Multifunctional nanoplatforms deciphering immune resistance in bone tumors: cooperative delivery, immune reprogramming and microenvironment remodeling.

Introduction
Bone tumors comprise a clinically heterogeneous group of neoplasms, spanning from benign lesions to highly aggressive malignancies. While primary malignant bone tumors are relatively rare, they impose a disproportionately heavy burden, particularly because they predominantly affect children and adolescents and are prone to metastasis [1, 2]. Osteosarcoma and Ewing sarcoma are the most common subtypes in younger patients, with incidence peaks that coincide with periods of rapid skeletal growth [1, 3]. Contemporary epidemiological studies continue to highlight demographic disparities, showing higher incidence and mortality rates among males and populations in rural areas [2, 4]. Despite considerable progress in diagnostic imaging and multimodal treatment protocols, survival outcomes for patients with these tumors have plateaued at concerning levels. The five-year survival rate for localized osteosarcoma is approximately 50–60%, and for Ewing sarcoma around 70%; however, these figures drop sharply to below 30% once the cancer has metastasized [1, 5]. The molecular pathogenesis of these tumors involves a complex interplay of genetic factors and microenvironmental cues. Notably, around 28% of osteosarcoma patients carry pathogenic germline variants in cancer susceptibility genes, pointing to underlying defects in DNA repair mechanisms, cell cycle control, and chromatin regulation [3, 6]. This progression is further driven by an immunosuppressive bone microenvironment, marked by dysregulated cytokine signaling, hypoxic niches, and osteolytic destruction, all of which contribute to therapeutic resistance [7]. The current standard of care typically involves surgical resection, supplemented by neoadjuvant and adjuvant chemotherapy or radiotherapy. Yet, these conventional approaches are hampered by significant limitations. Systemic chemotherapy, for instance, often shows poor biodistribution to skeletal sites and is limited by systemic toxicity [8, 9]. Extensive surgical resections, while necessary, often result in compromised musculoskeletal function and carry a substantial risk of infection, despite prolonged courses of antibiotic prophylaxis [10]. Most critically, metastatic and recurrent disease remains largely incurable, as demonstrated by the failure of more intensive treatment regimens to improve survival outcomes in metastatic Ewing sarcoma [11, 12]. The physical and psychosocial impact on patients is severe, with well-documented declines in health-related quality of life that include significant anxiety, chronic pain, and functional limitations, evident from the time of diagnosis [5, 13]. At the heart of this challenge lies the phenomenon of immune resistance, whereby advanced bone tumors evade immune destruction through mechanisms such as upregulating immune checkpoints, recruiting immunosuppressive cells, and altering local immunometabolism within the bone marrow niche [14].
Osteosarcoma is the most frequently diagnosed primary malignant bone tumor in adolescents and young adults, accounting for roughly 56% of all bone sarcomas; the median age at diagnosis is 16 years [15, 16]. This tumor originates from mesenchymal stem cells, typically developing in the metaphyseal regions of long bones, with a predilection for the distal femur, proximal tibia, and proximal humerus [17–19]. Although rare, osteosarcoma is the second leading cause of cancer-related deaths in adolescents, characterized by a tendency for early hematogenous spread, most commonly to the lungs [15, 20, 21]. Detectable metastases are present at diagnosis in 15–20% of patients, and an additional 30–50% will eventually develop recurrent disease, leading to five-year survival rates of only 20–30% in these cases [22, 23]. Current treatment involves neoadjuvant chemotherapy, radical surgical resection, and adjuvant chemotherapy. This approach yields five-year survival rates of up to 75% for patients with localized disease. However, for those with metastatic osteosarcoma, survival rates have seen little improvement over the past forty years [18, 24]. Among the key therapeutic challenges are the intrinsic chemoresistance of osteosarcoma stem cells, the persistence of micrometastases, and the non-specific biodistribution of cytotoxic drugs. Bone metastases, meanwhile, represent a devastating complication of advanced solid tumors. They are diagnosed in over 30% of patients with breast or prostate cancer, and in more than 40% of those with lung cancer [25]. The colonization of tumor cells in bone disrupts the normal balance of bone remodeling through the pathological activation of osteoclasts, frequently resulting in serious skeletal-related events such as pathological fractures, spinal cord compression, and severe bone pain. The dense, mineralized bone matrix presents a major physical barrier to effective drug delivery. Moreover, a dysregulated, bidirectional crosstalk between metastatic cells and the resident bone cells-osteoblasts and osteoclasts-promotes the development of treatment-resistant niches. These parallel challenges-the chemoresistance observed in primary bone tumors and the microenvironment-mediated treatment refractoriness seen in bone metastases-underscore the urgent need for innovative therapeutic strategies grounded in a deep understanding of the underlying biology.
Immunotherapy has undoubtedly transformed the landscape of oncology, yet its success in treating bone tumors has been limited. Clinical trials investigating immune checkpoint inhibitors (ICIs) in osteosarcoma have reported objective response rates below 20%, a stark contrast to the outcomes seen in melanoma or non-small cell lung cancer (NSCLC), for instance [26, 27]. This resistance stems from a profoundly immunosuppressive local microenvironment, which operates through several key mechanisms. These include impaired antigen presentation, a physical barrier that excludes cytotoxic T cells, and the dominance of immunosuppressive networks involving M2-polarized macrophages, regulatory T cells (Tregs), and co-inhibitory ligands such as Programmed death-ligand 1 (PD-L1) and Gal-9 [28–31]. Similarly, the bone metastatic niche utilizes distinct immune evasion tactics. Tumor-associated osteoclasts release factors like TGF-β and IL-6, which directly inhibit cytotoxic T cells while promoting the expansion of myeloid-derived suppressor cells (MDSCs) [32]. Chimeric antigen receptor T-cell therapy faces its own set of obstacles in this setting, including antigenic heterogeneity, inefficient homing to the bone marrow, and the rapid functional exhaustion of T cells within the bone microenvironment, as indicated by upregulation of T-cell immunoglobulin and mucin-domain containing-3 (TIM-3) and lymphocyte activation gene-3 (LAG-3) [33–35]. These formidable biological barriers highlight the critical need for smarter drug delivery systems-strategies capable of not only reaching the tumor site but also actively reversing the local immunosuppressive state.
The emergence of engineered nanomaterials offers a promising new approach to overcoming these barriers in bone tumor immunotherapy. Nanoparticles (NPs) can passively accumulate in tumors through the enhanced permeability and retention (EPR) effect, while surface engineering allows for active targeting through specific ligands [36, 37]. For example, NPs conjugated with bisphosphonates or aspartic acid show a strong, selective binding affinity for hydroxyapatite, achieving up to 30-fold higher accumulation in osteolytic lesions compared to their non-targeted counterparts [38–40]. Furthermore, the development of stimuli-responsive nanocarriers allows for precise control over drug release in response to specific signals within the tumor microenvironment (TME) [41]. A growing body of compelling preclinical data demonstrates that nanocarriers can significantly improve the pharmacokinetic profiles of various immunotherapeutic agents [42]. For instance, Mesoporous silica NPs (MSNs) co-delivering anti- Programmed Cell Death Protein 1 (PD-1) antibodies and doxorubicin (DOX) have demonstrated an 8-fold increase in drug accumulation within the tumor, with sustained release maintained for up to 96 h in osteosarcoma models, leading to a powerful synergistic antitumor effect and up to 80% reduction in tumor burden [43, 44]. Similarly, lipid-polymer hybrid NPs encapsulating IL-15 have been shown to maintain cytokine release for 14 days, thereby enhancing the persistence and killing capacity of CAR-T cells in models of bone metastasis [45]. In addition, NPs functionalized with Toll-like receptor agonists potently stimulate dendritic cell (DC) maturation within the bone marrow, increasing the number of tumor-infiltrating CD8⁺ T cells by as much as 6-fold [46–48]. Importantly, these advanced delivery systems collectively help to reduce off-target toxicity. Studies have shown that nanotherapeutic formulations can reduce drug exposure in the liver by over 90% compared to the free drug administered conventionally [49–51].
The integration of nanotechnology with immunotherapy represents a transformative approach for treating bone malignancies through three principal mechanisms. First, nanocarriers overcome biological delivery barriers by leveraging bone-specific pathophysiological features. Size-optimized particles extravasate through the permeable tumor vasculature, while mineral-affinity ligands facilitate selective accumulation within regions of heightened bone remodeling [52, 53]. Second, multifunctional NPs concurrently dismantle immunosuppressive circuits and potentiate anti-tumor immunity [54]. Co-encapsulation of Colony-stimulating factor 1 receptor (CSF-1R) inhibitors and OX40 agonists reprograms tumor-associated macrophages (TAMs) and activates effector T cells, thereby disrupting the feedforward loop of microenvironmental immunosuppression in osteosarcoma [55–57]. Third, nanotechnology enables temporally controlled combination therapy via engineered release kinetics, achieving rapid burst release of osteoclast inhibitors to restore bone homeostasis, followed by sustained release of immune checkpoint blockers [58]. Emerging theranostic platforms further unite diagnostic and therapeutic capabilities. Iron oxide NPs function dually as contrast agents for Magnetic Resonance Imaging (MRI) and as actuators of macrophage polarization, permitting longitudinal monitoring of treatment response [59, 60]. Biosensing dendrimers continuously quantify bone turnover biomarkers to inform dynamic therapeutic adaptation [61, 62]. For metastatic carcinoma, NPs delivering Receptor activator of nuclear factor kappa-B ligand (RANKL)-silencing RNAs in conjunction with PD-L1 inhibitors abrogate the self-perpetuating cycle of osteolytic destruction and reinstate immune surveillance [63]. Successful clinical translation will require addressing challenges in scalable manufacturing and regulatory approval; nevertheless, biologically informed nano-immunotherapies constitute a paradigm shift in targeting the dynamic interplay between osseous biology and tumor immunology.
This review navigates the landscape of nano-engineered solutions designed to meet three intertwined challenges in bone tumors: the mineralized matrix, the immunosuppressive TME, and the lack of post-treatment healing. We demonstrate that through the deployment of integrated strategies, these nanoplatforms effectively overcome the primary, secondary, and tertiary barriers. It posits that the true innovation lies not in any single function, but in the cooperative integration of delivery, reprogramming, and remodeling. We contend that this holistic approach, which seeks to simultaneously eliminate the tumor and rebuild the bone, represents a fundamental advance in the therapeutic paradigm.

Role of nanomaterials in reshaping the immunological landscape of bone malignancies

Immunosuppressive dynamics in osteosarcoma microenvironment
The osteosarcoma TME is characterized by profound immunosuppression orchestrated through intricate cellular and molecular networks [64, 65]. TAMs, constituting approximately 50% of the immune infiltrate, predominantly exhibit M2 polarization, secreting interleukin-10 (IL-10), TGF-β, and vascular endothelial growth factor (VEGF) to promote angiogenesis, extracellular matrix (ECM) remodeling, and T-cell suppression [32, 66–68]. MDSCs further attenuate antitumor immunity through arginase-1 (Arg-1) and inducible nitric oxide synthase (iNOS), depleting essential T-cell metabolites while generating reactive oxygen species (ROS) [69, 70]. Tregs amplify immunosuppression via direct cell-cell interactions and secretion of IL-35, which inhibits cytotoxic T lymphocyte (CTL) activity [71]. Hypoxia exacerbates this immunosuppressive milieu by downregulating immunostimulatory neutrophils and upregulating PD-L1 on tumor cells [72, 73]. Critically, cancer stem cells (CSCs) originating from mesenchymal precursors secrete extracellular vesicles (EVs) enriched with TGF-β and interferon-gamma (IFN-γ), facilitating the conversion of monocytes into Tregs and polarizing TAMs towards an M2 phenotype [74–76], with pulmonary lesions exhibiting elevated frequencies of exhausted CD8⁺ T cells and overexpression of immunosuppressive molecules including PD-L1, Cytotoxic T-lymphocyte-associated protein 4 (CTLA-4), Tim-3, CD163, indoleamine 2,3-dioxygenase (IDO), and LAG-3 [77, 78].

Unique immunological features of bone metastatic niches
Bone metastases develop within specialized microenvironments where tumor cells co-opt physiological bone remodeling pathways [79]. Colonizing tumor cells secrete parathyroid hormone-related peptide (PTHrP), IL-6, and macrophage colony-stimulating factor (M-CSF), stimulating osteoclast differentiation via RANKL activation [80, 81]. Osteoclast-mediated bone resorption releases embedded growth factors that fuel tumor proliferation, establishing a self-perpetuating “vicious cycle” [82]. In turn, this inflammatory setting actively recruits a suite of immunosuppressive cells. MDSCs proliferate in response to VEGF and G-CSF signaling, while Tregs are drawn in via the CCL22/CCR4 axis. Simultaneously, M2-like TAMs accumulate via CCL2/CCR2 signaling [83–86]. Together, these cells act in concert, releasing factors like IL-10, TGF-β, and adenosine that effectively dampen the activity of natural killer cells and CD8⁺ T cells [87]. Moreover, metastatic lesions upregulate “don’t-eat-me” signals such as CD47, and prominently express PD-L1 [88–90]. This collective effort establishes a protected niche, enabling disseminated tumor cells to evade immune destruction. Unlike the case in osteosarcoma, DC activation and antigen presentation are markedly suppressed in bone metastases, further contributing to this immune-privileged environment.

Harnessing nano-engineered strategies to counteract bone tumor immunosuppression
The distinct physicochemical properties of engineered nanomaterials-such as their capacity for EPR, precise surface engineering, and stimuli-responsive behavior-are being leveraged to disrupt the entrenched immunosuppressive networks in bone tumors. These multifunctional platforms operate through several complementary mechanisms. For instance, the strategic application of bone-homing ligands achieves remarkably specific tissue accumulation. This targeted approach is often coupled with the co-delivery of immunomodulatory agents, enabling a two-pronged attack on both suppressive circuits and effector cell functions. This coordinated approach allows for a more profound and sustained immunomodulatory effect than conventional modalities can achieve. Diagnostic-guided theranostics integrate real-time imaging with therapeutic intervention. Exemplifying these nano-enabled strategies: Hydroxyapatite-targeting NPs achieve 30-fold higher tumor accumulation [91], pH-responsive carriers release IL-12 in acidic niches [92, 93], and MMP-cleavable nanogels deliver TGF-β inhibitors in osteolytic regions [94]. Lipid-calcium-phosphate NPs co-deliver CSF-1R siRNA and OX40 agonists for dual immunomodulation [95, 96], while dendrimer platforms co-encapsulate oncolytic viruses and GM-CSF to promote DC maturation [97, 98]. MSNs sequentially release DOX and anti-PD-1 antibodies [99], and anti-CD47-conjugated gold nanocages enhance phagocytosis [100]. Iron oxide NPs serve as MRI contrast agents while repolarizing TAMs [101], and quantum dot-labeled nanobodies guide PD-L1+ metastasis resection [102]. Collectively, these nano-specific capabilities transform “cold” bone tumors into “hot” lesions, increasing tumor-infiltrating lymphocytes 6-fold in preclinical models [103].
The progression of osteosarcoma unfolds differently in its primary bone site versus its pulmonary metastatic deposits, each governed by distinct immunological principles. The primary TME is a composite where cellular elements like cancer-associated fibroblasts, osteoclasts, osteoblasts, and TAMs are enmeshed with the native bone structure. Central to this ecosystem is the RANK-RANKL signaling pathway, which acts as a master regulator of the cross-talk between malignant and bone cells, perpetuating the damaging cycle of osteolysis and tumor growth [104]. The functional polarization of macrophages underscores this spatial heterogeneity: those in the primary tumor often display a more pro-inflammatory profile compared to their counterparts in lung metastases. While the dense bone matrix inherently limits T-cell access to the primary site, targetable antigens for engineered T-cell approaches have been identified. Conversely, the lung metastases exhibit a form of vascular dysfunction that compromises immune cell delivery. In this niche, macrophages adopt an alternative, M2-like polarization that spatially confines T-cells to the edges of the lesions [105] (Fig. 1).

Bone metastasis represents a multi-step pathogenic journey. The process, termed the metastatic cascade to bone (Fig. 2A), involves tumor cell dissemination, intravasation, survival in the circulation, extravasation into the bone marrow, and subsequent colonization and proliferation. Within the bone microenvironment (Fig. 2B), key resident cell types-osteoblasts, osteoclasts, osteocytes, and mesenchymal stem cells-exert a critical influence on metastatic progression through bidirectional communication with tumor cells. This complex interplay drives the formation of self-sustaining pathological loops, often described as “vicious cycles”. In the context of breast cancer bone metastases, factors secreted by tumor cells drive excessive osteoclast-mediated bone resorption, forming osteolytic lesions. This resorption, in turn, liberates growth factors stored within the bone matrix, which further stimulate tumor growth and reinforce the osteolytic cycle (Fig. 2C). In contrast, prostate cancer bone metastases typically present with an osteoblastic vicious cycle (Fig. 2D), where tumor-derived mediators prompt dysregulated bone formation by osteoblasts. The resulting altered bone milieu then releases factors that support tumor proliferation and survival [106]. These distinct, self-reinforcing cycles highlight the critical role of tumor-stromal co-option in driving the characteristic osteolysis or osteosclerosis of metastases.

Advancing osteosarcoma therapy with immunomodulatory nanoplatforms
In the context of primary bone tumors, particularly osteosarcoma, a range of nanomaterial-based strategies have been tailored. These platforms are designed to engage specific immunological pathways within the TME, employing diverse mechanisms to counteract local immunosuppression and mount an effective anti-tumor response. The primary systems, their targets, and their modes of action are cataloged in Table 1.

Among the most promising developments are bone-targeted nanoplatforms that exert potent immunomodulatory effects. A case in point is the development of bone-seeking immunostimulatory metal-organic frameworks (MOFs) functionalized with zoledronic acid (ZOL) [107]. These NPs display a remarkable affinity for hydroxyapatite, with in vitro binding efficiencies to calcium phosphates (CaPs) exceeding 90%, enabling their precise localization within the bone TME-a key factor governing therapeutic success [107] (Fig. 3A). The ZOL component does more than anchor the particle to the bone matrix; it actively coordinates a broad immune response [108]. In vitro, these bone-targeting immunostimulatory MOFs (BT-isMOFs) potently suppress osteoclast formation and drive the repolarization of TAMs towards the pro-inflammatory M1 state, boosting the secretion of key inflammatory cytokines by more than 300% [109]. Importantly, in vivo studies in models of breast cancer bone metastasis confirm that BT-isMOFs effectively block osteoclast-mediated bone destruction while simultaneously enhancing the M1 polarization of macrophages within the tumor, substantially lowering the tumor burden and aiding bone repair [107]. This two-fold intervention breaks the destructive feedback loop between tumor growth and bone resorption.

This approach is complemented by systems such as zwitterion-functionalized dendrimer Gold NPs (AuNPs), which efficiently deliver CpG oligodeoxynucleotides to bone marrow DCs [110]. In vitro, this targeted delivery strongly upregulates DC maturation markers and robustly activates cytotoxic T cells, correlating with approximately 85% inhibition of tumor cell growth [111] (Fig. 3B). Corroborating these findings, a bone-targeted system using OVA-conjugated zoledronate (OVA-GdZol NPs) showed similar immunomodulatory effects in an osteosarcoma model, significantly promoting DC maturation and macrophage repolarization towards the M1 phenotype, and leading to marked suppression of primary tumor growth and metastasis in vivo [112].
The scope of this strategy is even broader. For example, the polymer carrier alendronate (ALN)-Pabol, designed for nucleic acid delivery, showed remarkable efficacy in a breast cancer bone metastasis model, reducing bone tumor weight by 79.1% compared to PBS controls and profoundly reducing osteolytic lesions, with micro-CT scans showing nearly full recovery of bone volume and structure [113]. In a similar vein, dual-targeted ZIF-8 NPs co-loaded with curcumin and functionalized with hyaluronic acid and ALN, exhibited strong inhibition of breast cancer cell proliferation and osteoclast formation in vitro. More significantly, it’s in vivo use triggered PD-1 checkpoint blockade, encouraged anti-tumor macrophage development, and effectively reigned in tumor progression and bone loss, leading to a substantial reconfiguration of the metastatic bone microenvironment [114]. Together, these varied nanoplatforms highlight the strong potential of combined bone targeting and immune reprogramming to overcome immunosuppression and restore bone health.
The efficacy of combinatorial nanomaterial approaches lies in their integration of targeted delivery, checkpoint modulation, and cellular engineering. This is embodied by NC-6300 epirubicin micelles, which are engineered for controlled drug release [115]. When paired with anti-PD-L1 antibodies, this platform prompts immunogenic cell death (ICD), recalibrates the TME, and significantly boosts the infiltration of CD8⁺ T cells, leading to better tumor control than standard regimens [115–117]. Beyond micelles, innovations like ROS-sensitive polymeric NPs co-delivering a Pt (IV) prodrug and an IDO inhibitor activate the cGAS-STING pathway and block suppressive tryptophan metabolism. This combined effect fosters CD8⁺ T-cell infiltration and works in concert with PD-1/PD-L1 blockade, achieving a powerful anti-tumor response in osteosarcoma models in vivo [118]. Similarly, pH-responsive ZIF-8 carriers loaded with curcumin and a small molecule inhibitor induce a form of ICD while blocking PD-1/PD-L1 interactions, yielding DC maturation, T-cell priming, and the establishment of lasting immunological memory [119].
Nanomaterials are proving equally vital for cellular engineering. Engineered M1 macrophage membranes that overexpress SIRPα interfere with the CD47 “don’t eat me” signal on tumor cells, re-establishing phagocytic function [120, 121]. Systems like ferumoxytol NPs function dually as imaging agents and as activators for these engineered macrophages, allowing for non-invasive tracking of TAM repolarization following anti-CD47 therapy [122, 123]. When used alongside microwave ablation, this system cuts tumor recurrence by 60% through combined ablation and immune stimulation [124] (Fig. 3C). Pushing this further, macrophage-mimetic NPs deliver pexidartinib-loaded ZIF-8 to osteosarcoma sites, concurrently reducing immunosuppressive cell numbers and improving the effectiveness of adoptive T-cell therapy [125]. Newer platforms further integrate chemo-immunotherapy. Platinum-based carriers with ALN achieve bone-specific delivery of oxaliplatin, which in turn promotes DC maturation and reprograms TAMs towards the M1 phenotype, reshaping the immunosuppressive TME and amplifying T-cell responses against the tumor while also inhibiting bone destruction [126]. Another significant advance is seen in dual pH-sensitive NPs co-loaded with chlorin e6 and anti-CD47 antibodies: here, Ce6-based photodynamic therapy (PDT) sets off ICD, while blocking CD47 works together to enhance macrophage phagocytosis and T-cell activation, setting in motion a powerful anti-tumor immune cascade [127]. Collectively, these nanomaterial-driven strategies-spanning micelles, MOFs, biomimetic carriers, and stimulus-responsive NPs-demonstrate that nanotechnology is indispensable for precise immune checkpoint modulation and cellular engineering. By enhancing targeted delivery, enabling real-time monitoring, and facilitating combinatorial immunotherapies, nanomaterials effectively dismantle immune resistance barriers in bone tumors.
The activation of innate immunity, especially via the cGAS-STING pathway, stands out as a key development. MnSₓ NPs, which release H₂S, are conceived to disrupt the tumor’s protective autophagic mechanisms while concurrently activating the cGAS-STING pathway [128]. This two-pronged action not only improves DC maturation but also activates CTLs, leading to tumor cell death [129](Fig. 3D). Another compelling illustration is found in Cell@CaP vaccines, which use a patient’s own tumor cells coated with a CaP layer and loaded with cGAMP and calcium ions. These vaccines work together to activate DCs through multiple routes, including cGAS-STING and PI3K-Akt signaling [130]. Combined with radiotherapy, this approach strongly curbs the return of the primary tumor and stops lung metastasis in post-surgery models, reinforcing the value of this approach [131](Fig. 3E).
TME-modulating nanomaterials have demonstrated significant potential in overcoming immunosuppressive barriers in osteosarcoma. Ketotifen remodels the osteosarcoma stroma by reducing tissue stiffness and enhancing perfusion, improving nanomedicine delivery [132, 133]. When combined with Polyethylene glycol (PEG)-modified liposomal DOX and anti-PD1, this strategy increases CD8⁺ T-cell infiltration by 50% and shifts macrophages from the immunosuppressive M2 towards the pro-inflammatory M1 phenotype [134–136]. Complementing this, mesoporous polydopamine (pDA) NPs co-loaded with bromelain (ECM-degrading enzyme) and immune adjuvant R848 (M@B/R) degrade collagen-rich extracellular matrix in vitro. Upon near-infrared (NIR) irradiation, photothermal therapy (PTT) induces ICD while activating bromelain to enhance CTL penetration, demonstrating potent in vivo anti-tumor efficacy in osteosarcoma-bearing mice [137]. Another innovative approach utilizes photothermal immunogels loaded with black phosphorus (BP) nanosheets and STING agonists. These nanogels generate hyperthermia and ROS to eradicate tumors while activating systemic immunity, confirmed by single-cell RNA sequencing showing cGAS-STING pathway activation, suppressed metastasis, and 35% increased bone volume [138] (Fig. 3F). A parallel in vitro and in vivo study employs Mn/Fe-gallate NPs coated with tumor membranes (MFG@TCM). Under mild PTT, MFG@TCM simultaneously induces pyroptosis in osteosarcoma cells and activates the cGAS-STING pathway in DCs, promoting DAMPs release and proinflammatory cytokines. This remodels the immunosuppressive TME, significantly inhibiting primary and distant tumor growth [139].
Theranostic platforms integrating microenvironment remodeling and bone regeneration show promise. 3D-printed bioactive glass scaffolds embedded with epigenetic regulator-loaded SrTiO₃ NPs utilize ultrasound and acidic TME to trigger tumor-specific PANoptosis. In vivo, this system induces DC maturation and cytotoxic T-cell activation, suppressing osteosarcoma growth by 73.47 ± 5.2% while Sr-doped components accelerate osteogenic differentiation and bone repair [140]. Similarly, a croconaine dye-based bilayer scaffold degrades under NIR irradiation, synergizing thermal ablation with hydroxyapatite-induced mitochondrial dysfunction to eradicate residual tumors. Concurrently, it upregulates collagen-I, osteopontin, and RUNX2 expression, facilitating in vivo osseointegration and osteochondral regeneration in rat models [141]. For metastatic bone tumors, bone-targeted nanoplatforms enable precision theranostics. ZIF-90 NPs modified with ZOL (Ce6@ZPZ) exhibit 4-fold higher accumulation in osteolytic lesions than non-targeted counterparts. In vitro assays demonstrated that when Ce6@ZPZ was combined with the glycolytic inhibitor 2-DG and then irradiated, the system generated a potent burst of ROS, effectively triggering programmed cell death in tumor cells. The in vivo application of this strategy proved highly effective, achieving maximal tumor suppression and significantly extending survival in models of bone metastasis. This was attributed to a significant amplification of the PDT effect via ROS overproduction [142, 143]. In a separate line of investigation, ALN-functionalized hollow MnO₂ nanocapsules (H-MnSiO/K&B-ALD) were shown to interfere with the circadian clock in osteosarcoma cells, thereby lowering mitochondrial oxygen consumption and mitigating the hypoxic conditions within the TME [144]. These collective findings underscore the capacity of nanoplatforms to operate on multiple fronts: reprogramming the immunosuppressive landscape while concurrently providing real-time, image-guided feedback on the treatment’s progression.
The inherent dual functionality of nanotheranostic platforms stems from their capacity to concurrently utilize physical properties for imaging and biological interactions for therapy. In osteosarcoma, donor-acceptor organic radicals exemplify this paradigm with intense NIR-I absorption enabling high-resolution tumor vasculature angiography. Critically, their 49% photothermal conversion efficiency mediates localized hyperthermia that induces irreversible protein denaturation and coagulative necrosis, directly accounting for 88% inhibition of osteosarcoma growth in vivo [145, 146]. Parallelly, Mn²⁺-pyroglutamate nanocomposites demonstrate osteosarcoma-specific theranostics: Released Mn²⁺ ions function as T₁-weighted MRI contrast agents with relaxivity r₁ = 6.8 mM⁻¹s⁻¹, enabling precise tumor localization. Concurrently, pyroglutamic acid generates cytotoxic ROS while Mn²⁺ activates the cGAS-STING pathway, achieving 93% suppression of primary osteosarcoma tumors through combined photodynamic cytotoxicity and CD8⁺ T-cell immune activation [147] (Fig. 3G).

Advanced targeted delivery systems for osteosarcoma therapy

Ligand-mediated active targeting
In recent years, targeted delivery systems have emerged as a promising approach for osteosarcoma treatment, taking advantage of ligand-receptor interactions to precisely deliver therapeutic agents to tumor cells. One example of such a system is ZOL-functionalized hybrid NPs [148, 149]. These NPs exploit the overexpression of the CD44 receptor on osteosarcoma cells, which is commonly found on tumor-associated cells, to enhance drug accumulation within the TME [150]. The rod-shaped nanostructures, approximately 150 nm in length and 40 nm in diameter, are designed to efficiently arrest the cell cycle at the S-phase and elevate the expression of pro-apoptotic proteins, which leads to potent anti-proliferative effects [151]. These nanocarriers function on multiple levels. Beyond simply enhancing drug delivery, they actively promote tumor cell death through precise interference with the cell cycle. In a similar vein, ALN-conjugated systems exemplify this ligand-mediated targeting, displaying a particular avidity for bone tissue. This enhanced targeting is evidenced by ALN-Pabol NPs, which bind to hydroxyapatite with an affinity over 30% greater than that of established commercial systems [152]. This property is key to their ability to localize within osseous sites. Once inside the TME and in the presence of high glutathione levels, these polyplexes dissociate, releasing their therapeutic miRNA cargo to curb cancer cell proliferation. At the same time, the intrinsic anti-resorptive action of the ALN itself helps to limit bone destruction [113](Fig. 4A). The strategy extends further to dual-targeting liposomes decorated with YSA peptides, which target EphA2 receptors. These liposomes co-deliver DOX and JIP1 siRNA, thereby achieving effective tumor suppression through the simultaneous targeting of both extracellular and intracellular signaling pathways [153, 154].

Biomimetic delivery platforms
Cell-derived nanovehicles are attracting increasing interest for osteosarcoma therapy, primarily due to their intrinsic tumor-homing properties. For instance, exosomes derived from mesenchymal stem cells (MSC-Exos) exhibit significantly higher uptake in osteosarcoma cells than in cardiomyocytes, a trait that helps mitigate cardiotoxicity while enhancing treatment efficacy [155]. These natural vesicles display surface proteins that promote specific binding to tumor cells, thereby enhancing drug accumulation at malignant sites. Loading DOX into such exosomes not only increases vesicle size but also favors selective tumor deposition, leading to improved therapeutic outcomes. In MG63 osteosarcoma xenograft models, DOX-loaded MSC-Exos achieved a markedly higher tumor inhibition rate compared to free DOX, while demonstrating a substantial reduction in cardiac toxicity markers [156, 157].
Recent engineering strategies further enhance the bone-targeting potential of exosomes. For instance, surface functionalization with chondrocyte-affinity peptide (CAP) enables precise vesicle homing to bone lesions, significantly improving their accumulation in osteolytic areas compared to unmodified counterparts [158, 159]. Co-delivery of HDAC1 siRNA and ZOL via CAP-modified exosomes has shown synergistic suppression of osteosarcoma progression in vivo, markedly reducing tumor volume and nearly completely inhibiting osteolytic lesion formation in a preclinical model [158]. Beyond natural exosomes, synthetic biomimetic systems also hold promise. Cancer cell membrane-camouflaged chitosan-polypyrrole nanogels retain original surface antigens, allowing homologous targeting toward osteosarcoma cells [160]. When loaded with docetaxel and RANK siRNA, these nanogels respond to acidic microenvironments to inhibit the RANK/RANKL signaling axis-a key driver of bone metastasis [161–163]. This approach has demonstrated notable anti-metastatic efficacy in prostate cancer models, substantially reducing bone metastatic burden and preserving trabecular bone structure relative to controls [160]. Further expanding the delivery toolkit, engineered bone marrow stromal cell-derived exosome mimetics have been designed to improve DOX delivery. These systems enhance tumor suppression compared to free DOX, significantly curbing tumor growth in a subcutaneous osteosarcoma model while reducing off-target adverse effects such as myelosuppression [164]. In regenerative applications, BMP-2 mRNA-enriched exosomes (ExoBMP2) embedded in gelatin methacryloyl (GelMA) hydrogels enable sustained release over an extended period, effectively promoting osteogenic differentiation and repairing critical-sized calvarial defects in rats. Micro-CT analysis confirmed substantially greater new bone volume in ExoBMP2-treated defects compared to hydrogel controls after several weeks [165]. Similarly, CAP-functionalized exosomes carrying RRAGD help restore lysosomal function in nucleus pulposus cells, highlighting their potential in treating degenerative spinal disorders. In an intervertebral disc degeneration model, these exosomes increased proteoglycan content and attenuated degenerative changes, supporting their broader applicability in skeletal tissue repair [166] (Fig. 4B). Together, these biomimetic platforms-through strategic modification and combinatorial cargo loading-overcome biological barriers to achieve precise osteotropic drug delivery and effective bone microenvironment remodeling. The consistent reporting of quantitative efficacy metrics, from tumor inhibition rates to bone formation indices, underscores the translational promise of these bio-inspired systems. However, it must be noted that major bottlenecks for the large-scale clinical translation of natural vesicles like exosomes remain, including production standardization, scalable drug-loading efficiency, and batch-to-batch consistency.

Stimuli-responsive nanosystems
Stimuli-responsive nanomaterials offer precise spatiotemporal control over drug delivery, enabling therapeutic release at defined anatomical sites and predetermined time points. This capability significantly enhances treatment efficacy while reducing systemic exposure. A representative example involves ultraviolet-sensitive PEG micelles that rapidly dissociate under photoirradiation, accelerating DOX release within tumor tissue. This strategy has been shown to limit off-target effects and mitigate cardiotoxicity [167–169] (Fig. 4C). In a similar vein, pH-responsive MSNs functionalized with hyaluronic acid as a “gatekeeper” can co-deliver DOX and selenium. These constructs leverage the acidic lysosomal environment of tumor cells to trigger synergistic drug release, resulting in superior osteosarcoma suppression both in vitro and in vivo [44, 170, 171] (Fig. 4D).
Notably, recent advances include magnetically guided hydrogel microrobots encapsulating the CDK1 inhibitor Ro-3306. Under external magnetic field control, these microrobots actively migrate to osteosarcoma sites, where tumor-specific stimuli subsequently induce on-demand drug release. This dual targeting strategy allows Ro-3306 to penetrate otherwise poorly accessible tumor niches and selectively trigger MYC-dependent apoptosis in chemotherapy-resistant populations-effectively overcoming the diffusion-limited targeting typical of conventional nanocarriers [172] (Fig. 4E).

Multifunctional inorganic carriers
Inorganic nanomaterials present a promising strategy for osteosarcoma treatment by integrating controlled drug delivery with bone-healing properties. For instance, layered zinc-titanate nanogrid coatings enable sustained Zn²⁺ release, which has been found to selectively inhibit MG63 osteosarcoma cell proliferation via mitochondrial apoptosis, while simultaneously supporting the growth of MC3T3-E1 osteoblasts [173, 174]. By fine-tuning zinc content, these materials achieve an optimal balance between antitumor efficacy and biocompatibility, rendering them particularly suitable for filling bone defects after tumor resection [175](Fig. 4F). Hydroxyapatite-based carriers represent another major category, where DOX is reversibly adsorbed onto the particle surface. Under the acidic lysosomal environment, these NPs trigger intracellular drug release, enhancing local retention while reducing systemic drug exposure. Furthermore, formulations combining nano- and micro-scale hydroxyapatite particles have demonstrated improved drug retention and overall treatment efficacy [176](Fig. 4G). Beyond these, AuNPs synthesized using chondroitin sulfate exemplify another functional design. These constructs selectively provoke ROS-mediated apoptosis in osteosarcoma cells through mitochondrial disruption, while showing negligible cytotoxicity toward normal fibroblasts. Such selectivity underscores the potential of gold-based platforms in targeting osteosarcoma with minimal off-target effects [177].

Localized deposition systems
Implantable and injectable systems represent a major advance in localized therapy, enabling sustained tumor suppression while actively supporting bone regeneration. For example, titanium implants modified with Fe-Ppy/CaO₂ coatings can continuously generate ROS via Fenton-like reactions. NIR light irradiation further augments this effect by inducing photothermal conversion, collectively enhancing antitumor activity [178, 179]. Beyond direct cytotoxicity, this system also promotes a shift in macrophage polarization from the pro-inflammatory M1 towards the pro-regenerative M2 phenotype, thereby facilitating bone tissue repair following tumor ablation [180](Fig. 4H). In the context of injectable carriers, Pickering emulsion hydrogels stabilized by nano-hydroxyapatite (nHAP) have been engineered for application in post-resection cavities. These hydrogels enable the sustained local release of RANKL inhibitors, which not only restrict tumor migration but also stimulate regeneration in cranial defect models [181](Fig. 4I). This strategy illustrates how localized nanomaterial systems can concurrently address therapeutic and regenerative challenges in osteosarcoma care. Further expanding the scope, porous crystals co-loaded with ALN and oleanolic acid (ALN-OA@PCs) demonstrate marked cytotoxicity against metastatic tumor cells while suppressing osteoclast differentiation. In advanced breast cancer models, such systems have shown efficacy in preserving bone mass and mitigating osteolytic damage [182](Fig. 5A). As summarized in Table 2, these localized deposition platforms exemplify a convergent strategy-merging persistent tumor inhibition with pro-regenerative signaling-thereby offering a promising therapeutic paradigm for osteosarcoma and other bone-infiltrating malignancies. It is worth noting that many implantable systems rely on surgical placement, which limits their application in multifocal or unresectable metastases and may increase surgery-associated complication risks.

Functionalized biomaterials for osteosarcoma therapy and bone regeneration
The treatment of osteosarcoma, a malignant bone tumor, requires innovative strategies that combine tumor suppression with bone regeneration [183, 184]. Functionalized biomaterials, particularly those designed for drug delivery and tissue engineering, have shown immense promise in addressing the complex needs of osteosarcoma therapy. These biomaterials are engineered to respond to specific TMEs or external stimuli, providing targeted and controlled release of therapeutic agents while simultaneously promoting bone healing and regeneration. In this context, composite scaffolds, surface-engineered implants, bioresponsive platforms, and nanocomposite systems play pivotal roles in advancing osteosarcoma treatment [185].

Composite scaffolds with dual therapeutic functions
Multifunctional scaffolds represent a pivotal innovation in osteosarcoma therapy, designed to concurrently suppress tumor growth and support functional bone repair. Among these, porous nHAP/collagen scaffolds incorporating adriamycin-loaded poly microspheres have demonstrated particular promise [186]. These scaffolds exhibit sustained drug release kinetics, effectively inhibiting the proliferation of MG63 osteosarcoma cells in vitro [187]. Critically, they also exhibit enhanced biocompatibility and reduced systemic toxicity relative to conventional intravenous administration-a benefit attributable to several key features: their native bone-mimetic composition enhances osteoblast adhesion and proliferation while mitigating foreign-body response; the gradual release of ADM from PLGA microspheres avoids toxic peak plasma concentrations and associated cardiotoxicity; and localized drug confinement substantially limits exposure to healthy organs [188].
In a related approach, three-dimensional printed bioglass (BG) scaffolds embedded with black phosphorus (BP) nanosheets provide dual-functionality. Under NIR irradiation, BP mediates efficient photothermal ablation of tumor cells [189]. Concurrently, BP oxidation releases phosphate ions, driving biomimetic mineralization and enhancing osseointegration [190]. This system not only aids in tumor eradication but also promotes the healing and regeneration of bone tissue, facilitating a comprehensive therapeutic approach [191](Fig. 5B). Another significant design involves β-tricalcium phosphate (β-TCP) scaffolds modified with Fe₃O₄-graphene oxide (Fe-GO) nanocomposites. These systems display strong magnetothermal conversion under alternating magnetic fields, enabling thermal ablation of osteosarcoma cells while stimulating the proliferation and osteogenic differentiation of rabbit bone marrow mesenchymal stem cells [192](Fig. 6A). A critical consideration is that the loading of chemotherapeutic agents may alter the inherent mechanical properties and degradation kinetics of these scaffolds, potentially affecting their long-term structural integrity in load-bearing bone defects. Together, these advanced scaffolds serve as a model for the integration of tumor therapy and bone regeneration in osteosarcoma treatment.

Surface-engineered implants for tumor suppression and osseointegration
Surface engineering of orthopedic implants offers a promising route to address two critical challenges in osteosarcoma management: preventing local tumor recurrence and promoting bone integration. Conventional unmodified implants possess limited drug delivery function and often present bioinert surfaces, which may hinder effective osseointegration. To overcome these limitations, titanium dioxide (TiO₂) nanorod arrays have been functionalized with pDA and cyclodextrin-based polymer (pCD) layers, creating a surface reservoir capable of high curcumin (CUR) loading with sustained release profiles. The released CUR, a natural polyphenol, triggers mitochondrial dysfunction in osteosarcoma cells via excessive ROS production, leading to selective apoptosis. Importantly, the biomimetic surface properties preserve osteoblast attachment and proliferation, maintaining the implant’s regenerative function [193]. In a similar strategy, bilayered TiO₂ nanogrids containing selenium-doped pores exhibit dose-dependent cytotoxicity against MG63 cells through ROS-mediated apoptotic pathways, while remaining fully compatible with MC3T3-E1 preosteoblast viability [194–196] (Fig. 6B). Zinc titanate nanogrid coatings (Zt-NGC) represent another successful example, where sustained Zn²⁺ release suppresses tumor proliferation via mitochondrial apoptosis while stimulating osteoblast activity [175, 197]. Together, these surface-modified systems illustrate how localized engineering can simultaneously tackle tumor suppression and bone regeneration in osteosarcoma therapy.

Bioresponsive platforms for sequential therapy
Stimuli-responsive systems are gaining prominence in osteosarcoma treatment owing to three core capabilities: autonomous adaptation to the local microenvironment, sequential switching between antitumor and regenerative phases, and self-regulating drug release kinetics aligned with physiological timelines. A representative example includes fructose-derived porous CaP microspheres, which respond to endogenous alkaline phosphatase (ALP) in bone tissue to trigger calcium ion (Ca²⁺) release. This mechanism enables a sequential therapeutic regimen: initial DOX release for tumor suppression is followed by sustained Ca²⁺ release that promotes MC3T3-E1 osteoblast differentiation, thereby integrating chemotherapy with osteoinduction [198] (Fig. 6C).
Coaxial gelatin/poly fiber membranes offer another form of spatiotemporally programmed delivery. In these constructs, outer layers containing DOX-loaded nHAP (DOX@nHAp) degrade rapidly in physiological fluids to provide immediate antitumor activity. Meanwhile, inner reservoirs of icariin (ICA) sustain release over 40 days, facilitating bone defect repair. This structural segregation ensures spatial control of drug action, with ICA simultaneously enhancing MC3T3-E1 proliferation and suppressing MG-63 cell viability by 50% [199]. Further extending this concept, TME-responsive hydrogels demonstrate autonomous phase transition. When copper-cysteamine-polyethylene (CCP) glycol NPs are encapsulated within polylactic acid (PLA) scaffolds, acidic tumor conditions drive Fenton-like reactions that generate ROS for early-stage ablation. Once the tumor is eradicated and physiological pH is restored, sustained release of CCP and nHAP promotes osteogenesis-enabling a self-regulated transition from cytotoxic to regenerative mode without external intervention [200].

Nanocomposite systems with synergistic mechanisms
Hybrid nanocomposites represent an emerging strategy in osteosarcoma treatment, integrating multimodal therapeutic actions to achieve enhanced antitumor efficacy. A notable example is the development of nickel-titanium (NiTi) alloy platforms structured with femtosecond-laser microgrooves and modified with BP nanosheets and pDA coatings. These systems generate synergistic photothermal-chemotherapy effects while exhibiting antibacterial properties and promoting osseointegration-characteristics particularly valuable in post-surgical settings [201, 202] (Fig. 6D). Another innovative design involves copper-ferrite nanoclusters (NCs) embedded in alginate hydrogels [203]. These NCs display multienzyme-mimetic activity that promotes ROS generation; when combined with artemether, these interactions amplify oxidative stress, inducing tumor cell necroptosis and cuproptosis for significantly improved suppression of osteosarcoma growth [204](Fig. 6E).
Further extending this paradigm, nano‑hydroxyapatite/silk fibroin (nHAP/SF) scaffolds functionalized with trimethine cyanine-modified pDA NPs enable combined chemo‑photothermal activity against tumor cells. At the same time, nHA-mediated signaling enhances osteogenic differentiation in bone marrow stromal cells, thereby enabling a dual therapeutic and regenerative function [205](Fig. 6F). As summarized in Table 3, these nanocomposite systems exemplify how multimodal therapeutic mechanisms-spanning photothermal, chemical, and oxidative stress pathways-can be unified within a single material platform to advance osteosarcoma therapy. Despite the promise, a limitation of such complex systems is that their design often involves multi-step synthesis and functionalization, posing significant challenges for reproducible scale-up and potentially increasing the final therapeutic cost substantially.

Combinatorial nanotherapeutic strategies for osteosarcoma
Osteosarcoma remains among the most difficult malignancies to treat, characterized by aggressive metastasis and frequent resistance to conventional therapies. Recent advances in nanotechnology, however, are opening new therapeutic avenues. Engineered NPs-with tunable size, surface properties, and active targeting capacity-enable a multimodal approach to addressing this complex disease [206, 207]. In particular, the integration of ion-mediated apoptosis, autophagy modulation, energy-based therapies, radiotherapy, stimuli-responsive systems, and biomimetic platforms has demonstrated promising potential in simultaneously curbing tumor progression and supporting bone regeneration.
Inorganic NPs that release therapeutic ions represent one such direction. For instance, selenium-doped hydroxyapatite NPs (Se-HANPs) merge the structural features of mineralized biomaterials with the controlled liberation of bioactive ions. These particles promote ROS generation, activating caspase-dependent apoptotic pathways that significantly inhibit tumor growth. Notably, Se-HANPs exhibit minimal systemic toxicity, positioning them as promising candidates for therapies that concurrently treat tumors and aid tissue repair [208](Fig. 7A). Another innovative ion-mediated therapeutic strategy involves Zn²⁺-releasing titanate nanogrid (NG-Zn) coatings [175, 209]. These coatings induce mitochondrial apoptosis in osteosarcoma cells while promoting the viability of MC3T3-E1 osteoblasts, demonstrating a dual therapeutic potential that balances tumor suppression with bone regeneration [210]. Among these, the NG-Zn0.15 formulation, with its carefully calibrated zinc content, has proven especially effective in achieving this therapeutic equilibrium [175].

Autophagy modulation offers another promising route. Fullerene C60 nanocrystals (nano-C60) generate substantial ROS levels, exerting cytotoxic effects on osteosarcoma cells. However, this also activates compensatory CaMKIIα signaling, which can diminish treatment efficacy. By selectively inhibiting CaMKIIα, researchers disrupted lysosomal function, causing autophagosome accumulation and markedly enhancing the antitumor effect of nano-C60 [211](Fig. 7B). In a related approach, hydroxyapatite NPs (nano-HAPs) suppress osteosarcoma metastasis by downregulating the FAK/PI3K/Akt signaling axis [212].
Energy-based modalities such as PDT and PTT, particularly when combined with chemotherapy, yield superior outcomes in osteosarcoma. Folic acid-functionalized graphene oxide NPs, for example, can co-deliver DOX, TH287, and indocyanine green, enabling targeted chemo-PDT. This combination not only inhibits MTH1 and elevates intracellular ROS but also triggers apoptosis and autophagy through endoplasmic reticulum stress, effectively bypassing common resistance mechanisms [213](Fig. 7C). PDA-modified ZIF-8 NPs further exemplify this trend, allowing NIR-triggered methotrexate release alongside photothermal ablation. The resulting synergy permits dose reduction while preserving cytotoxicity, offering a more efficient and tolerable regimen [214] (Fig. 7D).
NPs also function effectively as radiosensitizers. Iron oxide NPs conjugated with DOX, in particular, enhance radiotherapy precision: pre-exposing osteosarcoma cells to low-dose radiation induces G2/M arrest, which in turn promotes NP uptake and ultimately leads to mitotic catastrophe via unrepaired DNA damage [215]. This sequential strategy proves substantially more effective than radiotherapy or chemotherapy alone, while maintaining hematological biocompatibility [216].
Stimuli-responsive NPs represent another frontier, engineered to leverage features of the TME for spatiotemporally controlled therapy. A representative case involves ROS-triggered nitric oxide generators, which under NIR irradiation produce heat, ROS, and catalytic NO gas. Together, these outputs establish a triple-synergistic effect-uniting photothermal, ROS, and gas therapies-to mount a coordinated assault on osteosarcoma [217](Fig. 7E). Injectable Fe-gallic acid–DOX agarose hydrogels (FeGA-DOX-AG) provide another advanced responsive system. Upon NIR irradiation, the hydrogel dissolves, releasing components that drive a Fenton reaction, generating ROS to kill tumor cells while concurrently releasing chemotherapeutic agents [218](Fig. 7F).
Biomimetic and self-amplifying systems are increasingly designed to overcome two central obstacles in osteosarcoma treatment: tumor hypoxia and stromal penetration barriers. Self-oxygenating soft nanomotors, incorporating catalase, decompose endogenous hydrogen peroxide into oxygen, raising intratumoral partial pressure from hypoxic to normoxic levels. This oxygen surge not only alleviates hypoxia but also produces microbubbles that propel chlorin e6 diffusion, enhancing photosensitizer accumulation by 3.2-fold and improving penetration depth by 2.8-fold in dense osteosarcoma tissue. This process enhances PDT efficacy, suppressing primary 143B tumor growth by 78% and reducing pulmonary metastasis by 73% in mouse models [219] (Fig. 7G). Complementing this, pyrite NPs establish self-reinforcing therapeutic cycles under NIR-II irradiation, which accelerates Fenton reactions to increase hydroxyl radical yield 4.1-fold while depleting glutathione by 85%, thereby inducing dual apoptosis–ferroptosis in tumor cells. The integration of photothermal and chemodynamic actions achieves 92% regression in chemotherapy-resistant osteosarcoma [220] (Fig. 7H).
Together, these approaches illustrate how nanotechnology enables a multipronged assault on osteosarcoma. Their key characteristics and therapeutic–regenerative outcomes are systematically compared in Table 4.

Nanomaterial-mediated immunotherapy for bone metastasis
Bone metastasis is a frequent complication in several cancers, particularly breast, lung, and prostate cancer. The treatment of bone metastasis remains a significant challenge due to the complex nature of the bone microenvironment and the limitations of conventional therapies. Nanotechnology, particularly nanomaterial-mediated immunotherapy, offers a promising solution by improving the specificity, delivery, and efficacy of therapeutic agents [221]. These strategies focus on enhancing the immune response against metastatic tumors in bone tissue, overcoming the physical and biological barriers of the bone environment [222].
Among strategies designed for bone metastases, bone-targeting ligands conjugated to NPs represent a key strategy for enhancing immunotherapeutic delivery to bone metastases. For instance, ALN-functionalized NPs improve the bone-to-liver biodistribution ratio, facilitating the targeted delivery of Gli inhibitors such as GANT58 to metastatic sites. GANT58 specifically suppresses osteoclastogenesis, thereby limiting tumor-induced bone destruction. This dual-purpose design not only elevates local drug concentrations but also leverages ALN’s inherent anti-resorptive properties, further protecting bone integrity in breast cancer metastasis models [223](Fig. 8A). A complementary strategy uses RGDyK-modified MSNs to target integrin β3-overexpressing spinal metastases. This modification enables the NPs to selectively bind to metastatic tumor sites in the bone. The NPs induce PD-L1 ubiquitination, enhancing the effectiveness of PDT via ZnPP [224, 225]. This platform helps overcome resistance to PD-1/PD-L1 inhibitors in NSCLC spine metastases, thereby boosting the efficacy of immunotherapeutic treatments and providing a dual benefit of targeting the tumor while enhancing immune responses [226].

Genetically engineered cells represent another emerging modality for precision immunotherapy in metastatic niches, functioning as precision immunotherapy vectors through three core engineered functions: tissue-specific homing to bone metastatic niches, dual therapeutic cargo delivery, and sustained microenvironmental reprogramming. A representative example includes PD-1-overexpressing hematopoietic stem cells (HSCs-PD-1) loaded with the TGF-β inhibitor SB-505,124 [227, 228]. These engineered vectors exploit natural bone marrow tropism to accumulate in metastatic lesions at levels 5.8-fold higher than in peripheral tissues [227]. Within immunosuppressive niches, they simultaneously deliver membrane-anchored PD-1 to block PD-L1 on tumor and myeloid cells, while releasing SB-505,124 (0.5 pg/cell) to inhibit TGF-β signaling in CD4⁺ T cells [228]. Through self-renewal within the bone marrow, these cells sustain functionality for over 21 days, as confirmed by flow cytometry [229]. This dual-action mechanism reprograms the local immune milieu by promoting Th1/Th2 differentiation, reducing PD-L1 expression on malignant cells by 70%, and enhancing T-cell cytotoxicity 2.3-fold compared to conventional antibody therapies [230, 231]. Notably, this strategy surpasses anti-PD-L1 monotherapy by reactivating antitumor immunity specifically within bone metastases, while reducing systemic toxicity by 3.2-fold [230] (Fig. 8B). Overall, this precision vector approach circumvents key limitations of conventional immunotherapy by acting in a spatially confined manner, utilizing stem cell biology to achieve targeted modulation of the metastatic microenvironment.
In situ vaccination strategies provide another innovative route to remodel the TME. Biomimetic CaP nano-reagents co-delivering methotrexate, ZOL, and CpG-ODN have yielded encouraging results in controlling osteosarcoma progression and metastasis [232](Fig. 8C). These pH-responsive constructs induce ICD, activate DCs, and prime CD8⁺ T cells, collectively enhancing the immune system’s capacity to identify and eliminate osteosarcoma cells. Another technique employs degradable magnesium rods (MgR), which generate localized hyperthermia via eddy currents. This heating triggers ICD and promotes DC maturation. When integrated with immune checkpoint blockade, MgR-mediated thermotherapy enhances T-cell infiltration and promotes M1-like macrophage polarization-supporting both antitumor immunity and bone osteogenesis [233](Fig. 8D, E).
Ferroptosis, an iron-dependent, regulated form of cell death, has recently gained attention as a potential strategy for overcoming treatment resistance in bone metastasis. Bone-targeted exosomes loaded with capreomycin (BT-EXO-CAP) activate the Keap1/Nrf2/GPX4 axis, selectively inducing ferroptosis in tumor cells [234](Fig. 8F-H). In another ferroptosis-based system, mesoporous Fe₃O₄-based nanoplatforms (mFeB@PDA-FA) combine chemodynamic and photothermal therapies. Light-induced hyperthermia amplifies Fenton reaction kinetics, elevating ROS generation. Moreover, buthionine sulfoximine (BSO)-mediated glutathione depletion inactivates GPX4, further promoting ferroptosis. Together, these effects create a TME highly susceptible to ferroptotic death [235](Fig. 8I).
Enzyme-responsive systems that trigger tumor-specific mineralization offer another valuable tactic for targeting bone metastases. ALP-activated SAP-pY-PBA conjugates form supramolecular hydrogel layers selectively on metastatic tumor cells. This mineralized interface sequesters calcium ions, forming dense hydroxyapatite coatings that eradicate tumor cells without damaging healthy tissue or provoking multidrug resistance [236]. Additionally, activatable semiconducting polymer nanoferries (ASPNF) leverage sonodynamic effects to amplify ferroptosis. Ultrasound-triggered singlet oxygen (¹O₂) release liberates polyamine oxidase (PAO) and Fe₃O₄ NPs, which collaboratively generate cytotoxic aldehydes and hydroxyl radicals, suppressing metastatic outgrowth [237](Fig. 8J). As detailed in Table 5, these immunotherapeutic strategies illustrate how nanomaterials can be designed to address critical challenges in the bone metastatic microenvironment, such as immunosuppression, therapeutic resistance, and targeted delivery barriers.

It is important to emphasize that many of these encouraging results are derived from murine models with simplified immune systems. Translating these findings to patients with intact human immunity and heterogeneous metastatic deposits remains a significant, unvalidated leap.

Targeted nanodelivery systems for bone metastatic cancers
The treatment of bone metastasis-a frequent and formidable complication in breast, prostate, and lung cancers-is complicated by the bone microenvironment’s mineralized structure, the requirement for high targeting specificity, and the need to preserve surrounding healthy bone tissue [238, 239]. Nanotechnology is providing new strategies to address these constraints, enhancing the specificity, bioavailability, and overall efficacy of therapeutic agents [240].
Within the paradigm of bone-targeted therapy, bisphosphonate-functionalized nanocarriers represent a cornerstone in bone-targeted therapy. ALN-conjugated polymeric NPs, for example, optimize the bone-to-liver biodistribution ratio, enabling selective delivery of the Gli inhibitor GANT58 to osseous lesions [223]. This system suppresses tumor-driven osteoclast activation, leveraging both the enhanced local drug concentration and ALN’s intrinsic anti-resorptive properties [241, 242]. In breast cancer models, this dual-action design markedly reduces bone destruction [223]. A related approach employs ALN-decorated hybrid nanoconstructs (HNCs) with superparamagnetic iron oxide NPs (SPIONPs) embedded in PLGA cores. These constructs exhibit a 300% increase in MRI relaxivity relative to conventional contrast agents, supporting high-sensitivity detection and localization of osteosarcoma metastases [243]. Further extending this paradigm, coordination polymers such as nano-Ca@BBPA undergo pH-responsive degradation within the acidic TME. This strategy achieves 30% greater hydroxyapatite affinity than standard bisphosphonates, facilitating efficient 5-fluorouracil delivery to triple-negative breast cancer (TNBC) cells [244](Fig. 9A).

Hydroxyapatite-based systems capitalize on the natural affinity of bone mineral components to guide therapeutics to metastatic sites. Rod-shaped selenium-doped nHAP (Se-nHAP) carriers, for instance, achieve threefold higher selenium loading compared to mesoporous nHA, along with a 20% improvement in DOX encapsulation efficiency. The resulting sustained ion release profile not only inhibits osteosarcoma recurrence but also promotes osteogenic activity, yielding a combined antitumor and bone-forming outcome [245](Fig. 9B). PEG‑templated nHAP carriers further increase drug‑loading capacity, enabling targeted alpha radiotherapy using radium‑223 (²²³Ra) [246]. When combined with ⁹⁹ᵐTc-MDP labeling, this theranostic platform allows real-time tracking while inducing apoptosis in bone-localized malignancies [247]. In orthotopic osteosarcoma models, local administration of DOX-functionalized nHA (nHA-DOX) after tumor resection reduced tumor-mediated bone destruction by 70%, underscoring the advantages of localized nanodelivery [248].
Cell-derived membrane camouflage is increasingly used to improve tumor homing and immune evasion in the context of bone metastasis. Hybrid membrane vesicles-such as those derived from red blood cells (RBC) and WSU-HN6 cells, functionalized with Asp8 oligopeptides-enable dual bone and tumor targeting [249, 250]. These vesicles release tirapazamine and IR780 under hypoxic conditions, uniting photothermal and chemotherapeutic actions against osseous-invasive oral squamous cell carcinoma [251](Fig. 9C). In prostate cancer, PC-3 cell membrane-coated chitosan-polypyrrole nanogels (CH-PPy NGs) utilize homologous targeting and pH-responsive release to deliver docetaxel and RANK siRNA [160]. This combination downregulates RANK/RANKL signaling, inhibiting bone metastasis by 65% [252]. Furthermore, bioengineered erythrocyte-albumin NCs modified with ALN significantly enhance tumor accumulation, showing a 2.8-fold increase compared to unmodified NCs [253, 254]. This modification substantially improves chemotherapeutic efficacy in mineralized bone microenvironments, making it a highly effective approach for treating bone metastasis [255](Fig. 9D).
Stimuli-responsive nanoplatforms are designed to release therapeutic agents in response to specific microenvironmental triggers within metastatic lesions, allowing for precise control over drug delivery. One such example is the MnO₂@PA nanosheets, which release Mn²⁺ ions in acidic tumor conditions [256]. These nanosheets enable T₁-weighted MRI imaging while catalytically converting endogenous H₂O₂ into O₂, enhancing chemodynamic therapy. This theranostic approach has been shown to suppress over 90% of osteosarcoma growth in vivo, making it a promising strategy for targeted bone metastasis treatment [257]. Glutathione-responsive ALN-Pabol polyplexes also demonstrate efficacy in bone metastasis treatment. These polyplexes dissociate in metastatic niches, releasing therapeutic miRNAs that suppress cancer cell proliferation while ALN inhibits osteolysis [113]. The dual mechanism of action helps restore trabecular bone structure in breast cancer metastasis models, offering a comprehensive therapeutic approach [113]. Additionally, pH-degradable ZIF-90 frameworks functionalized with ZOL enable bone-specific delivery of chlorin e6. When paired with the glycolytic inhibitor 2-DG, the platform enhances both photodynamic and metabolic therapy, doubling apoptosis rates in bone metastases [142](Fig. 9E).
Recent advances in nanocomposite design have paved the way for multimodal therapies against bone metastasis. T-HCN@CuMS nanoheterojunctions modified with cRGDfk peptides induce cuproptosis, a novel form of cell death, through the aggregation of lipoylated mitochondrial proteins under NIR irradiation [258, 259]. This approach effectively suppresses metastatic osteosarcoma progression by inducing proteinotoxic stress, offering a new method for bone metastasis treatment [260](Fig. 9F). Similarly, GENP nanoplatforms self-assembled from GE11-stearate conjugates inhibit EGFR/PI3K/AKT signaling while co-delivering DOX. This approach reduces the burden of TNBC bone metastasis by 75%, with minimal off-target toxicity [261](Fig. 9G). Amphiphilic ALN-OA@NCs crystals also exhibit cytotoxic effects on metastatic cells and suppress osteoclastogenesis, preserving 40% more bone mass than monotherapies in advanced breast cancer models [182]. Table 6 summarizes cutting-edge targeted nanomaterial delivery systems developed for bone metastasis therapy, categorized by their design strategy. The table outlines the core materials, targeting mechanisms, and key therapeutic functions that enhance site-specific drug delivery while overcoming biological barriers in the bone metastatic microenvironment. A pervasive limitation in the field, however, is that the efficacy of these systems is often evaluated in established macrometastasis models. Their effectiveness for prophylactic strategies aimed at eradicating micrometastases or preventing disseminated tumor cell colonization remains largely unexplored.

Multifunctional nanoplatforms for bone metastasis therapy
Bone metastasis remains a frequent and difficult-to-treat complication of advanced breast, prostate, and lung cancers. Effective management requires approaches that not only eliminate tumor cells but also actively support bone regeneration. In this context, multifunctional nanoplatforms have gained considerable attention, as they can be engineered to combine antitumor, osteogenic, and immunomodulatory activities within a single system, offering a more integrated therapeutic strategy.
Integrated 3D-printed platforms enable simultaneous tumor eradication and bone reconstruction in metastatic settings. BP nanosheet-functionalized BG scaffolds exemplify this direction by integrating photothermal ablation with biomineralization. Under NIR irradiation, light penetrates the scaffold’s 200–500 μm macroporous network, producing localized hyperthermia that effectively eliminates osteosarcoma cells. Concurrently, dissolution of the bioactive glass releases phosphate ions, which extract calcium from surrounding physiological fluids to initiate hydroxyapatite crystallization at the bone–scaffold interface. This dual mechanism produces a corrugated microstructure resembling natural Haversian systems, which significantly improves osseointegration and increases bone formation by 2.1-fold in vivo [191]. A related system uses β-TCP scaffolds modified with iron-graphene oxide (Fe-GO) nanocomposites, achieving synergy between tumor suppression and bone regeneration through magnetothermal responsiveness and osteogenic stimulation. Under alternating magnetic fields, embedded Fe₃O₄ NPs generate targeted heat within the triangular pore structure, inducing death in over 75% of tumor cells. Meanwhile, graphene oxide layers and released Fe³⁺ ions establish an osteoinductive microenvironment, increasing ALP activity by 80% and upregulating key osteogenic transcription factors. With a triply periodic minimal surface design and interconnected 300–500 μm pores, the scaffold facilitates vascular infiltration and accelerates bone defect healing by 40% [192]. Collectively, these platforms demonstrate how 3D-printed geometries coordinate spatiotemporal functions: macroporous networks permit energy-based tumor ablation while bioactive components release ions to guide biomineralization, with surface topographies directly interfacing with host tissue to promote structural integration.
Surface-engineered implants represent another important direction for managing bone metastasis. These systems are designed to respond selectively to the TME while preserving their inherent osteogenic properties. Silicalite-1 zeolite coatings, for example, can be applied to orthopedic implants to enhance the adhesion and proliferation of MG-63 osteoblasts, despite their inherent hydrophobicity [262]. These aluminum-free coatings also provide superior corrosion resistance and mechanical stability compared to traditional materials, ensuring long-term functionality in bone regeneration [263]. Nickel-titanium (NiTi) alloy platforms structured with femtosecond-laser-generated microgrooves and modified with BP nanosheets and pDA deliver combined photothermal-chemotherapy, reducing tumor burden by 85%. The microgrooves direct bone ingrowth along the surface, and the coatings exhibit strong antibacterial activity against Staphylococcus aureus, reducing infection risk and supporting bone repair [201]. Additionally, iron-doped polypyrrole coatings on sulfonated polyetheretherketone implants continuously regenerate Fe²⁺ ions, which drive Fenton reactions to produce ROS. When combined with calcium peroxide (CaO₂), which supplies exogenous H₂O₂, these coatings sustain ROS generation for more than 72 h, contributing to tumor control and promoting a shift in macrophage polarization towards the pro-regenerative M2-like phenotype to aid bone healing [180].
Temporally controlled delivery systems are engineered to synchronize tumor suppression with subsequent regenerative phases, ensuring each therapeutic action occurs at the appropriate time. Coaxial gelatin-poly fiber membranes provide a compartmentalized drug release profile: the outer DOX-loaded nHAP (DOX@nHAP) layer inhibits tumor proliferation, while the inner ICA reservoir supports osteoblast function and bone reconstruction. This design increases scaffold tensile strength by 40% and enables spatially distinct cellular responses, making it a versatile platform for tumor therapy and bone regeneration [199]. TME-responsive hydrogels encapsulating copper-cysteamine-PEG NPs (CCP) offer a comparable mechanism. When incorporated into 3D-printed PLA scaffolds, these hydrogels initiate Fenton-like reactions in acidic regions, generating ROS to kill tumor cells. After tumor clearance, the local pH normalizes, allowing sustained release of CCP and nHAP to promote osteogenesis [200]. Injectable Pickering emulsion hydrogels stabilized by nHAP can locally deliver RANKL inhibitors, suppressing tumor migration while stimulating regeneration of cranial defects. This leads to a 2.3-fold increase in new bone volume relative to controls, underscoring the potential of such systems for combined antitumor and osteogenic therapy [181].
Immunomodulatory nanoreactors reconfigure immunosuppressive bone niches through spatially confined catalytic reactions that transform immunologically “cold” tumors into “hot” microenvironments. These systems operate via sustained biochemical amplification within tumor sites, initiating cascades that counteract immune evasion. Alginate hydrogel-encapsulated Cu-Fe₃O₄ nanoreactors illustrate this concept through self-sustaining enzymatic cycles. The Cu-Fe₃O₄ nanocatalysts continuously decompose tumor-derived H₂O₂ into ROS, increasing oxidative stress 5-fold to trigger RIPK1/RIPK3-dependent necroptosis. Simultaneously, released Fe²⁺/Cu²⁺ ions enhance artesunate-induced mitochondrial dysfunction, promoting cuproptosis through DLAT protein aggregation. This dual cell-death mechanism releases damage-associated molecular patterns-such as HMGB1 and ATP-that recruit DCs and expand CD8⁺ T-cell infiltration by 2.3-fold, effectively disrupting immunosuppressive barriers [204, 264]. Complementary platforms employ different strategies: MFeB@PDA-FA induces ferroptosis via glutathione depletion and GPX4 inactivation, though its immunomodulatory role appears secondary to direct cytotoxic effects [235]. Similarly, AuMnCO@BSA-N₃ NPs utilize hypoxic conditions mainly for gas therapy and osteogenic gene activation, without directly reprogramming immune responses [265]. Together, these nanoreactors remodel tumor-immune interactions through catalytic amplification of immunogenic signals combined with precise activation of death pathways, establishing microenvironments conducive to therapeutic efficacy (Table 7). Notably, these reactors depend on the availability of specific intratumoral metabolites to drive catalytic cycles. Variability in substrate availability within highly heterogeneous and potentially metabolically depleted advanced metastases could limit their efficacy and lead to heterogeneous responses.
Despite considerable advances in nanomaterial-based strategies for bone tumors, several inherent limitations remain that merit critical evaluation. Barriers to clinical translation are especially evident, given that systemic chemotherapy often shows poor biodistribution to skeletal sites, with dose-limiting toxicities [8, 9]. More fundamentally, intensified regimens have not improved survival in metastatic Ewing sarcoma [11, 12], highlighting fundamental deficiencies in conventional delivery approaches that require systematic addressing. These limitations arise partly from material-specific weaknesses. Inorganic NPs, while offering high stability and ease of functionalization, often exhibit slow degradation kinetics and possible bioaccumulation risks, particularly in long-term treatment contexts [173–177]. MOFs offer design flexibility and porosity benefits, but concerns over structural stability and metal ion leaching significantly limit their clinical potential [107–109]. Exosomal and bioinspired systems benefit from natural targeting capacity and low immunogenicity [155–157, 160–164]. However, they face challenges in production standardization and cost-effectiveness, creating substantial industrialization bottlenecks [155, 164]. Targeting efficiency and penetration barriers present additional hurdles. NPs primarily depend on the EPR effect for passive accumulation, while ligand-mediated approaches enable active targeting through surface engineering [36, 37]. The dense mineralized bone matrix itself constitutes a major physical barrier, substantially restricting drug penetration across all platform types. From a clinical application standpoint, comparative analysis reveals platform-specific challenges. Inorganic NPs raise long-term safety concerns due to their degradation behavior. Polymer carriers are limited by loading capacity and face scaling issues that increase treatment costs. Notably, exosomal systems encounter particular difficulties in production standardization and payload capacity, critically affecting their industrial feasibility. Liposomal and micellar systems, meanwhile, show uncertainties in in vivo stability and drug leakage that may compromise therapeutic durability. Engineering and manufacturing difficulties further compound these issues. Achieving consistent and stable surface functionalization remains a significant technical barrier [186–188], while multifunctional nanocomplexes confront unresolved scalability obstacles in manufacturing processes [201, 202]. These collective constraints necessitate focused future efforts on developing novel targeting strategies to overcome bone tissue physical barriers, while optimizing the delicate balance between material biocompatibility and degradation controllability remains paramount.
While the limitations outlined earlier provide a necessary overview of current challenges, a truly strategic assessment of clinical translation requires a more pragmatic and direct comparison of leading nanoplatform candidates. Here, we move beyond descriptive lists to critically evaluate four major platforms-MOFs, liposomes, polymeric NPs, and inorganic NPs-against a set of decisive translational criteria (Table 8). This comparison does not seek to crown a single winner, but rather to illuminate the inherent trade-offs that must guide platform selection for specific clinical scenarios.

The analysis presented in Table 8 reveals a landscape defined by compromise. MOFs excel in drug-loading capacity and functional versatility, making them powerful tools for complex theranostic applications. However, their path to the clinic remains fraught with uncertainties: long-term in-vivo stability, potential metal-ion leaching, and significant challenges in scaling up production while maintaining batch-to-batch consistency-all of which currently overshadow their impressive preclinical performance. Liposomes, by contrast, represent the most clinically mature platform, with well-established manufacturing scalability and a clearly understood safety profile. Yet their utility in dense bone tumors may be inherently limited by modest drug-loading capacity and potential stability issues, which can undermine therapeutic durability against aggressive malignancies. Polymeric NPs strike a different balance, offering excellent biocompatibility and controlled, predictable degradation-a crucial advantage for long-term implantable applications. Their main constraints lie in their relatively lower drug-loading capacities and the practical complexities of scaling up multi-step functionalization processes in a cost-effective manner. Finally, inorganic NPs provide unparalleled physicochemical stability and unique functionalities such as photothermal conversion and high-contrast imaging. Their Achilles’ heel remains the pervasive concern over non-biodegradability, raising legitimate questions about long-term tissue retention, inflammatory potential, and the fate of persistent materials in patients who may survive for decades.
This comparative perspective underscores a pivotal point: no universally superior platform exists. The choice ultimately depends on which translational criterion is most critical for a given therapeutic goal. For the rapid clinical deployment of a simple cytotoxic payload, liposomes or polymeric NPs may be optimal. For exploratory, multifunctional theranostics requiring high cargo capacity, MOFs offer an unparalleled design space-albeit with higher developmental risk. Any platform intended for curative use in young osteosarcoma patients must place long-term biocompatibility and clearance pathways at the forefront of design criteria, a consideration that currently favors biodegradable polymers over many inorganic alternatives. Future research must move beyond platform-centric promotion and embrace these comparative realities. Head-to-head studies should be designed to evaluate not only efficacy but also these critical translational parameters under standardized conditions.

Clinical translation: critical lessons from the vanguard
The compelling preclinical efficacy of multifunctional nanoplatforms is undeniably driving their progression toward clinical evaluation. Yet the path from promising animal data to human trials remains a formidable filter, shaped by rigorous regulatory, manufacturing, and safety requirements. Rather than merely listing platforms currently in trials, this section critically examines the specific attributes and pivotal datasets that have propelled a select few nanotherapies into the clinical arena for bone malignancies. Analyzing these front-runners offers invaluable lessons for the broader field-highlighting the non-negotiable benchmarks that underpin successful translation.
A cornerstone of any Investigational New Drug (IND) application is a comprehensive and reproducible preclinical package that convincingly addresses pharmacokinetics (PK), biodistribution, efficacy, and toxicology. The epirubicin-loaded polymeric micelle NC-6300 exemplifies this principle. Its clinical advancement was supported by a robust dataset demonstrating not only superior pharmacokinetics-including prolonged circulation and a reduced volume of distribution relative to free epirubicin-but also a compelling mechanistic rationale: the induction of ICD and subsequent T-cell activation [115]. This dual evidence of improved drug delivery and immune reprogramming provided a strong foundation for its IND submission. Its current investigation in combination with PD-L1 blockade represents a strategic clinical hypothesis that directly tests the preclinical observation that micellar delivery can convert a cytotoxic agent into an immune-sensitizing therapy [115, 116]. This case underscores a growing trend: clinical translation is increasingly driven by platforms that offer a clear mechanistic advantage beyond mere toxicity reduction [116].
Perhaps the most significant filter in nanomedicine translation is the leap from lab-scale synthesis to Good Manufacturing Practice (GMP)-compliant, reproducible production. Liposomal formulations historically paved the way by establishing scalable manufacturing processes such as thin-film hydration and extrusion, setting a gold standard for NP batch-to-batch consistency. Newer candidates must confront this challenge directly. For bone-targeted nanocarriers-such as ALN-conjugated systems-the transition requires not only scaling up NP synthesis but also ensuring batch-to-batch consistency in the conjugation of targeting ligands. Even slight variations in this step can drastically alter biodistribution profiles [38, 39]. The successful translation of any platform, including biologically derived exosomes, hinges on solving analogous challenges: developing standardized, scalable isolation methods that yield consistent vesicle size, purity, and cargo loading, all while controlling costs. This remains a key hurdle that currently limits the widespread clinical application of exosomes, despite their low immunogenicity [159].
The bone microenvironment introduces distinct safety considerations that clinical candidates must explicitly address. Long-term local biocompatibility is paramount, especially for non-biodegradable or slow-degrading components. Platforms that have advanced often incorporate materials with well-understood metabolic fates or demonstrate convincing local confinement. The progression of iron oxide NPs into oncology trials benefits from the body’s inherent iron-metabolism pathways, which help mitigate concerns about systemic accumulation [243]. For implantable systems, extensive biocompatibility testing in relevant bone-defect models over extended periods is essential to rule out chronic inflammation or interference with bone healing-a key dataset required for regulatory approval of such devices [180, 201]. A note of caution is warranted, however, as this safety extrapolation based on physiological metabolic pathways may not hold entirely for patients with altered iron homeostasis due to disease or therapy.
Despite these advances, the clinical translation of nano-immunotherapies for bone tumors continues to face persistent, overarching challenges. A major disconnect remains between the biological complexity of human bone metastases and the relative simplicity of current preclinical models. Future work must prioritize the development of more clinically relevant models that better recapitulate human immune responses and the mineralized metastatic niche. Moreover, the field should move toward adaptive theranostic platforms capable of responding to dynamic changes in the TME, thereby enabling real-time monitoring and treatment adjustment. Finally, the considerable cost of developing complex nanoplatforms calls for innovative approaches to manufacturing and a clearer regulatory pathway that acknowledges the unique properties of combination nanomedicines. Successfully addressing these strategic issues will ultimately determine whether the transformative potential of nanotechnology can be fully realized for patients with refractory bone cancer.

Biocompatibility and in vivo dynamics of nanomaterials
The successful translation of nanomaterial-based therapies from promising preclinical results into safe and effective clinical applications depends fundamentally on a rigorous and nuanced understanding of their biocompatibility, in vivo behavior, and ultimate fate. While the literature frequently cites “high biocompatibility” based largely on initial cytotoxicity assays, this section aims to move beyond such generalized claims. We provide here a critical, material-class-specific evaluation of safety profiles, degradation kinetics, and clearance mechanisms-with particular emphasis on the unresolved questions surrounding long-term tissue interactions, especially within the unique skeletal microenvironment.
Safety and initial biocompatibility: beyond cytotoxicity assays. Initial safety assessments appropriately focus on preventing acute off-target toxicity. Formulations such as deferoxamine-based nanochelators show improved safety profiles compared to their free-drug counterparts, owing to selective renal clearance and reduced systemic adverse effects [49]. Similarly, ligand-functionalized lipid NPs (LNPs) can be engineered to minimize nonspecific hepatic uptake-a common limitation-and instead promote bone marrow accumulation [50, 51]. The inherent metabolic pathways of certain components, like the iron in SPIONPs, also contribute to their historically favorable biocompatibility profile [243]. Nevertheless, these short-term studies represent only an initial step. True biocompatibility for bone-targeted applications must also account for the material’s interactions with bone cells, its potential to disrupt bone remodeling homeostasis, and its long-term effects on bone marrow hematopoiesis-evaluations that remain notably scarce.
Degradation behavior and drug release kinetics. Controlled degradation is essential for predictable drug release and acceptable long-term tissue responses. Biodegradable polymers such as PLGA and its copolymers are widely favored because they hydrolyze into metabolites that enter the Krebs cycle, a well-characterized metabolic pathway [187, 188]. This predictable breakdown enables tunable, sustained drug release, as demonstrated in bone-targeted polyplexes like ALN-Pabol, which dissociate under the reducing conditions of the TME [113]. In contrast, inorganic scaffolds such as vanadium-doped mesoporous bioactive glass (V-MBG) are designed for sustained ion release, yet their degradation kinetics within bone are often slower and less predictable than those of polymeric systems. This disparity raises practical questions about the timing of therapeutic ion availability relative to the potential for late inflammatory responses to residual particulates [186]. CaP-based systems offer an intermediate profile; their degradation is sensitive to local pH, remaining stable under physiological conditions but accelerating in acidic tumor niches, thereby supporting environment-responsive therapy [244].
Clearance pathways and long-term fate: a material-class divide. A critical yet frequently underexplored aspect of translational development is the definitive clearance pathway and long-term in vivo fate of nanomaterial components. These vary substantially across material classes and directly influence chronic safety considerations. Polymeric NPs and liposomes are typically cleared renally or via hepatobiliary routes through the mononuclear phagocyte system (MPS). Although biodegradable polymers are eventually eliminated, their degradation byproducts must be nontoxic at accumulated concentrations. Liposomes, while generally biocompatible, remain susceptible to opsonization and rapid MPS uptake, which can lead to an accelerated blood clearance (ABC) phenomenon upon repeated administration-a significant obstacle for chronic therapeutic regimens. The clearance of MOFs is perhaps the most complex and concerning. Their fate depends critically on the stability of metal–ligand coordination bonds in vivo. Some frameworks may disassemble, releasing organic linkers and metal ions, while others persist indefinitely. The long-term biodistribution, metabolism, and potential toxicity of these metal ions, particularly in organs such as the liver, spleen, and kidneys, are not yet fully understood and constitute a major barrier to clinical adoption [107–109].
Inorganic NPs, including gold, mesoporous silica, and BP, pose the most serious long-term safety concerns due to their limited or absent biodegradability. AuNPs, although often regarded as biocompatible, are effectively non‑degradable on biological timescales. They accumulate predominantly in the liver and spleen for the lifetime of the organism, raising legitimate concerns about delayed inflammation, granuloma formation, or unforeseen impacts on organ function. MSNs degrade extremely slowly, leading to similar uncertainties regarding long‑term persistence and fibrogenic potential. BP nanosheets, valued for their photothermal properties, degrade into phosphate ions. However, the kinetics and by‑products of this degradation within the hypoxic, inflammatory bone TME remain poorly mapped. The possibility of accelerated, uncontrolled degradation releasing sudden phosphate fluxes or generating inflammatory intermediates represents a real-yet still understudied-risk [189–191]. Likewise, other 2D materials such as transition‑metal dichalcogenides demand careful, long‑term toxicological investigation.
The Skeletal Context: A Unique Challenge The bone microenvironment introduces additional layers of complexity. NPs that accumulate at bone‑remodeling sites may become incorporated into newly formed bone or remain sequestered within Howship’s lacunae. For non‑biodegradable inorganic NPs, this equates to permanent residence within the skeletal architecture. The consequences of such lifelong incorporation-including interference with ongoing bone remodeling, mechanical weakening, or the establishment of a nidus for chronic inflammation-are largely unknown and warrant dedicated study. Moreover, bone marrow is a highly immunologically active site. Persistent nanomaterials could inadvertently alter hematopoietic stem cell niches or perturb local immune surveillance, with unpredictable long‑term repercussions.
Conclusion of the Section In summary, progress in nanotherapies for bone tumors requires a shift from reporting basic biocompatibility to conducting exhaustive, material‑specific pharmacokinetic and long‑term biodistribution studies. Future work must explicitly address the clearance pathways of each component, the chronic toxicity of degradation by‑products, and the lifelong implications of sequestering non‑degradable materials within the skeleton. Developing standardized protocols to evaluate these parameters in clinically relevant large‑animal models is not merely an academic pursuit but a prerequisite for responsible clinical translation. Only through such a rigorous and unflinching assessment of long‑term fate can the therapeutic promise of these sophisticated nanoplatforms be realized safely.

Discussion and prospects: from integrated platforms to clinical reality
The preceding sections have charted the remarkable evolution of nano-engineered strategies designed to meet three intertwined challenges in bone tumors: the mineralized matrix, the immunosuppressive TME, and the lack of post-treatment healing. This journey reveals a clear and powerful narrative: the field has matured from developing simple drug carriers to creating integrated systems that perform sequential, multi-modal interventions. True innovation lies not in any single component, but in the synergistic orchestration of targeting, immune reprogramming, and tissue remodeling within a unified platform. For instance, a single theranostic agent can now home in on bone lesions via bisphosphonate ligands, induce ICD via PTT, release an immune checkpoint blocker to sustain T-cell activity, and concurrently release osteoinductive ions to initiate bone repair. This holistic approach-aimed at eliminating the tumor while rebuilding the bone-marks a fundamental shift away from conventional, linear treatment models.
Despite these conceptual and preclinical advances, the path to clinical impact remains obstructed by a series of persistent, interconnected challenges that require honest recognition and strategic focus. A critical gap exists between the impressive efficacy observed in rodent models and the anticipated complexity of human disease. Most preclinical studies rely on subcutaneous or orthotopic xenografts in immunocompromised mice, which fail to capture the human immune landscape, the mineralized and hypoxic bone metastatic niche, and the dynamic, bidirectional “vicious cycle” of tumor–stromal co-evolution. Perhaps the most pressing unresolved issue is the lack of predictive models. Future progress will depend on adopting more physiologically relevant systems, such as humanized mouse models with functional immune systems, patient-derived organoids cultured within mineralized scaffolds, or sophisticated in vitro microfluidic models of the bone metastatic niche. These systems are essential for evaluating not only tumor kill rates but, more importantly, the durability of immune memory and the quality of bone regeneration in a human-relevant context.
The drive toward multifunctionality-integrating targeting, therapy, imaging, and regeneration-inevitably increases design complexity, which clashes with the demands of scalable, reproducible GMP production. The scalability and characterization hurdle is particularly acute for intricate platforms such as MOFs or engineered exosomes. A concrete next step would be to establish platform-agnostic standardization protocols early in development. For instance, defining critical quality attributes (CQAs) such as ligand density, drug-loading efficiency, and in vitro degradation kinetics should become as routine as measuring size and zeta potential. The field must shift from publishing “one-off” innovative designs to creating platforms in which key functional components are modular and interchangeable-a move that would ease regulatory scrutiny and facilitate manufacturing scale-up.
Moreover, the long-term biodistribution and fate of nanomaterials, especially inorganic and non-degradable components, remain a major concern for diseases affecting young populations or those that become chronic. The unique safety landscape of the skeleton, where particles may be sequestered for life, calls for a new standard of evaluation. Future research must mandate long-term studies in relevant bone-defect models to assess chronic inflammation, interference with bone remodeling, and effects on hematopoiesis. One tangible proposal is the community-wide adoption of a standardized “nanomaterial bone safety panel” in large animals, evaluating not only organ toxicity but also longitudinal micro-CT and histomorphometric analyses of bone incorporation and quality.
To translate current promise into patient benefit, the field must pivot toward strategic, solution-oriented research. We propose the following concrete pathways: Next-generation platforms should evolve from being merely “stimuli-responsive” to becoming truly adaptive. This could involve integrating biosensors with feedback-controlled drug-release systems. For instance, a NP might sense elevated adenosine levels in the TME and respond by releasing an A2A receptor antagonist, thereby dynamically countering immunosuppression in real time. Moving beyond generic calls for personalization, a tangible next step is to develop high-throughput screening platforms. Patient-derived tumor cells or exosomes could be screened against diverse libraries of targeting ligands to identify the most efficient homing molecule for an individual’s cancer-paving the way toward truly individualized nanocarrier design. The coupling of tumor eradication and bone healing must become more refined. Future “regenerative immunotherapies” might employ smart scaffolds that first release immunomodulators to clear the tumor and establish a pro-regenerative immune climate, followed by the staged release of osteogenic factors timed to this new microenvironment. Gene-activated matrices delivering osteogenic miRNAs or CRISPR-based systems could offer precise control over the regeneration process. The future of nano-immunotherapy lies in its convergence with digital technologies. Imaging data from nanotheranostic agents could feed into AI algorithms to predict treatment response, optimize dosing schedules, and detect early signs of recurrence or metastasis-enabling a truly dynamic and personalized treatment loop.
In summary, the field of nano-immunotherapy for bone tumors stands at a pivotal juncture. The foundational principles-integration, targeting, and reprogramming-are now firmly established. The challenge ahead is to navigate the translational “valley of death” through rigorous science, standardized methods, and an unwavering focus on the unresolved questions of model fidelity, manufacturing complexity, and long-term safety. By embracing these challenges as a roadmap rather than a barrier, the research community can accelerate the delivery of these transformative technologies to the patients who need them most.

Conclusion
Nanomaterials show substantial progress in addressing osteosarcoma and bone metastases by overcoming immunosuppressive microenvironments and structural barriers. Engineered platforms enable precise bone targeting, reprogramming of pathogenic cellular interactions, and disruption of tumor-stromal vicious cycles. Multimodal integration of photothermal, immunomodulatory, and cytotoxic strategies enhances therapeutic efficacy while minimizing systemic toxicity. Critical translational challenges persist, including clinical validation beyond preclinical models, biocompatibility optimization of inorganic components, and scalable manufacturing of multifunctional systems. Future advancement necessitates adaptive theranostic platforms responsive to dynamic tumor cues, standardized biocompatibility protocols, and regenerative designs synchronizing tumor eradication with functional bone reconstruction. Prioritizing these objectives will accelerate clinical deployment of nanomedicine against skeletal malignancies.