Theranostic Magnetic Nanostructures (MNS) for Cancer

Abstract

Despite the complexities of cancer, remarkable diagnostic and therapeutic advances have been made during the past decade, which include improved genetic, molecular, and nanoscale understanding of the disease. Physical science and engineering, and nanotechnology in particular, have contributed to these developments through out-of-the-box ideas and initiatives from perspectives that are far removed from classical biological and medicinal aspects of cancer. Nanostructures, in particular, are being effectively utilized in sensing/diagnostics of cancer while nanoscale carriers are able to deliver therapeutic cargo for timed and controlled release at localized tumor sites. Magnetic nanostructures (MNS) have especially attracted considerable attention of researchers to address cancer diagnostics and therapy. A significant part of the promise of MNS lies in their potential for “theranostic” applications, wherein diagnostics makes use of the enhanced localized contrast in magnetic resonance imaging (MRI) while therapy leverages the ability of MNS to heat under external radio frequency (RF) field for thermal therapy or use of thermal activation for release of therapy cargo. In this chapter, we report some of the key developments in recent years in regard to MNS as potential theranostic carriers. We describe that the r₂ relaxivity of MNS can be maximized by allowing water (proton) diffusion in the vicinity of MNS by polyethylene glycol (PEG) anchoring, which also facilitates excellent fluidic stability in various media and extended in vivo circulation while maintaining high r₂ values needed for T₂-weighted MRI contrast. Further, the specific absorption rate (SAR) required for thermal activation of MNS can be tailored by controlling composition and size of MNS. Together, emerging MNS show considerable promise to realize theranostic potential. We discuss that properly functionalized MNS can be designed to provide remarkable in vivo stability and accompanying pharmacokinetics exhibit organ localization that can be tailored for specific applications. In this context, even iron-based MNS show extended circulation as well as diverse organ accumulation beyond liver, which otherwise renders MNS potentially toxic to liver function. We believe that MNS, including those based on iron oxides, have entered a renaissance era where intelligent synthesis, functionalization, stabilization, and targeting provide ample evidence for applications in localized cancer theranostics.

1 Introduction

Magnetic nanostructures (MNS) have emerged as promising functional probes for simultaneous diagnostics and therapeutics (theranostic) applications. The diagnostics potential of MNS arises from their role in enhancing the contrast in magnetic resonance imaging (MRI). The therapeutic prospects of MNS stem from thermal activation under external applied radio frequency (RF) field and/or localized release of therapeutic cargo, either through diffusive processes or triggered by thermal activation. Both of these attributes of MNS are related to their unique size-dependent physical properties as well as compatibility of size to typical biomolecules (Fig. 1) [1–7]. The characteristics of MNS are typically measured by saturation magnetization (M_s), remanent magnetization (M_r), and coercivity (H_c). Saturation magnetization is the maximum magnetization value of MNS under an applied magnetic field while remanent magnetization is the magnetization after removing that field. Coercivity is the strength of the applied magnetic field that is necessary to reverse the remanent magnetization back to zero [8]. Superparamagnetism (SPM) occurs when particles are small enough for thermal fluctuations to cause random flipping of magnetic moments, resulting in no remanent magnetization and coercivity in the absence of an applied magnetic field, akin to para-magnetism. However, under external applied magnetic field, the nanostructures can be magnetized and manipulated for transport and thermal activation. The superparamagnetic nanostructures are essential for biomedical studies where no remanent magnetization is critical in preventing their coagulation and sustaining a long period of circulation in the body. Additional properties of MNS required for biomedicine include greater magnetic susceptibility and high saturation magnetization that results in faster and stronger response, even at low external applied magnetic field, and also enhances thermal activation. The magnetic properties of MNS are dictated by the composition, size, shape, and surface properties [9]. Hence, tuning control of these physical properties is essential for success of MNS for in vivo platform.

Fig. 1.

Size scale of MNS as compared to biomolecules. MNS can be adapted to include biomolecules, drugs, or targeting and imaging molecules to form targeted MNS theranostic agents

Typically, MNS comprise a magnetic core and biocompatible coating and/or surface functionalization that allows integration of targeting agents and bio/chemotherapeutics (Fig. 2). Targeting agents have been coupled with MNS for both diagnostic imaging and therapy of specific tumors [5, 10–12]. The diagnostic applications of MNS stem from the MRI [13]. MRI offers clinicians the ability to noninvasively obtain anatomic and metabolic/functional information with high spatial and temporal resolution [13–15]. The technique is based on the response of water proton spin in the presence of an applied magnetic field when triggered with a RF pulse. When external magnetic field is applied, protons align in one direction. Application of the RF pulse perturbs the alignment and the protons relax to the original state via two independent relaxation processes: longitudinal (T₁) and transverse (T₂) relaxation that are used to generate the MR images. The difference in water concentration and local environment between organs and tissues results in intrinsic contrast in MR images.

Fig. 2.

Functional architecture of MNS and theranostic modalities. MNS are comprised of thermally active magnetic core and biocompatible coating and/or functionalization that allows integration of targeting agents and bio/chemotherapeutics

The spatial resolution as well as the sensitivity (S/N) of the MR images can be enhanced with the use of contrast agents. Paramagnetic molecular complexes, such as Gd(III) chelates, are used as T₁ contrast agents that increase signal intensity, i.e., higher r₁ relaxation, and appear bright in T₁-weighted images [16]. T₁ contrast agents are not covered in this chapter, however, a number of recent reviews have appeared [15–20]. MNS are used as T₂ contrast agents that decrease the signal intensity, i.e., higher r₂ relaxation, and appear dark in T₂-weighted images. When water molecules (protons, more specifically) diffuse into the periphery of the induced dipole moment by MNS, the T₂ relaxation time of the protons is shortened, which enhances the negative contrast that helps in differentiating between pathogenic targets and normal tissues in T₂ weighted MRI images (Fig. 3).

Fig. 3.

T₂ contrast enhancement in water due to MNS. When water molecules diffuse into the periphery of the induced dipole moment by MNS, the T₂ relaxation time of the water protons is shortened which enhances the negative contrast

Several MNS-based T₂ contrast agents (e.g., Feridex and Resovist) have been clinically approved [21, 22]. The MRI contrast enhancement effect is measured by the relaxation rate R₁ (s⁻¹) and the relaxivity coefficient r₂, a slope of R₂ against MNS concentration. The R₂ relaxation rate of MNS is defined as

$$R_2=\frac{1}{T_2}=\frac{256{\pi }^2{\gamma }^2}{405}{M}_s^{2}V\frac{r^2}{D(1+\frac{L}{r})}$$ (1)

where T₂ is transverse relaxation time, γ is proton gyromagnetic ratio, M_s is saturation magnetization, V is volume of MNS, D is diffusion coefficient of water molecules, r is radius of MNS core, and L is thickness of MNS surface coating. [23]. The higher relaxivity corresponds to a better contrast effect. Based on Eq. (1), MNS should have high magnetization (M_s), large volume (V), and thin surface coating (small L) for better contrast effect.

MNS can generate heat under external RF field that make them very useful in cancer therapeutics. Under an external RF field (typically a few hundred kHz), superparamagnetic MNS switch their magnetization direction along the field directions, back and forth. The frictions caused by the physical rotation of the MNS (Brownian relaxation) and the magnetization reversal within the MNS (Neel relaxation) lead to the loss of magnetic energy and the generation of thermal energy [24]. The capability of generating heat at any targeted areas can be used for directly killing cancer cells via the thermal therapy and/or as an actuator for bio/chemo therapy cargo release. The thermal activation of MNS is measured by the specific absorption rate (SAR) that is measured as the initial temperature rise of the MNS solution per unit volume or mass.

$$\mathit{SAR}=\pi {\mu }_0{\chi }_0{H}_0^{2}f\frac{2\pi f\tau }{1+{\left(2\pi f\tau \right)}^2}$$ (2)

where μ₀ is vacuum permeability, χ₀ is equilibrium susceptibility, H₀ is RF field amplitude, and f is frequency of the external RF field. The higher SAR is crucial for clinical use since that would require a smaller amount of MNS to be injected into the patient. According to Eq. (2), SAR highly depends on various parameters such as the size, size distribution, shape, chemical composition and surface modification, and saturation magnetization of the particles [25]. In addition, it is clear that SAR values depend on the frequency f and the field amplitude H₀ of the applied field. However, in order to apply hyperthermia safely to patients and avoid any detrimental effect on healthy tissues due to electromagnetic radiation exposure, the H₀f factor should not exceed a threshold that was experimentally estimated to equal 5 × 10⁹ A m⁻¹ s⁻¹ [26]. Therefore, MNS with an exceptional SAR value that can generate heat under H₀f limit is highly desirable.

2 Synthesis and Characteristics of MNS: Prospects for Theranostics

For successful theranostic applications, MNS should be monodispersed and have uniform composition because the magnetic properties of MNS depend on the size, shape, and composition. It is clear from Eqs. (1) and (2) that the particles should possess high saturation magnetization and magnetic susceptibility, and be stable to a range of pH and salt concentrations. A key parameter for the magnetization of MNS is size. In a bulk magnetic material, all of the magnetic spins are aligned parallel to the applied magnetic field. However, in the nanoscale regime, a magnetically disordered spin-glass-like surface layer is formed. As the nanoparticle size decreases, such surface spin-canting effect becomes more pronounced and causes a drop in the saturation magnetization. While high saturation can be achieved with larger size particles, avoiding the surface-canting effect [27], the particle size should be under the superparamagnetic limit, which is typically less than ~20–30 nm for the majority of MNS. Further, the particles should have a coating or surface functional moieties that improve dispersion, biocompatibility, and provides a surface that can be functionalized. Strict attention to these parameters is essential during the design, synthesis, and formulation of MNS in order to be useful for in vivo applications.

MNS can be fabricated by either top-down (mechanical attrition) or bottom-up (chemical synthesis) approaches [9]. Since magnetic properties change with size and composition of MNS, chemical routes are preferred since they can synthesize MNS with uniform composition and size. The chemical methods include co-precipitation, microemulsion, thermal decomposition and/or reduction, hydrothermal synthesis, and polyol synthesis. Two excellent reviews describing MNS fabrication methods have recently appeared and we will provide only a brief summary here [28, 29].

The most common synthetic strategy involves aqueous precipitation of iron salts with in situ, or post-synthesis addition of surfactant [30]. This strategy has notable limitations yielding monodispersity [31]. The microemulsion method does produce MNS of narrower size distribution compared to aqueous precipitation, but suffers from low yields [32]. Recently, the thermal decomposition/reduction method has gained considerable attention since this technique offers fine control over the final particle size, shape, and crystal structure compared to other methods and is scalable [33, 34]. Monodispersed MNS are formed due to the reaction conditions that yield a quick nucleation step followed by slower growth phase. However, the reaction occurs in organic solvent containing hydrophobic stabilizers, which requires additional surface modifications to the MNS to impart aqueous stability.

Here, we discuss different types of MNS synthesized using chemical methods and their magnetic properties. Table 1 summarizes different MNS core materials, their magnetic properties, and r₂ relaxivity.

Table 1.

Summary of MNS with core diameter, surface coating, magnetic properties, and r₂ relaxivity

MNS core material	Core diameter (nm)	Surface coating	Magnetic moment (emu/g)	B0 (T)	r2 (mM−1 s−1)	References
Fe3O4 (Resovist)	4	Carboxy-Dextran	N/A	1.5	186	[35]
Fe3O4 (Feridex)	5	Dextran	45	1.5	120	[36]
Dy-SiO2-(Fe3O4)n	9	DMSA	N/A	9.4	397	[37]
Fe3O4	4–12	DMSA	25–101	1.5	78–218	[38]
Fe3O4	12	Nitrodopa-PEG600	N/A	1.5	396	Unpublished
Fe3O4	14	DSPE-mPEG1000	N/A	0.47	385	[39]
Fe3O4	58	DSPE-mPEG2000	132	1.5	324	[40]
MnFe2O4	6–12	DMSA	68–110	1.5	208–358	[38]
CoFe2O4	12	DMSA	99	1.5	172	[38]
NiFe2O4	12	DMSA	85	1.5	152	[38]
Zn0.34Fe0.66OFe2O3	5	DSPE-PEG	54.1	0.55	34.7	[41]
Zn0.4Fe0.6Fe2O4	15	DMSA	161	4.5	687	[42]
Zn0.4Mn0.6Fe2O4	15	DMSA	175	4.5	860	[42]
Fe40Co60	7	Phospholipid-PEG	215	1.5	644	[43]
Fe/Fe3O4	15	OAm-PEG	164	3	220	[44]
Fe/Fe3O4	16	DMSA	139	1.5	312	[45]
Fe/MnFe2O4	16	DMSA	149	0.47	356	[46]

r₂ transverse relaxivity; B₀ magnetic field strength; DSPE-PEG 1,2-Distearoyl-sn-glycero-3-phosphoethanolamine-N-[methoxy(polyethylene glycol)]; DMSA 2,3-dimercaptosuccinic acid; OAm-PEG oleylamine-α,ω-bis(2-carboxyethyl)poly(ethylene glycol)

2.1 Ferrite MNS

Fe₃O₄ MNS are extensively used in biomedicine because of their biocompatibility and ease of synthesis [47, 48]. The magnetic moment of superparamagnetic Fe₃O₄ MNS is dependent on the size with smaller particles producing lower magnetic moments [49]. Hence, the size of Fe₃O₄ MNS can be tuned by changing the reaction conditions such as reflux temperature, reflux time, and heating rate. The magnetic moment of Fe₃O₄ MNS was tuned from 25 to 43, 80, and 102 emu/(g Fe) with change in the size from 4, 6, 9, and 12 nm, respectively (resulting in r₂ values of 78, 106, 130, and 218 mM⁻¹ s⁻¹) [27]. Further, the magnetic moment of Fe₃O₄ MNS can be modified by doping transition divalent metal ions (Co²⁺, Ni²⁺, Mn²⁺, and Zn²⁺). By adding different metal precursors during Fe₃O₄ synthesis, monodisperse MFe₂O₄ MNS were synthesized [38]. The metal ferrite nanostructures have an inverse spinel crystal structure composed of face-centered cubic packed lattice of oxygen atoms with octahedral sites (O_h) occupied by Fe³⁺ and M²⁺ ions and tetrahedral sites (T_d) occupied by Fe³⁺ ions (Fig. 4a). The magnetic spins of the ions at the O_h and T_h sites align opposite to each other. Hence the spins of Fe³⁺ ions at O_h and T_h cancel each other and net magnetization of MFe₂O₄ MNS is decided by magnetic moment of M²⁺ ions (Fig. 4b). The magnetization of NiFe₂O₄, CoFe₂O₄, Fe₃O₄, and MnFe₂O₄ were found to be 85, 99, 101, 110 (emu/g metal ions), respectively, depending upon the magnetic moments of the M²⁺ ions. This resulted in r₂ values of 152, 172, 218, and 358 mM⁻¹ s⁻¹ (Fig. 4c) [38].

Fig. 4.

MFe₂O₄ (where M = Mn, Fe, Co, Ni) MNS with inverse spinel structure and its magnetic spin alignments. The mass magnetization values and r₂ relaxivity values of MFe₂O₄ MNS are proportional to the magnetic moments of the divalent ions (M²⁺) [38]

Interestingly, doping of nonmagnetic Zn in Fe₃O₄ and MnFe₂O₄ MNS resulted in (Zn_xFe_1−x)Fe₂O₄ and (Zn_xMn_1−x)Fe₂O₄ MNS, respectively, that exhibit extremely high net magnetic moment and r₂ relaxivity [42]. The magnetization of (Zn_xMn_1−x)Fe₂O₄ was dependent on the Zn doping level and were found to be 125, 140, 154, 166, 175, and 137 emu/g metal ions for x = 0, 0.1, 0.2, 0.3, 0.4, and 0.8, respectively, resulting in the r₂ values of 422, 516, 637, 754, 860, and 388 mM⁻¹ s⁻¹, respectively (Fig. 5). The r₂ value for Zn_0.4Mn_0.6Fe₂O₄ MNS is the maximum r₂ value of reported to date for MNS [42], eight times higher than r₂ of Feridex [36]. (Zn_xFe_1−x)Fe₂O₄ MNS exhibited a similar trend, but the magnetization and r₂ values were slightly lower than (Zn_xMn_1−x)Fe₂O₄ MNS (Fig. 5) [42].

Fig. 5.

Saturation magnetization and r₂ relaxivity (at 4.5 T) of (Zn_xMn_1–x)Fe₂O₄ MNS at different Zn²⁺ doping levels. The (Zn_xMn_1–x)Fe₂O₄ MNS showed significantly high r₂ relaxivities compared to conventional iron oxide MNS [42]

2.2 Metallic MNS

MNS based on transition metals of Fe, Co, Ni, and their alloys have higher magnetic moments than their oxide counterparts [44, 50–53]. Saturation magnetization of bulk FeCo (240 emu/g) and Fe (218 emu/g) is particularly high compared to bulk Fe₃O₄ (90 emu/g). Using the same mass of metallic MNS would then produce a far greater impact than their oxide counterparts, improving the T₂ contrast enhancement and therapeutic efficacy of drug delivery. However, the metallic MNS carries their own set of disadvantages like chemical instability, leaching of the noniron elements, and toxicity that renders them questionable for in vivo applications [9, 53, 54]. In addition, these pure metal nanoparticles are also ferromagnetic at room temperature, rather than superparamagnetic. This means that once they are magnetized, they will remain that way regardless of whether an external magnetic field is withdrawn, resulting in an aggregation. A number of reports have been published demonstrating coatings that prevent aggregation and ensure chemical stability of metallic MNS. Options under consideration include inert metals, such as Au and Ag, peptide capping ligands, ferrites, graphite, and silica [43, 44, 52, 55–57]. For example, after graphitic shell coating, FeCo MNS were stable up to 1 month and showed very high magnetic moment (215 emu/g) and r₂ relaxivity (644 mM⁻¹ s⁻¹ which is far superior to conventional ferrite MNS [43, 52]. Crystalline Fe₃O₄ shell was also used to protect metallic Fe MNS that resulted in M_s and r₂ of 164 emu/g and 220 mM⁻¹ s⁻¹, respectively [44, 56]. Co MNS were coated with Au shell to provide an inert, biocompatible and stable shell with a well-known surface chemistry. Though the bulk saturation magnetization of Co is ~160 emu/g, the measured value was found ~100 emu/g after Au coating, which was still higher than Fe₃O₄ MNS (75–80 emu/g) [55].

2.3 Multifunctional MNS

Hybrid MNS with two or more different functional units, such as Au–Fe₃O₄, FePt–CdS, and Fe₂O₃–carbon nanotube can be synthesized through seed mediated growth. In such a heterogeneous nanostructure, each unit exhibits its unique magnetic, optical, or electronic properties [10, 58–60]. Au–Fe₃O₄ nanostructures were prepared that preserved the optical property of Au (plasmonic absorption at ~530 nm) as well as the magnetic property of Fe₃O₄ MNS (M_s = 80 emu/g) [61]. This approach was extended to prepare semiconductor-metal alloy [62], semiconductor–metal oxide [63], and carbon nanotube–metal oxide complex [64]. MNS have been coupled with a wide range of fluorophores for multimodal imaging applications [65–67]. Lastly, T₁/T₂ MRI agent were prepared by conjugating Gd (III) based chelating agent with MNS [68].

3 Coating and Functionalization of MNS

In order to apply MNS in vitro and subsequently in vivo, the surface needs to be functionalized so that it; (i) protects against agglomeration; (ii) provides biocompatibility and chemical handles for the conjugation of drugs and targeting ligands; (iii) limits nonspecific cell interactions; and (iv) enhances MNS pharmacokinetics [69]. A diverse group of organic and inorganic coatings has been investigated including DMSA [27], PEG [70, 71], dextran [72], chitosan [73], liposomes [74], gold [75], and silica [76]. MNS coating can be achieved via a number of approaches, including in situ coating, post-synthesis adsorption, and post-synthesis end grafting [15]. Here, we discuss some of the most common coatings, their methods of attachment, and examples in cancer targeting.

3.1 Poly(Ethylene Glycol) (PEG)

PEG is a neutral and amphiphilic polymer that has been used clinically as excipients in FDA approved pharmaceutical formulations [77]. PEG coating of MNS improves their dispersion in biological media and increases blood circulation time since they are not readily recognized by the reticuloendothelial system (RES) [78]. Lutz et al. demonstrated in situ coating of PEG onto Fe₃O₄ MNS under aqueous conditions [79] while PEG grafting was achieved by single-point chemical anchoring through different functional groups including silanes [70], phosphate derivatives [80] and dopamine [81]. Peng and Sun reported ligand exchange with bifunctional PEG with dopamine [56]. Most recently, nitrodopamine has been proposed as an ultrastable chemical anchor for MNS [82]. We have developed MNS with high buffer stability by coating Fe₃O₄ with bifunctional PEG conjugated with nitrodopamine and carboxylate terminal groups [83]. The nirodopamine was covalently attached to the Fe₃O₄ surface by one end and the carboxylate group at the other end was kept open to functionalize with targeting ligands or therapeutic agents.

3.2 Dextran

Dextran is a branched polysaccharide comprised of glucose subunits and is widely used for MNS coatings because of its biocompatibility and polar interactions (chelation and hydrogen bonding). Addition of dextran during synthesis of Fe₃O₄ via the co-precipitation method resulted in dextran coated Fe₃O₄ MNS [84]. Subsequent iterations of this method produced clinically approved ferumoxtran-10 (AMI-277) and ferumoxides (AMI-25) [85–90]. These two structures have cores of ~5 nm, but differ significantly in dextran coating thickness (20–40 nm vs. 80–150 nm) which results in varying blood circulation times (24 h for ferumoxtran and 2 h for ferumoxides) [35]. Since the dextran molecules are adhered nonspecifically through hydroxyl interactions with the iron oxide core, there is always a possibility of desorption [91]. In order to prevent this, the dextran polymers were chemically cross-linked on MNS surface [92]. Using this strategy, clinically approved ferumoxytol [93] and ferucarbotran [22] have been synthesized.

3.3 Silica

Silica coating on MNS is popular because of the ease of synthesis and aqueous stability. By hydrolyzing silica precursors in basic solution, a uniform and thickness-controllable silica coating on MNS was obtained [76, 94, 95]. The silica shell have been used as a carrier for anticancer drugs (e.g., paclitaxel) and fluorescent molecules (e.g., fluorescein isothiocyanate (FITC)) [96]. Imparting additional functionalities to the silica coating have enabled targeting and labeling functionality. By addition of 3-aminopropyl-triethoxysilane (APS) to the silica precursors, we have coated silica shells with primary amine groups with controlled thickness [97]. Similarly, by reacting APS with isothiocyanate functionalized fluorescent dyes, Lu et al. were able to develop fluorescent MNS, a multimodal diagnostic agent [94]. Currently, silica coated MNS is available as ferumoxil (AMI-121), an orally ingested T₂ contrast agent for delineation of the intestinal loops from adjacent tissues and organs [21].

4 Pharmacokinetics and Biodistribution

The two most important factors that determine MNS pharmacokinetics are their surface characteristics and hydrodynamic size [98, 99]. Interplay of these properties with the reticuloendothelial system (RES) clearance determines plasma lifetime (blood circulation time). In RES clearance, circulating opsonin proteins adsorb to MNS surface (opsonization) that are recognized and removed from the bloodstream by tissue macrophages (Fig. 6). It has been shown that MNS with a hydrodynamic diameter of 10–100 nm are pharmacokinetically optimal for in vivo applications [100]. MNS smaller than 10 nm are subject to tissue extravasation and renal clearance, whereas those larger than 100 nm are quickly opsonized and eliminated from the circulation via the RES [31]. Decuzzi et al. produced models suggesting that within this range, smaller size nanostructures have longer blood circulation time [101].

Fig. 6.

Schematic illustration of RES clearance of MNS. MNS larger than 100 nm are absorbed by circulating opsonin proteins that are recognized by macrophages and removed from the bloodstream

With respect to the surface properties, charge can affect pharmacokinetics of MNS by enhancing their interactions with the plasma proteins and nontargeted cells, resulting in short blood circulation time [102]. In addition, hydrophobic groups on the surface of MNS induce the agglomeration of the MNS upon injection, leading to rapid removal by the RES. The rate of clearance, however, can be reduced by modification of MNS surfaces with coatings that resist RES interactions. As mentioned, surface modification with molecules such as the hydrophilic PEG has been a hallmark solution to many pharmacokinetic problems, including MNS [103, 104]. PEG chains linked to MNS reduce opsonization and macrophage uptake processes through steric repulsion, prolonging their circulation times [78, 105].

MNS biodistribution and cell uptake is significantly influenced by their physicochemical properties [99, 106, 107]. For example, it has been reported that MNS smaller than 150 nm accumulates in the bone marrow, heart, kidney, and stomach [108] while MNS larger than 150 nm are found in the liver and spleen [109]. Villanueva et al. showed that the charge and nature of surface functionalizing molecules on MNS affected their uptake of cancer cells [110]. They found that cells had effective uptake of positively charged aminodextran-MNS, minimal uptake of neutral charged dextran coated MNS, and low uptake of negatively charged DMSA-coated MNS [110]. Chouly et al. have found that negatively charged MNS gets opsonized quicker than neutral MNS and had greater liver uptake [99].

5 Targeting of MNS to Localized Cancer Tumors

The targeting of MNS for selected tumor tissues is critical in both diagnostic imaging and therapeutics [5, 10, 11]. Since nonspecific cell binding can place healthy tissue at risk, MNS have been engineered to target tumor tissues through passive and active targeting approaches.

5.1 Passive Targeting

Passive targeting uses the predetermined physicochemical properties of MNS to specifically migrate to selected tissues. The most common example of passive targeting is the enhanced permeability and retention (EPR) effect where MNS smaller than 200 nm can accumulate in many tumor tissues passively in solid tumors [111]. The compromised vasculature of a solid tumor facilitates passive MNS extravasation from the circulation into the tumor interstitium (Fig. 7) [112]. By contrast, endothelial cells of normal tissue vessels are closely packed and present a barrier for MNS penetration. However, passive targeting is limited to specific tumors since success of EPR effect depends on a number of factors such as lymphatic drainage rate, degree of capillary disorder, and blood flow which varies in different tumor types [113, 114].

Fig. 7.

Passive targeting of MNS via enhanced permeability and retention (EPR) effect. The compromised vasculature of a solid tumor facilitates extravasation of MNS of size less than 200 nm from the circulation into the tumor interstitium, while endothelial cells are closely packed and present a barrier for MNS penetration

5.2 Active Targeting with Targeting Agents

Because passive targeting is available for only certain types of tumors and does not necessarily guarantee internalization of MNS by targeted cells, MNS can be modified with tumor-selective agents to employ active targeting [115, 116]. These agents are complementary to unique receptors that are overexpressed or present on tumor cells. A variety of targeting agents have been used for MNS, depending on the specific target, and these are reviewed elsewhere [5, 10]. Some of the studies include: small organic molecules [115, 117, 118], peptides [119], proteins [120], and antibodies [27]. The density and molecular organization of these ligands significantly influence MNS binding to target cells due to the multivalency phenomenon [121]. Some of the targeting agents can be used to facilitate MNS internalization into cells, primarily via endocytosis [5]. However, synthesis of these targeting agents is expensive and involves complicated chemistry. Therefore, the process of scaling up the synthesis is challenging and may be a hurdle for clinical applications.

5.3 Active Targeting with External Magnetic Field

Accumulation of MNS can be realized by applying external magnetic field on the target site, a unique feature for MNS. Magnetic targeting has been studied for a number of tumor models [122, 123]. This technique was successfully implemented in a clinical trial to deliver the chemotherapeutic, doxorubicin, to hepatocarcinoma cells [124]. David et al. have explored magnetic targeting to brain tumors with PEI functionalized MNS [102, 122]. While successful, the efficacy of magnetic targeting is limited to target tissue that is close to the body's surface, since the magnetic field strength decreases with the distance from the magnetic source.

6 MNS for Diagnostic Imaging of Cancer

The diagnostic imaging applications of MNS have been realized as T₂ contrast agents in MR imaging over the past 20 years [13–15, 28, 125]. Owing to their significant deposition in liver, several MNS-based clinically approved T₂ contrast agents (Feridex I.V.®, Resovist®, and Gastromark®) have been used for liver imaging of humans [21, 22]. In order to extend MR visibility to image tumor at other parts, targeting agents have been coupled to MNS [85]. For example, Artemov et al. used Fe₃O₄ MNS conjugated with biotinylated Her-2/neu antibody Herceptin to generate strong T₂ contrast in breast cancer cell lines (AU-565, MCF-7, and MDA-MB-231) overexpressed with tyrosine kinase Her-2/neu receptors [126]. Contrast observed in MR images was found to be proportional to the expression level of kinase her-2/neu receptors for the given cell lines. Gao et al. successfully performed targeted MR imaging of human colon carcinoma xenograft tumors in mice by conjugating a cancer-targeting antibody (rch 24 mAb), to an 11 nm Fe₃O₄ MNS [127]. T₂ and T₂*-weighted MR images acquired before and after injection showed that the tumor site turned dark as early as 10 min after the injection of the rch 24 mAb conjugates and became darker and bigger until 24 h. In contrast, the Fe₃O₄ MNS without rch 24 mAb showed nearly no variation after injection [127]. Sun et al. reported c(RGDyK) peptide-coated Fe₃O₄ MNS and demonstrated their in vivo tumor-specific targeting capability [128]. When administrated intravenously in a mice bearing U87MG tumors, the c(RGDyK) peptide-coated Fe₃O₄ MNS accumulated preferentially in the integrin αvβ3-rich tumor area resulting in a significant drop in the tumor MR signal intensity [128].

Other than target-specific molecular imaging, MNS-based contrast agents have been used in MRI for cell-based therapy since cells must be tracked in vivo to optimize cell therapy [129–134]. MRI-based immune cell tracking using MNS has been applied to many types of preclinical studies, such as tumor targeting of cytotoxic T cells and natural killer cells [135, 136], organ-specific targeting of autoimmune T cells [137], and neural stem cells [138]. De Vries et al. have shown that in vivo magnetic resonance tracking of MNS labeled dendritic cells is feasible in humans in conjunction with detailed anatomical information in melanoma patients. In contrast to scintigraphic imaging, MRI allowed assessment of the accuracy of dendritic cell delivery and of inter- and intranodal cell migration patterns [139–143]. Recently, Bulte et al. evaluated the long-term clinical tracking of MNS labeled stem cells after intracerebroventricular transplantation in an 18-month-old patient with global cerebral ischemia [144]. Twenty-four hours post-transplantation, MRI was able to detect hypointense cells in the occipital horn of the lateral ventricle. The signal gradually decreased over 4 months and became undetectable at 33 months.

To improve T₂ contrast for advanced MR imaging, researchers have been studying various parameters that affect the r₂ relaxivity of MNS. In the following sections, we discuss the parameters that have improved T₂ contrast of MNS for cancer diagnostic imaging. They have been divided into four categories; (i) size, shape, and composition control; (ii) nanoassembly of MNS; (iii) coating of MNS; and (iv) MNS with metal core.

6.1 Size, Shape, and Composition Control

According to Eq. 1, the R₂ of MNS is proportional to saturation magnetization and volume. Since the saturation magnetization of an MNS is proportional to its size due to surface spin-canting effects, the r₂ value of MNS can be increased by increasing the size of MNS. However, for biological imaging applications, the hydrodynamic size of MNS should be below 100 nm in order to have longer circulation times and to avoid nonspecific uptake [145]. Cheon et al. investigated the size effect where Fe₃O₄ MNS of diameter 4, 6, 9, and 12 nm resulted in the r₂ relaxivity of 78, 106, 130, and 218 mM⁻¹ s⁻¹, respectively (Fig. 8). The MR contrast changed from light gray to black or from red to blue in color-coded images (Fig. 8) [27]. The 9 nm Fe₃O₄ MNS conjugated with Herceptin was used to image a breast cancer cell line, SK-BR-3, which possesses overexpressed HER2/neu cancer markers. Herceptin was selected due to its specific binding properties against a HER2/neu receptor. In the T₂-weighted MR images, treatment of 9 nm Fe₃O₄ MNS-Herceptin probe conjugates to the SK-BR03 breast cancer cell lines resulted in the significant negative contrast of the MR images compared to nontreated cell lines [27]. Chen et al. studied the size effect of polyvinylpyrrolidone (PVP)-coated Fe₃O₄ MNS (core size 8, 23, 37, and 65) on MRI of hepatic lesions in vivo [146]. PVP-Fe₃O₄ MNS with core size 37 and 65 nm showed higher r₂ relaxivity (239 and 248 mM⁻¹ s⁻¹, respectively) compared to other sizes. When administered in nude mice bearing orthotopic Huh7 liver cancer, PVP-Fe₃O₄ MNS with 37 nm core size showed higher contrast change compared to Feridex in T₂ and T₂* weighted MR images [146].

Fig. 8.

Size effects of Fe₃O₄ MNS on r₂ relaxivity. a TEM images, b saturation magnetization values, c T₂-weighted MR images (top black and white, bottom color), and d the r₂ relaxivity values of 4, 6, 9, and 12 nm sized Fe₃O₄ MNS. The r₂ relaxivity value increased with size of Fe₃O₄ MNS which resulted in the T₂ contrast change from light gray to black in T₂ weighted MR images or from red to blue in the corresponding color-coded images. Reprinted with permission from [27]. Copyright 2005 American Chemical Society

The change in shape of the MNS has been used to increase r₂ relaxivity. We reported size and shape effects of CoFe₂O₄ MNS on their r₂ relaxivity. Spherical CoFe₂O₄ MNS of various sizes were synthesized via seed mediated growth method, while faceted irregular CoFe₂O₄ MNS were synthesized via the same method but in the presence of a magnetic field [147]. While the r₂ relaxivity coefficient of CoFe₂O₄ MNS increased with an increase in size for spherical particles, faceted CoFe₂O₄ MNS showed higher r₂ relaxivity than spherical CoFe₂O₄ MNS [147]. Recently, Gao et al. reported octapod shape Fe₃O₄ MNS that exhibited highest relaxivity of 679.3+/30 mM⁻¹ s⁻¹ for Fe₃O₄ and demonstrated in vivo imaging and tumor detection [148].

As discussed earlier, the magnetization (and hence r₂) of MNS can be influenced by doping with magnetically susceptible elements. The effect of metal doping on the r₂ relaxivity was investigated for MFe₂O₄ in which Fe²⁺ ions were replaced by other transition metal dopants (Mn²⁺, Zn²⁺, Ni²⁺, and Co²⁺). The r₂ of 12 nm MnFe₂O₄ MNS was observed 358 mM⁻¹ s⁻¹ compared to 218 and 62 mM⁻¹ s⁻¹ from 12 nm Fe₃O₄ and dextran coated cross-linked 4 nm Fe₃O₄ (CLIO) MNS, respectively (Fig. 9) [38]. This increased MR contrast enhancement was tested to detect breast and ovarian cancer tumor in mice after Herceptin conjugation and intravenous injection (Fig. 9) [38]. MnFe₂O₄-Herceptin conjugates produced higher contrast than CLIO-Herceptin conjugates at the tumor site after 2 h (Fig. 9 a–d). Quantitatively, r₂ increase up to 34 % was observed for MnFe₂O₄-Herceptin conjugates in comparison to 5 and 13 % for CLIO-Herceptin conjugates and 12 nm Fe₃O₄-Herceptin conjugates, respectively (Fig. 9 e,f).

Fig. 9.

In vivo MR detection of cancer in a mouse implanted with the cancer cell line NIH3T6.7 using 12 nm MnFe₂O₄, 12 nm Fe₃O₄ and dextran coated cross-linked 4 nm Fe₃O₄ (CLIO) MNS. T₂-weighted MR images of the mouse (a) before injection, (b) after 1 h injection, and (c) after 2 h injection of MnFe₂O₄ in comparison to (d) after 2 h injection of CLIO. MnFe₂O₄-Herceptin conjugates produced higher contrast than CLIO-Herceptin conjugates at the tumor site after 2 h. e Plot of R₂ change versus time. Increase in R₂ up to 34 % was observed for MnFe₂O₄-Herceptin conjugates in comparison to 5 and 13 % for CLIO-Herceptin conjugate (dots) and 12 nm Fe₃O₄-Herceptin conjugates, respectively. f Change in R₂ values was confirmed in the ex vivo MR images of explanted tumors (8 h). Reprinted with permission from [38]. Copyright 2007 Nature Publishing Group

6.2 Nanoassembly of the MNS

Nanoassembly of MNS made up of numerous MNS have shown to increase r₂ relaxivity since an individual MNS in the nanoassembly is more efficient at dephasing the spins of the surrounding water protons [149]. For example, Feridex exhibits much higher r₂ relaxivity than monocrystalline iron oxide MNS because Feridex consists of several iron oxide nanoparticles embedded in a dextran shell [36]. Advances in chemical synthesis have enabled the preparation of nanoassemblies of tunable size and shape. Fe₃O₄ MNS have been assembled on dye-doped mesoporous silica nanoparticles (Fig. 10a) [37, 150]. The r₂ value of Fe₃O₄ coated silica nanoparticles was 2.8 times higher than that of well-dispersed Fe₃O₄ nano-particles (Fig. 10b) [150]. After subcutaneous injection into each dorsal shoulder of a nude mouse, MCF-7 cells labeled with Fe₃O₄-MNS were clearly detected as a dark volume of subcutaneous tumor in the T₂-weighted MR image, while the unlabeled MCF-7 cells did not show any MR contrast enhancement [150]. We have demonstrated controlled assembly of 6 nm amine functionalized Fe₃O₄ MNS [151]. The size of aggregates was ~40 nm and the r₂ value was 315 mM⁻¹ s⁻¹, which was significantly higher than r₂ of monodisperse 6 nm Fe₃O₄ (~100 mM⁻¹ s⁻¹) [38].

Fig. 10.

MR signal enhancement by assembly of Fe₃O₄ on SiO₂ nanoparticles. a Schematic illustration of the synthetic procedure for Fe₃O₄ decorated mesoporous silica nanoparticles. b Relaxivity values and T₂ weighted MR image of Fe₃O₄ decorated SiO₂ nanoparticles (Fe₃O₄-MSN) and free Fe₃O₄ nanoparticles. The r₂ relaxivity of Fe₃O₄ decorated SiO₂ nanoparticles was increased by 2.8 times as compared to free Fe₃O₄ nanoparticles, hence darker signal was observed in T₂ weighted MR image at the same concentration of Fe. Reprinted with permission from Ref. [150]. Copyright 2010 American Chemical Society

6.3 Coating of MNS

Typically, the role of the MNS coating is to provide stability, biocompatibility, and enhanced blood circulation times. In addition, the coating on MNS affect the r₂ relaxivity since it can increase the residence time of the surrounding water molecules by forming hydrogen bonds [152]. For example, polyethylene glycol (PEG) is one of the most common coatings that is used to make MNS stable in aqueous/biological media. Each ethylene glycol subunit in the PEG associates with two or three water molecules which slows water diffusion, resulting in a high r₂ relaxivity [152].

We have developed MNS with Fe₃O₄ core and PEG coating that shows excellent buffer stability [83]. It was found that the PEG coating not only provides stability but the thickness of PEG coating also affects the r₂ relaxivity of Fe₃O₄ MNS. The thickness was varied by changing the molecular weight of PEG that resulted in different r₂ relaxivities. The r₂ relaxivity of 12 nm Fe₃O₄ MNS was found at 160, 194, 277, and 396 mM⁻¹ s⁻¹ when the molecular weight of PEG was 200, 400, 500, and 600 Da (Fig. 11). Bao et al. reported different r₂ relaxivities for phospholipid (DSPE)-PEG-coated Fe₃O₄ MNS by changing their core size (5 and 14 nm) and PEG length (mol. wt. 550, 750, 1000, 2000, and 5000). The highest r₂ value of 385 mM⁻¹ s⁻¹ was obtained for 14 nm Fe₃O₄ MNS coated with DSPE-PEG1000 [39]. In vivo tumor imaging was done using Fe₃O₄ MNS (5 and 14 nm) coated with DSPE-PEG1000 and conjugated with antibodies against mouse VEGFR-1. Tumors were induced by implanting human U87 glioblastoma cells subcutaneously in the nude mice. Tail vein injection of the 14 nm Fe₃O₄ MNS with r₂ relaxivity ~385 mM⁻¹ s⁻¹ produced more T₂ contrast enhancement of the tumor tissues in comparison to 5 nm Fe₃O₄ whose relaxivity was 130 mM⁻¹ s⁻¹.

Fig. 11.

r₂ relaxivity values of 12 nm nitrodopamine-PEG functionalized Fe₃O₄ MNS with molecular weight of PEG 200 (EG2), 400 (EG4), 500 (EG5), and 600 Da (EG6) in comparison to Ferumoxytol and Ferumoxides (unpublished). It was found that the PEG coating not only provides stability but the thickness of PEG coating also affects the r₂ relaxivity of Fe₃O₄ MNS. The highest r value of 396 mM⁻¹ s⁻¹ with PEG 600 (EG6) was almost four times that of Feridex (Ferumoxide)

Silica coating on MNS is another common method to provide aqueous stability. It has been observed that r₂ relaxivity of MNS decreases with increment in silica shell thickness. Pinho et al. observed a systematic decrease in the r₂ relaxivity, from 228 to 23 mM⁻¹ s⁻¹, after coating silica shell up to 20 nm on a 10 nm γ-Fe₂O₃ MNS [95]. The authors attributed this effect to two regions of silica shell coating, an inner water-impermeable layer and an outer water-permeable layer. The outer layer provided colloidal stability while the impermeable layer reduced the interaction between the MNS and water protons significantly, resulting in a drop in r₂ relaxivity.

6.4 MNS with Metal Core

Although MNS with ferrite cores are promising MRI contrast agents, their saturation magnetization can be further improved since a portion of their magnetic spins cancels each other. MNS with metallic core such as Fe, Co, FeCo, FePt, and CoPt exhibit higher saturation magnetization since all their magnetic spins align in one direction and there is no canceled spin. Saturation magnetization of bulk FeCo and Fe is 240 and 218 emu/g, respectively compared to 90 emu/g of bulk Fe₃O₄ MNS. However, each metallic MNS has a specific limitation as MRI contrast agents. For example, FePt and CoPt are chemically stable but potentially toxic due to the possibility of leaching of Pt [153]. Fe, Co, and FeCo nanoparticles are biocompatible but are chemically unstable in air and prone to oxidation. To stabilize MNS with metallic core and maintain their magnetic properties, a number of coatings have been applied. MNS with Fe core and ferrite shell were reported that exhibited long-term stability in air [44–46]. Sun et al. reported bcc Fe/Fe₃O₄ core/shell MNS. Due to the high magnetization of Fe core, the r₂ relaxivity of Fe/Fe₃O₄ MNS (220 mM⁻¹ s⁻¹) was found ~10 times higher than r₂ of Fe₃O₄ MNS (24 mM⁻¹ s⁻¹) of the same size and coating [44, 56] and 2 times higher than the typical iron oxide NP contrast agent Feridex (110 mM⁻¹ s⁻¹). However, no in vitro or in vivo studies were reported. Weissleder et al. further improved the r₂ relaxivity up to 430 mM⁻¹ s⁻¹ by coating MnFe₂O₄ shell on Fe nanoparticles. The authors intravenously injected Fe/MnFe₂O₄ MNS, Fe₃O₄ MNS of the same size and cross-linked iron oxide (CLIO). Taken at 3 h postinjection, the images verified that Fe/MnFe₂O₄ MNS resulted in the most significant darkening compared to CLIO and Fe₃O₄ MNS [46]. Dai et al. developed FeCo nanoparticles embedded in graphitic carbon (GC) shell that were discrete, chemically functionalized, and water soluble as desired for biological applications. Due to the high saturation magnetization, FeCo/GC MNS exhibited very high r₂ values (644 mM⁻¹ s⁻¹). During in vivo intravascular MR imaging of the blood pool in the rabbit, mesenchymal stem cells labeled with FeCo/GC MNS showed significantly higher T₂ negative contrast enhancement compared to the ones labeled with Feridex [43, 52].

7 Thermally Activated MNS for Cancer Therapeutics

With the capability to generate thermal energy at targeted areas, MNS can be used in cancer therapeutics [5, 12, 154]. Compared with photodynamic therapy agents such as gold and graphene, MNS are advantageous for targets that reside deep inside the biological system without penetration depth problem. In addition, the fact that magnetic field causes no adverse effect on biological tissues serves as a distinctive benefit for noninvasive, in vivo applications.

The use of MNS in cancer therapeutic has been divided into three categories; (i) magnetic hyperthermia where MNS kill tumor cells via increase in tissue temperature; (ii) chemotherapy where MNS deliver a drug and trigger release at the tumor site; (iii) biotherapy where MNS are highly effective carrier platforms for bioactive molecules such as siRNA, oligos, and genes and facilitate transport of biomolecules in plasma membrane penetration necessary for cell internalization.

7.1 Magnetic Hyperthermia: MNS as Heat Generators

Targeted MNS can accumulate at the tumor site and increase the tissue temperature under an external RF field. Since the cancer tissues have higher heat sensitivity than normal tissues, thermal activation of MNS can be used to selectively kill tumor cells in the range of 41–47 °C [69]. As mentioned earlier, as MNS with high SAR have high efficacy for killing cancer cells, a variety of next-generation MNS with high SAR have been developed [155–157]. (Zn_0.4Mn_0.6)Fe₂O₄ MNS have shown high SAR value of 432 W/g which is ~4 times higher than SAR of Feridex (115 W/g). The high SAR of (Zn_0.4Mn_0.6)Fe₂O₄ MNS resulted in 84.4 % death of HeLa cancer cells in comparison to 13.5 % from Feridex (Fig. 12) [42]. Similarly, 82 % of Hela cells died when treated with CoFe₂O₄ MNS with SAR value of 238 W/g at 168 kHz [158]. Cheon et al. reported a significant increase in the SAR values of MNS by synthesizing core/shell MNS with hard ferromagnetic core and soft ferromagnetic shell [159]. Due to the exchange coupling between core and shell of MNS, the core/shell MNS showed SAR up to 3034 W/g which are an order of magnitude larger than SAR of conventional ferrite MNS (~300 W/g). Due to the high SAR values, the therapeutic efficacy of these MNS under RF was found superior to a common anticancer drug (doxorubicin). The same amounts (75 μg) of MNS and doxorubicin were injected into the tumor of a nude mice xenografted with cancer cells (U87MG). The tumor was clearly eliminated in 18 days in the group treated with the core/shell MNS, while in the doxorubicin-treated group tumor growth slowed initially, but then regrew after 18 days [159]. Recently, the same group reported Gd (III) texaphyrins (GdTx) conjugated ZnFe₂O₄ MNS with SAR of 471 W/g designed for hyperthermic treatment for apoptosis. This system is a double-effector MNS that generates heat as well as reactive oxygen species (ROS) which remarkably increased the degree of apoptotic cell death. Xenograft tumors in mice treated with the double-effector MNS were eliminated within 8 days whereas the tumor in untreated mice increased eightfold in 8 days [160].

Fig. 12.

SAR values and percentage of HeLa cells killed after treatment with (Zn_0.4Mn_0.6)Fe₂O₄ MNS or Feridex in AMF. Fluorescence microscopy images of HeLa cells treated with (Zn_0.4Mn_0.6) Fe₂O₄ nanoparticles (or Feridex) and stained with calcein show live cells as green fluorescence. (Zn_0.4Mn_0.6)Fe₂O₄ MNS have shown SAR value of 432 W/g, ~4 times higher than SAR of Feridex (115 W/g) which resulted in 84.4 % death of HeLa cancer cells in comparison to 13.5 % from Feridex. Reprinted with permission from Ref. [42] Copyright © 2009 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

7.2 Biotherapy: MNS as Carrier for Gene Therapeutics

Gene therapy is a technique that uses DNA and antisense RNA (siRNA) to treat and prevent disease via gene expression and gene silencing of defective genes [161, 162]. The coupling of nucleic acids with MNS improves the plasma pharmacokinetics and plasma membrane penetration of nucleic acids necessary for internalization into cells [162]. MNS designed for gene therapy have been coated with cationic polymers such as polyethylenimine (PEI), polyamidoamine, or chitosan in order to conjugate with negatively charged nucleic acids. While the cationic MNS have shown great success in vitro, their applicability in vivo has been limited because of toxicity and instability in biological media [163]. Zhang et al. coated MNS with a copolymer of PEI, PEG, and chitosan (NP-CP-PEI) [164]. The chitosan and PEG suppressed the PEI toxicity while PEG also provided the stability. NP-CP-PEI demonstrated an innocuous toxic profile and a high level of expression of the delivered plasmid DNA in a C6 xenograft mouse model, while MNS coated with only PEI or chitosan showed high toxicity or low gene transfection efficiency, respectively [164]. The attachment of the targeting ligand, chlorotoxin (CTX), to NP-CP-PEI enhanced the gene transfection efficiency. Histology analysis and confocal microscopy of the C6 xenograft tumor sections showed more cells expressing GFP in tumors treated with the NP-CP-PEI attached with CTX compared to NP-CP-PEI without CTX [165]. One alternative to the cationic coatings was offered by conjugating siRNA to MNS by covalent bonding. Medarova et al. developed a dual purpose probe for the simultaneous noninvasive imaging and delivery of siRNAs to tumors. This probe consisted of MNS labeled with Cy5.5 dye and conjugated to a synthetic siRNA duplex targeting a gene of interest. With use of model (green fluorescent protein, GFP) and therapeutic (surviving) genes, the authors demonstrated that the targeting and delivery of the probe could be monitored in vivo by MRI and optical imaging. In addition, they were able to follow the silencing process by optical imaging and to correlate it with histological data [86, 87].

7.3 Chemotherapy: MNS as Drug Carrier/Release Trigger for Chemotherapeutics

Chemotherapy focuses on the treatment of disease through delivery of small molecule drug formulations [88]. Most of the drugs do not have cell-targeting capabilities which results in undesirable side effects when internalized by healthy cells. The success of MNS in diagnostic imaging has generated considerable interest in their use as drug delivery vehicles. MNS coatings provide anchor points to which drug molecules can be coupled. Integrating the drugs into MNS improves their targeting abilities, limits their side effects, and allows increment of the drug dosage at the diseased tissue [5, 10]. Currently, several drugs have been combined with MNS for cancer chemotherapy, including paclitaxel (PTX), doxorubicin (DOX), and methotrexate (MTX) [89, 90]. The therapeutic moieties can be covalently bonded to MNS with cleavable linkages, encapsulated in the hydrophobic coating on MNS, or physically absorbed on the surface of MNS.

An ideal drug delivery vehicle should have efficient drug loading and controllable drug release. In applications where the drug (such as MTX) has an affinity for the target cell, it can be advantageous to graft the drug to the surface of the MNS. Kohler et al. demonstrated covalent attachment of MTX to the surface of a PEG-coated MNS via a cleavable amide linkage [89]. However, the drug loading capacity via direct conjugation was found low due to the small number of functional groups on the surface of MNS. Hollow MNS have been used to increase chemotherapeutic efficacy due to higher drug loading capacity [166]. Sun et al. utilized porous hollow Fe₃O₄ MNS with 5 times higher cisplatin loading compared to solid Fe₃O₄ MNS [167]. Once coupled with Herceptin to the surface, the cisplatin-loaded hollow NPs targeted breast cancer SK-BR-3 cells with IC₅₀ reaching 2.9 μM, much lower than 6.8 μM needed for free cisplatin. Labheshwar et al. coated a PEO-PPO diblock copolymer (Pluronic F127) on oleic acid coated MNS, where a hydro-phobic region in the oleic acid/PPO layer provided drug loading of 8.2 and 9.5 % for DOX and PTX, respectively [168]. They found that MNS loaded with both DOX and PTX in a 1:1 ratio demonstrated highly synergistic antiproliferative activity in MCF-7 breast cancer cells relative to MNS loaded with only DOX or PTX.

The thermal energy from MNS has been used as an external trigger for controlled drug release. Thomas et al. loaded mesoporous silica nanoparticles with DOX and 15 nm (Zn_0.4Fe_0.6)Fe₂O₄ MNS and capped the pores with cucurbit [6] uril that functioned as a heat labile molecular valve [169]. There was no drug release at the room temperature since the pores of the mesoporous silica remain capped, but under an external RF field, decapping occurred due to thermal activation of MNS, releasing most of the drug. In vitro, this controlled drug release killed 7-times more breast cancer cells. MNS coated with thermally responsive agents (e.g., hydrogels, thermosensitive polymers, lipids) have been explored where temperature works as a trigger for drug release [74, 170]. Recently, we reported poly(N-isopropylacrylamide) encapsulated Fe₃O₄ MNS in which DOX was loaded into the hydrogel via absorption. Drug release in presence of RF field was found 2 times higher than in absence of RF field due to thermal activation of MNS. In vitro localized drug delivery studies of the DOX loaded hydrogel-MNS composite with HeLa cell lines resulted in more than 80 % cell death under external RF field compared to 40 % cell death without RF field (Fig. 13) [83].

Fig. 13.

a Schematic illustration of drug (DOX) release from thermoresponsive hydrogel-MNS composite. b Percent DOX release and cell viability of HeLa cell lines treated with the hydrogel-MNS composite with and without external RF field [83]

8 Summary and Outlook

Magnetic nanostructures (MNS) truly represent a prototypical nanotechnology platform in the sense that their properties and phenomena are unquestionably size dependent in the nanoscale regime. Nominal ferrimagnetic behavior of MNS changes to superparamagnetism below ~10–15 nm size scale, which is essential for colloidal stability of MNS. The perturbations of protons in vicinity of the MNS provide the essential characteristics for contrast in MRI, while enthalpic contributions from external RF field generate localized thermal energy for therapeutic purpose. Thus, the combined theranostic attributes of MNS arise from diagnostic imaging and thermal therapy combination.

Over and beyond the technical and scientific aspects of theranostic administration of cancer, MNS also embody the other important attribute of nanotechnology in terms of complementarity, integration, and synergy of nominally disparate fields and subjects. For just MNS alone, these subjects and technical themes include: physics of magnetism, chemistry of synthesis, materials science of structure–property relationship, surface science of functionalization, biomedical engineering in MR imaging protocols and RF activation parameters, and the core biology and medical themes of cancer targeting, diagnostics imaging, and therapy. As a result, this has brought together scientists, engineers, and clinical practitioners from diverse backgrounds for more than a decade to advance biomedical sensing, diagnostics, and therapeutics.

As demonstrated by the examples highlighted in this chapter, remarkable advances have been made in the recent decade to harness the size, composition, and size-dependent properties of MNS for cancer diagnostics, diagnostic imaging, and localized therapy. MNS continue to exhibit realistic potential to address diagnostic imaging by MRI and localized therapy via thermal activation and/or timed therapeutic cargo release. In particular, it has been shown that appropriate choice of composition, size, and surface functionalization has the potential for synergistic combination of diagnostics MR imaging and thermally activated therapy.

Despite some promising results obtained so far, including in vivo animal studies, there are specific challenges for effective use of MNS in humans; the final objective for any cancer theranostic platform. Regulatory approval for use in humans will require further and extensive safety and toxicology studies. The composition, surface properties, drug loading, biodistribution, and pharmacokinetics are the diverse factors that may affect the toxicity of the MNS in a direct or indirect manner and need to be understood thoroughly. Limitations also exist for targeting efficiency, the lack of homogeneous MNS penetration, and inadequate delivery of therapeutics into the tumor volume.

Research is continuing in this regard; including development of new magnetic core materials with higher relaxivity and thermal activation properties, along with design of new coating materials to improve the pharmacokinetics, biodistribution, and biocompatibility. Success with MRI and progress over the past few years offer considerable prospects for eventual diagnostic and therapeutic translation of MNS technology. Indeed, several diagnostic clinical trials using MNS have been initiated over the past few years. The increasing trend toward in vivo studies in animals and subsequent escalation to clinical trials are expected to help translate MNS from the laboratory to the clinic.