) distant metastasis have a modestly better survival than those with distant metastases to other sites (Barth et al., 1995; Keung and Gershenwald, 2018).
Although prognostic biomarkers have been well-established for melanoma, predictive biomarkers are not yet routinely used for ICI in melanoma due to limited clinical utility or low sensitivity and specificity (Huang et al., 2022). However, multiple techniques have been utilized for the discovery of promising candidates which will be discussed.
Melanoma genetics and circulating tumor DNA
Recent advancements in genetic research have provided insight into the molecular landscape of melanoma, revealing potential genetic signatures associated with prognosis and therapeutic response. Notably, a genome-wide sequencing program of thousands of melanomas worldwide revealed that mutations in BRAF are most common, with reported frequencies of over 60% (Akbani 2015; Tímár and Ladányi, 2022). Specifically, a substitution of valine with glutamate at residue 600 (V600E) accounts for over 90% of mutations within that locus (Dahl and Guldberg, 2007). The prognostic significance of BRAF mutations has been widely controversial as some studies have found that BRAF-mutated melanomas may be associated with worse survival and higher risk of recurrence but other studies have shown no survival difference compared to BRAF wild-type (Meckbach et al., 2014; Adler et al., 2017; Ny et al., 2020; Naimy et al., 2023). Following BRAF, NRAS, and NF1 comprise the next most common mutations identified in melanoma. Patients with NRAS- and NF1- mutated melanomas tend to have a worse prognosis (Cirenajwis et al., 2017; Podlipnik et al., 2021; Randic et al., 2021). Mutations in BRAF, NRAS, and NF1 represent driver mutations that lead to aberrant activation of the MAPK pathway, which is important for cell proliferation and survival (Burotto et al., 2014).
The identification of mutations guide therapy for melanoma. BRAF inhibitors (BRAFi) such as dabrafenib and trametinib have been shown to improve progression-free survival (PFS) and OS in patients with these mutations (Long et al., 2017a; Long et al., 2017b; Yang et al., 2023). However, these clinical benefits of BRAFi are often short-lived given increasing drug resistance. Activating mutations in NRAS also confer resistance to BRAF-targeted therapy (Nikolaou et al., 2012; Hawryluk and Tsao, 2014). Additionally, while patients with melanomas harboring a different mutation, BRAF V600K, respond to BRAFi, they have been shown to have shorter PFS compared to BRAF V600E mutant melanomas with this therapy but exhibit superior clinical response to ICI (Akbani 2015; Pires Da Silva et al., 2019; Yang et al., 2023). Additional genetic markers that indicate favorable ICI response include high tumor mutational burden (TMB), increased expression of inflammatory mediators, BRAF wild type and BRCA2 mutants (Ho et al., 2013; Hugo et al., 2016; Goodman et al., 2017; Yang et al., 2023). Nonetheless, the key DREAMseq study found that initial treatment with combination ICI therapy is significantly better than initial treatment with targeted therapy against BRAF and MEK, even in patients with BRAF-mutated melanomas (Atkins et al., 2023).
Circulating tumor DNA (ctDNA) has also emerged as a promising prognostic and predictive blood-based biomarker to monitor disease status in advanced melanoma patients (Calapre et al., 2017). Melanoma, like many other solid tumors, releases DNA that may be isolated from peripheral blood which can then be analyzed using sensitive techniques such as next-generation sequencing (NGS) to recapitulate intratumoral heterogeneity, evaluate genomic evolution in response to treatment and reveal potential resistance mechanisms (Sacco et al., 2020). Prognostically, levels of ctDNA have been found to significantly correlate with clinically-relevant, serological markers of tumor burden such as S100 calcium-binding protein B (S100B), melanoma inhibitory activity (MIA) and lactate dehydrogenase (LDH) (Sanmamed et al., 2015; Calapre et al., 2017). In the predictive setting, multiple studies have shown that plasma ctDNA levels prior to the initiation of BRAFi therapy correlated with treatment response. Specifically, the BREAK trials revealed that high baseline ctDNA levels were reliably and significantly associated with lower PFS and overall response rate (ORR) to targeted therapy with dabrafenib (Ascierto et al., 2013; Santiag et al., 2016). A smaller number of studies have assessed the predictive value of ctDNA in patients treated with ICI. For example, a study in 2017 found that ctDNA levels at baseline and early during treatment with anti-PD-1 antibodies in metastatic melanoma patients accurately predicted tumor response, OS and PFS (Lee et al., 2017). More recently, another study showed that ctDNA levels may also inform treatment response to adjuvant ICI following curative resection (Tan et al., 2019; Tivey et al., 2022). Currently, ctDNA is being explored in clinical trials as a biomarker for melanoma recurrence and treatment response.
Genomic profiling
Immune surveillance has been shown to have potential value in prognostication for many solid cancers (Fridman et al., 2010; Bindea et al., 2011). Thus, the immunoscore, a scoring system that quantitatively classifies TIL density both at the tumor center and invasive margin, was proposed as a biomarker for cancer progression. In primary melanoma, the presence of a very high number of TILs confers a more favorable prognosis (Clemente et al., 1996; Azimi et al., 2012). However, the universal clinical application of TILs has been limited due to observer variability. Additionally, the majority of early stage melanoma patients have “non-brisk” TILs, an intermediate TIL group that does not provide much prognostic information (Busam et al., 2001; Azimi et al., 2012; Sivendran et al., 2014). Development of biomarkers beyond TILs in the clinical setting has also been challenging given the requirement of formalin-fixed and paraffin-embedded (FFPE) samples for melanoma diagnosis, which compromises RNA for transcriptomic analysis (Bogunovic et al., 2009).
To address this need, Nanostring transcriptomic technology, which analyzes the expression of multiple transcripts under varying pathological or physiological states, has been used to profile a group of 446 immune-associated candidate genes in primary melanoma. A study in 2014 found that 53 out of the 446 screened genes predicted non-progression, disease-specific survival (DSS) and prolonged recurrence-free survival (RFS) in two independent cohorts of patients with resectable stage II-III melanoma (Sivendran et al., 2014). This 53-immune gene signature panel, called the melanoma immune profile (MIP), was validated in a third independent cohort of stage II-III melanoma patients, further stratifying this patient population into low- and high-risk groups for enrollment in clinical trials and/or exposure to potentially toxic ICI (Gartrell et al., 2019). The MIP differs from other genomic signatures such as the Castle Biosciences signature, the latter of which evaluated a 31-gene expression profile (GEP) test in patients with stage I-II disease where risk is generally lower. The Castle Biosciences test is based on the mesenchymal to epithelial transition, hypothesized to play a role in melanoma genesis (Zager et al., 2018). Recently, a 2023 study has prioritized the co-extraction of quality DNA and RNA from FFPE melanoma sections for large scale multi-omic analysis for future clinical utility. The study described, for the first time, the optimal approach for the procurement and testing of nucleic acids for the screening of somatic mutations, miRNA and methylation that may identify new gene signatures in archival and limited tumor tissue (Orlow et al., 2023). Genomic tests are also being explored as companion biomarkers in clinical trial settings.
Single cell and spatial based genomics
Single-cell RNA-sequencing (scRNA-seq) analysis is a valuable method for obtaining gene expression profiles of individual cells which helps to identify different cell types and pathways involved in cancer progression and resistance (Lim et al., 2020; Maynard et al., 2020). However, the isolation of individual cells during the tissue dissociation step of scRNA-seq interferes with information regarding their native spatial organization within the tissue and relation to other neighboring cells. Spatial transcriptomics complements scRNA-seq by physically localizing gene sets upregulated by specific cell types thereby preserving spatial information (Yu et al., 2018; Longo et al., 2021). To achieve this aim, studies have found that messenger RNA (mRNA) can be captured on microarrays of spatially barcoded DNA capture probes. Complementary DNA (cDNA) can then be generated from the mRNA by reverse transcription and left affixed to the arrayed oligonucleotides on the slide, maintaining the RNA molecule’s original position in the tissue section using the unique positional molecular barcodes. Sequencing libraries and computational reconstruction usually follow to model the tissue’s spatial organization (Ståhl et al., 2016; Ahmed et al., 2022a; Piwecka et al., 2023).
A recent scRNA-seq study found that PRRT3-AS1, an important long non-coding RNA (lncRNA) that has been incorporated in prognostic models for prostate cancer, hepatocellular carcinoma and glioblastoma (GBM), may be required for tumor cell migration in melanoma, suggesting that PRRT3-AS1 is not only a potential prognostic biomarker but also a potential therapeutic target (Zhang et al., 2022; Liang et al., 2018; Fan et al., 2020; Zhang et al., 2021; Y et al., 2021). Additional lncRNA-based immune classes have been associated with survival and integrated into multi-omic panels for precision immunotherapy based on melanoma samples from The Cancer Genome Atlas (TCGA) (Yu et al., 2020). Aside from lncRNAs, studies have assessed the tumor ecosystem in primary melanoma to indicate prognosis. One study found that the composition of recurrent cellular neighborhoods (RCNs) involving stromal, tumor and immune cells significantly differs with disease stage. According to this model, a spatially confined suppressive TME develops in melanoma which is sustained by cytokine gradients upregulating MHC-II and IDO1 expression and by PD-1/PD-L1-mediated cell interactions (Nirmal et al., 2022).
Other cells in the TME analyzed using scRNA-seq include T cells and tumor-derived exosomes (TEXs). Studies have shown that the presence of CXCL13+ CD4+ T cells and CXCL13 expression broadly correlates with OS in a cohort of melanoma patients, independent of immunotherapy type (Litchfield et al., 2021; Veatch et al., 2022). A TEX-related signature, termed TEXscore, using scRNA-seq was associated with shorter OS across 12 cancer types, including melanoma (Wu et al., 2021).
Predictive biomarkers have also been proposed using sc-RNAseq and spatial profiling. One study’s spatial distribution analysis found that proximity of PD-L1+ cells to tumor cells and intratumoral CD8+ density predicts response to ICI in the metastatic setting (Gide et al., 2020). Similar analyses have provided insight into the mechanism for resistance of melanoma cells to ICI. For example, using single-cell functional proteomics, it was discovered that certain signaling networks become activated shortly after BRAF inhibition and before the emergence of drug-resistant phenotypes (Su et al., 2017). By leveraging single-cell profiles to understand tumoral heterogeneity and putative interactions between stromal-derived factors and immune mediators within melanoma, multiple studies have called for therapeutic strategies that account for specific tumor cell composition rather than bulk tumor expression (Tirosh et al., 2016).
Moreover, sc-RNAseq has been paired with single cell T cell receptor sequencing (sc-TCRseq) to elaborate additional predictive information in melanoma, which has recently been shown to be a feasible technique in both fresh and frozen tissue (Wang et al., 2023). This paired technique allows for the simultaneous analysis of T cell clones and phenotypes within single cells, which may provide information on T cell differentiation, specificity and activation to better understand underlying disease etiology and guide future treatment strategies (Pai and Satpathy, 2021). One study found that two pretreatment characteristics in the peripheral blood—activated CD4 memory T (TM) cell abundance and TCR diversity—constitute promising biomarkers of ICI-induced immune-related adverse events (irAEs) in metastatic melanoma patients. Additionally, the authors identified a notable correlation between early T cell clonal expansion and the onset of severe irAEs in patients treated with combination ICI (Lozano et al., 2022). A subsequent study found that metastatic melanoma patients who responded to anti-LAG-3 and anti-PD-1 combination therapy had higher baseline TCR clonality with CD8+LAG-3+ clones that expanded and shifted to a more cytotoxic phenotype resembling NK cells (Huuhtanen et al., 2023). These studies have exemplified the versatility of sc-RNAseq across different modalities for the management of melanoma.
Multiplexed IF
Typical approaches to immunohistochemistry (IHC) evaluation of tissue from melanoma patients have several limitations including inter-observer variability and the labeling of just a single biomarker for each tissue section. Emerging techniques, namely, multiplex IHC or immunofluorescence (mIHC/IF), have attempted to address these limitations by detecting multiple biomarkers in a single tissue section through high-throughput staining and quantitative analysis (Tan et al., 2020; Ugolini et al., 2022; Yaseen et al., 2022). This technology deters from using a cocktail of antibodies reared in separate hosts and instead relies on cycles of single antibody stains added in sequential order, which are subsequently removed in order for the next antibody to be added without cross-reaction (Nguyen et al., 2021). Studies utilizing mIHC/IF have focused largely on identification of specific cell populations in the melanoma tumor microenvironment (TME) to evaluate prognosis and assess response to melanoma immunotherapies.
TILs, a major component of the TME, have been implicated in the prevention or progression of tumor growth and invasion leading to significant interest in TILs as a potential prognostic biomarker (Oble et al., 2009; Gartrell et al., 2018; Rizk et al., 2019; Gartrell-Corrado et al., 2020). Conventional IHC methods have found that as melanocytic lesions transform from benign nevi to malignant melanomas, the absolute number of TILs rises (Hussein et al., 2006; Rizk et al., 2019). The use of mIHC has elaborated on these findings to show that the presence of TILs, particularly in the stroma, is a favorable prognostic indicator (Gartrell et al., 2018). Aside from TILs, melanoma-associated tertiary lymphoid structures (TLS) are associated with improved OS and lower risk of tumor recurrence following metastasectomy (Lynch et al., 2021; Mauldin et al., 2021).
Regarding predicting response to ICI, a multiplex chromogenic and IF study of melanoma samples showed that proximity between PD-1 and PD-L1+ cells was associated with response to anti-PD-1 therapy. Similarly, high co-localization of PD-L1 and CD8 expression was associated with increased response to targeted immunotherapy (Tumeh et al., 2014). In another study, depleting mast cells in the TME was found to improve responsiveness to anti-PD-1 therapy (Somasundaram et al., 2021).
Further, the AstroPath platform, a multistep framework for multispectral mIF, produces high quality datasets at the single cell level for biomarker development and quantitative pathology to inform precision ICI. Leveraging concepts drawn from the field of astronomy, this study was able to classify PD-1 and PD-L1 expression intensity on different cell types in the TME in situ on pretreatment melanoma specimens from advanced melanoma patients on ICI. In this study, higher density of early effector T cells (CD8+FoxP3+) correlated with response to anti-PD-1 therapy whereas the CD163+PD-L1- myeloid phenotype was associated with lack of response to PD-1 blockade (Berry et al., 2021).
Multiple studies have also integrated mIHC/IF with additional technologies to identify predictive biomarkers. Digital spatial profiling (DSP) with multiplex IF demonstrated that PD-L1 expression in macrophages but not tumor cells was a predictive marker for PFS, OS, and treatment response. Further, specific immune markers associated with PFS and OS, respectively (Toki et al., 2019). Cytometry time-of-flight imaging mass cytometry (CyTOF) is another tool that has been used in conjunction with multiplexing to show that proximity of antigen-experienced cytotoxic T cells (CD8+CD45RO + Ki67+) to melanoma cells was associated with positive response to ICI (Moldoveanu et al., 2022). In the metastatic setting, multiplexed mass cytometry-based imaging has shown that enrichment of B cell patches and follicles with naïve-like TCF7+ T cells is a favorable predictive indicator of ICI response (Hoch et al., 2022).
Artificial intelligence and multi-parameter biomarkers
While AI, a set of sophisticated algorithms and highly advanced machine learning tools to simulate some aspects of human intelligence, has greatly expanded its reach across all of medicine, it has demonstrated new potential horizons for melanoma biomarker development.
Machine learning, a subset of AI that involves computers improving performance from learned experience and pattern recognition, has been leveraged as an important tool for the identification of prognostic biomarkers. A study in 2022 utilized a machine learning classifier that accounted for multiple variables of TILs including cell type (e.g., tumor cells, immune cells) and area of interest (e.g., tumor, adjacent stroma) to validate the prognostic value of TILs for potential pathologist-independent use in future clinical trials. In this study, machine learning found that automated TIL score is prognostic in clinically-localized primary melanoma and may assist in isolating a subgroup of stage II patients with high recurrence risk. This will ultimately enable identification of patients who would likely benefit from adjuvant therapy (Aung et al., 2022). In another study, machine learning contributed to the development of immune diagnostic models to accurately classify melanoma patients from normal patients (Kulkarni et al., 2020; Du et al., 2022). Moreover, these authors could develop prognostic models to estimate composite risk score with clinical parameters to predict survival of over three to 5 years in melanoma patients. Patients can then be stratified based on these models into high versus low risk subgroups with different life expectancies (Du et al., 2022). A subsequent study, using machine learning, confirmed the prognostic value of TNM staging and also found that clinicopathological variables such as sex, tumor site, histotype, growth phase, and age, were linked to OS. The authors transformed their results into an online tool for prognostication for patients with melanoma (Cozzolino et al., 2023). Other AI techniques have been leveraged for biomarker discovery implicated in melanoma metastatic progression and have identified novel prognostic biomarkers (Miñoza et al., 2022).
Aside from prognostication, AI may process large amounts of available clinical and histopathologic data to aid physicians in determining the most favorable therapeutic choices for each patient and avoid treatments that are more likely to fail or lead to adverse events (Johnson et al., 2021; Guerrisi et al., 2022). A study has found that deep learning, another AI tool that uses algorithms modeled to operate similar to the human brain (i.e., artificial neural networks), applied to histology specimens and clinical data may predict ICI response in advanced melanoma (Johannet et al., 2021). Recent studies have integrated clinical outcomes and transcriptomic data from melanoma patients on ICI and have generated predictions for ICI treatment responses (Ahmed et al., 2022b; Kong et al., 2022). One study in particular was able to develop four machine learning models utilizing random-forest classification (RFC) incorporating clinical and genomic features (RFC7), differentially expressed genes (DEGs, RFC-Seq), survival-related DEGs (RFC-Surv) and a combination model. All models achieved high area under the curve (AUC), suggesting strong performances. These authors found that TMB, as well as the novel genes GSTA3 and VNN2, were important features in predicting ICI response (Ahmed et al., 2022b). Studies have also found that, in addition to clinical and transcriptomic data integration, simple segmentation of melanoma whole slide pathology images using machine learning can indicate ICI predictive biomarkers (Johannet et al., 2021; Li et al., 2021; Grossarth et al., 2023). Segmentation analyzes images at the pixel level to classify specific melanoma cells on the slide and ignore uninvolved tissue. This AI-based method has even achieved high sensitivity in detecting morphological changes in BRAF-mutated melanomas, providing additional information on targeted therapies (Kim et al., 2022).
Discussion
Melanoma is an aggressive skin cancer with rising yearly incidence. The growing field of biomarker detection in melanoma is very promising for determining prognosis and predicting treatment response. These biomarkers have tremendous implications for future therapeutic decision-making and drug development.
Currently, standard clinical care algorithms utilize TNM staging for prognosis. IHC has been able to elucidate many prognostic and predictive biomarkers including MART1/Ki-67, preferentially expressed antigen of melanoma (PRAME), makers of lymphovascular invasion (e.g., CD31/SOX-10) and mismatch repair (MMR) proteins, among many others (Torres-Cabala et al., 2020). However, a number of these markers are not routinely used in the clinic due to a variety of reasons, including lack of validation or accurate predictive potential (Diamandis, 2012). Thus, newer technologies are necessary for more robust analyses of biomarkers. For example, scRNA-seq and spatial transcriptomics have accounted for heterogeneity in melanoma which was a limitation of gene expression profiling of bulk tumor tissue. The latter technique has identified key genetic signatures such as BRAFV600E which have been considered when treating patients with ICI. However, it is possible that analysis of crosstalk between individual cells or the spatial influence of 1 cell on another may lead to identification of novel targets for treatment. Genomic immune-based (e.g., interferon) signatures have also stratified melanoma patients into low and high risk groups based on level of immune surveillance, which can further guide precision ICI. Subsetting these signatures based on single cell data may allow for improved accuracy.
Additional technologies that have elucidated cellular interactions in the TME include the use of multiplexed, quantitative IHC which has allowed for the analysis of multiple cellular phenotypes at a time, in addition to assessing proximity of individual cells to each other. In particular, multiplexing has better characterized TILs and their role in the TME.
AI-based analyses have also expanded biomarker discovery in melanoma. By simulating some aspects of human intelligence in a sophisticated and automated platform, these tools have the propensity to decrease inter-observer variability and error in order to more reliably quantify biomarker presence in patient samples based on integrated clinical and histopathologic data along with image analysis.
In the prognostic setting, the GEP test that classifies melanoma patients as Class 1 (low-risk) or Class 2 (high risk) for recurrence or metastasis is commercially available, which may allow clinicians to modify screening intervals and treatment regimens depending on a patient’s individual disease risk. However, studies have assessed the performance of this tool and while this GEP test generally identifies recurrence in patients with stage II disease, correctly identifying recurrence in stage I patients is poor, limiting its clinical utility (March et al., 2020). Thus, some of the aforementioned prognostic biomarkers will require further investigation for integration into standard AJCC staging and use in the clinic.
In the predictive setting, therapeutically targeting PD-1 and CTLA-4 correlate with clinical benefit. However, given intratumoral heterogeneity and limited ICI options, these markers are insufficient to capture the nature of all patient tumors. Additional biomarkers that have been well-explored such as the TMB and inflammatory mediators may soon be utilized in clinical settings. Newly discovered biomarkers such as antigen experienced cytotoxic T cells are likely to require additional evaluation, although preliminary data shows promise for predicting ICI treatment response.
While biomarkers may serve as independent prognostic or predictive indicators, a single biomarker is usually inadequate to precisely stratify patients. Thus, multimodal investigation of biomarkers using a combination of the techniques described while also prioritizing sensitivity, specificity and cost will be important for timely assessment of future patient risk and response.
Author contributions
OA: Conceptualization, Investigation, Writing–original draft, Writing–review and editing. EG: Writing–original draft, Writing–review and editing. DK: Writing–original draft. YB: Writing–original draft. AS: Writing–original draft. GE: Writing–original draft. LG: Writing–review and editing. YS: Writing–original draft, Conceptualization, Investigation, Supervision, Writing–review and editing.
Funding
The author(s) declare that no financial support was received for the research, authorship, and/or publication of this article.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
Abbas, O., Miller, D. D., and Bhawan, J. (2014). Cutaneous malignant melanoma: update on diagnostic and prognostic biomarkers. Am. J. Dermatopathol. 36, 363–379. doi:10.1097/DAD.0b013e31828a2ec5
Adler, N. R., Wolfe, R., Kelly, J. W., Haydon, A., McArthur, G. A., McLean, C. A., et al. (2017). Tumour mutation status and sites of metastasis in patients with cutaneous melanoma. Br. J. Cancer 117, 1026–1035. doi:10.1038/bjc.2017.254
Ahmed, R., Zaman, T., Chowdhury, F., Mraiche, F., Tariq, M., Ahmad, I. S., et al. (2022a). Single-cell RNA sequencing with spatial transcriptomics of cancer tissues. Int. J. Mol. Sci. 23, 3042. doi:10.3390/ijms23063042
Ahmed, Y. B., Al-Bzour, A. N., Ababneh, O. E., Abushukair, H. M., and Saeed, A. (2022b). Genomic and transcriptomic predictors of response to immune checkpoint inhibitors in melanoma patients: a machine learning approach. Cancers 14, 5605. doi:10.3390/cancers14225605
Akbani, R. (2015). Genomic classification of cutaneous melanoma. Cell 161, 1681–1696. doi:10.1016/j.cell.2015.05.044
Ascierto, P. A., Minor, D., Ribas, A., Lebbe, C., O'Hagan, A., Arya, N., et al. (2013). Phase II trial (BREAK-2) of the BRAF inhibitor dabrafenib (GSK2118436) in patients with metastatic melanoma. J. Clin. Oncol. 31, 3205–3211. doi:10.1200/JCO.2013.49.8691
Atkins, M. B., Lee, S. J., Chmielowski, B., Tarhini, A. A., Cohen, G. I., Truong, T. G., et al. (2023). Combination dabrafenib and trametinib versus combination nivolumab and ipilimumab for patients with advanced BRAF -mutant melanoma: the DREAMseq trial—ECOG-ACRIN EA6134. J. Clin. Oncol. 41, 186–197. doi:10.1200/JCO.22.01763
Aung, T. N., Shafi, S., Wilmott, J. S., Nourmohammadi, S., Vathiotis, I., Gavrielatou, N., et al. (2022). Objective assessment of tumor infiltrating lymphocytes as a prognostic marker in melanoma using machine learning algorithms. EBioMedicine 82, 104143. doi:10.1016/j.ebiom.2022.104143
Azimi, F., Scolyer, R. A., Rumcheva, P., Moncrieff, M., Murali, R., McCarthy, S. W., et al. (2012). Tumor-infiltrating lymphocyte grade is an independent predictor of sentinel lymph node status and survival in patients with cutaneous melanoma. J. Clin. Oncol. 30, 2678–2683. doi:10.1200/JCO.2011.37.8539
Balch, C. M., Gershenwald, J. E., Soong, S. J., Thompson, J. F., Ding, S., Byrd, D. R., et al. (2010). Multivariate analysis of prognostic factors among 2,313 patients with stage III melanoma: comparison of nodal micrometastases versus macrometastases. J. Clin. Oncol. Off. J. Am. Soc. Clin. Oncol. 28, 2452–2459. doi:10.1200/JCO.2009.27.1627
Balch, C. M., Soong, S. J., Gershenwald, J. E., Thompson, J. F., Reintgen, D. S., Cascinelli, N., et al. (2001). Prognostic factors analysis of 17,600 melanoma patients: validation of the American Joint committee on cancer melanoma staging system. J. Clin. Oncol. 19, 3622–3634. doi:10.1200/JCO.2001.19.16.3622
Barth, A., Wanek, L. A., and Morton, D. L. (1995). Prognostic factors in 1,521 melanoma patients with distant metastases. J. Am. Coll. Surg. 181, 193–201.
Berry, S., Giraldo, N. A., Green, B. F., Cottrell, T. R., Stein, J. E., Engle, E. L., et al. (2021). Analysis of multispectral imaging with the AstroPath platform informs efficacy of PD-1 blockade. Science 372, eaba2609. doi:10.1126/science.aba2609
Bindea, G., Mlecnik, B., Fridman, W.-H., and Galon, J. (2011). The prognostic impact of anti-cancer immune response: a novel classification of cancer patients. Semin. Immunopathol. 33, 335–340. doi:10.1007/s00281-011-0264-x
Bogunovic, D., O'Neill, D. W., Belitskaya-Levy, I., Vacic, V., Yu, Y. L., Adams, S., et al. (2009). Immune profile and mitotic index of metastatic melanoma lesions enhance clinical staging in predicting patient survival. Proc. Natl. Acad. Sci. 106, 20429–20434. doi:10.1073/pnas.0905139106
Burotto, M., Chiou, V. L., Lee, J.-M., and Kohn, E. C. (2014). The MAPK pathway across different malignancies: a new perspective. Cancer 120, 3446–3456. doi:10.1002/cncr.28864
Busam, K. J., Antonescu, C. R., Marghoob, A. A., Nehal, K. S., Sachs, D. L., Shia, J., et al. (2001). Histologic classification of tumor-infiltrating lymphocytes in primary cutaneous malignant melanoma: a study of interobserver agreement. Am. J. Clin. Pathol. 115, 856–860. doi:10.1309/G6EK-Y6EH-0LGY-6D6P
Calapre, L., Warburton, L., Millward, M., Ziman, M., and Gray, E. S. (2017). Circulating tumour DNA (ctDNA) as a liquid biopsy for melanoma. Cancer Lett. 404, 62–69. doi:10.1016/j.canlet.2017.06.030
Cascinelli, N., Belli, F., Santinami, M., Fait, V., Testori, A., Ruka, W., et al. (2000). Sentinel lymph node biopsy in cutaneous melanoma: the WHO Melanoma Program experience. Ann. Surg. Oncol. 7, 469–474. doi:10.1007/s10434-000-0469-z
Cirenajwis, H., Lauss, M., Ekedahl, H., Törngren, T., Kvist, A., Saal, L. H., et al. (2017). NF1-mutated melanoma tumors harbor distinct clinical and biological characteristics. Mol. Oncol. 11, 438–451. doi:10.1002/1878-0261.12050
Clemente, C. G., Mihm, M. C., Bufalino, R., Zurrida, S., Collini, P., and Cascinelli, N. (1996). Prognostic value of tumor infiltrating lymphocytes in the vertical growth phase of primary cutaneous melanoma. Cancer 77, 1303–1310. doi:10.1002/(SICI)1097-0142(19960401)77:7<1303:AID-CNCR12>3.0.CO;2-5
Cozzolino, C., Buja, A., Rugge, M., Miatton, A., Zorzi, M., Vecchiato, A., et al. (2023). Machine learning to predict overall short-term mortality in cutaneous melanoma. Discov. Oncol. 14, 13. doi:10.1007/s12672-023-00622-5
Dahl, C., and Guldberg, P. (2007). The genome and epigenome of malignant melanoma. APMIS 115, 1161–1176. doi:10.1111/j.1600-0463.2007.apm_855.xml.x
Diamandis, E. P. (2012). The failure of protein cancer biomarkers to reach the clinic: why, and what can be done to address the problem? BMC Med. 10, 87. doi:10.1186/1741-7015-10-87
Du, H., He, Y., Lu, W., Han, Y., and Wan, Q. (2022). Machine learning analysis of immune cells for diagnosis and prognosis of cutaneous melanoma. J. Oncol. 2022, 7357637–7357715. doi:10.1155/2022/7357637
Dzwierzynski, W. W. (2021). Melanoma risk factors and prevention. Clin. Plast. Surg. 48, 543–550. doi:10.1016/j.cps.2021.05.001
Fan, L., Li, H., and Wang, W. (2020). Long non-coding RNA PRRT3-AS1 silencing inhibits prostate cancer cell proliferation and promotes apoptosis and autophagy. Exp. Physiol. 105, 793–808. doi:10.1113/EP088011
Fridman, W. H., Galon, J., Dieu-Nosjean, M. C., Cremer, I., Fisson, S., Damotte, D., et al. (2010). “Immune infiltration in human cancer: prognostic significance and disease control,” in Cancer immunology and immunotherapy. Editor G. Dranoff (Springer Berlin Heidelberg), 344, 1–24.
Gartrell, R. D., Marks, D. K., Hart, T. D., Li, G., Davari, D. R., Wu, A., et al. (2018). Quantitative analysis of immune infiltrates in primary melanoma. Cancer Immunol. Res. 6, 481–493. doi:10.1158/2326-6066.CIR-17-0360
Gartrell, R. D., Marks, D. K., Rizk, E. M., Bogardus, M., Gérard, C. L., Barker, L. W., et al. (2019). Validation of melanoma immune profile (MIP), a prognostic immune gene prediction score for stage II–III melanoma. Clin. Cancer Res. 25, 2494–2502. doi:10.1158/1078-0432.CCR-18-2847
Gartrell-Corrado, R. D., Chen, A. X., Rizk, E. M., Marks, D. K., Bogardus, M. H., Hart, T. D., et al. (2020). Linking transcriptomic and imaging data defines features of a favorable tumor immune microenvironment and identifies a combination biomarker for primary melanoma. Cancer Res. 80, 1078–1087. doi:10.1158/0008-5472.CAN-19-2039
Gershenwald, J. E., Scolyer, R. A., Hess, K. R., Sondak, V. K., Long, G. V., Ross, M. I., et al. (2017). Melanoma staging: evidence-based changes in the American Joint Committee on Cancer eighth edition cancer staging manual. Ca. Cancer J. Clin. 67, 472–492. doi:10.3322/caac.21409
Gide, T. N., Silva, I. P., Quek, C., Ahmed, T., Menzies, A. M., Carlino, M. S., et al. (2020). Close proximity of immune and tumor cells underlies response to anti-PD-1 based therapies in metastatic melanoma patients. Oncoimmunology 9, 1659093. doi:10.1080/2162402X.2019.1659093
Goodman, A. M., Kato, S., Bazhenova, L., Patel, S. P., Frampton, G. M., Miller, V., et al. (2017). Tumor mutational burden as an independent predictor of response to immunotherapy in diverse cancers. Mol. Cancer Ther. 16, 2598–2608. doi:10.1158/1535-7163.MCT-17-0386
Grossarth, S., Mosley, D., Madden, C., Ike, J., Smith, I., Huo, Y., et al. (2023). Recent advances in melanoma diagnosis and prognosis using machine learning methods. Curr. Oncol. Rep. 25, 635–645. doi:10.1007/s11912-023-01407-3
Guerrisi, A., Falcone, I., Valenti, F., Rao, M., Gallo, E., Ungania, S., et al. (2022). Artificial intelligence and advanced melanoma: treatment management implications. Cells 11, 3965. doi:10.3390/cells11243965
Hawryluk, E. B., and Tsao, H. (2014). Melanoma: clinical features and genomic insights. Cold Spring Harb. Perspect. Med. 4, a015388. doi:10.1101/cshperspect.a015388
Hodi, F. S., Corless, C. L., Giobbie-Hurder, A., Fletcher, J. A., Zhu, M., Marino-Enriquez, A., et al. (2013). Imatinib for melanomas harboring mutationally activated or amplified KIT arising on mucosal, acral, and chronically sun-damaged skin. J. Clin. Oncol. 31, 3182–3190. doi:10.1200/JCO.2012.47.7836
Hoch, T., Schulz, D., Eling, N., Gómez, J. M., Levesque, M. P., and Bodenmiller, B. (2022). Multiplexed imaging mass cytometry of the chemokine milieus in melanoma characterizes features of the response to immunotherapy. Sci. Immunol. 7, eabk1692. doi:10.1126/sciimmunol.abk1692
Hodi, F. S., O'Day, S. J., McDermott, D. F., Weber, R. W., Sosman, J. A., Haanen, J. B., et al. (2010). Improved survival with ipilimumab in patients with metastatic melanoma. N. Engl. J. Med. 363, 711–723. doi:10.1056/NEJMoa1003466
Hu-Lieskovan, S., Bhaumik, S., Dhodapkar, K., Grivel, J. C. J. B., Gupta, S., Hanks, B. A., et al. (2020). SITC cancer immunotherapy resource document: a compass in the land of biomarker discovery. J. Immunother. Cancer 8, e000705. doi:10.1136/jitc-2020-000705
Huang, N., Lee, K. J., and Stark, M. S. (2022). Current trends in circulating biomarkers for melanoma detection. Front. Med. 9, 873728. doi:10.3389/fmed.2022.873728
Hugo, W., Zaretsky, J. M., Sun, L., Song, C., Moreno, B. H., Hu-Lieskovan, S., et al. (2016). Genomic and transcriptomic features of response to anti-PD-1 therapy in metastatic melanoma. Cell 165, 35–44. doi:10.1016/j.cell.2016.02.065
Hussein, M. R., Elsers, D. a. H., Fadel, S. A., and Omar, A.-E. M. (2006). Immunohistological characterisation of tumour infiltrating lymphocytes in melanocytic skin lesions. J. Clin. Pathol. 59, 316–324. doi:10.1136/jcp.2005.028860
Huuhtanen, J., Kasanen, H., Peltola, K., Lönnberg, T., Glumoff, V., Brück, O., et al. (2023). Single-cell characterization of anti–LAG-3 and anti–PD-1 combination treatment in patients with melanoma. J. Clin. Invest. 133, e164809. doi:10.1172/JCI164809
In ’t Hout, F. E., Haydu, L. E., Murali, R., Bonenkamp, J. J., Thompson, J. F., and Scolyer, R. A. (2012). Prognostic importance of the extent of ulceration in patients with clinically localized cutaneous melanoma. Ann. Surg. 255, 1165–1170. doi:10.1097/SLA.0b013e31824c4b0b
Johannet, P., Coudray, N., Donnelly, D. M., Jour, G., Illa-Bochaca, I., Xia, Y., et al. (2021). Using machine learning algorithms to predict immunotherapy response in patients with advanced melanoma. Clin. Cancer Res. 27, 131–140. doi:10.1158/1078-0432.CCR-20-2415
Johnson, K. B., Wei, W. Q., Weeraratne, D., Frisse, M. E., Misulis, K., Rhee, K., et al. (2021). Precision medicine, AI, and the future of personalized health care. Clin. Transl. Sci. 14, 86–93. doi:10.1111/cts.12884
Keung, E. Z., and Gershenwald, J. E. (2018). The eighth edition American Joint Committee on Cancer (AJCC) melanoma staging system: implications for melanoma treatment and care. Expert Rev. Anticancer Ther. 18, 775–784. doi:10.1080/14737140.2018.1489246
Kim, R. H., Nomikou, S., Coudray, N., Jour, G., Dawood, Z., Hong, R., et al. (2022). Deep learning and pathomics analyses reveal cell nuclei as important features for mutation prediction of BRAF-mutated melanomas. J. Invest. Dermatol. 142, 1650–1658.e6. doi:10.1016/j.jid.2021.09.034
Kong, J., Ha, D., Lee, J., Kim, I., Park, M., Im, S. H., et al. (2022). Network-based machine learning approach to predict immunotherapy response in cancer patients. Nat. Commun. 13, 3703. doi:10.1038/s41467-022-31535-6
Kulkarni, P. M., Robinson, E. J., Sarin Pradhan, J., Gartrell-Corrado, R. D., Rohr, B. R., Trager, M. H., et al. (2020). Deep learning based on standard H&E images of primary melanoma tumors identifies patients at risk for visceral recurrence and death. Clin. Cancer Res. Off. J. Am. Assoc. Cancer Res. 26, 1126–1134. doi:10.1158/1078-0432.CCR-19-1495
Lee, J. H., Long, G. V., Boyd, S., Lo, S., Menzies, A. M., Tembe, V., et al. (2017). Circulating tumour DNA predicts response to anti-PD1 antibodies in metastatic melanoma. Ann. Oncol. 28, 1130–1136. doi:10.1093/annonc/mdx026
Li, T., Xie, P., Liu, J., Chen, M., Zhao, S., Kang, W., et al. (2021). Automated diagnosis and localization of melanoma from skin histopathology slides using deep learning: a multicenter study. J. Healthc. Eng. 2021, 5972962. doi:10.1155/2021/5972962
Liang, R., Zhi, Y., Zheng, G., Zhang, B., Zhu, H., and Wang, M. (2018). Analysis of long non-coding RNAs in glioblastoma for prognosis prediction using weighted gene co-expression network analysis, Cox regression, and L1-LASSO penalization. OncoTargets Ther. 12, 157–168. doi:10.2147/OTT.S171957
Lim, B., Lin, Y., and Navin, N. (2020). Advancing cancer research and medicine with single-cell genomics. Cancer Cell 37, 456–470. doi:10.1016/j.ccell.2020.03.008
Litchfield, K., Reading, J. L., Puttick, C., Thakkar, K., Abbosh, C., Bentham, R., et al. (2021). Meta-analysis of tumor- and T cell-intrinsic mechanisms of sensitization to checkpoint inhibition. Cell 184, 596–614.e14. doi:10.1016/j.cell.2021.01.002
Long, G. V., Flaherty, K. T., Stroyakovskiy, D., Gogas, H., Levchenko, E., de Braud, F., et al. (2017b). Dabrafenib plus trametinib versus dabrafenib monotherapy in patients with metastatic BRAF V600E/K-mutant melanoma: long-term survival and safety analysis of a phase 3 study. Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 28, 1631–1639. doi:10.1093/annonc/mdx176
Long, G. V., Hauschild, A., Santinami, M., Atkinson, V., Mandalà, M., Chiarion-Sileni, V., et al. (2017a). Adjuvant dabrafenib plus trametinib in stage III BRAF-mutated melanoma. N. Engl. J. Med. 377, 1813–1823. doi:10.1056/NEJMoa1708539
Longo, S. K., Guo, M. G., Ji, A. L., and Khavari, P. A. (2021). Integrating single-cell and spatial transcriptomics to elucidate intercellular tissue dynamics. Nat. Rev. Genet. 22, 627–644. doi:10.1038/s41576-021-00370-8
Lozano, A. X., Chaudhuri, A. A., Nene, A., Bacchiocchi, A., Earland, N., Vesely, M. D., et al. (2022). T cell characteristics associated with toxicity to immune checkpoint blockade in patients with melanoma. Nat. Med. 28, 353–362. doi:10.1038/s41591-021-01623-z
Luke, J. J., Rutkowski, P., Queirolo, P., Del Vecchio, M., Mackiewicz, J., Chiarion-Sileni, V., et al. (2022). Pembrolizumab versus placebo as adjuvant therapy in completely resected stage IIB or IIC melanoma (KEYNOTE-716): a randomised, double-blind, phase 3 trial. Lancet 399, 1718–1729. doi:10.1016/S0140-6736(22)00562-1
Lynch, K. T., Young, S. J., Meneveau, M. O., Wages, N. A., Engelhard, V. H., Slingluff, C. L., et al. (2021). Heterogeneity in tertiary lymphoid structure B-cells correlates with patient survival in metastatic melanoma. J. Immunother. Cancer 9, e002273. doi:10.1136/jitc-2020-002273
Marchetti, M. A., Coit, D. G., Dusza, S. W., Yu, A., McLean, L., Hu, Y., et al. (2020). Performance of gene expression profile tests for prognosis in patients with localized cutaneous melanoma: a systematic review and meta-analysis. JAMA Dermatol 156, 953–962. doi:10.1001/jamadermatol.2020.1731
Mauldin, I. S., Mahmutovic, A., Young, S. J., and Slingluff, C. L. (2021). “Multiplex immunofluorescence histology for immune cell infiltrates in melanoma-associated tertiary lymphoid structures,” in Melanoma. Editor K. M. Hargadon (Springer US), 2265, 573–587.
Maynard, A., McCoach, C. E., Rotow, J. K., Harris, L., Haderk, F., Kerr, D. L., et al. (2020). Therapy-induced evolution of human lung cancer revealed by single-cell RNA sequencing. Cell 182, 1232–1251. doi:10.1016/j.cell.2020.07.017
Meckbach, D., Bauer, J., Pflugfelder, A., Meier, F., Busch, C., Eigentler, T. K., et al. (2014). Survival according to BRAF-V600 tumor mutations-an analysis of 437 patients with primary melanoma. PloS One 9, e86194. doi:10.1371/journal.pone.0086194
Miñoza, J. M. A., Rico, J. A., Zamora, P. R. F., Bacolod, M., Laubenbacher, R., Dumancas, G. G., et al. (2022). Biomarker discovery for meta-classification of melanoma metastatic progression using transfer learning. Genes 13, 2303. doi:10.3390/genes13122303
Moldoveanu, D., Ramsay, L., Lajoie, M., Anderson-Trocme, L., Lingrand, M., Berry, D., et al. (2022). Spatially mapping the immune landscape of melanoma using imaging mass cytometry. Sci. Immunol. 7, eabi5072. doi:10.1126/sciimmunol.abi5072
Monberg, T. J., Borch, T. H., Svane, I. M., and Donia, M. (2023). TIL therapy: facts and hopes. Clin. Cancer Res. 29, 3275–3283. doi:10.1158/1078-0432.CCR-22-2428
Naimy, S., Bzorek, M., Eriksen, J. O., Dyring-Andersen, B., and Rahbek Gjerdrum, L. M. (2023). BRAFV600E expression is homogenous and associated with nonrecurrent disease and better survival in primary melanoma. Dermatology 239, 409–421. doi:10.1159/000528159
Nalejska, E., Mączyńska, E., and Lewandowska, M. A. (2014). Prognostic and predictive biomarkers: tools in personalized oncology. Mol. Diagn. Ther. 18, 273–284. doi:10.1007/s40291-013-0077-9
Nguyen, T., Kocovski, N., Macdonald, S., Yeang, H. X. A., Wang, M., and Neeson, P. J. (2021). Multiplex immunohistochemistry analysis of melanoma tumor-infiltrating lymphocytes. Methods Mol. Biol. Clifton N. J. 2265, 557–572. doi:10.1007/978-1-0716-1205-7_39
Nikolaou, V. A., Stratigos, A. J., Flaherty, K. T., and Tsao, H. (2012). Melanoma: new insights and new therapies. J. Invest. Dermatol. 132, 854–863. doi:10.1038/jid.2011.421
Nirmal, A. J., Maliga, Z., Vallius, T., Quattrochi, B., Chen, A. A., Jacobson, C. A., et al. (2022). The spatial landscape of progression and immunoediting in primary melanoma at single-cell resolution. Cancer Discov. 12, 1518–1541. doi:10.1158/2159-8290.CD-21-1357
Ny, L., Hernberg, M., Nyakas, M., Koivunen, J., Oddershede, L., Yoon, M., et al. (2020). BRAF mutational status as a prognostic marker for survival in malignant melanoma: a systematic review and meta-analysis. Acta Oncol. 59, 833–844. doi:10.1080/0284186X.2020.1747636
Oble, D. A., Loewe, R., Yu, P., and Mihm, M. C. (2009). Focus on TILs: prognostic significance of tumor infiltrating lymphocytes in human melanoma. Cancer Immun. 9, 3.
O’Neill, C. H., and Scoggins, C. R. (2019). Melanoma. J. Surg. Oncol. 120, 873–881. doi:10.1002/jso.25604
Orlow, I., Sadeghi, K. D., Edmiston, S. N., Kenney, J. M., Lezcano, C., Wilmott, J. S., et al. (2023). InterMEL: an international biorepository and clinical database to uncover predictors of survival in early-stage melanoma. PLOS ONE 18, e0269324. doi:10.1371/journal.pone.0269324
Pai, J. A., and Satpathy, A. T. (2021). High-throughput and single-cell T cell receptor sequencing technologies. Nat. Methods 18, 881–892. doi:10.1038/s41592-021-01201-8
Patel, S. P., Othus, M., Chen, Y., Wright, G. P., Yost, K. J., Hyngstrom, J. R., et al. (2023). Neoadjuvant–adjuvant or adjuvant-only pembrolizumab in advanced melanoma. N. Engl. J. Med. 388, 813–823. doi:10.1056/NEJMoa2211437
Pires Da Silva, I., Wang, K. Y. X., Wilmott, J. S., Holst, J., Carlino, M. S., Park, J. J., et al. (2019). Distinct molecular profiles and immunotherapy treatment outcomes of V600E and V600K BRAF -mutant melanoma. Clin. Cancer Res. 25, 1272–1279. doi:10.1158/1078-0432.CCR-18-1680
Piwecka, M., Rajewsky, N., and Rybak-Wolf, A. (2023). Single-cell and spatial transcriptomics: deciphering brain complexity in health and disease. Nat. Rev. Neurol. 19, 346–362. doi:10.1038/s41582-023-00809-y
Podlipnik, S., Potrony, M., and Puig, S. (2021). Genetic markers for characterization and prediction of prognosis of melanoma subtypes: a 2021 update. Ital. J. Dermatol. Venereol. 156, 322–330. doi:10.23736/S2784-8671.21.06957-1
Randic, T., Kozar, I., Margue, C., Utikal, J., and Kreis, S. (2021). NRAS mutant melanoma: towards better therapies. Cancer Treat. Rev. 99, 102238. doi:10.1016/j.ctrv.2021.102238
Rao, U. N. M., Ibrahim, J., Flaherty, L. E., Richards, J., and Kirkwood, J. M. (2002). Implications of microscopic satellites of the primary and extracapsular lymph node spread in patients with high-risk melanoma: pathologic corollary of Eastern Cooperative Oncology Group Trial E1690. J. Clin. Oncol. Off. J. Am. Soc. Clin. Oncol. 20, 2053–2057. doi:10.1200/JCO.2002.08.024
Rastrelli, M., Tropea, S., Rossi, C. R., and Alaibac, M. (2014). Melanoma: epidemiology, risk factors, pathogenesis, diagnosis and classification. Vivo Athens Greece 28, 1005–1011.
Read, R. L., Haydu, L., Saw, R. P. M., Quinn, M. J., Shannon, K., Spillane, A. J., et al. (2015). In-transit melanoma metastases: incidence, prognosis, and the role of lymphadenectomy. Ann. Surg. Oncol. 22, 475–481. doi:10.1245/s10434-014-4100-0
Rizk, E. M., Gartrell, R. D., Barker, L. W., Esancy, C. L., Finkel, G. G., Bordbar, D. D., et al. (2019). Prognostic and predictive immunohistochemistry-based biomarkers in cancer and immunotherapy. Hematol. Oncol. Clin. North Am. 33, 291–299. doi:10.1016/j.hoc.2018.12.005
Robert, C., Thomas, L., Bondarenko, I., O'Day, S., Weber, J., Garbe, C., et al. (2011). Ipilimumab plus dacarbazine for previously untreated metastatic melanoma. N. Engl. J. Med. 364, 2517–2526. doi:10.1056/NEJMoa1104621
Rohaan, M. W., Borch, T. H., van den Berg, J. H., Met, Ö., Kessels, R., Geukes Foppen, M. H., et al. (2022). Tumor-infiltrating lymphocyte therapy or ipilimumab in advanced melanoma. N. Engl. J. Med. 387, 2113–2125. doi:10.1056/NEJMoa2210233
Sacco, A., Forgione, L., Carotenuto, M., Luca, A. D., Ascierto, P. A., Botti, G., et al. (2020). Circulating tumor DNA testing opens new perspectives in melanoma management. Cancers 12, 2914. doi:10.3390/cancers12102914
Sanlorenzo, M., Vujic, I., Posch, C., Dajee, A., Yen, A., Kim, S., et al. (2014). Melanoma immunotherapy. Cancer Biol. Ther. 15, 665–674. doi:10.4161/cbt.28555
Sanmamed, M. F., Fernández-Landázuri, S., Rodríguez, C., Zárate, R., Lozano, M. D., Zubiri, L., et al. (2015). Quantitative cell-free circulating BRAFV600E mutation analysis by use of droplet digital PCR in the follow-up of patients with melanoma being treated with BRAF inhibitors. Clin. Chem. 61, 297–304. doi:10.1373/clinchem.2014.230235
Santiago-Walker, A., Gagnon, R., Mazumdar, J., Casey, M., Long, G. V., Schadendorf, D., et al. (2016). Correlation of BRAF mutation status in circulating-free DNA and tumor and association with clinical outcome across four BRAFi and MEKi clinical trials. Clin. Cancer Res. 22, 567–574. doi:10.1158/1078-0432.CCR-15-0321
Sarnaik, A. A., Yu, B., Yu, D., Morelli, D., Hall, M., Bogle, D., et al. (2011). Extended dose ipilimumab with a peptide vaccine: immune correlates associated with clinical benefit in patients with resected high-risk stage IIIc/IV melanoma. Clin. Cancer Res. 17, 896–906. doi:10.1158/1078-0432.CCR-10-2463
Sivendran, S., Chang, R., Pham, L., Phelps, R. G., Harcharik, S. T., Hall, L. D., et al. (2014). Dissection of immune gene networks in primary melanoma tumors critical for antitumor surveillance of patients with stage II–III resectable disease. J. Invest. Dermatol. 134, 2202–2211. doi:10.1038/jid.2014.85
Somasundaram, R., Connelly, T., Choi, R., Choi, H., Samarkina, A., Li, L., et al. (2021). Tumor-infiltrating mast cells are associated with resistance to anti-PD-1 therapy. Nat. Commun. 12, 346. doi:10.1038/s41467-020-20600-7
Ståhl, P. L., Salmén, F., Vickovic, S., Lundmark, A., Navarro, J. F., Magnusson, J., et al. (2016). Visualization and analysis of gene expression in tissue sections by spatial transcriptomics. Science 353, 78–82. doi:10.1126/science.aaf2403
Su, Y., Wei, W., Robert, L., Xue, M., Tsoi, J., Garcia-Diaz, A., et al. (2017). Single-cell analysis resolves the cell state transition and signaling dynamics associated with melanoma drug-induced resistance. Proc. Natl. Acad. Sci. 114, 13679–13684. doi:10.1073/pnas.1712064115
Tan, L., Sandhu, S., Lee, R. J., Li, J., Callahan, J., Ftouni, S., et al. (2019). Prediction and monitoring of relapse in stage III melanoma using circulating tumor DNA. Ann. Oncol. 30, 804–814. doi:10.1093/annonc/mdz048
Tan, W. C. C., Nerurkar, S. N., Cai, H. Y., Ng, H. H. M., Wu, D., Wee, Y. T. F., et al. (2020). Overview of multiplex immunohistochemistry/immunofluorescence techniques in the era of cancer immunotherapy. Cancer Commun. 40, 135–153. doi:10.1002/cac2.12023
Tímár, J., and Ladányi, A. (2022). Molecular pathology of skin melanoma: epidemiology, differential diagnostics, prognosis and therapy prediction. Int. J. Mol. Sci. 23, 5384. doi:10.3390/ijms23105384
Tirosh, I., Izar, B., Prakadan, S. M., Wadsworth, M. H., Treacy, D., Trombetta, J. J., et al. (2016). Dissecting the multicellular ecosystem of metastatic melanoma by single-cell RNA-seq. Science 352, 189–196. doi:10.1126/science.aad0501
Tivey, A., Britton, F., Scott, J. A., Rothwell, D., Lorigan, P., and Lee, R. (2022). Circulating tumour DNA in melanoma—clinic ready? Curr. Oncol. Rep. 24, 363–373. doi:10.1007/s11912-021-01151-6
Toki, M. I., Merritt, C. R., Wong, P. F., Smithy, J. W., Kluger, H. M., Syrigos, K. N., et al. (2019). High-plex predictive marker discovery for melanoma immunotherapy–treated patients using digital spatial profiling. Clin. Cancer Res. 25, 5503–5512. doi:10.1158/1078-0432.CCR-19-0104
Topalian, S. L., Hodi, F. S., Brahmer, J. R., Gettinger, S. N., Smith, D. C., McDermott, D. F., et al. (2012). Safety, activity, and immune correlates of anti-PD-1 antibody in cancer. N. Engl. J. Med. 366, 2443–2454. doi:10.1056/NEJMoa1200690
Torres-Cabala, C., Li-Ning-Tapia, E., and Hwu, W.-J. (2020). Pathology-based biomarkers useful for clinical decisions in melanoma. Arch. Med. Res. 51, 827–838. doi:10.1016/j.arcmed.2020.09.008
Tumeh, P. C., Harview, C. L., Yearley, J. H., Shintaku, I. P., Taylor, E. J. M., Robert, L., et al. (2014). PD-1 blockade induces responses by inhibiting adaptive immune resistance. Nature 515, 568–571. doi:10.1038/nature13954
Ugolini, F., Pasqualini, E., Simi, S., Baroni, G., and Massi, D. (2022). Bright-field multiplex immunohistochemistry assay for tumor microenvironment evaluation in melanoma tissues. Cancers 14, 3682. doi:10.3390/cancers14153682
Uhara, H. (2019). Recent advances in therapeutic strategies for unresectable or metastatic melanoma and real-world data in Japan. Int. J. Clin. Oncol. 24, 1508–1514. doi:10.1007/s10147-018-1246-y
van Akkooi, A. C. J., Nowecki, Z. I., Voit, C., Schäfer-Hesterberg, G., Michej, W., de Wilt, J. H. W., et al. (2008). Sentinel node tumor burden according to the Rotterdam criteria is the most important prognostic factor for survival in melanoma patients: a multicenter study in 388 patients with positive sentinel nodes. Ann. Surg. 248, 949–955. doi:10.1097/SLA.0b013e31818fefe0
Van Es, S. L., Colman, M., Thompson, J. F., McCarthy, S. W., and Scolyer, R. A. (2008). Angiotropism is an independent predictor of local recurrence and in-transit metastasis in primary cutaneous melanoma. Am. J. Surg. Pathol. 32, 1396–1403. doi:10.1097/PAS.0b013e3181753a8e
Veatch, J. R., Lee, S. M., Shasha, C., Singhi, N., Szeto, J. L., Moshiri, A. S., et al. (2022). Neoantigen-specific CD4+ T cells in human melanoma have diverse differentiation states and correlate with CD8+ T cell, macrophage, and B cell function. Cancer Cell 40, 393–409.e9. doi:10.1016/j.ccell.2022.03.006
Volkovova, K., Bilanicova, D., Bartonova, A., Letašiová, S., and Dusinska, M. (2012). Associations between environmental factors and incidence of cutaneous melanoma. Review. Environ. Health Glob. Access Sci. Source 11 (1), S12. doi:10.1186/1476-069X-11-S1-S12
Wang, Y., Fan, J. L., Melms, J. C., Amin, A. D., Georgis, Y., Barrera, I., et al. (2023). Multimodal single-cell and whole-genome sequencing of small, frozen clinical specimens. Nat. Genet. 55, 19–25. doi:10.1038/s41588-022-01268-9
Wilmott, J., Haydu, L., Bagot, M., Zhang, Y., Jakrot, V., McCarthy, S., et al. (2012). Angiotropism is an independent predictor of microscopic satellites in primary cutaneous melanoma. Histopathology 61, 889–898. doi:10.1111/j.1365-2559.2012.04279.x
Wolchok, J. D., Kluger, H., Callahan, M. K., Postow, M. A., Rizvi, N. A., Lesokhin, A. M., et al. (2013). Nivolumab plus ipilimumab in advanced melanoma. N. Engl. J. Med. 369, 122–133. doi:10.1056/NEJMoa1302369
Wolchok, J. D., Neyns, B., Linette, G., Negrier, S., Lutzky, J., Thomas, L., et al. (2010). Ipilimumab monotherapy in patients with pretreated advanced melanoma: a randomised, double-blind, multicentre, phase 2, dose-ranging study. Lancet Oncol. 11, 155–164. doi:10.1016/S1470-2045(09)70334-1
Wu, J., Zeng, D., Zhi, S., Ye, Z., Qiu, W., Huang, N., et al. (2021). Single-cell analysis of a tumor-derived exosome signature correlates with prognosis and immunotherapy response. J. Transl. Med. 19, 381. doi:10.1186/s12967-021-03053-4
Yang, S., Zhou, Y., Zhang, X., Wang, L., Fu, J., Zhao, X., et al. (2021). The prognostic value of an autophagy-related lncRNA signature in hepatocellular carcinoma. BMC Bioinforma. 22, 217. doi:10.1186/s12859-021-04123-6
Yang, T.-T., Yu, S., Ke, C.-L. K., and Cheng, S.-T. (2023). The genomic landscape of melanoma and its therapeutic implications. Genes 14, 1021. doi:10.3390/genes14051021
Yaseen, Z., Gide, T. N., Conway, J. W., Potter, A. J., Quek, C., Hong, A. M., et al. (2022). Validation of an accurate automated multiplex immunofluorescence method for immuno-profiling melanoma. Front. Mol. Biosci. 9, 810858. doi:10.3389/fmolb.2022.810858
Yu, X., Zheng, H., Tse, G., Chan, M. T., and Wu, W. K. (2018). Long non-coding RNAs in melanoma. Cell Prolif. 51, e12457. doi:10.1111/cpr.12457
Yu, Y., Zhang, W., Chen, Y., Ou, Q., He, Z., Zhang, Y., et al. (2020). Association of long noncoding RNA biomarkers with clinical immune subtype and prediction of immunotherapy response in patients with cancer. JAMA Netw. Open 3, e202149. doi:10.1001/jamanetworkopen.2020.2149
Zager, J. S., Gastman, B. R., Leachman, S., Gonzalez, R. C., Fleming, M. D., Ferris, L. K., et al. (2018). Performance of a prognostic 31-gene expression profile in an independent cohort of 523 cutaneous melanoma patients. BMC Cancer 18, 130. doi:10.1186/s12885-018-4016-3
Zhang, P., Tan, X., Zhang, D., Gong, Q., and Zhang, X. (2021). Development and validation of a set of novel and robust 4-lncRNA-based nomogram predicting prostate cancer survival by bioinformatics analysis. PLOS ONE 16, e0249951. doi:10.1371/journal.pone.0249951
Keywords: immunotherapy, checkpoint inhibition, predictive biomarker, prognostic biomarker, melanoma
Citation: Adeuyan O, Gordon ER, Kenchappa D, Bracero Y, Singh A, Espinoza G, Geskin LJ and Saenger YM (2023) An update on methods for detection of prognostic and predictive biomarkers in melanoma. Front. Cell Dev. Biol. 11:1290696. doi: 10.3389/fcell.2023.1290696
Received: 07 September 2023; Accepted: 04 October 2023;
Published: 13 October 2023.
Edited by:
Donghua Zou, The Second Affiliated Hospital of Guangxi Medical University, ChinaReviewed by:
Himasha Perera, Novartis Institutes for BioMedical Research, United StatesThomas Kleen, Immodulon Therapeutics Ltd., United Kingdom
Copyright © 2023 Adeuyan, Gordon, Kenchappa, Bracero, Singh, Espinoza, Geskin and Saenger. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Oluwaseyi Adeuyan, ooa2134@cumc.columbia.edu; Yvonne M. Saenger, yvonne.saenger@einsteinmed.edu