Application of adoptive cell therapy in malignant melanoma

Abstract

Cutaneous melanoma is one of the most aggressive skin cancers originating from skin pigment cells. Patients with advanced melanoma suffer a poor prognosis and generally cannot benefit well from surgical resection and chemo/target therapy due to metastasis and drug resistance. Thus, adoptive cell therapy (ACT), employing immune cells with specific tumor-recognizing receptors, has emerged as a promising therapeutic approach to display on-tumor toxicity. This review discusses the application, efficacy, limitations, as well as future prospects of four commonly utilized approaches -including tumor-infiltrating lymphocytes, chimeric antigen receptor (CAR) T cell, engineered T-cell receptor T cells, and chimeric antigen receptor NK cells- in the context of malignant melanoma.

Background

Melanoma is one of the most aggressive and deadly types of skin cancer. According to global cancer statistics, the incidence of cutaneous melanoma accounts for 1.6% of all kinds and caused more than 60,000 deaths in 2018 [1]. Despite the steady or decreasing trends of the majority of other cancers, the incidence rate of cutaneous melanoma continues to increase [2, 3]. As superficial soft tissue tumor, cutaneous melanoma offers unique advantages for its early diagnosis and detection. Skin biopsy remains the primary method for confirming cutaneous melanoma (CM), while various molecular and imaging techniques are also recognized for diagnosis [4]. However, approximately 3% of melanomas lack a discernible primary site, referred to as melanoma of unknown primary (MUP). This atypical melanoma subtype remains poorly understood biologically, compared to the more established melanoma of known primary (MKP) [5].

Checkpoint inhibitor immunotherapy characterized by anti-PD1 (nivolumab, pembrolizumab, etc.) and anti-CTLA4 (ipilimumab) antibodies has been widely used in advanced melanoma [6,7,8]. Although a combination of nivolumab and ipilimumab has been proven efficient in patients and leads to improved survival [9, 10], the development of required resistance over time remains a major hurdle [11, 12]. Of note, recent decades have seen microRNAs (miRNA) emerge as a novel therapeutic target, with the potential to impact the natural progression of the disease by enhancing sensitivity to immune checkpoint inhibitors [13]. However, low response rates and resistance can still occur, even with combination therapy, due to the loss of SOX10 expression and suppressive tumor microenvironment (TME) [14,15,16]. M2 macrophages, Tregs, myeloid-derived suppressor cells (MDSCs), and other cells are key components of the suppressive tumor microenvironment (TME). In addition to these cellular components, soluble factors—such as IDO (indoleamine 2,3-dioxygenase), IL-10, and TGF-β—are crucial in modulating immune responses. Soluble factors are bioactive peptides that can bind to cellular receptors, influencing immune cell function by either promoting or inhibiting differentiation, proliferation, and cytotoxic activity [17]. Taken together, these factors contribute to the limited efficacy of conventional systemic treatment and the high recurrence rates observed in advanced melanoma patients. Consequently, there is a pressing need for novel therapies to improve survival and reduce mortality in this population.

Adoptive cell therapy (ACT) is a novel immunotherapeutic approach, involving the extraction, modification, and expansion of a patient’s own immune cells to enhance their anti-tumor toxicity, followed by reinfusion into the patient. This therapy aims to overcome immunosuppressive mechanisms such as insufficient T cell generation, inadequate function of these cells, and impaired formation of memory T cells [18, 19]. ACT has undoubtedly revolutionized the management of hematologic malignancies, whereas its usage in solid tumors remains largely unexploited. ACT harnesses the on-target toxicity of T cells and other immune cells to augment their immune response against cancer. It can be classified into three main types, depending on their specificity, clonality, and target recognition: tumor-infiltrating lymphocyte (TIL) therapy, chimeric antigen receptor (CAR) T-cell therapy, and T-cell receptor (TCR)-based therapy [20, 21]. Other forms of ACT include B cells, NK cells, macrophages, and dendritic cells. Melanoma is among the most immunogenic malignancies expressing melanoma-differentiation antigens and neoantigens. Since the immunogenicity of malignant melanoma has been elucidated [22, 23], the application of ACT in advanced melanoma invoked unprecedented research enthusiasm. This article aims to elucidate the mechanism, manufacturing process, clinical application, advantages, and limitations of the four most promising strategies (Fig. 1).

Fig. 1

Common approaches of adoptive cell therapy for advanced melanoma A Tumor-infiltrating cell therapy. B Chimeric antigen receptors (CARs) T cells therapy. C T-cell receptor (TCR)-based therapy. D Chimeric antigen receptors (CARs)-Natural Killer (NK) cells therapy

Tumor-infiltrating lymphocytes (TIL)

In the 1980s, Rosenberg et al. first proposed the potential application of tumor-infiltrating lymphocytes (TILs) in advanced cancer and succeeded in expanding and maintaining TILs in recombinant IL-2 media [24, 25]. By definition, TILs refer to those lymphocytes penetrating into tumor stroma or intraepithelial after being released by bone marrow and other lymphoid organs. Intraepithelial TILs, compared to those TILs in the tumor stroma, are regarded to play a crucial role in controlling tumor growth in almost all solid tumors [26]. It is renowned that CD8+ and CD4+ T cell populations play a critical role in tumor control through recognizing tumor-associated antigen (TAA) or over-expressing their self-antigens to hinder cancer development [27, 28]. These cells promote inflammatory signals as soon as the immune system recognizes the growing tumor in tissue. Activated CD8+ T cells exhibit direct cytotoxicity toward tumor cells, while CD4+ T cells induce remote inflammatory cell death, which indirectly eradicates interferon-unresponsive and MHC-deficient tumors [29]. This highlights the long-known presence of both CD8+ and CD4+ tumor-infiltrating lymphocytes (TILs) in melanoma [30, 31].

As a therapeutic method, TILs undergo a multistep procedure ex vivo to circumvent the immunosuppressive TME and tumor tolerance. The process starts with a tumor biopsy or surgical resection of patients with advanced melanoma. The tissue is subsequently sent to a laboratory to isolate the TILs. The next stage is called expansion, where the TILs are co-cultured with high-dose of recombinant interleukin-2 (rIL-2) media focusing on amplifying their anti-tumor potency [32,33,34]. Interferon-gamma secretion can be used to measure the reactivity of TILs against tumors by enzyme-linked immunosorbent assay (ELISA) [35]. Simultaneously, patients are given non-myeloablative lymphodepletion chemotherapy or radiotherapy prior to reinfusion to patients with high-dose IL-2 therapy as a preparative regimen [36]. Studies have shown that lymphodepletion ahead of T cell transfer enhances persistence compared with those without lymphodepletion. Lymphodepleting chemotherapy with cyclophosphamide plus fludarabine garnered a response rate of 51% [37, 38], while chemotherapy plus total-body irradiation was not efficient in achieving deep bone marrow suppression in a recent research [39]. High-dose IL-2 therapy has been corroborated to benefit metastatic melanoma, leading to sustained partial responses (PR) and complete responses (CR) in patients. Additionally, post-infusion of high-dose IL-2 can promote the survival and expansion of infused T cells [40, 41].

A clinical study of TIL therapy over a 62-month follow-up period reported a 56% (52 in 93) objective response rate, among whom 20 patients underwent complete remission. This landmark study laid the foundation for the use of TILs therapy in melanoma patients [42]. Subsequently, there was an explosion of research focused on the industrialization of this therapy [35]. In 2017, the FDA granted the TIL-based therapy lifleucel (LN-144) as a fast-track designation for the treatment of patients with advanced melanoma. A phase 2, international, single-arm trial of the TIL treatment lifleucel (LN-144) evaluated patients with stage IIIC or IV unresectable or metastatic melanoma [NCT02360579] in 2021. The results posed a durable response with an objective response rate (ORR) of 36% and a disease control rate of 80% [43]. Subsequently, another TILs treatment, ITIL-168, was granted an orphan drug designation by the FDA for the treatment of patients with stage IIB to IV melanoma in 2021. An international, phase 2 clinical trial on ITIL-168 is currently underway in advanced melanoma [NCT05050006]. TIL therapy has shown significant advantages over the immune checkpoint inhibitor. Rohaan et al. assigned patients with unresectable stage IIIC or IV melanoma to either the TIL therapy group or the ipilimumab group. The median progression-free survival was 7.2 months in the TILs group, almost twice the survival of patients in the ipilimumab group. The patients who received TILs therapy yielded a 49% objective response rate, with the figure for whom received ipilimumab therapy being 21% [44]. In addition to the above-mentioned, this review provides a brief summary of clinical trials involving TIL therapy in melanoma over the last decade (Table 1).

Table 1 Clinical trials evaluating TIL therapy in melanoma

In addition, the application and manufacturing of TILs continue to face numerous formidable challenges. Hence, modified strategies for TIL manufacturing are under exploration to improve response rates and accelerate the cell production process [45,46,47,48]. Due to the rarity of T cells in tumor tissue, the selection and isolation of reactive TILs present a significant challenge. In response to this issue, Wang et al. developed a device called magnetic affinity targeting of infiltrating cells (MATIC), which is sandwiched by permanent magnets, balances magnetic forces and fluidic drag forces to sort cells labeled with magnetic nanoparticles conjugated with antibodies for the target markers. TILs sorted by MATIC require only 5 days of expansion to reach the necessary cell numbers for therapeutic use, with a purity of up to 95% and an isolation efficiency of 90% [49]. However, these TILs yield minimal effect in virtue of the tumor microenvironments enriched with myeloid-derived suppressor cells (MDSCs) and T regulatory cells (Tregs), as well as tumor-associated macrophages, which release reactive oxygen species (ROS) amongst other factors, effectively inhibiting NK cell response [30, 31].

One of the pivotal advantages of TIL therapy is that it offers a terminal treatment corroborated to be more potent in anti-tumor toxicity than those derived from peripheral blood lymphocytes [25]. Specifically, using T cells extracted from the patient’s lesions reduces the likelihood of infection or allergic reactions. However, numerous challenges remain in the manufacturing and delivery of TIL therapy. The high cost of the production process, resulting from personalized manufacturing procedures, highly specialized good manufacturing practice (GMP) facilities, and trained staff cannot be overlooked.

Chimeric antigen receptor (CAR) T therapy

CAR-T cells are genetically engineered antigen receptors that combine features of both antibodies and T-cell receptors (TCRs). This enables them to recognize and bind to specific tumor surface antigens, activating T cells to exert anti-tumor effects. The realization of this CAR function does not depend on the binding of TCR to human leukocyte antigen (HLA) molecules; thus, it is not limited to the HLA type [50, 51]. CAR-T cells exhibit anti-tumor activity through various mechanisms, such as releasing granzymes and perforins to directly kill tumor cells or secreting pro-inflammatory cytokines like IL-2 and IFN-γ, which alter the tumor microenvironment and promote apoptosis and necrosis of tumor cells [52].

The effectiveness of CAR-T-cell therapy depends on the design of CAR proteins, which typically consist of three components: an extracellular domain that binds to tumor antigens, a transmembrane domain for structural stability, and an intracellular domain for signal transduction. The development of CAR-T has gone through the inexhaustible efforts and exploration of several generations of scholars and can be roughly divided into five generations based on the number of intracellular domains and the number of transgenes in the CAR gene. The first reports of tumor-targeting CARs demonstrated that an scFv (single chain fragment variable) recognizing antigens such as human epidermal growth factor receptor 2 (HER2) fused to the CD3ζ signaling domain can elicit tumor-specific cytotoxicity [53, 54]. However, T cells expressing these “first-generation” CARs, which only included the CD3ζ chain for T-cell signaling, generally failed to produce strong anti-tumor responses.

In subsequent years, second- and third-generation CARs were developed, incorporating one or two costimulatory domains, respectively. This advancement was based on the understanding that the endogenous TCR requires association with other costimulatory or accessory molecules for effective signaling [55]. Typically derived from CD28 or 4-1BB, these co-stimulatory domains enhanced antitumor cytotoxicity, boosted cytokine production, and improved the proliferation and persistence of CAR-T cells [56, 57]. The fourth-generation CAR is designed based on the second-generation CAR, which can induce or continuously express cytokines or chemokines in CAR-T cells, commonly referred to as “armored” CARs. For instance, Chmielewski et al. [58] developed “T cells redirected for universal cytokine-mediated killing” (TRUCK), a type of armored CAR-T cell engineered to secrete the proinflammatory cytokine IL-12. The fifth generation of CAR-T is designed to simultaneously activate TCR, co-stimulatory domains CD28, and cytokines. This approach seeks to overcome individual limitations, enable large-scale production and treatment, and reduce costs with the ease and accessibility of CRISPR-Cas9-based gene editing [59].

Currently, clinical trials exclusively studying the efficacy of CAR-T therapy in melanoma patients are still quite scarce. However, its application in certain cell and animal experiments has shown promising prospects. For instance, Forsberg et al. used the second-generation CAR-T cells, with tandem CD28 and CD3ζ CAR chains, to target HER-2+ melanoma tumor cells and confirmed their strong scavenging effect on these melanoma cells in vitro and in vivo [60]. Mishra et al. [61] performed cytotoxicity experiments on third-generation CAR-T cells targeting CD126 in vitro to verify their strong killing effect in 624-mel metastatic melanoma cell lines. In addition, this review also summarizes potential targets or biomarkers for CAR-T cell-mediated melanoma cell killing in other preclinical studies, including B7-H3, TYRP1, MET, VEGFR2, CD19, CD70, and gp100 [62,63,64,65,66,67].

T-cell receptor (TCR)-based therapy

Unlike TIL therapy, TCR-T therapy relies on the isolation and genetic modification of peripheral blood T cells via apheresis, followed by ex vivo transduction with a tumor-specific receptor targeting melanoma-associated antigens [68]. TCR-T has elicited intense investigation interest since CAR-T therapy failed to meet the need for most solid tumors [69]. Therefore, TCR-T is considered another promising ACT modality for advanced melanoma, due to its highly immunogenetic next to the TIL therapy.

After leukapheresis of peripheral blood lymphocytes, antigen-specific T cells are isolated and then expanded in vitro in the presence of γ-chain cytokines, including IL-2, IL-7, IL-15, and IL-21 [70, 71]. The prior identification of tumor-associated antigens underlies the foundation of this process. The melanoma-associated antigen recognized by T cells (MART-1) and glycoprotein 100 (gp100) were among the foremost ascertained melanoma differentiation antigens [72, 73]. Thereafter, cancer/testis antigens, including NY-ESO-1 and MAGE family proteins, have been extensively studied on account of their attributes of being expressed primarily in germline cells and rarely in normal somatic cells [68, 74,75,76]. This restricted expression greatly reduces the risk of on-target, off-tumor toxicity. The next stage involves the transduction of gene-encoding TCR. Currently, most gene-encoding TCR are transduced through gamma-retroviral or lentiviral vectors before being introduced into isolated peripheral blood T lymphocytes [77, 78]. However, the optimization of more productive and efficient T-cell transduction methods is actively being explored. For example, the Sleeping Beauty transposon-mediated TCR gene transfer has demonstrated sustained expression and functional activity [79, 80]. Another study demonstrated that knocking out endogenous TCRs using CRISPR/Cas9 technology increased TCR expression, thereby preventing mixed dimer formation that occurs in the original transduction strategy [81,82,83].

TCR-T therapy was initially proven effective for the treatment of metastatic melanoma by Morgan and his group members in 2006, which achieved a 13% (2 out of 15) response rate [78]. This method shows unrivaled potential in patients who cannot benefit from TIL therapy. Johnson et al. conducted another clinical trial using TCR targeting of MART-1 and gp100, where the results indicated an objective response rate (ORR) of 30% (6/20) for patients treated with high-avidity MART-1 and 19% (3/16) for those treated with gp100 TCR, as defined by RECIST criteria [84]. Besides, NY-ESO-1 and MAGE-A1 have been estimated to express in approximately 33% [85] and 35% [86] melanoma regardless of metastatic status, while MAGE-A3 has been detected in 62% of metastatic lesions [87]. In a clinical trial with TCR transduced to target NY-ESO-1, the objective clinical response was observed in five of 11 (45%) patients with metastatic melanoma [88]. Unfortunately, a phase I/II clinical trial centered on TCR targeting MAGE-A3 antigen in patients with advanced melanoma has shown a predisposition to neurotoxicity. While four in seven patients (57%) diagnosed with melanoma obtained a clinical response to a different extent, three of them suffered severe neurological adverse effects, including TIA, seizures, coma, and even mortality [89]. This phenomenon may result from the recognition of the MAGE-A12 protein expressed in a subset of neurons in the human brain, leading to a calamitous immune response to the white matter. In summary, Table 2 highlights some clinical trials of melanoma involving TCR-T therapy and the antigen targets of TCRs.

Table 2 Clinical trials evaluating TCR-T therapy in melanoma

The clinical application of TCR-T therapy still has limitations and obstacles due to significant adverse events and the risk of resistance. On-target off-tumor toxicities are linked to target antigen expression in normal tissues and are mostly associated with TAAs. The assessment of TCR-T cell-related on-target off-tumor toxicity involves bioinformatic analysis of transcriptomic and proteomic databases, immunopeptidomics, and in vitro or ex vivo assays to evaluate the ability of TCR-T cells to recognize normal cells or tissues [90]. As a reaction to off-tumor toxicities and cross-reactivity, groups [91] have developed a mutational positioning scan (X-scan) with a peptide library where each epitope residue is consecutively replaced by all other amino acids. For each positive result, peptide profiles are then compared to the human proteome to identify potential cross-reactive peptides. The resistance to TCR-T therapy has emerged as a big challenge. A recent clinical study using TCR-T cells targeting HPV-16 E7 antigen in HPV-16+ epithelial cancers showed that resistance to T cell therapies involved several actors of the antigen presentation process and of the interferon pathway, with a patient demonstrating loss of TAP1, TAP2, and IFNGR [92].

Although many hurdles remain to be overcome in the clinical application of TCR-T therapy, it is more satisfying in solid tumors than CAR-T therapy [93, 94]. In conclusion, TCR therapy presents a promising field for advanced melanoma treatment. Current clinical trials for TCR-T therapies have not achieved clinical response rates as high as those for TILs therapies, but there is still a need to explore more effective target antigens with higher affinities. With the rapid development of genetic engineering and enhancing technologies, and promise in the field of not only advanced melanoma but also other solid tumors.

Chimeric antigen receptors (CARs)-natural killer (NK) cells therapy

Chimeric antigen receptor T cells (CAR-T) therapy has prompted tremendous clinical success in patients with leukemia and lymphoma [94, 95]. However, its application in solid tumors remains particularly challenging. Multi-mechanisms are attributable to this hurdle, including the heterogeneity of antigens or lack of specific antigens, highly suppressive TME, difficulties penetrating solid tumors’ physical barrier, and the lack of chemokines to attract CAR-T cells [95, 96]. Therefore, the defects of CAR-T in solid tumors have stimulated much effort to explore other types of immune cells, for instance, the replacement of T lymphocytes with NK cells for CAR expression has posted a promising prospect for solid tumors.

Natural killer cells (NK cells), originating from CD34+ hematopoietic progenitors, exert anti-tumor activity independent of TCR-MHC restriction, without the risk of causing graft-versus-host disease (GVHD). Instead, a set of stimulatory and inhibitory receptors are involved in NK cell activation and targeted cell killing as well as cytokine production [97,98,99,100]. One of the key functions of NK cells is to recognize and attack cancer cells or virally infected cells that down-regulate HLA class I molecules or express stress markers [101]. NK cells can synergize with many other immune cells against cancer cells, as shown in Fig. 2. The major benefit of NK cells lies in their intrinsic anti-tumor capability, by inducing cell lysis through the release of perforin and granzyme and the degranulation [97, 102]. The latter acts by inducing apoptosis in cancer cells or virus-infected cells via the expression of FasL (ligand of Fas receptor, or CD95, tumor necrosis factor receptor superfamily member 6) or TRAIL (TNF-related apoptosis-inducing ligand) molecules on their surface. Antibody-dependent cellular cytotoxicity (ADCC) mediated by the Fc receptor CD16 (FcγRIII) also plays an important role in NK-mediated cell lysis. The direct cytotoxicity toward tumor cells, together with the potential for bridging a memory adaptive immune response have rendered NK cells a potential “off-the-shelf” modality for non-hematologic solid tumor cell therapies.

Fig. 2

Tumor killing function of NK cells and cross-talk with other immune cells. Natural killer (NK) cells interact with various immune cells to collectively enhance the immune response against tumors. NK cells can support the anti-tumor activity of T cells in association with TCR-transducted T-cell infusions and tumor-infiltrating lymphocytes, as well as immune checkpoint inhibitors via enhancing the presentation of tumor antigens to T cells. Furthermore, NK cells directly recruit dendritic cells (DCs) to the tumor microenvironment and stimulate their maturation via CC-chemokine ligand 5 (CCL5), XC-chemokine ligand 1 (XCL1), and FMS-related tyrosine kinase 3 ligand (FLT3L) [96]. DCs stimulate NK and T cells via membrane-bound IL-15 (mbIL-15) and 4-1BBL expression. Lysed tumor cells release cancer antigens, which are then presented by DCs, promoting specific T cell proliferation. Finally, NK cells can kill myeloid-derived suppressor cells (MDSCs), which suppress the antitumor T cell

Another superiority of NK cells over T cells resides in their availability, with multiple potential sources where they can be derived. Immortalized human NK cell lines, apheresis of peripheral blood lymphocytes, archived samples of umbilical cord blood, and the supervised differentiation of iPSCs constitute the multiple sources of CAR-NK cell manufacture [103]. On top of optimizing the multiple sources of NK cells, the maintenance of NK cells in circulation became a major concern for further application. Liu et al. have demonstrated that membrane-bound IL-15 and inducible caspase-9-based suicide gene (iC9) expressed on cord blood NK cells via a retroviral vector remarkably augment the tumor-killing capability with prolonged persistence [104]. Furthermore, CAR-engineered and CISH depletion via CRISPR/Cas9 system, leading to enhancement in IL-15 signaling, has substantially improved CAR-NK cell persistence and cytotoxic function [105]. Thereby, a combination of CAR-NK cells secreting IL-15 and knock-out of the cytokine immune checkpoint CISH surprisingly presents as a promising therapeutic strategy to enhance the anti-tumor capability of CAR-NK cells.

CAR-NK therapy has sparked numerous investigations as an emerging field of adoptive cell therapy. Several successful clinical trials have been conducted in hematological malignancies or solid tumors. For instance, Liu et al. designed a phase 1/2 clinical trial using NK cells transduced with anti-CD19 CAR on 11 patients with relapsed or refractory CD19-positive cancers (non-Hodgkin’s lymphoma or chronic lymphocytic leukemia [CLL]). The results showed that eight (73%) patients received a clinical response, seven of whom (4 with lymphoma and 3 with CLL) underwent a complete remission [101]. Current clinical trials for solid tumors aim to establish a better understanding of the efficacy and safety of CAR-NK therapy.

There is a scarcity of clinical evidence centered on the efficacy of CAR-NK therapy in patients with advanced melanoma. Nonetheless, its future use has become progressively brighter as the efficacy of CAR-NK cell therapy was elucidated in other solid tumor clinical trials. In humans, NK cells can be classified into two main subsets on the basis of their expression of CD56 and CD16, including the poorly cytotoxic but highly proliferative, cytokine-producing CD56^brightCD16⁻, and the cytolytic, weakly cytokine-producing CD56^dimCD16+ NK cells [102, 106,107,108]. A cohort of 29 stage III/IV melanoma patients was analyzed to investigate the anti-tumor potential of NK cells in melanoma. It unraveled that CD56^bright NK cells adversely correlated with overall survival and progression-free survival of patients with advanced melanoma [109]. Besides, an abundance of NK cells (CD56^dimCD57^dimCD69 + CCR7 + KIR + NK cells) has been detected in the tumor-infiltrating lymph node of melanoma patients, suggesting the anti-tumor potential of NK cells against melanoma cells deserved further investigation [110, 111].

Conclusions

In conclusion, adoptive cell therapy (ACT) presents as an essential hallmark of melanoma immunotherapy, mainly comprised of TIL, CAR-T, TCR-T, and CAR-T (NK) cell therapy (Table 3). There are, however, loads of limitations that need overcoming. The time and labor-consuming manufacturing procedure and cumbersome logistics add up to financial hardship. The multi-mechanisms of tumor escape as well as the immunosuppressive TME in turn handicap the clinical translational application. Hence, there is a pressing need for improved selection of TIL populations to improve overall responsiveness and mitigate adverse events, such as cytokine release syndrome (CRS) and neurotoxicity of CAR-NK cells and TCR T cell therapies [89, 112, 113]. Advances in cell engineering and gene editing techniques that alter T-cell function—not only to evade resistance mechanisms but also to subdue exhaustion- may amplify the implication of ACTs. Innovations in CAR design, transduction methodologies, and selection of the optimal antigens are bound to lead to improved responses and reach more patients with advanced melanoma.

Table 3 Comparison among ACT modalities