Immunotherapy for pediatric solid tumors: overcoming biological barriers through rational multimodal combinations.

Introduction: the urgent need for novel pediatric solid tumor therapies

Despite decades of progress, childhood solid tumors remain a leading cause of disease-related mortality in high-income settings [ 1 ]. Relapsed or refractory disease continues to carry a poor prognosis: In high-risk neuroblastoma, event-free survival after relapse remains persistently low; in osteosarcoma, 5-year overall survival (OS) after relapse remains below 30%; and recurrent medulloblastoma is associated with limited survival and substantial late toxicity [ 2 – 4 ].

Immune checkpoint blockade has transformed outcomes in several adult cancers, epitomized by durable survival benefits with nivolumab ± ipilimumab in advanced melanoma [ 5 ]. Pediatric oncology has similarly seen success in hematologic malignancies, where anti-CD19 CAR T cell therapy (tisagenlecleucel) achieves high complete remission rates and durable disease control in children with relapsed/refractory B cell acute lymphoblastic leukemia [ 6 ]. These advances provide the rationale to extend immunotherapy to pediatric solid tumors, though distinct tumor biology and host immunity in children necessitate tailored approaches [ 7 ]. (Fig. 1 )

Two key barriers restrict efficacy in pediatric solid tumors. First, most have exceptionally low tumor mutational burdens (TMB) with correspondingly sparse neoantigen landscapes, which reduces spontaneous and therapy-induced T cell priming [ 8 , 9 ]. Second, the tumor microenvironment (TME) is profoundly immunosuppressive, rich in myeloid-derived suppressor cells, tumor-associated macrophages, and stromal elements that concertedly blunt effector cell trafficking, priming, and persistence [ 10 – 12 ]. Recent immune profiling highlights that age, tumor type, and systemic immune tone differentially shape antitumor responses in adults, underscoring the need for pediatric-specific strategies [ 7 ].

This review summarizes recent advances in pediatric solid tumor immunotherapy, emphasizing validated and emerging targets, such as GD2, B7-H3, and GPC2, strategies to remodel the suppressive TME, and novel platforms including next-generation CAR/T cell receptor designs and rational combinations, while integrating early clinical signals and translational insights unique to pediatric tumors.

The unique immune landscape of pediatric solid tumors

Most pediatric solid tumors (PSTs) are immune-cold, characterized by a low tumor mutational burden (TMB), frequent structural and epigenetic drivers rather than point mutations, and infrequent preexisting antitumor T cell activity [ 13 ]. Genomically stable landscapes with lower TMB characterize pediatric tumors compared to adult cancers, typically driven by transcriptional alterations, including chromosomal rearrangements, rather than accumulation of nonsynonymous mutations [ 14 ]. Highly complex structural variation, including chromoplexy, chromothripsis, and extrachromosomal DNA, is common and pathogenic, rationalizing why neoantigen load remains typically low despite aggressive biology [ 14 ].

Genomic characteristics and neoantigen burden across tumor types

The mutational burden and neoantigen landscape vary substantially across pediatric solid tumor histologies, fundamentally shaping their immunotherapeutic responsiveness. Compared with “hot” adult tumors such as melanoma or non-small cell lung cancer, where high TMB predicts superior immune checkpoint inhibitor (ICI) outcomes, PSTs rarely reach TMB-high thresholds [ 15 ].

Neuroblastoma

Neuroblastoma, the most common extracranial solid tumor in children, exhibits an exceptionally low mutational burden with a mean of approximately 3 mutations per tumor (range 0–56). The TMB in neuroblastoma is estimated at 0.24–0.38 mutations per megabase, resulting in a sparse neoantigen landscape with limited immunogenic tumor neoantigens [ 16 ]. Despite this low overall mutational burden, patients with higher mutational burden (>3 mutations) demonstrated significantly worse overall survival, with high TMB, unfavorable histology, and advanced age representing independent prognostic factors [ 16 ]. Even in relapsed neuroblastoma, where RAS-MAPK pathway mutations become more frequent (occurring in ~78% of relapsed tumors vs. lower frequency at diagnosis [ 17 ]), the mutation burden remains relatively low (~2.1 variants/Mb at relapse vs. 0.7 variants/Mb at diagnosis [ 18 ]). This low immunogenicity profile is further characterized by low expression of major histocompatibility complex I (MHC-I), overexpression of CD47, and downregulated antigen processing and presentation pathways [ 19 ].

Osteosarcoma

Osteosarcoma exhibits the highest mutational burden among pediatric solid tumors, with a point mutation burden of approximately 1.5 mutations per megabase [ 20 ]. The median TMB in osteosarcoma is 1.4 per megabase, with no significant differences among adolescent, young adult, and elderly patients [ 21 ]. Integrated immunogenomic analysis revealed that neoantigen expression in osteosarcoma is significantly limited by nonsense-mediated decay (NMD), with samples with low immune infiltrate, characterized by higher numbers of deleted genes, showing elevated levels of adaptive resistance pathways. In contrast, those with high immune infiltrate expressed elevated levels of adaptive resistance pathways [ 22 ]. Even osteosarcomas with TMB exceeding 10 mutations/megabase do not necessarily show improved responses to ICIs, suggesting that additional factors beyond neoantigen load contribute to immune evasion [ 22 ].

Ewing Sarcoma

Ewing sarcoma represents an extreme example of low mutational burden among pediatric sarcomas, with approximately 0.24 mutations per megabase [ 23 ]. Even among osteosarcoma, which has relatively high levels of copy-number variations and structural rearrangements, fewer than 30% of expressed nonsynonymous mutations are predicted to be immunogenic neoantigens (5–30% per patient), and the majority of predicted neoepitopes from genomic rearrangements remain unexpressed due to nonsense-mediated decay, while Ewing sarcoma has an extremely low mutation rate of approximately 0.15 mutations/megabase and generates essentially no neoantigens [ 22 ]. Single-cell RNA sequencing of 18 Ewing sarcoma primary tissue samples confirmed that Ewing sarcoma is characterized by low immunogenicity due to its exceptionally low mutational burden [ 24 ]. However, the mechanism of translocation formation may influence prognosis, with chromoplexy occurring in 42% of cases and associated with higher mutational burden, TP53 mutations, and increased likelihood of relapse (54% vs 30%, p <0.05) [ 25 , 26 ]. The paucity of conventional tumor-specific neoantigens derived from coding mutations has led investigators to explore unconventional sources, including intron retention from alternative splicing, gene fusions creating novel open reading frames, and endogenous retroelements as expanded sources of targetable tumor antigens for cancer immunotherapy [ 27 ].

Rhabdomyosarcoma

Rhabdomyosarcoma, the most frequent pediatric soft tissue sarcoma, is characterized by a low tumor mutational burden (TMB) [ 28 ]. As a pediatric malignancy driven primarily by chromosomal translocations (particularly PAX3/7-FOXO1 fusions) and RAS pathway mutations rather than high mutational burden, rhabdomyosarcoma expresses few tumor-specific neoantigens detectable by the immune system, with fusion-positive tumors averaging only 6.4 nonsynonymous mutations per tumor and fusion-negative tumors averaging 17.8 mutations per tumor, both substantially below the threshold typically associated with immunotherapy responses [ 29 , 30 ].

Medulloblastoma and pediatric brain tumors

Median TMB in pediatric CNS tumors is one mutation per megabase (range 0–132), with higher TMB independently predictive of worse survival outcomes [ 31 ]. Pediatric brain tumors, including medulloblastoma, ependymoma, and H3K27M-mutant gliomas, are characterized by low mutational burden and limited neoantigen presentation [ 32 ]. Despite this low mutational load, tumor-specific neoantigens derived from mutations in genes such as Histidine Ammonia-Lyase (HAL), Neuraminidase 2 (NEU2), and Programmed Cell Death 10 (PDCD10) have been successfully identified and shown to induce de novo cytotoxic T cell responses in medulloblastoma [ 33 ]. Alternative splicing events, which join non-exonic sequences to create aberrant splice junctions, represent an important source of neoantigens in pediatric brain tumors with extremely low mutational burden [ 34 ]. Myeloid and lymphocyte infiltration levels are significantly lower in pediatric glioblastoma and medulloblastoma compared to pilocytic astrocytoma and ependymoma, with medulloblastoma exhibiting low expression of immune markers. Additionally, immune-stromal microenvironmental patterns differ between medulloblastoma molecular subgroups (WNT, SHH, Group 3, Group 4), with SHH tumors showing relatively higher immune cell infiltration [ 11 ].

Mismatch repair-deficient tumors

A key exception to the low TMB paradigm includes pediatric cancers with biallelic mismatch repair deficiency (bMMRD/CMMRD) and other hypermutated tumors, where PD-1 blockade can induce durable responses, as demonstrated in prospective pediatric nivolumab trials including NCT02992964 , the first trial exclusively enrolling patients with high tumor mutational burden (≥5 mutations/megabase) and/or mismatch repair deficiency [ 35 – 37 ]. Genomic predictors of response to PD-1 inhibition in children with germline DNA replication repair deficiency include high TMB, MSI-high status, and specific mutational signatures [ 38 ]. These observations position hypermutation/MSI as rare but actionable pediatric biomarkers for ICI benefit, contrasting the broader adult paradigm in which TMB-high status serves as a pan-tumor predictor [ 35 ].

Tumor microenvironment architecture across pediatric solid tumors

Myeloid-dominated immunosuppression

Infiltrating myeloid cells dominate the microenvironments of many pediatric solid tumors. Myeloid-derived suppressor cells (MDSCs) and M2-polarized tumor-associated macrophages (TAMs) suppress effector T cell activation through multiple mechanisms: MDSCs primarily operate through arginase-1-mediated L-arginine depletion, generation of reactive oxygen and nitrogen species, PD-L1 upregulation, and impairment of dendritic cell antigen presentation, while collectively these populations attenuate cytotoxic T cell responses, promote immune evasion, drive pathological angiogenesis via VEGF and bFGF secretion, and remodel the extracellular matrix through MMP activity and TGF-β signaling [ 12 , 39 ].

In neuroblastoma, myeloid components, including M2-polarized macrophages and tumor-associated neutrophils, are enriched in aggressive disease and associate with poor prognosis and treatment resistance. The stromal compartment, including cancer-associated fibroblasts (CAFs) and mesenchymal stem cells (MSCs), secretes immunosuppressive cytokines such as CXCL12, CSF-1, and CCL2, with distinct CAF subsets (myCAFs, iCAFs, and apCAFs) exerting differential effects on immune suppression, angiogenesis, and tumor invasion [ 40 ].

In osteosarcoma, the TME analysis revealed that patients with non-metastatic disease showed higher immune scores and better prognosis compared to those with metastatic osteosarcoma, with high levels of tumor-infiltrating immune cells potentially suppressing metastasis [ 41 ]. In Ewing sarcoma, single-cell transcriptomic analysis revealed that antigen-presenting cells lack costimulatory gene expression, implying functional impairment in priming effective T cell responses [ 24 ].

T-cell dysfunction and exclusion

Across pediatric and adult sarcomas, macrophage-rich tumor microenvironments correlate with a non-inflamed immunosuppressive phenotype and ICI resistance. Sarcomas are traditionally characterized by low tumor mutational burden, decreased T cell infiltration, and elevated levels of immunosuppressive tumor-associated macrophages (TAMs), highlighting the rationale for macrophage-targeted combination therapies [ 42 , 43 ]. In rhabdomyosarcoma, the tumor microenvironment is characterized as ‘cold’ or non-inflammatory, defined by very few infiltrating T cells and low levels of chemokines/cytokines. While the presence of regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs) has been reported to contribute to the highly immunosuppressive microenvironment that limits immunotherapy efficacy, comprehensive single-cell characterization of these populations in RMS remains limited and warrants further investigation [ 44 ]. In rhabdomyosarcoma, granulocytic myeloid-derived suppressor cells (G-MDSCs) expressing CXCR2 accumulate in the tumor microenvironment and mediate local immunosuppression, thereby limiting ICI efficacy. While these cells utilize suppressive mechanisms characteristic of MDSCs across cancer types, including arginase-1 (ARG1) expression and production of reactive oxygen species (ROS) and nitric oxide (NO), comprehensive characterization of their specific immunosuppressive mechanisms in rhabdomyosarcoma remains limited. Human pediatric sarcoma patients with elevated serum CXCR2 ligands (including CXCL8) have decreased survival, and MDSC accumulation in the tumor bed limits the efficacy of immunotherapy [ 45 ].

Pediatric brain tumors have traditionally been considered immunologically ‘cold,’ characterized by limited T cell infiltration. However, recent single-cell analyses reveal substantial heterogeneity: While many pediatric brain tumors exhibit sparse immune infiltrates, others contain clonally expanded, tumor-reactive T cells. The mechanisms underlying limited T cell infiltration in immunologically cold tumors include low tumor mutational burden (fewer neoantigens), deficiencies in the antigen presentation machinery (including HLA class I and II downregulation), impaired T cell priming and activation, and physical barriers to T cell trafficking into the tumor bed. The immunosuppressive tumor microenvironment features tumor-associated macrophages as the dominant immune cell population, which actively downregulates MHC molecules and creates a feedback loop that further impairs T cell infiltration and function [ 32 ].(Fig. 2 )

Monoclonal antibody (mAb) and antibody-derived platforms

Anti-GD2 antibodies

Anti-GD2 monoclonal antibodies represent the most clinically validated immunotherapy for pediatric solid tumors, having fundamentally transformed outcomes in high-risk neuroblastoma. The target antigen, disialoganglioside GD2, exhibits abundant expression on neuroblastoma cells with restricted normal tissue distribution limited predominantly to peripheral nerve fibers, providing a favorable therapeutic window [ 46 ]. Dinutuximab, a chimeric anti-GD2 monoclonal antibody, received FDA approval in 2015 following the landmark Children's Oncology Group (COG) ANBL0032 trial, which demonstrated significantly improved event-free survival and OS when combined with granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-2, and isotretinoin in the maintenance phase of therapy [ 47 ]. The clinical efficacy of anti-GD2 therapy is mediated through two principal mechanisms: antibody-dependent cellular cytotoxicity (ADCC) via natural killer (NK) cell activation via FcγRIIIA (CD16a) receptors and complement-dependent cytotoxicity (CDC) via C1q-mediated activation of the complement cascade, leading to direct tumor cell lysis. The predominant mechanism is considered to be ADCC, wherein effector cells release perforins and granzymes upon antibody engagement, resulting in targeted neuroblastoma cell death [ 48 ].

Naxitamab, a humanized anti-GD2 antibody with high binding affinity, received accelerated FDA approval in 2020 for relapsed/refractory disease with bone or bone marrow involvement. A recent phase 2 trial ( NCT03363373 ) demonstrated remarkable efficacy in this challenging patient population, achieving an overall response rate (ORR) of 50% with a complete response (CR) rate of 38%, and notably strong responses in both the bone compartment (58%) and the bone marrow compartment (74%) disease [ 46 ]. One-year OS reached an impressive 93%, establishing naxitamab plus GM-CSF as a viable salvage option for patients with osteomedullary disease [ 46 ].

The addition of cytokines, particularly GM-CSF, has proven critical for optimizing the efficacy of anti-GD2 therapies by enhancing ADCC through the expansion and activation of effector cell populations [ 49 ]. Recent comparative analyses have demonstrated that, although dinutuximab was approved with both GM-CSF and IL-2, accumulating evidence suggests that dinutuximab beta is less toxic and achieves similar survival outcomes when administered without IL-2, leading to its removal from contemporary protocols [ 49 ].

Despite these successes, anti-GD2 immunotherapy is associated with substantial treatment-related toxicities that can be dose-limiting. Neuropathic pain represents the most characteristic adverse effect, affecting over 50% of patients with Grade 3 severity, attributed to CDC-mediated complement lysis of GD2-expressing peripheral afferent nerve fibers [ 50 ]. This pain typically manifests during infusion but resolves rapidly post-treatment, with median pain scores returning to zero within 15 minutes of infusion completion in recent trials [ 46 ]. Additional significant adverse events include infusion-related hypotension (58% Grade 3 incidence), bronchospasm, and allergic reactions, necessitating comprehensive supportive care protocols and multidisciplinary team coordination for safe administration [ 46 ]. Uncommon but notable complications include transient mydriasis and loss of accommodation, occurring in up to 34% of patients receiving high-dose therapy [ 51 ], which typically resolve upon treatment cessation. Recent innovations in administration strategies, including prolonged 10-day continuous infusions of dinutuximab beta, have demonstrated reduced pain intensity while maintaining ADCC efficacy and immunogenicity, suggesting that infusion schedule modifications may improve tolerability without compromising therapeutic benefit [ 52 ].

Emerging mAb strategies

Antibody–drug conjugates (ADCs): The antibody–drug conjugate paradigm represents a transformative approach to selective delivery of cytotoxic payloads in pediatric solid tumors, leveraging antibody specificity to achieve targeted tumor destruction while minimizing systemic toxicity. Global ADC development has accelerated dramatically, with 15 ADCs approved for oncologic indications as of 2024 and over 400 candidates in various developmental stages, reflecting the maturation of linker technology, payload optimization, and target selection strategies [ 53 ]. While the majority of approved ADCs target adult malignancies, the validated platform technology, refined through trials in lymphoma, breast cancer, and other adult solid tumors, now provides a robust foundation for pediatric oncology, where tumor-selective antigens such as GD2, B7-H3, and GPC2 offer opportunities to leverage proven ADC architectures against treatment-refractory childhood cancers. The fundamental ADC architecture comprises a tumor-targeting monoclonal antibody conjugated via a cleavable or non-cleavable chemical linker to a highly potent cytotoxic payload, enabling precise intracellular drug delivery following antigen-mediated internalization [ 54 ].

B7-H3 (CD276) has emerged as a particularly attractive target for ADC development in pediatric malignancies due to its high expression across multiple tumor types, including neuroblastoma, osteosarcoma, rhabdomyosarcoma, Ewing sarcoma, and brain tumors, with minimal expression in normal tissues [ 55 ]. Recent preclinical evaluations have demonstrated profound anti-tumor activity of B7-H3-targeted ADCs across pediatric solid tumor models, with the investigational agent vobramitamab duocarmazine showing remarkable efficacy in neuroblastoma xenograft systems [ 56 ]. Multiple B7-H3-directed ADCs have now entered clinical evaluation, with YL201 demonstrating encouraging safety profiles and preliminary clinical efficacy in a large-scale phase 1/1b international trial for advanced solid tumors [ 57 ]. The Pediatric Preclinical Testing Consortium has validated the therapeutic potential of B7-H3 ADCs across diverse histologies, with robust responses observed in osteosarcoma models, supporting continued clinical translation for pediatric applications [ 58 ].

GD2-targeted ADCs represent a logical evolution of anti-GD2 immunotherapy, designed to deliver cytotoxic payloads directly to tumor cells while potentially circumventing the Fc-mediated toxicities associated with conventional antibodies. Novel constructs, including M3554, an anti-GD2 ADC incorporating the topoisomerase I inhibitor exatecan linked via a cleavable β-glucuronide linker to a humanized ch14.18 antibody backbone with engineered Fc mutations to eliminate ADCC and CDC functions, have demonstrated subnanomolar potency and tumor regressions in neuroblastoma, osteosarcoma, and glioma patient-derived xenografts at doses of 3-10 mg/kg [ 59 ]. Smaller format ADC constructs, including minibody and single-chain variable fragment (scFv) derivatives targeting GD2, have shown selective cytotoxic and cytostatic effects specifically in GD2-positive cell lines, demonstrating proof-of-principle for alternative antibody architectures [ 60 ].

Alternative tumor-associated antigens are also being exploited for ADC development in neuroblastoma. GPC2, a cell-surface heparan sulfate proteoglycan highly and differentially expressed in neuroblastoma with limited regular tissue expression, has been successfully targeted with pyrrolobenzodiazepine dimer-based ADCs that induce immunogenic cell death, reprogram the tumor immune microenvironment through upregulation of innate immune activation pathways, and enhance the efficacy of macrophage-stimulating immunoregulators, including anti-CD40 agonists and anti-CD47 antagonists in preclinical models [ 61 ]. These findings suggest that ADCs may function not merely as targeted cytotoxics but also as immunomodulatory agents that can reshape the TME to enhance combinatorial immunotherapeutic strategies.

Bispecific T cell Engagers (BiTEs): Bispecific T cell engagers are an innovative immunotherapeutic modality that simultaneously binds CD3ε on T cells and tumor-associated antigens on malignant cells, thereby inducing T cell activation and target cell lysis, independent of significant histocompatibility complex restriction or T cell receptor specificity. This mechanism enables BiTEs to overcome tumor immune escape mechanisms, including major histocompatibility complex (MHC) downregulation and low-affinity TCR binding that limit conventional T cell recognition [ 62 ]. The archetypal BiTE architecture comprises two single-chain variable fragments (scFvs) connected by a flexible peptide linker, lacking Fc regions to minimize off-target effects and optimize pharmacokinetics. While BiTEs have achieved remarkable success in hematologic malignancies, with blinatumomab approved for B cell acute lymphoblastic leukemia, their efficacy in solid tumors, including pediatric malignancies, has been more limited due to challenges including inadequate T cell infiltration, immunosuppressive TME, antigen heterogeneity, and suboptimal pharmacokinetic properties [ 62 , 63 ].

Nevertheless, preclinical development of BiTEs for neuroblastoma and other pediatric solid tumors is actively progressing. GD2×CD3 bispecific constructs, including the humanized 3F8 bispecific antibody (Hu3F8-BsAb), are under clinical investigation for relapsed/refractory neuroblastoma, osteosarcoma, and other GD2-expressing solid tumors ( NCT03860207 ) [ 64 ]. Innovative next-generation approaches are incorporating BiTEs into combination platforms to overcome barriers in solid tumors. For example, engineered CAR T cells have been designed to simultaneously express a GPC2-directed chimeric antigen receptor while secreting bispecific innate immune cell engagers (BiCEs) targeting GD2 and CD16a, demonstrating that CAR T cell-mediated delivery of GD2-BiCEs enhances GPC2-CAR antitumor efficacy through increased intratumoral NK cell retention and activation in neuroblastoma patient-derived xenograft models with reconstituted human immune cells [ 65 ]. Novel nanotechnology-based approaches utilizing liposomal bispecific T cell engagers functionalized with neuroblastoma-targeting moieties have shown promise in redirecting CAR T cells to tumor sites, suggesting that delivery platform innovations may address penetration limitations [ 66 ]. These evolving strategies highlight the potential for BiTEs to be integrated into multi-modal immunotherapeutic regimens designed to overcome the complex immunosuppressive barriers characteristic of pediatric solid tumors. However, clinical translation remains in early phases and substantial challenges persist regarding efficacy, safety, and optimal combinatorial approaches [ 67 ]. (Table 1 )

Advancements and hurdles in CAR T cell therapy for PSTs

Clinical experience with target antigens

Chimeric antigen receptor (CAR) T cells are genetically engineered lymphocytes that express synthetic receptors comprising an extracellular antigen-recognition domain (typically a single-chain variable fragment), costimulatory signaling domains, and CD3ζ activation machinery, enabling MHC-independent recognition and cytotoxic elimination of tumor cells expressing target surface antigens [ 78 ]. While CAR T cell therapy has achieved remarkable success in hematologic malignancies, particularly with anti-CD19 CARs in B cell leukemias, translating it to solid tumors faces unique challenges, including limited T cell trafficking, immunosuppressive tumor microenvironments, and antigen heterogeneity [ 79 ].

The translation of CAR T cell therapy from hematologic malignancies to pediatric solid tumors has gained considerable momentum through the identification of tumor-associated antigens with favorable expression profiles. Among the most extensively studied targets in clinical trials are GD2, HER2, B7-H3, GPC2, and WT1, each presenting unique opportunities and challenges in the solid TME.

GD2, a disialoganglioside highly expressed on neuroblastoma cells, has emerged as the most clinically advanced target in pediatric solid tumors. Final results from a phase 1/2 trial of third-generation GD2-CART01 cells incorporating CD28 and 4-1BB costimulatory domains, along with an inducible caspase-9 (iCasp9) safety switch, in 35 children with relapsed or refractory high-risk neuroblastoma, demonstrated an ORR of 66% with manageable toxicity [ 68 ]. Importantly, long-term follow-up studies have revealed exceptional durability in select patients, with sustained complete remissions exceeding 18 years in patients treated with first-generation GD2 CARs, highlighting the potential for durable disease control [ 69 ]. The updated final analysis from the Italian cohort confirmed these promising results, reporting grade 3 immune effector cell-associated neurotoxicity syndrome in only four children, which was rapidly controlled through iCasp9 activation with rimiducid, demonstrating the value of incorporating safety switches into CAR designs [ 68 ]. Beyond neuroblastoma, GD2 CAR T cells have shown initial efficacy in diffuse midline gliomas, with combined intravenous and intracranial administration resulting in tumor regression in children with H3K27M+ diffuse midline gliomas, a historically lethal diagnosis [ 80 ].

HER2 has demonstrated clinical activity across multiple pediatric solid tumor histologies. The HEROS 2.0 phase 1 trial evaluated autologous HER2-specific CAR T cells in 13 patients with advanced sarcomas, including osteosarcoma and rhabdomyosarcoma [ 76 ]. Clinical benefit was observed in 50% of treated patients, with one metastatic rhabdomyosarcoma patient achieving a CR sustained for over five years, demonstrating the potential for exceptional responses in carefully selected patients. However, dose-limiting cytokine release syndrome (CRS) emerged at higher dose levels, highlighting the need for continued optimization of dosing strategies to enable safer delivery of therapeutically effective cell doses. A remarkable case of a pediatric patient with refractory bone marrow metastatic rhabdomyosarcoma, a participant in the Phase 1 trial, achieved durable remission following HER2 CAR T cell therapy, with longitudinal immune monitoring revealing remodeling of the T cell receptor repertoire and emergence of endogenous antitumor immunity [ 77 ].

B7-H3 (CD276) represents a desirable target given its broad expression across pediatric solid tumors and limited presence in normal tissues. Multiple early-phase clinical trials have demonstrated the feasibility of B7-H3-directed CAR T cells, with the STRIvE-02 trial representing the first-in-human phase 1 study of systemically administered B7-H3 CAR T cells for relapsed/refractory solid tumors [ 70 ]. B7-H3 is highly expressed in osteosarcoma, rhabdomyosarcoma, Ewing sarcoma, Wilms tumor, and neuroblastoma, making it an ideal pan-tumor target for basket trial designs [ 71 ]. Notably, locoregional delivery approaches have shown particular promise, with repeated intracerebroventricular administration of B7-H3 CAR T cells in children with diffuse intrinsic pontine glioma (DIPG) achieving a median OS of 19.8 months from the time of diagnosis, with three patients surviving over 40 months from diagnosis, a dramatic improvement over historical outcomes [ 72 ].

Glypican-2 (GPC2), an oncofetal antigen, has emerged as a promising neuroblastoma-specific target with exceptional tumor selectivity. Preclinical optimization studies demonstrated that a GPC2 CAR incorporating CD28 hinge, CD28 transmembrane, and 4-1BB costimulatory domains (CT3.28H.BBζ) exhibited superior preclinical efficacy against neuroblastoma patient-derived xenografts [ 73 ]. Iterative engineering of transmembrane and costimulatory domains, combined with c-Jun overexpression, enabled GPC2 CARs to effectively target tumors expressing clinically relevant antigen densities (~5,000 molecules/cell) without toxicity in preclinical models [ 74 ]. These studies highlight the critical interplay between CAR design and antigen density threshold, a consideration particularly relevant for solid tumor targets. GPC2-targeted CAR T cells engineered with nuclear factor of activated T cells (NFAT)-inducible membrane-tethered IL-15/IL-21 exhibited enhanced activity against high tumor burden neuroblastoma with acceptable tolerability in murine models, suggesting pathways to improve efficacy in patients with bulky disease [ 75 ].

Engineering strategies to enhance efficacy in solid tumors

The limited success of conventional CAR T cell approaches in solid tumors has driven intensive investigation into novel engineering strategies that address fundamental barriers, including inadequate tumor infiltration, immunosuppressive microenvironments, and antigen heterogeneity. These next-generation approaches encompass sophisticated CAR designs, cytokine armoring, and multi-targeting strategies that collectively aim to recapitulate the remarkable efficacy observed in hematologic malignancies.

Third-generation CARs incorporating dual costimulatory domains have emerged as a leading strategy to enhance T cell activation and persistence in the hostile solid TME. While second-generation CARs contain either CD28 or 4-1BB costimulatory domains, third-generation constructs combine CD28 with either 4-1BB or OX40 (CD134) to provide synergistic costimulatory signals. Comparative preclinical studies of different costimulatory domain combinations in GD2-directed CARs for neuroblastoma demonstrated that CD28.4-1BB constructs provided superior antitumor efficacy compared to CD28.OX40 configurations, with enhanced cytokine production (significantly higher IFN-γ levels and a less exhausted secretory profile) and more sustained, homogenous in vivo persistence; notably, direct cytolytic activity against neuroblastoma tumor cells was comparable between the two constructs in vitro [ 81 ]. The incorporation of safety switches, particularly inducible caspase-9 (iCasp9), has become standard practice in third-generation designs for pediatric solid tumors. Administration of the chemical inducer of dimerization AP1903 triggers rapid apoptosis of CAR T cells expressing iCasp9, providing a critical safety mechanism to manage on-target, off-tumor toxicities [ 82 ].

Armored CAR T cells, also known as fourth-generation CARs or “TRUCKs” (T cells Redirected for Antigen-Unrestricted Cytokine-initiated Killing), represent an innovative approach wherein CAR T cells are engineered to autonomously secrete immunostimulatory cytokines that reshape the immunosuppressive TME in situ. These fourth-generation CARs incorporate genes encoding cytokines such as IL-12, IL-15, or IL-18, enabling localized immunomodulation. For example, FAP-targeting CAR T cells engineered with a P2A self-cleaving peptide system to constitutively secrete IL-15 have demonstrated enhanced anti-tumor efficacy in preclinical solid tumor models. This approach promotes CAR T cell proliferation, activation, and differentiation toward a T stem cell memory phenotype with superior self-renewal capacity, while simultaneously targeting cancer-associated fibroblasts to disrupt the stromal barrier and reduce tumor angiogenesis, thereby improving therapeutic outcomes while minimizing systemic toxicity [ 83 ]. IL-12-secreting CAR T cells have demonstrated particular promise, with preclinical studies showing enhanced infiltration of endogenous immune cells and polarization of the TME toward an inflammatory phenotype. CAR T cells engineered to secrete bifunctional αPD-L1–IL-12 fusion proteins demonstrated superior safety and efficacy compared to standalone cytokine-armed CARs, achieving localized interferon-gamma (IFN-γ) production and remodeling of the TME without systemic toxicity in prostate and ovarian cancer models [ 84 ]. IL-18 armoring has similarly shown promise, with IL-18-armed CAR T cells inducing NK cell recruitment and M1 macrophage polarization in ovarian and pancreatic cancer models, effectively shifting the immune landscape from immunosuppressive to immunostimulatory [ 85 , 86 ]. Membrane-anchored cytokines offer advantages over secreted forms by enabling targeted activation through cis-presentation while limiting systemic exposure [ 75 ].

Bispecific and tandem CAR strategies have emerged as sophisticated solutions to address tumor antigen heterogeneity and prevent immune escape through antigen loss or downregulation. These approaches can be implemented through multiple configurations: bicistronic vectors encoding two independent CARs, tandem CARs where two single-chain variable fragments are linked within a single CAR molecule, or co-administration of distinct CAR T cell populations. T cells engineered to express a GPC2-directed CAR while simultaneously secreting a bispecific innate immune cell engager (BiCE) targeting GD2 and CD16a achieved enhanced antitumor efficacy in neuroblastoma patient-derived xenograft models through coordinated adaptive and innate immunity [ 65 ]. Dual-target CAR T cells with split costimulatory signaling, in which the CD28 and 4-1BB domains are distributed across two different CARs, can provide both rapid tumor regression and long-term persistence, combining the complementary advantages of each costimulatory domain [ 87 ]. For pediatric B cell acute lymphoblastic leukemia, bicistronic CARs targeting both CD19 and CD22 have shown clinical efficacy, with results from the AUTO3 trial demonstrating that dual targeting can reduce antigen escape-mediated relapse. However, limited CAR T cell persistence remains a challenge [ 88 ]. Comparisons of different tandem CAR configurations noted that loop structures, in which variable light and heavy chain sequences are intercalated, can provide greater structural stability than classical tandem designs [ 89 ]. (Fig. 3 )

Delivery and administration

The route and method of CAR T cell delivery have emerged as critical determinants of therapeutic efficacy and toxicity profiles in solid tumors, with accumulating evidence suggesting that locoregional administration may offer substantial advantages over systemic infusion for specific tumor locations, particularly in the central nervous system. Local delivery methods, including intraventricular, intratumoral, and intracranial administration, ensure higher concentrations of cells at the tumor site while minimizing systemic toxicity [ 90 ].

For brain tumors, intracerebroventricular delivery has demonstrated remarkable advantages by circumventing the blood–brain barrier and enabling sustained exposure of tumor cells to CAR T cells through cerebrospinal fluid circulation. The BrainChild-03 trial showed that repeated intracerebroventricular doses of B7-H3 CAR T cells in children with DIPG were feasible and well-tolerated, with patients receiving multiple infusions over weeks to months without dose-limiting toxicity [ 72 ]. Similarly, GD2 CAR T cells augmented with constitutive IL-7 receptor and delivered via combined intravenous and intracranial routes for high-grade pediatric Central Nervous System (CNS) tumors resulted in detectable CAR T cells within tumor tissue and cerebrospinal fluid, with antitumor activity observed in several patients [ 91 ]. Comparative assessment of intravenous versus intracranial delivery of GD2 CAR T cells for H3K27M+ diffuse midline gliomas demonstrated that both routes achieved tumor responses, though intracranial delivery provided more consistent tumor infiltration [ 80 ].

For extracranial solid tumors, the benefits of locoregional delivery are less uniformly established, with ongoing debate regarding optimal administration strategies. Preclinical studies demonstrated that intraperitoneal delivery of ErbB2 (HER2)-targeted CAR T cells in ovarian cancer models achieved disease remission and improved survival compared to intravenous administration, while also demonstrating a superior safety profile attributable to reduced systemic circulation of CAR T cells and decreased off-tumor tissue exposure [ 92 ]. A notable clinical success with locoregional delivery was observed in hepatocellular carcinoma, where intratumoral injection of GPC3-targeted CAR T cells engineered to secrete IL-7 and C-C motif chemokine ligand 19 (CCL19) led to complete tumor regression within 30 days ( NCT03198546 ). This case demonstrates that for accessible tumors, direct intratumoral injection can achieve high local concentrations of CAR T cells that overcome barriers to tumor infiltration while limiting systemic exposure and associated toxicities [ 93 ].

The choice between systemic and locoregional delivery must consider tumor location, disease burden, and the potential for both local and distant disease control. For disseminated disease or multiple tumor sites, intravenous delivery remains essential to ensure broad biodistribution. However, repeat infusions, whether systemic or locoregional, may be necessary to maintain therapeutic CAR T cell levels and prolong responses, as single infusions often result in limited persistence in the solid tumor context [ 94 ]. An emerging strategy involves combining delivery routes, with initial locoregional delivery to establish a local antitumor response, followed by systemic administration to address micrometastatic disease. However, this approach requires further clinical validation. Ultimately, optimizing delivery strategies requires consideration of tumor-specific factors, CAR T cell trafficking capabilities, and the immunologic landscape of individual patients, with personalized approaches likely to yield the best outcomes in this heterogeneous patient population.

Immune checkpoint blockade: from monotherapy to combination strategies

Limited efficacy of single-agent ICIs

The translation of ICI success from adult oncology to pediatric solid tumors has proven more challenging than anticipated, with single-agent therapies demonstrating disappointing results in unselected cohorts. A multicenter phase I study of sintilimab in pediatric patients with advanced or recurrent malignancies revealed that although the agent was well-tolerated, single-agent efficacy was not evident in most solid tumors, with anti-tumor activity observed primarily in Hodgkin lymphoma rather than the solid tumor cohorts [ 95 ]. This pattern of limited response has been consistently observed across multiple single-agent ICI trials, with neither nivolumab nor pembrolizumab demonstrating significant activity in sporadic pediatric solid tumor histotypes when used as monotherapy [ 96 ]. The disappointing efficacy of single-agent ICIs in pediatric solid tumors is fundamentally attributable to the characteristically low TMB observed in these malignancies, which translates to reduced neoantigen generation and diminished tumor immunogenicity [ 97 ]. Comprehensive genomic profiling of 723 pediatric brain tumors demonstrated that TMB remains low in the vast majority of cases, with 91.8% classified as low TMB and only 2.1% as high TMB; notably, 93% of high TMB tumors harbored a driver mutation, with high TMB cases predominantly comprising high-grade gliomas frequently associated with TP53 alterations and homozygous mismatch repair gene deletions, collectively underscoring the molecular basis for resistance to checkpoint inhibition in this population [ 98 ]. This low mutational landscape represents a critical biological obstacle, as the generation of tumor-specific neoantigens depends on somatic mutations that can be recognized by the immune system following ICI [ 99 ].

Identifying responsive subsets

Despite the generally poor response to ICIs in pediatric solid tumors, molecularly defined subsets with specific genomic characteristics have demonstrated remarkable sensitivity to these agents. The most compelling evidence comes from studies of tumors harboring mismatch repair deficiency or high microsatellite instability, which exhibit a characteristic elevated TMB and robust responses to PD-1 blockade [ 38 ]. Cancers arising from germline DNA mismatch repair deficiency or polymerase proofreading deficiency in children harbor the highest mutational and microsatellite insertion–deletion burden in humans, and treatment with ICIs in an international consortium registry study of 38 patients demonstrated durable objective responses in most patients, culminating in a 3-year survival of 41.4% [ 38 ]. The FDA’s tissue-agnostic approval of pembrolizumab for adult and pediatric patients with unresectable or metastatic microsatellite instability-high or mismatch repair-deficient solid tumors represents a paradigm shift in oncology, marking the first cancer-site agnostic regulatory approval and establishing biomarker-driven therapy as a viable strategy in pediatric oncology [ 100 , 101 ]. Beyond mismatch repair deficiency, high TMB itself has emerged as a predictive biomarker across tumor types, with pembrolizumab receiving FDA approval for tumors with TMB ≥10 mutations per megabase, regardless of tumor origin [ 102 ]. These findings establish that while ICI monotherapy fails in unselected pediatric solid tumor populations, carefully selected molecular subsets defined by hypermutation, mismatch repair deficiency, or elevated TMB can achieve meaningful clinical benefit with single-agent therapy.

The current trend: ICI combination trials

The limited efficacy of ICI monotherapy in pediatric solid tumors has catalyzed an intensive investigation of combination strategies designed to overcome the inherent immunologic barriers of these malignancies. Dual ICI represents the most extensively studied combination approach [ 103 – 105 ]. The Children’s Oncology Group ADVL1412 trial established that nivolumab 3 mg/kg plus ipilimumab 1 mg/kg, administered every 21 days for 4 cycles, followed by nivolumab monotherapy every 14 days, is a tolerable regimen in children and young adults with relapsed or refractory solid tumors. However, clinical activity remained modest, with only two sustained partial responses observed among 41 patients treated at the recommended phase II dose [ 103 ]. Recent regulatory developments have expanded access to dual ICI in specific populations, with the FDA’s April 2025 approval of nivolumab with ipilimumab for adult and pediatric patients 12 years of age and older with unresectable or metastatic microsatellite instability-high or mismatch repair-deficient colorectal cancer, reflecting growing recognition of the role of combination immunotherapy in biomarker-selected cohorts. The combination of ICIs with conventional chemotherapy represents an alternative strategy predicated on the immunomodulatory properties of specific cytotoxic agents, though systematic evaluation in pediatric populations remains limited. Clinical trials are actively investigating nivolumab in combination with cyclophosphamide, vinblastine, and capecitabine in pediatric patients, based on evidence that metronomic chemotherapy may enhance ICI efficacy through immunomodulatory mechanisms, including depletion of immunosuppressive cell populations [ 106 , 107 ]. Combining local therapies with ICI has garnered substantial interest due to the potential to convert tumors into in situ vaccines by releasing tumor-associated antigens and danger-associated molecular patterns. The abscopal effect, whereby local radiation therapy induces regression of distant non-irradiated metastatic lesions through systemic immune activation, remains rare with radiation alone but occurs with increased frequency when radiation is combined with ICIs [ 108 ]. Preclinical investigations in pediatric tumor models have demonstrated that low-dose targeted radionuclide therapy, in a stimulator of interferon genes (STING)-dependent manner, synergizes with ICI therapy to induce CR in immunologically inert tumors, including neuroblastoma, underscoring the importance of immune-activating signals in the TME for successful combination therapy [ 108 ]. Multiple phase I, II, and III studies are actively investigating the safety and efficacy of combined radiation and ICI approaches in pediatric populations, with emerging results suggesting that careful optimization of radiation dose, fractionation, and timing relative to immunotherapy administration will be critical to maximizing abscopal responses [ 109 ]. (Table 2 )

Novel approaches to modulate the TME and innate immunity

The development of strategies to transform immunologically “cold” tumors into “hot,” immune-inflamed phenotypes represents a critical frontier in cancer immunotherapy. Poorly responsive solid tumors often exhibit minimal T cell infiltration, robust immunosuppressive networks, and limited neoantigen expression, creating substantial barriers to the efficacy of conventional ICIs. Emerging therapeutic modalities targeting the TME and innate immune pathways offer promising approaches to overcome these obstacles.

Oncolytic virotherapy (OVs): reprogramming the immunosuppressive microenvironment

Oncolytic viruses have emerged as powerful TME modulators through their dual mechanism of action encompassing direct tumor lysis and induction of immunogenic cell death. These genetically engineered viruses, including modified Herpes Simplex Virus (HSV), Measles Virus (MeV), and adenoviruses, selectively replicate within tumor cells, triggering the release of damage-associated molecular patterns (DAMPs), pathogen-associated molecular patterns (PAMPs), and tumor-associated antigens (TAAs) that activate both innate and adaptive immune responses [ 110 ]. The lytic process converts immunologically cold tumors into inflamed phenotypes characterized by enhanced dendritic cell maturation, increased CD8+ T cell infiltration, and upregulated expression of pro-inflammatory cytokines [ 111 ]. Recent clinical advances have demonstrated that Oncolytic Viruses (OVs) engineered to express immunostimulatory transgenes, such as GM-CSF, interleukin-12 (IL-12), and tumor necrosis factor-alpha (TNF-α), further amplify antitumor immunity by restructuring the immunosuppressive TME [ 112 ]. These transgene-armed platforms are of direct relevance to pediatric solid tumors, whose densely immunosuppressive TMEs, dominated by MDSCs and M2-polarized TAMs, mirror the barriers that OV-expressed IL-12 and GM-CSF are designed to dismantle; indeed, GM-CSF is already clinically validated as an ADCC enhancer in neuroblastoma, lending biological precedent for its incorporation into OV payloads targeting pediatric histologies. Critically, the therapeutic potential of OVs is substantially enhanced when combined with ICIs, as viral infection induces compensatory upregulation of PD-L1 on tumor cells, providing a rationale for combinatorial approaches [ 110 , 113 ]. Phase I and II clinical trials combining oncolytic viruses with pembrolizumab or nivolumab have shown encouraging objective response rates and improved OS in patients with refractory hepatocellular carcinoma and head and neck squamous cell carcinoma, with manageable toxicity profiles [ 114 , 115 ]. These adult trials establish proof-of-concept that oncolytic viruses can safely augment ICI efficacy in immunologically cold tumors, a biological barrier shared by most pediatric solid tumors, supporting the translation of OV-ICI combinations to pediatric populations where conventional immunotherapy has failed. Notably, intratumoral administration of oncolytic Newcastle Disease Virus has been shown to induce systemic antitumor immunity in distant, non-virally injected lesions through mechanisms independent of viral spread, rendering metastatic tumors susceptible to CTLA-4 blockade and achieving long-term survival in preclinical models [ 116 ]. This capacity to induce abscopal immunity is particularly relevant for pediatric solid tumors with metastatic dissemination, where locoregional OV delivery could prime systemic antitumor responses to control distant lesions, a strategy urgently needed for neuroblastoma, rhabdomyosarcoma, and osteosarcoma with refractory metastatic disease.

Targeting innate immunity: NK cell-based therapies

NK cells are emerging as an attractive therapeutic modality for solid tumors because they can recognize and kill tumor cells via mechanisms that do not depend on antigen presentation via MHC-I, thus overcoming key limitations of T cell-based therapies. Clinical data from the meta-analysis show promising objective response rates in solid tumors while maintaining a favorable safety profile, with markedly lower incidence of cytokine-release syndrome (CRS) and negligible graft-versus-host disease (GVHD), supporting the feasibility of ‘off-the-shelf’ allogeneic NK cell products [ 117 ]. This toxicity profile is particularly compelling for children, in whom dose-limiting CRS and neurotoxicity from CAR-T therapy have directly constrained the achievable therapeutic dose; an allogeneic NK platform that maintains cytolytic potency while eliminating these toxicities could enable safe dose escalation to clinically effective levels in pediatric populations. Recent innovations in NK cell engineering have yielded multiple promising platforms, including CAR-NK cells, bispecific NK cell engagers (NKCEs), and trispecific killer cell engagers (TriKEs) that simultaneously target tumor antigens and activating NK receptors while delivering immunostimulatory cytokines [ 118 ]. In patients with relapsed or refractory CD30⁺ Hodgkin and non-Hodgkin lymphomas, bispecific innate cell engagers such as AFM13, which simultaneously target CD30 on tumor cells and CD16A on NK cells, have demonstrated remarkable clinical efficacy when pre-complexed with ex vivo-expanded, cytokine-activated cord blood-derived NK cells, achieving an ORR of approximately 93% and a CR rate of 67% in a heavily pretreated population, with minimal toxicity and no CRS or neurotoxicity [ 119 ]. Although this trial enrolled lymphoma patients, the platform's exceptional safety profile, with no CRS or neurotoxicity even in heavily pretreated patients, positions NK cell engagers as attractive candidates for pediatric solid tumors, where CAR-T toxicity remains dose-limiting. Bispecific NK engagers targeting B7-H3, GD2, or GPC2 could leverage this favorable toxicity profile while addressing the challenge of tumor antigen heterogeneity through "off-the-shelf" allogeneic NK products. The development of B cell maturation antigen (BCMA)-targeted NKCEs for multiple myeloma offers advantages over T cell engagers by circumventing severe CRS and neurotoxicity while maintaining potent cytolytic activity [ 120 ]. This therapeutic window, potent tumor killing without severe toxicity, has direct implications for pediatric solid tumors, where age-related physiologic vulnerabilities necessitate platforms with manageable safety profiles. The NKCE platform validated in myeloma provides a blueprint for engineering pediatric-specific NK engagers against established targets such as GD2, B7-H3, and GPC2, potentially enabling safe dose escalation to therapeutically effective levels in children. Emerging strategies to enhance NK cell function include cytokine armoring with IL-15, combination with ICI targeting T cell immunoreceptor with Ig and ITIM domains (TIGIT) or NK group 2 member A (NKG2A), and engineering of induced pluripotent stem cell (iPSC)-derived universal NK cell products [ 121 ].

Vaccine strategies: priming antitumor immunity in combination regimens

Cancer vaccines, particularly those targeting patient-specific neoantigens, have demonstrated renewed clinical promise when combined with ICI to overcome immune tolerance in PSTs. Personalized neoantigen vaccines leverage whole-exome sequencing and computational algorithms to identify tumor-specific mutations with high MHC-binding affinity, enabling the generation of peptide-based, mRNA-encoded, or dendritic cell-loaded vaccine formulations [ 122 ]. In the phase 2b KEYNOTE-942 trial, individualized neoantigen therapy mRNA-4157 (V940) in combination with pembrolizumab significantly improved recurrence-free survival compared with pembrolizumab alone in patients with resected high-risk melanoma. At a median follow-up of 18 months, the combination reduced the risk of recurrence or death by 44% (hazard ratio 0.56; 95% CI 0.31–1.00; p = 0.049), with 18-month recurrence-free survival rates of 79% for the combination and 62% for pembrolizumab monotherapy [ 123 ]. Importantly, this combinatorial approach has shown efficacy in traditionally immunoresistant tumors, with recent Phase I data in pancreatic ductal adenocarcinoma, a prototypical “cold” tumor, revealing significant delays in recurrence following adjuvant neoantigen-pulsed dendritic cell vaccination [ 124 ]. These adult trials provide critical proof-of-principle that neoantigen vaccines can induce meaningful clinical benefit even in tumors with low baseline immunogenicity, precisely the biological constraint that limits immunotherapy efficacy in pediatric solid tumors. The ability of personalized vaccines to prime tumor-reactive T cell responses in melanoma and pancreatic cancer, both characterized by hostile microenvironments, provides a translational roadmap for pediatric cancers: Whole-exome sequencing can identify rare neoantigens arising from structural variants or fusion proteins (the predominant drivers in pediatric tumors), with mRNA or peptide vaccine formulations tailored to each patient’s mutational landscape. When combined with ICIs to prevent T cell exhaustion, this platform addresses the neoantigen-poor phenotype of neuroblastoma, osteosarcoma, and Ewing sarcoma, offering a rational strategy to convert immunologically inert pediatric tumors into vaccine-responsive diseases. Mechanistically, therapeutic vaccines enhance the pool of tumor-reactive T cells and establish immune memory, while concurrent ICI prevents vaccine-induced T cell exhaustion and maintains effector function [ 125 ]. This mechanism of de novo T cell priming is of particular importance in pediatric solid tumors, where preexisting tumor-reactive T cell responses are characteristically absent due to the near-total absence of conventional neoantigens; in this setting, the vaccine itself, rather than checkpoint release, must serve as the primary driver of antitumor immunity [ 104 ]. Dendritic cell-based vaccines loaded with synthetic long peptides or tumor mRNA demonstrate superior immunogenicity by engaging both MHC class I and class II pathways, thereby activating CD8+ cytotoxic and CD4+ helper T cell responses [ 126 ]. Emerging data suggest that even non-personalized mRNA vaccines, such as SARS-CoV-2 mRNA formulations, can sensitize tumors to ICIs through type I interferon-mediated innate immune activation, expanding the therapeutic window for ICI in patients with low baseline immune infiltration [ 127 ]. The integration of cancer vaccines into multimodal regimens that combine ICIs, adoptive cell therapy, or oncolytic viruses represents a rational strategy to achieve durable antitumor immunity in PSTs, with numerous Phase II and III trials currently underway to explore these synergistic combinations [ 128 ].(Fig. 4 )

Conclusion and future directions

Despite transformative advances in adult oncology and pediatric hematologic malignancies, immunotherapy for pediatric solid tumors faces distinct biological challenges that demand tailored solutions. Two fundamental barriers define this therapeutic landscape: exceptionally low TMB with sparse neoantigen generation, and profoundly immunosuppressive TME dominated by myeloid suppressors, regulatory T cells, and stromal barriers that exclude effector cells and blunt therapeutic responses.

The field has nonetheless achieved meaningful progress across multiple platforms. Anti-GD2 monoclonal antibodies have established a validated paradigm in high-risk neuroblastoma, demonstrating that antibody-mediated cytotoxicity can improve survival when integrated into multimodal regimens. Next-generation antibody platforms, including B7-H3 and GD2-targeted ADCs, and bispecific T cell engagers, are translating preclinical promise into early clinical trials, with engineered payloads and dual-targeting strategies addressing limitations of conventional antibodies. CAR T cell therapy has shown clinical activity against GD2, HER2, and B7-H3 targets, with locoregional delivery and advanced engineering (third-generation CARs, cytokine armoring, bispecific constructs) enabling responses in select patients, though tumor infiltration, persistence, and antigen escape remain rate-limiting. ICIs, while largely ineffective as monotherapy in unselected populations, induce durable responses in molecularly defined subsets, hypermutated tumors, mismatch repair deficiency, and high TMB, establishing biomarker-driven selection as essential for rational ICI deployment.

The future of pediatric solid tumor immunotherapy lies in rational combinations that simultaneously address multiple resistance mechanisms. Emerging data support integrating TME modulators, oncolytic viruses, NK cell engagers, myeloid-reprogramming agents, and TGF-β inhibitors with ICI or cellular therapies to convert cold tumors into immunologically active lesions. Personalized neoantigen vaccines combined with ICIs represent a compelling strategy to expand tumor-reactive T cell repertoires in otherwise non-immunogenic cancers. Critically, comprehensive immune profiling using spatial transcriptomics, multiplex imaging, and longitudinal immunomonitoring will be essential for stratifying patients, predicting responses, and guiding adaptive treatment strategies.

Realizing the full potential of immunotherapy in pediatric solid tumors requires continued investment in pediatric-specific trials, novel target discovery through systematic antigen screening, optimization of delivery routes and dosing schedules, and mechanistic studies elucidating age-dependent immune ontogeny and tumor–host interactions. As platforms mature and combinations are refined through iterative clinical–translational cycles, immunotherapy stands poised to fundamentally alter outcomes for children with these historically refractory malignancies.

Acknowledgments

Not applicable

Author contributions

Mahsa Fatahichegeni: Substantial contributions to conception and design, acquisition of data, and analysis and interpretation of data; drafting the article. Mohammad Amin Ansarian: Substantial contributions to acquisition of data and analysis of data; drafting the article. Mi Xiao: Acquisition of data and analysis of data; revising the article critically for important intellectual content. Wenjuan Gao: Analysis and interpretation of data; revising the article critically for important intellectual content. Ruiming Shi: Substantial contributions to conception and design, and interpretation of data; revising the article critically for important intellectual content; final approval of the version to be published.

Funding

No financial assistance was received in support of the study.

Data availability

The data supporting this study's findings are included within the article and are available upon request from the corresponding author.

Declarations

Conflict of interests

The authors declare no competing interests.

Informed consent

Patient consent was not required.

Footnotes

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Mahsa Fatahichegeni and Mohammad Amin Ansarian have contributed equally to this work and share first authorship.

Contributor Information

Mohammad Amin Ansarian, Email: maansarian01@gmail.com.

Ruiming Shi, Email: shiruiming75@126.com.

References

1. Force LM, Abdollahpour I, Advani SM, Agius D, Ahmadian E, Alahdab F, Alam T, Alebel A, Alipour V, Allen CA (2019) The global burden of childhood and adolescent cancer in 2017: an analysis of the Global Burden of Disease Study 2017. Lancet Oncol 20(9):1211–1225

2. Moreno L, Weston R, Owens C, Valteau-Couanet D, Gambart M, Castel V, Zwaan CM, Nysom K, Gerber N, Castellano A (2024) Bevacizumab, irinotecan, or topotecan added to temozolomide for children with relapsed and refractory neuroblastoma: results of the ITCC-SIOPEN BEACON-neuroblastoma trial. J Clin Oncol 42(10):1135–1145

3. Rubio-San-Simón A, Wilson W, Sironi G, Le Deley MC, Palmerini E, Gaspar N, Meazza C (2025) Prognostic factors in patients with relapsed high-grade osteosarcoma: a systematic review: Clinical Study. Br j cancer 133(7):1020–1028

4. Peyrl A, Chocholous M, Sabel M, Lassaletta A, Sterba J, Leblond P, Nysom K, Torsvik I, Chi SN, Perwein T (2023) Sustained survival benefit in recurrent medulloblastoma by a metronomic antiangiogenic regimen: a nonrandomized controlled trial. JAMA Oncol 9(12):1688–1695

5. Wolchok JD, Chiarion-Sileni V, Rutkowski P, Cowey CL, Schadendorf D, Wagstaff J, Queirolo P, Dummer R, Butler MO, Hill AG (2024) Final, 10-year outcomes with nivolumab plus ipilimumab in advanced melanoma. N Engl J Med 392(1):11

6. Laetsch TW, Maude SL, Rives S, Hiramatsu H, Bittencourt H, Bader P, Baruchel A, Boyer M, De Moerloose B, Qayed M (2023) Three-year update of tisagenlecleucel in pediatric and young adult patients with relapsed/refractory acute lymphoblastic leukemia in the ELIANA trial. J Clin Oncol 41(9):1664–1669

7. Chen Q, Zhao B, Tan Z, Hedberg G, Wang J, Gonzalez L, Brodin P (2025) Systems-level immunomonitoring in children with solid tumors to enable precision medicine. Cell 188(5):1425–1440

8. Forrest SJ, Gupta H, Ward A, Li YY, Doan D, Al-Ibraheemi A, Alexandrescu S, Bandopadhayay P, Shusterman S, Mullen EA (2024) Molecular profiling of 888 pediatric tumors informs future precision trials and data-sharing initiatives in pediatric cancer. Nat Commun 15(1):5837

9. Sun S, Liu L, Zhang J, Sun L, Shu W, Yang Z, Yao H, Zhang Z (2025) The role of neoantigens and tumor mutational burden in cancer immunotherapy: advances, mechanisms, and perspectives. J Hematol Oncol 18(1):84

10. Ramirez CF, Taranto D, Ando-Kuri M, de Groot MH, Tsouri E, Huang Z, de Groot D, Kluin RJ, Kloosterman DJ, Verheij J (2024) Cancer cell genetics shaping of the tumor microenvironment reveals myeloid cell-centric exploitable vulnerabilities in hepatocellular carcinoma. Nat Commun 15(1):2581

11. Griesinger AM, Birks DK, Donson AM, Amani V, Hoffman LM, Waziri A, Wang M, Handler MH, Foreman NK (2013) Characterization of distinct immunophenotypes across pediatric brain tumor types. J Immunol 191(9):4880–4888

12. Ishfaq M, Pham T, Beaman C, Tamayo P, Yu AL, Joshi S (2021) BTK inhibition reverses MDSC-mediated immunosuppression and enhances response to anti-PDL1 therapy in neuroblastoma. Cancers 13(4):817

13. Byron SA, Hendricks WP, Nagulapally AB, Kraveka JM, Ferguson WS, Brown VI, Eslin DE, Mitchell D, Cornelius A, Roberts W (2021) Genomic and transcriptomic analysis of relapsed and refractory childhood solid tumors reveals a diverse molecular landscape and mechanisms of immune evasion. Cancer Res 81(23):5818–5832

14. van Belzen IA, van Tuil M, Badloe S, Janse A, Verwiel ET, Santoso M, de Vos S, Baker-Hernandez J, Kerstens HH, Solleveld-Westerink N (2024) Complex structural variation is prevalent and highly pathogenic in pediatric solid tumors. Cell genomics. 10.1016/j.xgen.2024.100675

15. Chen X, Yang W, Roberts CW, Zhang J (2024) Developmental origins shape the paediatric cancer genome. Nat Rev Cancer 24(6):382–398

16. Hwang WL, Wolfson RL, Niemierko A, Marcus KJ, DuBois SG, Haas-Kogan D (2019) Clinical impact of tumor mutational burden in neuroblastoma. JNCI: J Natl Cancer Inst 111(7):695–699

17. Eleveld TF, Oldridge DA, Bernard V, Koster J, Daage LC, Diskin SJ, Schild L, Bentahar NB, Bellini A, Chicard M (2015) Relapsed neuroblastomas show frequent RAS-MAPK pathway mutations. Nat Genet 47(8):864–871

18. Fransson S, Martinez-Monleon A, Johansson M, Sjöberg R-M, Björklund C, Ljungman G, Ek T, Kogner P, Martinsson T (2020) Whole-genome sequencing of recurrent neuroblastoma reveals somatic mutations that affect key players in cancer progression and telomere maintenance. Sci Rep 10(1):22432

19. Persaud NV, Park JA, Cheung NKV (2024) High-risk neuroblastoma challenges and opportunities for antibody-based cellular immunotherapy. J Clin Med 13(16):4765

20. Orrapin S, Moonmuang S, Udomruk S, Yongpitakwattana P, Pruksakorn D, Chaiyawat P (2024) Unlocking the tumor-immune microenvironment in osteosarcoma: insights into the immune landscape and mechanisms. Front Immunol 15:1394284

21. Jiang Y, Wang J, Sun M, Zuo D, Wang H, Shen J, Jiang W, Mu H, Ma X, Yin F (2022) Multi-omics analysis identifies osteosarcoma subtypes with distinct prognosis indicating stratified treatment. Nat Commun 13(1):7207

22. Wu C-C, Beird HC, Andrew Livingston J, Advani S, Mitra A, Cao S, Reuben A, Ingram D, Wang W-L, Ju Z (2020) Immuno-genomic landscape of osteosarcoma. Nat Commun 11(1):1008

23. Tirode F, Surdez D, Ma X, Parker M, Le Deley MC, Bahrami A, Zhang Z, Lapouble E, Grossetête-Lalami S, Rusch M (2014) Genomic landscape of Ewing sarcoma defines an aggressive subtype with co-association of STAG2 and TP53 mutations. Cancer Discov 4(11):1342–1353

24. Visser LL, Bleijs M, Margaritis T, van de Wetering M, Holstege FC, Clevers H (2023) Ewing sarcoma single-cell transcriptome analysis reveals functionally impaired antigen-presenting cells. Cancer Res Commun 3(10):2158–2169

25. Shulman DS, Whittle SB, Surdez D, Bailey KM, de Álava E, Yustein JT, Shlien A, Hayashi M, Bishop AJ, Crompton BD (2022) An international working group consensus report for the prioritization of molecular biomarkers for Ewing sarcoma. NPJ precis oncol. 6(1):65

26. Anderson ND, de Borja R, Young MD, Fuligni F, Rosic A, Roberts ND, Hajjar S, Layeghifard M, Novokmet A, Kowalski PE (2018) Rearrangement bursts generate canonical gene fusions in bone and soft tissue tumors. Science 361(6405):eaam8419

27. Smart AC, Margolis CA, Pimentel H, He MX, Miao D, Adeegbe D, Fugmann T, Wong K-K, Van Allen EM (2018) Intron retention is a source of neoepitopes in cancer. Nat biotechnol. 36(11):1056–1058

28. El Meligy OA, Elemam NM, Hassan WA, Talaat IM (2025) Tumor-immune interactions in pediatric oral rhabdomyosarcoma: a narrative review on immuno-oncology and emerging therapies. Children. 12(9):1249

29. Shern JF, Chen L, Chmielecki J, Wei JS, Patidar R, Rosenberg M, Ambrogio L, Auclair D, Wang J, Song YK (2014) Comprehensive genomic analysis of rhabdomyosarcoma reveals a landscape of alterations affecting a common genetic axis in fusion-positive and fusion-negative tumors. Cancer discov. 4(2):216–231

30. Milewski D, Tian M, Kim Y, Wei J, Khan J (2023) Suppression of antigen presentation is a hallmark of pediatric rhabdomyosarcoma. Cancer Res 83(7_Supplement):6736–6736

31. Parisi R, Patel RR, Rood G, Bowden A, Turco G, Korones DN, Andolina JR, Comito M, Barth M, Weintraub L (2023) Multi-institution analysis of tumor mutational burden and outcomes in pediatric central nervous system tumor patients. Pediatr Blood Cancer. 70(3):e30139

32. Upadhye A, Meza Landeros KE, Ramírez-Suástegui C, Schmiedel BJ, Woo E, Chee SJ, Malicki D, Coufal NG, Gonda D, Levy ML (2024) Intra-tumoral T cells in pediatric brain tumors display clonal expansion and effector properties. Nat cancer. 5(5):791–807

33. Blaeschke F, Paul MC, Schuhmann MU, Rabsteyn A, Schroeder C, Casadei N, Matthes J, Mohr C, Lotfi R, Wagner B (2019) Low mutational load in pediatric medulloblastoma still translates into neoantigens as targets for specific T-cell immunotherapy. Cytotherapy. 21(9):973–986

34. Rivero-Hinojosa S, Grant M, Panigrahi A, Zhang H, Caisova V, Bollard CM, Rood BR (2021) Proteogenomic discovery of neoantigens facilitates personalized multi-antigen targeted T cell immunotherapy for brain tumors. Nat commun. 12(1):6689

35. Das A, Tabori U, Sambira Nahum LC, Collins NB, Deyell R, Dvir R, Faure-Conter C, Hassall TE, Minturn JE, Edwards M (2023) Efficacy of nivolumab in pediatric cancers with high mutation burden and mismatch repair deficiency. Clin Cancer Res. 29(23):4770–4783

36. Bouffet E, Larouche V, Campbell BB, Merico D, De Borja R, Aronson M, Durno C, Krueger J, Cabric V, Ramaswamy V (2016) Immune checkpoint inhibition for hypermutant glioblastoma multiforme resulting from germline biallelic mismatch repair deficiency. J clin oncol. 34(19):2206–2211

37. AlHarbi M, Ali Mobark N, AlMubarak L, Aljelaify R, AlSaeed M, Almutairi A, Alqubaishi F, Hussain ME, Balbaid AAO, Said Marie A (2018) Durable response to nivolumab in a pediatric patient with refractory glioblastoma and constitutional biallelic mismatch repair deficiency. oncologist. 23(12):1401–1406

38. Das A, Sudhaman S, Morgenstern D, Coblentz A, Chung J, Stone SC, Alsafwani N, Liu ZA, Karsaneh OAA, Soleimani S (2022) Genomic predictors of response to PD-1 inhibition in children with germline DNA replication repair deficiency. Nat med. 28(1):125–135

39. Wang H, Zhou F, Qin W, Yang Y, Li X, Liu R (2025) Metabolic regulation of myeloid-derived suppressor cells in tumor immune microenvironment: targets and therapeutic strategies. Theranostics. 15(6):2159

40. Yu W, Biyik-Sit R, Uzun Y, Chen C-H, Thadi A, Sussman JH, Pang M, Wu C-Y, Grossmann LD, Gao P (2025) Longitudinal single-cell multiomic atlas of high-risk neuroblastoma reveals chemotherapy-induced tumor microenvironment rewiring. Nat Genet. 57(5):1142–1154

41. Yu S, Yao X (2024) Advances on immunotherapy for osteosarcoma. Mol cancer. 23(1):192

42. Roland CL, Nassif Haddad EF, Keung EZ, Wang W-L, Lazar AJ, Lin H, Chelvanambi M, Parra ER, Wani K, Guadagnolo BA (2024) A randomized, non-comparative phase 2 study of neoadjuvant immune-checkpoint blockade in retroperitoneal dedifferentiated liposarcoma and extremity/truncal undifferentiated pleomorphic sarcoma. Nat Cancer. 5(4):625–641

43. Petitprez F, De Reyniés A, Keung EZ, Chen TW-W, Sun C-M, Calderaro J, Jeng Y-M, Hsiao L-P, Lacroix L, Bougoüin A (2020) B cells are associated with survival and immunotherapy response in sarcoma. Nature 577(7791):556–560

44. Danielli SG, Wei Y, Dyer MA, Stewart E, Sheppard H, Wachtel M, Schäfer BW, Patel AG, Langenau DM (2024) Single cell transcriptomic profiling identifies tumor-acquired and therapy-resistant cell states in pediatric rhabdomyosarcoma. Nat commun. 15(1):6307

45. Highfill SL, Cui Y, Giles AJ, Smith JP, Zhang H, Morse E, Kaplan RN, Mackall CL (2014) Disruption of CXCR2-mediated MDSC tumor trafficking enhances anti-PD1 efficacy. Sci transl med 6(237):237ra67-237ra67

46. Mora J, Chan GC, Morgenstern DA, Amoroso L, Nysom K, Faber J, Wingerter A, Bear MK, Rubio-San-Simon A, de Las Heras BM (2025) The anti-GD2 monoclonal antibody naxitamab plus GM-CSF for relapsed or refractory high-risk neuroblastoma: a phase 2 clinical trial. Nat Commun. 16(1):1636

47. Yu AL, Gilman AL, Ozkaynak MF, Naranjo A, Diccianni MB, Gan J, Hank JA, Batova A, London WB, Tenney SC (2021) Long-term follow-up of a phase III study of ch14. 18 (dinutuximab)+ cytokine immunotherapy in children with high-risk neuroblastoma: COG study ANBL0032. Clin Cancer Res 27(8):2179–2189

48. Galassi L, Rossi M, Lodeserto P, Lenzi M, Borsetti F, Voltattorni M, Farruggia G, Blasi P, Orienti I (2023) Naxitamab activity in neuroblastoma cells is enhanced by nanofenretinide and nanospermidine. Pharmaceutics. 15(2):648

49. Mora J, Modak S, Kinsey J, Ragsdale CE, Lazarus HM (2024) GM-CSF, G-CSF or no cytokine therapy with anti-GD2 immunotherapy for high-risk neuroblastoma. Int j cancer. 154(8):1340–1364

50. Nysom K, Morad AG, Rafael MS, Zier J, Marachelian A, Watt T, Morgenstern DA (2023) Pain mitigation and management strategies for anti‐GD2 infusions: an expert consensus. Pediatr Blood Cancer 70(5):e30217

51. Brian CT, Navid F, Billups CA, O’Donnell T, Hoehn ME (2015) Ocular abnormalities in patients treated with a novel anti-GD2 monoclonal antibody, hu14. 18K322A. J Am Assoc Pediatr Ophthalmol Strabismus 19(2):112–115

52. Lode HN, Ehlert K, Huber S, Troschke-Meurer S, Siebert N, Zumpe M, Loibner H, Ladenstein R (2023) Long-term, continuous infusion of single-agent dinutuximab beta for relapsed/refractory neuroblastoma: an open-label, single-arm, Phase 2 study. Br J Cancer 129(11):1780–1786

53. Wang R, Hu B, Pan Z, Mo C, Zhao X, Liu G, Hou P, Cui Q, Xu Z, Wang W (2025) Antibody–drug conjugates (ADCs): current and future biopharmaceuticals. J Hematol Oncol 18(1):51

54. Tian H, Pan W, Shi H, Shao X, Sun J, Liang J, Chen B, Cheng B, Liu K, Yao G (2026) Coibamide analogue as a novel-class payload for antibody-drug conjugate. J Med Chem. 10.1021/acs.jmedchem.5c02941

55. Bianchi G, Pastorino F, Rolandi G, Ciampi E, Segalerba D, De Giovanni B, Cafferata B, Balbi M, Ravera S, Vellone VG (2026) The investigational anti-B7-H3 antibody-drug conjugate vobramitamab duocarmazine exerts anti-tumor activity in vitro and in vivo in pediatric sarcoma preclinical models. Cell Death Dis. 10.1038/s41419-025-08397-z

56. Brignole C, Calarco E, Bensa V, Giusto E, Perri P, Ciampi E, Corrias MV, Astigiano S, Cilli M, Loo D (2023) Antitumor activity of the investigational B7-H3 antibody-drug conjugate, vobramitamab duocarmazine, in preclinical models of neuroblastoma. J Immunother Cancer 11(9):e007174

57. Ma Y, Yang Y, Huang Y, Fang W, Xue J, Meng X, Fan Y, Fu S, Wu L, Zheng Y (2025) A B7H3-targeting antibody–drug conjugate in advanced solid tumors: a phase 1/1b trial. Nat med. 10.1038/s41591-025-03600-2

58. Kurmasheva R, Mosse YP, Del Pozo V, Earley EJ, Erickson SW, Groff D, Kolb EA, Krytska K, Smith MA, Tsang M. Testing of B7-H3 targeting antibody-drug conjugate (ADC) MGC018 in models of pediatric solid tumors by the Pediatric Preclinical Testing Consortium (PPTC). Wolters Kluwer Health; 2021.

59. Amendt C, Knuehl C, Crandall T, Sloot W, Toister-Achituv M, Sweeney-Lasch S, Pereira JN, Dotterweich J, Boehm HH, Ma J (2024) Abstract ND08: M3554, a novel anti-GD2 antibody drug conjugate. Cancer Res 84(7_Supplement):ND08–ND08

60. Ma X, Wang M, Ying T, Wu Y (2024) Reforming solid tumor treatment: the emerging potential of smaller format antibody-drug conjugate. Antibody Ther 7(2):114–122

61. Pascual-Pasto G, McIntyre B, Shraim R, Buongervino SN, Erbe AK, Zhelev DV, Sadirova S, Giudice AM, Martinez D, Garcia-Gerique L (2022) GPC2 antibody–drug conjugate reprograms the neuroblastoma immune milieu to enhance macrophage-driven therapies. J Immunother Cancer 10(12):e004704

62. Wang J, Huang X, Shi Q, Swingle KL, Hamilton AG, Gong N, Mitchell MJ (2025) Drug-loaded bispecific T cell nanoengager overcomes T cell exhaustion for potent cancer immunotherapy. Proc Natl Acad Sci U S A 122(45):e2409564122

63. Gupta S, Rau RE, Kairalla JA, Rabin KR, Wang C, Angiolillo AL, Alexander S, Carroll AJ, Conway S, Gore L (2025) Blinatumomab in standard-risk B-cell acute lymphoblastic leukemia in children. N Engl J Med 392(9):875–891

64. Yankelevich M, Thakur A, Modak S, Chu R, Taub J, Martin A, Schalk D, Schienshang A, Whitaker S, Rea K (2024) Targeting refractory/recurrent neuroblastoma and osteosarcoma with anti-CD3× anti-GD2 bispecific antibody armed T cells. J Immunother Cancer 12(3):e008744

65. Pascual-Pasto G, McIntyre B, Hines MG, Giudice AM, Garcia-Gerique L, Hoffmann J, Mishra P, Matlaga S, Lombardi S, Shraim R (2024) CAR T-cell-mediated delivery of bispecific innate immune cell engagers for neuroblastoma. Nat Commun 15(1):7141

66. Parra-Nieto J, Hidalgo L, Márquez-Cantudo M, García-Castro J, Megias D, Ramirez M, Baeza A (2025) Liposomal-based nanoarchitectonics as bispecific T Cell engagers in neuroblastoma therapy. ACS Appl Mater Interfaces 17(8):11937–11945

67. Park AK, Monroy I, Ren Y, Lu C, Chaurasiya S, Valencia H, Lent-Koop J, Cook C, Kang S, Lopez L (2025) Universal off-the-shelf combination immunotherapy using oncolytic viruses to redirect T cell engagers to target solid tumors. J Immunother Cancer 13(8):e011051

68. Locatelli F, Pagliara D, De Ioris MA, Becilli M, Del Baldo G, Serra A, del Bufalo F (2025) GD2-targeting CAR T cells in high-risk neuroblastoma: a phase 1/2 trial. Nat med 31(11):3689–3699

69. Li C-H, Sharma S, Heczey AA, Woods ML, Steffin DH, Louis CU, Grilley BJ, Thakkar SG, Wu M, Wang T (2025) Long-term outcomes of GD2-directed CAR-T cell therapy in patients with neuroblastoma. Nat Med. 10.1038/s41591-025-03513-0

70. Pinto N, Albert CM, Taylor MR, Ullom HB, Wilson AL, Huang W, Wendler J, Pattabhi S, Seidel K, Brown C (2024) STRIvE-02: a first-in-human phase I study of systemically administered B7-H3 chimeric antigen receptor T cells for patients with relapsed/refractory solid tumors. J Clin Oncol 42(35):4163–4172

71. Epperly R, Gottschalk S, DeRenzo C (2024) CAR T cells redirected to B7-H3 for pediatric solid tumors: current status and future perspectives. EJC Paediatr Oncol 3:100160

72. Vitanza NA, Ronsley R, Choe M, Seidel K, Huang W, Rawlings-Rhea SD, Beam M, Steinmetzer L, Wilson AL, Brown C (2025) Intracerebroventricular B7-H3-targeting CAR T cells for diffuse intrinsic pontine glioma: a phase 1 trial. Nat Med 31(3):861–868

73. Sun M, Cao Y, Okada R, Reyes-González JM, Stack HG, Qin H, Li N, Seibert C, Kelly MC, Ruppin E (2023) Preclinical optimization of a GPC2-targeting CAR T-cell therapy for neuroblastoma. J Immunother Cancer 11(1):e005881

74. Heitzeneder S, Bosse KR, Zhu Z, Zhelev D, Majzner RG, Radosevich MT, Dhingra S, Sotillo E, Buongervino S, Pascual-Pasto G (2022) GPC2-CAR T cells tuned for low antigen density mediate potent activity against neuroblastoma without toxicity. Cancer Cell 40(1):53-69.e9

75. Okada R, Reyes-González JM, Rodriguez C, Kondo T, Oh J, Sun M, Nguyen R (2025) GPC2-targeted CAR T cells engineered with NFAT-inducible membrane-tethered IL15/IL21 exhibit enhanced activity against neuroblastoma. Cancer Immunol Res 13(9):1363–1373

76. Hegde M, Navai S, DeRenzo C, Joseph SK, Sanber K, Wu M, Gad AZ, Janeway KA, Campbell M, Mullikin D (2024) Autologous HER2-specific CAR T cells after lymphodepletion for advanced sarcoma: a phase 1 trial. Nat Cancer 5(6):880–894

77. Hegde M, Joseph SK, Pashankar F, DeRenzo C, Sanber K, Navai S, Byrd TT, Hicks J, Xu ML, Gerken C (2020) Tumor response and endogenous immune reactivity after administration of HER2 CAR T cells in a child with metastatic rhabdomyosarcoma. Nat Commun 11(1):3549

78. Feins S, Kong W, Williams EF, Milone MC, Fraietta JA (2019) An introduction to chimeric antigen receptor (CAR) T-cell immunotherapy for human cancer. Am J Hematol 94(S1):S3-s9. 10.1002/ajh.25418.

79. Sterner RC, Sterner RM (2021) CAR-T cell therapy: current limitations and potential strategies. Blood Cancer J 11(4):69. 10.1038/s41408-021-00459-7

80. Monje M, Mahdi J, Majzner R, Yeom KW, Schultz LM, Richards RM, Barsan V, Song K-W, Kamens J, Baggott C (2025) Intravenous and intracranial GD2-CAR T cells for H3K27M+ diffuse midline gliomas. Nature 637(8046):708–715

81. Quintarelli C, Orlando D, Boffa I, Guercio M, Polito VA, Petretto A, Lavarello C, Sinibaldi M, Weber G, Del Bufalo F (2018) Choice of costimulatory domains and of cytokines determines CAR T-cell activity in neuroblastoma. Oncoimmunology 7(6):e1433518

82. Xiao W, Xu L, Wang J, Yu K, Xu B, Que Y, Zhao J, Pan Q, Gao C, Zhou P (2024) FGFR4-specific CAR-T cells with inducible caspase-9 suicide gene as an approach to treat rhabdomyosarcoma. Cancer Gene Ther 31(10):1571–1584

83. Pang Y, Chen L, Sun Q, He N, Lai Z, Wang X, Xie Z, Yang W, Wang W. A new strategy for CAR-T therapy in solid tumors: IL-15-autocrine signaling augments tumor stroma depletion and promotes a TSCM subset in the TME. Cell Death & Disease. 2025.

84. Murad JP, Christian L, Rosa R, Ren Y, Buckley AJ, Lee EHJ, Lopez LS, Park AK, Yang J, Yamaguchi Y (2025) Solid tumour CAR-T cells engineered with fusion proteins targeting PD-L1 for localized IL-12 delivery. Nat Biomed Eng. 10.1038/s41551-025-01509-2

85. Chmielewski M, Abken H (2017) CAR T cells releasing IL-18 convert to T-Bethigh FoxO1low effectors that exhibit augmented activity against advanced solid tumors. Cell Rep 21(11):3205–3219

86. Avanzi MP, Yeku O, Li X, Wijewarnasuriya DP, van Leeuwen DG, Cheung K, Park H, Purdon TJ, Daniyan AF, Spitzer MH (2018) Engineered tumor-targeted T cells mediate enhanced anti-tumor efficacy both directly and through activation of the endogenous immune system. Cell Rep 23(7):2130–2141

87. Ai K, Liu B, Chen X, Huang C, Yang L, Zhang W, Weng J, Du X, Wu K, Lai P (2024) Optimizing CAR-T cell therapy for solid tumors: current challenges and potential strategies. J Hematol Oncol 17(1):105

88. Cordoba S, Onuoha S, Thomas S, Pignataro DS, Hough R, Ghorashian S, Vora A, Bonney D, Veys P, Rao K (2021) CAR T cells with dual targeting of CD19 and CD22 in pediatric and young adult patients with relapsed or refractory B cell acute lymphoblastic leukemia: a phase 1 trial. Nat Med 27(10):1797–1805

89. Qin H, Ramakrishna S, Nguyen S, Fountaine TJ, Ponduri A, Stetler-Stevenson M, Yuan CM, Haso W, Shern JF, Shah NN (2018) Preclinical development of bivalent chimeric antigen receptors targeting both CD19 and CD22. Mol Ther Oncolytics 11:127–137

90. Bagley SJ, Logun M, Fraietta JA, Wang X, Desai AS, Bagley LJ, Nabavizadeh A, Jarocha D, Martins R, Maloney E (2024) Intrathecal bivalent CAR T cells targeting EGFR and IL13Rα2 in recurrent glioblastoma: phase 1 trial interim results. Nat Med 30(5):1320–1329

91. Lin FY, Stuckert A, Tat C, White M, Ruggieri L, Zhang H, Mehta B, Lapteva N, Mei Z, Major A (2024) Phase I trial of GD2. CART cells augmented with constitutive interleukin-7 receptor for treatment of high-grade pediatric CNS tumors. J Clin Oncol 42(23):2769–2779

92. Deshet-Unger N, Horn G, Rawet-Slobodkin M, Waks T, Laskov I, Michaan N, Raz Y, Bar V, Zundelevich A, Aharon S (2022) Comparing intraperitoneal and intravenous personalized ErbB2CAR-T for the treatment of epithelial ovarian cancer. Biomedicines 10(9):2216

93. Pang N, Shi J, Qin L, Chen A, Tang Y, Yang H, Huang Y, Wu Q, Li X, He B (2021) IL-7 and CCL19-secreting CAR-T cell therapy for tumors with positive glypican-3 or mesothelin. J Hematol Oncol 14(1):118

94. Li J, Liu C, Zhang P, Shen L, Qi C (2025) Optimizing CAR T cell therapy for solid tumours: a clinical perspective. Nat Rev Clin Oncol. 10.1038/s41571-025-01075-1

95. Que Y, Wang J, Sun F, Wang S, Zhu J, Huang J, Zhao Z, Zhang L, Liu J, Xu J (2023) Safety and clinical efficacy of sintilimab (anti-PD-1) in pediatric patients with advanced or recurrent malignancies in a phase I study. Signal Transduct Target Ther 8(1):392

96. Davis KL, Fox E, Merchant MS, Reid JM, Kudgus RA, Liu X, Minard CG, Voss S, Berg SL, Weigel BJ (2020) Nivolumab in children and young adults with relapsed or refractory solid tumours or lymphoma (ADVL1412): a multicentre, open-label, single-arm, phase 1–2 trial. Lancet Oncol. 21(4):541–550

97. Bertacca I, Pegoraro F, Tondo A, Favre C (2023) Targeted treatment of solid tumors in pediatric precision oncology. Front oncol. 13:1176790

98. Patel RR, Ramkissoon SH, Ross J, Weintraub L (2020) Tumor mutational burden and driver mutations: Characterizing the genomic landscape of pediatric brain tumors. Pediatr Blood Cancer. 67(7):e28338

99. Tomlins SA, Khazanov NA, Bulen BJ, Hovelson DH, Shreve MJ, Lamb LE, Matrana MR, Burkard ME, Yang ES-H, Edenfield WJ (2023) Development and validation of an integrative pan-solid tumor predictor of PD-1/PD-L1 blockade benefit. Commun med 3(1):14

100. Ciombor KK, Eng C (2023) Immunotherapy in Localized Microsatellite Instability–High/Mismatch Repair Deficient Solid Tumors: Are We Ready for a New Standard of Care? Wolters Kluwer Health, pp 2138–2140

101. Lemery S, Keegan P, Pazdur R (2017) First FDA approval agnostic of cancer site—when a biomarker defines the indication. New Engl J Med. 377(15):1409–1412

102. He Y, Zhang L, Zhou R, Wang Y, Chen H (2022) The role of DNA mismatch repair in immunotherapy of human cancer. Int J Biol Sci. 18(7):2821

103. Davis KL, Fox E, Isikwei E, Reid JM, Liu X, Minard CG, Voss S, Berg SL, Weigel BJ, Mackall CL (2022) A phase I/II trial of nivolumab plus ipilimumab in children and young adults with relapsed/refractory solid tumors: a Children’s Oncology Group Study ADVL1412. Clin Cancer Res. 28(23):5088–5097

104. Long AH, Morgenstern DA, Leruste A, Bourdeaut F, Davis KL. Checkpoint Immunotherapy in Pediatrics: Here, Gone, and Back Again. American Society of Clinical Oncology Educational book. American Society of Clinical Oncology. Annual Meeting; 2022. 1-14 p. (vol. 42).

105. Jaing T-H, Wang Y-L, Chiu C-C (2024) Immune checkpoint inhibitors for pediatric cancers: is it still a stalemate? Pharmaceuticals. 17(8):991

106. André N, Le Deley MC, Léguillette C, Probst A, Willems L, Travers R, Aerts I, Faure-Conter C, Revond-Riviere G, Min V (2024) METRO-PD1: Phase 1 study of nivolumab in combination with metronomic chemotherapy in children and adolescents with relapsing/refractory solid tumors. Eur J Cancer. 198:113525

107. Pasquier E, Kavallaris M, André N (2010) Metronomic chemotherapy: new rationale for new directions. Nat rev Clin oncol. 7(8):455–465

108. Vonderhaar EP, Dwinell MB, Craig BT (2023) Targeted immune activation in pediatric solid tumors: opportunities to complement local control approaches. Front Immunol. 14:1202169

109. Qiu B, Kline C, Mueller S (2021) Radiation in combination with targeted agents and immunotherapies for pediatric central nervous system tumors-progress, opportunities, and challenges. Front Oncol 11:674596

110. Cheng C, Wang Q, Zhang S (2025) Synergy of oncolytic adenovirus and immune checkpoint inhibitors: transforming cancer immunotherapy paradigms. Front Immunol 16:1610858

111. Rivera-Orellana S, Bautista J, Palacios-Zavala D, Ojeda-Mosquera S, Altamirano-Colina A, Alcocer-Veintimilla M, Parrales-Rosales G, Izquierdo-Condoy JS, Vásconez-González J, Ortiz-Prado E (2025) Oncolytic virotherapy and tumor microenvironment modulation. Clin Exp Med 25(1):256

112. Shen Y, Bai X, Zhang Q, Liang X, Jin X, Zhao Z, Song W, Tan Q, Zhao R, Jia W (2025) Oncolytic virus VG161 in refractory Hepatocellular Carcinoma. Nature 641(8062):503–511

113. Schoeps B, Lauer UM, Elbers K (2025) Deciphering permissivity of human tumor ecosystems to oncolytic viruses. Oncogene 44(16):1069–1077

114. Yi L, Ning Z, Xu L, Shen Y, Zhu X, Yu W, Xie J, Meng Z (2024) The combination treatment of oncolytic adenovirus H101 with Nivolumab for refractory advanced Hepatocellular Carcinoma: An open-label, single-arm, pilot study. ESMO Open 9(2):102239

115. Zhang T, Jou TH-T, Hsin J, Wang Z, Huang K, Ye J, Yin H, Xing Y (2023) Talimogene laherparepvec (T-VEC): a review of the recent advances in cancer therapy. J clin med 12(3):1098

116. Zamarin D, Holmgaard RB, Subudhi SK, Park JS, Mansour M, Palese P, Merghoub T, Wolchok JD, Allison JP. Localized oncolytic virotherapy overcomes systemic tumor resistance to immune checkpoint blockade immunotherapy. Science translational medicine. 2014;6(226):226ra32-226ra32.

117. Park H, Kim G, Kim N, Ha S, Yim H (2024) Efficacy and safety of natural killer cell therapy in patients with solid tumors: a systematic review and meta-analysis. Front Immunol 15:1454427

118. Balkhi S, Zuccolotto G, Di Spirito A, Rosato A, Mortara L (2025) CAR-NK cell therapy: promise and challenges in solid tumors. Front Immunol 16:1574742

119. Nieto Y, Banerjee P, Kaur I, Basar R, Li Y, Daher M, Rezvani K (2025) Allogeneic NK cells with a bispecific innate cell engager in refractory relapsed lymphoma: a phase 1 trial. Nat med 31(6):1987–1993

120. Lin H, Case R, Wei KY, Eres I, Alba BM, Kaner J, Wu Y-C, Yamawaki TM, Mostafavi M, Zhou H. (2025) Characterization and comparative analysis of multifunctional natural killer cell engagers during antitumor responses. Cell Reports Medicine. 6(5).

121. Fan Y, Duan Y, Chen J, Wang Y, Shang K, Jiang J, Su L, Zhou C, Sadelain M, Huang H. (2025) Bispecific killer cell engager-secreting CAR-T cells redirect natural killer specificity to enhance antitumour responses. Nature Biomedical Engineering.1-14.

122. Lei W, Zhou K, Lei Y, Li Q, Zhu H (2025) Cancer vaccines: platforms and current progress. Mol Biomed 6(1):3

123. Weber JS, Carlino MS, Khattak A, Meniawy T, Ansstas G, Taylor MH, Kim KB, McKean M, Long GV, Sullivan RJ (2024) Individualised neoantigen therapy mRNA-4157 (V940) plus Pembrolizumab versus Pembrolizumab monotherapy in resected Melanoma (KEYNOTE-942): a randomised, phase 2b study. Lancet 403(10427):632–644

124. Oyama K, Nakata K, Abe T, Hirotaka K, Fujimori N, Kiyotani K, Iwamoto C, Ikenaga N, Morisaki S, Umebayashi M (2025) Neoantigen peptide-pulsed dendritic cell vaccine therapy after surgical treatment of pancreatic cancer: a retrospective study. Front Immunol 16:1571182

125. Zhou Y, Wei Y, Tian X, Wei X (2025) Cancer vaccines: current status and future directions. J Hematol Oncol 18(1):18

126. Morisaki T, Morisaki T, Kubo M, Onishi H, Hirano T, Morisaki S, Eto M, Monji K, Takeuchi A, Nakagawa S (2021) Efficacy of intranodal neoantigen peptide-pulsed dendritic cell vaccine monotherapy in patients with advanced solid tumors: A retrospective analysis. Anticancer Res 41(8):4101–4115

127. Grippin AJ, Marconi C, Copling S, Li N, Braun C, Woody C, Young E, Gupta P, Wang M, Wu A. (2025) SARS-CoV-2 mRNA vaccines sensitize tumours to immune checkpoint blockade. Nature. 1-10.

128. Zhang Y, Li Y, Chen K, Qian L, Wang P (2021) Oncolytic virotherapy reverses the immunosuppressive tumor microenvironment and its potential in combination with immunotherapy. Cancer Cell Int 21(1):262

Associated Data

Data Availability Statement

The data supporting this study's findings are included within the article and are available upon request from the corresponding author.