Magnetic Hyperthermia with Iron Oxide Nanoparticles: From Toxicity Challenges to Theranostic Cancer Applications

1. Introduction

As one of the leading global health challenges, malignant diseases remain a major cause of mortality, primarily arising from the progressive accumulation of genetic mutations in normal cells that promote unchecked cell division and tumor development. Despite decades of research and therapeutic advances, the overall mortality rates for several cancer types have only modestly declined [1]. Conventional cancer treatment strategies typically include surgical resection, radiotherapy, and systemic therapies such as chemotherapy, hormonal therapy, and targeted biological agents [2]. While these modalities have demonstrated clinical efficacy, they are frequently associated with significant adverse effects. Surgical interventions, for instance, can result in postoperative complications, tissue damage, and the potential dissemination of malignant cells, thereby increasing the risk of metastasis [3]. Chemotherapy involves the administration of cytotoxic agents that often lack tumor specificity, leading to collateral damage to healthy tissues and systemic toxicity [4]. Similarly, radiotherapy employs ionizing radiation which, when applied over extended periods, may compromise the structural and functional integrity of surrounding normal tissues [5].

Hyperthermia treatment, or thermotherapy, has emerged as a promising adjunct or alternative to conventional treatments. This approach involves elevating the temperature of body tissues to induce cancer cell apoptosis while sparing normal cells [6]. The therapeutic application of heat dates back to ancient civilizations, including those of Greece, Egypt, Rome, and India [7]. In the 19^th century, spontaneous tumor regression in febrile patients was documented, prompting early investigations into hyperthermia for oncological applications, such as in the treatment of cervical cancer [7,8]. Since the 1970s, hyperthermia has garnered renewed clinical interest, with controlled trials exploring its efficacy in cancer treatment. Studies have demonstrated that cancer cells are more susceptible to temperatures between 42–45 °C, undergoing apoptosis, whereas normal cells exhibit greater thermal resilience [9].

Depending on the tumor’s location, depth, and stage of progression, three main hyperthermia strategies have been established for clinical application: local, regional, and whole-body hyperthermia. Whole-body hyperthermia is typically employed in cases involving deep-seated tumors or disseminated metastases, where the entire body is uniformly heated using methods such as hot water baths, thermal chambers, or infrared radiation [10,11,12,13]. For advanced-stage malignancies confined to specific areas, regional hyperthermia is applied through techniques like thermal perfusion, external applicators, or microwave antennas to deliver targeted heat [10,11,12,13]. Local hyperthermia, the least invasive approach, is primarily used for treating localized tumors situated either superficially or within accessible body cavities.

Over the past century, substantial technological progress has led to the development of magnetic nanoparticles (MNPs), which have attracted growing interest in biomedical research, particularly in oncology. These nanomaterials possess the unique ability to convert electromagnetic energy into heat when exposed to an external alternating magnetic field (AMF) [14,15]. Importantly, the penetration depth of AMF is not significantly attenuated by biological tissues, allowing effective activation of MNPs even within deep-seated tumors [15,16]. Once internalized by cancer cells, MNPs function as localized heat sources, raising the temperature of tumor tissue to levels sufficient to induce apoptosis. This approach, known as magnetic hyperthermia (MH), has emerged as a promising and innovative therapeutic strategy with the potential to enhance cancer treatment outcomes while minimizing damage to surrounding healthy tissues [16,17].

The first experimental evidence supporting the use of MH for cancer treatment was reported in 1957 by Gilchrist et al. [18], who conducted an in vitro study involving lymph nodes containing colon and rectal cancer metastases. In this pioneering work, maghemite (γ-Fe₂O₃) nanoparticles, ranging in size from 20 to 100 nm, were introduced into the lymph nodes and subjected to an AMF with an amplitude (H) of 16 - 19.2 kA/m at a frequency (f) of 1.2 MHz. The resulting temperature increases to 43 - 46 °C successfully eradicated carcinoma cells by destroying the metastatic tissue [18]. As a result of extensive research and due to their favorable biocompatibility and biodegradability, both maghemite (γ-Fe₂O₃) and its reduced form, magnetite (Fe₃O₄), were approved by the U.S. Food and Drug Administration (FDA) for clinical trials [19]. Consequently, iron oxide nanoparticles have become the most widely used agents in magnetic hyperthermia (MH) applications [20,21]. The therapeutic potential of MH reached a significant milestone with the advancement of the German company MagForce AG, which received regulatory approval from the European Union to clinically treat glioma patients [22,23,24,25].

The interaction between AMFs and biological tissues generates non-specific heating through the induction of eddy currents. This can activate the body’s thermoregulatory responses and produce complex thermal gradients throughout the patient’s body [26,27]. To ensure safety, a limit has been established for human exposure to AMFs by restricting the product of H and f to a maximum of 5 × 10⁹ A·m⁻¹·s⁻¹ [28]. However, the heat dissipation capabilities of commercially available magnetic nanoparticles (MNPs), such as Nanotherm, Feridex, and Resovist, remain inadequate within physiologically safe AMF parameters. As a result, the therapeutic effect is often insufficient for complete tumor ablation, limiting the widespread adoption of MH as a standalone treatment option in clinical settings.

To advance the clinical application of MH in cancer therapy, two principal strategies have been identified in scientific literature. The first approach focuses on the design and synthesis of magnetic nanoparticles (MNPs) with enhanced intrinsic magnetic properties, such as increased saturation magnetization (M_s) and magnetic anisotropy (K), to improve heat generation efficiency even at low concentrations. This can be achieved by manipulating parameters such as particle size, shape, chemical composition, and surface morphology. As these aspects have been extensively covered in previous reviews [29,30,31], therefore we will briefly summarize the most significant developments in this area in the first part of our review. Notably, numerous studies have reported that only a small fraction of systemically administered MNPs effectively accumulate at the tumor site [32]. Consequently, clinical efficacy often relies on direct intratumoral injections, which restrict treatment to tumors that are accessible [33]. To address this limitation, recent research has increasingly focused on combining MH with other anticancer modalities within a single, multifunctional MNP-based nanoplatform [34]. The aim of this review is to highlight the advancements and the challenges that the magnetic hyperthermia faces from the design of magnetic nanoparticles to ensure biocompatibility, tumor specificity and clean biodegradation in the living organisms, while still preserving the hyperthermic and drug delivery properties, towards the biological testing on different in vitro, respectively in vivo models and eventually clinical patients. This integrative strategy represents the second major approach in the field and will be explored in detail in the second part of this review. Clinical advancements in the field, based on clinical studies of MH efficacy in oncologic patients, will be further discussed in the third part of this review. Finally, the challenges, outlook and conclusions regarding the advances of magnetic hyperthermia in oncological research and the prospects of MH to become an important adjuvant therapy in the oncologic patients are described.

2. Relevant Sections and Discussions

2.1. Toxicity Issues

Magnetite (Fe₃O₄) nanoparticles (MNPs) are among the most widely investigated nanomaterials in biomedicine due to their superparamagnetic (SP) properties, ease of functionalization, and intrinsic biocompatibility. They are extensively employed in applications such as magnetic resonance imaging (MRI) contrast enhancement, targeted drug delivery, and MH therapy. However, despite their promising biomedical potential, MNPs are not inherently risk-free. Concerns about both acute and chronic toxicity have emerged, particularly under conditions of MH, where deliberate nanoparticle heating can inadvertently damage surrounding healthy tissues [35,36,37].

In vitro cytotoxicity studies have revealed that Fe₃O₄ nanoparticle-induced toxicity is both dose- and time-dependent, typically associated with increased production of reactive oxygen species (ROS) and lipid peroxidation, evidenced by elevated malondialdehyde (MDA) levels. This oxidative stress can impair cellular enzymes, damage membranes, and compromise cell viability. For instance, Ahamed et al. [38] demonstrated significant ROS and MDA elevation in A431 and A549 cell lines exposed to 25–100 µg/mL MNPs, correlating with reduced cell viability. Similarly, recent findings on human umbilical vein endothelial cells (HUVECs) identified oxidative stress and genotoxic effects at an IC₅₀ of approximately 79 µg/mL [39]. Iron overload from internalized nanoparticles may further exacerbate toxicity by triggering ferroptosis—a form of programmed cell death driven by lipid peroxidation—and catalyzing Fenton-type reactions that generate hydroxyl radicals. These highly reactive species induce DNA strand breaks, protein carbonylation, and membrane damage. Such mechanisms have been implicated in neuropathological contexts like intracerebral hemorrhage, and may similarly contribute to nanotoxicity in non-neuronal tissues [40,41]. The surface chemistry of Fe₃O₄ nanoparticles plays a pivotal role in modulating their biocompatibility. For example, in porcine aortic endothelial cells, dextran- or polyethylene glycol (PEG)-coated particles (5 nm and 30 nm) induced no significant loss of viability or increase in ROS generation even at 0.5 mg/mL after 24 h exposure. Notably, PEG reduced ROS production by 62.6%, and dextran by 35.2% compared to uncoated cores, with apoptosis levels remaining below 10% [42]. In contrast, polyethylenimine (PEI)-functionalized Fe₃O₄ particles (30 nm) displayed pronounced cytotoxicity in SH-SY5Y, MCF-7, and U937 cell lines, decreasing viability by up to 50% within 24 hours at 100 µg/mL, and even further after 168 hours. These effects, marked by increased ROS, lipid peroxidation, and lactate dehydrogenase (LDH) release, were largely mitigated upon PEGylation of the PEI coating [43].

In vivo, the toxicity profile of MNPs is influenced by parameters such as dose, administration route, biodistribution, and particle size. Following systemic administration, MNPs tend to accumulate in organs of the mononuclear phagocyte system, notably the liver and spleen, where they may trigger inflammation and tissue injury. For instance, high-dose administration of dextran-coated iron oxide particles in mice, followed by AMF exposure, resulted in severe hepatosplenic damage or mortality, whereas lower-dose groups survived but still exhibited signs of tissue stress, including elevated liver enzymes and splenic necrosis [44]. Particle size is equally critical: ultrasmall MNPs (2.3 and 4.2 nm) administered intravenously at 100 mg/kg induced fatal multiorgan oxidative stress, particularly affecting cardiac tissue, whereas 9.3 nm particles of identical composition showed no overt toxicity at the same dose [45]. Long-term studies suggest that MNPs are only partially cleared from the body. Residual iron, often sequestered in ferritin-like structures, can persist in the liver and spleen for several months post-administration. This has been associated with chronic low-grade hepatic inflammation and disruption of iron metabolism, although conclusive evidence linking this persistence to long-term health consequences remains limited [46,47].

2.2. Methods to Enhance the Hyperthermic Capability of Iron Oxide Nanoparticles

The thermal effect generated by MNPs under AMF stimulation is quantified by a physical parameter known as the specific absorption rate (SAR), also referred to as specific loss power (SLP). SAR represents the amount of heat released per unit time per unit mass of MNPs and is typically expressed in watts per gram (W·g⁻¹) [48]. The SAR value is influenced by both the intrinsic properties of the MNPs—such as particle volume and saturation magnetization—and the extrinsic parameters of the applied AMF. To enhance the induction heating performance, efforts have been directed toward optimizing the intrinsic magnetic properties of MNPs. Simultaneously, thermal efficiency has been externally improved by increasing the frequency and amplitude of the AMF in different MH setups.

2.2.1. Formulation

For biomedical applications requiring "injectable" nanoprobes, superparamagnetic nanoparticles—commonly referred to as SPIONs (superparamagnetic iron oxide nanoparticles)—are generally preferred. Their lack of remanent magnetization in the absence of an external magnetic field facilitates colloidal stability, enhances dispersion in biological fluids, and minimizes the risk of particle aggregation. Among these, ultra-small SPIONs (typically <5 nm in core diameter) have emerged as promising candidates for magnetic resonance imaging (MRI) contrast enhancement due to their excellent magnetic relaxation properties [49,50]. However, in the context of magnetic hyperthermia (MH), such ultra-small SPIONs often exhibit low specific absorption rate (SAR) values, limiting their heating efficiency [51]. Furthermore, a significant reduction in heating performance is commonly observed when SPIONs are internalized into cells or embedded in tissues, likely due to restrictions in Brownian motion and changes in local viscosity [52]. As a result, considerable research efforts have been directed toward optimizing nanoparticle properties within the superparamagnetic regime to enhance the efficacy of magnetically induced hyperthermia under physiological conditions.

The magnetic properties of MNPs are strongly influenced by their size. An increase in the size or volume of MNPs typically results in a higher M_s, which reflects the net alignment of all magnetic spins within the particle. This increase continues up to a critical threshold, beyond which M_s stabilizes and approaches the bulk material value. Numerous studies have reported a significant increase in the SAR with the growth of spherical SPIONs, ranging from several tens to several hundreds of W/g [53,54,55,56,57,58]. As the diameter of MNPs increases, their magnetic anisotropy energy — the energy responsible for maintaining the magnetic moment in a preferred orientation — also increases. For each MNP composition, there exists a characteristic size at which the anisotropy energy surpasses the thermal energy, stabilizing the magnetic moment along a preferred axis known as the easy axis of magnetization. This transition drives MNPs from a superparamagnetic to a ferromagnetic regime, characterized by the appearance of hysteresis loops. These loops exhibit remanent magnetization (M_r), representing the residual magnetization at zero external field, and coercivity (H_c), the magnetic field required to bring the magnetization to zero. The magnetic MH efficiency of MNPs is governed by their dynamic hysteresis behavior [59], which is influenced not only by Neel and Brownian relaxation mechanisms but also by DC magnetic hysteresis. As a result, SAR values in ferromagnetic particles can be nearly an order of magnitude higher than in their superparamagnetic counterparts [60,61,62,63]. However, despite their high heating efficiency, ferromagnetic nanoparticles are generally less suitable for biomedical applications due to their colloidal instability and finite coercive field, which promote aggregation and reduce biocompatibility [64,65,66].

Individual SPIONs often exhibit limited magnetic moments, which restrict their efficiency in MH. However, when these SPIONs are organized into clusters through self-assembly or aggregation, magnetic interactions between the closely packed cores can induce collective magnetic behaviors, resulting in enhanced net magnetic moments [67]. This clustering significantly improves key magnetic properties, such as M_s and magnetic susceptibility, thereby increasing their responsiveness to external magnetic fields. Moreover, the clustered architecture provides improved colloidal stability and resistance to uncontrolled aggregation, preserving SP behavior while ensuring long-term performance under physiological conditions [68,69]. Cluster formation can occur via two primary strategies. In a two-step process, SPIONs are first synthesized as discrete particles, followed by their assembly into clusters mediated by ligand-induced colloidal interactions, such as hydrophobic or electrostatic forces [70]. Alternatively, clustering can occur in a single-step synthetic route, wherein the nanoparticles aggregate during formation [71,72]. The polyol method has been widely employed for this purpose in the past decade due to its adaptability, scalability, and ability to control particle morphology. This method enables the formation of various hierarchical structures, including nanoclusters, nanoflowers, and hollow spheres, by tuning reaction parameters such as temperature, solvent polarity, and precursor concentration [73,74,75,76,77,78]. Particularly, flower-like magnetic nanoparticles (nanoflowers) with coherent crystallographic orientation between cores have garnered attention for their superior magnetic heating performance [79,80,81]. The improved SAR values are thought to arise from collective spin dynamics and magnetic coupling effects within the multicore structure, which favor more efficient energy dissipation under AMF [82,83].

2.2.2. Shape

In the case of MNPs, surface atoms represent a significant proportion of the total atomic content, and their magnetic and chemical behavior often diverges from that of the bulk material. This is primarily due to the intrinsically high surface-to-volume ratio of MNPs, which causes surface effects to dominate their overall magnetic properties. Notably, the asymmetric coordination of surface atoms gives rise to spin disorder or spin canting, ultimately reducing the M_s of the nanoparticles [84]. This phenomenon is especially pronounced in spherical SPIONs, which expose multiple crystallographic facets with numerous edges and corners [85]. These structural features enhance surface anisotropy and require greater energy to reorient surface magnetic moments, negatively impacting their heat dissipation efficiency under AMF. Consequently, the synthesis of anisotropic SPIONs has gained considerable interest as a strategy to enhance MH performance [86]. Various non-spherical morphologies, including nanocubes [87,88,89,90,91], octopods [92], octahedrons [93,94], nanorods [95,96,97], nanodiscs [98], nanorings [99], and polyhedral structures [100], have demonstrated improved heating efficiency compared to their spherical counterparts (Figure 1).

2.2.3. Doping with Metallic Ions

Magnetite (Fe₃O₄) exhibits an inverse cubic spinel structure, in which O²⁻ anions form a face-centered cubic (FCC) lattice that accommodates two distinct cationic sublattices: tetrahedral (A) sites, exclusively occupied by trivalent iron ions (Fe³⁺), and octahedral (B) sites, shared by both divalent (Fe²⁺) and trivalent (Fe³⁺) cations. Superexchange interactions mediated by O²⁻ anions govern the magnetic coupling between these cations, resulting in three main interaction types: A–O–A, B–O–B (intra-sublattice), and A–O–B (inter-sublattice). While intra-sublattice interactions tend to be ferromagnetic, the inter-sublattice (A–O–B) interactions are antiferromagnetic in nature, giving rise to the characteristic ferrimagnetism of magnetite. The net magnetic moment per formula unit is determined by the difference between the magnetic moments at B and A sites (μ_oct – μ_tet), primarily attributed to the presence of Fe²⁺ ions on B sites, resulting in a net moment of approximately 4 μ_B (Bohr magnetons) per formula unit.

Tailoring the magnetic properties of Fe₃O₄ nanoparticles (MNPs), can be effectively achieved through substitutional doping of Fe²⁺ ions with other divalent transition metal cations such as Mn²⁺, Co²⁺, Ni²⁺, Cu²⁺, and Zn²⁺ (Figure 1). For instance, replacing Fe²⁺ (3d⁶) with Mn²⁺ (3d⁵), which possesses a higher magnetic moment, enhances the overall magnetic moment to 5 μ_B per formula unit. This substitution has been shown to significantly increase the saturation magnetization (M_s), reaching values up to 110 emu/g_metal, and consequently, improve both heating efficiency in MH [101,102,103,104,105,106]. Interestingly, zinc doping—despite Zn²⁺ (3d¹⁰) having zero magnetic moment—also leads to modulation of M_s due to site-specific cation rearrangement within the spinel lattice. At low doping concentrations (x < 0.5 in Zn_xFe₃₋ₓO₄), Zn²⁺ preferentially occupies A sites, displacing Fe³⁺ cations to B sites. This redistribution increases the magnetic moment at B sites (μ_oct) and reduces that at A sites (μ_tet), resulting in a substantial increase in M_s (161 emu/g_metal), displaying approximately four times greater heating efficiency compared to conventional SPIONs [85,107,108,109]. Another example is the use of magnesium as a dopant. MgₓFe₂O₃ nanoparticles (x = 0 – 0.15) with an average size of 7 nm demonstrated exceptional heating power — approximately 100 times higher than commercial Fe₃O₄ (Feridex) — attributed to enhanced magnetic susceptibility and ~50% octahedral Fe³⁺ vacancy occupation by Mg²⁺ ions, as supported by atomic structural modeling [110]. Although M_xFe_3-xO₄ nanoferrites are often doped to M_s and SAR, results vary significantly depending on synthesis parameters and dopant distribution across T_d and O_h sites. Thus, precise control of dopant incorporation within the Fe₃O₄ lattice is essential to optimize magnetic and thermal performance.

2.2.4. Controlled Nanoscale Assembly of MNPs

It has been demonstrated that during MH experiments, the application of an AMF promotes the organization of MNPs into elongated assemblies or chains. This field-induced structuring significantly influences the SAR and overall heat generation capabilities of the MNPs [111,112,113]. Chain formation behavior has also been observed intracellularly, where MNPs internalized by cells tend to align in response to the AMF [114]. Further studies have revealed that such alignment occurs within intracellular vesicles and does not compromise cellular morphology or nuclear integrity [115]. In a related context, magnetosomes—magnetic nanoparticles biosynthesized by magnetotactic bacteria—exhibit superior heating efficiency compared to their synthetic analogs, primarily due to their intrinsic chain-like organization [116,117,118]. Consequently, the controlled nanoscale assembly of MNPs to enhance SAR represents an important topic in MH research.

Several groups have investigated the effect of pre-aligning MNPs in a static magnetic field (H_DC) on their heating performance. They have demonstrated that MNPs pre-aligned under a H_DC before immobilization (e.g., via gelation in a solid matrix) produce significantly higher SAR values when aligned parallel to the AMF, compared to randomly oriented nanoparticles [119,120,121,122,123]. Furthermore, two in vitro studies have shown that either culturing cancer cells with MNPs under a H_DC or pre-aligning incubated MNPs significantly improves MH efficiency and enhances cancer cell destruction, compared to the non-aligned scenario [124,125].

Another promising strategy for boosting MNP heating efficacy involves the superposition of a s H_DC on the AMF during MH treatment. Experimental evidence indicates that for SPIONs, this approach can lead to SAR enhancements of up to 40% relative to AMF-only conditions [16]. Chain formation in this system was confirmed by atomic force microscopy [126]. In contrast, ferromagnetic MNPs, when aligned under an H_DC in low-viscosity agar matrices (0.10–2.00 wt%), exhibit even more pronounced SAR increases—up to threefold—especially at low agar concentrations (0.1 wt%), where particle mobility is less restricted [10]. Similarly, the application of static fields (H_DC = 10-20 kA/m) parallel to the AMF during MH measurements has been shown to significantly increase the SAR of zinc ferrite nanoparticles in a concentration-dependent manner, with greater effects observed at lower particle concentrations [127].

2.2.5. Controlled Nanoscale Assembly of MNPs

An effective strategy to enhance the specific absorption rate (SAR) of superparamagnetic iron oxide nanoparticles (SPIONs) lies in tailoring their surface coating properties. Surface coatings influence both magnetic behavior and colloidal stability, ultimately impacting heating efficiency under an alternating magnetic field (AMF) [128].

Liu et al. investigated the influence of coating thickness on the SAR of Fe₃O₄ magnetic nanoparticles (MNPs) coated with phosphorylated methoxy polyethylene glycol 2000 (PEG2000). Their findings revealed that for smaller-sized nanoparticles (e.g., 9 nm and 19 nm), SAR increased as coating thickness decreased, an effect attributed to enhanced Brownian relaxation losses. Notably, the PEGylated SPIONs retained high SAR values under various physiological conditions, indicating strong colloidal and functional stability [129].

In another study, Fe₃O₄ nanoparticles coated with PEG of different molecular weights demonstrated resistance to the formation of collective coatings. This prevented the agglomeration of nanoparticles into large clusters and preserved their high SAR across environments with varying ionic strengths and viscosities, including distilled water, physiological saline, agar, and cell culture media [130]. Further investigations into the impact of surface functionalization demonstrated that hydrophilic SPIONs synthesized via oleate capping and subsequently modified with diverse ligands (PEG, dimercaptosuccinic acid – DMSA, cetrimonium bromide – CTAB, stearic acid – SA, and poloxamer 188 – P188) exhibited different heating profiles [131]. Ligand exchange with PEG and DMSA promoted nanoparticle dispersion, whereas intercalation with CTAB and SA or encapsulation with P188 led to agglomeration into spherical clusters. Magnetic hyperthermia experiments showed significantly higher SAR for the PEG- and DMSA-modified samples, emphasizing the detrimental effect of aggregation on heating performance [131]. Additionally, dextran-coated SPIONs with a diameter of 7 nm have also demonstrated high SAR values [132], supporting the notion that both organic and inorganic surface coatings can substantially enhance magnetic heating efficiency.

Inorganic coatings, such as gold or silica shells, have also proven effective [133]. Mohammad et al. [134] reported a 4–5-fold enhancement in SAR when SPIONs were coated with a thin gold shell (0.5 nm). Moreover, a maximum SAR value of 1300 W/g_Fe was achieved for dumbbell-shaped hybrid nanostructures comprising a 24 nm Fe₃O₄ domain attached to a 9 nm gold seed [135]. These structures benefit from synergistic effects between magnetic and plasmonic components, improving thermal response. Silica coating represents another widely used approach for surface functionalization of SPIONs [136,137,138]. Individual SPIONs coated with a silica shell were shown to maintain colloidal stability and avoid magnetic dipolar interactions, particularly under AMF exposure [139]. This led to superior heating performance compared to uncoated SPIONs or clusters encapsulated within a common silica shell [140,141,142,143]. These findings collectively emphasize the critical role of surface chemistry and nanoparticle architecture in optimizing the magnetic hyperthermia potential of SPIONs.

2.2.6. AFM Characteristics

In general, MH experiments demonstrate that SAR tends to increase with both the frequency (f) and amplitude (H) of the applied alternating magnetic field (AMF). However, the heat released during MH cannot be indefinitely enhanced solely by tuning these two external AMF parameters, due to both biological safety limits and intrinsic magnetic nanoparticle (MNP) properties.

First, to prevent overheating of healthy tissues due to eddy currents, safety guidelines such as the well-known Brezovich limit (H·f ≤ 5 × 10⁹ A·m⁻¹·s⁻¹) have been established [28]. Although some recent studies suggest that higher limits could be acceptable under certain conditions [144,145], García-Alonso et al. propose a more permissive threshold of H·f ≤ 9.6 × 10⁹ A·m⁻¹·s⁻¹ [146].

Second, while SAR is often assumed to scale linearly with frequency across various MNP types and sizes, its dependence on field amplitude H is more complex. For SPIONs smaller than ~10 nm, SAR typically follows a quadratic relationship with H (i.e., SAR ∝ H²) [48,147,148]. In contrast, for larger nanoparticles, SAR exhibits more complex field dependencies, sometimes deviating from simple power-law behavior [63]. Importantly, the quadratic dependence is generally observed only at low H; beyond this range, SAR tends to saturate [149]. This saturation effect has been correlated with the nanoparticle saturation magnetization (M_s) and is supported by numerical simulations that incorporate the field dependence of both Néel and Brown relaxation times [149]. For ferromagnetic particles, SAR saturation follows a sigmoidal trend as a function of H, a behavior that has been reported in several experimental and theoretical studies [74,100,122,127,150,151].

The saturation of SAR with increasing field amplitude (H), together with the safety constraints imposed by the H·f product limit, highlights a fundamental limitation of MH: the amount of heat that can be safely and effectively delivered to deep-seated tumors is inherently restricted. Consequently, MH alone is unlikely to achieve full tumor eradication, particularly in aggressive or resistant cancer types. For this reason, MH is more appropriately used as an adjuvant strategy, enhancing the efficacy of conventional treatments such as chemotherapy, radiotherapy, or immunotherapy. When applied in combination, MH can sensitize tumor cells to these therapies by promoting localized hyperthermia, thereby offering a synergistic anticancer approach [152]

2.3. Organic Coating

A great deal of research was conducted on biocompatibility and cellular uptake enhancement by using synthetic polymer coating, such as polyethylene glycol (PEG), polyvinylpyrrolidone (PVP), polydopamine (PDA) and PDA analogues or natural derived coatings like chitosan and dextran (Figure 1).

2.3.1. Synthetic Polymers

PEG is one of the most common nanoparticle formulations, having wide-ranging applications by decreasing clearance and increasing water solubility. PEG forms a “stealth” hydration layer, reducing opsonization and MPS capture along with preventing hydrophobic aggregation of hydrophobic particles, shielding the NP surface from enzymes and antibodies [153,154,155,156]. Polyvinylpyrrolidone (PVP) is a colorless, water-soluble, biocompatible polymer known for its exceptional pH stability and binding capabilities, which aid in drug solubility and dispersion. Its amphiphilic structure allows it to interact effectively with solvents of varying polarities, making it versatile for constructing complex macromolecules, often via conjugation with polyacids like polyvinyl alcohol PVA or PAA. PVA-PVP composites improve mechanical strength and thermal stability along with boosting ferrimagnetic performance [157,158].

Polymerizing catecholamines leads to polydopamine (PDA) and PDA analogue polymers on the surface of the nanoparticles, particularly nanoclusters, which allow many potential applications due to their multiple surface functions. Dopamine and L-DOPA from the catecholamine class can act as surfactants and therefore can be used for making core-shell structures, in a single step by using the solvothermal/hydrothermal method of synthesis. Magnetic nanoclusters containing a magnetite core and a polymeric shell synthesized by in situ solvothermal process, using, 3,4-dihydroxybenzhydrazide (DHBH) and poly[3,4-dihydroxybenzhydrazide] (PDHBH) as stabilizers showed biocompatibility, antitumor efficacy and tumor selectivity against colon cancer cells (CACO2), melanoma cells (A375) when used for MH in vitro [159]. Other MNC synthetized by solvothermal polyol reaction and using as a coating dopamine, 3,4-dihydroxybenzylamine, 2-aminomethyl-3-(3,4-dihydroxyphenyl) propionamide and 3,4-dihydroxybenzylidenehydrazine yielded PDA and PDA analogues coating onto the core-shell MNCs. These nanoparticles showed good biocompatibility in normal cells (fibroblasts and endothelial cells) and melanoma (A375), and emitted a fluorescent signal, which can be used for tumor imaging purposes [160].

2.3.2. Natural Polymers

Chitosan, derived from the deacetylation of chitin (found in insect exoskeletons), carries multiple hydroxyl (–OH) and amino (–NH₂) functional groups, which facilitate the binding of antitumor drugs such as paclitaxel (PTX). Chitosan is biocompatible, biodegradable, and antibacterial, making it a leading candidate for nanoscale drug delivery systems. Magnetite nanoparticles coated with a chitosan shell demonstrated an 18% increase in paclitaxel adherence compared to uncoated particles [161,162]. Chitosan also exhibits stimuli-responsive release patterns due to its pH-dependent solubility, and its mucoadhesive properties, while useful in mucosal delivery, can reduce specificity by increasing the risk of off-target accumulation in normal tissues [163].

Dextran-coated IONPs are widely used due to their exceptional biocompatibility, and enhanced magnetic performance, it provides a stabilizing effect inhibiting aggregation and preserving superparamagnetic behavior. Dextran-coated IONPs have been successfully employed in drug delivery applications, with notable effects due to controlled drug release [164,165].

2.4. Combination with Other Therapeutic Methods

Although MH is apt for the induction of cell death by itself, it is particularly effective by sensitizing tumors to other treatments, its results proving significant as an adjunct therapy. For instance, studies point to the ability of MH to increase the proportion of complete responders to radiation therapy by 20 percentage points or more in breast, cervical and head and neck cancers [166], while a 2010 phase III trial in high-risk soft-tissue sarcoma showed that adding regional hyperthermia to chemotherapy nearly doubled the response rate: 28.8% versus 12.7% with chemotherapy alone (p = 0.002) [167].

Combining MH with modalities such as chemotherapy, radiotherapy, immunotherapy, and advanced drug carriers or natural compounds has shown synergistic efficacy in preclinical studies and emerging clinical applications [168]. IONPs are a central focus due to their biocompatibility and strong heat induction, but similar combination approaches are being explored with other nanoparticle types as well (e.g. gold nanostructures for photothermal therapy or high-Z radiosensitizers) [169,170].

Although both combination therapies effectively induce the death of tumor cells and stromal components, there is a risk of selecting resistant tumor cell subpopulations [171], such as anastatic cells[172], blebbishield emergency program cells [173], phoenix rising [174,175], CASP3+ islands of cells [176], nuclear expulsion cells [177], or senescence reversal [178] capable of repopulating the tumor. However, due to their multi-targeted approach, combination therapies are potentially less likely to promote emergence of resistant cells. Therefore, the most promising advancement in oncotherapy is immunotherapy and the downstream awakening of the body’s own immune defense to tumor antigens released by the MH, or other therapies which can detect and attack distant and dormant tumor cells, eventually leading to a stable antitumor immunity (Figure 2).

2.4.1. Chemotherapy

Moderate hyperthermia increases tumor blood flow and cell membrane permeability, improving drug delivery to the tumor and potentiating drug uptake [179]. Furthermore, heating can interfere with DNA repair and induce apoptosis, making cancer cells more susceptible to chemotherapeutic agents. In practice, IONPs have been engineered as dual hyperthermia and drug-delivery agents: they can be loaded or coated with chemotherapeutics (like doxorubicin or paclitaxel) and then remotely heated to simultaneously release drugs and damage cancer cells. For example, in a murine model, intratumoral injection of doxorubicin-conjugated iron oxide nanoparticles in a breast cancer xenograft led to greater tumor regression under an AMF than either MHT alone or doxorubicin alone, indicating a potentiating effect of the combination [180]. Similarly, paclitaxel-loaded magnetic nanoparticles have shown synergistic efficacy with MHT: one study reported both in vitro and in vivo results that supported the conclusion that paclitaxel-bearing IONPs under AMF produced significantly higher cancer cell death and tumor growth inhibition compared to either paclitaxel or hyperthermia alone [181]. This combined approach can reduce the required drug dose (mitigating systemic side effects) while achieving enhanced tumor response (Figure 3).

Thermosensitive carriers can further refine chemo-hyperthermia synergy. For instance, magneto-liposomes or polymer-encapsulated IONPs can be designed to release a payload when the local temperature rises during MHT. In one design, thermosensitive magnetic liposomes loaded with doxorubicin, and a cell-penetrating peptide achieved targeted drug release upon heating and significantly improved therapeutic outcomes both in vitro and in an MCF-7 xenograft murine model [182]. Thus, magnetic IONPs offer a promising prospect in the shape of thermo-chemotherapy, including noninvasive control and deep tissue penetration of the activating magnetic field, making them especially suitable for treating hard-to-reach tumors [30].

2.4.2. Radiotherapy

Heat can radiosensitize tumor cells through several mechanisms. Hyperthermia induces protein and DNA damage and can impede the repair of radiation-induced DNA breaks, thereby increasing radiation efficacy. It also improves tumor oxygenation by increasing blood flow at mild heating, countering hypoxia-driven radio resistance, until higher temperatures cause vascular collapse and direct cell killing [166]. Furthermore, while traditional hyperthermia techniques struggled to heat deep or irregularly shaped tumors uniformly [12], magnetic hyperthermia via IONPs offers a more targeted solution: IONPs can be delivered into the tumor (systemically or via direct injection) and then heated in situ by an external magnetic field, focusing thermal damage within the tumor and sparing surrounding tissue [166].

In human glioblastoma xenograft models, for instance, adjuvant magnetic hyperthermia elevated tumor cell DNA damage (γ-H2AX levels) and apoptosis when combined with radiation, translating into delayed tumor growth and longer survival in treated animals [179]. These benefits are being explored clinically. In a single-arm pilot clinical study, intratumoral MHT plus radiotherapy was tested in patients with recurrent glioblastoma multiforme. Iron oxide nanoparticles (aminosilane-coated magnetite) were injected into the tumor and an AMF applied to produce heating alongside fractionated RT. The combined treatment was found to be safe and led to prolonged survival in these patients compared to historical controls (median overall survival ~13–14 months after recurrence, which was notably longer than with radiotherapy alone) [183]. This approach has since received regulatory approval in Europe as an adjunct therapy for glioblastoma, validating the potential of clinically exploiting the properties of magnetite nanoparticles [184].

2.4.3. Immunotherapy

MH can trigger immunogenic cell death (ICD), through the release of DAMPs such as ATP, HMGB1, and calreticulin that activate dendritic and T cells, turning dying tumor cells into immunostimulants. Intracellular heating from IONPs induced broader ICD marker expression than external heating, highlighting MH’s unique immunological effect [185]. Thus, MH subsumes mechanisms which, beside local tumor control, offer a pathway to systemic immune activation. Recent reviews highlight that nanoparticle-mediated hyperthermia boosts tumor immunogenicity and permeability, thereby enhancing immune cell infiltration and increases responsiveness to checkpoint inhibitors [185,186]. In one model, MH combined with anti-CTLA-4 therapy suppressed both primary and metastatic tumors, inducing long-term immune memory [187]. Other studies showed that iron in superparamagnetic iron-oxide nanoparticle can shift macrophages to a tumor-suppressive phenotype [188]. Combining MH with immunotherapy may also reduce immune escape by increasing antigen presentation and induce inflammation into the tumor microenvironment [189,190]. Although still preclinical, these findings suggest MH may help turn immunologically “cold” tumors into “hot,” responsive ones.

2.4.4. Role of Natural Compounds and Polymer-Based Carriers in MH

Polymer and natural-compound–based carriers enhance IONPs function through stability, targeted delivery, and stimuli-responsive payload release. For example, quercetin-loaded chitosan-coated magnetic nanoparticles improved stability and tumor targeting in colon cancer models [191]. Similarly, polysaccharide-based magnetic hydrogels – such as chitosan-alginate combined with PNIPAM – exhibited efficient on-demand drug delivery and controlled release under AMF heating [192]. These systems improved both therapeutic specificity and systemic toxicity, making them promising platforms for combining hyperthermia and pharmacotherapy.

Polymer coatings like chitosan and PEG improve IONP stability, circulation, and drug loading. Chitosan-coated IONPs, for example, loaded ~3.2 mg doxorubicin/mg nanoparticle – six times more than uncoated versions – yielded higher in vitro cancer cell death [193]. PEGylated magnetite with folate or peptide ligands improved tumor uptake and PEG-stability [194]. These systems can also reduce premature clearance and minimize systemic toxicity.

Natural compounds may also synergize with MH. One PLA–PEG–curcumin–Fe₃O₄ formulation allowed AMF-triggered curcumin release and tumor shrinkage in vivo, outperforming curcumin or hyperthermia alone [195]. Other natural agents (e.g. resveratrol) and biodegradable hydrogels further expand MHT’s versatility by enabling controlled release and retention at specific sites [192,196,197,198]. These strategies added multifunctionality while maintaining biocompatibility.

Another strategy for designing specific nanoparticles revolves around targeting specific receptors on the cellular membrane, leading to highly specific therapeutic benefits. Using an engineered antibody fragment, Christian Ndong and his team managed to target folate receptor alpha overexpressing cancer cells, which leading to high accumulation intracellularly [199]. A similar formulation was used by Monty Liong, achieving drug delivery, magnetic resonance, fluorescence imaging and cell targeting with the same formulation [200]. Formulations targeting the transferrin receptor have also been used [201,202].

2.5. Experimental Studies of Biocompatibility and Oncologic Efficiency of IONPs In Vitro

2.5.1. In Vitro Cancer Models Used for Testing of IONPs

Modern in vitro models are being used more frequently to explore novel approaches for cancer detection and treatment, including the use of iron oxide nanoparticles. These models range from 2D cell cultures, where cancer cells are grown in flat monolayers to more advanced 3D systems like tumor spheroids and organoids, which better mimic the tumor microenvironment. In addition, more complex simulated models are also employed (Table 1).

Certain 2D models have been widely adopted and have produced promising outcomes while remaining reproducible and accurate. Therefore, cancer cell lines, whether human: glioblastoma [203], lung cancer [204], breast cancer [205], cervical cancer [206], pancreatic cancer [200], hepatic cancer [207]; or murine: breast cancer/colon carcinoma [208], fibrosarcoma [209], were used to test nanoparticle cytotoxicity, cellular uptake, and drug delivery efficacy. Fibroblast-like cell lines from humans [203] and mice [209,210]) were also employed, primarily as controls. Bacteria such as Staphylococcus aureus, Proteus vulgaris, and Pseudomonas aeruginosa were used as models for the testing of the antibacterial activity of MNPs [204].

3D tumor models are generally considered more accurate than monolayer-based systems in replicating tumor physiology and predicting the response to chemotherapeutic agents. A study conducted on breast cancer spheroids revealed that MCF-7 spheroids exhibited considerable heterogeneity, with notable differences in spheroid morphology. This variability suggests that these spheroids may not be ideal for evaluating the cytotoxicity or resistance of anticancer drugs [205]. Porcine aortic endothelial cells (PAEC) were exposed to superparamagnetic iron oxide nanoparticles to assess reactive oxygen species (ROS) levels, cytoskeletal structure, and cell stiffness, yielding significant and consistent results [211].

Norouzi et al. developed an MDCK-MDR1-GBM co-culture model to replicate the human blood-brain barrier (BBB) and GBM tumor interface. As a result, the MDCK-MDR1 layer, which consisted of kidney epithelial cells genetically engineered to overexpress the human MDR1 gene, mimicked the BBB, whereas the GBM layer, which comprised human glioblastoma U251 cells, was used to assess nanoparticle uptake and cytotoxicity [212].

Table 1. Biocompatibility and anti-cancer efficacy of iron nanoparticles w/wt MH in vitro.

Nanoparticles	Model	Main results
IONPs with PEG coating/IONPs with PEI coating	SKOV-3 human ovarian cancer/ RAW 264.7 murine macrophages	Cytotoxic effects by ROS production and apoptosis induction [156]
SPIONs loaded with curcumin, coated with poly (lactic-co-glycolic acid)-poly (ethylene glycol) di-block copolymer (PLGA-b-PEG)conjugated with glycine-arginine-glycine-aspartic acid-serine (GRGDS)	T98G- glioblastoma multiforme, fibroblast -like cell line	Induced cytotoxic effects increased by exposure to radiofrequency hyperthermia application [203]
IONPs	A549 human lung cancer cell line Staphylococcus aureus, Proteus vulgaris, Pseudomonas aeruginosa	Cytotoxic effect Antibacterial effect through ROS generation [204]
SPIONS- functionalized with SDS and loaded with curcumin and coated with chitosan SPIONs-SDS-CU-CHIT	HeLAa human cervical cancer	Decreased viability in a dose and time related manner related to drug release in the medium [206]
green iron nanoparticles (Rosemary-FeNPs)	4T1 murine breast cancer C26 cancer cell lines	Cytotoxic effect against cancer cells, efficient intracellular delivery of the rosemary flavonoid components [208]
Bare superparamagnetic iron oxide nanoparticles (SPIONs)	Porcine aortic endothelial cells (PAEC)	ROS formation leads to morphological changes and forms actin stress fibers; blocking ROS formation by functionalization could increase medical applications [211]
IONPs coated with chitosan IONPs coated with polyvinyl alcohol (PVA)	Human fibroblasts	IONPs coated with chitosan induced mild toxicity, IONPs coated with PVA were well tolerated [222]
ferumoxytol carboxymethyldextran coating	mammary adeno carcinoma cells incubated with macrophages	Macrophages showed pro-inflammatory M1 phenotype upon ferumoxytol exposure Increased caspase -3 in mammary tumor cells [305]
IONPs load with LLY-507 (inhibitor of SMYD2), coated with PVA	A549 human non-small cell lung cancer cell line RBC- human	Efficient delivery of the SMYD2 inhibitor by the IONPs , dose dependent decrease in viability, hemolysis below 5%[306]
poly(ethylene glycol)-block-poly(lactic-co-glycolic acid) copolymer-encapsulated Fe3O4 superparticles (SPs), loaded with imiquimod (R837 a Toll-like receptor 7 agonist)	4T1 triple negative human breast cancer cells	Efficient photothermal ablation of 4T1 cells by apoptosis/necrosis upon PTT irradiation, efficient delivery of R837 in vivo against primary tumors to enhance immune response [307]
Fe3O4@PDA SPs	HeLa human cervical cancer cell line, mice bearing tumor (HeLa)	Biocompatible, increased efficacy of photothermal therapy against tumors in vivo [308]
IONPs - loaded with curcumin and coated with dextran CUR/DEX/Fe3O4-NPs	MCF-7 human breast cancer	Decreased cell viability in a dose and time related manner [309]
SPIONS- functionalized with SDS and loaded with curcumin and coated with chitosan SPIONs-SDS-CU-CHIT	HeLAa human cervical cancer	Decreased viability in a dose and time related manner related to drug release in the medium [310]

Table 1. Biocompatibility and anti-cancer efficacy of iron nanoparticles w/wt MH in vitro.

Nanoparticles	Model	Main results
IONPs with PEG coating/IONPs with PEI coating	SKOV-3 human ovarian cancer/ RAW 264.7 murine macrophages	Cytotoxic effects by ROS production and apoptosis induction [156]
SPIONs loaded with curcumin, coated with poly (lactic-co-glycolic acid)-poly (ethylene glycol) di-block copolymer (PLGA-b-PEG)conjugated with glycine-arginine-glycine-aspartic acid-serine (GRGDS)	T98G- glioblastoma multiforme, fibroblast -like cell line	Induced cytotoxic effects increased by exposure to radiofrequency hyperthermia application [203]
IONPs	A549 human lung cancer cell line Staphylococcus aureus, Proteus vulgaris, Pseudomonas aeruginosa	Cytotoxic effect Antibacterial effect through ROS generation [204]
SPIONS- functionalized with SDS and loaded with curcumin and coated with chitosan SPIONs-SDS-CU-CHIT	HeLAa human cervical cancer	Decreased viability in a dose and time related manner related to drug release in the medium [206]
green iron nanoparticles (Rosemary-FeNPs)	4T1 murine breast cancer C26 cancer cell lines	Cytotoxic effect against cancer cells, efficient intracellular delivery of the rosemary flavonoid components [208]
Bare superparamagnetic iron oxide nanoparticles (SPIONs)	Porcine aortic endothelial cells (PAEC)	ROS formation leads to morphological changes and forms actin stress fibers; blocking ROS formation by functionalization could increase medical applications [211]
IONPs coated with chitosan IONPs coated with polyvinyl alcohol (PVA)	Human fibroblasts	IONPs coated with chitosan induced mild toxicity, IONPs coated with PVA were well tolerated [222]
ferumoxytol carboxymethyldextran coating	mammary adeno carcinoma cells incubated with macrophages	Macrophages showed pro-inflammatory M1 phenotype upon ferumoxytol exposure Increased caspase -3 in mammary tumor cells [305]
IONPs load with LLY-507 (inhibitor of SMYD2), coated with PVA	A549 human non-small cell lung cancer cell line RBC- human	Efficient delivery of the SMYD2 inhibitor by the IONPs , dose dependent decrease in viability, hemolysis below 5%[306]
poly(ethylene glycol)-block-poly(lactic-co-glycolic acid) copolymer-encapsulated Fe3O4 superparticles (SPs), loaded with imiquimod (R837 a Toll-like receptor 7 agonist)	4T1 triple negative human breast cancer cells	Efficient photothermal ablation of 4T1 cells by apoptosis/necrosis upon PTT irradiation, efficient delivery of R837 in vivo against primary tumors to enhance immune response [307]
Fe3O4@PDA SPs	HeLa human cervical cancer cell line, mice bearing tumor (HeLa)	Biocompatible, increased efficacy of photothermal therapy against tumors in vivo [308]
IONPs - loaded with curcumin and coated with dextran CUR/DEX/Fe3O4-NPs	MCF-7 human breast cancer	Decreased cell viability in a dose and time related manner [309]
SPIONS- functionalized with SDS and loaded with curcumin and coated with chitosan SPIONs-SDS-CU-CHIT	HeLAa human cervical cancer	Decreased viability in a dose and time related manner related to drug release in the medium [310]

2.5.2. MNP Formulation

The primary type of iron-based nanoparticles used in cancer therapy are iron oxide nanoparticles (IONPs), known for their magnetic properties and biocompatibility. Among these, superparamagnetic iron oxide nanoparticles (SPIONs) are particularly significant, as they facilitate magnetic drug targeting, MH, and function as contrast agents in MRI [200]. There are two main approaches in the study of iron nanoparticles used in cancer therapy research, (1) functionalized bare nanoparticles, evaluated for their biocompatibility, tumor-targeting ability, and cytotoxic effects w/wt MH and (2) drug-loaded nanoparticles, designed for delivering chemotherapeutic agents, photosensitizers, or for use in combined MH and chemotherapy [217].

Natural compounds rich in antioxidants serve a dual purpose in nanoparticle synthesis: they act as reducing agents that promote nanoparticle formation and prevent aggregation, and they function as coatings that enhance the biocompatibility of magnetic nanoparticles (MNPs) by forming a natural antioxidant shell. Additionally, these compounds facilitate the targeted delivery of bioactive agents into tumor cells. Many cancer cells are particularly sensitive to polyphenols, resveratrol, flavonoids or anthocyanins [213] present in natural extracts, which can trigger apoptosis or increase the cells’ responsiveness to other treatments, such as chemotherapy, photodynamic therapy (PDT), or hyperthermia. Natural compounds, such as polyphenols, synergize with anticancer drugs like cisplatin, doxorubicin, and 5-fluorouracil [214]. Rosemary terpenes also showed antitumoral activity on colon cancer cells in vitro inducing necrosis by an acute ROS increase and on in vivo colon cancer models, they inhibit proliferation and increased animal survival [215]. Green iron nanoparticles (Rosemary-FeNPs) phyto synthesized by using the natural antioxidants from the rosemary extract showed an average diameter range of 50-100 nm and excellent homogenization [208]. Turmeric extracts and their key compounds, carnosic acid and curcumin also showed antiproliferative effects of cancer cells [216]. SPIONs loaded with curcumin and coated with organic polymers, poly (lactic-co-glycolic acid)-poly (ethyleneglycol) di-block copolymer (PLGA-b-PEG) conjugated with glycine-arginine-glycine-aspartic acid-serine (GRGDS). GRGDS peptide has been found to allow targeting of integrins, typically overexpressed in cancer cells, moreover the combined delivery of curcumin enhanced therapy efficiency and can serve as a drug delivery platform for a chemotherapeutic in view of a synergistic effect [203]. Multiple chemotherapeutic drugs, such as paclitaxel (PTX) [200,217], doxorubicin (DOX) [207,212,218], camptothecin (CPT) [200], gemcitabine [219], sorafenib [220], temozolomide [179] were loaded on magnetic nanoparticles. Despite its efficiency as an antitumor drug, Paclitaxel (PTX) administration is difficult due to its hydrophobic nature. To address this issue, MNPs loaded with PTX were synthesized using chitosan as coating: Fe₃O₄@LaF₃: Ce^3+,Tb^3+/chi NPs associated with Paclitaxel (PTX). This formulation increased the water-solubility, of PTX while preserving the superparamagnetic behavior of the MNPs, and provided a biocompatible method for Paclitaxel administration with reduced side-effects and possibility for a synergic therapy using MH and chemotherapy (Table 2) [217].

Folic acid receptors are overexpressed in many cancers; therefore folic acid can be used for selective targeting of the malignant cells cancers including ovary, kidney, uterus, testis, brain, colon, lung, myelocytic blood cells [221]. This strategy was used for the synthesis of PTX-loaded nanoparticles, such as SPIONs with PTX-loaded chitosan (Cs), polyethylene glycol (PEG), and folic acid (FA), and yielded improved tumor targeting and PTX uptake in the malignant cells [209]. A study on SPIONs- PLGA core / poly(N-isopropylacrylamide)-carboxymethyl chitosan shell with NU7441/Gemcitabine found that targeting folate receptors increased specific uptake, while the pH-sensitive shell ensured gemcitabine was preferentially released in the tumor environment. The nanoplatform retained magnetic properties, which make it suitable for combined MH and chemotherapy combined therapy [219]. Another coating, such as lauric acid (LA) and human serum albumin (HSA) was used for the synthesis of SPIONs (SPION-LA-HSA-Ptx) for the delivery of PTX. The presence of lauric acid improved the PTX hydrophobic drug loading and nanoparticle stability, while LA and HSA increased MNPs biocompatibility and colloidal stability [205]. Magnetic nanoparticles coated with an amphiphilic polymer containing disulphide linkages (hyaluronic acid-disulfide bond-polylactic acid) loaded with PTX showed efficient drug delivery by combining magnetic tumor targeting and redox-triggered specific release of paclitaxel, leading to improved therapeutic efficacy and minimizing side effects [206]. Mesoporous silica-coated SPIONs with fluorescent dyes, hydrophilic groups, cancer-specific targeting ligands, and co-loaded with camptothecin (CPT) and paclitaxel (PTX) showed advantages in magnetic manipulation, targeted drug delivery, and efficient drug loading and release [200].

A54-Dex-PLGA micelles with DOX exhibited strong encapsulation efficiency (approximately 80%) and prolonged release (up to 72 hours). MNPs demonstrated tumor targeting and enhanced efficacy compared to free medication [207]. Doxorubicin-loaded IONPs with surface coatings such as trimethoxysilylpropyl-ethylenediamine triacetic acid (EDT) were also effective, as EDT coating had a significant impact on blood-brain barrier penetration. Furthermore, this formulation achieved sustained DOX release, with quicker release in acidic conditions (tumor microenvironment), allowing for more tailored therapeutic action [212]. The DOX-loaded Fe₃O₄@MnO₂@PPy nanocomposite improved hypoxia tolerance and PDT efficiency by integrating photothermal, photodynamic, and chemotherapeutic treatments [218].

SPIONs synthesized by using a double coating of polyvinyl alcohol (PVA) or polyethylene glycol (PEG) and magnesium–aluminum-layered double hydroxide (MLDH) were loaded with Sorafenib. The resulting nanoparticles were spherical, with an average diameter of 17nm and released sorafenib over a period of 72 hours, more effectively when exposed to an acidic pH (4.8), simulating tumor microenvironment. This system showed increased toxicity towards HepG2 hepatoma cells and decreased towards fibroblast 3T3 cells, which served as controls, compared to sorafenib [220].

Table 2. Biocompatibility and anti-cancer efficacy of iron nanoparticles w/wt MH in vitro.

PACLITAXEL
Nanoparticles	Model	Main results
Multifunctional mesoporous silica nanoparticles SPIONs Surface modifications: Fluorescent dye molecules/ Hydrophilic groups / Cancer-specific targeting ligands – folate (FA), Drugs: Camptothecin (CPT) /Paclitaxel (PTX)	Human pancreatic cancer cell lines: PANC-1, BxPC3, human foreskin fibroblasts (HFF) as control	Selective cytotoxicity; dual imaging capability; targeted drug delivery through ligands (FA) [200]
SPIONs coated with lauric acid and human serum albumin as carriers for paclitaxel (SPION-LA-HSA-Ptx)	human breast cancer cell lines (T-47D, BT-474, MCF-7, and MDA-MB-231 cells)	High potential for magnetically targeted drug delivery in breast cancer Similar effects on human breast cancer as PTX alone [205]
SPION@Cs-PTX-PEG-FA SPIONs with paclitaxel (PTX)-loaded chitosan (Cs), polyethylene glycol (PEG) and receptors that target folate (FA)	WEHI-164: Mouse fibrosarcoma; MEF: Mouse embryonic fibroblast (normal) cell line	High nanoparticle stability, selective uptake, reduced systemic toxicity due to the FA receptors, apoptosis of cancer cells [209]
Fe3O4@LaF3:Ce3+,Tb3+/chi NPs bonded with Paclitaxel (PTX)	A549 human lung cancer cell line	Increased cell toxicity compared to free paclitaxel; efficient imaging (MRI and fluorescence imaging); reduced side effects [217]
MNPs coated with an amphiphilic polymer containing disulfide linkages (Hyaluronic Acid–disulfide bond–Polylactic Acid, HA-SS-PLA), loaded with PTX	HeLa cells human cervical cancer cell line)	Targeted delivery, through magnetism and redox response; improved cytotoxicity, and biocompatibility [311]
DOXORUBICIN
A54 peptide-functionalized poly(lactic-co-glycolic acid)-grafted dextran (A54-Dex-PLGA) micelles with DOX/ SPIO	BEL-7402, HepG2 hepatic cancer	MNPs easy synthesis of SPIO, low off-target distribution and toxicity; controlled drug release; dual imaging/ therapy function [207]
Electro-spun fibers co-loaded with magnetic IONPs, cubic shaped loaded with doxorubicin	Mouse embryonic fibroblast cell line (NIH 3 T3 cells), DOX-sensitive HeLa-WT cervical cancer cells and the DOX-resistant MCF7 breast cancer cells	Hyperthermia combined with enhanced diffusion of doxorubicin - effective oncotherapy [210]
Doxorubicin-loaded IONPs with surface coatings like trimethoxysilylpropyl-ethylenediamine triacetic acid (EDT)	MDCK-MDR1-GBM co-culture model	High DOX penetration through BBB; effective magnetic targeting and reduced systemic toxicity; possibly overcoming MDR cancer cells [212]
Fe₃O₄@MnO₂@PPy nanocomposite loaded with DOX; Fe₃O₄ (Iron oxide) core; MnO₂ (Manganese dioxide) shell; PPy (Polypyrrole) outer layer	human hepatoma (HepG2)	Good magnetic targeting delivery and enhanced cancer toxicity improved photodynamic (PDT) /photothermal therapy (PTT) reduced side effects and better tolerance to hypoxia induced by PDT/PTT [218]
IONP DOX: PEG-coated, doxorubicin-loaded nanoparticles	HeLa cells (human cervical cancer cell line)	Delivery of DOX directly into the cytoplasm trough macro pinocytosis and endocytosis; high biocompatibility [223]
PEG-coated Fe3O4 luteininzing hormone-releasing hormone (LHRH) ligand containing doxorubicin	A549 and MCF-7 cancer cells	Theranostic nanoparticle formulation using LHRH ligand with individual chemotherapy and thermotherapy, effective on both cell lines [312]
OTHERS
Magnetic IONPs/ temozolomide	SD3, G-16, G-302, GL-261 cell lines	Combined hyperthermia using magnetic IONPs with temozolomide and radiation showed synergistic anti-glioblastoma effects [179]
SPIONs- PLGA core / poly(N-isopropylacrylamide)-carboxymethyl chitosan shell with NU7441/Gemcitabine	A549 and H460 lung cancer cells	Approach for simultaneous radiotherapy and chemotherapy, Folate receptor targeting increased specific uptake [219]
SPIONs- (PVA/LDH-coated and PEG/LDH-coated) with Sorafenib	HepG2 human hepatoma/ 3T3 mouse fibroblast cell line	Strong superparamagnetic behavior; enhanced anticancer activity and selectivity; minimal side effects [220]
Magnetic-core silica nanoparticles with nano valves and loaded with cucurbituril	MDA-MB-231 breast cancer cells	Targeted delivery using a nano valve system and hyperthermia [313]
Fe-NP2 coated with PEI conjugated with cisplatin (IV) prodrug	Human ovarian carcinoma A2780 cells / cisplatin-resistant A2780DDP cells	Efficient drug delivery overcoming cisplatin resistance through unique internalization pathway of nanoparticles/ increased production of ROS [314]
Phospholipid-modified Pt(IV) prodrug-loaded IONP-filled micelles	B16-F10 melanoma cells	Redox-triggered release of cisplatin, ferroptosis of melanoma cells, lower concentration threshold, lymphatic delivery [315]
Nanoflowers MoS2@Fe3O4- loaded with ICG/Pt(IV) indocyanine green (ICG) and platinum (IV) prodrugs {c,c,t-Pt(NH3)2Cl2(OOCCH2CH2COOH)2}	I.929 fibroblasts, HeLa, H22 tumor-bearing Balb/c mice	Biocompatible, theranostics bioimaging capabilities and laser-induced cytotoxicity [316]
Fe(Salen) nanoparticles with μ-oxo N,N′-bis (salicylidene) ethylene diamine	tongue cancer VX2 (rabbit), HSC-3 (human), and OSC-19 (human)	Hyperthermia-guided, temperature stable cytotoxic effects, even at low concentrations [317]

Table 2. Biocompatibility and anti-cancer efficacy of iron nanoparticles w/wt MH in vitro.

PACLITAXEL
Nanoparticles	Model	Main results
Multifunctional mesoporous silica nanoparticles SPIONs Surface modifications: Fluorescent dye molecules/ Hydrophilic groups / Cancer-specific targeting ligands – folate (FA), Drugs: Camptothecin (CPT) /Paclitaxel (PTX)	Human pancreatic cancer cell lines: PANC-1, BxPC3, human foreskin fibroblasts (HFF) as control	Selective cytotoxicity; dual imaging capability; targeted drug delivery through ligands (FA) [200]
SPIONs coated with lauric acid and human serum albumin as carriers for paclitaxel (SPION-LA-HSA-Ptx)	human breast cancer cell lines (T-47D, BT-474, MCF-7, and MDA-MB-231 cells)	High potential for magnetically targeted drug delivery in breast cancer Similar effects on human breast cancer as PTX alone [205]
SPION@Cs-PTX-PEG-FA SPIONs with paclitaxel (PTX)-loaded chitosan (Cs), polyethylene glycol (PEG) and receptors that target folate (FA)	WEHI-164: Mouse fibrosarcoma; MEF: Mouse embryonic fibroblast (normal) cell line	High nanoparticle stability, selective uptake, reduced systemic toxicity due to the FA receptors, apoptosis of cancer cells [209]
Fe3O4@LaF3:Ce3+,Tb3+/chi NPs bonded with Paclitaxel (PTX)	A549 human lung cancer cell line	Increased cell toxicity compared to free paclitaxel; efficient imaging (MRI and fluorescence imaging); reduced side effects [217]
MNPs coated with an amphiphilic polymer containing disulfide linkages (Hyaluronic Acid–disulfide bond–Polylactic Acid, HA-SS-PLA), loaded with PTX	HeLa cells human cervical cancer cell line)	Targeted delivery, through magnetism and redox response; improved cytotoxicity, and biocompatibility [311]
DOXORUBICIN
A54 peptide-functionalized poly(lactic-co-glycolic acid)-grafted dextran (A54-Dex-PLGA) micelles with DOX/ SPIO	BEL-7402, HepG2 hepatic cancer	MNPs easy synthesis of SPIO, low off-target distribution and toxicity; controlled drug release; dual imaging/ therapy function [207]
Electro-spun fibers co-loaded with magnetic IONPs, cubic shaped loaded with doxorubicin	Mouse embryonic fibroblast cell line (NIH 3 T3 cells), DOX-sensitive HeLa-WT cervical cancer cells and the DOX-resistant MCF7 breast cancer cells	Hyperthermia combined with enhanced diffusion of doxorubicin - effective oncotherapy [210]
Doxorubicin-loaded IONPs with surface coatings like trimethoxysilylpropyl-ethylenediamine triacetic acid (EDT)	MDCK-MDR1-GBM co-culture model	High DOX penetration through BBB; effective magnetic targeting and reduced systemic toxicity; possibly overcoming MDR cancer cells [212]
Fe₃O₄@MnO₂@PPy nanocomposite loaded with DOX; Fe₃O₄ (Iron oxide) core; MnO₂ (Manganese dioxide) shell; PPy (Polypyrrole) outer layer	human hepatoma (HepG2)	Good magnetic targeting delivery and enhanced cancer toxicity improved photodynamic (PDT) /photothermal therapy (PTT) reduced side effects and better tolerance to hypoxia induced by PDT/PTT [218]
IONP DOX: PEG-coated, doxorubicin-loaded nanoparticles	HeLa cells (human cervical cancer cell line)	Delivery of DOX directly into the cytoplasm trough macro pinocytosis and endocytosis; high biocompatibility [223]
PEG-coated Fe3O4 luteininzing hormone-releasing hormone (LHRH) ligand containing doxorubicin	A549 and MCF-7 cancer cells	Theranostic nanoparticle formulation using LHRH ligand with individual chemotherapy and thermotherapy, effective on both cell lines [312]
OTHERS
Magnetic IONPs/ temozolomide	SD3, G-16, G-302, GL-261 cell lines	Combined hyperthermia using magnetic IONPs with temozolomide and radiation showed synergistic anti-glioblastoma effects [179]
SPIONs- PLGA core / poly(N-isopropylacrylamide)-carboxymethyl chitosan shell with NU7441/Gemcitabine	A549 and H460 lung cancer cells	Approach for simultaneous radiotherapy and chemotherapy, Folate receptor targeting increased specific uptake [219]
SPIONs- (PVA/LDH-coated and PEG/LDH-coated) with Sorafenib	HepG2 human hepatoma/ 3T3 mouse fibroblast cell line	Strong superparamagnetic behavior; enhanced anticancer activity and selectivity; minimal side effects [220]
Magnetic-core silica nanoparticles with nano valves and loaded with cucurbituril	MDA-MB-231 breast cancer cells	Targeted delivery using a nano valve system and hyperthermia [313]
Fe-NP2 coated with PEI conjugated with cisplatin (IV) prodrug	Human ovarian carcinoma A2780 cells / cisplatin-resistant A2780DDP cells	Efficient drug delivery overcoming cisplatin resistance through unique internalization pathway of nanoparticles/ increased production of ROS [314]
Phospholipid-modified Pt(IV) prodrug-loaded IONP-filled micelles	B16-F10 melanoma cells	Redox-triggered release of cisplatin, ferroptosis of melanoma cells, lower concentration threshold, lymphatic delivery [315]
Nanoflowers MoS2@Fe3O4- loaded with ICG/Pt(IV) indocyanine green (ICG) and platinum (IV) prodrugs {c,c,t-Pt(NH3)2Cl2(OOCCH2CH2COOH)2}	I.929 fibroblasts, HeLa, H22 tumor-bearing Balb/c mice	Biocompatible, theranostics bioimaging capabilities and laser-induced cytotoxicity [316]
Fe(Salen) nanoparticles with μ-oxo N,N′-bis (salicylidene) ethylene diamine	tongue cancer VX2 (rabbit), HSC-3 (human), and OSC-19 (human)	Hyperthermia-guided, temperature stable cytotoxic effects, even at low concentrations [317]

2.5.3. Eficiency, Side Effects

Iron nanoparticles have been widely studied and applied in cancer therapy due to their effectiveness in targeted treatment, imaging, and hyperthermia, along with their generally favorable biocompatibility.

MNPs cytotoxic effects may arise either through hyperthermia-induced mechanisms [203] or from the intrinsic properties of the nanoparticles themselves [204]. Functionalization has been shown to enhance both cytotoxicity against cancer cells and intracellular targeting. MNPs incorporating rosemary flavonoid compounds demonstrated improved efficacy [208]. Tran et al showed that chitosan or PVA coating of iron oxide nanoparticles reduced the cell toxicity towards normal mouse fibroblasts, with PVA coating having a better result and also reduced nanoparticle aggregation, underscoring the role of surface coating for biocompatibility [222].

The generation of reactive oxygen species (ROS) has been identified as a key mechanism driving the biological activity and toxicity of nanoparticles, contributing not only to their antibacterial effects [204] but also to cellular morphological changes and the formation of actin stress fibers [211]. Overall, the use of iron nanoparticles loaded with various anticancer agents has proven beneficial, improving therapeutic effectiveness and targeting specificity while minimizing adverse side effects. MNPs loaded with PTX demonstrated enhanced tumor cell toxicity compared to free paclitaxel, with less side effects and increased imaging properties, by both MRI and fluorescence imaging [217], induced apoptosis in cancer cells [209], demonstrated a high potential of PTX-loaded SPIONs for magnetically based targeted drug delivery in breast cancer, but the effects were similar to those of PTX alone [205]. PTX-loaded MNPs were used for targeted administration, which combined magnetic drug delivery with redox dependent release to increase cytotoxicity [206].

DOX-loaded nanoparticles also demonstrated significant potential for cancer treatment. Doxorubicin-loaded IONPs with surface coatings such as trimethoxysilylpropyl-ethylenediamine triacetic acid (EDT) were able to overcome MDR cancer cells in a GBM model, combined with magnetic targeting and low systemic toxicity [212]. Popescu et al. presented direct delivery of DOX loaded nanoparticles into the cytoplasm via macropinocytosis and endocytosis, with promising future possibilities [223]. Fe₃O₄@MnO₂@PPy nanocomposites used DOX-loaded nanoparticles to deliver the chemotherapeutic and enhance two types of phototherapy (PDT and PTT) at the tumor site, resulting in more effective cancer treatment [218].

2.5.4. Type of Cell Death

• Apoptosis

Nanoparticles, depending on their dose and physicochemical properties, can influence various cell fates (Figure 4), including necrosis and apoptosis [224]. Apoptosis is a distinct type of cell death characterized by particular morphological changes, such as membrane blebbing, cell shrinkage, chromatin condensation, and tiny vesicles (apoptotic bodies) [225]. The most important apoptosis mechanisms occur via three different pathways involving death receptors, mitochondria, or the endoplasmic reticulum, with caspases mediating all morphological and biochemical changes [226,227]. As a result, multiple studies indicate the necessity of linking nanoparticle dosage and exposure time with apoptotic intensity, such as the one published by Naqvi et al., who examined SPIONs coated with Tween 80 on murine macrophage (J774) cells. Cell viability was higher at lower concentrations (25-200 μg/mL) and up to three hours of exposure but decreased to 55-65% at higher concentrations (300-500 μg/mL) and longer exposure (six hours). According to the same study, apoptosis was the main registered cause of cell death, with oxidative stress serving as the primary toxicity mechanism. [228] Functionalized iron oxide nanoparticles, particularly those conjugated with therapeutic agents (such as, IONs conjugated with lysine and methotrexate tested on breast cancer), can effectively induce apoptosis in cancer cells in vitro [229]. Tousi et al. found that mPEG-b-PLGA coated IONs loaded with the flavonoid eupatorin increased apoptosis and decreased necrosis in prostate cancer cells compared to free eupatorin or uncoated nanoparticles, suggesting that they could be an effective drug delivery system for cancer therapy [230].

• Necrosis

Necrosis has long been thought to be the outcome of general cell injury caused by trauma. As a result, it is seen as an uncontrolled form of cell death that is not caused by specific signaling processes. Various clinical circumstances, including toxin exposure, ischemia, viral or bacterial infection can cause necrotic cell death [224]. Also in this case, the key mechanism appears to be reactive oxygen species (ROS), as supported by Khan et al. The study reveals that ROS generated by iron oxide nanoparticles lead to necrosis and cell death in lung cancer cells (Figure 4). The type of cell death (necrosis vs. apoptosis) is determined by the level of oxidative stress and the cellular antioxidant capability [231]. Another study found that certain coatings and greater doses of IONPs can cause necrosis, notably PEI-coating, which is known to be cytotoxic and damage the cell membrane [156]. There are just a few in vitro studies that particularly address necrosis caused by iron nanoparticles (figure 4). Most of the studies focus on apoptosis, with necrosis as a secondary consequence at higher doses, frequently related to increased oxidative stress or membrane damage [228,232].

Figure 4. Mechanisms of cell death induced by magnetic hyperthermia. Following cellular uptake, and exposure to an alternating magnetic field (AMF), MHT triggers apoptosis via either the membrane pathway, through death receptor activation, or the mitochondrial pathway, both leading to caspase cascade activation. Higher temperatures, nutrient deprivation, or combined therapies favor necrosis, typically caused by rapid ROS accumulation that overwhelms antioxidant defenses, leading to terminal oxidation of cellular components. ROS can also damage nuclear DNA, leading to necroptosis or activating DNA repair. Ferroptosis results from intracellular iron buildup, which generates ROS through Fenton reactions, and causes lipid peroxidation. This process can culminate in cell death or trigger autophagy, particularly described on in vivo models, with lysosomal activation. Autophagy may enable cell survival by recycling damaged organelles to provide energy and restore cellular functions acting as a tumor escape mechanism. However, when damage is extensive, autophagy serves as a programmed cell death pathway.

Figure 4. Mechanisms of cell death induced by magnetic hyperthermia. Following cellular uptake, and exposure to an alternating magnetic field (AMF), MHT triggers apoptosis via either the membrane pathway, through death receptor activation, or the mitochondrial pathway, both leading to caspase cascade activation. Higher temperatures, nutrient deprivation, or combined therapies favor necrosis, typically caused by rapid ROS accumulation that overwhelms antioxidant defenses, leading to terminal oxidation of cellular components. ROS can also damage nuclear DNA, leading to necroptosis or activating DNA repair. Ferroptosis results from intracellular iron buildup, which generates ROS through Fenton reactions, and causes lipid peroxidation. This process can culminate in cell death or trigger autophagy, particularly described on in vivo models, with lysosomal activation. Autophagy may enable cell survival by recycling damaged organelles to provide energy and restore cellular functions acting as a tumor escape mechanism. However, when damage is extensive, autophagy serves as a programmed cell death pathway.

• Ferroptosis

Ferroptosis represents a type of cellular death due to lethal accumulation of iron intracellularly. Ferroptosis is mainly caused by accumulation of reactive oxygen species, that lead to lipid peroxidation, cellular membrane instability and eventually, cellular death. One of the main pathways in which IONPs act to induce ferroptosis is by intracellular accumulation, internalization in lysosomes which act to dissolve the formulations, leading to release of iron ions intracellularly (figure 4). Ferroptosis is driven by ferrous iron through Fenton reactions, generating reactive oxygen species (ROS, among these, radical hydroxyl, HOž, responsible for lipid peroxidation intensifying) [233], disruption of mitochondria functions [234,235], cellular membrane rupture [236], and other alteration processes at ultrastructural levels (endoplasmic reticulum, peroxisomes, etc.) [237,238]. Antioxidant defenses, through the cysteine/glutamate antiporter and FSP1/ubiquinol systems are overloaded. Ferroptosis represents an attractive target for current and future anticancer therapies as tumor cells are surprisingly susceptible to its effects. [239,240,241]. In vivo, ferroptosis (iron-dependent cell death) was considered a helpful mechanism that may destroy the tumors but the accumulation of IONPs in healthy tissues limited the usage of this therapeutic method. Medication associated with ferroptosis such as Lanperisone, Sorafenib, Trigonelline, Cisplatin, Ferumoxytol etc. is used in cancer [233], as a therapeutic method only affecting tumors. IONPs are especially suitable for inducing ferroptosis through several mechanisms and may act as theranostic, providing multiple capabilities at the same time. Another strategy concentrates on creating synergistic therapies, for example, Qi Nie and her research team used IONPs loaded with paclitaxel (PTX) to increase intracellular concentration of iron ions, with higher ROS formation and confirmed ferroptosis by evaluating cellular upregulation as a response to ferroptosis [242], results that were echoed by a similar article focusing on inhibiting non-small cell lung cancer cells. Ferroptosis appears to be associated with autophagy, as several studies presented this possible correlation [243,244]. Autophagy, the physiological cellular removal and replacement of degraded organelles represents a mechanism that protects against tumor development or destroys healthy cells. When the stressors, like IONPs, accumulate in healthy tissues, intracellular environment is modified toward the production of reactive oxygen species in high concentration that induce damage inside the cells and transforms autophagy into a cancer promoter [245].

2.6. Biocompatibility and Oncologic Efficiency of IONPs In Vivo

In vivo studies were realized for theranostic purposes of magnetic iron oxide nanoparticles (IONPs), superparamagnetic iron oxide nanoparticles (SPIONs), surface-coated IONPs, charged polyvinyl alcohol-coated SPIONs (PVA-coated SPIONs), protein-coated IONPs, SPIONs coated with anti-biofouling polymers etc. Technological advances permit the drug delivery at nanoscale inside the tumor, the small dimensions and coating giving the possibility of their transport even through tumor stromal components.

2.6.1. Biodistribution

IONPs biodistribution depends on several factors such as the nanoparticle’s dimension and shape the type, chemical composition, or electric charges of their coating, properties that make them able to also migrate in healthy tissues. In living organisms, iron oxide nanoparticles may follow different pathways, especially depending on their dimension (Figure 5). Nanoparticles with diameters higher than 100 nm are rapidly captured in spleen and liver, and their penetration inside the tumor is limited by the tumor pathological characteristics of the vessels’ wall, that differ with the cancer type and stage [246]. Wang et al. showed that after oral administration of IONPs with dimensions lower than 100 nm in mice, the liver was exposed to two picks of nanoparticles concentration, in the first day and in the seventh day. These particles also accumulated in other organs, with maximum levels: at 6 hours after gavage (lungs, kidneys), in the first day (stomach, small intestine, bone marrow) and during the first 3 days (heart, spleen, brain) [247].

IONPs with dimensions between 20 nm and 100 nm are preferred for cancer treatments, to avoid IONPs urinary excretion (< 20 nm) and spleen/liver capture (> 100 nm) [248]. Several studies presented their role on vascular permeability. In tumors with reduced vascularization, administration of IONPs coupled with external magnetic field exposure leads to endothelial layer alterations, increasing the accumulation of drugs inside the cancer area [249]. Endothelium permeability may also be increased by IONPs through the oxidative stress that they generate, reorganizing the microtubules position in the cellular cytoskeleton [250].

Small IONPs, with dimensions lower than 20 nm, may pass easily the endothelium toward different organs and may be filtered by the kidneys, processes that occur when they are administered intravenously, orally or when they migrate from the tumor back into the blood flow [251]. Since the endothelial glycocalyx presents 20 nm gaps between proteoglycan chains, the small nanoparticles can pass freely the healthy endothelium layer [252]. At tumor level, endothelial layer develops pores of 100 nm till 1 µm (pores dimensions depend on cancer type and stage) that are passed easily by the small IONPs (<20 nm) [253].

Figure 5. Pharmacokinetics of IONPs in vivo after systemic administration or vascular leakage post-intratumoral injection. Biodistribution of the MNPs strongly depends on their size, with nanoparticles between 20-100 nm diameter being the most suitable for MHT, due to their ability to pass through the endothelial layer into the tissue leading to selective tumor accumulation (EPR effect), that can be enhanced by application of an external magnetic field. Smaller particles (<1 nm) are fast eliminated by the kidney or can cluster into the glomerular cells glycocalyx leading to impaired nephron function, while bigger MNPs (>100nm) can be stored in the internal organs, leading to medium/long term toxicity.

Figure 5. Pharmacokinetics of IONPs in vivo after systemic administration or vascular leakage post-intratumoral injection. Biodistribution of the MNPs strongly depends on their size, with nanoparticles between 20-100 nm diameter being the most suitable for MHT, due to their ability to pass through the endothelial layer into the tissue leading to selective tumor accumulation (EPR effect), that can be enhanced by application of an external magnetic field. Smaller particles (<1 nm) are fast eliminated by the kidney or can cluster into the glomerular cells glycocalyx leading to impaired nephron function, while bigger MNPs (>100nm) can be stored in the internal organs, leading to medium/long term toxicity.

Injected IONPs with sizes lower than 10 nm are excreted in large amounts through urine. Studies showed that more than 40% of administered IONPs were eliminated within 24 hours after administration. PEGylated IONPs with dimensions around 10 nm are transported easily inside the cells, accumulate in high concentrations inside the tumors, but also in spleen and liver where their degradation is realized very slowly (Figure 5). The PEGylated IONPs are toxic at high concentrations and may trigger the autophagy [156].

IONPs with sizes lower than 3 nm can pass through the vessel wall also paracellular [254], while those of around 1 nm are blocked in the glycocalyx of the glomerular filtration membrane, forming nanoclusters that remain for a long time at kidney level [255].

2.6.2. Coating

For drug delivery, IONPs are coated with different materials: natural polymers (dextran, chitosan, starch, etc.) [256], synthetic polymers (polyethylene glycol -PEG; polyvinyl alcohol -PVA; polyvinyl pyrrolidone -PVP; etc.) [256,257], proteins (albumin) [258], lipids [259], silane [260], silica [261] or combination of synergistic materials [262]. To avoid the IONPs agglomeration that can produce embolism inside the capillaries [263] and to prevent the nanoparticles rapid systemic dispersion, amines, aldehydes, thiols or the carboxylic acids are used for IONPs synthesis [264]. Iron nanoparticles can be synthesized with specific coating that may increase the cytotoxicity of the transported drug, like gold-coated IONPs [265], or with coverings that may present a prolonged drug release, even over 4 days after their administration, like hyaluronic acid-coated IONPs [266]. Intraperitoneal administration of gold-coated IONPs in mice with melanoma tumor leaded to nanoparticles accumulation in high amounts in tumor area but also in spleen, liver, kidney, lungs and brain, with ultrastructural modification at tissue levels [267].

2.6.3. Shape

Iron nanoparticles were developed in different shapes: cubes, concave cubes, spheres, tetrahedrons, hexagons, octahedrons, octapods, polypods [268], ellipsoids, discs, cylinders, cones, hemispheres, etc. [248]. The shape of IONPs is very important for their transport through the circulatory system and for the drug delivery to the target tissue. The elongated drug carriers travel through the blood stream closer to the vessel wall, compared to the spherical nanosystems, margination property that permits the easier transport of these nanoparticles into the adjacent tissues [269]. The spike-shaped IONPs adhere easily to the endothelium while the lengthen-shaped IONPs interact with the vessel wall through the long axis, mechanisms that keep these drug delivery systems attached to areas that may not be of treatment interest, blocking their transport toward the tumor zone, when they are administered through intravenous or intraarterial injections [270]. The spherical nanoparticles are transported easier along the circulatory system because of their small area of contact [248].

2.6.4. Electrical Charge

Zeta potential of iron nanoparticles affects their cellular uptake. Several studies present a higher internalization when IONPs zeta potential is positive [271]. IONPs penetration inside the cells is affected by their size, chemical composition of coating, hydrophobicity, or by the proteins that can attach to them. Administration of IONPs with positive zeta potential (30 nm, 10 mg/kg, for 8 days) in pregnant mice produced fetal death and accumulation of nanoparticles in fetal liver and placenta [272]. PEGylated IONPs, through its hydrophilic properties, has prolonged systemic circulation, while PVP-coated IONPs has anti-opsonization properties [248]. The cell type exposed to IONPs is also important, affecting the nanoparticles internalization [273]. Cancer tissue is more acidic than the healthy zones of the body, a tumor feature that stimulated the development of IONPs that can attach and release the drug in a pH-controlled manner [274].

2.6.5. IONPs Internalization

Inside the cell, IONPs can bind different molecules or ions, leading to reactions that may affect the desired result [275]. Nanoparticles are captured by lysosomes (spherical IONPs) or are transported in endosomes (spherical, elongated, or spiked IONPs along the microtubules till the storage and processing area. Inside the cell, the ultrasmall spherical IONPs are rapidly transported to the nucleus, inhibit DNA synthesis and activate apoptosis [276]. Similar effects on DNA were in a study performed in mice with intravenous administration of ultrasmall IONPs (4 - 6 nm) [277]. The hexagonal-shaped nanoparticles remain in the cytoplasm [248] where they initiate oxidative stress.

2.6.6. Immune Response Following IONPs Administration

Systemic administration of IONPs triggers complex responses of the organism. The administration of iron nanoparticles through inhalations, injections (intraperitoneal, intravenous, intraarterial) or through oral gavage decreases the immune responses, suppressing the activity of helper T lymphocytes [278,279]. Macrophages easily phagocytize iron nanoparticles of large dimensions [280], with positive zeta potential [281] or spherical-shaped. The elliptical nanoparticles are phagocytized by macrophages in less than 6 minutes if the first contact with the cell is realized with the major axis perpendicular on the phagocyte membrane [282], while other approaches delay the phagocytosis for hours [283]. The worm-shaped IONPs avoid the macrophages phagocytosis [284] and present a higher accumulation inside the tumor compared to spherical nanoparticles [285].

2.6.7. Routes of Administration and Toxicity

Intratumoral injection of IONPs could be considered the most efficient because it avoids the systemic complex responses, but the possibility of migration from the cancer area into the circulatory system or into the surrounding areas cannot be neglected and must be explored. The iron nanoparticles can be synthesized in shapes, dimensions and chemical compositions for the transport and release of the drug in a controlled manner and can be used for magnetic hyperthermia to induce the apoptosis of tumoral cells. Wojtera et al. investigated the importance of iron content in nanostructures exposed to radio electromagnetic fields and demonstrated that a higher iron concentration produced a higher heat [286].

Applying an electromagnetic field for the transport of iron nanoparticles toward the tumor area can interfere with other electromagnetic fields (wi-fi, static magnetic fields, etc.) [287,288] and the response of IONPs depends not only on their dimension but also on their composition, local structure of IONPs assemble organized under magnetic field effects, etc. [289]. The number and the arrangement of blood vessels inside the tumor can also affect the drug delivery [290]. The IONPs transport through tissues toward the tumor area using an external electromagnetic field may disrupt healthy organs and vessels functions, processes that can be avoided by choosing the intratumorally administration route. Johannsen et al. injected intratumorally 13 nm IONPs in rats with prostate cancer for magnetic hyperthermia (45 °C or 50°C, for 30 minutes) that led to homogenously distributed nanoparticles inside the tumor, but not all the animals survived [291].

Toxicity of IONPs depends on different factors related to the nanoparticle’s properties, healthy tissues characteristics, tumor specificity, electromagnetic fields, etc. [292]. Wu et al. presented the important role of nanoparticles dimensions in IONPs toxicity. The intravenous administration of 2.3 nm or of 4.2 nm IONPs, 100 mg/kg, produced the mice death, probably of cardiac cause, the results showing increased oxidative stress in heart, lung, liver, spleen and serum. The death of mice was not recorded when the same dose was administered of 9.3 nm IONPs size [45]. Oral administration of 30 nm IONPs, through gavage for 5 days, in Wistar rats, showed that 500 mg IONPs/kg leaded to anorexia and lethargy while 5000 mg IONPs/kg had severe effects like ataxia, respiratory arrhythmia with hemorrhages in the lungs and in the heart, and liver degeneration [293]. A previous study noticed the noxious effects on liver of Wistar rats when a lower doze of IONPs (150 µg/kg) but of the same size (30 nm) was orally administered for longer period (15 days) [294].

2.6.8. Elimination

IONPs can be excreted from the body if several conditions are accomplished: the size, shape, zeta potential and other characteristics inhibit the nanoparticles dispersion, agglomeration and storage but permit their glomerular filtration. The systemic administration of IONPs permits the storage of nanoparticles in organs, according to their properties and tissues characteristics [295]. IONPs can interfere with iron metabolism, participate in electron transfer reactions or can combine with proteins, they are immunogenic and promoters of oxidative stress that initiates ferroptosis. In vivo or in clinical trials, IONPs were investigated for the treatment of several cancer types, with hyperthermia (induced by radiofrequencies, microwaves, magnetic field excitation or ultrasounds) or followed by radiotherapy.

2.6.9. Combined Radiotherapy and MH

Magnetic hyperthermia may be used combined with radiotherapy; the effects of latter being enhanced by heating of administered IONPs [166]. Several iron nanoparticles were developed and studied for radio sensitization: dextran-coated IONPs (in prostate carcinoma, glioblastoma), gold coating (in melanoma), sodium citrate coating IONPs (in breast adenocarcinoma) [279]. Li et al. studied, in their experiment performed in mice with non-small cell lung cancer, the effects of superparamagnetic iron oxide nanoclusters (60 nm), able to respond to different pH by self-assembling, administered intratracheally, followed by radiotherapy, and noticed that all the animals survived and presented reduced tumor areas [296]. Zhu et al. investigated the distribution of intratracheally administrated IONPs (22 nm) in Sprague-Dawley rats and found these nanoparticles in systemic circulatory system at 10 minutes after instillation, and stored inside the liver, kidney, and spleen even after 50 days [297]. Hyaluronic acid-based IONPs (40 mg/kg, peritumoral injections) were administered for radiosensitization in mice with subcutaneous tumors, combined with X-rays irradiation, in an experiment performed by Bae et al. that showed significant decreases of tumor dimensions with 100% survival rate [298]. The literature data provides conflicting results related to the effects of iron nanoparticles on cancer tumors.

2.7. Clinical Studies

The importance and relevance of the topic of this study are supported by the large number and variety of clinical studies conducted in recent years, and especially by the fact that some iron oxide nanoparticles have already been approved for clinical use. Therefore, Nanotherm (MagForce Ag, Berlin, Germany), Superparamagnetic iron oxide nanoparticle, Fe₃O₄ or γ-Fe₂O₃ with aminosilane coating), approved by both EMA and FDA, is used for magnetic hyperthermia therapy in recurrent Glioblastoma and prostate cancer. [19,24] Other FDA approved nanoparticles, with no use in cancer therapy are Ferumoxytol (Feraheme, Superparamagnetic iron oxide (magnetite) coated with Polyglucose sorbitol carboxymethylether produced by AMAG Pharmaceuticals, Inc. (Cambridge, MA, UK), for the treatment of iron deficiency anemia in patients with chronic kidney disease; Fermoxtran-10 (Sinerem, produced by Guerbet, Saint-Ouen, France, or named as Combidex produced by Advanced Magnetics, USA), Superparamagnetic iron oxide (magnetite) coated with Dextran, as an MRI contrast agent; and Ferumoxsil (Lumirem- Guerbet), Superparamagnetic iron oxide (magnetite) with Siloxane coating as an oral gastrointestinal tract imaging agent [19].

The clinical trials aimed to investigate the diagnostic and therapeutic properties of iron nanoparticles in various malignancies. In a phase 1 clinical trial, Carbon Nanoparticle-Loaded Iron CNSI-Fe(II) in doses of 30 mg, 60 mg, 90 mg, and 150 mg was tested on breast cancer and other advanced solid tumors, with a partial response of 25% for the 60mg dose and serious adverse events of 33.33% for 30mg, 25% for 60mg, 83.33% for 90mg, and 0% for 150 mg [299]. When it comes to the diagnostic role of Superparamagnetic Iron Oxide Nanoparticles (SPION), two clinical trials highlight the advantages of using them for sentinel lymph node detection: the procedure using MagTrace (SPIONs coated with carboxydextran) (Sysmex Europe SE, Norderstedt, Germany) was successful in all patients (100%), and no adverse effects were reported [300]; and it detected even more sentinel lymph nodes (97.4% vs 91.2%) than radioisotope (P = 0.057) [301]. NanoTherm is also the first nanoparticle-based therapy approved for the treatment of Glioblastoma Multiforme. An ongoing clinical trial, ANCHIALE, is recruiting in Poland to examine the efficacy and tolerability of this medication in the treatment of glioblastomas [302].

Intra-venous injection of SPION- contrast agent Ferumoxytol proved effective in the identification of liver neoplasms as it improved liver heterogeneity in MRI scans, allowing for more accurate characterization of liver function and tumours [303].

In rectal cancer, clinical trials are now investigating a NanoEcho Particle-1 (NEP-1, Ferumoxtran Lyophilisate 20 mg Fe/mL) based contrast agent to improve the diagnosis of lymph node metastases and staging in patients [304]. The combination therapy of MH and radiotherapy was also tested in a clinical study performed in 66 patients with glioblastoma. The pacients received intratumoral IONPs followed by magnetic hyperthermia and radiotherapy showed adverse effects during thermotherapy: the increase of body temperature at 38°C, headaches (probably because of the increase of intracranial pressure), convulsions, motor impairments, tachycardia, and blood pressure variations [183].

3. Conclusions

Although superparamagnetic iron oxide nanoparticles are widely regarded as biocompatible and clinically translatable, their safe application requires a nuanced understanding of the parameters influencing toxicity. Optimization of surface coatings, control of administered dose and core diameter, and careful design of external stimuli (e.g., magnetic fields) are essential to mitigate adverse effects. Further research, particularly on long-term biodistribution and chronic toxicity—is warranted to ensure their safe and effective use in clinical settings. In the synthesis of IONPs all the factors that interfere with the treatment, factors related to the MNPs, targeted tissue, to the systemic body responses, to the environmental electromagnetic field, etc. must be considered. Magnetic hyperthermia’s integration with other therapies – particularly immunotherapy, chemotherapy, and smart carriers – marks a shift from single-modality treatment to multifunctional platforms. With growing preclinical validation and early clinical success, iron oxide-based MH is positioned as a powerful adjunct strategy in cancer therapy, offering both targeted thermal control and synergistic therapeutic potential.

4. Future Directions

Magnetic nanoparticles (MNPs) are emerging as promising agents for enhancing the efficacy of cancer treatment. Among the innovative approaches involving MNPs, magnetic hyperthermia (MH) stands out due to its minimally invasive nature, ability to penetrate deep tissues, and potential to selectively induce various cell death pathways—including apoptosis, ferroptosis, necrosis and pyroptosis. MH allows for localized thermal ablation of tumor tissues, minimizing damage to surrounding healthy tissues. Despite numerous preclinical studies demonstrating the therapeutic potential of MH, its clinical translation remains limited. While MH has been approved for the treatment of recurrent glioblastoma, its safety and effectiveness for other types of malignancies still require validation through comprehensive clinical trials. Consequently, the transition of MH from an experimental therapeutic platform to a widely accepted clinical modality—used in conjunction with surgery, radiotherapy, and chemotherapy—will depend on future advancements in interdisciplinary materials science and the development of intelligent, adaptive treatment systems.

Significant progress was made in tailoring the intrinsic properties of MNPs, such as magnetic responsiveness, biocompatibility, and surface functionality. However, further refinement is needed to optimize their performance under clinically relevant alternating magnetic field (AMF) conditions. Standardization of experimental parameters—such as MNP concentration, exposure duration, AMF strength, and route of administration—is essential to ensure reproducibility and facilitate clinical translation.

Many MH studies were conducted on in vitro 2D models, containing tumor cells w/wt co-cultured stromal and/or vascular cells. The 2D models present certain advantages such as easy maintenance, reliability of the results, they usually involve human cell lines, and are suitable for toxicity screening of MNPs, researching mechanisms of cell death induction or tumor escape but they lack the complex spatial tumor structure. There is a great need for better tumor models that match the clinical scenario, particularly the tumor -stroma- immune system interactions, which is the key to understanding the mechanisms of tumor destruction generated by MH as a single therapy or combined with other options, such as radiotherapy, immunotherapy and/or chemotherapy. In view of this, the most reliable models, so far, have been developed in vivo, on lab animals, particularly rodents. There are, however, many limitations related to animal physiology, the ability to generate a certain type of tumor or to the ethical and financial considerations.

Recent advancements in the development of in vitro 3D models, such as the spheroids, organoids, microfluidics and the possibility of bioprinting creates opportunities for the design of human tumor models, that can comprise multiple human cell types grown on organic 3D scaffolds like collagen, Matrigel or others. These models can develop ingrown tumor vascularization and by adding immune cells can generate a certain immune response. Therefore, the 3D models can represent a step forward in standardized testing of MH to generate more reliable data in the preclinical testing. However, the creation of these models requires time, knowledge and financial resources, also reproducibility can be an issue, depending on the donor, media and reactives used for the model creation.

Future research should prioritize the design of multifunctional MNPs capable of integrating diagnostic and therapeutic modalities (so-called "theranostics") into a single nanoplatform. Such platforms would allow for precise in vivo tumor localization, real-time imaging, and patient-specific treatment, contributing to the realization of personalized oncology. A key challenge remains the efficient targeting and accumulation of MNPs at tumor sites. Passive targeting via the enhanced permeability and retention (EPR) effect often results in significant off-target deposition, particularly in organs such as the liver, spleen, and kidneys. To overcome this, advanced targeting strategies—including ligand-mediated active targeting and magnetic field-assisted navigation—should be further explored. In parallel, the development of next-generation imaging and tracking technologies will be critical to monitor in vivo distribution and enhance tumor-specific accumulation of MNPs via systemic administration. Looking forward, MNPs may also play a pivotal role in preventing metastasis. One conceptual application involves engineering MNPs to circulate within the bloodstream and capture circulating tumor cells, directing them toward an implanted magnetic device for sequestration and removal—a novel approach for metastasis interception. Combination of MH and immunotherapy has the potential to inhibit the suppressive effect of the tumor cytokines, particularly towards stroma infiltrating macrophages and trigger a phenotype switch. Combined with enhanced tumor antigen release by MH induced cell killing it can lead to an effective immune response against tumor antigens, leading to local and distant tumor destruction. Additionally, MNPs could be engineered to facilitate non-invasive biopsies of tumors that are otherwise inaccessible through conventional methods, offering new possibilities for early diagnosis and molecular profiling. In cancer immunotherapy, MNPs may serve as potent carriers for vaccine delivery, improving antigen presentation and immune activation. The continued convergence of nanotechnology, immunology, and bioengineering will likely unlock new therapeutic paradigms that exploit the full potential of MNPs in precision oncology.