Читать книгу Magnetic Nanoparticles in Human Health and Medicine - Группа авторов - Страница 33

The current standard of cancer care comprises the elimination of solid tumors by surgery followed by treatment with chemotherapeutic drugs (Bhattacharjee et al. 2010). However, it has been shown that most of anticancer drugs used in chemotherapy also target other healthy tissues in the body, causing major toxicity problems to vital organs. A typical example is the treatment with doxorubicin that provokes major heart problems as side effect (Minotti et al. 2004). Hyperthermia therapy or thermotherapy was considered a potentially useful alternative of chemotherapy. In this case, the body tissues are exposed to higher temperatures in order to damage or kill cancer cells by inducing cell apoptosis (Kerr et al. 1994). The concept of hyperthermia has been introduced in “clinical” practice many centuries ago by Greeks, Egyptians, Romans, and Indians. During the nineteenth century, it was observed that fever can cause tumor regression (Moyer and Delman 2008) and scientific studies were performed to treat cervical cancer by hyperthermia (Baronzio and Hager 2006). Beginning with 1970, the cancer treatment by hyperthermia was taken more seriously and controlled clinical trials started to be conducted. It has been discovered that cancer cells undergo apoptosis at temperatures of 42–45 °C in contrast to healthy cells that are able to withstand those temperatures (Cavaliere et al. 1967).

Depending on the location, depth, and stage of the malignancy, three main types of hyperthermia have been developed for clinical applications: whole body, regional, and local hyperthermia. In the case of deep‐seated and propagated metastasis, whole‐body hyperthermia is used. In this case, the entire body is heated up through hot water baths, electric blankets, hot wax, thermal chambers, or infrared radiators (Chicheł et al. 2007). Heat delivering to advanced stage tumors is realized by regional hyperthermia by means of thermal perfusion or external arrays of applicators (Falk and Issels 2001). The treatment of localized superficial tumors is often carried out by local hyperthermia, in which the heat is applied using electromagnetic waves such as radio waves, microwaves, and ultrasound, generated by applicators placed at the surface or under the skin of superficial cancer. Although in local hyperthermia, there is a better control of the area exposed to heat and a better heat uniformity, canceling the drawbacks of the first two types of hyperthermia, it is mainly focused on small and superficial cancer regions. The small penetration depth of the generated heat (in the order of a few centimeters) hampers its utilization for the cure of deep cancerous regions. However, deep cancer cell therapy can be achieved by replacing the heating sources with MNPs delivered only to the cancer cells. The capability of MNPs, once internalized in cancer cells, to convert electromagnetic energy into thermal energy, upon their exposure to an external alternating magnetic field (AMF), allows to locally raise the temperature of the cancerous region up to a level at which cellular apoptosis can be initiated.

As such, the new noninvasive local hyperthermia method, called magnetic hyperthermia (MH), has become one of the most promising innovative cancer therapies (Piñeiro et al. 2015). It possesses numerous advantages over “traditional” therapies, mostly relying on the physico‐chemical properties of MNPs. First, the MNPs can be potentially injected anywhere in the body, the injection being less invasive and allowing the treatment of all kinds of tumors with limited side effects (Nedelcu 2008). Second, the MNPs can be functionalized with a recognition moiety (i.e. antibodies, proteins) in order to increase the selectivity to malignant cells, therefore, increasing the internalization of the MNPs in a specific type of cancer cells (Cherukuri et al. 2010). Nevertheless, the MNPs can be magnetically targeted toward the cancer region by using an external magnetic field gradient. Once the MNPs have been internalized into the cancer cells, they can remain inside the cells even after their multiplication, meaning that a subsequent hyperthermic treatment can be applied without large reinjection of MNPs (Jordan et al. 1999).

It is known that at room temperature, the macroscopic magnetic materials hold a permanent magnetic dipole moment. All the individual magnetic spins originating from electrons movements are aligned along a particular direction called the easy axis. As a result, the individual magnetic spins can possess two opposite orientations (up and down). These two orientations are separated by a barrier of magnetocrystalline anisotropy energy. When the size of the magnetic materials is reduced to the nanoscale, up to several tens of nanometers, this energetic barrier can be overcome with the aid of the thermal energy. Other ways saying, the magnetic dipole moment of MNPs is not anymore fixed along the easy axis as in the case of macroscopic magnetic materials. This new magnetic property is known as superparamagnetism (SP). In this novel state, the value of the magnetization (which basically represents the sum of all individual magnetic spins) is randomly distributed by the thermal effect. Upon the application of an external AMF, all the individual magnetic spins are flipped, while the magnetization direction is reversed. The provided magnetic energy is released as heat in a phenomenon known as Neel relaxation. The delay between the application of the AMF and the magnetic spins flipping gives rise to a torque that leads to the rotation of MNPs in liquids. The rotational friction between MNPs and the liquid environment also produces heat in a process described as Brownian relaxation.

The capacity of MNPs to generate heat under an external AMF is quantified by the specific absorption rate (SAR) or specific loss power (SLP). This parameter provides a measure of the rate at which energy is absorbed per unit mass of MNPs. The SAR values strongly depend on the magnetic properties of MNPs (saturation magnetization, coercive field, and magnetic anisotropy) which in turn are governed by the structure, size, size distribution, shape, and composition of MNPs (Périgo et al. 2015). On the other hand, the SAR values can theoretically be increased as much as necessary, by increasing the frequency (f) and the amplitude (H) of the applied AMF (Glöckl et al. 2006). From the practical point of view, this approach is limited by difficulties in designing equipment able to generate large f and high H and, more importantly, by the increased harm produced to healthy cells as a consequence of the occurrence of eddy currents in conducting media (Spirou et al. 2018). For clinical applications, several safety conditions in terms of the H × f product were proposed (Mamiya and Jeyadevan 2019). Based on real tests on patients who were exposed to AMF for a duration that exceeds one hour, according to Atkinson–Brezovich criterion, it was largely accepted that the product H × f should be limited to 5 × 10⁸ A m⁻¹ s⁻¹ (Hergt and Dutz 2007). This limit can be increased 10 times if the treatment is applied to small body regions (Hergt and Dutz 2007).

Two classes of SPIONs, magnetite (Fe₃O₄), and maghemite (Fe₂O₃) have been approved for clinical use by FDA for MRI applications. They have been both tested in vivo for clinical MH therapy (Maier‐Hauff et al. 2011; Wilczewska et al. 2012). In Europe, this form of therapy was clinically approved in the case of glioblastoma treatment. A clinical trial was also performed on prostate cancer (Maier‐Hauff et al. 2011). Since the SAR values of spherical SPIONs (having a diameter of ~10 nm) are very low, drastically decreasing when they are localized into cells or tissues as a consequence of the intracellular clustering (Hilger et al. 2005), the SPIONs were not able to deliver sufficient heat to completely destroy the tumors. As such, for a complete elimination of the tumor, the MH has been used in conjunction with other therapies (chemo‐ and/or radiotherapies). In this case, aggressive side effects have been observed.

As a result, the scientific community involved in MH research has focused on the elaboration of biocompatible iron oxide MNPs (IOMNPs) with enhanced magnetic properties and better MH performance. They have to be capable of completely destroying the tumors at doses below their intrinsic toxicity and safety levels of AMF. In order to accomplish this goal, two major scientific strategies emerged. The first strategy consisted in increasing the size of SPIONs, by keeping them in the SP limit, thereby increasing the Ms and consequently their SAR values. The Neel relaxation, which prevails when SPIONs are confined inside cellular endosomal compartments due to the considerable reduction of Brownian relaxation, is governed by the magnetic anisotropy. In the case of MNPs with large surface to volume ratio, surface contributions to magnetic properties become significant. Hence, the magnetic anisotropy of MNPs is dominated by the surface anisotropy, which originates from the spin direction discrepancy between core and surface. As such, it is directly associated with MNPs shape. This is the reason why the second strategy focused on controlling the shape of SPIONs by using different synthesis methods.

The influence of the mean size of SPIONs on SAR values has been the subject of many studies. For instance, Jeun et al. reported an increase of the SAR values from 45 to 322 W g_Fe⁻¹ H × f = 1.3 × 10⁹ A m⁻¹ s⁻¹) when the SPIONs diameter is increased from 4.5 to 22.5 nm (Jeun et al. 2012). Other studies reported maximum SAR values of 447 W g_Fe⁻¹ (H × f = 9.8 × 10⁹ A m⁻¹ s⁻¹) (Gonzales‐Weimuller et al. 2009), 702 W g_Fe⁻¹ (H × f = 6.3 × 10⁹ A m⁻¹ s⁻¹) (Müller et al. 2013) and 950 W g_Fe⁻¹ (H × f = 18.9 × 10⁹ A m⁻¹ s⁻¹) (Lévy et al. 2008) for SPIONs with mean diameter of 14, 15.2 and 17.7 nm, respectively, synthesized using different protocols. Finally, Fortin et al. demonstrated both experimentally and theoretically that the SAR values of SPIONs increase from 4 to 1650 W g_Fe⁻¹ (H × f = 17.5×10⁹ A m⁻¹ s⁻¹) when the mean SPIONs diameter is increased from 5 to 16.5 nm (Fortin et al. 2007). Besides the mean size of SPIONs, their size distribution can have a significant impact on the SAR value. As shown by Gazeau et al., the size selection of SPIONs with 30 nm in diameter increases their heating performance to 600 W g_Fe⁻¹, at clinically relevant conditions (H × f = 4.1 × 10⁹ A m⁻¹ s⁻¹) (Gazeau et al. 2008).

An alternative approach for increasing SPIONs heating rates turned out to be the control of their surface coating. It has been shown that dextran‐coated SPIONs with a diameter of 7 nm display a SAR value of 626 W gFe⁻¹ (H × f = 6.25 × 10⁹ A m⁻¹ s⁻¹) (Mornet et al. 2004). Liu et al. reported a maximum SAR value of 930 W g_Fe⁻¹ (H × f = 10.8×10⁹ A m⁻¹ s⁻¹) when SPIONs with a diameter of 19 nm are coated with a 6 nm shell of phosphorylated methoxy polyethylene glycol 2000 (Liu et al. 2012). Interestingly, the high heating capacity of PEGylated SPIONs is maintained in various physiological conditions. On the other hand, the inorganic coating can also improve the SAR value of SPIONs. For example, Mohammad et al. found that the hyperthermic effect of SPIONs is four‐ to fivefold enhanced (920 W g_Fe⁻¹) on coating with a gold shell of 0.5 thickness (Mohammad et al. 2010). A maximum SAR value of 1300 W g_Fe⁻¹ (H × f = 7.9×10⁹ A m⁻¹ s⁻¹) was measured for dumbbell‐like shaped dimers formed by an iron oxide domain of 24 nm in size and a gold seed of 9 nm in diameter (Guardia et al. 2017).

When the SPIONs size exceeds the SP limit, the thermal energy cannot overcome anymore the barrier of magneto‐crystalline anisotropy, and the MNPs acquire a permanent magnetic dipole moment, pointing in the direction of the easy axis. The MNPs become ferromagnetic at room temperature, developing hysteresis loops, which are characterized by M_r and coercivity (H_c – magnetic field strength required for demagnetization). The MH properties of ferromagnetic IOMNPs (FeMIONs) are now dictated by their dynamic hysteresis behavior (Carrey et al. 2011). The synergistic contribution from the hysteresis and susceptibility (Neel and Brown) loss enhance the SAR values of FeMIONs in comparison with SPIONs. Several research teams have obtained the following maximum SAR values for FeMIONs of different diameters synthesized via thermal decomposition of magnetic precursors in organic solvent: 650 W g_Fe⁻¹ (H × f = 19.1 × 10⁹ A m⁻¹ s⁻¹) for 52 nm FeMIONs (Nemati et al. 2018); 716 W g_Fe⁻¹ (H × f = 7.75 × 10⁹ A m⁻¹ s⁻¹) for 22 nm FeMIONs (Chen et al. 2013); 801 W g_Fe⁻¹ (H × f = 13.2 × 10⁹ A m⁻¹ s⁻¹) for 28 nm FeMIONs (Mohapatra et al. 2018) and 2560 W g_Fe⁻¹ (H × f = 6.7 × 10⁹ A m⁻¹ s⁻¹) for 40 nm FeMIONs (Tong et al. 2017). The increase of FeMIONs sizes can be realized up to a certain threshold value; above it, the FeMIONs might enter into a multimagnetic domains state that will lead to a decrease in their SAR values (Chen et al. 2013; Mohapatra et al. 2018; Nemati et al. 2018). Despite the large amounts of generated heat, FeMIONs are less favorable for biomedical applications due to their colloidal instability and Hc that facilitate their aggregation. The dipole–dipole interactions manifested between FeMIONs significantly influence their heating efficiency (Serantes et al. 2010; Salas et al. 2014; Coral et al. 2016). But when these interactions, coupled with the uniaxial shape anisotropy, will arrange (Toulemon et al. 2016) or align (Jiang et al. 2016) the FeMIONPs into chain‐like superstructures, their Hc and Ms will increase. Consequently, the heating performances will be enhanced as it was shown theoretically (Serantes et al. 2014) and experimentally (Myrovali et al. 2016). Through polyol‐mediated synthesis, the stabilization of the FeMIONs may occur as multicore aggregates: nanoclusters (Sakellari et al. 2016), hollow nanospheres (Gavilán et al. 2017), and nanoflowers (Hugounenq et al. 2012; Gavilán et al. 2017). Due to the collective spin rotation, the SAR value of these aggregates is much greater than of their constituents. For instance, nanoflowers of 50 nm consisting of spherical FeMIONs of 11 nm displayed SAR values of 1790 W g_Fe⁻¹ (H × f = 7.7 × 10⁹ A m⁻¹ s⁻¹) (Hugounenq et al. 2012), while nanoflowers of 60 nm consisting of spherical FeMIONs of 22 nm presented SAR values of 1180 W g_Fe⁻¹ (H × f = 17 × 10⁹ A m⁻¹ s⁻¹) (Gavilán et al. 2017).

A particular class of iron oxide MNPs that hold great potential for MH applications are the magnetosomes, produced by magnetostatic bacteria. Unlike chemically synthesized iron oxide MNPs, the magnetosomes are directly synthesized in chain‐like structures consisting of perfectly stoichiometric nanocrystals with controlled sizes and shapes, surrounded by a biocompatible membrane. These characteristics reduce magnetosomes' cellular toxicity. At the same time, they are able to deliver large amounts of heat (Prabhu 2016; Cypriano et al. 2019). For instance, two studies reported SAR values of 880 W g_Fe⁻¹ (H × f = 5 × 10⁹ A m⁻¹ s⁻¹) (Muela et al. 2016) and 960 W g_Fe⁻¹ (H × f = 4 × 10⁹ A m⁻¹ s⁻¹) (Hergt et al. 2005) for magnetosomes produced by the same bacteria: Magnetospirillum gryphiswaldense. As shown by Alphandery et al., chains of magnetosomes inside Magnetospirillum magneticum strain AMB‐1 bacteria yielded a SAR value of 864 W g_Fe⁻¹ (H × f = 7.6 × 10⁹ A m⁻¹ s⁻¹) (Alphandéry et al. 2011). It was enhanced up to 1242 W g_Fe⁻¹, upon the extraction of magnetosomes chains from bacteria, potentially due to the Brownian friction within the liquid. The removal of the membrane surrounding the magnetosomes followed by the disruption of the chains resulted in a decrease of the SAR value down to 950 W g_Fe⁻¹. The highest SAR values reported so far were obtained for magnetosomes inside M. gryphiswaldense bacteria dispersed in water: 2400 W g_Fe⁻¹ (H × f = 9 × 10⁹ A m⁻¹ s⁻¹) (Gandia et al. 2019). The heating efficiency of the magnetosomes was reduced by half (~1200 W g_Fe⁻¹), when the bacteria were randomly distributed in 2% agar medium. When bacteria were aligned parallel to the AFM, the SAR values almost returned to its initial value (2100 W g_Fe⁻¹). This could be a strong evidence of the fact that in water, the bacteria align parallel with the AFM. In this case, the Brownian relaxation of the magnetosomal chains played a minor role being embedded in bacterial matrix. As a consequence, their mutual magnetic interactions are strongly reduced. When these bacteria have been internalized in human lung carcinoma cells A549, the cellular viability and growth were not affected. But the MH experiments, performed on these cells, strongly affected the cancer cell proliferation, making these bacteria promising candidates for cancer applications. Spherical IOMNPs exhibit multiple facets featuring many edges and corners. This type of curved morphology displays many disordered surface spins. The large surface canting effects and high‐surface anisotropy strongly affect the heat dissipation properties of spherical IOMNPs (Noh et al. 2017).

As it was pointed out before, the second strategy for heat generation improvement consisted in tuning the effective anisotropy of IOMNPs by modifying their shape. It has been theoretically demonstrated that cubic MNPs have lower surface anisotropy compared to spheres due to a smaller amount of disordered spins (4 vs. 8%). Several experimental studies have confirmed this phenomenon. The comparison between cubic and spherical IOMNPs, with similar magnetic volumes, show an important increase of SAR values in the case of cubic IOMNPs: 356.2 vs. 189.6 W g_Fe⁻¹ (H × f = 6 × 10⁹ A m⁻¹ s⁻¹) (Bauer et al. 2016); 314 vs. 140 W g_Fe⁻¹ (H × f = 20 × 10⁹ A m⁻¹ s⁻¹) (Das et al. 2016); 395 vs. 150 W g_Fe⁻¹ (H × f = 19.1 × 10⁹ A m⁻¹ s⁻¹) (Nemati et al. 2018) and 1963 vs. 410 W g⁻¹ (H × f = 6.6 × 10⁹ A m⁻¹ s⁻¹) (Elsayed et al. 2017). An extensive research on the heating properties of cubic IOMNPs with sizes ranging from 13 to 38 nm has been performed by Guardia et al. under different conditions of field and frequency (Guardia et al. 2012, 2014). They found that the nanocubes with a mean size of 19 nm exhibit SAR value as high as 2453 W g_Fe⁻¹ (H × f = 15 × 10⁹ A m⁻¹ s⁻¹). Smaller SPIONs nanocubes (13 nm) and larger FeMIONs nanocubes (38 nm) exhibited lower heating performances (<300 W g_Fe⁻¹). This finding can be explained by the lack of hysteresis losses contribution, the existence of large anisotropic fields, and the formation of 3D aggregates. The same trend was observed by Nemati et al. (2018). In their study, the SAR dropped from 800 to 300 W g_Fe⁻¹ as the size of FeMIONs nanocubes was increased from 30 to 106 nm. The coating of large FeMIONs nanocubes with gallol‐polyethylene glycol prevents their aggregation in big clusters and hence increases their SAR value up to 1400 W g_Fe⁻¹ (H × f = 7.7 × 10⁹ A m⁻¹ s⁻¹). A very high SAR (2614 W g_Fe⁻¹, H × f = 0.66 × 10⁹ A m⁻¹ s⁻¹) has been reported for FeMIONs nanocubes of 30 nm in size coated with chitosan, probably as a consequence of the hysteresis losses contribution (Bae et al. 2012). Local symmetry breaking as a result of structural defects, broken symmetry bonds, and surface strain causes the deformation of the nanocubes into nanooctopods (Nemati et al. 2016). Similar to cubic MNPs, the enhanced shape anisotropy of the nanooctopods with size ranging from 17 to 47 nm led to an increase in the heating efficiency from 285 to 415 W g_Fe⁻¹ (H × f = 20 × 10⁹ A m⁻¹ s⁻¹) with respect to spherical MNPs.

Octahedral SPIONs presented also enhanced SAR values compared to the spherical MNPs owing to their higher Ms (Mohapatra et al. 2015; Lv et al. 2015). It was found that octahedrons with size between 6 and 12 nm exhibited SAR values between 163 and 275 W g_Fe⁻¹ (H × f = 6.1 × 10⁹ A m⁻¹ s⁻¹) (Mohapatra et al. 2015). The hysteresis losses generated by the octahedral FeMIONs with size between 40 and 98 nm have boosted the SAR values in the 2480–2630 W g_Fe⁻¹ range (H × f = 23 × 10⁹ A m⁻¹ s⁻¹) (Lv et al. 2015). In this category, one has to mention the polyhedral FeMIONs with a mean size of 34 nm, who displayed a saturation SAR value around 1900 W g_Fe⁻¹ (H × f = 14.2 × 10⁹ A m⁻¹ s⁻¹) (Iacovita et al. 2016).

In the last years, elongated IOMNPs with tunable aspect ratio, the so‐called “nanorods,” received a particular attention (Geng et al. 2016; Das et al. 2016). The pronounced unidirectional shape anisotropy of nanorods having an aspect ratio of 11 (65 nm length × 5.7 nm width) led to an increase in the saturation magnetization and effective anisotropy that improved their heating efficiency up to 862 W g_Fe⁻¹ (H × f = 20 × 10⁹ A m⁻¹ s⁻¹). For comparison, their cubic and spherical counterparts of similar magnetic volumes presented much lower values (314 W g_Fe⁻¹ for cubes and 140 W g_Fe⁻¹ for spheres) (Das et al. 2016). Moreover, the rapid alignment of nanorods, parallel with the external applied AMF, increases appreciably the SAR values up to 1300 W g_Fe⁻¹ (Das et al. 2016). Nanorods with an aspect ratio of 4.5 (45 nm length × 10 nm width), synthesized by the reduction of akaganeite (β‐FeOOH) in organic solvents, exhibited SAR value up to 1072 W g_Fe⁻¹ (H × f = 13 × 10⁹ A m⁻¹ s⁻¹) (Carrey et al. 2011). Instead, nanorods with a larger aspect ratio (10 : 400 nm length × 40 nm width) displayed low SAR value at the same AMF parameters probably due to a higher switching field necessary to reverse their magnetization (Geng et al. 2016).

Among the different classes of iron oxide MNPs presented so far, vortex iron oxide MNPs (VIONs) present outstanding magnetic hyperthermia response, when compared to standard SPIONs, while preserving low cytotoxicity. The VIONs explore a peculiar magnetic configuration known as a magnetic vortex, in which the iron magnetic moment curls in concentric circles, confining the magnetic flux within MNPs. Hence, the VIONs have no magnetic pole and a negligible remanent magnetization, avoiding thus the aggregation due to dipolar interaction. It has been shown that nanorings with an average outer diameter of 70 nm, heights of 50 nm, and an inner to outer diameter ratio of ~0.6, displayed a saturation SAR value of 3050 W g_Fe⁻¹ (H × f = 23.6 × 10⁹ A m⁻¹ s⁻¹) (Liu et al. 2015), while the SAR value of hexagonal nanodiscs with 225 nm in diameter and 26 nm in thickness can be increased up to 4400 W g_Fe⁻¹ by increasing the AMF to 47.8 kA m⁻¹ (H × f = 23.3 × 10⁹ A m⁻¹ s⁻¹) (Yang et al. 2015b). The parallel alignment of both types of VIONs with AMF and the hysteresis losses induced by the vortex domain structure contribute to the high SAR values observed for these morphologies. Nanodiscs with smaller dimensions (12 nm in diameter and 3 nm in thickness) exhibit a very low SAR value (125 W g_Fe⁻¹) despite the high H × f factor used (20 × 10⁹ A m⁻¹ s⁻¹). This is due to the lack of vortex domain configuration (Nemati et al. 2017). For H × f factors below the imposed limit, nanodiscs (150–200 nm in diameter and 10–15 nm in thickness), nanorings (average outer diameter of 165 nm, heights of 75 nm, and an inner to outer diameter ratio of ~0.4), elongated nanorings (average outer diameter of 155 nm, heights of 170 nm, and an inner to outer diameter ratio of ~1), and nanotubes (average outer diameter of 130 nm, heights of 250 nm, and an inner to outer diameter ratio of ~0.45) displayed SAR values of 245 W g_Fe⁻¹ (Ma et al. 2013), 426, 368 and 401 W g_Fe⁻¹, respectively (Dias et al. 2017).