The Role of Systemic Therapy in Melanoma Brain Metastases

A Narrative Review

Kainat Saleem; Diwakar Davar


Chin Clin Oncol. 2022;11(3):24 

In This Article


Epidemiology and Etiopathogenesis

Cutaneous melanoma is the third most common cause of metastases to the brain, exceeded only by lung cancer and breast cancer.[7] Brain metastases can be seen in patients with non-cutaneous melanoma including mucosal and uveal variants.[8,9] However, immunogenomic characterization and translational clinical trial efforts have focused upon cutaneous MBM; hence, we have sought to limit our discussion in this narrative review to MBM arising in the setting of cutaneous melanoma only. Several risk factors are associated with MBM development including male gender, increasing age, primary tumor location (head and neck vs. trunk), primary tumor depth (Breslow depth), ulceration of primary tumor, presence of visceral metastases, elevated serum lactate dehydrogenase (LDH) level, and CCR4 expression.[10–14] CCR4 PKB/Akt leads to activation of the PI3K pathway,[15] and is linked to MBM formation in preclinical models.[10,15]

Melanoma cells primarily spread to the brain hematogenously. Under normal circumstances, the brain is a "sanctuary" site protected from external influences by the BBB that comprises endothelial cells connected by tight junctions and surrounded by a basal lamina composed of pericytes and astrocyte endfeet. The BBB is uniquely equipped with special efflux pumps for drugs and has slow rates of transcytosis.[16] This complex, multi-layer system limits the entry of pathogens, drugs, and toxins into the CNS, although melanoma cells have unique adaptations that facilitate BBB penetrance. Upon arrival at the microcapillaries of the brain, they arrest due to their size and remain quiescent.[17] Tumor cell extravasation is mediated through various mechanisms: (I) physical disruption of the BBB by the development of cytoplasmic processes that push endothelial cells apart, (II) loosening of tight junctions through the release of angiopoietin-2 and expression of other pro-invasive integrins by the melanoma cells, and (III) cleavage of the basement membrane of BBB by various proteases expressed by the tumor cells.[17,18] After extravasation, tumor cells use the blood vessels as tracks for invasion (vessel cooption).[19] This pattern of invasion explains the higher incidence of MBM in regions of the brain with greater blood flow such as frontal lobes. The immune response to tumor invasion in the brain is also complex. Under homeostatic conditions, the brain is an immune-privileged tissue, and the BBB has low expression of adhesion molecules and low leukocyte traffic.[20] However, BBB disruption by tumor cells results in increased penetration of CNS by lymphocytes which may be responsible for the unique microenvironment of MBM.[21–23]

MBM Tumor Microenvironment (TME) and Role of PI3K/AKT and Mitogen-activated Protein Kinase (MAPK) Pathways in MBM

While the TME of MBMs is heterogenous, subgroups differentially enriched in T cells, monocytes and myeloid dendritic cells with commensurately altered survival can be identified.[24] The immune infiltrate of MBMs comprises a mixed population of tumor-infiltrating lymphocytes (TILs), and the TIL density in MBM is the highest of all solid tumor CNS metastases.[25] Paired analyses of intra- and extra-cranial metastases reveals that the TME of MBMs is significantly enriched with macrophages and gamma delta T cells (γδ T cells), although the frequency of NK cells is reportedly lower in MBM compared to matched primary tumors.[23,26]

The genetics of MBM are similar to other systemic metastatic sites of melanoma: activating point mutations in BRAF are found in 42–50% of MBM, while other driver mutations include NRAS (15–28%), and NF1 (22%).[27,28] However, MBM is uniquely characterized by oncogenic activation of the PI3K-AKT pathway,[29,30] typically through loss of function mutations in the tumor suppressor gene PTEN, found in 20–30% of melanoma and typically associated with concurrent activating mutations in BRAF.[31,32] PI3K/AKT pathway activation facilitates MBM development via upregulation of CCR4, heparanase, VEGF, and STAT3,[15,33,34] and PTEN loss is an adverse prognostic factor in MBM.[35]

Recent work has shed light on the role of tumor-intrinsic metabolism on the development and response of MBM to ICIs and targeted therapy. Compared to extracranial metastases and primary melanomas, gene expression and metabolite profiling reveal that MBMs demonstrate increased oxidative phosphorylation (OXPHOS), in turn associated with lower immune infiltrates and reversed, in preclinical studies by addition of direct OXPHOS inhibitor IACS-010759.[24] OXPHOS induction appears to occur secondary to a shift in nutrient dependence from glucose to other amino acids such as glutamine in metastatic cells,[36] leading to an upregulation of de novo amino acid synthesis to overcome the dearth of extracellular amino acids in the brain—therapy facilitating cerebral tropism, a phenomenon that has been recapitulated in brain metastases from multiple tumor types.[37]

Systemic Therapy in MBM

Optimal management of MBM requires a multidisciplinary approach that considers all available treatment modalities that balances the factors that need to be considered in determining a comprehensive treatment approach including: nature of brain disease (number and size of metastases, location of metastases, presence or absence of neurological symptoms and presence or absence of LMD), extracranial disease burden, tumor mutation status (BRAF mutant or wildtype), patient's functional status, and prior therapy for MBM. In addition, the context of MBM development (i.e., MBM development prior to vs. while on systemic therapy) also helps determine the order of local and systemic therapy. There remain clear roles for radiation therapy,[38–40] and surgery[41] in the management of MBM, but these are well-discussed elsewhere. Below, we focus on the systemic therapy options for MBM patients.

Principles of Systemic Therapy

Historically, patients with untreated MBM were excluded from systemic therapy clinical trials due to concerns about CNS drug penetrance, neurotoxicity and poor prognosis.[42,43] In the context of untreated MBM, several case reports unexpectedly suggested that both ICIs and targeted BRAF/MEK had intracranial efficacy. These led to formal phase II studies evaluating BRAF/MEK targeted therapy and ICIs in patients with untreated MBM that are further elaborated upon below. The majority of these studies had fairly stringent criteria, and limited patients based on size of MBM (0.5–3.0 cm), degree of symptoms (asymptomatic or minimally symptomatic), and systemic glucocorticoid use (absent or minimal) due to concerns over rapid disease progression in symptomatic patients,[44] and lack of efficacy with concomitant steroid use particularly in patients treated with ICIs.[45] LMD was and remains a poorly studied population, and these patients continue to be excluded from clinical trials.

The use of ICIs in MBM

Programmed death-1 (PD-1) is a receptor expressed by activated T cells which binds to programmed death ligand 1 (PD-L1, B7-H1),[46,47] and PD-L2 (B7-DC).[48,49] PD-1 negatively regulates T cell function primarily through the engagement of PD-L1, which is expressed by a wide variety of tissues[46–49] and human tumors, including melanoma, either constitutively or after treatment with IFN-γ.[50,51]

Separately, cytotoxic T lymphocyte associated antigen-4 (CTLA-4, CD152) is an activation-induced glycoprotein that belongs to the Immunoglobulin (Ig) superfamily. CTLA-4 is homologous to the T cell co-stimulatory protein CD28; but where CD28 provides the co-stimulatory signal required for antigen-specific T cell activation and expansion after the initial interaction between T cell receptor (TCR) and antigen presenting cells (APCs), CTLA-4 down-regulates T cell responses by acting as a decoy receptor.[52–54] CTLA-4 is constitutively expressed on regulatory T cells (Tregs) while expression on CD8+ T cells occurs rapidly following TCR engagement (signal 1).[55,56] Both CD28 and CTLA-4 have two natural ligands found on APCs: CD80 (B7.1) or CD86 (B7.2),[57–59] although CTLA-4 has higher avidity and affinity for both compared to CD28.[60–62] Because B7.1/B7.2 provide the positive costimulatory signal (signal 2) through CD28 required for TCR activation, competitive inhibition of CD80 and CD86 by CTLA-4 effectively attenuates T cell activation. CD80 and CD86 are primarily expressed at sites of T-cell priming (e.g., secondary lymphoid organs), and to a lesser extent constitutively expressed to varying degrees on antigen-presenting cells (APC) and activated T cells. Hence, CTLA-4 blockade primarily increases T cell priming, and secondarily reduces Treg-mediated suppression of T cell responses.

Ex vivo, PD-1/PD-L1 pathway blockade in combination with prolonged antigen stimulation augments the frequencies of cytokine-producing, proliferating tumor antigen (TA)-specific CD8+ T cells which express markers of T cell activation/exhaustion including PD-1 and CTLA-4.[63,64] Separately, CTLA-4 blockade primarily induces expansion of ICOS+ Th1-like CD4 effector population, and secondarily increases the number of exhausted/activated CD8+ T cells.[65] Conversely, combination CTLA-4/PD-1 dual blockade induces distinct changes with increased expression of terminally differentiated CD8+ T cells expressing Tbet and EOMES in addition to the afore noted findings with either PD-1 or CTLA-4 blockade singly.[66,67]

ICIs targeting the PD-1/PD-L1 and CTLA-4 pathways have transformed the management of melanoma with approvals in the high-risk node negative, node positive, and metastatic settings. In advanced melanoma, blockade of PD-1 singly or in combination with CTLA-4, results in objective response rate (ORR) of 35–41% (pembrolizumab or nivolumab),[68–72] 43% (nivolumab/relatlimab combination), and 55% (nivolumab-1/ipilimumab-3 combination),[73–77] with grade 3–4 immune-related adverse event (irAE) rates of 10%, 19% and 55% respectively. Dropping the dose of ipilimumab (nivolumab-3/ipilimumab-1 combination) is associated with a lower dose of grade 3–4 irAEs but undiminished efficacy.[78]

The first report of clinical efficacy of ICIs in untreated MBMs occurred more than a decade ago in two separate patients with advanced metastatic melanoma.[79,80] In both instances, patients with untreated, progressive MBM responded favorably to ipilimumab with systemic and CNS responses.[79,80] These results were buttressed by data from an Italian Expanded Access Program (EAP) that evaluated ipilimumab in 146 asymptomatic MBM patients of whom 145 were evaluable for response, and reported global ORR of 11% with durable benefit observed in a subset.[81] Interesting developments in these reports that foretold future developments in this arena included the development of edema surround MBM (and consequent need for corticosteroids and anti-epileptic drugs), tumor necrosis, delayed regression of MBM, discordant responses in MBM vs. systemic lesions and yet, the possibility of durable benefit in a small subset of treated patients.[79–81] These observations led to a plethora of trials, summarized in Table 2 and described below.

These results led to the formal evaluation of ipilimumab 10 mg/kg in untreated MBM in a phase II single-arm trial (CA184-042) that included 2 cohorts: Cohort A with 51 asymptomatic MBM patients who not on corticosteroids, and Cohort B with 21 symptomatic MBM patients maintained on a stable corticosteroid dose.[82] The intra-cranial ORR of 16% (Cohort A) and 5% (Cohort B), and median OS of 7.0 months and 3.7 months confirmed the safety and intracranial efficacy of ipilimumab in patients with untreated MBM; but the striking difference in ORR and OS between the two cohorts formed the basis to exclude patients requiring corticosteroids in future MBM trials of ICIs. Given the dose-response relationship observed with ipilimumab at 10 vs. 3 mg/kg, this question was tested in a phase III trial (CA184-169) that included 127 patients with asymptomatic untreated MBM where the results confirmed the superiority of ipilimumab 10 mg/kg (over 3 mg/kg) with significantly greater 1-year progression-free survival (PFS). ORR was similarly greater for ipilimumab 10 mg/kg (over 3 mg/kg), although this difference was not statistically significant.[83] Subset analyses revealed that MBM patients derived durable benefit that held up at later follow-up.[84] Collectively, these studies established a clear paradigm that ICIs were effective in untreated MBM and primed the field for future developments.

The efficacy of anti-PD-1 ICIs in melanoma spurred the evaluation of anti-PD-1 ICI in MBM patients, initially in the setting of a non-randomized phase II trial that studied pembrolizumab in patients with untreated MBM.[85] Driven by the prior experiences with ipilimumab, enrollment was limited by size of intra-parenchymal lesion (5–20 mm) and excluded patients who were either symptomatic and/or needed corticosteroids. The intra-cranial ORR was 22% (4/18) in MBM with concordant brain and systemic responses, and updated reporting demonstrated that responses were durable with all intra-cranial responses ongoing at 24 months—statistics that compared favorably to the single-agent ORRs in either disease.[86]

Prompted by the development of dual PD-1/CTLA-4 blockade in advanced melanoma, the efficacy of the ipilimumab/nivolumab combination in MBM was evaluated in two clinical trials: CheckMate-204 and ABC[44,87–89] (Table 2). CheckMate-204 was a phase II study that evaluated ipilimumab/nivolumab at the approved doses (ipilimumab 1 mg/kg with nivolumab 2 mg/kg every 3 weeks for 4 doses; then nivolumab 2 mg/kg every 2 weeks for 2 years) in two cohorts: Cohort A (101 patients) with asymptomatic brain metastases, and Cohort B (18 patients) with symptomatic disease. In a notable departure from prior studies, the primary endpoint of the trial was an efficacy one—specifically, intracranial clinical benefit rate which was defined as the aggregate sum of patients with complete or partial response, and stable disease lasting at least 6 months. The intracranial clinical benefit rate was greater in cohort A compared to B (57% vs. 22%); and while the rate of grade 3/4 adverse events was comparable in both arms, the rate of CNS adverse events was much higher in cohort B (6.9 vs. 16.7%).[44,87] Unlike the CheckMate-204 study which evaluated only the ipilimumab/nivolumab combination albeit in asymptomatic and symptomatic MBM, the ABC study compared the ipilimumab/nivolumab combination (Cohort A, 36 patients) to nivolumab monotherapy (Cohort B, 27 patients) although the study was not designed for a formal comparison between cohorts. Valuably, the ABC study additionally included a 3rd arm comprising patients with symptomatic MBM, LMD and/or MBM that failed radiation therapy or surgery with a similar primary endpoint (Cohort C, 16 patients).[88,89] The intracranial clinical benefit rate was numerically greater in Cohort A compared to B (46% vs. 20%), clearly demonstrating the superiority of the ipilimumab/nivolumab combination in this patient population, even with extended follow-up.[89] Conversely, the intracranial clinical benefit rate was very low (6%, 1 patient) in Cohort C, underscoring the limitations of systemic immunotherapy even with dual PD-1/CTLA-4 blockade in this highly-refractory patient population. Concordant with CheckMate-204, the grade 3/4 adverse event rate was high with ipilimumab/nivolumab combination (Cohort A—46%, Cohort B—4%, Cohort C—13%). Collectively, both CheckMate-204 and ABC studies clearly established the efficacy of ipilimumab/nivolumab combination in patients with asymptomatic first-line MBM.

Targeted Therapy use in MBM

The commonest oncogenic alterations in melanoma are in BRAF codon V600 (40–50%), NRAS codons 12, 13, or 61 (15–23%), NF1 (24%) and CKIT (3%) Of these, mutations in the RAF-MEK-ERK MAPK in BRAF and NRAS are typically seen in in melanomas without chronic sun damage (CSD) compared with those associated with CSD (BRAF: 60% vs. 6–22%; NRAS: 20–22% vs. 0–15%), while KIT mutations are significantly more common with CSD associated melanomas (15–30% vs. <1%) and in patients of Asian descent.[90–93] Outside of rare instances or in the setting of acquired resistance following exposure to BRAF (or BRAF/MEK) inhibitor therapy, these driver mutations are typically mutually exclusive. BRAF V600E mutations are more common in females, younger patients; while NRAS mutations are commoner in older and male patients.[94,95] Targeted kinase inhibitors of both BRAF (vemurafenib, and dabrafenib) and MEK (cobimetinib, and trametinib) demonstrated high radiographic response rates with improved PFS and OS in BRAF V600 mutant melanoma.[96–98] However, resistance inevitably develops secondary to BRAF inhibitor-induced paradoxical reactivation of MAPK pathway.[99–101] Preclinically, combination BRAF and MEK inhibition (vemurafenib/cobimetinib, dabrafenib/trametinib, and encorafenib/binimetinib) enhances the inhibitory effect on MAPK pathway signaling in BRAF mutant cells and delays emergence of resistance,[99–101] and clinically is associated with even greater benefit (improved ORR, PFS and OS) compared to BRAF inhibitor monotherapy without a substantial increase in toxicity.[102–105]

Similar to the experience with ICIs, the use of BRAF and BRAF/MEK inhibitors in untreated MBM was triggered by several favorable case reports that hinted at intra-cranial efficacy.[106,107] These results led to a surfeit of clinical trials evaluating BRAF inhibitors singly and combination BRAF/MEK inhibitors in MBM, summarized in Table 3.

Single agent BRAF inhibitor therapy was evaluated in 3 trials: 1 each involving vemurafenib (MO25743) and dabrafenib (BREAK-MB) in asymptomatic MBM, and a separate study that evaluated vemurafenib in symptomatic MBM patients.[108–110] Both studies enrolled patients with BRAF V600 codon mutant and single or multiple MBM between 0.5–4 cm into separate cohorts depending on prior MBM-directed therapy, although MO25743 permitted LMD that BREAK-MB excluded. The former studies enrolled MBM patients and had relatively similar enrollment criteria: presence of activating BRAF V600 codon mutant, single or multiple MBM between 0.5–4 cm. In general, the studies revealed that MBM patients exhibited clinically meaningful response rates to single agent BRAF inhibitors that were well tolerated and without significant CNS toxicity, although unlike with ICIs, the intra-cranial and extra-cranial response rates are not commensurate, with the intra-cranial rates typically lower than the extra-cranial rates, although the reason for this is not clear as the aforenoted CNS penetration of BRAF (and MEK) inhibitors and their efficacy in primary brain tumors including pediatric gliomas and glioblastoma multiforme.[111] Of note, single agent vemurafenib demonstrated safety and efficacy in symptomatic MBM, unlike the experience with ICIs as noted earlier.[110]

A separate trial (BUMPER) evaluating the CNS-penetrable pan-PI3K inhibitor buparlisib in untreated MBM demonstrated that buparlisib monotherapy produced no intracranial responses in MBM patients unselected for PI3K pathway activation, although the drug itself was well tolerated.[112] The MEK inhibitor binimetinib showed modest efficacy in a pilot study of 11 patients with NRAS mutated MBM who had failed multiple previous lines of therapy.[113] ORR was 18%, although interestingly, in 2 patients, responses were deeper intracranially than extracranially.

Based on the enhanced anti-tumor activity and improved OS of combined BRAF/MEK inhibition compared to BRAF monotherapy alone, combination BRAF/MEK inhibitor therapy became the approved standard for BRAF V600 mutant melanoma over BRAF inhibition alone and prompted evaluation in MBM. COMBI-BM evaluated the dabrafenib/trametinib combination in 125 MBM patients in 4 cohorts: BRAF V600E mutant asymptomatic, untreated MBM (cohort A); BRAF V600E mutant asymptomatic, previously treated MBM (cohort B); BRAF non-V600 mutant (V600D/K/R mutant) asymptomatic, previously treated/untreated MBM (cohort C); and BRAF non-V600 mutant symptomatic, previously treated/untreated MBM (cohort D).[114] The majority of patients in each cohort had one or two brain metastases and had extracranial metastatic disease at the time of recruitment. The intra-cranial response rates were similar in all cohorts (cohort A 58%; cohort B 56%; cohort C 44%; cohort D 59%), and typically lower than the commensurate extra-cranial response rates (cohort A 55%; cohort B 44%; cohort C 75%; cohort D 41%), underscoring the trend seen with single agent BRAF inhibitor use. Concordantly, the duration of intra-cranial response was typically lower than for extra-cranial response (6.5 vs. 10.2 months), which pointed towards early CNS failure with BRAF/MEK inhibition—another distinction vis ICI use, wherein intra- and extra- cranial responses were concordant, and responses were typically long-lasting.

Combined TT and ICI Therapy in MBM

Preclinical studies demonstrated potential synergies between combined BRAF/MEK inhibition and ICIs for several reasons including:

  1. Increased T cell effector and DC function following combined BRAF/MEK inhibition in vitro in co-culture experiments.[115–117]

  2. Independent beneficial effects upon multiple components of the tumor immune microenvironment, including increased frequencies and enhanced activation status of TIL, and inhibition of suppressive immune cells including tumor-associated macrophages, and myeloid-derived suppressor cells, and consequent synergy with ICI.[118,119]

Based on the above, randomized phase II trials evaluated the addition of anti-PD-1 ICI to BRAF/MEK inhibitor combination in BRAF mutant melanoma and reported higher PFS and greater durations of response with combined immunotherapy and targeted therapy, although notably the toxicity of the triplet regimens was considerable.[120–123] Two separate phase III trials reported modest improvements in PFS with combined immunotherapy and targeted therapy compared with targeted therapy alone, and while this was statistically significant in IMspire150,[124] it was not in COMBI-I,[125] and overall has not resulted in widespread adoption of the atezolizumab vemurafenib/cobimetinib triplet despite regulatory approval. However, untreated MBM patients were excluded from the studies, and the evidence for efficacy in active MBM is limited.

The combination of ipilimumab and vemurafenib was tested in a small phase II study in advanced melanoma, 6 of whom had MBM.[126] Sequential administration of vemurafenib and ipilimumab led to disease control in 7 patients (5 patients with partial response and 2 with stable disease). The median PFS was 8 months, which was an improvement compared to the PFS seen with either drug, however, the combination was particularly toxic leading to discontinuation.

NIBIT-M1 was a phase II study which evaluated the efficacy of ipilimumab in combination with fotemustine, a chemotherapeutic agent with known CNS penetrance.[127,128] This combination was based on the hypothesis that fotemustine would generate TA and consequently TA-specific T cells, a response that could be amplified by ipilimumab. 86 patients with advanced melanoma were enrolled, 20 out of which had brain disease. The intra-cranial RR was 50%, with 25% patients achieving partial response or stable CNS disease, and 25% without detectable disease on scans. Median OS in patients with brain disease was 12.7 months and 2- and 3-year survival rates were 38.9% and 27.8% respectively. Median PFS was 3.4 months in this subset of patients. NIBIT-M2 is an ongoing phase III study (Table 3) which plans to enroll 168 patients with MBM wherein fotemustine with or without ipilimumab will be evaluated against the combination of ipilimumab/nivolumab.

Combined Local and Systemic Therapy

Despite the encouraging results seen in the clinical trials, a significant proportion of MBM patients are either resistant to systemic treatment or acquire resistance to systemic treatment over time.[129] One of the treatment strategies to overcome this obstacle is combining local (neurosurgery vs. radiosurgery) and systemic therapies. Pooled analyses of various studies have shown improved 1-year survival and improved regional control in patients who receive SRS and ICI concurrently compared to SRS alone.[130] When it comes to the synergistic effect seen between BRAF inhibitors and SRS, the radiosensitizing effect of this drug class on melanoma cells plays an important role. This effect, originally seen in preclinical studies,[131] has been validated in the clinical setting by pooled analyses which have shown a significant survival advantage with the combination of BRAFi and SRS compared to SRS alone.[132] Combination therapy, however, carries the potential for increased toxicities such as the increased risk of radiation necrosis with ICI/SRS combination[133] and intracranial hemorrhage with BRAF inhibitor/SRS combination.[132] More data from other retrospective and prospective studies is needed to further characterize adverse events from combined local and systemic therapy (Table 4).

Adoptive Cellular Therapy (ACT) in MBM

ACT is demonstrably effective in highly refractory melanoma with ORR 36–41%.[134,135] Despite early reports of efficacy in MBM,[136] the evaluation of ACT in MBM has been unexpectedly slow due to the ACT-specific limitations, in particular the complexity entailed in generation and administration of autologous TIL which has restricted this therapy to specialized centers with expertise in this area.

A phase II study of ACT from NIH that included MBM reported intracranial responses in 5 out of 15 patients with previously treated CNS disease and 6 out of 18 patients with previously untreated CNS disease.[137] Of note, the PFS was significantly less in patients with CNS disease compared to patients without CNS disease. In a small, randomized phase II study of ACT with or without antigen-loaded dendritic cells (DC) in a mostly ICI-naïve melanoma patient population, a small fraction of patients had durable responses including patients with untreated MBM. Pooled analyses have showed a significant response rate (32–38%) in patients who received TIL as a second line therapy after ICI failure. However, emerging data is showing poorer response in patients who receive TIL after ICI failure compared to ICI-naïve patients.[134]

Sustained response after ACT is dependent on various factors, including higher doses of TIL (≥50 billion), higher absolute number of CD8+ cells in transfused cells, "young" TIL, high dose IL-2 and the in vitro conditions during expansion and priming of T-cells which affect T-cell functionality and longevity.[134,138] These in vitro conditions are a focus of ongoing research and include potassium concentration,[134,139] intracellular concentrations of L-arginine,[140] rate of T-cell metabolic activity during in vitro expansion and priming,[141] and cholesterol metabolism in T-cells.[142] Transduction of T-cells with chimeric antigen receptors (CAR-T) and TCR gene modified T cells are other forms of adoptive cell therapies that are currently under investigation. The one-off nature of ACT and its tolerable toxicities compared to other systemic therapies makes it ACT a promising mode of therapy for MBM, although improving the survival of transduced T cells remains an ongoing challenge. The use of novel IL-2 pathway agonists, and other cytokines such as IL-7, IL-15, IL-18 to expand TIL ex vivo is promising, although it remains unclear if these data can be translated into the MBM setting.