ICANCER RESEARCH56. 4694-4701.

October IS, 1996]

Predinical Experiences with Magnetic Drug Targeting: Tolerance and Efficacy

Andreas Stephan Lübbe,'Christian Bergemann, Winfried HUhnt, Thomas Fricke, Hanno Riess,
Jeffery Walter Brock, and Dieter Huhn
Department of Medicine (Hematology and Oncology). Virchow Medical School. Humboldt-Universität zu Berlin, Augustenburger Platz I. 13353 Berlin (A. S. L, C. B., W. H.,
7,.F.. H. R., D. H.J and Department ofBiological Sciences. Murray State University, Murray, Kentucky42071 (J. W. B.)

ABSTRACT of the particle size and the surface characteristics. Many attempts have
been made to formulate various forms of alternative drug delivery
Although slte-spedflc direction of drugs within an organism would methods, yet they are all in preclinical stages.
benefit patients with many diseases, active drug targeting Is dinically not
Although magnetically controlled, targeted chemotherapy had been
yet possible. To overcome some of the problems associated with active
experimentally tried with various systems (magnetic emulsions, mag
drug targeting, we have developed a magnetic fluid to which drugs,
cytokines, and other molecules can be chemically bound to enable those netic starch microspheres, magnetic erythrocytes, and magnetic albu
agents to be directed within an organism by high-energy magnetic fields. mm microspheres), never has a patient been treated with such a
In the first part of this study, various concentrations ofthe magnetic fluid system (8—10).There have been too many problems that needed to be
were tested in rats and Immunosuppressed nude mice with regard to overcome; apart from a lack of data suggesting easy large-scale
subjective and objective tolerance. In the second part, the same parame production of the magnetic drug carrier, a lack of data supporting
ters were evaluated after administration of the ferrofluid to which epiru reproducibility in the making of the system, aggregation of some
bicin (4'-epldoxorublcin) was chemically bound. Finally, two forms of magnetic carriers in vivo, and the need of strong-enough magnets with
therapy with the magnetic fluid were tested: tumor treatment by median constant field gradients, to name but a few, there was concern regard
ical occlusion with the ferrofluid in high concentradonsi and magnetic
ing the long-term deposition of magnetic particles in the organism, in
drug targeting, using small amounts of the ferrofluid as a vehicle to
addition to discouraging data from large animal experiments and
concentrate epirubicin locally in tumors. As a result, the ferrofluld did not
cause major laboratory abnormalities there was no LD,@. With very high general obstacles with regard to regulatory approval and the econom
concentrations of the ferrofluid, animals showed lethargy for 1—2
days. ics of the therapy (8). Taken together, numerous experiments on small
There were no Intolerances with the epirubicin-bound ferrofiuld as well. animals have not lead to further exploration of the idea to direct drugs
Both forms of treatment led to complete tumor responses in an experl within large organisms by means of magnetic fluids in conjunction
mental human kidney as well as in a xenotransplanted colon carcinoma with magnetic fields.
modeL Thus, the magnetic fluid is a safe agent, which can be used in Some of the above-named problems were directly associated with
different ways for local forms of cancer treatment In conjunction with the magnetic fluid or ferrofluid itself. Through a series of experimen
high-energy magnetic fields.
tal steps, we have produced the first ferrofluid to which drugs can be
bound directly. Its advantage is the independence of a third vehicle
and the excellent in vivo stability under the influence of magnetic
INTRODUCTION
fields.
There is a long-existing wish to treat local problems locally and If the ferrofluid were physiologically well tolerated and directable
systemic problems with systemic medical therapy. However, if the within the organism, it should be possible to concentrate it in tissues
local disease problem is within an organism and inaccessible to or tumors that are located at the body surface by externally applied,
conventional local treatment forms (surgery, radiation therapy, and high-energy magnetic fields. By these means, two modes of action
topical application of antibiotics) systemic treatment is often used to seem possible; therefore, we investigated: (a) if the amount of the
treat local problems. This, in turn, requires administration of large ferrofluid were high enough, one could expect mechanical occlusions
amounts of drugs, most of which are metabolized by normal tissues of tumor blood vessels and some therapeutic benefit; and (b) if the
(1 , 2). For drugs with a low therapeutic index, this causes a series of minimal amount of the ferrofluid was used as a vehicle to direct a
problems. Site-directed drug targeting could circumvent this problem. cytostatic drug to a tumor, then drug concentrations could result in
Other advantages of drug targeting are the potential to reduce the tumor responses that would otherwise not cause those responses with
amount of drugs needed (biological and economic impact) and the systemic application.
opportunity to use drugs that otherwise would not be able to be given Therefore, it was the purpose of the following studies to test
to the patient (pharmacological impact; Ref. 3). whether the magnetic fluid in physiological as well as supraphysi
Although many attempts have been made to find a clinically ap ological concentrations was tolerated well in two animal models,
plicable way to perform drug targeting, so-called active drug targeting Sprague-Dawley rats and immunosuppressed mice. We examined
is currently not available for any medical indication. It requires behavior, laboratory values, tissue specimens, and survival. Also, we
guidance of the particular drug to the target cells in a manner that wanted to know whether the ferrofluid could be used for magnetically
differs from its normal distribution characteristics. Conjugation of controlled (i.e., magnetic field-dependent) localization within the or
drugs to antibodies specific to target cell antigens is one example that ganism and, more specifically, whether magnetically bound epirubi
is clinically being explored (4, 5). Passive drug targeting, on the other cm, a well-known and widely used anticancer drug, could be directed
hand, is used clinically in the form of liposomes and other vehicles to tumors that were localized at the body surface.
that incorporate the drug component and release it according to
passive rules of distribution (6, 7). Those rules are mainly a function MATERIALS AND METHODS
The Ferrofluid
Received2/21/96;accepted8/15/96.
The costs of publication of this article were defrayed in part by the payment of page The ferrofluid was obtained from Nano-Technologies GBR (Berlin, Ocr
charges. This article must therefore be hereby marked advertisement in accordance with
18 U.S.C. Section 1734 solely to indicate this fact.
many). It was a colloidal dispersion, made by wet chemical methods from iron
I To whom requests for reprints should be addressed, at Cecilien-Klinik, Cecilienallee oxides and hydroxides into special multidomain particles. Those particles were
6-8, 33175 Bad Lippspringe, Germany. specially (not randomly) arranged to possess advantages with respect to
4694

Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
 PRECLINICAL EXPERIENCES WITH MAGNETIC DRUG TARGETING
directability under the influence of magnetic fields in vivo. The particles were
surrounded with anhydroglucose polymers to stabilize the magnetic particles
under various physiological conditions. In addition, the surrounding polymer
was used for chemoadsorptive binding. Because the binding was reversible,
desorption of the drug that had been bound to the surface occurred according
to the physiological environment (pH, osmolality, and temperature). Thus, the
desorption of drugs took place by competitive forces through blood electro
lytes and could be varied according to the specific need. In the experiments
described below, desorption ofepirubicin took place 30 mm after the measured
intravasal availability of the magnetic particles. Thus, it was assured that the
drug could act freely once it had been placed in the tumor by the magnetic
field. The ferrofluid was isotonic, with a viscosity close to that of water. The
dispersion contained 1.5% particles, i.e., 15 mg (0.24 mM)iron oxide in 1 ml.
The particle diameter was 50—150am. The pH was 7.4, the color was black,
and the odor was neutral. The magnetic particles occupied 1.5% of the total
weight; the rest of the ferrofluid was water. The particles consisted of Fe304
and contained 60% pure iron, such that there was a total iron content of roughly
6 mg/ml ferrofluid (German potent 19624426.9).
The Magnetic Field
For the animal experiments, high-energy permanent magnets were used that
were made of rare earths (neodymium being the primary agent). They were
column or block shaped and consisted of discs or blocks with variable thick
nesses. Thus, the magnets could be arranged most tightly around the individual
configuration of the tumors. The magnets never compressed the tumors and
created a magnetic strength of up to 0.5 tesla but at least 0.2 tesla (depending
on the size of the tumor).
Epirubicin
Epirubicin (4'-epidoxorubicin; Farmorubicin) was a gift from Pharmacia. In
the respective experiments, the dry substance was diluted with the ferrofluid
instead of isotonic saline solution according to standard procedures within 15
mm prior to administration.
Animals
Overall, 71 male Sprague-Dawley rats were used for the studies. They were
obtained from Bomholtgard (Ry, Denmark). Animals were maintained on a
controlled diet of Purina Rat Chow and were given water ad libitwn in an
American Association for the Accreditation of Laboratory Animal Care
approved animal care center. They were housed in an environmentally con
trolled room with a 12-h light-dark cycle until body weights ranged between
180 and 280 g.
Overall, 166 male athymic nude mice (NMRI-nu/nu), weighing between 20
and 28 g at the age of 6—12weeks, were used for the studies. These animals
were purchased from the same company as above 1—2 weeks prior to the
experiments and were acclimated under laminar-flow conditions
temperatures between 24-26°Cand a humidity of6O% in the same animal care
center. The mice were allowed a standard laboratory diet (3-Altromin 1320;
Altromin, Lage, Germany) and acidified water ad libitum. For the mouse as
well as rat experiments, food but not water was withdrawn 12 h before each
experiment. All animals were handled according to the NIH 1978 Guide for the
Care and Use of Laboratory Animals. For tumor implantation, mice and rats
were anesthetized with sodium pentobarbital (45 mg/kg i.p.). For all other
experiments, animals were lightly anesthetized with ketamine-HCI (45 mg/kg
body weight) and xylacin (6 mg/kg body weight). Rectal temperature was
continuously monitored and maintained between 36 and 37°Cby a heating
lamp, which was positioned above the animal. Respiration rates were also
monitored.
The various compounds (epirubicin, the ferrofluid, and the magnetic epiru
bicin) were injected into the tail veins of the animals, which sometimes had
been lightly anesthetized. Histological specimens were sometimes
after injection of an overdose of the anesthetic. Specifically, organs were
quickly removed after the animals had been sacrificed, embedded in formalin,
and processed according to standard procedures for staining with H&E or
potassium hexacyanoferrat for iron, usually within the next 48 h.
Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
Tumors and Technique of Implantation
The tumors that were used in the experiments were either a malignant and
aggressive metastasizing adenocarcinoma of the colon or a hypernephroma.
Both tumorshadbeen removedfrompatientsandcultivatedwithinnudemice
for up to 20 generations. Both tumors were rather undifferentiated, according
to 03. As described in detail in other articles (11, 12), small tumor pieces
(volume, 5—15 p3) of a donor animal were prepared and put into Ringer lactate
solution. After injection of 10 @.tl
isotonic saline under the dermis of the dorsal
ear surface, a small s.c. pocket (3 mm long and 1 mm wide) was created. One
of the tumor pieces was cut to a size of approximately 1 p3, a volume
equivalent to a sphere with a diameter of 1.25 mm, and was introduced into the
ear pocket before the 1-mm incision was sewn with a 10-0 silk suture. In other
experiments, larger tumor pieces (50 @sl)
were transplanted s.c. bilaterally into
the abdominal region.
Experimental PrOtOCOlS
Rat Experiment 1: Weight and Behavior. Of 23 rats, four subgroups
were formed. Subgroup 1 (n 4) received isotonic saline solution (5% of the
estimated blood volume, which was estimated to be 10% of the body weight).
Subgroup 2 (n = 7) received the ferrofluid (5% of estimated blood volume).
Subgroup 3 (n = 6) received the ferrofluid (5% of estimated blood volume)
bound to epirubicin (1 mg/kg body weight). Subgroup 4 (n 6) received an
injection of 1 mg/kg body weight epirubicin only (volume also 5% of the
estimated blood volume). Animal behavior and weight were documented daily
for up to 4 weeks.
Rat Experiment 2: Laboratory Values. In six subgroups (36 rats), em
phasis was put on hematological and chemical laboratory values. Subgroup 1
(n 6) received the ferrofluid in a low dose (0.5% of estimated blood
volume). Subgroup 2 (n = 6) received the ferrofluid in a high dose (5% of
blood volume). Subgroup 3 (n 6) received 1 mg/kg body weight epirubicin;
subgroup 4 (n 6) received 10 mg/kg body weight epirubicin; subgroup 5
(n 6) received 1 mg/kg body weight epirubicin bound to the ferrofluid; and
subgroup 6 (n = 6) received 10 mg/kg body weight epirubicin bound to the
ferrofluid. The volume of injection in the last four subgroups was 0.5% of the
estimated blood volume in each experiment. Animal behavior, weight (not
shown), and laboratory values were obtained prior to and I, 2, 7, 14, and 35
days after therapy.
Rat Experiment 3: Histology. In 12 rats, the ferrofluid (2% of the esti
mated blood volume) was injected i.v. over 1 mm into the tail vein. Four, 12,
18, and 65 days later, three of the animals were sacrificed in each subgroup by
an overdose of ketamine-HC1, and relevant organs were histologically exam
med with standard procedures for H&E or potassium hexacyanoferrat (staining
for iron).
Mouse Experiment 1: Survival. In 13 subgroups, animals were subjected
to an injection of either isotonic saline (5% of estimated blood volume; n 4)
or the ferrofluid [ranging from 0.1% of the blood volume (n = 4) to 1%
(n 5), 5% (n 6), 10% (n = 5), and 20% (n 4) of the blood volumej.
Subgroup 7 (n = 5) received 0.5 mg/kg body weight epirubicin; subgroup 8
at room
(n 4) received 0.9% saline solution; and subgroup 9 (n 6) received 1
mg/kg body weight epirubicin bound to the ferrofluid. To subgroup 10 (n 6),
0.5 mg/kg epirubicin bound to the magnetic fluid was given; subgroup 11
(n 5) received 1 mg/kg body weight epirubicin; subgroup 12 (n 4)
received 10 mg epirubicin/kg body weight bound to the ferrofluid; and sub
group 13 (n = 4) received 10 mg epirubicinlkg body weight only. The volume
of each injectant in the last 7 subgroups was 0.5% of the estimated blood
volume. Overall, 62 mice were used for those survival studies. The observation
period was terminated after 12 weeks.
Mouse Experiment 2: Tumor Embolization. Each mouse of the two
subgroups (subgroup I, colon carcinoma, n 6; subgroup 2, hypernephroma,
n = 6, implanted into the outer ear) received the ferrofluid (10% of the
estimated blood volume). A magnetic field was applied over the tumor for 20
mm, including the time of injection, roughly 1 mm. Then, the tumor was
measured manually in three axes over the next 14 days.
obtained Mouse Experiment 3: Tumor Treatment with Epirubicin. In four sub
groups with a total of 16 animals, two concentrations of epirubicin (I and 10
mg/kg body weight) were administered i.v. into the tail veins of the animals,
which were carrying large tumors (colon and kidney cancers) under the
abdominal skin.
4695
 PRECLINICAL EXPERIENCES WITH MAGNETIC DRUG TARGETING

.n
Sal
Fig. 1. Weight of Sprague-Dawley rats >1@@
(378 ±12 g baseline weight) after application of
isotonic saline solution (0), magnetic fluid (MF;
5% of the estimated blood volume; •),epirubicin all

(Epi: 1 mg/kg body weight; 0), and magnetic epi

rubicin (5% of estimated blood volume; 1 mg/kg a
.@ @-
body weight, @).Data presented as means; bars, U
SEM. No differences among the groups were found.

0 2 4 6 8 10 12 14 16 18

flue (days aftertrealment)

Mouse Experiment 4: Magnetic Drug Targeting. The volume of xeno rats, as indicated for each group, increased or remained at baseline
transplanted colon and kidney cancers was measured daily over 10 days in six levels according to normal growth rates over the period of 18 days, no
subgroups after treatment with I mg epirubicin/kg body weight bound to the matter whether the animals received saline solution (sham-treated
ferrofluid and directed to the tumor with a magnetic field (colon, n = 6; kidney rats), the ferrofluid in high doses, or epirubicin with and without being
cancer, n = 5) or not directed to the tumor with a magnetic field (colon,
attached to the magnetic fluid (experiment 1; Fig. 1). Overall, there
n = 10; kidney cancer, n = 6) or with no treatment of the two tumor entities
were no statistical differences among the groups.
at all.
Mouse Experiment 5: Control Studies. In four subgroups, the ferrofluid Laboratory values for the animals of experiment 2 demonstrate that
(0.5% of the estimated blood volume) was administered in the two tumor relevant hematological and blood chemical values did not change
entities (colon, n 7; kidney cancer, n = 6) while a magnetic field was from baseline after injection of different amounts of the ferrofluid (0.5
applied for 20 mm, in contrast to two other subgroups, in which the procedure and 5% of the estimated blood volume; Table 1). Epirubicin, however,
was identical, except that a magnetic field was not applied (n 6 in both tumor did cause changes in hematological parameters, as expected. Although
groups). this was not particularly striking in the low-dose range (1 mg/kg body
Statistics weight epirubicin), not only was there a significant mortality in the
high-dose group (10 mg/kg body weight epirubicin), but in the sur
In general, data were reported as means ±SEM. Means ±SEM of the viving animals there were also significant drops in the number of
tumor volumes were calculated as baseline values and for changes over time. WBC, platelets, and hemoglobin (Table 2). This basic pattern did not
Baseline data were analyzed by one-way ANOVA to test for any changes at the
change significantly with regard to the absolute values in the groups
P < 0.05 level and the Bonferroniprocedure for multiple comparisons.
in which epirubicin at both concentrations was bound to the ferrofluid
(Table 3). All other organ functions we measured seemed to be intact.
RESULTS
The histological data revealed that the magnetic particles accumu
The first part of the study put emphasis on the tolerance of the lated in the liver and spleen, as expected, without causing significant
ferrofluid or the magnetically bound epirubicin. Body weights of the hepatosplenomegaly (by wet weight analysis), whereas they were not

ferrofluidLaboratory Table 1 Meanlaboratory values over time prior to and after application of the
value
(0.5/5%)Sodium Day 0 (0.5/5%)―Day 1 (0.5/5%)Day 2 (0.5/5%)Day 7 (0.5/5%)Day 14 (0.5/5%)Day 35
133/143141/146143/142141/145141/145146/141Glucose
(M)
247/248252/194261/180276/229264/189169/195Creatinine
(mg/dl)
0.4/0.450.4/0.450.4/0.40.3/0.450.35/0.40.3/0.4Albumin
(mg/dI)
2.3/2.92.2/2.92.5/2.92.7/2.73.0/2.92.7/2.7Iron
(g/dl)
30.7130.238.7/25.829.7/33.734.0/30.839.0/39.032.0/38.0Bilirubin
(psi)
0.3/0.150.6/0.20.1/0.20.1/0.20.1/0.20.15/0.9Cholesterol(mg/dl)
(mg/dl)
96/11698/106117/126110/110112/116100/110ALT
(units/liter) 44/39
3/4AP'(units/liter)
GOT6 2/334/38 7/430/44 4/442/44 2/349/45 2/445/38
263/239210/249184/233195/213216/247180/222Lipase
(units/liter)
<50/<50<50/<50<50/<50<50/<50<50/<50<50/<50Hemoglobin(g/dl)
(units/liter)
14.0/14.912.6/13.411.0/12511.6/14.313.3/12.315.1/15.0Leukocytes
(0/liter) 17.0/14.01
1.0/17.716.0/17.015.1/16.512.1/10.68.3/14.6l'hrombocytes
951/1006706/1048834/6421258/12221321/727848/936a
(G/liter)
Ferrofluid as percentages of estimated blood volume.
b
Gamma-glutamyltransferase.C
Alkaline phosphate.

4696

Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
 PRECLINICAL EXPERIENCES WITH MAGNETIC DRUG TARGETING

epirubicinLaboratory Table 2 Mean laboratory values over time prior to and after application of
mg)Sodium value Day 0 (1/10 mg)― Day I (1/10 mg) Day 2 (1/10 mg) Day 7 (1/10 mg) Day 14 (1/10 mg) Day 35 (1/10
139/143Glucose
(mM) 145/139 140/138 141/140 138/140 140/138
187/—Creatinine
(mg/dl) 221/177 186/166 176/227 171/185 167/155
0.5/0.5Bilirubin(mg/dl)
(mg/dl) 0.4/0.4 0.5/0.4 0.5/0.5 0.4/0.4 0.4/0.5
0.1/0.1ALT 0.8/0 0.1/0 0.1/0 0.1/0.1 0.1/0.1
(units/liter) 34/33 34/18 46/26 30/29 39/22 35/23
256/—Lipase
APb (units/liter) 184/254 250/206 247/— 215/196 229/206
<50/<50Hemoglobin
(units/liter) <501<50 <50/<50 <501<50 <50/<50 <50/<50
14.9/13.7Leukocytes
(g/dI) 14.9/14.8 14.1/13.9 13.2/13.1 13.2/10.9 13.6/8.7
8.0/8.5l'hrombocytes
(G/liter) 9697.5 8.8/8.6 9.2/5.7 9.0/2.9 7.4/4.3
866/1600a (G/liter) 731/844 666/818 744/678 972/529 792/428
Epimbicin doses in mg/kg body weight
phosphatase.found
b Alkaline

themeasured;
in other organs. We noted a significant (semiquantitatively over 7 days compared with treatment with no magnetic field (i.e.,
tumor)those data not shown) decrease in the number of particles in magnetic drug was given, but no magnetic field attached to the
betweenfar
two organs over the 65 days. Organ structure was not altered, as or no treatment whatsoever (Fig. 6). There was no difference
types.No
as histological criteria were concerned (experiment 3; Table 4), the two tested tumor
causedmice,
matter which concentration of the ferrofluid we injected into the Fig. 7 shows that the same amount of the magnetic fluid that
bloodobservation,
there was no increased mortality rate. After the sixth week of successful magnetic drug targeting (0.5% of the estimated
epirubicincardiovascular
one animal in each high-dose group died, probably of volume) did not cause significant tumor responses when
nonresponse,those
collapse due to sepsis (mouse experiment 1; Fig. 2). In had not been attached to the ferrofluid. Thus, for this
role.animals
groups in which the high dose of epirubicin was administered, application of a magnetic field did not play any
somewhatlater,
died rapidly; in the low-dose groups, they died
around 4—6weeks (Fig. 3). This fact was independent of DISCUSSION
epirubicin
includingthebeing bound to the ferrofluid. In all other groups,
by(0.5
ones in which a low concentration of epirubicin was administered The present study demonstrates good tolerance of the ferrofluid
mg/kg body weight), animals survived over the observation the animals and effective tumor therapies with both mechanical ob
period
weThe (Fig. 3). struction by high concentrations of magnetic particles and what
magneticlization
second part of this research focused on the mechanical embo- call magnetic drug targeting, using low concentrations of the
well?by by the ferrofluid after injection and concentration in the tumor fluid. Is it surprising that the ferrofluid was tolerated
showtype,
means of an external magnetic field. Independent of the tumor Both the laboratory values and the histological observations
there was a rapid and consistent decrease of the tumor volume that organ function was not significantly altered acutely and chroni
within
specialIt
14 days after treatment (Fig. 4). cally. The magnetic fluid that was used in this study was a
thatinwas impossible to reproduce this tumor response in the animals colloidal dispersion, which consisted of multidomain particles
specialthe
which only epirubicin was given. Although tumors did respond to were manufactured such that the particles were aligned in a
irona
high dose of epirubicin (10 mg/kg body weight), this was only for manner on an external magnetic force. The particles consisted of
whichshortly
brief period (several days), and most animals of this group died oxide and hydroxide and, therefore, were made of material for
similarbasically
thereafter. According to the baseline volumes, there was the effect on live organisms is well known. Specifically,
forthe an unaltered growth characteristic of the tumors over time in ferrofluids have been developed as ferrimagnetic contrast agents
Althoughall
groups that received the low dose of epirubicin or no treatment at diagnostic nuclear magnetic resonance purposes (13, 14).
wasMagnetic
(Fig. 5). the latter compounds are considerably different from the fluid that
drug targeting demonstrated
studies,Table
significant tumor responses used in this study, they also contain iron cores. In preclinical

3 Mean laboratory values over time prior to and after application of magnetic epirubicin
DayO Day] Day2 Day7 Dayl4 Day35
Laboratory value (1/10 mg) + 0.5% MFa (1/10 mg) + 0.5% MF (1/10 mg) + 0.5% MF (1/10 mg) + 0.5% MF (1/10 mg) + 0.5% MF (1/10 mg) + 0.5% MF
Sodium (mM) 142/140 141/144 145/142 145/143 140/142 142/139
Glucose(mg/dl) 170/0 195/0 168/0 0/0 212/0 0/0
Creatinine (mg/dl) 0.4/0.4 0.4/0.4 0.4/0.4 0.3/0.3 0.3/0.3 0.3/0.4
Iron (@sM) 28/29 33/41 35/57 37/36 24/28 34/29
ALT (units/liter) 40/43 36/37 33/26 40/43 38/43 41/42
GGTb (units/liter) 2/3 2/2 3/3 4/3 4/4 3/1
Hemoglobin (g/dI) 13.4/14.8 13.7/12.8 13.3/12.5 12.8/12.9 13.5/0 15.1/14.2
Leukocytes(G/liter) 10.5/11.0 11.1/16.3 9.7/8.7 8.5/11.5 10.1/0 13.6/10.7
Thrombocytes (G/liter) 981/809 909/628 1022/749 775/696 123/0 515/694
a Epirubicin doses in mg/kg body weight. MFF, magnetic ferrofluid given at 0.5% of the estimated blood volume.
b Gamma-glutamyltransferase.

Table 4 Histological iron deposition in some organs at varioustimes after injection ofthe ftrrofluid
dataLiverLungKidneySpleenHeartAnalysis (2% of the estimated blood volume): semiquantitative

after 4 days + + +(‘ ++

Analysis after 12 days +++ +++ —
Analysis after 18 days ++ — — +++ —
—a —— —+ + +—
Analysis after 65 days ++—
@ ÷ @,much; + +, moderate; and —,little iron deposition.

4697

Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
 PRECUNICAL EXPERIENCES WITH MAGNETIC DRUG TARGETING

90

90

70

60

Fig. 2. Long-term survival of nude mice (percentage of respective a90

group) after increasing volumes (percentage of estimated blood vol
ume) of magnetic fluid (MF) or isotonic saline solution (NaCl). Data
40 —0—NaCI
represent the means,.
——MF0.1%
30 —D--MF1%
——MF5%
20 —@—MF
10%
—@—MF
Z@%
10

0
0 1 2 3 4 6 6 7 8 9 10 11
Thtie (viaeks after treabnent)

dose escalation experiments with those contrast agents have shown no
harmful effects on organ function, as reflected by laboratory values
and other organ function tests (15, 16). In animal and preclinical
studies, iron concentrations far higher than those used in this study
had been tested. One main difference between ferrofluids used for
diagnostic purposes and the one described in this article was the
particle size. Although the particle size in the ferrofluid used in this
study is 10 to 50 times higher than that of ferrimagnetic contrast
fluids, it was safe. We saw no accumulation in the lungs or any
clinical signs of respiratory problems. Only in supraphysiological
dose ranges, in which 10—20% of the blood volume
within a relatively short time, subjective responses were obtained in
the animals, characterized by lethargy for 12—24h and resistance of
food uptake. Also, because of the relatively enormous iron load,
discoloration of the animals was noticed for about 1 week, symptoms
quite similar to those noted in preclinical studies by Van Hecke et al.
(16). Taken together, the ferrofluid that was used in this study was
tolerated well from a clinical standpoint, as shown by the laboratory
and histological data. Uptake and successive elimination of iron
particles by the reticuloendothelial system are well-known character
istics of this system, which were confirmed by the histological data.
One crucial characteristic of the ferrofluid was its carbohydrate
coating and thus, on the one hand, its in vitro stability over months
(i.e. , no sedimentation of the magnetic particles) and, on the other, the
capacity for adsorptive binding of many different drugs (German
patent 19624426.9). An important feature of the ferrofluid was its
was infused ability to separate from the drug whenever necessary. For drug
targeting purposes, it was the intention to use the ferrofluid as a
vehicle to achieve relevant concentrations of the drug within the
tumor tissue. Once localized at the site of choice, desorption must
occur, by which the drug leaves the vehicle and can freely act on the
target tissue. Because anthracyclines treat a wide spectrum of ma

I
Fig. 3. Long-term survival (percentage of respective group) of nude mice after
increasing doses of epirubicin (Epi) or epirubicin (all applications in a volume
estimated to be 0.5% of the blood volume) bound to the magnetic fluid (MF).
High and intermediate mortality resulted with 10 and I mg epinibicin/kg body
weight, respectively. Data represent means; bars. SEM.

0 1 2 3 4 6 6 7 8 9 10
Time (v@aeksaft@ trealrnent)
4698

Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
 PRECLINICAL
EXPERIENCES
WITHMAGNETICDRUGTARGETING

10000

1000

Fig. 4. Tumor regression after mechanical obstruction

by the ferrofluid (10% of the estimated blood volume) in
100
conjunctionwitha magneticfieldlocalizedat the tumor
at the body surface (20 mm). Co colon cancer (•,U);
RC, renal cell carcinoma (0, 0). Data represent means;
bars, SEM; significant difference, P < 0.05 from base I
line.

10 —0-- RC, therapy

—4—-Go,
therapy
—O--RC,control
—U--Co control

I —
12 14 16 18 20 22 24 26 28 30
line (days after tianor @npIantatIon)

lignomas, and because there is a positive dose-response relationship in 7 and 14 days. These data confirm what is known from the literature
some malignant diseases, such as breast carcinoma and soft-tissue sar (19, 20), and it did not make any difference whether the ferrofluid was
comas, epirubicin (4'-epidoxorubicin) was used as the anticancer agent in bound to epirubicin or not. For ethical reasons, a magnetic field was
this study (17, 18). Therefore, epirubicin and magnetically bound epiru not applied in rats that were not carrying malignant tumors. Thus,
bicin were also tested in rats and mice for clinical andlaboratory function. drug-targeting experiments with potentially high toxic concentrations
Epirubicin was used in two standard (19, 20) concentrations (a low in healthy tissues of live animals were not performed in this animal
one of 1 mg/kg body weight and a high one of 10 mgfkg body weight). model. It was, therefore, not possible to demonstrate any change in
Although the LD50 for this substance lies around 3 mg/kg body blood cell counts as a response to drug targeting with epirubicin doses,
weight, the low dose did not cause major abnormalities and was which, given systemically, would cause blood cell nadirs.
tolerated well by the animals. The high dose not only caused a high Through a series of calculations, the most practical route of appli
mortality rate, but it also led to a typical hematological nadir between cation of the ferrofluid was found to be i.v. injection resulting in its

Fig. 5. Tumor response of colon cancer (Co. •,U) and

renal cell carcinoma (RC. 0, 0) after one treatment with 1
and 10 mg epinibicin/kg body weight (Ep:). Data represent
means; bars, SEM.
I

0 2 4 6 8 10 12 14 16 18 20 22
Time (days after treatment)
4699

Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
 PRECLINICAL EXPERIENCES WITH MAGNETIC DRUG TARGETING

10000

1000@

Fig. 6. Tumor regression with magnetic drug-targeting [magnetic

fluid (MF), 0.5% of the estimated blood volume; I mg epirubicin
(Epi)/kg body weight; 20-mm magnetic field application (Dl)]. Colon 100
cancer (Co. •,U) and renal cell carcinoma (RC. 0, 0) completely
disappear within 7 days. Control studies were conducted with no
magnetic field but the same injected drug versus no therapy at all I
(control). Data represent means; bars. SEM. There was no difference
between the two tumors (P > 0.05).

10

0 2 4 6 8 10
Time (days after treatment)

localization at different sites in the organism. Also,
appropriate concentrations of the compound at which drug targeting
and mechanical obstruction of tumor blood vessels predominantly
occur. In the latter experiments, 10% of the estimated blood volume
had been injected into the animals, and a magnetic field had been
applied to 3-week-old tumors (estimated volume, 10% of the body
volume). In 7 to 14 days, the tumors underwent a series of reproduc
ible changes, leading from initial discoloration to a continuous shrink
ing with a dry and black appearance at the surface to a complete loss
of the tumors. We interpret those stages as a concentration of the
magnetic particles in the tumor-feeding vessels and the successive
necrosis over the next several days. Therefore, it was possible to cause
we arrived at complete and lasting tumor remissions by mechanical obstruction in
two different malignant human tumors. In contrast to other forms of
mechanical obstruction (such as the direct injection of particles into
the surgically prepared tumor-feeding artery), the compound was
injected systemically and allowed to distribute within the organism.
During this time and the following 15 minutes, a magnetic field that
was applied to the tumor was able to recruit at least that many
magnetic particles that were sufficient to obstruct tumor vessels and
cause necroses. Those tumor obstructions have been confirmed his
tologically and under intravital microscopy (data not shown here).
Roughly 30 years ago, the first attempts were made with vascular
occlusion of tumors and organs by mechanical obstruction with mag

10000-

1000

Fig. 7. Normal growth behavior of colon cancer (Co. closed sym @I00
bols) and renal cell carcinoma (RC. open symbols) to application of the
magnetic fluid (MF: 5% of the estimated blood volume) with and b
without a magnetic field (Dl) or compared with untreated controls (A,
L@).Data represent means; bars. SEM.
—O—RC,
MF0.5% + DT
——Co,MF0.5%9-DT
10
—D--RC,MFO.5%
—U--Co,MFO.5%
—6-—
RC, control
—è--Cocontrol

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28
line (days after treatment)

4700

Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
 PRECLINICAL EXPERIENCES WITH MAGNETIC DRUG TARGETING
netic fluids (21—24). Some of the authors of those studies used
ferrosilicone preparations with limited in vivo stability and low di
rectability with magnetic fields, and others used simple iron particle
suspensions that were placed in mannitol to obliterate cochlear blood
vessels. All of those materials have not fulfilled essential criteria for
long-term in vivo compatibility after systemic injection.
Drug targeting, i.e., the phenomenon of site-directed concentration of
drugs within the organism, has been tried in different diseases and with
many different methods. Especially in cancer therapy, in which chemo
therapeutic agents with many and severe associated side effects and low
therapeutic indices are available, the protection of normal, healthy tissues
from those drugs and the concentrations of the drugs exclusively in the
tumor tissues form an important goal. Among those methods are tumor
antigen-directed and liposomal-encapsulated drugs (4—7).However,
there are some significant drawbacks with those techniques. Either the
drugs are injected in the neighborhood of the target area (sthrch-encap
sulated drugs) and are not available at other body sites, or they are limited
to the liver and spleen, organs of the reticuloendothelial system that
phagocytose drug-containing liposomes. Both approaches are character
ized as so-called passive drug targeting.
Active drug targeting, i.e., the concentration of drugs within the
organism against the local distribution characteristics of the drug, by
magnetically controlled techniques is a roughly 20-year-old attempt to
bind anticancer drugs to magnetically active compounds and to direct
those by means of high-energy magnetic fields to the tumor (8, 25).
Until now, it has not been possible to accomplish that in a satisfactory
way. Either there were technical difficulties in the making of the
magnetic compound, or a third intermediate compound had to be used
to combine the drug with the ferrofluid. This intermediate compound,
however, usually large, denaturated albumin molecules, created many
problems, especially in larger animals; consequently, those experi
ments were stopped in the mid-l980s (8, 9, 25).
The ferrofluid that was used in this study did not necessitate a third
compound as an intermediate. Rather, with this system drugs can be
bound directly to the coating surface ofthe particles. Thus, the drugs can
be directly moved by magnetic forces together with the magnetic parti
des. In addition, the process of desorption of the drug from the magnetic
vehicle within the organism can be precisely determined. This is impor
(ant, because once the magnetic drug is injected into the body and
directed to the tumor by the magnetic field, the drug must dissociatefrom
the magnetic compound to act freely in the tumor. Through a series of
experiments, we have shown that epirubicin desorbed from the ferrofluid
at 30 mm (half-life). In other words, once the drug was injected into the
organism and a magnetic field was applied to the tumor, 50% ofthe drug
was released under the influence of the magnetic field and was free to act
on the tumor cells. Through anothertest series, we arrivedat a dose of the
ferrofluid of 0.5% of the estimated blood volume. This is considerably
lower than the 10% necessary to mechanically obstruct the blood vessels.
There is no tumor-obstructing effect with the low concentration of the
ferrofluid (0.5% ofthe estimated blood volume). Fig. 7 demonstrates that
tumor size did not shrink after injection ofthat amount and application of
the magnetic field for 15 minutes. Also, we showed that epirubicin at the
dose of 1 mg/body weight injected i.v. into rats did not cause tumor
responses.
However, when the two treatments were combined, 1 mg epirubi
cm/kg body weight was bound to the ferrofluid (0.5% of the estimated
blood volume), and this magnetic epirubicin was then injected into the
animals while a magnetic field was applied to the tumor for 15
minutes, reproducible tumor regressions occurred over the next 10 24.
days, no matter which type of tumor we tested. Thus, with this
technique and those concentrations, it was possible to successfully 25.
target the tumors.
Taken together, this is a description of a promising and potentially
4701
Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.
new form of localized cancer chemotherapy. If it were possible to
obtain high-energy permanent magnets strong enough to direct the
ferrofluid within large organisms, this therapy could be used in
patients with locally advanced and otherwise nontreatable
malignomas.
ACKNOWLEDGMENTS
This work was made possible by a gift from W. Miehrendorff and the
personaleffortof E. Bergemann.
REFERENCES
1. Poste, G., and Kirsh, R. Site specific (targeted) drug delivery in cancer therapy.
Biotechnology, JO: 869—885,1983
2. Gurney, H., Dodwell; D., Thatcher, N., and Tattersall, M. H. N. Escalating drug
delivery in cancer chemotherapy: a review of concepts and practice—partI. Ann.
Oncol., 4: 23—34,1993.
3. Gupta, P. K. Drug targeting in chemotherapy: a clinical perspective. J. Pharm. Sci.,
79: 949—962,1990.
4. Goustein, C., Winlder, U., Bohlen, H., Diehl, V., and Emgert, A. Immunotoxins: is
there a clinical value? Ann. Oncol., 5: 97—103,1994.
5. Siegall, C. B. Targeted toxins as anticancer agents. Cancer (Phila.), 74: 1006—1012,
1994.
6. Gregoriadis, 0., and Florence, A. T. Liposomes in drug delivery. Clinical, diagnostic
and ophthalmic potential. Drugs, 45: 15—28,1993.
7. Gabizon, A., Isacson, R., Libson, E., Kaufmann, B., Uziely, B., Catane, R., Ben-Dor,
C. G..Rabello,E.,Cass,P.,andPeretz,T. Clinicalstudiesof liposome-encapsulated
doxorubicin. Acta Oncol., 33: 779—786,1994.
8. Gupta, P. K., and Hung, C. T. Magnetically controlled targeted chemotherapy. In: N.
Willmott and J. Daly (eds.), Microspheres and Regional Cancer Therapy, pp 1—59.
Boca Raton, FL: CRC Press, Inc., 1993.
9. Widder, K. J., Senyei, A. E., and Scarpelli, D. G. Magnetic microspheres: a model
system for site specific drug delivery in vivo. Proc. Soc. Exp. Biol. Med., 58:
141—146, 1978.
10. Widder, K. J., Morris, R. M., Poore, G. A., Howards, D. P., and Senyei, A. E.
Selective targeting of magnetic albumin microspheres containing low-dose doxoru
bicin: total remission in Yoshida sarcoma-bearing rats. Eur. J. Cancer Clin. Oncol.,
19: 135—139,
1983.
11. Huhnt, W., and LUbbe,A. S. Growth, microvessel density and tumour cell invasion
of human colon adenocarcinoma under repeated treatment with hyperthermia and
serotonin. J. Cancer Res. Cliii. Oncol., 121: 423—428,1995.
12. LUbbe,A. S., and Huhnt, W. Microvessel diameter of human colon adenocarcinoma
during acute treatment with serotonin. mt. J. Microcirc. Clin. Exp., /4: 218—225,
1994.
13. Ferrucci, J. T., and Stark, D. D. Iron oxide-enhanced MR imaging of the liver and
spleen: review of the first 5 years. AiR Am. J. Roentgenol., 155: 943—950,1990.
14. Bacon, B. R., Stark, D. D., Park, C. H., Salni, S., Groman, E. V., Hahn, P. F.,
Compton, C. C., and Ferrucci, J. T. Ferrite particles: a new magnetic resonance
imaging contrast agent. Lack of acute or chronic hepatotoxicity after intravenous
administration. J. Lab. Clin. Med., 110: 164—171,1987.
15. Weisleder, R., Stark, D. D., Engelstad, B. L., Bacon, B. R., Compton, C. C., White,
D. L.,Jacobs,P.,andLewis,J. Superparamagnetic
ironoxide:pharmacokinetics
and
toxicity. AiR Am. J. Roentgenol., 152: 167—173, 1989.
16. Van Hecke, P., Marchal, G., Decrop, E., and Brett, A. L. Experimental study of the
pharmacokinetics and dose response of ferrite particles used as contrast agent in MRI
of the normal liver of the rabbit. Invest. Radiol., 24: 397—399,1989.
17. Plosker, G. L., and Faulds, D. Epirubicin. A review of its pharmacodynamic and
pharmacokinetic properties, and therapeutic use in cancer chemotherapy. Drugs, 45:
788—856,
1983.
18. Bonadonna, G., Gianni, L., Santora, A., Bonfante, V., Bidoli, P., Casali, P.,
Demichelis, R., and Valagussa, P. Drugs ten years later: epirubicin. Ann. Oncol., 4:
359—369,
1993.
19. Ltibbe, A. S. Doxorubicin and local hyperthermia in the microcirculation of skeletal
muscle. Cancer Chemother. Pharmacol., 31: 295—300,1993.
20. Casazza, A. M., and Giuliani, F. S. Preclinical properties of epirubicin. In: G.
Bonadonna (ed), Advances in Anthracycline Chemotherapy: Epirubicin, pp. 31—40.
Milan, Italy: Masson Italia Editon SrI, 1984.
21. Mosso, J. A., and Rand, R. W. Ferromagnetic silicone vascular occlusion: a technique
for selective infarction of tumors and organs. Ann. Surg., 178: 663—668,1973.
22. Giebel, W., Wagner, H., and Scheibe, F. Preliminary electrophysiological data after
the obliteration of cochlear blood vessels by the action of a magnetic field on
circulating iron particles. Arch. Otorhinolaryngol., 242: 337—341,1985.
23. Meyers, P. H., Cronic, F., and Nice, C. M. Experimental approach in the use and
magnetic control of metallic iron particles in the lymphatic and vascular system of
dogsas a contrastand isotopicagent.Am.J. Radiol.,90: 1068—1077,
1963.
Nakamura, T., Konno, K., and Morone, T. Magneto-medicine: biological aspects of
ferromagnetic fine particles. J. Appl. Physiol., 42: 1320—1324,1971.
Widder, K. J., Senyei, A. E., and Ranney D. F. Magnetically responsive microspheres
and other carriers for the biophysical targeting of antitumor agents. In: S. Gavattini,
A. 001dm, F. Howking, I. J. Kopin, and R. J. Schnitzer (eds.), Advances in Pharma
cology and Chemotherapy, pp 213-239. New York: Academic Press, 1979.
Preclinical Experiences with Magnetic Drug Targeting:
Tolerance and Efficacy
Andreas Stephan Lübbe, Christian Bergemann, Winfried Huhnt, et al.

Cancer Res 1996;56:4694-4701.

Updated version Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/56/20/4694

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Subscriptions Department at pubs@aacr.org.

Permissions To request permission to re-use all or part of this article, use this link
http://cancerres.aacrjournals.org/content/56/20/4694.
Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)
Rightslink site.

Downloaded from cancerres.aacrjournals.org on November 18, 2019. © 1996 American Association for Cancer
Research.