Harnessing for quantitative MRI, bioluminescence, chemo-magnetic hyperthermia, and near-infrared optical imaging of a magnetic nanotheranostic for in vivo passive targeted monitoring in colorectal cancer models

Abstract

Colorectal cancer is the second foremost cause of cancer-related fatalities, but it is currently missing an effective treatment. For researchers, the ability to monitor therapeutic nanoparticles of interest within a living animal using non-invasive observation tools has been a futuristic idea for a long time. In this work, a multifunctional magnetic nanotheranostic functionalized with Cyanine 5.5 and Doxorubicin was tested to enhance bio-monitoring and therapy efficiency in an in vivo CT26 murine colon model via retro-orbital injection. The in vivo biodistribution was measured by near-infrared optical imaging, which showed that the nanotheranostic largely accumulated in the tumor with maximum uptake observed at the 3 h postinjection time point. This nanotheranostic was then metabolized and eliminated via the kidneys after 6 h and both the liver and spleen after 72 h of injection. Besides, the magnetic resonance imaging (MRI) modality with a specific T2*-weighted sequence demonstrates efficient nanoparticle accumulation within the tumor by the %I_0.25 quantitative method of hyposignal processing. Moreover, bioluminescence imaging demonstrated a significant chemotherapeutic effect after 72 h of injection of a 15 mg kg⁻¹ single dose. Hyperthermia treatment by AMF significantly impacts the synergistic efficiency of this second therapy provided by the Fe₃O₄ nanoparticle (NP) content. The resulting nanotheranostic demonstrated enhanced passive accumulation and selectively accumulated in the tumor with negligible distribution to adjacent healthy tissue, effectively suppressing tumor proliferation when combined with alternating magnetic field (AMF) stimulation. This powerful synergistic approach has proven to be a robust and versatile nanotheranostic for the effective treatment of colorectal cancer.

---

Introduction

Globally, colorectal cancer (CRC) is the second foremost cause of cancer-related fatalities, with its incidence positioned third.^1,2 In addition to age, other factors influence the onset of CRC, including a history of colon polyps, diabetes mellitus, inflammatory bowel illnesses or cancers, and detrimental lifestyle choices such as obesity, alcohol intake, smoking, and inadequate nutritional practices. The increasing prevalence of CRC in those under 50 years of age is a notable trend.³ The current clinical paradigm for CRC detection comprises various modalities, each with significant limitations. Colonoscopy remains the gold standard, offering direct visualization, diagnosis, and therapeutic removal of precancerous polyps in a single procedure.⁴ However, its widespread adoption is hindered by its invasive nature, rigorous bowel preparation, and procedural risks (e.g., perforation), which contribute to suboptimal patient compliance.^5,6 In contrast, non-invasive alternatives like the fecal immunochemical test (FIT) are widely used for population-level screening due to their low cost and ease of use. Yet, their limited sensitivity for detecting early-stage, non-bleeding adenomas frequently results in false negatives and delayed diagnosis.⁷ While stool DNA tests offer improved sensitivity, their higher cost and notable false-positive rates often necessitate follow-up invasive colonoscopies.^8,9 Similarly, traditional serum biomarkers like carcinoembryonic antigen (CEA) are unsuitable for primary screening due to poor sensitivity and specificity for early-stage disease, limiting their use to recurrence monitoring.^10,11 Collectively, these limitations highlight a critical unmet need for non-invasive tools that can detect early-stage CRC with high accuracy and enable real-time tracking of therapeutic response.

Surgery, chemotherapy, and radiation have demonstrated efficacy in enhancing survival rates; nevertheless, they are accompanied by disadvantages, including adverse effects and damage to healthy tissues.¹² As such, during diagnosis and monitoring of treatment progress, knowing the extent of the disease is crucial for planning appropriate therapy and evaluating prognosis. Cancer progression is the result of a combination of factors, often not tumor-specific but involved in various pathophysiological processes. Furthermore, tumor cells are highly heterogeneous and have high mutation rates, leading to pathological differences between different cancer types, individuals, and regions within tumors. To address this heterogeneity, multifunctional combination therapies in cancer diagnosis and treatment based on nanomaterials are being focused on. Using nanomedicine in association with imaging probes as nanotheranostics offers a potentially efficient strategy to develop, optimize, and monitor therapeutics.^13,14 Imaging technologies are very beneficial for detecting the stage of cancer. Furthermore, imaging is essential for assessing medication efficacy and identifying tumor recurrence and metastases.^15,16 However, a critical gap remains in the literature: most studies rely on conventional tail-vein administration and often lack a comprehensive, quantitative assessment of the nanocarrier's dynamic in vivo journey, from the initial biodistribution and tumor accumulation to long-term clearance and true therapeutic synergy. The novelty of our work lies in addressing this gap by evaluating a magnetic nanotheranostic through the underexplored retro-orbital route, coupled with a powerful suite of quantitative, multi-modal imaging techniques to provide an unprecedentedly detailed picture of its performance and fate.

Nanomedicine is one of the efficient strategies for tackling this health issue. The limitation of chemotherapy in terms of drug tolerance can be overcome when combined with other modalities, such as magnetic hyperthermia. This strategy will increase the response of cancer cells to anti-cancer drugs, and immunotherapy can increase the body's defence. Combining both therapies results in a lower dose for the same therapeutic effect or the same dose with a higher therapeutic effect and fewer side effects.¹⁷

Fe₃O₄ nanoparticles (NPs) have the potential to increase the contrast of MRI images and exploit the magnetic hyperthermia effect to treat cancer when placed in an external magnetic field.^18,19 Several of our previous studies have explored the in vivo MRI contrast enhancement effect of Fe₃O₄ NPs coated with polymers.^20,21 However, the results were limited to qualitative in vivo MRI signal quantification in a mouse model, which provided a preliminary picture of the distribution of Fe₃O₄-coated NPs. Moreover, Cyanine 5.5 is a near-infrared fluorophore that facilitates the in vivo monitoring of drug distribution during treatment by optical imaging.²² The special feature of the strong fluorescence in the near infrared region is that it provides fluorescence images with little background scattering as well as the ability to penetrate deep tissue.²³ Simultaneously, Cyanine 5.5 has significant protein-binding properties, allowing it to be retained and accumulated along with the multifunctional nanotheranostic at the tissue site.²⁴ This advantage can create suitable conditions for detecting and monitoring the biological process of the drug delivery system, promising to bring high efficiency in monitoring and treating cancer.^24–26 Previous in vitro studies by our group using Fe₃O₄ NPs modified with Cyanine 5.5 or Doxorubicin have shown that this nanoplatform specifically has a cytostatic effect, inducing cell cycle arrest at the G2/M phase in human brain adenocarcinoma (CCF–STTG1) cells.^27,28

However, the quenching of the fluorescence signal caused by Fe₃O₄ NPs makes it difficult to obtain strong fluorescence signals.²⁹ Several previous reports have shown that fluorescence can be quenched by Fe₃O₄ NPs when their spacing is less than 10 nm.³⁰ This is a common problem faced by all nanotheranostics that combine Fe₃O₄ NPs and NIR agents. Therefore, effectively mitigating fluorescence quenching while harnessing the expression of the Fe₃O₄ NP core and elucidating the in vivo fate following injection via NIR optical imaging becomes a critical area of focus in advancing this nanotheranostic modality.

Typically, the lateral tail vein is used to deliver nanoparticles intravascularly to the adult mouse. This procedure is technically challenging and often has a high rate of failure, as mice's tail veins are fragile and difficult to cannulate.³¹ The mouse is usually put under a heat lamp to enhance peripheral vasodilation before being physically restrained for tail vein venipuncture. This procedure might cause suffering to the animals, especially if the initial venipuncture fails and further tries are performed.³² Furthermore, repeated injections might develop vein fibrosis, making the procedure more difficult.³³ Retro-orbital injection is an alternate delivery route that is technically easier, quicker, and repeatable with no complications associated with repetition.³⁴ Mice's retro-orbital region is rich in capillary microcirculation, which increases absorption and can avoid the first-pass metabolism occurring in the liver.³⁵

To date, no comprehensive and quantitative studies that focus on the biodistribution of nanotheranostics via retro-orbital administration in organisms and their tumor-targeting effectiveness in colorectal cancer have been reported. In order to fill this gap, we aim to perform physicochemical characterization, along with quantitative MRI, bioluminescence, chemo-magnetic hyperthermia, and NIR optical imaging of PLA-TPGS-Fe₃O₄-Doxorubicin-Cyanine 5.5 NPs, as a bimodal nanotheranostic scaffold against tumors on CT26 colon carcinoma-bearing murine models. To combine MRI and NIR optical imaging functions in a multimodal imaging system, Cyanine 5.5 was attached to the surface of Fe₃O₄ NPs. This dual NIR/MRI functionality offers precise monitoring and multimodal capabilities that are challenging to achieve with conventional nanoparticles. Once the response of the Fe₃O₄ NPs’ surface-adorned Cyanine 5.5 was optimized by varying the proportions of components, pH, temperature, and reaction time, these materials were integrated along with Doxorubicin within the polymeric micelle aggregates. Doxorubicin was encapsulated in micelle aggregates decorated with Fe₃O₄ NPs to develop a multi-responsive drug delivery magnetic hyperthermia nanotheranostic. Fig. 1 illustrates the multimodal nanotheranostic strategy. The platform is administered via retro-orbital injection for passive tumor targeting, which is quantified by MRI and NIR optical bioimaging. It then serves a dual therapeutic and diagnostic purpose: achieving a potent synergistic effect through chemo-magnetic hyperthermia and enabling quantitative monitoring of the chemotherapeutic response via bioluminescence.

Fig. 1 Schematic representation of PLA-TPGS-Fe₃O₄-Doxorubicin-Cyanine 5.5 nanotheranostic treatment for magnetic hyperthermia and chemotherapy, passive targeting by quantitative MRI/bioluminescence, and NIR optical bioimaging monitoring.

Results and discussion

A strategic element of this study is the evaluation of our nanotheranostic following retro-orbital administration. While a direct comparison with conventional tail-vein injection was not performed, this route was selected for its significant technical and pharmacokinetic advantages. Retro-orbital injection not only reduces procedural stress and failure rates but also offers a distinct pharmacokinetic profile, enabling rapid systemic entry that may partially bypass immediate first-pass hepatic metabolism. This profile is theorized to enhance nanoparticle bioavailability, potentially leading to a greater tumor area under the curve (AUC) and improved therapeutic outcomes compared to routes where RES sequestration is more pronounced. Our findings thus provide a crucial performance benchmark for complex nanotheranostics delivered via this potent, underutilized route.

Synthesis and physicochemical characterization of the PLA-TPGS-Fe₃O₄-Doxorubicin-Cyanine 5.5 nanotheranostic

Initially, 3-aminopropyl triethoxysilane (APTES) was attached to the surface of Fe₃O₄ NPs and Cyanine 5.5 was then incorporated by the reaction between Cyanine 5.5 NHS ester and APTES. Several nanotheranostic samples were prepared with varied temperatures, reaction times, pH, and component ratios to optimize the emission intensity of the obtained samples. As illustrated in Fig. 2A, APTES molecules act as an intermediate layer to bind the Fe₃O₄ NPs’ surface to Cyanine 5.5 molecules. APTES is an organosilane molecule commonly used in silane-based functionalization processes to attach biomolecules to the surface. The surface of Fe₃O₄ NPs has hydroxyl groups having high surface energy, which can rapidly interact and form covalent bonds with silane molecules. Therefore, Cyanine 5.5 molecules can be chemically attached to the silane layer via covalent bonding (NH–CO) without disrupting the silane structure. Fig. 2B shows that the fluorescence intensity increases with the decrease in the Fe₃O₄ NPs : APTES (w/w) ratio. However, in cases of a too high concentration of APTES, specifically at an Fe₃O₄ to APTES (w/w) ratio of 5 : 3, the fluorescence intensity decreased compared to the ratio of 10 : 3. Therefore, the ratio of Fe₃O₄ NPs to APTES (w/w) = 10 : 3 was used for the next steps. In this step, a condensation reaction can occur between two APTES molecules and/or between the surface of Fe₃O₄ NPs and the APTES molecule. The condensation reaction between silanol groups on APTES molecules and –OH groups on the Fe₃O₄ NPs’ surface occurs to form siloxane bonds (Fig. S1A). Siloxane linkages can covalently attach silane molecules to surfaces in various ways. In addition, the silanol groups of different silane molecules can interact, forming polymeric structures.³⁶

Fig. 2 (A) Schematic diagram of the synthetic procedure for Cyanine 5.5-attached surface Fe₃O₄ NPs, and (B–F) fluorescence spectra under different conditions of temperature, pH, reaction time, and component ratios.

Moreover, during the phase separation process, the reaction media become heterogeneous, resulting in the formation of more substantial polymer structures.³⁷ Hence, altering the temperature during APTES grafting can enhance the response rate and the effectiveness of APTES grafting. As the temperature of the APTES attachment reaction increased from 25 to 80 °C, the fluorescence intensity gradually increased with the emission maximum at 712 nm (Fig. 2C). There was no difference in maximum fluorescence intensity between 70 °C and 80 °C, so the 70 °C temperature condition was used for the next steps. Elevated temperature incubations facilitated the establishment of stable covalent siloxane bonds while breaking hydrogen bonds. The desorption of APTES molecules at higher temperatures occurred more slowly than that from APTES annealed at room temperature.³⁸ Thus, the efficiency of the APTES grafting reaction is higher at high temperatures. Another report by Yang et al. showed that the reaction temperature not only promoted the grafting reaction but also promoted the APTES arrangement.³⁹ As a result, the alternating pseudo-bilayer arrangement of APTES molecules was rearranged into a monolayer arrangement.

For the second step of Cyanine 5.5 incorporation, the fluorescence intensity of the samples directly depended on the concentration of Cyanine 5.5, reaching the highest intensity at a Cyanine 5.5 : Fe₃O₄ NP ratio of (w/w) = 0.05 : 1 (Fig. 2D). However, if the Cyanine 5.5 concentration was too high in the ratio of Cyanine 5.5 : Fe₃O₄ NPs (w/w) = 0.1 : 1, it gave a much lower fluorescence intensity than that for the ratio of 0.05 : 1. This phenomenon may be related to the number of Cyanine 5.5-NHS ester molecules that actually reacted with the NH₂ groups on the surface of Fe₃O₄ NPs to convert Cyanine 5.5 NHS ester to free Cyanine 5.5. The amount of NH₂ on the Fe₃O₄ NPs’ surface depended on the concentration of APTES added to the previous reaction at the functionalization step of the Fe₃O₄ NP surface. Therefore, the ratio of Cyanine 5.5 : Fe₃O₄ NPs (w/w) = 0.05 : 1 was maintained in the following steps.

On the APTES-modified surface of Fe₃O₄ NPs, Cyanine 5.5 molecules can be bound via activation of their carboxy groups with carbodiimides (N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride, EDC) and N-hydroxysuccinimide (NHS), which are converted into more stable peptide bonds via reductive amination using APTES, and then activated with EDC/NHS (Fig. S1B). The efficiency of this reaction is directly dependent on pH. The ability to attach Cyanine 5.5 to the Fe₃O₄ NPs increases with increasing pH and the highest fluorescence intensity is achieved at pH 8.5 (Fig. 2E). In an acidic environment (pH 5.5), NH₂ groups are protonated to NH₃⁺, which diminishes the capacity to conjugate Cyanine 5.5 to the Fe₃O₄ NP surface, resulting in a significantly low fluorescence signal from the NPs. In a strong alkaline (pH 10) environment, the hydrolysis of NHS ester occurs rapidly, and the denaturation efficiency decreases. Hence, the fluorescence intensity at pH 10 is lower than that at pH 8.5. Therefore, the optimal pH condition of 8.5 was maintained in subsequent experiments. Finally, the reaction time significantly affects the efficiency of Cyanine 5.5 attachment to the Fe₃O₄–NH₂'s surface under EDC/NHS catalytic conditions (Fig. 2F). At longer reaction times, the fluorescence intensity increased. The fluorescence intensity grew dramatically by 3.5 times between 1 hour and 2 hours and by 8 times after 5 hours. When the time was further increased to 12 hours, the increase was marginal. Therefore, a reaction time of 5 hours for the Cyanine 5.5 attachment step was selected for the following experiments.

The optimized PLA-TPGS-Fe₃O₄-Cyanine 5.5 NPs were then used to prepare PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs. Field Emission Scanning Electron Microscopy (FESEM) and Transmission Electron Microscopy (TEM) images of this nanotheranostic exhibited a uniform spherical shape (Fig. 3A and B), with the TEM average particle sizes of approximately 11.8 ± 0.2 nm (Fig. 3C). Fig. S2 reveals an average hydrodynamic diameter of approximately 127.5 nm, as determined by Dynamic Light Scattering (DLS), and a narrow particle size distribution, indicating the high uniformity of the NPs’ decentralized system. This hydrodynamic diameter of 127.5 nm is of critical biological importance, as it is ideally situated within the optimal window for exploiting the Enhanced Permeability and Retention (EPR) effect. Particles of this size are large enough to prevent rapid renal filtration yet small enough to effectively permeate the leaky, fenestrated vasculature unique to tumor tissues.⁴⁰ Crucially, upon extravasation, the tumor's impaired lymphatic drainage is expected to trap these nanoparticles, leading to prolonged retention.^41,42 Therefore, the physicochemical properties of our nanotheranostic are intentionally engineered to leverage the physiological hallmarks of the tumor microenvironment, setting the stage for the efficient passive accumulation demonstrated in subsequent in vivo analyses. This characteristic confers upon the NPs a passive targeting capability to accumulate in tumors via the permeability and retention (EPR) effect. A zeta potential value of −32.2 mV (Fig. S3) indicates high stability of the synthesized NPs, great solubility and dispersion while mitigating aggregation tendencies. Remarkably, the stability of NPs was maintained over a three-month storage period, with minimal alterations observed in solutions across various media (pure water, phosphate-buffered saline (PBS), high-glucose Dulbecco's modified Eagle's medium (DMEM), a mixture of PBS + 10% fetal bovine serum (FBS), and a mixture of DMEM + 10% FBS at 37 °C (Fig. 3D and E)). Most importantly, we found that PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs had greatly improved stability in the five treated media, such that only very minimal Cyanine 5.5 detachment was observed when non-denatured serum was added to a solution of the nanotheranostic in PBS and DMEM, with under 20% release for 72 h (Fig. 3F). There is no statistical difference in drug release between the five groups over time. This robust stability profile highlights the suitability for prolonged storage and application under diverse experimental conditions of synthesized NPs. Assessment of the drug loading capacity (DLC) and encapsulation efficiency (EE) of Dox in NPs (0.25 mg mL⁻¹) revealed the DLC and EE values of 10.67 ± 1.25% and 87.4 ± 1.83%, respectively (Fig. 3G).

Fig. 3 (A) FESEM and (B) TEM images of PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs, (C) TEM particle size distribution histogram of (B) obtained using ImageJ; the changes of size (D), zeta potential (E) of PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs within 3 months, and (F) the release of Cyanine 5.5 from NPs monitored in pure water, PBS, DMEM, PBS + 10% FBS, and DMEM + 10% FBS; (G) the encapsulation efficiency and loading efficiency of Dox and Cyanine 5.5 in NPs; (H) full X-ray photoelectron spectroscopy (XPS) spectra of NPs, (I–L) high-resolution XPS spectra of Fe 2p, O1s, N 1s, and C 1s in PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs.

X-ray diffraction (XRD) measurements were used to characterize the structure of the Fe₃O₄ NPs and PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs. The XRD pattern showed six well-resolved diffraction peaks (Fig. S4), which were indexed to the (220), (311), (400), (422), (511), and (440) planes of the cubic Fe₃O₄ phase.^43,44 The average nanocrystal size was ∼10.3 nm, calculated using the Scherrer formula for the (311) peak; this agreed well with the TEM data. The chemical properties of NPs were further studied using X-ray Photoelectron Spectroscopy (XPS). As shown in Fig. 3H, the XPS spectra of NPs show the existence of Fe 2p, O 1s, N 1s, and C 1s. Fig. 3I shows the XPS spectra of the Fe 2p region that can be assigned to the presence of iron on the structure of NPs. The binding energies of Fe 2p in NPs were fitted to two characteristic peaks of 2p_1/2 and 2p_3/2 at 725 eV and 711.6 eV, respectively. These peaks are consistent with the Fe₃O₄ form, according to another research, close to the values in the published literature about Fe₃O₄ NPs.^45,46 The O 1s profile of XPS spectra was fitted to three different characteristic peaks (Fig. 3J). The peak at 531 eV is due to the Fe–O bond of the Fe₃O₄ nanoparticles. The C–OH peak was observed at 531.9 eV, and the O–C=O peak was observed at 533 eV, which is regarded as the indicator of the Doxorubicin molecule. The analysis of the high-resolution N 1s core-level spectra furnished three different contributions (Fig. 3K). The first deconvoluted peak observed at 398.5 eV could be attributed to the N–H peak. The other two contributions found at lower binding energies could be attributed to the C–N peak and the pyrrolic N of the Cyanine 5.5 core, which have slightly different binding energies in the range from 399.9 to 401.2 eV because they are differently involved in the stabilization of the negative charge density of Cyanine 5.5.⁴⁷ Furthermore, the C 1s profile of XPS spectra was fitted to four characteristic peaks (Fig. 3L). The C–O–C=O peak was observed at 288.8 eV, and the O–C=O peak was observed at 290.8 eV. The C 1s peaks at 284.6 and 286.4 eV are due to C–C/C–H and C–O–C bonds, respectively, that arise from the PEG component of the PLA-TPGS copolymer.⁴⁸ Considered as a whole, XPS analysis of PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs proves the presence of Fe₃O₄ NPs, Doxorubicin, Cyanine 5.5, and ultimately, the PLA-TPGS copolymer.

Measurement of the heating ability and Doxorubicin release study in vitro

One of the most important criteria for practical applications is the ability of the synthesized nanoparticles to respond to an external magnetic field, which would encourage their biomedical applications such as magnetic hyperthermia and magnetic resonance imaging. All of the samples showed that the magnetic field was saturated at 11 kOe. The saturation magnetization (M_s) of the Fe₃O₄, PLA-TPGS-Fe₃O₄-Cyanine 5.5, PLA-TPGS-Fe₃O₄-Dox, and PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs was found to be ∼64.9 emu g⁻¹, 58.7 emu g⁻¹, 45.3 emu g⁻¹, and 41.1 emu g⁻¹, respectively (Fig. S5). The comparatively lower saturation magnetization for the three multicomponent NPs, as compared to the bare Fe₃O₄ NPs, was attributed to the presence of the non-magnetic (PLA-TPGS, Cyanine 5.5, and Doxorubicin) components. Thus, the total magnetic moment per unit mass decreases, leading to a decrease in the saturation magnetization. While a reduction in M_s is noted, a pivotal consideration for clinical translation is the system's functional performance. The final M_s value of 41.1 emu g⁻¹ is, in fact, robust and highly competitive, placing it well within the efficacious range of other advanced nanotheranostic platforms that have demonstrated significant preclinical success.^49,50 This value was higher than that of similar polymer-coated and drug-loaded Fe₃O₄ NPs.^51–53 Crucially, the decisive metrics for therapeutic and diagnostic utility are not the M_s value in isolation, but the functional outputs under applied fields. The excellent specific absorption rate (SAR) of NPs and the strong T2* MRI contrast enhancement observed in vivo (as detailed later) provide definitive proof that the magnetic response is sufficient for its intended applications. This confirms that an optimal balance was achieved, enabling potent therapeutic effects at a well-tolerated dose and thus mitigating a key translational hurdle.

Furthermore, the observed room temperature coercivity was found to be insignificant (the bottom inset of Fig. S5), owing to inter-particle interaction and size polydispersity.^20,54 Conversely, low H_c not only enhances the contributions of Néel and Brownian relaxation processes but also demonstrates the superparamagnetic properties of the Fe₃O₄ core nanoparticles, aligning with one of the essential criteria for clinical treatments.⁵⁵ Otherwise, the aggregation of Fe₃O₄ core NPs would form a clot in the blood circulation system.⁵⁶ The magnetic characteristics of NPs are derived from the electron orbit and spin motion, and their intensity is proportional to the intensity of electron interactions. High M_s causes NPs to interact strongly in an alternating magnetic field during magnetization, demagnetization, and reverse magnetization, converting more electromagnetic energy into heat energy.^57,58 The high saturation magnetization of the synthesized NPs was found to influence the AMF-induced heating efficiency significantly, as discussed subsequently.

Fig. 4A shows the time dependence of temperature at different alternating magnetic fields (AMFs) (100, 150, 200, and 250 Oe) and a constant frequency of 450 kHz in the inductive heating experiment of each nanosystem. As can be seen, the more significant the variation in AMF, the greater the temperature increase over time. The highest temperature rise is obtained for PLA-TPGS-Fe₃O₄-Cyanine 5.5 NPs, with the temperature rising sharply with time, up to 800 s. At the same magnetic field amplitude, the presence of Doxorubicin and Cyanine 5.5 molecules lowered the heating capacity of the NPs. This observation is a direct functional consequence of the successful encapsulation of the therapeutic and imaging agents. For PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs, the change in temperature increased with the amplitude of the AMF, from a maximum of 60 °C at 100 Oe up to a maximum of 93.3 °C at 250 Oe. After about 1500 s of heating, the energy loss to the surrounding medium became equal to the energy generated by the AMF, and the sample temperature stabilized. Temperature regulation is critical during hyperthermia to keep it within the vicinity of the hyperthermia limit. Moderate temperatures can selectively harm tumors or aberrant cells while protecting healthy cells.⁵⁹ The obtained specific absorption rate (SAR) values were calculated from the Box–Lucas fitting equation⁶⁰ and entered into the following equation:

where C, m_NPs, and dT/dt are the specific heat capacity of water, the NPs’ mass, and the temperature change slope, respectively. The substantial rise in temperature of the magnetic nanoparticles resulted in high SAR values in all the cases, which agreed with Rosensweig's model.⁶¹ From Fig. 4B, SAR values were found to increase along with AMF amplitude. As can be seen, the SAR value of PLA-TPGS-Fe₃O₄-Cyanine 5.5 nanoparticles was improved from 200.3 ± 6.6 to 386.8 ± 10.4 W g⁻¹ by increasing the AMF from 100 to 250 Oe. For PLA-TPGS-Fe₃O₄-Dox NPs, the SAR value was lower, with 108.3 ± 4.2 W g⁻¹ at 100 Oe and 244.7 ± 7.5 W g⁻¹ at 250 Oe. It should be noted that the amount of Dox in the system was 0.25 mg ml⁻¹, five times higher than the Cyanine 5.5 concentration. This could lead to a reduction in the SAR value. In the case of PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs, SAR values were almost the same as the value of PLA-TPGS-Fe₃O₄-Dox NPs. In addition, the lowest and highest SAR values for PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs were 115.1 ± 3.8 W g⁻¹ under an AMF strength of 100 Oe and 256.2 ± 7.8 W g⁻¹ under an AMF strength of 250 Oe, respectively. Besides, the addition of these diamagnetic components, particularly Doxorubicin, leads to a reduction in the overall saturation magnetization per unit mass of the nanoparticle. This lower magnetic moment per particle directly translates into a reduced capacity for energy dissipation under an AMF, thus explaining the lower SAR values for the fully loaded nanotheranostic. However, it is critical to note that the final SAR values remain robust and are proven to be sufficient for inducing effective hyperthermia in vivo, highlighting the successful optimization for synergistic therapy.^62,63 To compare SAR values derived from the literature under various extrinsic parameters such as magnetic field, frequency, size, and concentration of MNPs, intrinsic loss power (ILP, in nHm² per kg) is a more appropriate term.where f and H are the frequency and the intensity of the applied magnetic field, respectively. The maximum ILP value was found for the PLA-TPGS-Fe₃O₄-Cyanine 5.5 NPs (7.0 ± 0.3 nHm² per kg at 100 Oe). It is obvious that the ILP values decrease with increasing magnetic field strength (2.2 ± 0.1 nHm² per kg at 250 Oe), similar to that previously reported by Radoń et al.⁶⁴ Anyway, the ILP values of PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs were in the range of 1.4–4.0 nHm² per kg under different magnetic field intensities. ILP values of these NPs are much higher than those reported in the literature, which indicates that the prepared samples have potential application in magnetic hyperthermia.^57,65,66

Fig. 4 (A) Magneto-inductive heating experiments, and (B) SAR and ILP values of PLA-TPGS-Fe₃O₄-Cyanine 5.5, PLA-TPGS-Fe₃O₄-Dox, and PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs using different AC magnetic fields; (C) drug release triggered by the different AMFs, and (D) a schematic of the triggering mechanism of Dox from PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs.

To evaluate whether Dox release could be modulated by altering AMF in the PBS medium, the PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs were placed at 37 °C in dialysis membranes with a molecular weight cut-off of 3.5 kDa. After monitoring the release of Dox every 1 h for 4 h with a UV-VIS spectrophotometer, the tube was exposed to different AMFs (100–250 Oe, 450 kHz) for 25 min (Fig. 4C). In the first 4 h, about 4.6% of Dox was released before exposure to different AMFs, showing the slow release capability of NPs under physiological pH conditions. After exposure to AMF, the burst increased in Dox release at 5 h. About 10.4% cumulative release of the Dox under the AMF (100 Oe, 450 kHz) was significantly higher than that obtained by incubating the NPs in an external bath at the same temperature. The results of the linear fit of the increments of Dox release controlled by AMF suggested that the release of the drug was correlated with the intensity of AMF (Fig. S6). The above results showed that the Dox released from PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs could be controlled by adjusting the AMF strength (Fig. 4D), with a higher rate of drug diffusion (following Fick's laws) through the PLA-TPGS copolymer matrix.^67,68 The major mechanisms of drug diffusion might be erosion/swelling of the PLA-TPGS copolymer caused by heat production (by Brownian and Néel relaxation) and mechanical force (Brownian motion) from densely loaded Fe₃O₄ NPs under AMF stimulation.⁶⁹ Moreover, the observed drug release behavior may be influenced by the stimulation of water molecule absorption in the polymeric matrix by amphiphilic TPGS of the NP shells.⁷⁰ This mechanism effectively enables a significantly enhanced Dox release to be realized with the simultaneous application of AMF at the tumor site, providing the potential to minimize the side effects of chemotherapy substantially. Hu et al. revealed that the release of Dox facilitated by high-frequency AMF was observed, with the release profile categorized into two phases.⁷¹ A gradual and steady release phase was sustained when the microcapsule structure remained intact. However, a rapid release transpired when the microcapsules were disrupted by high-frequency alternating magnetic fields. Therefore, the frequency of AMF is crucial to control the release behaviors of Doxorubicin-loaded nanocapsules.

Passive targeting monitoring by quantitative in vivo MRI

Nanoparticles have the ability to accumulate in tumor tissues by passive targeting mechanisms due to the EPR effect. However, the tumor microenvironment is characterized by a heterogeneous structure, and depends on the type, size, and location of the tumor; it responds differently to nanoscale material systems.⁷² Besides, the EPR effect depends on many factors, such as (i) vascular leakage, (ii) tumor microenvironment density/composition, (iii) interstitial fluid pressure, (iv) tumor growth rate, (v) necrotic area volume, and (vi) NP dosage used.⁷³ Therefore, when designing a multifunctional nanotheranostic for cancer diagnosis and treatment, the ability of the nanotheranostic to be distributed to the tumor site plays an important role in treatment efficacy.

A critical prerequisite for the clinical translation of any nanotheranostic is the establishment of a favorable safety and efficacy profile in vitro. To this end, we first evaluated the cytotoxicity of the PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs against a standard non-cancerous cell line. We utilized Vero cells, a model recommended by ISO 10993-5 for biocompatibility screening. As shown in Fig. S7, the nanotheranostic exhibited excellent biocompatibility, with cell viability remaining above 70% even at the highest tested concentration (125 μg mL⁻¹) after 48 hours. This demonstrated that the low cytotoxicity against normal cells directly complements our previously reported findings,^27,28 which established the platform's potent cytostatic effect on cancer cells. Together, these results confirm a desirable therapeutic window, highlighting the nanotheranostic's selective toxicity. Having established this robust in vitro safety and efficacy profile, we proceeded to investigate the platform's performance in a complex in vivo colorectal cancer model, focusing on its biodistribution, passive targeting via MRI and NIR bioimaging capabilities, and synergistic therapeutic effects.

In the next step, T2*-weighted MR images of the normal and tumor bearing mice with and without injection of the nanotheranostic were taken with a Fast Low Angle Shot (FLASH) sequence using a 7 T MRI instrument (300WB, Bruker, Avance II, Wissembourg, France). Fig. 5 shows in vivo MRI images of a Balb/C mouse model with a CT26 colon tumor treated with NPs and the control mouse at the 4 h post-retro-orbital-injection time point (dosage: 15 mg kg⁻¹ in 150 μL of the nanotheranostic). Tumor contrast can be clearly seen, and the nanoparticles can be seen dispersed in the tumor (red arrow), demonstrating their ability to enhance the MRI contrast of the solid tumor with T2*-weighted imaging mode when retro-orbitally injected. This effect allows more accurate assessment of the size, shape, and internal structure of the tumor, as well as makes it easier to assess the presence or absence of a tumor inside the body.⁷⁴ Bilateral tumor MRI images display a well-delineated tumor, enabling quantification of both whole 3D tumor dark signals from Fe₃O₄ NP uptake, as visualized in the T2* gradient echo pictures of the tumor. Indeed, as previously demonstrated in studies by Thébault et al., Fe₃O₄ NPs act as T2 type MRI contrast agents, exhibiting strong r₂ relaxivities and producing hyposignals in T2*-weighted images.^75,76

Fig. 5 The axial in vivo T2*-weighted MRI images of CT26 tumors of the control mouse and Balb/C mouse after 4 h of retro-orbital injection.

To quantify the MRI signal and assess the in vivo accumulation of NPs in tumors, we used a previously developed %I_0.25 treatment method.⁷⁶ Regions of interest (ROIs) were drawn on each slice of the tumor on the MRI image (Fig. 6A and C), and pixel intensity distributions were obtained for each slice. The sum of these pixel intensity distribution plots gives a unique pixel intensity distribution for each tumor, as shown in Fig. 6B and D. Fig. 6E shows that the percentage of pixels below I_0.25 was significantly different (p < 0.001) between the uninjected mice and the mice 4 h after nanotheranostic injection, which were 3.35 and 13.82%, respectively. This reflects the effective passive accumulation of the nanotheranostic in these tumors.⁷⁷ This %I_0.25 method is based on intensity distribution correction in standard T2*-weighted MRI images. The NPs containing Fe₃O₄ NPs reduce the signal, causing a shift of the intensity histogram to the low intensity region. Therefore, the percentage of pixels below I_0.25 is a relevant parameter to quantify the accumulation of Fe₃O₄-containing multifunctional nanomaterials at the tumor site. These results demonstrated the in vivo MRI contrast enhancement and effective passive accumulation at the tumor site of NPs in a CT26 tumor-bearing murine model.

Fig. 6 %I_0.25 methodology of data processing from the T2*-weighted MRI images for the evaluation of in vivo passive accumulation in CT26 tumors. (A) T2*-Weighted MRI images and (B) MRI intensity distribution associated with the tumor after retro-orbital injection (dosage: 15 mg kg⁻¹); (C) T2*-weighted MRI images of control mice; (D) MRI intensity distribution associated with the tumor of the control mice; (E) comparison of %I_0.25 values (mean ± SD) in MRI tumor images for two groups on the day of the treatment (n = 6). Non-Gaussian distribution Mann–Whitney test: *p < 0.05; **p < 0.01; ***p < 0.001; nonsignificant p > 0.5.

In vivo biodistribution by NIR optical imaging

In vivo fluorescence optical bioimaging has enabled the monitoring and tracking of therapeutic nanoparticles thanks to the emission of near-infrared (NIR) light from imaging agents grafted to the drug nanovector.⁷⁸ Moreover, to assess targeting efficiency and to follow the progress of treatment in vivo, the biodistribution parameters of therapeutic agents advantageously associated with imaging agents should be controlled. To evaluate the in vivo solid tumor imaging and biodistribution of the multifunctional nanodrug delivery system with near-infrared fluorescence, CT26 tumor-bearing mice were retro-orbitally injected with a single dose of 15 mg kg⁻¹ in 150 μL of NPs, as shown in Fig. 7A. Fluorescence images were acquired using a photon counting system based on a cold GaAs ICCD camera (Optima, Biospace, Paris, France) at excitation/emission wavelengths of 662/722 nm.

Fig. 7 (A) Schematic illustration showing the timeline for optical bioimaging (λ_exc = 662 nm, λ_em = 722 nm) of the CT26 tumor bearing murine model and the mice in both views; (B) longitudinal in vivo whole body optical imaging from 5 min up to 24 h post-retro-orbital-injection of one representative mouse treated with PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs (dose: 15 mg kg⁻¹) (n = 5); and (C) quantitative analysis of the biodistribution in (B); and (D) determining kinetic changes in tumor : liver and liver : kidney ratios.

Fluorescence imaging in vivo showed the clear kinetic biodistribution and uptake of the NPs in the mouse body after retro-orbital injection of the formulations. Fig. 7B depicts the in vivo optical temporal distribution of NPs at the following time points: 5 min, 15 min, 30 min, 1 h, 3 h, 6 h, and 24 h post-injection. Fluorescence signals were analyzed using M3 vision software (BioSpace Lab, France) and expressed as the average radiance (ph s⁻¹ cm⁻² sr⁻¹). Regions of interest (ROIs) were drawn around the subcutaneous tumors on both flanks and surrounding tissues to determine the average radioactivity. The signal was normalized relative to control mice to assess the degree of signal enhancement relative to the background (corresponding biological tissue). Immediately after 5 min of injection, fluorescence signals were observed, highest in the tumor, followed by the liver, kidneys, and spleen (Fig. 7C). A detailed quantitative analysis revealed that fluorescence intensity in the tumor progressively increased, culminating at 3 h post-injection with a peak radiance of 63.52 × 10⁶ ph s⁻¹ cm⁻² sr⁻¹ (Fig. S9). At this time point, tumor accumulation was markedly higher than that in the spleen (27.37 × 10⁶ ph s⁻¹ cm⁻² sr⁻¹) and coincided with high signal in the liver (57.8 × 10⁶ ph s⁻¹ cm⁻² sr⁻¹), a primary organ for nanoparticle clearance. To further validate targeting specificity, the tumor-to-background (T/B) ratio was calculated, reaching a robust peak of 13.2 at the 3-hour mark (Fig. S10). This high T/B ratio not only confirms the efficient passive accumulation of the nanotheranostic, consistent with our MRI quantification, but also defines an optimal window for bioimaging.

Following the signal intensity in the tumor, a quite strong signal was recorded for the first 5 h after injection in the liver, where NPs were expected to accumulate for clearance by the macrophages and Kupffer cells. The fluorescence signal in the spleen was very low, not changing significantly over the 24 h period. After 24 h, the fluorescence signal in the tumor was low, significantly reduced (5.6 times compared to 3 h). The highest fluorescence intensity was observed in the kidneys of the mouse, followed by the liver and spleen. Moreover, it should be noticed that the kidneys also display high fluorescence imaging (FLI) signals showing that the glomerular elimination pathway is the other way of the imaging probe clearance. It is worth noticing that major clearance (80%) is registered after 72 h of injection. This is a remarkable behavior of these theranostic NPs in providing biosafety. In another study, Shuai et al. conjugated Cyanine 5.5 with Paclitaxel (PTX) and mPEG-PLA coating to monitor the distribution of PTX in a Bel-7402 tumor-bearing mouse model.⁷⁹ The results showed that Cyanine 5.5 covalently conjugated with the PTX drug gave better fluorescence imaging simulating the in vivo distribution of the NPs in the body than the physically combined mixture of Cyanine 5.5 and PTX.

Additionally, the tumor/liver signal ratio was determined as an indicator of tumor-specific nanoparticle distribution relative to the RES. Fig. 7D shows that the tumor/liver signal ratio increased (≥1) and reached its peak 30 minutes after injection (= 2.1). This showed that the NPs were concentrated at the dominant tumor and demonstrated the passive targeting effect of the NPs to the tumor site. Further analysis of the liver/kidney signal ratio allowed assessment of the role of each organ in the metabolism and elimination of NPs. During the first 3 hours, the liver/kidney signal ratio was always >1, indicating that hepatic metabolism was dominant. At 6 and 24 hours post injection, the liver/kidney signal ratio was <1.0, and this value was <0.5 at 24 hours post injection, indicating that the kidney signal was dominant.

After 36 hours, the tumor and organs such as the lungs, spleen, liver, and kidneys were dissected, and ex vivo fluorescence imaging was performed to quantify the fluorescence signal in these organs (Fig. 8A and B). Fig. 8C shows that the highest fluorescence signal was observed in the kidneys, followed by the liver and tumor. Meanwhile, no significant fluorescence signal was observed in the spleen and lungs. The tumor/liver signal ratio <0.5 and liver/kidney <1.0 showed that after 36 hours of injection, the nanosystem was mainly distributed in the kidney (Fig. 8D). This also suggests that after 36 hours the NPs have undergone elimination, concentrating heavily in the kidneys. The high fluorescence signal in the kidneys may also be the result of some free Cyanine 5.5 molecules dissociating from the NPs. It has been demonstrated that nanomaterials can be distributed in the liver, spleen, kidneys, and central nervous system after retro-orbital injection.⁸⁰ Furthermore, the size of NPs affects cellular uptake and elimination from the body. Spherical NPs with diameters greater than 150 nm are absorbed by the spleen and liver, while particles smaller than 100 nm are selectively filtered and eliminated from the body by the kidneys.^81,82

Fig. 8 (A and B) Fluorescence and light imaging of major organs in mice after 24 h of injection; (C) average fluorescence intensity of each organ in (A); and (D) the tumor : liver and liver : kidney ratios of ex vivo FLI intensity (n = 5).

Ex vivo quantification of iron after retro-orbital injection

After 24 hours of retro-orbital injection, the in vivo fluorescence signal in optical imaging was very low and difficult to observe. Therefore, to further evaluate the distribution of NPs at longer times, we determined the iron content by inductively coupled plasma mass spectrometry (ICP-MS) after 24, 48, and 72 hours of retro-orbital injection in the tumor and other organs of mice, such as the liver, kidneys, spleen, lungs, and heart (Fig. 9A).

Fig. 9 (A) Overall timetable for animal experiments ex vivo quantification of iron after retro-orbital injection (dosage: 15 mg kg⁻¹); (B) the bioaccumulation of Fe in the tumor and other organs was quantified by ICP-MS. Data are expressed as mean ± SD (n = 7–8 mice per group). Statistical significance relative to the control group was determined using one-way ANOVA with Dunnett's post-hoc test; *p < 0.05, **p < 0.01, ***p < 0.001, p > 0.05 no significance with respect to the control group.

After 24 hours, the highest iron content was found in the kidneys (209.07 ± 9.98 mg kg⁻¹), which was significantly different (p < 0.001) from the endogenous Fe content in the control mice (53.04 ± 1.14 mg kg⁻¹). This was followed by the liver (177.33 ± 11.92 mg kg⁻¹), tumor (147.03 ± 13.21 mg kg⁻¹), and spleen (134.14 ± 18.28 mg kg⁻¹). After 48 hours, the Fe content was mainly concentrated in the liver (198.56 ± 16.32 mg kg⁻¹) and spleen (195.34 ± 16.32 mg kg⁻¹). After 72 hours, the highest Fe content was in the spleen (277.43 ± 14.86 mg kg⁻¹), followed by the liver, kidneys, and tumor. However, at all three time points, the Fe content in the lungs and heart was very low (<50 mg kg⁻¹), with no difference compared to the endogenous Fe content in the control group (Fig. 9B). Thus, from 24 hours after injection, the Fe content in the tumor gradually decreased. Meanwhile, the Fe content in the liver and spleen gradually increased, showing that at this stage, the liver and spleen play a leading role in the metabolism and excretion of NPs. NPs can be taken up by hepatocytes or Kupffer cells, which are specialized macrophages in the liver, after entering the liver via the blood circulation.^83,84 Additionally, Fe₃O₄ in the core of the nanosystem can be metabolized in the liver by lysosomal enzymes to release iron ions (like ferritin) and reused in the bone marrow to generate new red blood cells.^85,86 Other studies have also shown that although the liver and spleen are the main sites of nanoparticle uptake, when high doses of nanoparticles are injected, their excess will accumulate in other tissues and organs such as the lungs, heart, or adipose tissue.⁸⁷ Several studies have shown that Fe₃O₄ nanoparticles have low toxicity, causing only short-term, transient changes, such as oxidative stress, but do not contribute to permanent damage to organs. Only very high doses (500 mg Fe per kg) cause pathological changes in the liver and spleen.⁸⁸ Based on ICP-MS results, together with the in vivo and ex vivo fluorescence distribution imaging results above, the passive tumor targeting effect and the distribution of NPs in the organs of CT26 tumor-bearing mice were confirmed once again. The liver and spleen play a role in the metabolism and excretion of this nanotheranostic at 72 h postinjection. This 72-hour observation period was strategically chosen to capture the critical transition from systemic circulation to RES-mediated sequestration, which is the gateway to the well-documented, long-term biodegradation pathway for IONPs.^87,89 While comprehensive, long-term toxicological studies extending over several weeks are an indispensable next step for future preclinical development, our short-term data, showing no signs of acute toxicity and aligning with established metabolic principles, provide a robust and sufficient basis for the proof of concept demonstrated in this work.

In vivo anti-tumor efficiency by bioluminescence

In this experiment, luciferase bioluminescence imaging was used to evaluate the tumor response to the chemotherapy effect of NPs (Fig. 10A). Since the luciferase gene was inserted into the CT26 tumor marker gene, they would always be expressed together. Therefore, after luciferin injection, the bioluminescence signal was only generated in the cancer cells at the tumor site due to the activity of luciferase (Fig. 10B). Compared with optical bioimaging, in vivo bioluminescence imaging has the distinct advantage that the background signals due to tissue autofluorescence are almost zero since no external excitation is required in bioluminescence imaging.⁹⁰ In addition, bioluminescence imaging without irradiation with excitation light results in the absence of photobleaching and photocytotoxicity.⁹¹ Bioluminescence signal intensity allows the assessment of cancer cell viability and tumor growth. In mice injected with PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs, after 72 hours the fluorescence signal decreased 7.98 times compared to the PLA-TPGS-Fe₃O₄-Cyanine 5.5 NP group and decreased 9.5 times compared to the control group; the difference was statistically significant, p < 0.001 (Fig. 10C). This result shows the inactivation of luciferase and the tumor inhibitory effect of the nanotheranostic loading Doxorubicin (PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5). In contrast, the bioluminescence intensity in the group of mice injected with PLA-TPGS-Fe₃O₄-Cyanine 5.5 NPs decreased insignificantly (1.19 times) compared with the control group injected with PBS solution (the difference was not statistically significant, p > 0.05). This clearly demonstrated that the chemotherapeutic effect of the PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs effectively inhibited tumor growth, with the effect clearly visible after 72 hours of retro-orbital injection.

Fig. 10 (A) Schematic diagram of the experimental treatment efficacy; (B) tumor growth was monitored by bioluminescence imaging during treatment once every day; (C) the normalized bioluminescence flux in the metastatic tumor region in the mice. The Luciferase signal was normalized with the signal before treatment (n = 6). Non-Gaussian distribution Mann–Whitney test: *p < 0.05; **p < 0.01; ***p < 0.001; nonsignificant p > 0.05; (D) body weight profiles of CT26 tumor-bearing mice across 3 days after various treatments; (E) tumor growth curves of CT26 tumor-bearing mice during treatment once every day. Statistical comparisons of tumor volumes between groups were performed at the final time point using the non-Gaussian distribution Mann–Whitney test. No statistically significant differences were observed (p > 0.05).

During the 72-hour post-treatment period, the body weight (Fig. 10D) and tumor volume (Fig. 10E) values of mice treated with PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs were not significantly reduced (the difference was not statistically significant, p > 0.05). However, by quantifying the biofluorescence signal through the bioluminescence technique, after 72 hours of injection, a clear difference was shown between the group treated with PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs and the group treated with PLA-TPGS-Fe₃O₄-Cyanine 5.5 NPs and the control group (statistically significant difference, p < 0.001). These results demonstrated the chemopreventive effect of PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs and the advantage of the bioluminescence technique in evaluating the efficacy of therapeutics. In another study, Tamarov et al. evaluated the anti-tumor efficacy of anti-PD-L1 peptide–glycol chitosan-conjugated nanomaterials loading Doxorubicin in a CT26 tumor-induced mouse model. The results showed that this nanomaterial effectively inhibited the growth of metastatic tumors for 15 days, in which the bioluminescence intensity was significantly reduced to 0.5–0.52% compared to the control group on day 5.⁹² By recording the bioluminescence intensity and local distribution sequentially during the experiment, tumor progression can be monitored, thereby determining the therapeutic efficacy of NPs. Because bioluminescence is energy (ATP) dependent, bioluminescence imaging is quantitative, which can be related to important biological processes in experimental tumors, including growth, invasion, and metastasis.⁹³ Thus, this technique provides an effective approach for evaluating the chemotherapeutic effects of test materials and monitoring tumor progression in in vivo models.

In vivo magnetic hyperthermia and chemotherapy

The in vivo magnetic hyperthermia and chemotherapy effects were evaluated in a 7-day-old CT26 tumor by injecting the nanotheranostic directly into the tumor at a dose of 150 μL per mouse, equivalent to 15 mg kg⁻¹ (Fig. 11A). Thirty-five mice were divided into 7 groups; a scheme of the experimental design can be seen in Fig. 11B. Two control groups with no NPs and no AMF were used: negative control with PBS (i) and positive control with free Dox (ii). A third group of mice helped in the analysis of the tumor response to an individually magnetic hyperthermia therapy corresponding to a dose of PLA-TPGS-Fe₃O₄ NPs with AMF exposure (iii). Another two groups of mice were injected with PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs, one without AMF exposure (iv) while the other group was exposed to the AMF (v). The last two groups of mice were injected with PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs, one without exposure to the AMF (vi), and the other was exposed to the AMF (vii) to test the efficacy of the NPs due to the heat generation per se. The magnetic hyperthermia treatment was applied to those selected groups for 30 minutes on the day of the injection and two more consecutive days. For this treatment, a magnetic field with an alternating frequency of 450 kHz and an intensity of 150 Oe was utilized. It is critical to note that under these exact conditions, our nanotheranostic exhibits a potent SAR of 183.8 ± 6.8 W g⁻¹ (as characterized in Fig. 4B). This value confirms the system's capacity to generate sufficient localized heat, providing a strong physical rationale for the observed hyperthermia-induced therapeutic efficacy and its synergy with chemotherapy.

Fig. 11 (A) Scheme of the in vivo magnetic hyperthermia tests in the murine model; (B) experimental design of the in vivo magnetic hyperthermia tests; (C) tumor size measurement in a mouse; (D) mice weight monitoring; (E) tumor volume in all groups after treatment; (F) mechanism illustration of the synergistic effects of magnetic hyperthermia and chemotherapy. All data are expressed as mean ± SD (n = 5 mice per group; non-Gaussian distribution Mann–Whitney test: *p < 0.05; **p < 0.01; ***p < 0.001, p > 0.05 no significance with respect to the control group).

Magnetotherapy was used in four applications, which were performed in a magnetic field with an alternating frequency of 450 kHz and an intensity of 150 Oe with an interval of 36 h between two consecutive treatments (Fig. 11C). The local temperature at the tumor site was monitored in real time and was confirmed to reach the therapeutic window of 42.8–43.5 °C during AMF application (Fig. S8). During the treatment period (12 days), the change in the weight of the mice after each treatment in the groups was negligible (Fig. 11D). The treatment effect was reflected through the tumor size before and after treatment (Fig. 11E). In the negative control group, the tumor was very large and grew rapidly. At the initial time (day 7 after tumor induction), the average tumor volume was 345 ± 146 mm³. After 12 days, the tumor had grown very large, increasing 21.4 times, reaching a size of 7377 ± 623 mm³. In the positive control group injected with free Doxorubicin and the (iii) group injected with PLA-TPGS-Fe₃O₄ + magnetic field irradiation, tumor size increased half as slowly as in the negative control group. The tumor volumes in these two groups were almost the same, reaching 4060 ± 690 mm³ and 4827 ± 882 mm³, respectively, a 12-fold increase compared to the initial time. The tumor sizes in these two groups of mice were 50% smaller than those in the negative control group; the difference was statistically significant (p < 0.05). This confirms the chemotherapeutic effect of Doxorubicin and the magnetic hyperthermia effect of the Fe₃O₄ core nanotheranostic coated with the PLA-TPGS copolymer, inhibiting tumor growth.

In the (v) group treated with PLA-TPGS-Fe₃O₄-Cyanine 5.5 and subjected to magnetic field irradiation, the tumor size was similar to that of the (iii) group; the tumor volume reached 4884 ± 695 mm³. In the (iv) group, which was also injected with PLA-TPGS-Fe₃O₄-Cyanine 5.5 but not subjected to magnetic field irradiation, the tumor volume reached 6027 ± 685 mm³ after 12 days, which is 1.23 times higher than when irradiated with the magnetic field; the difference was statistically significant (p < 0.05). This clearly demonstrates the thermogenic role of Fe₃O₄ NPs in the nanotheranostic, and the attachment of the near-infrared fluorescent agent Cyanine 5.5 did not influence the magnetic hyperthermia effect of Fe₃O₄ NPs. The magnetic hyperthermia effect of Fe₃O₄ nanoparticles in cancer treatment not only damages tumor cells, but also increases the expression of tumor antigens, activates dendritic cells and natural killer cells, and transports leukocytes through the vascular endothelium. Moreover, magnetic hyperthermia also changes the stability of DNA, protein structure, and the production of free radicals in cancer cells. The effects of magnetic hyperthermia on these processes can make tumors more sensitive to chemotherapy, leading to improved treatment efficacy.⁹⁴ In addition, cancer cells absorb 8–400 times more Fe₃O₄ nanoparticles than normal cells, so cancer cells are more susceptible to thermomagnetic destruction. Side effects are limited because normal cells are less sensitive to heat than cancer cells, which are characterized by hypoxia and lower pH.⁹⁵ Simultaneously, the effectiveness of thermotherapy on tumors is affected by a number of factors, such as the dispersion of nanomaterials throughout the tumor (which affects the heat generated when irradiated by an external magnetic field), the metabolic process at the tumor site, blood perfusion rate, heat transfer coefficient in cancerous tissues, blood density, etc.⁹⁶

Magnetic hyperthermia significantly impacts tumor growth compared to control untreated pools and, interestingly, free Dox. After 4 treatments, the tumor size in the (vii) group (PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 injection + magnetic field irradiation) significantly differed from that in the control group. The tumor volume reached 3077 ± 280 mm³, 58.3% lower than the negative control group (p < 0.001), and significantly different from the other groups (p < 0.05). The therapeutic efficacy also supports that the drug release takes place from the inside of the PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs only when AMF was applied; otherwise the Dox-loaded sample counterpart (without AMF application) must have the same therapeutic effect. Remarkably, this point supports the synergistic effect when combining chemotherapy (Dox) with magnetic hyperthermia therapy (Fe₃O₄ + magnetic field irradiation) in the multifunctional nanotheranostic attached to Cyanine 5.5. The first possible reason for such a synergistic effect is that the magnetic hyperthermia therapy increases the tumor temperature (40–43 °C), which accelerates the drug release from the nanoparticles, leading to higher Doxorubicin concentrations in the cancer cells (Fig. 11F). Another reason is that when tumor temperature increases, it increases blood flow to the tumor, leading to increased drug delivery and high oxygen concentrations. In vivo experiments have shown that when tumor temperature increases from 39 °C to 43 °C, acute oxygenation and long-term reoxygenation occur at the tumor site, increasing the tumor's response to chemotherapy.⁹⁷ In the study by Piehler et al., Fe₃O₄ nanoparticles coated with dimercaptosuccinic acid and loaded with doxorubicin showed synergistic effects of doxorubicin and thermotherapy in reducing detectable tumor volume up to day 17 post-treatment.⁹⁸ On the other hand, magnetic hyperthermia also induced tumor cell apoptosis.⁹⁹ Finally, we can conclude that the proposed treatment with PLA-TPGS-Fe₃O₄-Dox-Cyanine 5.5 NPs could be a very useful tool to combat accessible tumors. To study the growth inhibition effect at longer times, an animal model with a slower tumor development can be used to enhance the final effect by several more AMF applications. Regarding the administration route, retro-orbital injection was strategically chosen as a standard and highly efficient method for systemic delivery, ensuring rapid entry of the nanotheranostics into general circulation.^32,34 Following all injections, animals were closely monitored by trained personnel. This monitoring revealed no signs of acute local toxicity, such as inflammation, edema, or ocular discharge, thereby suggesting good local tolerance. While detailed ophthalmological examinations like slit-lamp analysis were beyond the scope of this efficacy-focused study, we acknowledge their importance. A thorough evaluation of local safety and tolerance at the administration site will be an essential component of future comprehensive, IND-enabling toxicology studies.

Experimental

Cyanine 5.5 NHS ester (C₄₄H₄₆N₃BF₄O₄) was provided by Lumiprobe GmbH (Europe). Lactide (3,6-dimethyl-1,4-dioxane-2,5-dione, C₆H₈O₄) and vitamin E TPGS (D-α-tocopheryl polyethylene glycol 1000 succinate, C₃₃O₅H₅₄ (CH₂CH₂O)₂₃) were purchased from Sigma Aldrich. The used lactide was recrystallized twice from ethyl acetate before the reaction. Vitamin E TPGS was freeze-dried for two days before use. Stannous octoate (Sn(OOCC₇H₁₅)₂) was purchased from Sigma and was used as 1% distilled toluene solution. Ferric chloride hexahydrate (FeCl₃·6H₂O), ferrous chloride tetra-hydrate (FeCl₂·4H₂O), NaOH, HCl, NH₄OH (26% of ammonia), N-(3-dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC), N-hydroxysuccinimide (NHS), Dulbecco's modifed Eagle's medium (DMEM, purity, 99%), fetal bovine serum (FBS, purity, 99%), and doxorubicin (Dox) hydrochloride were also purchased from Sigma Aldrich. Moreover, 3-aminopropyl triethoxysilane (APTES), tetrahydrofuran (THF), dichloromethane (DCM), triethylamine and methanol were obtained from Merck. All chemicals were used without further purification. Distilled water was used for all experiments. The FESEM and TEM images were obtained from a Hitachi S-4800 system and a JEOL JEM-1010 equipment. Particle sizes were measured using ImageJ software and the resulting histograms were constructed using the OriginPro 8.5 (Origin-Lab, Northampton, MA, USA) software package. The hydrodynamic size and zeta potential of the nanotheranostics in different solutions (pure water, PBS, DMEM, PBS + 100% FBS, DMEM + 100% FBS) were measured by dynamic light scattering using a Malvern Nano-ZS ZEN 3600 zeta sizer instrument. The measurement was performed in triplicate. Fluorescence emission spectra were acquired using an iHR550 (Horiba) spectrometer equipped with a thermoelectrically cooled Si-CCD camera (Synapse) and a 355 nm diode laser as an excitation source. X-ray diffraction (XRD) was used to identify the crystal structure of materials (SIEMENS-D5000). The static hysteresis loops of dry samples were measured at room temperature by vibrating sample magnetometry (VSM). X-ray photoelectron spectroscopy (XPS) was performed using a Thermo VG Multilab 2000 spectrometer. The chemical synthesis, in vivo MRI, bioluminescence, and NIR bioimaging as well as magnetic hyperthermia in murine models are detailed in the SI.

Conclusions

In summary, we have successfully developed high fluorescence-intensity NPs incorporating Cyanine 5.5, Fe₃O₄, and Doxorubicin with a dual effect of therapeutic and MRI/NIR bioimaging functions. The innovation lies not merely in the composition of the nanoparticles but also in the comprehensive quantitative framework we employed to demonstrate its efficacy. Our results confirm that retro-orbital administration leads to effective passive tumor accumulation, which we quantified with high precision using both MRI (%I_0.25 analysis) and NIR optical bioimaging. We established a favorable biodistribution and clearance profile, a critical step for biosafety and clinical translation. Furthermore, by using bioluminescence imaging, we monitored the therapeutic response at the cellular level, demonstrating significant tumor inhibition within 72 hours. The meticulously designed in vivo study unequivocally proved the powerful synergistic effect of combining AMF-induced hyperthermia with chemotherapy, resulting in superior tumor growth suppression. This holistic approach, combining optimized synthesis with quantitative, multi-modal validation of targeting, efficacy, and clearance, provides a robust blueprint for the preclinical development of next-generation nanotheranostics.

Author contributions

Ke Son Phan: writing – original draft, methodology, investigation, formal analysis, data curation, and conceptualization. Bich Thuy Doan and Yiqian Wang: methodology and investigation. Thi Thu Huong Le, and Thi Thu Trang Mai: writing – original draft, methodology, investigation, formal analysis, and data curation. Ha Bao Hung Bui: methodology and investigation. Hong Nam Pham: methodology, formal analysis, and data curation. Le Hang Dang: software and methodology. Ngoc Quyen Tran: validation, supervision, project administration, and funding acquisition. Phuong Thu Ha: writing – review & editing, visualization, validation, supervision, resources, project administration, and funding acquisition.

Conflicts of interest

There are no conflicts to declare.

Ethical statement

All animal procedures were performed in accordance with the Guidelines for Care and Use of Laboratory Animals of the University of Paris, referral CEEA34.JS.142.1 and approved by the Animal Ethics Committee of the University.

Data availability

The data that support the findings of this study are available on request from the corresponding author. Supplementary information is available, containing detailed experimental procedures and additional physicochemical and biological characterization data (Figs. S1–S10). See DOI: https://doi.org/10.1039/d5bm00659g.

Acknowledgements

This work was funded by the Vietnam Academy of Science and Technology (VAST) under Grant Number NCXS.01.01/23-25.

In addition, we also thank Dr J. Seguin for the CT26 biological model supply. We are also grateful to the Plateformes d'imagerie du Vivant PIV, in particular the LIOPA where the in vivo imaging experiments were performed. This project has also received support from the CNRS Centre National de la Recherche Scientifique (France) through the MITI interdisciplinary programs. We are thankful to the CSC program for PhD support (YW) as well as Campus France.

References

1. F. Bray, M. Laversanne, H. Sung, J. Ferlay, R. L. Siegel, I. Soerjomataram and A. Jemal, CA-Cancer J. Clin., 2024, 74, 229–263.
2. F. Bray, M. Laversanne, H. Sung, J. Ferlay, R. L. Siegel, I. Soerjomataram and A. Jemal, CA-Cancer J. Clin., 2024, 74, 229–263.
3. N. Keum and E. Giovannucci, Nat. Rev. Gastroenterol. Hepatol., 2019, 16, 713–732.
4. A. M. D. Wolf, E. T. H. Fontham, T. R. Church, C. R. Flowers, C. E. Guerra, S. J. LaMonte, R. Etzioni, M. T. McKenna, K. C. Oeffinger, Y.-C. T. Shih, L. C. Walter, K. S. Andrews, O. W. Brawley, D. Brooks, S. A. Fedewa, D. Manassaram-Baptiste, R. L. Siegel, R. C. Wender and R. A. Smith, CA-Cancer J. Clin., 2018, 68, 250–281.
5. S. R. Brown, N. S. Fearnhead, O. D. Faiz, J. F. Abercrombie, A. G. Acheson, R. G. Arnott, S. K. Clark, S. Clifford, R. J. Davies, M. M. Davies, W. J. P. Douie, M. G. Dunlop, J. C. Epstein, M. D. Evans, B. D. George, R. J. Guy, R. Hargest, A. B. Hawthorne, J. Hill, G. W. Hughes, J. K. Limdi, C. A. Maxwell-Armstrong, P. R. O'Connell, T. D. Pinkney, J. Pipe, P. M. Sagar, B. Singh, M. Soop, H. Terry, J. Torkington, A. Verjee, C. J. Walsh, J. H. Warusavitarne, A. B. Williams, G. L. Williams, R. G. Wilson and ACPGBI IBD Surgery Consensus Collaboration, Colorectal Dis., 2018, 20, 3–117.
6. D. K. Rex, R. C. Boland, J. A. Dominitz, F. M. Giardiello, D. A. Johnson, T. Kaltenbach, T. R. Levin, D. Lieberman and D. J. Robertson, Am. J. Gastroenterol., 2017, 112, 1016–1030.
7. G. P. Young, S. C. Benton, R. S. Bresalier, H.-M. Chiu, E. Dekker, C. G. Fraser, M. A. M. Frasa, S. P. Halloran, M. Hoffmeister, S. Parry, K. Selby, C. Senore, H. Singh and E. L. Symonds, Dig. Dis. Sci., 2025, 70, 1637–1645.
8. R. Gómez-Molina, M. Suárez, R. Martínez, M. Chilet, J. M. Bauça and J. Mateo, Healthcare, 2024, 12, 1645.
9. S. S. Beniwal, P. Lamo, A. Kaushik, D. L. Lorenzo-Villegas, Y. Liu and A. MohanaSundaram, Biosensors, 2023, 13, 926.
10. J. Niedzielska and T. Jastrzębski, J. Clin. Med., 2025, 14, 3189.
11. L. Ma, H. Guo, Y. Zhao, Z. Liu, C. Wang, J. Bu, T. Sun and J. Wei, Signal Transduction Targeted Ther., 2024, 9, 336.
12. S. Li, Z. Chen, L. Tan, Q. Wu, X. Ren, C. Fu, M. Niu, H. Li and X. Meng, Biomaterials, 2022, 283, 121472.
13. B. K. Kashyap, V. V. Singh, M. K. Solanki, A. Kumar, J. Ruokolainen and K. K. Kesari, ACS Omega, 2023, 8, 14290–14320.
14. P.-Y. Wu, Z.-C. Shen, J.-L. Jiang, B.-C. Zhang, W.-Z. Zhang, J.-J. Zou, J.-F. Lin, C. Li and J.-W. Shao, Biomater. Sci., 2022, 10, 6267–6281.
15. A. Pulumati, A. Pulumati, B. S. Dwarakanath, A. Verma and R. V. L. Papineni, Cancer Rep., 2023, 6, e1764.
16. J.-W. Bai, S.-Q. Qiu and G.-J. Zhang, Signal Transduction Targeted Ther., 2023, 8, 1–32.
17. M. Dunne, M. Regenold and C. Allen, Adv. Drug Delivery Rev., 2020, 163–164, 98–124.
18. M. J. Ko, S. Min, H. Hong, W. Yoo, J. Joo, Y. S. Zhang, H. Kang and D.-H. Kim, Bioact. Mater., 2024, 32, 66–97.
19. H. Zhu, J. Tao, W. Wang, Y. Zhou, P. Li, Z. Li, K. Yan, S. Wu, K. W. K. Yeung, Z. Xu, H. Xu and P. K. Chu, Biomaterials, 2013, 34, 2296–2306.
20. K. S. Phan, B. T. Doan, T. T. H. Le, T. T. T. Mai, T. D. T. Ung, T. Q. Bui, S. Boumati and P. T. Ha, ChemistrySelect, 2023, 8, e202303580.
21. P. T. Ha, T. T. H. Le, T. D. T. Ung, H. D. Do, B. T. Doan, T. T. T. Mai, H. N. Pham, T. M. N. Hoang, K. S. Phan and T. Q. Bui, New J. Chem., 2020, 44, 17277–17288.
22. K. C. On, J. Rho, H. Y. Yoon, H. Chang, J. Y. Yhee, J. S. Yoon, S. Y. Jeong, H. K. Kim and K. Kim, Pharmaceutics, 2020, 12, 621.
23. S. H. Kim, J. H. Park, J. S. Kwon, J. G. Cho, K. G. Park, C. H. Park, J. J. Yoo, A. Atala, H. S. Choi, M. S. Kim and S. J. Lee, Biomaterials, 2020, 258, 120267.
24. C.-Y. Chen, C.-B. Jian, H.-D. Gao, X.-E. Yu, Y.-C. Chang, S. K. Leong, J.-J. Shie and H.-M. Lee, Sens. Diagn., 2024, 3, 1028–1038.
25. Y. Ji, C. Jones, Y. Baek, G. K. Park, S. Kashiwagi and H. S. Choi, Adv. Drug Delivery Rev., 2020, 167, 121–134.
26. Y. Tang and G. Wang, J. Photochem. Photobiol., C, 2021, 47, 100420.
27. K. S. Phan, H. D. Do, B. T. Doan, T. T. H. Le, T. T. T. Mai, Q. B. N. Nguyen, T. N. Dong, H. B. H. Bui, V. D. Dang, L. H. Dang, N. Q. Tran and P. T. Ha, ChemMedChem, 2025, 20, e202400586.
28. K. S. Phan, P. T. Ha, H. N. Do, T. A. Nguyen, T. Q. Bui, H. N. Pham, M. H. Le and T. T. H. Le, ChemMedChem, 2021, 16, 3615–3625.
29. J. Yu, L. Yang, X. Liang, T. Dong and H. Liu, Analyst, 2015, 140, 4114–4120.
30. L. Cursi, S. Vercellino, M. M. McCafferty, E. Sheridan, V. Petseva, L. Adumeau and K. A. Dawson, Nanoscale, 2021, 13, 16324–16338.
31. A. Schoch, I. S. Thorey, J. Engert, G. Winter and T. Emrich, Lab Anim., 2014, 43, 95–99.
32. T. Yardeni, M. Eckhaus, H. D. Morris, M. Huizing and S. Hoogstraten-Miller, Lab Anim., 2011, 40, 155–160.
33. C. Nanni, C. Pettinato, V. Ambrosini, A. Spinelli, S. Trespidi, D. Rubello, A. Al-Nahhas, R. Franchi, A. Alavi and S. Fanti, Nucl. Med. Commun., 2007, 28, 547–553.
34. J. Li, Y. Wu, B. Liu, Y. Huang, Q. Wu, H. Li, S. Xiao, Y. Lin and T. Li, BMC Ophthalmol., 2021, 21, 208.
35. F. Rizvi, E. Everton, A. R. Smith, H. Liu, E. Osota, M. Beattie, Y. Tam, N. Pardi, D. Weissman and V. Gouon-Evans, Nat. Commun., 2021, 12, 613.
36. A. A. Issa, M. S. Elazazy and A. S. Luyt, Int. J. Chem. Kinet., 2018, 50, 846–855.
37. A. A. Issa and A. S. Luyt, Polymers, 2019, 11, 537.
38. M. Sypabekova, A. Hagemann, D. Rho and S. Kim, Biosensors, 2023, 13, 36.
39. S. Yang, P. Yuan, H. He, Z. Qin, Q. Zhou, J. Zhu and D. Liu, Appl. Clay Sci., 2012, 62–63, 8–14.
40. S. J. Dumas, E. Meta, M. Borri, Y. Luo, X. Li, T. J. Rabelink and P. Carmeliet, Nat. Rev. Nephrol., 2021, 17, 441–464.
41. M. Izci, C. Maksoudian, B. B. Manshian and S. J. Soenen, Chem. Rev., 2021, 121, 1746–1803.
42. M. Chehelgerdi, M. Chehelgerdi, O. Q. B. Allela, R. D. C. Pecho, N. Jayasankar, D. P. Rao, T. Thamaraikani, M. Vasanthan, P. Viktor, N. Lakshmaiya, M. J. Saadh, A. Amajd, M. A. Abo-Zaid, R. Y. Castillo-Acobo, A. H. Ismail, A. H. Amin and R. Akhavan-Sigari, Mol. Cancer, 2023, 22, 169.
43. P. Torkaman, R. Karimzadeh and A. Jafari, Sci. Rep., 2023, 13, 18151.
44. J. Mu, B. Chen, Z. Guo, M. Zhang, Z. Zhang, P. Zhang, C. Shao and Y. Liu,  10.1039/C1NR10972C.
45. H. Simonyan, C. Hurr and C. N. Young, Physiol. Genomics, 2016, 48, 762–770.
46. K. Rekha, R. Ezhil Vizhi, B. B. Lahiri and J. Philip, Inorg. Chem. Commun., 2024, 165, 112466.
47. L. Rodríguez-Pérez, C. Villegas, M. Á. Herranz, J. L. Delgado and N. Martín, ACS Omega, 2017, 2, 9164–9170.
48. Z. Zhang and S.-S. Feng, Biomaterials, 2006, 27, 262–270.
49. C. Pucci, A. Degl'Innocenti, M. Belenli Gümüş and G. Ciofani, Biomater. Sci., 2022, 10, 2103–2121.
50. S. Alzola-Aldamizetxebarria, L. Fernández-Méndez, D. Padro, J. Ruíz-Cabello and P. Ramos-Cabrer, ACS Omega, 2022, 7, 36905–36917.
51. G. Prabha and V. Raj, Mater. Sci. Eng., C, 2017, 79, 410–422.
52. R. Al-Obaidy, A. J. Haider, S. Al-Musawi and N. Arsad, Sci. Rep., 2023, 13, 3180.
53. G. Wang, D. Zhao, N. Li, X. Wang and Y. Ma, J. Magn. Magn. Mater., 2018, 456, 316–323.
54. S. Ranoo, B. Lahiri and J. Philip, J. Magn. Magn. Mater., 2020, 498, 166138.
55. L. H. Nguyen, P. T. Phong, P. H. Nam, D. H. Manh, N. T. K. Thanh, L. D. Tung and N. X. Phuc, Materials, 2021, 14, 1875.
56. A. R. Yasemian, M. Almasi Kashi and A. Ramazani, Mater. Chem. Phys., 2019, 230, 9–16.
57. W. Ji, P. Hu, X. Wang, B. Chen, T. Chang, F. F. Yang, Q. Cao, W. Zhang, R. Dang and K. Wang, J. Alloys Compd., 2021, 866, 158952.
58. P. Lemal, S. Balog, C. Geers, P. Taladriz-Blanco, A. Palumbo, A. M. Hirt, B. Rothen-Rutishauser and A. Petri-Fink, J. Magn. Magn. Mater., 2019, 474, 637–642.
59. G. Hannon, F. L. Tansi, I. Hilger and A. Prina-Mello, Adv. Ther., 2021, 4, 2000267.
60. C. Papadopoulos, E. K. Efthimiadou, M. Pissas, D. Fuentes, N. Boukos, V. Psycharis, G. Kordas, V. C. Loukopoulos and G. C. Kagadis, Phys. Med., 2020, 71, 39–52.
61. R. E. Rosensweig, J. Magn. Magn. Mater., 2002, 252, 370–374.
62. E. Natividad and I. Andreu, in Magnetic Characterization Techniques for Nanomaterials, ed. C. S. S. R. Kumar, Springer Berlin Heidelberg, Berlin, Heidelberg, 2017, pp. 261–303.
63. P. Guardia, R. Di Corato, L. Lartigue, C. Wilhelm, A. Espinosa, M. Garcia-Hernandez, F. Gazeau, L. Manna and T. Pellegrino, ACS Nano, 2012, 6, 3080–3091.
64. A. Radoń, A. Włodarczyk, Ł. Sieroń, M. Rost-Roszkowska, Ł. Chajec, D. Łukowiec, A. Ciuraszkiewicz, P. Gębara, S. Wacławek and A. Kolano-Burian, Sci. Rep., 2023, 13, 7860.
65. A. Rajan, M. Sharma and N. K. Sahu, Sci. Rep., 2020, 10, 15045.
66. H. Xu and Y. Pan, Nanomaterials, 2019, 9, 1457.
67. C. Yin and X. Li, Int. J. Pharm., 2011, 418, 78–87.
68. F. Zamani, F. Jahanmard, F. Ghasemkhah, S. Amjad-Iranagh, R. Bagherzadeh, M. Amani-Tehran and M. Latifi, in Nanostructures for Drug Delivery, ed. E. Andronescu and A. M. Grumezescu, Elsevier, 2017, pp. 239–270.
69. Y. Xu, C. Kim, D. M. Saylor and D. Koo, J. Biomed. Mater. Res., Part B, 2017, 105, 1692–1716.
70. A. Sudame, G. Kandasamy, D. Singh, C. V. Tomy and D. Maity, Colloids Surf., A, 2020, 606, 125355.
71. S.-H. Hu, C.-H. Tsai, C.-F. Liao, D.-M. Liu and S.-Y. Chen, Langmuir, 2008, 24, 11811–11818.
72. S. K. Golombek, J.-N. May, B. Theek, L. Appold, N. Drude, F. Kiessling and T. Lammers, Adv. Drug Delivery Rev., 2018, 130, 17–38.
73. R. Baghban, L. Roshangar, R. Jahanban-Esfahlan, K. Seidi, A. Ebrahimi-Kalan, M. Jaymand, S. Kolahian, T. Javaheri and P. Zare, Cell Commun. Signaling, 2020, 18, 1–19.
74. A. Nomani, S. Yousefi, D. Sargsyan and A. Hatefi, J. Controlled Release, 2024, 368, 728–739.
75. C. J. Thébault, G. Ramniceanu, S. Boumati, A. Michel, J. Seguin, B. Larrat, N. Mignet, C. Ménager and B.-T. Doan, J. Controlled Release, 2020, 322, 137–148.
76. C. J. Thébault, G. Ramniceanu, A. Michel, C. Beauvineau, C. Girard, J. Seguin, N. Mignet, C. Ménager and B.-T. Doan, Mol. Imaging Biol., 2019, 21, 269–278.
77. J. Seguin, C. Nicolazzi, N. Mignet, D. Scherman and G. G. Chabot, Tumor Biol., 2012, 33, 1709–1717.
78. T. Etrych, H. Lucas, O. Janoušková, P. Chytil, T. Mueller and K. Mäder, J. Controlled Release, 2016, 226, 168–181.
79. Q. Shuai, G. Zhao, X. Zhang, B. Yu, R. J. Lee and W.-K. Su, Colloids Surf., B, 2019, 181, 872–878.
80. J. Bourquin, A. Milosevic, D. Hauser, R. Lehner, F. Blank, A. Petri-Fink and B. Rothen-Rutishauser, Adv. Mater., 2018, 30, 1704307.
81. M. J. Mitchell, M. M. Billingsley, R. M. Haley, M. E. Wechsler, N. A. Peppas and R. Langer, Nat. Rev. Drug Discovery, 2021, 20, 101–124.
82. M. Zhang, S. Gao, D. Yang, Y. Fang, X. Lin, X. Jin, Y. Liu, X. Liu, K. Su and K. Shi, Acta Pharm. Sin. B, 2021, 11, 2265–2285.
83. J. Li, C. Chen and T. Xia, Adv. Mater., 2022, 34, 2106456.
84. H. B. Haroon, A. C. Hunter, Z. S. Farhangrazi and S. M. Moghimi, Adv. Drug Delivery Rev., 2022, 188, 114396.
85. N. Sukhbaatar and T. Weichhart, Pharmaceuticals, 2018, 11, 137.
86. A.-C. S. Vogt, T. Arsiwala, M. Mohsen, M. Vogel, V. Manolova and M. F. Bachmann, Int. J. Mol. Sci., 2021, 22, 4591.
87. M. Levy, N. Luciani, D. Alloyeau, D. Elgrabli, V. Deveaux, C. Pechoux, S. Chat, G. Wang, N. Vats and F. Gendron, Biomaterials, 2011, 32, 3988–3999.
88. B. A. Katsnelson, T. D. Degtyareva, I. I. Minigalieva, L. I. Privalova, S. V. Kuzmin, O. S. Yeremenko, E. P. Kireyeva, M. P. Sutunkova, I. I. Valamina and M. Y. Khodos, Int. J. Toxicol., 2011, 30, 59–68.
89. A.-C. S. Vogt, T. Arsiwala, M. Mohsen, M. Vogel, V. Manolova and M. F. Bachmann, Int. J. Mol. Sci., 2021, 22, 4591.
90. C. E. Badr and B. A. Tannous, Trends Biotechnol., 2011, 29, 624–633.
91. S. B. Mali, Oral Oncol. Rep., 2024, 9, 100127.
92. K. Tamarov, J. T.-W. Wang, J. Kari, E. Happonen, I. Vesavaara, M. Niemelä, P. Perämäki, K. T. Al-Jamal, W. Xu and V.-P. Lehto, ACS Appl. Mater. Interfaces, 2021, 13, 40392–40400.
93. S. B. Mali, Oral Oncol. Rep., 2024, 9, 100127.
94. D. Egea-Benavente, J. G. Ovejero, M. del P. Morales and D. F. Barber, Cancers, 2021, 13, 4583.
95. M.-X. Ren, Y.-Q. Wang, B.-Y. Lei, X.-X. Yang, Y.-L. Hou, W.-J. Meng and D.-L. Zhao, Ceram. Int., 2021, 47, 20686–20692.
96. N. Pazouki, S. Irani, N. Olov, S. M. Atyabi and S. Bagheri-Khoulenjani, Prog. Biomater., 2022, 11, 43–54.
97. S. A. Salim, T. A. Salaheldin, M. M. Elmazar, A. F. Abdel-Aziz and E. A. Kamoun, RSC Adv., 2022, 12, 33835–33851.
98. S. Piehler, H. Dähring, J. Grandke, J. Göring, P. Couleaud, A. Aires, A. L. Cortajarena, J. Courty, A. Latorre, Á. Somoza, U. Teichgräber and I. Hilger, Nanomaterials, 2020, 10, 1016.
99. B. Sanz, M. P. Calatayud, T. E. Torres, M. L. Fanarraga, M. R. Ibarra and G. F. Goya, Biomaterials, 2017, 114, 62–70.

---