Graphene and Graphene Oxide-Based Nanocarriers for Targeted Delivery of Mitoxantrone in Cancer Therapy

Abstract

Mitoxantrone (MTX) is limited by cardiotoxicity and MDR. At the same time, graphene oxide (GO) and nano-graphene oxide (NGO) offer ultrahigh surface area, strong π–π interactions with aromatic drugs, and versatile surface chemistry, making them superior to many conventional nanocarriers for high-capacity, stimuli-responsive MTX delivery. Conventional chemotherapy remains constrained by nonspecific biodistribution, dose-limiting toxicity, and the frequent emergence of multidrug resistance. This review surveys recent advances in GO- and NGO-based nanocarriers developed for targeted delivery of mitoxantrone. The platforms examined include hyaluronic acid–Pluronic functionalization for CD44-mediated targeting and P-glycoprotein inhibition, biomimetic coatings derived from mesenchymal stem cells or cancer-cell exosomes, magnetic graphene oxide systems designed for mitochondrial delivery, hybrid gold–graphene constructs, and multi-stimuli-responsive designs that respond to acidic pH, near-infrared light, or external magnetic fields. Combination strategies that integrate chemotherapy with photothermal therapy, nitric oxide release, or immunotherapy are also considered. Molecular dynamics simulations together with in vitro and limited in vivo studies indicate high drug-loading capacities (commonly reaching 40–45 wt% and higher in selected systems), stimulus-triggered release within acidic microenvironments, improved intracellular accumulation in resistant cell lines, and enhanced antitumor activity relative to free MTX in preclinical models, frequently accompanied by reduced systemic exposure. Nevertheless, long-term biocompatibility, immunogenicity, biodegradation and clearance pathways, batch-to-batch reproducibility, scalable manufacturing, and the complete absence of clinical data remain major translational barriers. Further progress will require safer, more biodegradable graphene derivatives and rigorous toxicological characterization before these platforms can advance toward clinical evaluation.

Graphical Abstract

Introduction

Cancer remains the second leading cause of death worldwide. According to GLOBOCAN estimates, approximately 20 million new cases and 9.7 million deaths occurred in 2022, with projections indicating a substantial rise by 2050 driven largely by population aging.¹ Chemotherapy continues to play a central role in cancer treatment by targeting proliferating cells rapidly; however, its lack of selectivity damages healthy tissues with high turnover rates and produces well-recognized adverse effects.^2–4

Among clinically used chemotherapeutic agents, mitoxantrone (MTX), a synthetic anthraquinone derivative, was selected as the focus of this review because of its established efficacy against breast cancer, advanced prostate cancer, acute leukemias, and non-Hodgkin lymphoma, combined with a well-documented set of pharmacological limitations that directly motivate the development of advanced delivery systems.^5–14 MTX stabilizes the topoisomerase II–DNA complex, leading to DNA strand breaks, cell-cycle arrest, and apoptosis.^6,7 Despite these therapeutic benefits, its clinical utility is restricted by high plasma-protein binding (>95%), hepatic metabolism, a prolonged half-life (24–40 h), cumulative dose-dependent cardiotoxicity that can progress to cardiomyopathy and heart failure, nonspecific biodistribution, and the frequent development of multidrug resistance (MDR) mediated primarily by efflux transporters such as P-glycoprotein.^11,15–24 These constraints create a clear need for carriers capable of high drug loading, protection of the active molecule, controlled release at the tumor site, and strategies to circumvent efflux-mediated resistance.

Graphene oxide (GO) and its nanoscale form (nano-graphene oxide, NGO), together with reduced graphene oxide (rGO), were examined for MTX delivery because their physicochemical features align closely with the requirements imposed by the drug’s aromatic structure and toxicity profile. The extended sp² carbon lattice provides a large planar surface area that supports strong π–π stacking interactions with the anthraquinone rings of MTX, while abundant oxygen-containing groups (hydroxyl, carboxyl, and epoxide) enable hydrogen bonding and further chemical functionalization.^15,25–31 These interactions have been shown in both computational and experimental studies to support high loading capacities, often exceeding 40 wt%.¹⁵ Reduction of GO to rGO partially restores the conjugated π-system, increasing near-infrared (NIR) absorption and thereby facilitating photothermal or light-triggered release strategies.^32–35 At the same time, the two materials differ in hydrophilicity, colloidal stability, and surface reactivity: GO is more hydrophilic and easier to functionalize in aqueous media, whereas rGO offers enhanced photothermal conversion but generally lower dispersibility unless further modified. Examining both families side-by-side allows assessment of how oxidation state influences loading efficiency, release kinetics, and biological behavior—questions that remain only partially resolved in the current literature.

While liposomes, polymeric nanoparticles, dendrimers, and metallic nanoparticles have also been investigated for MTX delivery,^5,23–27 graphene-based platforms provide distinctive opportunities for combining high loading capacity with multi-stimuli responsiveness and straightforward incorporation of targeting ligands or biomimetic coatings.^28,29

The present review focuses on the rational design and preclinical evaluation of GO-, NGO-, and rGO-based nanocarriers for targeted MTX delivery. It moves from molecular-dynamics insights into drug–carrier interactions to experimental systems that employ active targeting, biomimetic coatings, magnetic guidance, or combination with photothermal, gas, or immunotherapy. Emphasis is placed on the mechanistic basis of loading and release, the strength of the available preclinical evidence, and the principal barriers—long-term biocompatibility, immunogenicity, manufacturing reproducibility, and the absence of clinical data—that currently limit translation.

Physicochemical Properties of NGO and Interaction with MTX

Graphene oxide (GO) and its nanoscale form (nano-graphene oxide, NGO) have emerged as highly attractive two-dimensional nanocarriers for drug delivery due to their unique physicochemical properties.^22,32–39 The honeycomb lattice of sp²-hybridized carbon atoms provides a large planar surface area (theoretically up to 2630 m²/g), enabling efficient non-covalent interactions with aromatic drug molecules (Figure 1). In particular, NGO possesses abundant oxygen-containing functional groups, including hydroxyl (–OH), carboxyl (–COOH), and epoxide (–O–) groups, primarily located at the edges and basal plane. These groups confer excellent aqueous dispersibility and serve as versatile sites for further chemical functionalization.^31,33

Figure 1.

Schematic illustration of the main interactions between nano-graphene oxide (NGO) and mitoxantrone (MTX), highlighting π–π stacking between the anthraquinone rings of MTX and the sp² domains of NGO, together with hydrogen bonding involving the hydroxyl/amine groups of MTX and the oxygen functionalities of NGO. This figure supports Physicochemical Properties of NGO and Interaction with MTX by visualizing the molecular basis for the high drug-loading capacities (commonly 40–45 wt%) achieved with GO/NGO carriers.

The interaction between mitoxantrone (MTX) and NGO is primarily driven by two synergistic mechanisms: π–π stacking between the aromatic rings of MTX and the sp² domains of graphene oxide, and hydrogen bonding between the hydroxyl and amine groups of MTX and the oxygenated functional groups on the NGO surface.^15,35 These interactions result in exceptionally high drug loading capacities, with some systems achieving up to 45 wt%.¹⁵ The planar structure of NGO also allows for compact stacking upon drug loading, often leading to a reduction in hydrodynamic diameter compared to unloaded carriers.

Reduction of GO to reduced nano-graphene oxide (r-NGO) partially restores the conjugated π-system, enhancing electrical conductivity and near-infrared (NIR) light absorption. This property is particularly valuable for photothermal therapy and NIR-triggered drug release.^32,40 Additionally, NGO nanosheets smaller than 100 nm exhibit superior colloidal stability and cellular uptake due to their high edge-to-surface ratio and negative surface charge.³⁴

Surface modification of NGO with biocompatible polymers such as polyethylene glycol (PEG), chitosan, or Pluronic further improves stability in physiological media, reduces opsonization, and prolongs blood circulation time.^15,22 Moreover, the pH-responsive nature of these systems is highly advantageous for cancer therapy. In the acidic tumor microenvironment (pH ≈ 5.0–6.5), protonation of functional groups weakens the hydrogen bonds and π–π interactions, facilitating controlled and targeted release of MTX while minimizing premature leakage in neutral physiological conditions (pH 7.4).^15,16,41

Molecular dynamics (MD) simulations have provided valuable atomic-level insights into these interactions. For instance, Mirhosseini et al demonstrated that MTX forms more stable hydrogen bonds and exhibits lower mobility on hydroxyl-functionalized nanographene compared to other anticancer drugs, confirming its high affinity for graphene-based carriers.³⁵ These computational findings have been experimentally validated in subsequent studies, establishing a strong foundation for the rational design of advanced NGO-based delivery systems.

Functional Basis of GO/rGO Nanocarriers for Mitoxantrone Delivery

Beyond the specific physicochemical interactions of MTX with GO/rGO, the broader functional advantages of nanocarrier systems—drug encapsulation and protection, prolonged systemic circulation, active or biomimetic targeting, enhanced cellular uptake, and stimuli-controlled release—collectively determine therapeutic outcomes.^42–46 These general attributes are particularly relevant for overcoming the dense extracellular matrix, elevated interstitial fluid pressure, and immunosuppressive microenvironment that limit conventional chemotherapy (Figure 2).^42,43 In the following subsections, we link each of these attributes to the experimental evidence available for MTX–GO platforms.

Figure 2.

Schematic overview of programmed nanoparticle delivery strategies for overcoming barriers to deep tumor penetration. Collectively, these programmed design strategies—size transformation combined with surface functionalization and biomimetic coatings—address the key physiological barriers that limit conventional nanoparticle delivery and enable more efficient drug accumulation in poorly accessible tumor regions.⁴⁶ (A) Major biophysical and cellular barriers within the solid tumor microenvironment (TME), including elevated interstitial fluid pressure (IFP), dense collagen-rich extracellular matrix (ECM), and a hypoxic core (low pO₂). Conventional large nanoparticles (~180 nm) remain confined near the tumor periphery due to these barriers, resulting in limited penetration. (B) Size-switching (size-transformable) strategy: nanoparticles remain large during systemic circulation to exploit the enhanced permeability and retention (EPR) effect, then rapidly shrink to approximately 10 nm in response to tumor-specific stimuli (acidic pH, elevated glutathione, or enzymatic activity). The resulting small nanoparticles achieve significantly deeper penetration into the tumor core. (C) Complementary surface-engineering approaches, including ligand decoration (eg, hyaluronic acid, folic acid, or antibodies) for active receptor-mediated targeting and biomimetic “Trojan-horse” cellular hitchhiking using macrophages or neutrophils to enhance immune evasion and further improve intratumoral delivery.

The utility of GO and rGO as carriers for MTX rests on a set of interrelated physicochemical and biological functions that can be linked, where data exist, to measurable experimental outcomes.

Drug Loading and Protection

The large planar surface of GO/NGO supports non-covalent loading of MTX primarily through π–π stacking between the drug’s anthraquinone core and the sp² domains of the graphene sheet, supplemented by hydrogen bonding involving the drug’s hydroxyl and amine groups and the oxygen functionalities of GO.^15,35 Loading capacities reaching 45 wt% have been reported for HA-NGO systems,¹⁵ and even higher relative loadings have been achieved with nitrogen-doped GO.¹⁶ Once adsorbed, the drug is partially shielded from the external aqueous environment, which may reduce premature degradation or interactions with plasma components, although direct quantification of chemical stability under physiological conditions remains limited in the cited studies.

Intermolecular Interactions

π–π stacking, electrostatic forces, and hydrophobic contributions act synergistically. Molecular-dynamics simulations indicate that hydroxyl-functionalized nanographene forms a greater number of hydrogen bonds with MTX than with certain other aromatic drugs, resulting in lower drug mobility on the surface.³⁵ These interactions are pH-sensitive: under acidic conditions typical of the tumor microenvironment or endosomal compartments, protonation weakens the binding and facilitates release.^15,16,41

Circulation and Stability

Surface modification with PEG, Pluronic, or other hydrophilic polymers improves colloidal stability and reduces opsonization in vitro and in short-term in vivo models.^15,22 Nevertheless, the circulation half-life of even PEGylated graphene oxide remains substantially shorter than that of clinically approved liposomal formulations. Recent advances in “stealth and pseudo-stealth” strategies that go beyond conventional steric stabilization⁴⁷ suggest that denser hydrophilic brushes, zwitterionic coatings, or biomimetic cloaks capable of actively suppressing protein-corona formation may be required to achieve clinically relevant blood residence times.

Rapid formation of a protein corona, residual recognition by the reticuloendothelial system, and incomplete shielding of the hydrophobic basal plane continue to limit blood residence time.^36,37 As a result, a significant fraction of the injected dose is still cleared within hours rather than days. Alternative or complementary strategies are therefore required. Biomimetic membrane coatings (mesenchymal-stem-cell or cancer-cell-derived membranes), precise control of lateral size below 100 nm, fine-tuning of surface charge, and the use of more effective stealth polymers or zwitterionic coatings have shown promise in prolonging circulation and reducing non-specific uptake. Systematic head-to-head comparisons of these approaches under identical conditions, together with quantitative pharmacokinetic studies, will be essential to determine which strategy most effectively overcomes the inherent circulation limitations of graphene-based nanocarriers.

Targeting and Cellular Uptake

Passive accumulation via the enhanced permeability and retention (EPR) effect is supplemented in several systems by active targeting ligands (hyaluronic acid for CD44, folic acid for folate receptors) or biomimetic coatings (MSC membranes, cancer-cell-derived exosomes).^15,36,37,41 These modifications have been associated with increased cellular internalization in receptor-positive cell lines and, in selected xenograft models, higher tumor accumulation relative to non-targeted counterparts.

Stimuli-Responsive Release

pH-triggered release is the most consistently demonstrated mechanism, with accelerated MTX liberation observed at pH 4.5–5.5 versus pH 7.4.^15,36,37,41 Additional triggers—NIR irradiation (enabled by the photothermal properties of rGO or N-GO) and alternating magnetic fields—have been shown to further enhance release on demand in vitro and in limited in vivo settings.^15,16,38

Link to Measured Therapeutic Outcomes

In the systems reviewed, higher loading and stimuli-responsive release have been correlated with increased intracellular MTX concentrations in both drug-sensitive and resistant cell lines, reduced IC₅₀ values relative to free MTX, and, in several xenograft models, greater tumor-growth inhibition and lower systemic toxicity markers.^15,16,41,48 These outcomes remain at the preclinical level; no clinical efficacy data are available (Table 1).

Table 1.

Functional Attributes of GO/rGO Nanocarriers for MTX Delivery and Associated Evidence Levels

Functional Attribute	Primary Mechanism	Representative Experimental Observation	Key Limitation	Evidence Level
High drug loading	π–π stacking + H-bonding	Up to 45 wt.% (HA-NGO); higher relative loading (N-GO)	Batch-to-batch variability	In vitro/computational
Drug protection	Surface adsorption/steric shielding	Reduced premature release at pH 7.4	Limited chemical-stability data	In vitro
Stimuli-responsive release	pH-dependent weakening of interactions; NIR/magnetic triggers	Accelerated release at acidic pH or under NIR/AMF	Incomplete release in some hybrid systems	In vitro/limited in vivo
Active/biomimetic targeting	Ligand–receptor or membrane homology	Increased uptake in CD44+ or homologous cells; higher tumor accumulation in selected models	Heterogeneity of receptor expression	In vitro/limited in vivo
Circulation prolongation	PEGylation or biomimetic coating	Improved stability and reduced opsonization short-term	Still shorter than many approved liposomes	In vitro/short-term in vivo
Overcoming MDR	P-gp inhibition (Pluronic) or NO-mediated down-regulation	Higher intracellular MTX in resistant lines; tumor suppression in MDR models	Mechanism and durability incompletely defined	In vitro/limited in vivo

The attributes summarized above provide a mechanistic rationale for the use of GO/rGO platforms with MTX. At the same time, Table 1 underscores that most supporting data remain preclinical and that critical parameters (long-term circulation, biodegradation, and chronic toxicity) require further systematic investigation.

Mechanistic Basis of Mitoxantrone Loading and Release on GO/rGO Surfaces

The loading of mitoxantrone (MTX) onto graphene oxide (GO) and nano-graphene oxide (NGO) is governed by two synergistic non-covalent interactions. π–π stacking occurs between the anthraquinone rings of MTX and the sp²-hybridized domains of the graphene lattice, while hydrogen bonding forms between the hydroxyl and amine groups of the drug and the oxygen-containing functionalities (hydroxyl, carboxyl, and epoxide) present on the GO surface.^15,35 Molecular dynamics simulations using the COMPASS force field demonstrated that MTX establishes a greater number of hydrogen bonds and exhibits lower surface mobility than camptothecin on hydroxyl-functionalized nanographene, providing a molecular-level explanation for the high loading capacities frequently observed (up to 45 wt% in HA-NGO systems).^15,35

Reduction of GO to reduced graphene oxide (rGO) partially restores the conjugated π-system. This structural change enhances near-infrared absorption and photothermal conversion but simultaneously decreases the density of hydrogen-bonding sites. Consequently, GO is generally preferred when stable aqueous loading is prioritized, whereas rGO is advantageous for light-triggered applications.^32,33

Release of MTX is predominantly pH-responsive. Under acidic conditions characteristic of the tumor microenvironment or endosomal compartments (pH ≈ 5.0–6.5), protonation of surface functional groups weakens both hydrogen bonding and electrostatic interactions, thereby facilitating drug liberation while minimizing premature release at physiological pH (7.4). This behavior has been consistently observed across multiple platforms, including HA-NGO/Pluronic, MSC-NGO, exosome-coated NGO, and FA-PEG-cGQD systems.^15,36,37,41 In certain hybrid constructs, such as SMTX-AuNPs/RGO, residual strong π–π interactions can retard release even under acidic conditions, underscoring the need to balance loading stability with triggered liberation.³⁹

Additional external stimuli further refine temporal control. Near-infrared irradiation exploits the photothermal properties of rGO or nitrogen-doped GO to accelerate release, while alternating magnetic fields enable on-demand liberation in magnetic GO systems.^15,16,38 Combination of pH sensitivity with these external triggers allows spatiotemporally controlled delivery.

Two complementary strategies have been employed to address multidrug resistance. Pluronic F68 directly inhibits P-glycoprotein efflux, whereas near-infrared-triggered nitric oxide release from BNN6 down-regulates P-glycoprotein expression, both resulting in increased intracellular MTX accumulation in resistant cell lines.^15,16 These mechanistic features collectively support the design of GO/rGO carriers capable of high loading, tumor-selective release, and partial reversal of resistance, while remaining at the preclinical stage of validation.

Evolution of Graphene Drug Delivery: From Simulation to Biological Application

The development of graphene oxide-based nanocarriers for mitoxantrone delivery has progressed from theoretical understanding through molecular dynamics (MD) simulations to experimentally validated smart delivery systems. This evolution has been instrumental in elucidating the key molecular interactions governing drug loading and release, thereby guiding the rational design of advanced nanocarriers.

Molecular Insights from Computational Simulations

Molecular dynamics (MD) simulations have provided atomic-level understanding of the interactions between mitoxantrone and graphene-based surfaces. The most detailed study focused on MTX remains that of Mirhosseini et al (2019), who used the COMPASS force field to examine loading of mitoxantrone and camptothecin onto a hydroxyl-functionalized nanographene sheet (approximately 3×3 nm, ∼11% hydroxyl groups).³⁵ MTX formed a larger number of hydrogen bonds and displayed lower surface mobility (mean-square displacement analysis) than camptothecin, indicating stronger binding affinity driven by both π–π stacking and polar interactions.

Although systematic computational studies that specifically vary oxidation degree, lateral size, solvent, or pH for the MTX–GO pair are still scarce, related graphene–aromatic-drug simulations offer complementary insights that have guided experimental design (Table 2). Higher surface oxidation generally increases the density of hydrogen-bond donors/acceptors and improves aqueous dispersibility, while excessive oxidation can reduce the area available for π–π stacking. Smaller lateral sizes increase edge-to-surface ratio and colloidal stability but may lower absolute loading capacity. Solvent and pH strongly modulate the balance between hydrophobic and electrostatic contributions; acidic conditions weaken hydrogen bonding and favor release—observations that align with the pH-responsive behavior later confirmed experimentally for several MTX–GO systems.^15,16,41

Table 2.

Selected Computational Studies Relevant to MTX–Graphene Interactions and Their Experimental Implications

Study/Method	Variable Examined	Key Finding	Implication for Experimental Design	Main Limitation
Mirhosseini et al (2019) COMPASS MD35	Drug type (MTX vs camptothecin), H-bonding	MTX forms more H-bonds and shows lower mobility	Prefer surfaces with polar groups for stable MTX loading	Small flake size; no explicit solvent effects on release
Related GO–aromatic drug MD (general literature)	Oxidation degree, sheet size	Moderate oxidation balances π–π and H-bonding	Avoid over-oxidation if high π–π loading is required	Rarely MTX-specific; limited experimental cross-validation
pH/solvent effects (inferred)	Protonation state	Acidic pH weakens binding	Design for tumor-microenvironment-triggered release	Few direct MTX simulations under varying pH

Collectively, these simulations have supported three practical design choices observed in the experimental literature: (i) retention of a moderate density of oxygen functionalities to balance loading capacity and aqueous stability, (ii) use of hydroxyl- or carboxyl-rich surfaces when hydrogen bonding is desired, and (iii) expectation of accelerated release under acidic conditions. Important limitations remain: most models employ simplified, periodic, or small graphene flakes in pure solvent, omit the full biological corona, and operate on nanosecond timescales that cannot capture long-term desorption or degradation. Direct quantitative validation against experimental loading efficiencies or release kinetics for MTX is still limited to a small number of systems.

Experimental Validation: HA-NGO/Pluronic Multifunctional System

Building upon these computational predictions, Hou et al (2015) developed a practical multifunctional nanocarrier that successfully translated theoretical insights into a biologically functional platform.¹⁵ They fabricated hyaluronic acid-modified nano-graphene oxide (HA-NGO) conjugated with Pluronic F68 to achieve both active targeting and reversal of multidrug resistance. Hyaluronic acid served as a ligand for CD44 receptors, which are overexpressed on many cancer cells, while Pluronic F68 acted as a P-glycoprotein (P-gp) inhibitor to reduce drug efflux (Figure 3).

Figure 3.

Schematic illustration of the HA-NGO/Pluronic–MTX smart nanocarrier (adapted from).¹⁵ Panel (A) shows how the system combines EPR effect + active CD44 targeting via hyaluronic acid, P-glycoprotein inhibition by Pluronic F68 (blocking drug efflux), receptor-mediated endocytosis, and dual pH/NIR-triggered release inside the tumor cell, ultimately delivering MTX to the nucleus. Panel (B) is a hyperspectral microscopy image confirming the presence of the loaded nanocarrier. This figure directly illustrates the multifunctional design discussed in Experimental Validation: HA-NGO/Pluronic Multifunctional System and the mechanisms of MDR reversal emphasized throughout the review.

This system demonstrated remarkable drug loading capacity, achieving up to 45 wt% mitoxantrone at a carrier concentration of 2.5 mg/mL. Upon drug loading, π–π stacking and hydrogen bonding interactions led to structural compaction, resulting in a smaller hydrodynamic diameter compared to unloaded HA-NGO. This observation was consistent with the simulation results regarding enhanced stability through intermolecular interactions.^15,35

In vitro drug release studies showed a clear dual-responsive behavior: significantly accelerated mitoxantrone release at acidic pH (4.5, mimicking the tumor microenvironment) and under near-infrared (NIR) laser irradiation. The combination of pH and NIR stimuli enabled precise, on-demand drug release. Cellular uptake and cytotoxicity experiments on both sensitive (MCF-7) and doxorubicin-resistant (MCF-7/ADR) breast cancer cell lines demonstrated markedly higher intracellular drug accumulation and enhanced cytotoxicity compared to free mitoxantrone or non-targeted carriers. In vivo studies using xenograft mouse models further confirmed improved tumor accumulation, reduced systemic toxicity, and superior therapeutic efficacy in drug-resistant tumors.¹⁵

Implications and Limitations

The successful translation of molecular-dynamics predictions into the multifunctional HA-NGO/Pluronic platform demonstrates that a rational, interaction-driven design strategy can yield nanocarriers with both high drug-loading capacity and biologically relevant performance. The computational observation that mitoxantrone forms a greater number of hydrogen bonds and exhibits lower surface mobility on hydroxyl-functionalized graphene directly informed the retention of a moderate density of oxygen-containing groups on the NGO surface. This structural choice preserved sufficient sp² domains for π–π stacking while providing polar sites for hydrogen bonding, resulting in the experimentally observed loading of up to 45 wt%. Concurrently, the incorporation of hyaluronic acid and Pluronic F68 converted a purely physicochemical loading advantage into a dual biological function—CD44-mediated targeting and P-glycoprotein inhibition—thereby addressing two of the principal pharmacological limitations of free mitoxantrone.

Nevertheless, several important limitations persist and constrain the immediate translational value of this and related systems. First, the majority of supporting data remain confined to short-term in-vitro assays and single-dose xenograft studies; long-term toxicity, immunogenicity, and chronic biodistribution have not been systematically evaluated. Second, the precise contribution of each component (NGO scaffold, hyaluronic acid, and Pluronic) to the observed reversal of multidrug resistance has not been fully dissected through appropriate control experiments. Third, batch-to-batch variability in lateral size, oxidation degree, and degree of functionalization continues to hinder reproducibility—an issue that becomes critical when moving from laboratory-scale synthesis to larger production. Finally, the circulation half-life of even PEGylated or Pluronic-modified graphene oxide remains substantially shorter than that of clinically approved liposomal formulations, indicating that additional strategies to minimize protein-corona formation and accelerate clearance of degraded fragments will be required.

Taken together, the HA-NGO/Pluronic system serves as a valuable proof-of-concept that computational insight can guide the rational design of high-capacity, stimuli-responsive, and resistance-overcoming carriers. At the same time, it highlights the gap that still separates sophisticated preclinical performance from the rigorous safety, manufacturing, and regulatory standards demanded for clinical evaluation.

Targeting Strategies for MTX Delivery

Receptor-mediated and biomimetic targeting represent the main strategies used to improve the selectivity of GO/rGO-based MTX carriers.

Active Ligand-Based Targeting

The most extensively studied active-targeting approach employs hyaluronic acid for CD44 recognition. In the HA-NGO/Pluronic system, hyaluronic acid enabled preferential uptake by CD44-overexpressing cells while Pluronic contributed to P-glycoprotein inhibition (Figure 3).¹⁵ A parallel strategy using folic acid on carboxylated graphene quantum dots (FA-PEG-cGQD) exploited folate-receptor overexpression and achieved high encapsulation efficiency together with pH-dependent release (Figure 4).⁴¹ Both systems increased intracellular MTX accumulation relative to non-targeted controls in receptor-positive cell lines. The HA-NGO platform additionally demonstrated improved tumor accumulation in xenograft models. Limitations common to ligand-based approaches include receptor heterogeneity across tumors and the still-limited number of direct comparisons with non-targeted graphene carriers.

Figure 4.

Fabrication process and function of the FA-PEG-cGQD-MTX smart nanocarrier (adapted and redrawn from).⁴¹ (A) Carboxylated graphene quantum dots (cGQD) are covalently attached to PEG and folic acid. Then, the drug mitoxantrone (MTX) is loaded onto the surface of these nanoparticles through π-π interactions and hydrogen bonding. (B) The folic acid (FA) ligand enables the targeted entry of the nanocarrier into cancer cells that have folate receptors. In the acidic endosomal-lysosomal environment of tumor cells, the drug is released in a controlled and pH-dependent manner. This smart mechanism enables precise drug release and increases treatment efficacy on cervical tumors in vivo. The figure clarifies the covalent PEGylation/folate conjugation steps and the subsequent receptor-mediated uptake and pH-triggered release pathway described in the targeting subsection of the review.(adapted and redrawn from).⁴¹

Biomimetic and Cell-Derived Coatings

To improve tumor homing and reduce immune recognition, researchers have coated GO carriers with biological membranes. Mesenchymal-stem-cell (MSC)–NGO complexes retained high MSC viability while selectively reducing cancer-cell viability in co-culture (Figure 5).³⁶

Figure 5.

Schematic representation of the MSC-NGO-MTX biomimetic system (adapted from).³⁶ Graphene oxide is loaded with MTX and then internalized by mesenchymal stem cells (MSCs), which act as living “Trojan-horse” carriers. The MSCs actively home to tumor sites via their natural tropism; once there, the acidic tumor microenvironment triggers release of MTX, preferentially killing cancer cells while sparing healthy tissue. This figure visualizes the biomimetic targeting strategy discussed in the text as an alternative to conventional ligand-based active targeting.

Exosome-coated NGO-CO-γ-PGA increased loading efficiency and showed markedly higher release at acidic pH than at physiological pH (Figure 6).³⁷ These coatings confer homologous targeting and short-term immune-evasive properties; however, both platforms remain largely confined to in-vitro evaluation, and the relative contribution of the graphene core versus the biological membrane has not been fully separated.

Figure 6.

Characterization and mitoxantrone (MIT) loading performance of the exosome-coated NGO-CO-γ-PGA nanocarrier (adapted from).³⁷ (A) FTIR spectra of GO, GO-CO-γ-PGA, and EXO-GO-CO-γ-PGA, showing the characteristic vibrational bands associated with GO functionalization and subsequent exosome-membrane coating. (B) TEM images of EXO-GO-CO-γ-PGA at low and high magnifications, showing the morphology and nanosheet structure of the hybrid carrier; scale bars, 500 and 200 nm, respectively. (C) UV–Vis absorption spectra of GO, EXO, GO-CO-γ-PGA, and EXO-GO-CO-γ-PGA, showing the spectral changes associated with covalent modification and exosome coating. (D) MIT loading capacity of GO-CO-γ-PGA and EXO-GO-CO-γ-PGA as a function of MIT concentration. The exosome-coated nanocarrier shows higher loading at the tested concentrations, reaching approximately 1.4 mg MIT per mg carrier at the highest concentration shown.

Subcellular (Mitochondrial) Targeting

While receptor-mediated and biomimetic strategies aim to improve tumor selectivity at the tissue and cellular levels, organelle-level targeting offers an additional mechanistic advantage by delivering the cytotoxic payload directly to a critical subcellular compartment. Zhu et al conjugated a mitochondria-targeting peptide (MitP) to magnetic graphene oxide nanoparticles (GOMNP-MitP) and loaded the construct with mitoxantrone (Figure 7).³⁸ The peptide markedly increased both drug-loading efficiency and preferential accumulation within mitochondria. Upon application of an alternating magnetic field (375 kHz), rapid on-demand release of mitoxantrone was achieved (approximately 34% within 30 min and 38% within 60 min), while negligible leakage occurred in the absence of the field. The localized release induced mitochondrial membrane depolarization, disruption of ATP synthesis, and high levels of apoptosis in cancer cells.

Figure 7.

Schematic design of the mitochondrion-targeting magnetic graphene oxide system (GOMNP-MitP-MTX) (adapted from).³⁸ Magnetic graphene oxide nanoparticles (GOMNP) are first functionalized with the mitochondria-targeting peptide MitP and then loaded with MTX. After cellular uptake, the peptide directs the nanocarrier to mitochondria; subsequent application of an alternating magnetic field (AMF) triggers rapid local release of MTX, impairing mitochondrial function and inducing apoptosis. SEM images confirm the morphology of GOMNP and GOMNP-MitP. This figure illustrates the organelle-level targeting and external-stimuli control strategies highlighted in the magnetic graphene oxide subsection.

This approach is conceptually attractive for two reasons. First, mitoxantrone’s primary mechanism of action—stabilization of the topoisomerase II–DNA complex—ultimately leads to mitochondrial-dependent apoptosis; delivering the drug directly to mitochondria may therefore amplify its cytotoxic effect and potentially reduce the required systemic dose. Second, the combination of magnetic guidance with a molecular targeting peptide provides both spatial control (external magnetic field) and molecular specificity (MitP), thereby adding a layer of selectivity that is difficult to achieve with conventional nanocarriers.

Despite these mechanistic advantages, the strategy remains at an early stage of development. All reported data are limited to in-vitro cell-culture models; no in-vivo biodistribution, tumor-accumulation, or efficacy studies have yet been published. In addition, the dual functionalization (magnetic nanoparticles plus peptide) increases synthetic complexity and raises questions about long-term colloidal stability, potential peptide degradation in biological fluids, and the immunogenicity of the hybrid construct. Finally, the requirement for specialized alternating-magnetic-field equipment limits the immediate practicality of the approach in clinical settings.

In summary, mitochondrial targeting illustrates how the high surface area and versatile chemistry of graphene oxide can be exploited to achieve organelle-level precision. However, substantial additional work—particularly in vivo validation, simplification of the multi-component design, and thorough safety evaluation—is required before this strategy can be considered a realistic candidate for translational development.

Active ligand-based targeting, biomimetic membrane coating, and subcellular mitochondrial delivery represent three complementary yet distinct strategies for improving the selectivity of graphene oxide–based mitoxantrone carriers. Ligand-mediated approaches (hyaluronic acid for CD44 or folic acid for folate receptors) are the most straightforward to implement and have already demonstrated improved cellular uptake and, in selected cases, enhanced tumor accumulation in xenograft models. Their principal limitation remains the heterogeneous and often transient expression of the target receptor across different tumor types and even within the same tumor mass. Biomimetic coatings derived from mesenchymal stem cells or cancer-cell exosomes offer the additional advantages of homologous targeting and short-term immune evasion; however, they substantially increase manufacturing complexity, raise concerns about batch-to-batch reproducibility of the biological membrane, and have so far been evaluated almost exclusively in vitro. Mitochondrial targeting, although the least mature of the three strategies, provides a unique mechanistic benefit by directing the cytotoxic payload to an organelle whose dysfunction amplifies mitoxantrone-induced apoptosis. The requirement for specialized external stimuli (alternating magnetic fields) and the current absence of in-vivo data nevertheless restrict its near-term translational potential.

Across all three approaches the graphene oxide sheet functions primarily as a high-capacity, easily functionalized scaffold rather than as an intrinsically targeting entity. The therapeutic gain therefore depends less on the intrinsic properties of the two-dimensional material itself and more on the quality of the surface engineering and the biological relevance of the chosen targeting moiety. Future progress will require head-to-head comparisons of these strategies under identical experimental conditions, systematic evaluation of long-term immunogenicity and clearance, and the development of simpler, more scalable functionalization methods that preserve the high drug-loading capacity of graphene oxide while improving its biological performance.

Loading Mechanisms and Stimuli-Responsive Release

High drug-loading capacity and controlled, on-demand release are central performance attributes that distinguish GO/rGO platforms for mitoxantrone delivery.

Mechanisms of MTX Loading

Loading of mitoxantrone onto GO and NGO is driven primarily by π–π stacking between the anthraquinone core of the drug and the sp² domains of the graphene sheet, reinforced by hydrogen bonding between the drug’s hydroxyl and amine groups and the oxygen-containing functionalities (–OH, –COOH, epoxide) on the GO surface.^15,35 These synergistic interactions enable high loading capacities. The HA-NGO/Pluronic system achieved up to 45 wt% MTX (Figure 3),¹⁵ while nitrogen-doped GO (N-GO) reached substantially higher relative loadings (up to 424% with respect to carrier weight) owing to its large surface area and positive surface charge (Figure 8).¹⁶ Upon loading, the planar geometry of the nanosheets often leads to structural compaction and a reduction in hydrodynamic diameter, consistent with strong intermolecular interactions observed in both experimental and molecular-dynamics studies.^15,35

Figure 8.

Schematic of the N-GO-MTX-BNN6 nanosystem (adapted from).¹⁶ Nitrogen-doped graphene oxide is sequentially loaded with MTX and the nitric-oxide donor BNN6. Upon NIR irradiation, BNN6 releases NO, which down-regulates P-glycoprotein and sensitizes multidrug-resistant cells to MTX; simultaneously the photothermal effect of N-GO contributes to direct cell killing. The inset shows the chemical conversion of BNN6 to NO. This figure captures the multi-modal (chemo + gas + photothermal) strategy used to overcome MDR that is discussed in the combination-therapy section.

In hybrid systems the balance between loading stability and releasability becomes more complex. When MTX was first covalently conjugated to gold nanoparticles and then immobilized on reduced graphene oxide (SMTX-AuNPs/RGO), the strong residual π–π interactions with the RGO surface markedly slowed release even under acidic conditions (Figure 9).³⁹ This observation highlights a recurring design tension: the same interactions that confer high loading can hinder triggered liberation if not carefully modulated.

Figure 9.

Schematic representation of hybrid graphene-gold and graphene-magnetic nanocomposites for mitoxantrone delivery (adapted from).^39,40 (A) Synthesis of SMTX-AuNPs/RGO: oleylamine-capped gold nanoparticles are first conjugated with a thiolated MTX derivative (SMTX) via a pH-sensitive linker, then immobilized on reduced graphene oxide; a parallel MPA-AuNPs/RGO route is also shown. (B) Preparation of GO-PEG/Fe₃O₄: graphite is oxidized to GO, PEGylated, and finally decorated with magnetic Fe₃O₄ nanoparticles. These hybrids illustrate the design trade-offs between high loading (via π–π or covalent attachment) and controlled release that are analyzed in the hybrid-systems subsection.

Stimuli-Responsive Release

The pH-triggered release is the most consistently demonstrated mechanism across the reviewed platforms. Accelerated MTX liberation at acidic pH (4.5–5.5), which mimics the tumor microenvironment or endosomal compartments, compared with physiological pH (7.4) has been reported for HA-NGO (Figure 2),¹⁵ FA-PEG-cGQD (Figure 3),⁴¹ MSC-NGO complexes (Figure 5),³⁶ and exosome-coated NGO (Figure 6).³⁷ Protonation of surface functional groups weakens both hydrogen bonding and π–π interactions, facilitating drug detachment.

External physical triggers provide an additional layer of temporal control. Near-infrared (NIR) irradiation exploits the intrinsic photothermal properties of reduced or nitrogen-doped graphene. In the HA-NGO/Pluronic system, NIR laser irradiation further accelerated release beyond the effect of pH alone (Figure 3).¹⁵ Similarly, the N-GO-MTX-BNN6 platform exhibited multi-stimuli responsiveness, with both pH and NIR (660 nm) enhancing simultaneous liberation of MTX and nitric oxide (Figure 8).¹⁶ Magnetic field-triggered release was demonstrated with the mitochondrion-targeted GOMNP-MitP carrier: application of an alternating magnetic field (375 kHz) produced rapid MTX release (34% at 30 min, 38% at 60 min) while minimal leakage occurred in the absence of the field (Figure 7).³⁸

Analytical Synthesis (Loading and Release)

A comparative examination of the loading and release behavior across the reviewed platforms reveals several consistent structure–property relationships. The highest absolute and relative loading capacities are achieved with moderately oxidized or nitrogen-doped graphene surfaces that simultaneously preserve a sufficient area of sp² domains for π–π stacking and retain polar oxygen- or nitrogen-containing groups for hydrogen bonding and electrostatic interactions. Purely reduced graphene oxide, while advantageous for photothermal conversion, frequently over-stabilizes mitoxantrone through residual strong π–π interactions, resulting in incomplete or delayed release even under acidic conditions. Conversely, heavily oxidized surfaces improve aqueous dispersibility but can reduce the effective area available for aromatic stacking and thereby lower loading efficiency.

Among internal stimuli, pH-triggered release remains the most reproducible and broadly applicable mechanism; accelerated liberation at pH 4.5–5.5 versus pH 7.4 has been demonstrated across chemically diverse constructs (HA-NGO, FA-PEG-cGQD, MSC-NGO, exosome-coated NGO, and N-GO). External triggers—near-infrared irradiation and alternating magnetic fields—provide superior temporal and spatial control, yet they introduce additional layers of complexity: the former requires an intrinsically photothermal material (rGO or N-GO) and appropriate laser dosimetry, while the latter necessitates specialized instrumentation and magnetic functionalization. Hybrid systems that combine covalent drug conjugation with graphene immobilization (eg, SMTX-AuNPs/RGO) illustrate a recurring design tension: the same interactions that confer high loading can compromise the efficiency of triggered release if not carefully balanced.

Collectively, these observations indicate that optimal performance is obtained when loading stability and releasability are deliberately tuned rather than maximized in isolation. Quantitative structure–property relationships that systematically link oxidation degree, lateral size, surface charge, and release kinetics under physiologically relevant conditions are still lacking and constitute a critical priority for future experimental and computational work.

Multifunctional and Combination Platforms

Several GO/rGO systems integrate MTX chemotherapy with additional therapeutic modalities, exploiting the intrinsic photothermal properties of reduced or doped graphene and the high loading capacity of the two-dimensional sheet.

Photo-Chemo-Immunotherapy

Zhou et al developed a PEGylated reduced nano-graphene oxide (r-NGO) platform co-loaded with MTX and the TGF-β inhibitor SB-431542 (Figure 10).⁴⁸ Upon NIR irradiation the system simultaneously released both agents, induced immunogenic cell death, increased CD8⁺ T-cell infiltration, reduced regulatory T cells, and generated systemic antitumor immunity. In the 4T1 metastatic breast-cancer model, 70% of treated mice achieved long-term survival and protection against tumor rechallenge. This platform illustrates how the photothermal conversion capacity of r-NGO can be combined with immune-modulatory chemotherapy to produce abscopal effects.

Figure 10.

(A) Schematic diagram of a hybrid nanosystem based on reduced nano-graphene oxide (r-NGO) for the treatment of metastatic tumors using photo-chemo-immunotherapy (adapted and redrawn from).⁴⁸ In this system, r-NGO acts as a drug carrier and photo thermal agent. The chemotherapeutic drug mitoxantrone (MTX) and a TGF-β pathway inhibitor SB-431542 are loaded on the surface of this nanocarrier. After laser irradiation, the system simultaneously releases these two agents. This release leads to primary tumor destruction, immunogenic cell death (ICD), and release of tumor antigens. Subsequently, increased infiltration of CD8⁺ cells, inhibition of Treg cells, and systemic antitumor immunity such as an abscopal effect and prevention of metastasis, were observed. This approach demonstrates that the combination of phototherapy (PTT) and r-NGO-based chemotherapy can, like an intratumoral vaccine, modulate the immune microenvironment and induce a potent and sustained anticancer response. (B) Histology of excised lungs 10 days after different treatments shows the superior suppression of metastatic nodules by the full rGO/MTX/SB + laser combination. The figure links the multi-modal design to the observed systemic antitumor immunity and abscopal effect discussed in the immunotherapy combination section. Histology of excised lungs 10 days after different treatments (adapted and redrawn from).⁴⁸

Nitric-Oxide-Sensitized Photothermal Chemotherapy

Huang et al introduced a nitrogen-doped graphene oxide (N-GO) carrier co-loaded with MTX and the nitric-oxide donor BNN6 (Figure 8).¹⁶ NIR irradiation (660 nm) triggered concurrent release of MTX and NO. Nitric oxide down-regulated P-glycoprotein expression, thereby reversing multidrug resistance and increasing intracellular MTX accumulation. In MCF-7/ADR tumor-bearing mice the triple combination (chemotherapy + photothermal therapy + gas therapy) produced near-complete tumor-growth suppression. The exceptionally high loading capacity of N-GO (up to 424% relative to carrier weight) was critical to delivering effective doses of both the chemotherapeutic and the gas-generating agent.

Hybrid Metal–Graphene Constructs

Hybrid systems that combine graphene with gold or magnetic nanoparticles aim to add optical or magnetic functionality. In the SMTX-AuNPs/RGO platform, covalent attachment of MTX to gold nanoparticles via a pH-sensitive linker was followed by immobilization on RGO (Figure 9).³⁹ Although the gold component enabled high initial loading, strong π–π interactions with RGO substantially reduced the overall release rate. The PEGylated magnetic GO-PEG/Fe₃O₄ nanocomposite offered the potential for magnetic guidance in addition to pH-responsive release,⁴⁰ yet detailed therapeutic efficacy data in MTX-loaded form remain limited.

Comparative Assessment of Multifunctional Platforms

When the various multifunctional and combination strategies are viewed side by side, a clear hierarchy of therapeutic impact emerges. The most pronounced gains in multidrug-resistant and metastatic models have been achieved by systems that simultaneously exploit three intrinsic advantages of reduced or doped graphene: (i) high mitoxantrone loading capacity, (ii) efficient photothermal conversion under near-infrared irradiation, and (iii) the ability to co-deliver an additional biological effector—either an immune-modulatory agent (TGF-β inhibitor) or a gasotransmitter (nitric oxide). Platforms that combine only two of these elements (for example, chemotherapy plus photothermal therapy without an immunomodulatory or MDR-reversing component) generally produce more modest improvements. Hybrid metal–graphene constructs add optical or magnetic functionality but frequently suffer from reduced release efficiency or limited in-vivo validation, thereby diminishing their overall therapeutic advantage.

These multimodal designs capitalize on physicochemical properties of graphene that are difficult to replicate with purely organic nanocarriers. At the same time, the progressive increase in functional complexity exacts a measurable cost: synthetic routes become longer and less reproducible, batch-to-batch variability rises, and comprehensive safety evaluation becomes substantially more demanding. Long-term biocompatibility, potential immunogenicity of multi-component assemblies, and the still-complete absence of clinical data remain the principal barriers that separate the impressive preclinical efficacy observed in resistant and metastatic models from realistic translational prospects. Future development should therefore prioritize simplification of the most effective multi-stimuli architectures while retaining their synergistic therapeutic mechanisms.

Comparative Overview of Graphene-Based Nanocarriers for MTX Delivery

To facilitate direct comparison of the diverse strategies reviewed above, the principal GO-, NGO-, and rGO-based systems developed for MTX delivery are summarized in Table 3. The table systematically records targeting strategy, stimuli-responsiveness, drug-loading capacity, key design features, experimental models, main outcomes, and principal limitations.

Table 3.

Comparative Overview of Major Graphene and Graphene Oxide-Based Nanocarriers for Mitoxantrone (MTX) Delivery

System/Year	Targeting Strategy	Stimuli-Responsiveness	Drug Loading	Key Features	Experimental Model	Main Outcomes	Limitations
HA-NGO/Pluronic15 (2015)	HA (CD44) + Pluronic (P-gp inhibition)	pH + NIR	45 wt%	High loading, MDR reversal	MCF-7 and MCF-7/ADR (in vitro and in vivo)	Enhanced intracellular accumulation; improved efficacy in resistant tumors	Limited long-term toxicity data
MSC-NGO-Drug36 (2018)	Mesenchymal stem cells (tumor-homing)	pH	>30%	Living carrier, low immunogenicity	In vitro co-culture	Selective cancer-cell killing with preserved MSC viability	Only in vitro; no in vivo data
EXO-GO-CO-γ-PGA37 (2022)	Cancer-cell-derived exosomes (homologous targeting)	pH	73% (with exosome)	Biomimetic coating, immune evasion	MDA-MB-231 (in vitro)	Improved uptake and selectivity	Lack of in vivo and toxicity studies
GOMNP-MitP38 (2021)	Mitochondrial targeting peptide + Magnetic	AMF + pH	19%	Mitochondria-specific delivery	In vitro cancer cells	>80% apoptosis; mitochondrial dysfunction	In vitro only; complex magnetic setup
SMTX-AuNPs/RGO39 (2017)	Covalent linkage + EPR	pH	High (covalent)	Hybrid gold–graphene	MCF-7 (in vitro)	High release from AuNPs at acidic pH	Strongly reduced release after RGO immobilization
FA-PEG-cGQD41 (2019)	Folic acid (folate receptor) + PEG	pH	40.1% (97.5% EE)	Quantum dots, active targeting	HeLa (in vitro and in vivo)	Significant tumor-growth inhibition; low systemic toxicity	Limited to folate-receptor-positive tumors
r-NGO-PEG/MTX/SB42 (2020)	PEG (EPR) + SB-431542 (TGF-β inhibition)	NIR (PTT)	High	Photo-chemo-immunotherapy	4T1 (in vitro and in vivo)	70% long-term survival; systemic immunity	Complex multi-component system
N-GO-MTX-BNN616 (2021)	Nitrogen doping + EPR	pH + NIR + NO release	Up to 424% (relative)	Gas therapy + PTT + Chemo; MDR reversal	MCF-7/ADR (in vitro and in vivo)	Near-complete tumor suppression in resistant model	Scalability and precise NO-control challenges

Systems that combine active or biomimetic targeting with multi-stimuli responsiveness (particularly HA-NGO/Pluronic and N-GO-MTX-BNN6) have produced the most consistent improvements in intracellular accumulation and antitumor activity in multidrug-resistant models. Biomimetic coatings improve short-term immune evasion and homologous targeting but remain largely confined to in-vitro evaluation. Mitochondrial and hybrid metal–graphene constructs demonstrate elegant mechanistic concepts yet have not progressed beyond cell-culture studies. Across all platforms the graphene sheet functions primarily as a high-capacity, easily functionalized scaffold; the greatest therapeutic gains arise when this scaffold is paired with an additional biological effector (P-gp inhibition, NO-mediated MDR reversal, or immune modulation). Virtually all evidence remains preclinical. Long-term biocompatibility, immunogenicity, batch-to-batch reproducibility, scalable manufacturing, and the complete absence of clinical data constitute the principal barriers to translation.

Safety, Biocompatibility, and Biological Fate of GO/rGO Nanocarriers

Although many of the MTX–GO systems reviewed above report favorable short-term cytotoxicity profiles in cancer cell lines and, in selected cases, reduced systemic toxicity markers in xenograft models, a systematic evaluation of long-term biocompatibility remains largely absent from the primary literature. The following subsections summarize the limited available evidence and highlight critical knowledge gaps.

Hemocompatibility and Short-Term Cytotoxicity

Most published MTX–GO studies include only basic in-vitro viability assays (typically MTT or similar) performed on cancer cell lines. Hemocompatibility data (hemolysis, platelet activation, complement activation) are rarely reported for the specific formulations discussed in this review. Where PEGylation or biomimetic coatings are employed, short-term blood compatibility is generally improved relative to pristine GO, yet standardized protocols and quantitative thresholds are seldom provided.

Oxidative Stress, Inflammation, and Immunogenicity

Pristine graphene oxide is known from the broader nanotoxicology literature to be capable of inducing reactive oxygen species, inflammatory cytokine release, and activation of innate immune pathways. None of the MTX-loaded systems examined here have systematically quantified these responses in relevant immune-cell populations or in vivo over extended periods. Biomimetic coatings (MSC membranes, exosomes) are intended to reduce immune recognition, but direct comparative immunogenicity data versus uncoated GO carriers are lacking.

Genotoxicity

Potential genotoxic effects of GO and rGO (DNA damage, micronucleus formation, etc.) have been reported in the general literature and appear to depend strongly on lateral size, oxidation degree, and surface coating. No genotoxicity assessment has been published for the MTX-loaded formulations reviewed in this article.

Biodistribution, Accumulation, and Clearance

Short-term biodistribution studies in the MTX–GO literature are limited and typically focus on tumor accumulation. Long-term organ retention (particularly in liver, spleen, and lung), biodegradation pathways, and routes of clearance (renal versus hepatobiliary versus macrophage-mediated) remain poorly characterized for these specific systems. The broader graphene literature indicates that larger or poorly functionalized sheets can persist in the reticuloendothelial system for weeks to months. Whether the surface modifications used in the MTX carriers (PEG, HA, exosomes, etc.) substantially accelerate clearance has not been rigorously demonstrated.

Chronic Toxicity and Translational Implications

No chronic toxicity studies (repeat-dose, reproductive, or carcinogenic assessments) have been reported for any of the MTX–GO platforms discussed. Given that no graphene oxide-based drug-delivery system has yet received clinical approval for cancer therapy, the absence of such data represents a major translational barrier. Future work should prioritize standardized hemocompatibility testing, quantitative biodistribution and clearance studies with appropriately labeled carriers, evaluation of oxidative stress and inflammatory markers in vivo, and, ultimately, long-term safety packages that meet regulatory expectations for novel inorganic nanocarriers.

In summary, while the reviewed MTX–GO systems frequently demonstrate acceptable acute cytotoxicity profiles and, in some cases, reduced systemic toxicity relative to free MTX in preclinical models, the long-term safety, immunogenicity, biodegradation, and clearance of these materials remain inadequately defined. Addressing these gaps is essential before clinical translation can be realistically considered.

Future Perspectives

Progress toward clinical translation will require coordinated advances on several fronts:

Material Standardization and Characterization

Reproducible control of lateral size, oxidation degree, layer number, and surface chemistry must be established, together with validated analytical methods for batch release.

Comprehensive Safety Package

Systematic evaluation of hemocompatibility, oxidative stress, inflammatory responses, genotoxicity, long-term biodistribution, biodegradation, and chronic toxicity is essential. Comparative studies against clinically approved liposomal and polymeric carriers would help position GO/rGO systems realistically.

Regulatory challenges specific to graphene-based materials remain substantial. Unlike liposomal formulations that benefit from established CMC (Chemistry, Manufacturing and Controls) frameworks and decades of clinical precedent, GO/rGO nanocarriers are classified as novel inorganic nanomaterials. Key unresolved issues include: (i) the absence of validated methods for quantifying residual graphene sheets in biological matrices, (ii) uncertainty regarding long-term organ accumulation and the potential for delayed toxicity, and (iii) the lack of clear guidance from regulatory agencies (FDA/EMA) on acceptable levels of graphene-related impurities and degradation products. Manufacturing cost and complexity are currently higher than those of approved liposomal systems (eg, Doxil^® or Onivyde^®) owing to batch-to-batch variability in lateral size and oxidation degree and the need for multi-step surface functionalization. Comparative head-to-head studies evaluating cost-of-goods, scalability under GMP conditions, and long-term clearance kinetics versus clinically approved polymer- and lipid-based carriers are therefore essential before any realistic translational pathway can be defined.

Improved Circulation and Clearance

Although PEGylation and biomimetic coatings improve short-term stability, circulation half-lives remain modest. Strategies that more effectively reduce protein-corona formation and accelerate renal or hepatobiliary clearance of degraded fragments should be prioritized.

Head-to-Head and Mechanistic Studies

Direct comparisons of GO versus rGO, of different targeting ligands, and of single- versus multi-stimuli systems under identical experimental conditions are needed to identify the design features that most reliably improve therapeutic index.

Translational Development Path Scalable

GMP-compatible manufacturing routes, robust quality-control specifications, and early regulatory engagement will be required before first-in-human studies can be contemplated.

Only when these scientific and technical gaps are addressed can the preclinical advantages observed with MTX–GO/rGO nanocarriers be rigorously tested in the clinical setting.

Conclusion

Graphene oxide (GO), nano-graphene oxide (NGO), and reduced graphene oxide (rGO) have been extensively explored as nanocarriers for mitoxantrone (MTX) because their large surface area, capacity for π–π stacking with the drug’s anthraquinone core, and versatile surface chemistry enable high loading and stimuli-responsive release. The systems reviewed in this article demonstrate that these materials can achieve drug-loading capacities frequently exceeding 40 wt%, support pH-, NIR-, or magnetic-field-triggered release, and, when appropriately functionalized, improve intracellular accumulation in both drug-sensitive and multidrug-resistant cancer cell lines. Selected platforms that combine active or biomimetic targeting with additional therapeutic modalities (photothermal therapy, nitric-oxide release, or immune modulation) have produced enhanced tumor-growth inhibition and reduced systemic toxicity markers relative to free MTX in preclinical xenograft models.

These findings remain confined to the preclinical stage. The majority of studies report only short-term in-vitro cytotoxicity or limited in-vivo efficacy data. Long-term biocompatibility, immunogenicity, biodegradation pathways, organ accumulation, and clearance kinetics have not been systematically characterized for the specific MTX-loaded formulations. No graphene-based MTX delivery system has entered clinical evaluation, and batch-to-batch reproducibility and scalable manufacturing processes are still underdeveloped.

In their present form, GO/rGO nanocarriers therefore offer a flexible preclinical platform for addressing key pharmacological limitations of MTX—particularly nonspecific biodistribution, dose-limiting cardiotoxicity, and efflux-mediated resistance—but they cannot yet be regarded as clinically validated alternatives to existing drug-delivery technologies.

Funding Statement

The authors have no affiliations with any organization that has a direct or indirect financial interest in the subject matter discussed in the manuscript.

Author Contributions

All authors made a significant contribution to the work reported, whether that is in the conception, study design, execution, acquisition of data, analysis and interpretation, or in all these areas; took part in drafting, revising or critically reviewing the article; gave final approval of the version to be published; have agreed on the journal to which the article has been submitted; and agree to be accountable for all aspects of the work.

Disclosure

The authors report no conflicts of interest in this work.