Magnetic nanoparticles

for drug delivery

Controlled release of drugs from nanostructured functional materials,
especially nanoparticles (NPs), is attracting increasing attention
because of the opportunities in cancer therapy and the treatment of
other ailments. The potential of magnetic NPs stems from the intrinsic
properties of their magnetic cores combined with their drug loading
capability and the biochemical properties that can be bestowed on them
by means of a suitable coating. Here we review the problems and recent
advances in the development of magnetic NPs for drug delivery, focusing
particularly on the materials involved.

Manuel Arruebo, Rodrigo Fernández-Pacheco, M. Ricardo Ibarra, and Jesús Santamaría*

Nanoscience Institute of Aragon (INA), Pedro Cerbuna 12, University of Zaragoza, 50009 Zaragoza, Spain
*E-mail: [email protected]

NPs are submicron moieties (between 1 nm and 100 nm according coating that renders the particles biocompatible, stable, and may serve
to the usual definition, although there are examples of NPs as a support for biomolecules. Their magnetic properties enable these
several hundreds of nanometers in size) made of inorganic or particles to be used in numerous applications, belonging to one or
organic (e.g. polymeric) materials, which may or may not be more of the following groups:
biodegradable. Their importance relates to the fact that the (i) Magnetic contrast agents in magnetic resonance imaging (MRI)2;
characteristics of NPs are different from those of bulk materials of (ii) Hyperthermia agents, where the magnetic particles are heated
the same composition, which is mainly because of size effects, the selectively by application of an high frequency magnetic field.
magnetic and electronic properties, and the role played by surface (e.g. in thermal ablation/hyperthermia of tumors3); and
phenomena as the size is reduced. (iii) Magnetic vectors that can be directed by means of a magnetic
Preparation methods for NPs generally fall into the category of so- field gradient towards a certain location, such as in the case of the
called ‘bottom-up’ methods, where nanomaterials are fabricated from targeted drug delivery4.
atoms or molecules in a controlled manner that is thermodynamically The scientific community is seeking to exploit the intrinsic properties
regulated by means such as self-assembly1. Some biomedical of magnetic NPs to obtain medical breakthroughs in diagnosis, and
applications require core-shell magnetic NPs. They consist of a metal drug delivery. Perhaps the most promising applications relate to the
or metallic oxide core, encapsulated in an inorganic or a polymeric diagnosis and treatment of cancer.

22 JUNE 2007 | VOLUME 2 | NUMBER 3 ISSN:1748 0132 © Elsevier Ltd 2007

 Magnetic nanoparticles for drug delivery REVIEW

Even though, according to the American Cancer Society, cancer

deaths in the US have dropped for a second straight year, which is
attributed to the decrease in smoking rates and to earlier detection and
more effective treatment of tumors5, cancer is still one of the leading
causes of death in developed countries. Conventional treatments,
including surgery, radiation, chemotherapy, and biologic therapies
(immunotherapy) are limited by the accessibility to the tumor, the risk
of operating on a vital organ, the spread of cancer cells throughout the
body, and the lack of selectivity toward tumor cells. Immunotherapy is
still relatively recent, and is most likely to be applied to small tumors,
since its effectiveness seems to decrease for more advanced stages
of cancer. Multimodal therapy that uses radiotherapy, chemotherapy,
immunotherapy, and other forms of treatment in combination with
surgery provides a better chance of survival6.
The potential of drug delivery systems based on the use of nano-
and microparticles stems from significant advantages such as: (i) the
ability to target specific locations in the body; (ii) the reduction of the
quantity of drug needed to attain a particular concentration in the
vicinity of the target; and (iii) the reduction of the concentration of Fig. 1 Temporal evolution in the number of scientific papers published
involving drug delivery using NPs. (Source: ISI Web of Knowledge © The
the drug at nontarget sites7 minimizing severe side effects. All these Thomson Corporation. Search terms: ‘drug delivery’ and ‘nanoparticles’. Date
benefits justify the exponential growth in the number of publications of search: December 2006.)
dealing with NPs for drug delivery applications (Fig. 1).
NPs can act at the tissular or cellular level. The latter implies that pore cut-off size between 380 nm and 780 nm, although vasculature
they can be endocytosed or phagocytosed (i.e. by dendritic cells, organization may differ depending on the tumor type, its growth rate,
macrophages) resulting in internalization of the NP. In this process, the and microenvironment9. Apart from tumors, size-dependent removal
NP can reach beyond the cytoplasmatic membrane and, in some cases, of NPs is a common occurrence in healthy capillaries. Table 110 shows
also beyond the nuclear membrane (i.e. transfection applications). the morphological pore sizes contributing to diffusive permeability
Tumor targeting with magnetic NPs may use passive or active in the capillaries of the human vascular system. It can be seen that,
strategies. Passive targeting occurs as a result of extravasation of from the delivery point of view, there is practically no limitation as
the NPs at the diseased site (tumor) where the microvasculature is the diameters of typical NPs are well below that of the narrowest
hyperpermeable and leaky, a process aided by tumor-limited lymphatic capillaries. Instead, the main limitation concerns the residence time
drainage. Combined, these factors lead to the selective accumulation of NPs in the bloodstream. Thus, the use of conventional NPs for
of NPs in tumor tissue, a phenomenon known as enhanced permeation drug delivery by passive targeting would be limited to tumors in
and retention (EPR)8. The majority of solid tumors exhibit a vascular mononuclear phagocyte system (MPS) organs (liver, spleen, and bone

Table 1 Relevant sizes regarding particle distribution through and removal from the capillaries of the human vascular
system.

Size System/Organ

Particle removal* Tight-junction capillaries < 1 nm Central nervous system, blood-brain barrier
Continuous capillaries ~ 6 nm Tissues such as muscle, skin, and lung
Fenestrated capillaries ~ 50-60 nm Kidney, intestine, and some endocrine and exocrine glands
Sinusoidal capillaries ~ 100-1000 nm Liver, spleen, and bone marrow
Particle delivery Arteriole radius 0.005-0.07 mm Circulatory system. Particles supplied by intravenous
Artery radius 0.08-7.5 mm administration. Elimination involves opsonization and removal
Venule radius 8-100 µm by monocytes in blood
*It is noted that this table expresses only the morphological pores contributing to diffusive permeability. Actual transcapillary exchanges are
modified by vesicular transports, which are able to internalize particles with sizes up to 20-30 nm.
(Adapted with permission from10. © 1999 Elsevier.)

JUNE 2007 | VOLUME 2 | NUMBER 3 23

REVIEW Magnetic nanoparticles for drug delivery

marrow). Addressing other tumoral tissues does not seem feasible factor affecting the fate of NPs in passive targeting processes using the
without active targeting strategies because of the short circulation permeability of the capillary vessels, as discussed above.
times involved and the low concentration of NPs that is achieved in Here we shall concentrate on the therapeutic applications of
the tumor area (despite the EPR effect), leading to drug concentrations magnetic drug targeting using NPs. Hyperthermia/thermal ablation
below the therapeutic level11. will not be addressed, although the general implications between
Active targeting is based on the over or exclusive expression both therapies based on magnetism can be inferred. Applications in
of different epitopes or receptors in tumoral cells, and on specific diagnosis, where magnetic NPs are widely used as contrast agents, will
physical characteristics. Thus, vectors sensitive to physical stimuli (e.g. not be addressed here either.
temperature, pH, electric charge, light, sound, magnetism) have been
developed and conjugated to drugs. Alternatively, active targeting Magnetic drug delivery
may be based on over-expressed species such as low molecular weight The development of magnetic drug delivery
ligands (folic acid, thiamine, sugars), peptides (RGD, LHRD), proteins Any overview on drug delivery should start with the deserved
(transferrin, antibodies, lectins), polysaccharides (hyaluronic acid), recognition of Paul Ehrlich (1854-1915), who proposed that if an agent
polyunsaturated fatty acids, peptides, DNA, etc. could selectively target a disease-causing organism, then a toxin for
Different moieties including dendrimers, micelles, emulsions, that organism could be delivered along with the agent of selectivity.
nanoparticulated drugs, and liposomes are used to target specific Hence, a ‘magic bullet’ would be created able to kill the targeted
areas in the body (Fig. 2)12,13. The NPs must be endowed with the organism exclusively. Ehrlich received the 1908 Nobel Prize in Medicine
specific characteristics needed to reach a given target, which means for his work in the field of immunity, and the magic bullet idea was
attaining a suitable combination of nature, size, way of conjugating even used as a script for the 1940 movie Dr. Ehrlich’s Magic Bullet.
the drug to the NP (attached, adsorbed, encapsulated), surface Since then, various strategies have been proposed to deliver a drug
chemistry, hydrophilicity/hydrophobicity, surface functionalization, to the vicinity of a tumor including, as mentioned above, the use of
biodegradability, and physical response properties (temperature, pH, vectors sensitive to physical stimuli and tumor-recognition moieties
electric charge, light, sound, magnetism). Among these, size is the main conjugated to a drug.

Fig. 2 NP systems for drug delivery applications. (Adapted with permission from119 © 2005 Elsevier; and from120 © 2005 PharmaVentures Ltd.)

24 JUNE 2007 | VOLUME 2 | NUMBER 3

 Magnetic nanoparticles for drug delivery REVIEW

Prior to their use for drug delivery, magnetic microparticles were the drug or give rise to a local effect (irradiation from radioactive
proposed as contrast agents for localized radiation therapy14 and to microspheres or hyperthermia with magnetic NPs)33. Drug release
induce vascular occlusion of the tumors (antiangiogenic therapy)15,16. can proceed by simple diffusion or take place through mechanisms
Freeman et al.17 proposed in 1960 that magnetic NPs could be requiring enzymatic activity or changes in physiological conditions
transported through the vascular system and concentrated in a specific such as pH, osmolality, or temperature34; drug release can also be
part of the body with the aid of a magnetic field. magnetically triggered from the drug-conjugated magnetic NPs.
The use of magnetic micro- and NPs for the delivery of
chemotherapeutics has evolved since the 1970s. Zimmermann and Drug delivery with magnetic NPs
Pilwat18 in 1976 used magnetic erythrocytes for the delivery of Different organic materials (polymeric NPs, liposomes, micelles) have
cytotoxic drugs. Widder et al.19 described the targeting of magnetic been investigated as drug delivery nanovectors using passive targeting,
albumin microspheres encapsulating an anticancer drug (doxorubicin) active targeting with a recognition moiety (e.g. antibody), or active
in animal models. In the 1980s, several authors developed this targeting by a physical stimulus (e.g. magnetism in magnetoliposomes).
strategy to deliver different drugs using magnetic microcapsules and However, these organic systems still present limited chemical and
microspheres20-23. In 1994, Häfeli et al.24 prepared biodegradable mechanic stability, swelling, susceptibility to microbiological attack,
poly(lactic acid) microspheres that incorporated magnetite and the inadequate control over the drug release rate35, and high cost.
beta-emitter 90Y for targeted radiotherapy, and successfully applied Polymer NPs also suffer from the problem of high polydispersity.
them to subcutaneous tumors25. Synthesis produces particles with a broad size distribution and irregular
However, all these initial approaches were microsized. Magnetic branching, which could lead to heterogeneous pharmacological
NPs were used for the first time in animal models by Lübbe et al.26. In properties. One alternative is to use dendrimers, which have a
1996, the first Phase I clinical trial was carried out by the same group monodisperse character and globular architecture resulting from
in patients with advanced and unsuccessfully pretreated cancers using their stepwise synthesis and can be purified at each step of growth36.
magnetic NPs loaded with epirubicin27. However, in that first trial, Visualization of dendrimers requires tagging with a specific moiety
more than 50% of the NPs ended up in the liver. (i.e. a fluorophore or metal). A major drawback of dendrimers and
Since then, a number of groups around the world have synthesized dendritic polymers, however, is their high cost. The preparation
magnetic vectors and shown potential applications. Different start-ups of dendritic polymers that circulate in the blood long enough to
now manufacture magnetic micro- and NPs, which are used in MRI, accumulate at target sites but that can also be removed from the body
magnetic fluid hyperthermia, cell sorting and targeting, bioseparation, at a reasonable rate to avoid long-term accumulation also remains a
sensing, enzyme immobilization, immunoassays, and gene transfection challenge.
and detection systems. Passive targeting using drug-conjugated dendrimers and dendritic
FeRx, Inc. (founded in 1997) produced doxorubicin-loaded polymers has been widely studied, mainly using the EPR effect.
magnetic NPs consisting of metallic Fe ground together with activated Therapies based on active targeting, such as antibody-conjugated
carbon28,29. A Phase II clinical study in patients with primary liver dendrimers, constitute a promising alternative in view of the potential
cancer was conducted using these NPs, although the trial was not of antibodies for selective targeting37. Because of the disadvantages
successful. Chemicell GmbH currently commercializes TargetMAG- of organic NPs for drug delivery, inorganic vectors constitute an
doxorubicin NPs involving a multidomain magnetite core and a interesting option and are the subject of intense research. Some
cross-linked starch matrix with terminal cations that can be reversibly examples of inorganic magnetic NPs will be given below.
exchanged by the positively charged doxorubicin. The particles have The main advantages of magnetic (organic or inorganic) NPs are
a hydrodynamic diameter of 50 nm and are coated with 3 mg/ml that they can be: (i) visualized (superparamagnetic NPs are used in
doxorubicin30. These NPs loaded with mitoxantrone have already MRI); (ii) guided or held in place by means of a magnetic field; and
been used in animal models with successful results31. Chemicell also (iii) heated in a magnetic field to trigger drug release or to produce
commercializes FluidMAG® for drug delivery applications. Magnetic hyperthermia/ablation of tissue. It is important to point out that the
NP hydro-gel (MagNaGel®) from Alnis Biosciences, Inc. is a material latter capability is not restricted to magnetic NPs, but also to other
comprising chemotherapeutic agents, Fe oxide colloids, and targeting particles capable of absorbing near-infrared, microwave, and ultrasound
ligands32. radiation.
In summary, for magnetic targeting, a drug or therapeutic Depending on the synthesis procedure, magnetic NPs or
radionuclide is bound to a magnetic compound, introduced in the body, nanocapsules can be obtained. We refer to NPs when the drug is
and then concentrated in the target area by means of a magnetic field covalently attached to the surface or entrapped or adsorbed within
(using an internally implanted permanent magnet or an externally the pores of the magnetic carrier (polymer, mesoporous silica, etc.).
applied field). Depending on the application, the particles then release Nanocapsules (‘reservoirs’) designate magnetic vesicular systems where

JUNE 2007 | VOLUME 2 | NUMBER 3 25

REVIEW Magnetic nanoparticles for drug delivery

Magnetic nanoparticles

Therapy Diagnosis

Hyperthermia/ In vivo In vitro

Drug
thermal
delivery
ablation Sensing
MRI
Radiotherapy Musculoskeletal
combined with system associated Cell sorting
MRI diseases Bioseparation
Anemic chronic * Enzyme
kidney disease immobilization

Immunoassays

Transfection

Purification

Fig. 3 Biomedical applications of magnetic NPs.

*Ferumoxytol® (Advanced Magnetics, Inc.) is in Phase III multicenter clinical studies for use as an intravenous Fe replacement therapeutic for patients with anemic
chronic kidney disease.

the drug is confined to an aqueous or oily cavity, usually prepared by area or which triggers the drug desorption. Permanent Nd-Fe-B
the reverse micelle procedure, and surrounded by an organic membrane magnets in combination with SPION, which have excellent magnetic
(magnetoliposomes) or encapsulated within a hollow inorganic properties, can reach effective magnetic field depths up to 10-15 cm
capsule35. in the body39. However, it must be noted that the magnetic carriers
The key parameters in the behavior of magnetic NPs are related to accumulate not only at the desired site but also throughout the cross-
surface chemistry, size (magnetic core, hydrodynamic volume, and size section from the external source to the depth marking the effective
distribution), and magnetic properties (magnetic moment, remanence, field limit. Obviously, the geometry of the magnetic field is extremely
coercivity). The surface chemistry is especially important to avoid the important and must be taken into account when designing a magnetic
action of the reticuloendothelial system (RES), which is part of the targeting process.
immune system, and increase the half-life in the blood stream. Coating As a means to elude the limitations of using external magnetic
the NPs with a neutral and hydrophilic compound (i.e. polyethylene fields, internal magnets can be located in the vicinity of the target by
glycol (PEG), polysaccharides, dysopsonins (HSA), etc.) increases the using minimally invasive surgery40. Several studies have simulated the
circulatory half-life from minutes to hours or days. Another possibility interaction between a magnetic implant and magnetic NPs, enabling
is to reduce the particle size; however, despite all efforts, complete drug delivery6,41,42. In addition, work in several laboratories40,43,44 is
evasion of the RES does not seem feasible38 and unwanted migration addressing targeted drug delivery with magnetic implants.
to other areas in the body could cause toxicological problems. Another limitation relates to the small size of NPs, a requisite
In addition to cancer treatment, magnetic NPs can also be used for superparamagnetism, which is in turn needed to avoid magnetic
in anemic chronic kidney disease and disorders associated with agglomeration once the magnetic field is removed (see below). A small
the musculoskeletal system (i.e. local inflammatory processes, side size implies a magnetic response of reduced strength, making it difficult
effects) (Fig. 3). For those disorders, superparamagnetic Fe oxide NPs to direct particles and keep them in the proximity of the target while
(SPION), in conjunction with external magnetic fields, seem a suitable withstanding the drag of blood flow45. Targeting is likely to be more
alternative for drug delivery to inflammatory sites by maintaining effective in regions of slower blood velocity, and particularly when the
appropriate local concentrations while reducing overall dosage and side magnetic field source is close to the target site.
effects39. As for all biomedical applications, limitations also arise in
extrapolating from animal models to humans. There are many
Limitations of magnetic drug delivery physiological parameters to consider, ranging from differences in
Since the magnetic gradient decreases with the distance to the target, weight, blood volume, cardiac output, and circulation time to tumor
the main limitation of magnetic drug delivery relates to the strength volume/location/blood flow, complicating the extrapolation of data
of the external field that can be applied to obtain the necessary obtained in animal models46,47. Related to this point is the fact
magnetic gradient to control the residence time of NPs in the desired that studies on toxicity (not only direct toxicity, but also toxicity

26 JUNE 2007 | VOLUME 2 | NUMBER 3

 Magnetic nanoparticles for drug delivery REVIEW

of the degradation products and induced responses48) and the fate magnetic field, the magnetic moment of entire crystallites aligns with
of magnetic carriers are insufficient and, in many cases, there is the magnetic field (Fig. 4).
insufficient characterization. In large NPs, energetic considerations favor the formation of
Finally, state-of-the-art magnetic drug delivery seems mainly domain walls. However, when the particle size decreases below a
applicable to well-defined tumors, as treatment of metastatic certain value, the formation of domain walls becomes unfavorable
neoplasms and small tumors in the early stages of their growth and each particle comprises a single domain. This is the case for
still remains a challenge. Treating emerging tumors will involve superparamagnetic NPs. Superparamagnetism in drug delivery is
the development of a new generation of seek-and-destroy NPs, necessary because once the external magnetic field is removed,
which specifically recognize small clusters of cancer cells and carry magnetization disappears (negligible remanence and coercivity, see
the necessary elements (drugs or hyperthermia agents) for their Fig. 4), and thus agglomeration (and the possible embolization of
destruction. A strong interest continues in this field given the capability capillary vessels) is avoided.
of NPs to access tumors in regions where conventional surgery cannot Another key requirement is the biodegradability or intact excretion
be applied. of the magnetic core. Thus, SPION are considered to be biodegradable
with Fe being reused/recycled by cells using normal biochemical
Tailoring magnetic NPs pathways for Fe metabolism50,51. For nonbiodegradable cores, a specific
Essential requisites coating is needed to avoid exposure (and possible leaching) of the
Magnetic NPs for biomedical applications must be endowed with magnetic core and to facilitate intact excretion through the kidneys,
the specific characteristics required. As mentioned above, the first so that the half-life of the agent in the blood is determined by the
requirement is often superparamagnetism. glomerular filtration rate (e.g. contrast agents based on gadolinium)51.
Superparamagnetism occurs in magnetic materials composed of Coatings on magnetic NPs
very small crystallites (threshold size depends on the nature of the The coatings on magnetic NPs often serve multiple purposes. Their
material, for instance, Fe-based NPs become superparamagnetic at role in reducing leaching of the cores has already been mentioned. The
sizes <25 nm49). In a paramagnetic material, the thermal energy coating also often facilitates the stabilization of NPs in an environment
overcomes the coupling forces between neighboring atoms above the with a slightly alkaline pH or a significant salt concentration. For
Curie temperature, causing random fluctuations in the magnetization instance, the isoelectrical point of SiO2 is reached at pH 2-3, meaning
direction that result in a null overall magnetic moment. However, that silica-coated NPs are negatively charged at the pH of blood,
in superparamagnetic materials, the fluctuations affect the direction inducing electrostatic repulsion that helps avoid aggregate formation.
of magnetization of entire crystallites. The magnetic moments of Silica coatings also have additional advantages. On the one hand,
individual crystallites compensate for each other and the overall the external surface of silica coatings can be functionalized to allow
magnetic moment becomes null. When an external magnetic field is the binding of biomolecules. This is mainly related to the presence
applied, the behavior is similar to paramagnetism except that, instead of hydroxyl surface groups in significant concentrations that provide
of each individual atom being independently influenced by an external intrinsic hydrophilicity and allow surface attachment by covalent

Fig. 4 Hysteresis loops (magnetization versus applied magnetic field) characteristic of ferromagnetic and superparamagnetic NPs. For comparison, para- and
diamagnetic behavior are also shown. The figure also indicates the values of the remanence, Mr, and coercive field, Hc.

JUNE 2007 | VOLUME 2 | NUMBER 3 27

REVIEW Magnetic nanoparticles for drug delivery

linkage of specific biomolecules. On the other hand, the internal

porosity of silica can be used to host a specific drug, a feat achievable (a) (b)
while avoiding the unwanted physical adsorption of larger molecules.
Thus, according to Ambrose and Fritz52, the smallest major protein in
serum is serum albumin, with a molecular mass of approximately
66 000 Da and an effective spherical radius of ~40 Å. This means that
serum albumin, and larger molecules, would be excluded from the
channels of microporous and mesoporous (with pore sizes <4 nm)
coatings (e.g. silica, carbon, zeolites) on magnetic cores. Finally, silica
and other microporous inorganic materials are heat resistant, with (c) (d)
high surface areas and good mechanical strength. Figs. 5 and 6 show
different types of inorganic coatings on metal cores developed in our
laboratory.
In addition, coatings play an essential role in retarding clearance
by the RES. Depending on their size, surface functionalization, and
hydrophilicity, a rapid uptake of uncoated NPs by the mononuclear
phagocyte system (MPS) is likely after systemic administration, Fig. 5. High resolution transmission electron microscopy (HRTEM) images of
followed by clearance to the liver, spleen, and bone marrow. Different (a) magnetite NP encapsulated in silica; (b) magnetite NPs embedded in a
zeolitic (aluminosilicate) matrix; (c) Fe NPs encapsulated in silica (energy-
proteins (antibodies) of the blood serum (opsonins) bind to the surface
filtering TEM, EFTEM, color map); (d) magnetite NP encapsulated in graphite.
of foreign bodies, accelerating phagocitation of the particles. To avoid
this, biodegradable (e.g. dextran) and nonbiodegradable organic and
(a)
inorganic coatings can be used as a means to retard detection and
uptake by the macrophages of the RES. Perhaps the most widely used
coating for this purpose is PEG, a linear neutral polyether, whose
attachment to NP surfaces provides a ‘stealth’ shielding effect, delaying
the action of the RES53. PEG shows little toxicity and immunogenicity,
and intact excretion is possible, either via the kidneys (for PEG <30
kDa) or in the feces (for PEG >20 kDa)54. Avoiding detection by the
RES stems from the protein-resistant character of PEGylated surfaces,
which has been ascribed to the combination of a low interfacial
(b)
energy in water and the steric stabilization effect55. Unfortunately, the
‘immunostealthing’ function provided by PEG is frequently concurrent
with the loss of biomolecular targeting capabilities; therefore, it is
necessary to optimize the coating.
The nature of the coating is also important in those cases where the
surface functionalization might cause hydrogen bonding and, therefore,
agglomeration of NPs. For such situations, surface modification can
reduce the aggregation and increase the stability of NPs in body fluids.
A wide variety of molecules has been loaded onto organic
and inorganic shells, e.g. by chemical functionalization or physical
absorption. The list includes tumor-recognition moieties such as Fig. 6 HRTEM photographs of: (a) magnetite NPs encapsulated in a silica
antibodies in ‘smart’ contrast agents56-58 and cell-penetrating matrix; and (b) EFTEM color map of Au NPs encapsulated in a silica matrix.
peptides for MRI applications59,60; and enzymes61, toxins62, genes63-68
(transfection), growth factors69,70, radionucleotides4,71-73, folic acid74, Fate
and drugs (mitoxantrone4,75, tamoxifen76, cefradine77, doxorubicin78-83, The distribution of NPs and their loads throughout the body depends
ammonium glycyrrhizinate84, fludarabine85, danorubicin86, cisplatin on numerous physicochemical factors: size of particles, toxicity, surface
and gemcitabine87, pingyangmycin88, nonsteroidal anti-inflammatory charge, capacity for protein adsorption, surface hydrophobicity, drug
drugs89, amethopterin90,91, mitomycin92, paclitaxel21,93, diclofenac loading and release kinetics, stability, degeneration of carrier systems,
sodium94,95, and adriamycin96) for drug delivery applications. hydration behavior, electrophoretic mobility, porosity, specific surface

28 JUNE 2007 | VOLUME 2 | NUMBER 3

 Magnetic nanoparticles for drug delivery REVIEW

characteristics, density, crystallinity, contact angle, and molecular spleen (3-10%). Particles larger than 200 nm are usually filtered to the
weight39. Nevertheless, the fate (and also the possible toxicity) of spleen, whose cut-off point extends up to 250 nm104, while particles up
magnetic NPs also depends strongly on the dose and administration to 100 nm are mainly phagocytosed through liver cells. In general, the
route (oral or parenteral: including delivery routes such as intravenous, larger the particles are, the shorter their plasma half-life-period101. This
pulmonary, transdermal, and ocular, in addition to less conventional clearance of particles by Kupffer cells can be, on the other hand, useful
routes, e.g. when used as scaffold coatings). Some of the implications for the treatment of liver diseases, such as cancer or leishmania105,
for the three most common ways of administration are now discussed. tuberculosis, listeria, leprosy, etc.; although it is important to consider
that it would simultaneously entail the depletion of a significant
Intravenous administration number of the patient’s own defense cells.
The general rule for magnetic NPs in parenteral applications it that
the carrier is nontoxic, nonimmunogenic, and of a size that avoids Subcutaneous or intratumoral administration
embolization of capillary ducts. Once a NP enters the bloodstream, Unlike water-soluble molecules, which are rapidly absorbed through
opsonization processes activate the RES system response. Circulating the blood capillary walls and pass into the circulatory system, small
mononuclear phagocytes (monocytes) clear the NPs to the liver, particles injected locally infiltrate into the interstitial spaces around
spleen, and bone marrow where resident cells (e.g. Kupffer cells in the the injection site and are gradually absorbed by the lymphatic capillary
liver) capture the NPs prior to degradation, if possible. Depending on system11. For this reason, subcutaneously or locally injected NPs
biodegradability and size, some of the NPs present in the lysosomal can be used for lymphatic targeting, i.e. as a tool for chemotherapy
vesicles of Kupffer cells may be incorporated into the bile and be for lymphatic tumors, although this route is rarely used in clinical
removed in the feces. Other NPs will be filtered by the kidneys and practice as it is not useful for targeting metastatic tumors. In this case,
incorporated into the urine. In general, smaller NPs are subject to rapid magnetism is not required but may be a useful property to hold the
renal elimination, while larger ones show uptake by the liver, spleen, particles in place or to enable hyperthermia. As a general rule, colloidal
and bone marrow (Fig. 7)97-103. Large particles will be removed by cells carriers aimed at regional lymph nodes through subcutaneous injection
capable of phagocytosis (i.e. by macrophages or dendritic cells), while need to be small (60 nm or less)106. Another limitation relates to
small particles can be removed by cells capable of endocytosis (i.e. by intratumoral pressure gradients caused by the fast cell proliferation in
B and T lymphocytes). Finally, if the magnetic NPs are biodegradable, solid tumors.
the decomposition products can be taken up by any cell by means of
pinocytosis. Oral administration
According to Neuberger et al.39, magnetic particles smaller than Several publications describe the use of magnetic NPs coated with an
4 µm are removed by cells of the RES, mainly in the liver (60-90%) and organic shell as oral drug delivery vectors107, although most concern

Monocrystalline
iron oxide nanoparticles
< 10 nm 101-103
Blood residence time

Ultrasmall superparamagnetic
iron oxide nanoparticles
5-20 nm 100
Ferromagnetic
particles
200 nm 98 Non-biodegradable
inorganic and polymeric
97
micrometric spheres

Large superparamagnetic
iron oxide nanoparticles
> 40 nm 99

Plasma half life = 2 h Removal by liver (80%) Opsonization.

Accumulation in lymph and spleen (15%) Removal by liver, spleen,
nodes. Excreted by urine lungs, and bone marrow
and feces macrophages

Particle size

Fig. 7 Qualitative diagram showing the evolution of blood residence time with particle size.

JUNE 2007 | VOLUME 2 | NUMBER 3 29

REVIEW Magnetic nanoparticles for drug delivery

magnetic NPs used as MRI contrast agents for the gastrointestinal It is well documented that the large surface-to-volume ratio of
tract. The main problem here is that oral delivery of peptides and all nanosized particles can potentially lead to unfavorable biological
proteins is hampered by their degradation in gastrointestinal acid, responses if they are inhaled and subsequently absorbed via the lung
low absorption, first-pass metabolism by the liver, and a significant or swallowed and then absorbed across the gastrointestinal tract115.
initial increase in drug concentration. Feng et al.108 describe the fate Interestingly, it has also been reported that, in 20-100 mg/ml
of chemotherapeutical NPs in oral delivery: particles under 5 µm can concentrations, large magnetic particles show higher cytotoxicity
be removed via lymphatic drainage, particles up to 500 nm can cross than smaller ones even after normalizing for surface area116 despite
the membrane of epithelial cells through endocytosis, and particles the lower surface-to-volume ratio, although it is difficult to perform
less than 50 nm can achieve paracellular passage between intestinal comparable experiments with differently sized particles. In any case,
epithelial cells. toxicity studies should consider not only acute toxicity but also that of
degradation products, the possible stimulation of cells with subsequent
Safety: influence of the magnetic field release of inflammatory mediators39, and long term toxicity.
Although all the components of the body are either dia-, para-, For a magnetic carrier with potential as a drug delivery vector, it is
superpara-, ferri-, or ferromagnetic, the magnetic fields required to necessary, at the very least, to analyze its: (i) toxicity (acute, subacute,
produce obvious effect in the body are very large. Even red blood cells, and chronic toxicity, teratogenicity, and mutagenicity) in cellular
which each contain micrograms of the Fe protein hemoglobin, show a and animal models; (ii) hematocompatibility; (iii) biodegradation
relatively low response to large fields or steep field gradients, although (whenever possible); (iv) immunogenicity; and (v) pharmacokinetics
this low value is enough to be used in functional MRI (fMRI). The other (body distribution, metabolism, bioavailability, elimination, organ-
natural Fe-containing compounds in the body are hemosiderin, ferritin, specific toxicity) before the start of preclinical testing. The contrast
transferrin, and the cytochromes. agents based on magnetic NPs currently on the market, either
According to Schenck109, the US Food and Drug Administration involving superparamagnetic Fe oxides (i.e. Feridex®, Endorem™,
(FDA) initially considered applications for approval to market MRI GastroMARK®, Lumirem®, Sinerem®, or Resovist®) or paramagnetic
scanners on a case-by-case evaluation of the safety and efficacy metals encapsulated in a chelating agent (i.e. Magnevist®, Dotarem®,
information provided. Based on positive clinical and safety experience, Gadovist®, Teslacan®) have satisfied current regulations regarding use
the FDA classified magnets with field strength of less than 2 T as in patients. The same applies to the magnetic drug delivery systems
nonsignificant risk devices in 1987. Further positive experiences led the that have already been commercialized, MagNaGel®32 and FluidMAG®
FDA to increase this threshold to 4 T in 1996 and again in 2003 (for and TargetMAG® (Table 2).
adults) to 8 T. Even though experiments with strong static magnetic Recently, the US Environmental Protection Agency has started
fields (8 T) have been shown to reduce the flow rate of human blood to regulate a large class of products made with Ag NPs using the
by 30% in in vitro tests110 and it has been reported that magnetic fields legislation developed for pesticides. This is the first federal restriction
above 3 T might affect the normal behavior of erythrocytes, recent to focus largely on nanotechnology117.
studies evaluating human subjects for adverse effects in physiological
or neurocognitive functions resulting from exposure to static magnetic Perspectives and future challenges
fields (up to 8 T) from MRI systems have not shown any clinically In therapy, we are witnessing the early use of magnetic NPs as drug
relevant effects111. According to Kangarlu and Robitaille112, within the delivery vectors and as tools for hyperthermia/thermal ablation.
decade, human imaging at fields in excess of 10 T will most probably Magnetic drug delivery constitutes a promising technology to treat
be achieved and such projects are now being planned. cancer, and several products are already on the market. The limitations
inherent in the use of external magnetic fields can, in some cases, be
Toxicity circumvented by means of internal magnets located in the proximity
When discussing the toxicity of NPs, generalization becomes difficult of the target by minimally invasive surgery40,43,44,98,118. Magnetic fluid
because their toxicity depends on numerous factors including the dose, hyperthermia/thermal ablation is also promising and is currently being
chemical composition, method of administration, size, biodegradability, applied (i.e. MagForce Nanotecnologies AG), but is limited by the fact
solubility, pharmacokinetics, biodistribution, surface chemistry, that the tumor needs to be localized. This route, therefore, cannot be
shape, and structure, to name but a few. With NPs, as with any new used in preventive medicine, or for treating early-stage tumors.
biomedical discovery, the risk-benefit trade-off must be considered to The greatest therapeutic potential is probably associated with
assess whether the risks can be justified. In general, the size, surface applications involving ‘intelligent’ particles with a magnetic core
area, shape, composition, and coating of an NP are the most important (to direct the particles to the vicinity of the target and also for
characteristics regarding cytotoxicity113, and modifications of the NP hyperthermia or for temperature-enhanced release of the drug), a
surface are a key tool to minimize toxicological effects114. recognition layer (to which suitable receptors are attached), and a

30 JUNE 2007 | VOLUME 2 | NUMBER 3

 Magnetic nanoparticles for drug delivery REVIEW

Table 2 Some companies involved in the development and production of magnetic micro- and nanoparticles.

Company Application Website

Bangs Laboratories, Inc. Cell separation, DNA and RNA purification, immunoassays www.bangslabs.com
Polysciences, Inc. Magnetic separation, cell sorting, nucleic acid purification, flow cytometry, www.polysciences.com
calibration, immunoassay, fluorescent microscopy, diagnostic assays
Micromod Drug delivery, biomagnetic separation, nucleic acid purification www.micromod.de
Partikeltechnologie GmbH
Guerbet SA X-ray and MRI contrast agents www.guerbet.com
Ademtech SA Cell sorting, biomagnetic separation www.ademtech.com
Advanced Magnetics, Inc. Treatment of anemia, MRI contrast agents www.advancedmagnetics.com
Invitrogen Corp. Immunoassay and nucleic acid in vitro diagnostics, DNA and RNA www.invitrogen.com
isolation, protein purification, cell separation and expansion, food and
environmental testing
Estapor (Merck & Co. Inc.) Chemiluminescent and radio-immunoassays, cell separation, protein www.estapor.com
purification, immuno-precipitation, bacteria detection, immuno- www.merck.com
chromatographic assays
MagForce Hyperthermia www.magforce.com
Nanotecnologies AG
Polymicrospheres (division Drug delivery, diagnostic assays www.polymicrospheres.com
of Vasmo, Inc.)
Spherotech, Inc. Cell separation, enzyme immunoassay www.spherotech.com
Alnis Biosciences, Inc. Drug delivery for anti-cancer and -infective treatments MagNaGel®105 www.alnis.com
and smart contrast agents
Triton Biosystems, Inc. Hyperthermia www.tritonbiosystems.com
Sirtex Medical Ltd. Radiation therapy www.sirtex.com
Biophan Technologies, Inc. Drug delivery www.biophan.com
Magnamedics GmbH Drug delivery, in vitro diagnostics www.magnamedics.com
Chemicell GmbH Drug delivery (FluidMAG®), bioseparation, gene transfection and detection www.chemicell.com

therapeutic load (adsorbed inside the pores or hosted within internal sensing, enzyme immobilization, immunoassays, and gene transfection/
cavities of the particles). The challenges are formidable, especially detection systems.
those related to the development of suitable recognition layers. Not
only must useful recognition moieties be identified and attached to the Acknowledgments
particles, but they must be loaded to a high density while maintaining
Support from the Spanish Nanoscience Action NAN200409270-C3-1/2 and
their desired characteristics. from the Consolider CSD2006-00012 and Ciber Ingenio 2010 CB06/01/0026
Finally, as already noted, the biomedical uses of magnetic NPs programs, is gratefully acknowledged. MA acknowledges support of a contract
are not restricted to drug delivery, and indeed new applications for from the Juan de la Cierva program (project PPQ2003-04986). The authors
also gratefully acknowledge the Serveis Cientificotècnics of the University
magnetic NPs are also likely in MRI, where contrast agents could be of Barcelona for the use of the TEM, and Jordi Arbiol i Cobos for the TEM
tagged with a recognition moiety, cell sorting/targeting, bioseparation, photographs and helpful discussions.

REFERENCES 7. Ritter, J. A., et al., J. Magn. Magn. Mater. (2004) 280, 184
1. Ferrari, M., Nat. Rev. Cancer (2005) 5, 161 8. Bogdanov, A., et al., J. Drug Targeting (1997) 4, 321
2. Cunningham, C. H., et al., Magn. Reson. Med. (2005) 53, 999 9. Hobbs, S. K., et al., Proc. Natl. Acad. Sci., USA (1998) 95, 4607
3. Johannsen, M., et al., Int. J. Hyperthermia (2005) 21, 637 10. Okuhata, Y., Adv. Drug Delivery Rev. (1999) 37, 121
4. Jurgons, R., et al., J. Phys.: Condens. Matter (2006) 18, S2893 11. Brigger, I., et al., Adv. Drug Delivery Rev. (2002) 54, 631
5. Jemal, A., CA: Cancer J. Clin. (2007) 57, 43 12. Tomalia, D. A., Prog. Polym. Sci. (2005) 30, 294
6. Cancer Reference Information, American Cancer Society, Inc. (2006) 13. Boyd, B. J., Drug delivery report Autumn/Winter, PharmaVentures Ltd., Oxford,
UK (2005)

JUNE 2007 | VOLUME 2 | NUMBER 3 31

REVIEW Magnetic nanoparticles for drug delivery

14. Meyers, P. H., et al., Am. J. Roentgenol. Radium Ther. Nucl. Med. (1963) 90, 68. Scherer, F., et al., Gene Therapy (2002) 9, 102
1068 69. Tanaka, H., et al., J. Biomed. Mater. Res. (2005) 73A, 255
15. Alksne, J. F., and Fingerhut, A. G., Bull. LA Neurol. Soc. (1965) 30,153 70. Kullberg, M. et al., Med. Hyp. (2005) 64, 468
16. Turner, R. D., et al., J. Urology (1975) 113, 455 71. Cao, J., et al., J. Magn. Magn. Mater. (2004) 277, 165
17. Freeman, M. W., et al., J. Appl. Phys. (1960) 31, S404 72. Liang, S., et al., J. Radioanal. Nucl. Chem. (2006) 269, 3
18. Zimmermann, U., and Pilwat, G., J. Biosci. (1976) 31, 732 73. Chen, J., et al., Cancer Lett. (2006) 231, 169
19. Widder, K. J., et al., Proc. Soc. Exp. Biol. Med.(1978) 158, 141 74. Kohler, N., et al., J. Am. Chem. Soc. (2004) 126, 7206
20. Kato, T., et al., Appl. Biochem. Biotechnol. (1984) 10, 199 75. Alexiou, C., et al., J. Magn. Magn. Mater. (2005) 293, 389
21. Widder, K. J., et al., Proc. Natl. Acad. Sci., USA (1981) 78, 579 76. Hu, F. X., et al., Biomaterials (2006) 27, 5725
22. Gupta, P. K., et al., J. Pharmacol. Sci. (1989) 78, 290 77. Zhang, Y. Q., et al., J. Nanosci. Nanotechnol. (2006) 6, 3210
23. Morimoto, Y., et al., J. Pharmacobio-Dynamics (1980) 3, 264 78. Arruebo, M., et al., Nanotechnology (2006) 17, 4057
24. Häfeli, U. O., et al., J. Biomed. Mater. Res. (1994) 28, 901 79. Ma, Y., et al., J. Mater. Eng. Perform. (2006) 15, 376
25. Häfeli, U. O., et al., Nucl. Med. Biol. (1995) 22, 147 80. Ma, Y., et al., J. Biomater. Sci.-Polym. Ed. (2004) 15, 1033
26. Lübbe, A. S., et al., Cancer Res. (1996) 56, 4694 81. Nobuto, H., et al., Int. J. Cancer (2004) 109, 627
27. Lübbe, A. S., et al., Cancer Res. (1996) 56, 4686 82. Leakakos, T., et al., Cancer Chemother. Pharmacol. (2003) 51, 445
28. Goodwin, S., et al., J. Magn. Magn. Mater. (1999) 194, 132 83. Rudge, S., et al., J. Controlled Release (2001) 74, 335
29. Goodwin, S. C., et al., Toxicol. Sci. (2001) 60, 177 84. Wu, Y., et al., Polymer (2006) 47, 5287
30. Steinfeld, U., et al., Int. J. Pharm. (2006) 311, 229 85. Port, R. E., et al., Cancer Chemother. Pharmacol. (2006) 58, 607
31. Wiekhorst, F., et al., J. Nanosci. Nanotechnol. (2006) 6, 3222 86. Zhang, R.,et al., Nanotechnology (2006) 17, 3622
32. Sunderland, C. J., et al., Drug Develop. Res. (2006) 67, 70 87. Yang, J., et al., J. Microencapsulation (2006) 23, 203
33. Häfeli, U. O., Int. J. Pharm. (2004) 277, 19 88. Adriane, K., et al., J. Drug Targeting (2006) 14, 243
34. Alexiou, C., et al., Cancer Res. (2000) 60, 6641 89. Taepaiboon, P., et al., Nanotechnology (2006) 17, 2317
35. Arruebo, M., et al., Chem. Mater. (2006) 18, 1911 90. Kohler, N., et al., Small (2006) 2, 785
36. Fréchet, J. M. J., Prog. Polym. Sci. (2005) 30, 844 91. Kohler, N., et al., Langmuir (2005) 21, 8858.
37. Duncan, R., Materials Today (2005) 8 (7-8, Suppl. 1), 16 92. Cheung, R.Y., et al., Biomaterials (2005) 26, 5375
38. Gaur, U., et al., Int. J. Pharmacol. (2000) 202, 1 93. Zhang, J.Q., et al., Pharm. Res. (2005) 22, 573
39. Neuberger, T., et al., J. Magn. Magn. Mater. (2005) 293, 483 94. Saravanan, M., et al., Int. J. Pharm. (2004) 283, 71
40. Ibarra, M. R., et al., Spanish Patent Application 200302803, 2006 95. Jain, S., et al., Int. J. Pharm. (2003) 261, 43
41. Rotariu, O., and Strachan N. J. C., J. Magn. Magn. Mater. (2005) 293, 639 96. Kubo, T., et al., Int. J. Oncology (2001) 18, 121
42. Iacob, G., et al., Biorheology (2004) 41, 599 97. Vonarbourg, A., et al., Biomaterials (2006) 27, 4356
43. Yellen, B. B., et al., J. Magn. Magn. Mater. (2005) 293, 647 98. Fernández-Pacheco, R., et al., J. Magn. Magn. Mater. (2006) 311, 318
44. Rosengart, A. J., et al., J. Magn. Magn. Mater. (2005) 293, 633 99. Mornet, S., et al., J. Mater. Chem. (2004) 14, 2161
45. Pankhurst, Q. A., et al., J. Phys. D: Appl. Phys. (2003) 36, R167 100 Gould, P., Nano Today (2006) 1 (2), 34
46. Lübbe, A. S., et al., J. Surgical Res. (2001) 95, 200 101. Lu, J., et al., Nanotechnology (2006) 17, 5812
47. Lübbe, A. S., et al., J. Magn. Magn. Mater. (1999) 194, 149 102. Bonnemain, B., J. Drug Targeting (1998) 6, 167
48. Maaßen, S., et al., STP Pharma Sci. (1993) 3, 11 103. Weissleder, R., et al., Radiology (1990) 175, 489
49. Lee, J., et al., J. Colloid Interface Sci. (1996) 177, 490 104. Moghimi, S. M., et al., FASEB J. (2005) 19, 311
50. Bulte, J. W. M., and Kraitchman, D. L., NMR Biomed. (2004) 17, 484. 105. Chellat, F., et al., Biomaterials (2005) 26, 7260
51. Saebo, K. B., (ed.), Comprehensive summaries of Uppsala Dissertations form the 106. Butterworth, M. D., et al., Colloids Surf., A (2001) 179, 93
Faculty of Medicine, Uppsala University, Sweden, (2004)
107. Cheng, J., et al., Pharm. Res. (2006) 23, 557
52. Ambrose, D. L., and Fritz, J. S., J. Chromatogr., B (1998) 709, 89
108. Feng, S.-S., and Chien, S., Chem. Eng. Sci. (2003) 58, 4087
53. Ferrari, M., Curr. Opin. Chem. Biol. (2005) 9, 343
109. Schenck, J. F., Prog. Biophys. Mol. Biol. (2005) 87, 185
54. Yamaoka, T., et al., J. Pharm. Sci. (1994) 83, 601
110. Haik, Y., et al., J. Magn. Magn. Mater. (2001) 225, 180
55. Meng, F., et al., J. Biomed. Mater. Res. (2004) A70, 49
111. Chakeres, D. W., and de Vocht, F., Prog. Biophys. Molec. Biol. (2005) 87, 255
56. Moats, R. A., et al., Angew. Chem. Int. Ed. (1997) 36, 726
112. Kangarlu, A., and Robitaille P. M. L., Concepts Magn. Reson. (2000) 12, 321
57. Jacques, V., and Desreux, J. F., Top. Curr. Chem. (2002) 221, 123.
113. Macaroff, P. P., et al., J. Appl. Phys. (2006) 99, S102
58. Arruebo, M., et al., Adv. Funct. Mater. (2007), in press
114. Park, S. I., et al., J. Magn. Magn. Mater. (2006) 304, e406
59. Bullok, K. E., et al., Mol. Imaging (2006) 5, 1
115. Duncan, R., and Izzo, L., Adv. Drug Delivery Rev. (2005) 57, 2215
60. Nitin, N., et al., J. Biol. Inorg. Chem. (2004) 9, 706
116. Yin, H., et al., Biomaterials (2005) 26, 5818
61. Qasba, P. K., et al., AAPS J.(2006) 8, E190
117. Morris, J., Nanotechnology White Paper. US Environmental Protection Agency,
62. James, J., and Kullberg, M., Ethnic. Disease (2005) 15, 113 2007
63. Morishita, N., et al., Biochem. Biophys. Res. Commun. (2005) 334, 1121 118. Fernández-Pacheco, R., et al., Nanotech (2005) 1, 144
64. Schillinger, U., et al., J. Magn. Magn. Mater. (2005) 293, 501 119. Tomalia, D. A., Prog. Polym. Sci. (2005) 30, 294
65. Yoon, T.-J., et al., Angew. Chem. Int. Ed. (2005) 44, 1068 120. Boyd, B. J., Technology overviews, Colloidal drug delivery. Drug Delivery Report,
66. Gersting, S. W., et al., J. Gene Med. (2004) 6, 913 Autumn/Winter 2005, 63
67. Huth, S., et al., J. Gene Med. (2004) 6, 923

32 JUNE 2007 | VOLUME 2 | NUMBER 3