1990;50:1061-1070. Cancer Res
Rolf F. Barth, Albert H. Soloway and Ralph G. Fairchild
Boron Neutron Capture Therapy of Cancer
Access the most recent version of this article at:
E-mail alerts related to this article or journal.Sign up to receive free email-alerts
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
To request permission to re-use all or part of this article, contact the AACR Publications
on May 29, 2013. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
[CANCER RESEARCH 50. 1061-1070. February 15. 1990)
Perspectives in Cancer Research
Boron Neutron Capture Therapy of Cancer1
Rolf F. Barth,2 Albert H. Soloway, and Ralph G. Fairchild
Department of Pathology and College of Pharmacy, The Ohio State L'nirersily. Columbus, Ohio 43210 Â¡R.F. B., A. H. S./, and Medical Department, Brookhaven
National Laboratory, L'pton. Long Island, New York 11973 [R. G. F.J
Boron neutron capture therapy is based on the nuclear reac
tion that occurs when a stable isotope, '"B, is irradiated with
low energy (0.025 eV) or thermal neutrons to yield stripped
down helium nuclei (Â«-particles)and 7Li nuclei.
/He + 7Li + 2.79 MeV (6%)
4He + 7Li + 0.48 MeV 7 + 2.31 MeV (94%)
The therapeutic potential of this reaction was recognized by
Locher over 50 years ago (1), but it was Sweet (2-4), who first
suggested that BNCT1 might be useful for the treatment of
brain tumors. Shortly thereafter, a clinical trial was initiated at
the Brookhaven National Laboratory in cooperation with Sweet
and others at the Massachusetts General Hospital utilizing
borax as the capture agent (5, 6). The objective at that time was
to use BNCT as an adjunct to surgery for the treatment of
patients with the most highly malignant and therapeutically
refractory of all brain tumors, glioblastoma multiforme. Further
trials were carried out in the early 1960s, but as will be described
in more detail later on, these failed to show any evidence of
therapeutic efficacy (5-7) and were associated with adverse
effects in normal tissues (7). Stimulated by the more encour
aging clinical studies of Hatanaka et al. (8, 9) for the treatment
of malignant gliomas and those of Mishima et al. (10) for
melanoma, there has been renewed national and international
interest in BNCT. The theoretical advantage of BNCT is that
it is a two component or binary system, consisting of 10Band
thermal neutrons, which when combined together generate high
LET radiation capable of selectively destroying tumor cells
without significant damage to normal tissues. In order for
BNCT to succeed a critical amount of 10B and a sufficient
number of thermal neutrons must be delivered to individual
tumor cells. Over the past few years the Department of Energy
and the NIH have renewed funding for BNCT-related research,
and this has supported a growing number of investigators in
many different disciplines. Advances in BNCT in the areas of
compound distribution and pharmacokinetics compare favora
bly with other emerging modalities such as photon activation
therapy, photodynamic therapy, and the use of radiolabeled
antibodies for cancer treatment in which physiological targeting
There are a number of nuclides that have a high propensity
for absorbing low energy or thermal neutrons (Table 1), and
this property, referred to as the neutron capture cross-section
(<r),is measured in barns (1 b = 10~24cm2). Of the various
Received 1/22/88; revised 8/1/89, 11/13/89; accepted 11/15/89.
'Supported by Grants 5R01CA41288-03 and P-30 CA16058-15 from the
National Cancer Institute; by Grant DE-AC02-82ERG0040 and contract DE-
AC02-76CHOOD16 from the Department of Energy; and by The Ohio State
University, Office of Research and Graduate Studies.
1To whom requests for reprints should be addressed, at The Ohio State
University. Department of Pathology, 4170 Graves Hall, 333 W. 10th Avenue.
Columbus, OH 43210.
' The abbreviations used are: BNCT, boron neutron capture therapy; NCT,
neutron capture therapy; LET, linear energy transfer; BNL, Brookhaven National
Laboratory; BMRR. Brookhaven medical research reactor; PBF. Power Burst
Facility; n,h, thermal neutrons; RBE, relative biological effectiveness; BPA. bo-
ronophenylalanine: i.e.. intracranial; i.g.. intragastric.
nuclides that have high neutron capture cross-sections, "'B is
the most attractive for the following reasons: (a) it is nonra-
dioactive and readily available, comprising approximately 20%
of naturally occurring boron; (b) the particles emitted by the
capture reaction [l('B(n,Â«)7Li]are largely high LET; (c) their
path lengths are approximately 1 cell diameter (10-14 /jm),
theoretically limiting the radiation effect to those tumor cells
that have taken up a sufficient amount of IOBand simultane
ously sparing normal cells and (d) the extensive chemistry of
boron is such that it can be incorporated into a multitude of
different chemical structures. Although the neutron capture
cross-sections for the elements in normal tissues (Table 2) are
several orders of magnitude lower than boron, two of these,
hydrogen and nitrogen, are present in such high concentrations
that their capture of neutrons contributes significantly to the
total radiation absorbed dose. In order to reduce this it is
essential that the tumor have high 10Bconcentrations so that
the neutron dose or fluence (ncm~2) can be held to a minimum,
thereby maximizing the "'B(n,Â«)7Lireaction and minimizing
the n,p reaction with nitrogen [14N(n,p)'4C] and the n,7 reaction
with hydrogen ['H(n,-y)2HJ. It has been estimated that with a
tumor "'B concentration of 50 Â¿Â¿g/g.86% of the total radiation
dose would result from the capture reaction (2).
7Li and Â«-particles are the primary fission product of the
neutron capture reaction with 10B. Â«-Particles are relatively
slow and give rise to closely spaced ionizing events that consist
of tracks of sharply defined columns. They have a path length
of approximately 10 urn, are high LET, and destroy a wide
variety of biologically active molecules including DNA, RNA,
and proteins. For these reasons there is little, if any, cellular
repair from Â«-particle-induced radiation injury. Since the
10B(n,Â«)7Lireaction will produce a significant biological effect
only when there is a sufficient fluence of thermal neutrons and
a critical amount of 10Blocalized around, on, or within the cell,
Table 1 Thermal neutron capture cross-section values of potential nuclides for
neutron capture therapy
Â°Asterisk (*) indicates that the nuclides are radioactive.
* Capture cross-sections (a) are given in barns, where 1 barn = 10~" cm2.
Table 2 Thermal neutron capture cross-section values of normal tissue elements
' Capture cross-sections (a) are given in barns, where 1 barn = 10 24cm.
on May 29, 2013. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
BORON NEUTRON CAPTURE THERAPY
the radiation produced can be extremely localized thereby spar
ing normal tissue components. Thus, selectivity is simultane
ously one of the advantages and disadvantages of BNCT, since
it requires delivery of boron-10 to tumor cells in greater
amounts than normal cells. In contrast to the ionizing radiation
produced by radionuclides, little or no radiation is delivered to
bystander cells by the 'Â°B(n,Â«)7Lireaction, if the 10Bis selec
tively localized on or within tumor cells. Otherwise, adverse
effects may be produced in surrounding normal tissues (11-13).
A major advantage of a binary system is that each component
can be manipulated independently of the other. With BNCT
one can adjust the interval between administration of the cap
ture agent and neutron irradiation to an optimum time when
there is the highest differential between normal tissues and the
tumor. Furthermore, the neutron beam itself can be collimated
so that the field of radiation is circumscribed and normal tissues
with high '"B concentration can be excluded from the treatment
volume. Protection of normal tissues near and within the treat
ment volume can be achieved by selective targeting of 'Â°Bto
the tumor. Following the early clinical trials at the Massachu
setts General Hospital and the Brookhaven National Labora
tory it became apparent that there were two major reasons for
their lack of success, (a) Thermal neutrons are attenuated
rapidly in tissue due to absorption and scattering, and their
effective depth of penetration is limited to 3-4 cm. This means
that only superficial tumors would be destroyed by the capture
reaction, (b) The boron compounds that were used were freely
diffusible, low molecular weight substances that did not achieve
selective localization in the tumor. Those which did had high
blood values, and this explains why so much radiation was
delivered to adjacent normal brain.
Boron Chemistry and Compound Development
Ideally, boron compounds to be used for BNCT (Fig. 1)
should have a high specificity for malignant cells with concom-
itantly low concentrations in adjacent normal tissues and blood.
Since it is desirable to confine the radiation solely to these cells,
an intracellular and optimally intranuclear localization of boron
would be preferred. Initially, boron compounds were not spe
cifically designed for use in BNCT but, rather, were selected
because of their ready availability, known pharmacology, and
lack of toxicity (13). Because of these considerations, sodium
borate, boric acid, and their derivatives were chosen for evalu
ation. Time course studies in mice that had been inoculated
either s.c. or i.e. with a transplantable ependymoblastoma were
used to evaluate the clinical potential of a compound for BNCT
(14). It was postulated that these inorganic boron compounds
would not penetrate normal brain tissue to the same degree as
brain tumors, where the blood-brain barrier was absent or
severely compromised. Differences in the concentration of bo
ron in tumor and brain were detected, but these were transient
and not very large, and within a period of 1 to 2 h had decreased
to unity (2, 4, 15). This limitation prompted a major effort in
compound development prior to further clinical trials. From
more than 100 compounds that were screened, p-carboxyben-
zeneboronic acid and sodium decahydrodecaborate (Na^B^M^)
were selected. These attained tumonbrain boron ratios of 5-
8:1 which persisted for 2-3 h (16). Subsequently these were
synthesized at a IOBenrichment level of 92-94% (16) and used
for another clinical trial, the results of which were disappointing
In the 1960s the basis for achieving selective delivery of boron
compounds was unknown, since by their very nature they were
not naturally occurring. The clinical results of Sweet et al.
provided the impetus for a pragmatic approach to compound
development. Following the cessation of clinical trials in 1961,
new compounds were developed and screened. These were
administered several times daily, followed by a 2-day interval
in order to allow blood boron values to fall. From the many
compounds tested, two sulfhydryl-containing boron hydride
anions, B12H|]SH2~ and B10Cli((SH)22~,initially synthesized at
E. I. duPont de Nemours Co., were chosen for further study.
These had tumor:blood boron ratios in mice that ranged from
1.7-20:1 (17). In retrospect, based upon the methods of synthe
sis that were used at that time, it is likely that the initial
preparations contained mixtures of the disulfide analogues as
well as the mercapto compounds. There are major biological
differences between the B,2HI22~aniÃ³nand its mercapto coun
terpart, B|2H,|SH2~, which has the potential to form mixed
disulfides with disulfide groups on various plasma proteins (18).
Such covalent linkage of sulfur-containing polyhedral boranes
bound to proteins has been demonstrated by their cleavage with
dithiothreitol. It remains to be determined whether the incor
poration of mercapto compounds into proteins of tumor cells
is the basis for their selective uptake. There also has been
increasing interest in the disulfide and its further oxidation
product: B12H,,S-SB,2H,,4 and B12H,,S(O)SB,2H,,4- (19).
The disulfide attained higher concentrations in gliomas than
did the parent mercaptoborane, but at the same time the liver
enzyme levels were elevated suggesting hepatotoxicity. The
mechanism for the increased uptake of the dimer is also un
known. One possibility is the facile generation of stable free
radicals by homolytic cleavage of disulfide groups (20) and their
subsequent incorporation into proteins. Alternatively, the di
sulfide may react directly with sulfhydryl-containing constitu
ents of tissue. The mechanism by which sulfur-containing com
pounds achieve greater selectivity for brain tumors and possibly
other malignancies is not only important per se but may also
provide the rationale for the design and development of other
capture agents for BNCT.
Because of high and persistent boron concentrations in tumor
and low systemic toxicity, Na^BuHuSH appeared to be a
particularly attractive compound for BNCT. For these reasons
Hatanaka (8) initiated a clinical trial in the late 1960s prior to
the standard requirement of in depth pharmacokinetic studies
in humans. Nevertheless, boron uptake data have been accu
mulated in 57 patients with surgically resected brain tumors
who received NaiB^HuSH at doses ranging from 30-80 mg of
10Benriched Na2B12H,iSH/kg of body weight by intracarotid
infusion approximately 12 h prior to neutron irradiation. The
average concentration was 26.3 tig/g tumor and 18.2 ng/g
blood, and the mean tumorblood ratio obtained from 48 pa
tients was 1.69 (21). The lack of toxicity of this compound in
nearly 100 patients, together with a suggestion of therapeutic
efficacy, has provided the impetus for more detailed pharma
cokinetic and tissue distribution studies, which will be initiated
shortly at several institutions.
Malignant melanoma is another tumor that is a candidate
for treatment by means of BNCT. Although melanoma cells
are variably resistant to photon irradiation (22), they are highly
sensitive to Â«-particles (23). Mishima (24) in Japan first pro
posed the incorporation of boron into chlorpromazine as a
capture agent for the treatment of melanomas. Fairchild et al.
(25) have shown that this compound is selectively accumulated
in murine and hamster melanomas in amounts exceeding 100
Mg/g of tumor. Based on these data several boron-containing
derivatives of this drug have been synthesized (26, 27) and are
on May 29, 2013. © 1990 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from
BORON NEITRON CAPTURE THERAPY
Fig. 1. Boron compounds that either are
being used or potentially could be used as
capture agents are shown above. Compound 1.
a sulfhydryl-containing polyhedral borane
(Na2B12H,,SH), was first shown by Soloway et
al. ( 17) to have tumor-localizing properties and
currently is being used by Hatanaka in Japan
as a capture agent to treat patients with glio-
blastoma (8. 9). Compounds 2 and 3 are deriv
atives of alanine and phcnylalanine. respec
tively. Based on their selective incorporation
into melanin. Mishima in Japan first suggested
that phenylalanine derivatives might be useful
in the treatment of melanomas (28). Com
pounds 4 and 5 are promazines and the ration
ale for their use is based on the observations
of Mishima (24) and Fairchild et al. (25) that
chlorpromazine exhibits a significant degree of
localization in melanoma. Compound 6 is a
carboranylporphyrin. Hematoporphyrin has
an affinity for tumors, which has led to its use
in photodynamic therapy. Kahl in the United
States has synthesized boron-containing ana
logues (36), such as the one shown here. Com
pound 7 is a boron-containing nucleoside syn
thesized by Schinazi and Prusoff (41). Such
analogues of nucleic acid precursors may be
incorporated into rapidly dividing malignant
HC CHCH2CH(NH3)COO" CH2CH(NH3)COO
\ // NH
being evaluated for their in vivo tumor localizing properties. /?-
Boronophenylalanine is another compound that is being studied
as a potential capture agent for the treatment of melanoma.
The rationale for its use is the avidity of melanomas for aro
matic amino acids and their subsequent incorporation into
melanin (28, 29). Tumor localization has been demonstrated
following i.g. administration by means of whole body autora-
diography (30) and in several patients with cutaneous mela
noma following pn ilesionai injection (31). A clinical trial of
this compound is currently under way in Japan under the
direction of Mishima and his promising results (10) will be
discussed in more detail later in this review. Stimulated by
Mishima's experience, a number of other boron-containing
amino acids have been synthesized that potentially could be
incorporated in larger amounts into proteins of malignant cells
(32). Another approach to the selective targeting of boron to
melanomas is based on the observation that thiouracil is pref
erentially incorporated into melanotic melanomas during me-
lanogenesis (33). This observation provided the impetus for the
synthesis of several boron-containing thiouracils (34), and these
currently are being evaluated in animals.
Two other classes of compounds with a propensity for local
izing in malignant tumors are the porphyrins and the related
phthalocyanines. The biochemical basis by which these com
pounds achieve elevated concentration in malignant tumors is
unknown, but this observation has served as the rationale for
the use of hematoporphyrin derivative in the photodynamic
therapy of cancer (35). The high concentration of these com
pounds in tumors and their intracellular localization and per
sistence have stimulated several groups of investigators to syn
thesize boronated porphyrins (36) and phthalocyanines (37) as
potential capture agents. Boronated porphyrins appear to be 3-
4 times more effective per unit dose in cell culture than the
monomeric or dimeric form of Na2B,2HnSH (38). Although
liver concentrations of these compounds are also high (36) this
would not limit their use as a capture agent for the treatment
of brain tumors. Key questions that must be answered for all of
these compounds include: (a) the biochemical and physiological
mechanisms by which they concentrate in tumors; (b) their
toxicity; and (c) their photosensitizing potential in humans.
One final category of low molecular weight boron compounds
that should be mentioned are boron-containing purines and
pyrimidines and their nucleosides. The rationale for their de
velopment is that such compounds may be selectively incorpo
rated into rapidly proliferating tumor cells and trapped within
the cell following their conversion to the corresponding nucleo-
tide. Alternatively, these bases and their nucleosides may func
tion as analogues of naturally occurring precursors of nucleic
acids and become incorporated into nuclear DNA. Cytoplasmic
or preferably a nuclear localization of all of these boron com
pounds would be advantageous since the heavy particles result
ing from the capture reaction would deliver a greater proportion
of their energy to intranuclear targets, thereby permitting lower
boron concentrations than would have been required if the
compounds were located extracellularly (39, 40). Schinazi and
Prusoff (41) have synthesized the first boron-containing nucle
oside, 5-dihydroxyboryl-2'-deoxyuridine, an analogue of thy-
midine, and have shown that it was not cytotoxic to African
green monkey (Vero) cells at a concentration level of 1600 Â¿Â¿M
(42). In vitro neutron radiation studies of cells grown in the
BORON NEl'TRON CAPTURE THERAPY
presence of 5-dihydroxy-2'-deoxyuridine produced a biological
effect that was equivalent to a concentration of ~6 /ug '"B/g,
which, if attainable in vivo, would be sufficient for BNCT.
All of the compounds described in the preceding section are
low molecular weight substances, but in most instances, it is
unclear whether these compounds remain unchanged in vivo. It
is conceivable that these structures could interact with other
molecular species such as serum or cellular proteins through
the formation of ionic, hydrophobic, or covalent linkages to
yield conjugates that would alter either the transport or cellular
uptake of the capture agent. To date, very little has been done
to determine whether such reactions occur and what effects, if
any, they may have on their selective concentration in tumors.
Antibodies and Other Macromolecular Species
During the 1960s and early 1970s interest developed in the
potential use of polyclonal antibodies directed against tumor-
associated antigens for the delivery of drugs and radioisotopes
to tumors (43-45). In 1964 Soloway suggested that antibodies
might be used for the selective targeting of "'B to tumors (15).
Hawthorne et al. (46) reported on the incorporation of the
diazonium salt from l-(4-aminophenyl)-l,2-dicarbo-c/os0-
dodecaborane into antibodies directed against bovine serum
albumin and polyclonal antibodies directed against human and
mouse histocompatibility antigens (46). It was claimed from in
vitro experiments that these immunoconjugates were capable
of delivering enough boron to human and murine lymphocytes
to sustain a lethal 'Â°B(n,Â«)7Lireaction, as evidenced by reduced
viability following neutron irradiation. However, the immuno
conjugates contained only 0.2% natural boron by weight, which
was equal to 6 atoms of "'B/molecule of antibody. In retrospect,
it appears that there must have been some other explanation
for the reduced cell viability that was observed. Sneath et al.
(47) showed that water-solubilizing groups had to be incorpo
rated into protein-binding polyhedral boranes if protein solu
bility in aqueous systems was to be maintained. Subsequently,
a group of polyhedral borane derivatives containing protein-
binding functional groups were linked to IgG molecules by
means of the carbodiimide reaction without evidence of precip
With the advent of hybridoma technology and the develop
ment of monoclonal antibodies directed against a wide variety
of tumor-associated antigens, new possibilities opened up for
the targeting of IOB.Our own studies initially focused on the
linkage of the polyhedral borane disodium mercaptoundeca-
hydro-iVoso-dodecaborate (Na2B|2HiiSH) to antibodies either
by thiol disulfide exchange (49) or by means of the heterobi-
functional reagent A'-succinimidyl-3-(undecahydro-c/oio-dode-
caboranyldithio)propionate (50). Using a polyclonal antibody
directed against human thymocytes (anti-thymocyte globulin)
and 95% 10B-enriched B|2HnSH2~, we were able to incorporate
approximately 140 '"B atoms/molecule of antibody (51). This
was associated with a slight reduction in immunoreactivity.
yldithio)propionate, we were able to incorporate as many as
1500 atoms of 'Â°B/molecule of antibody, but there was a 90%
reduction in immunoreactivity (50). which was attributed to the
large number of sites that were modified.
It has been estimated that ~35-50 ng of 1UBmust be delivered
per g of tumor in order to achieve a tumoricidal effect (2, 52,
53). If this is extrapolated to the cellular level, then ~ 109atoms
of IOB must be delivered per cell (53). If boronated antibody
alone is to be used as the delivery system, then a very large
number of boron atoms must be linked per molecule. The exact
number would depend upon a number of parameters, including
the antigen site density on the target cell and the affinity
constant of the antibody molecule (54). We have used a mono
clonal antibody, designated 17-1A, which was produced against
a human colorectal cancer-derived cell line to produce boron-
containing immunoconjugates (55). This antibody recognizes
an epitope expressed with a density of 10* antigenic sites/cell
and has an affinity constant of 1.05 x IO8ivT1.Assuming that
all antigenic receptor sites could be saturated, this would require
that each molecule of antibody must deliver 1000 atoms of
boron in order to attain the critical number. This is what we
have set as a minimum requirement.
Mizusawa et al. (56) and Goldenberg et al. (57) conjugated
antibodies directed against carcinoembryonic antigen with p-
[1,2-dicarba-c/oio-[ 1-//']dodecaboran( 12)-2-yI] benzenediazon-
ium ion. The resulting immunoconjugates were estimated to
have 30-50 atoms of l()B/molecule of antibody. Although reten
tion of immunoreactivity and selective in vivo localization were
observed in hamsters carrying human colon cancer xenografts
(57), the small number of boron atoms per molecule of antibody
would preclude the delivery of a sufficient amount of boron to
sustain a lethal 'Â°B(n,Â«)7Lireaction. In order to minimize the
number of sites on an antibody molecule that would be modified
during boronation, and to maximize the number of "'B atoms
that could be linked to an antibody molecule, we have synthe
sized water-soluble, boron-containing macromolecules which
can be attached to one or two sites on the antibody molecule.
A polyhedral borane isocyanate decaborate was linked to poly-
lysine and the resulting macromolecule contained 21 to 28%
boron by weight and up to 2000 boron atoms/molecule of
polymer (58). This in turn has been linked to monoclonal
antibodies 17-1A and IB 16-6, which is directed against the B16
melanoma, utilizing two heterobifunctional reagents, Â¿V-succi-
nimidyl-3-(2-pyridyldithio)propionate and w-maleimidoben-
zoyl jV-hydroxysuccinimide ester (59, 60). More than 1000
atoms of boron have been incorporated per antibody molecule
by modifying only one site and the resulting immunoconjugates
retained 40-90% of the immunoreactivity of the native antibody
(60). The in vitro cellular uptake of the boronated 17-1A was
studied by means of electron energy loss spectroscopy utilizing
a Zeiss 902 microscope. This instrument can detect elemental
boron with a high degree of sensitivity and spatial resolution
(61). Preliminary observations suggest that there is intracellular
uptake of boron by colorectal cancer cells exposed in vitro to
the immunoconjugate (60). These observations have important
dosimetrie implications, since intracellular uptake of "'B would
increase the selective therapeutic effect achieved as a result of
the 'Â°B(n,Â«)7Lireaction (39, 62). This is illustrated by Monte
Carlo calculations for hamster V-79 cells, which show that 10B
located external to the cell will produce -10% of the dose
delivered by a uniform distribution while cytoplasmic and nu
clear locations would each deliver 2.5x the dose from a uniform
distribution (39). If boronated antibodies are not internalized,
this clearly would be a disadvantage.
In vivo distribution studies have shown that there was a
marked reduction in the amount of boronated 17-1A localized
in human colon cancer tumors that have been implanted s.c.
into nude mice and a corresponding increase in the amount in
the liver compared to the native antibody (60). The problem of
altered distribution is one that has been encountered with a
variety of immunoconjugates including immunotoxins, radio-
labeled antibodies, and drug-antibody conjugates, and meth
odology will have to be developed that produces less modifica-
BORON NEl'TRON CAPTI RE THERAPY
tion of the antibody molecule. New developments in hybridoma
technology such as the production of recombinant antibodies
that have two combining sites, one specific for the tumor-
associated antigen and the other for the tumoricidal agent,
potentially might have applicability to BNCT, if the boronated
species itself were immunogenic and reactive with one of the
combining sites on the hybrid molecule. Advances in the chem
istry of immunoconjugation, especially the ability to increase
the distance between the combining site for the tumor-associ
ated antigen and the tumoricidal moiety, also might have ap
plicability to developing better boronated immunoconjugates.
The boronated polylysine that we have synthesized has a high
negative charge, and this may have adversely affected its tumor-
localizing properties and increased its clearance by the reticu-
lodothelial system. The production of antibodies directed
against more universally expressed tumor-associated antigens
would reduce the problem of antigenic heterogeneity of tumor
cells. The potential for using monoclonal antibodies for target
ing tumoricidal agents including '"H is great, and although
significant problems exist, these theoretically are not insoluble.
One final category of macromolecular species that potentially
may be useful for the delivery of 10B is what may be termed
"encapsulating complexes," such as liposomes, microspheres,
and low density lipoproteins (63). Theoretically large amounts
of 10Bcould be encapsulated, and if these encapsulating com
plexes could be targeted to the tumor by linkage to a monoclo
nal antibody using existing methodology or targeting an endog-
enously expressed cell surface receptor, they might be powerful
delivery systems. Again, there may be preferential localization
in the reticuloendothelial system, and methodology would have
to be developed to minimize this and maximize tumor uptake.
Nuclear Reactors. It is thought that a fluence of 5 x IO'2 n,h
cm"2 will be needed for successful NCT. At the present time
only nuclear reactors are capable of generating such beams,
although accelerator-based neutron sources are being investi
gated as less expensive and more practical for hospital environ
ments. Approximately 35 research and test reactors with powers
of ~1 M W now exist in the United States that potentially could
produce beams of therapeutic intensity (64). In particular, the
Brookhaven Medical Research Reactor, the MIT Research
Reactor, and the Georgia Institute of Technology Research
Reactor have irradiation facilities that were designed for medi
cal and biological research. In addition, extensive work has been
done on the design of a proposed clinical facility for NCT at
the Power Burst Facility at the Idaho National Engineering
Laboratory. This reactor, with a steady state of power of 20
MW, would provide a beam of greater intensity than any other
currently available. The patient irradiation ports of all of these
reactors have a geometry that reduces fast neutron and 7-
photon contamination of the neutron beam thereby enhancing
its clinical potential.
Beam Types. Neutrons with an energy of ~1 MeV are "born"
in the fission reaction within the reactor core. Low energy or
thermal beams (0.025 eV), epithermal beams (1-10,000 eV) or
fast neutron beams (> 10,000 keV) may be extracted from
nuclear reactors for use in radiation therapy, by varying the
amount of slowing down or "moderation" that occurs. Fast
neutrons can be obtained by extracting a beam of neutrons that
has little or no moderation. Scattering media such as light
(H2O) or heavy (D2O) water or graphite can slow down or
"moderate" fast neutrons so that they lose energy and become
thermalized (64-66). The latter "thermal" or room temperature
neutrons are the ones that are utilized in the l()B(n,Â«)7Lireac
tion. Thermal neutrons are rapidly attenuated by tissue with a
half-value layer (distance to reduce beam intensity by a factor
of 2) of ~ 1.5 cm (40), and consequently it is difficult to obtain
sufficient neutron fluence rates at increasing depth without
heavily irradiating surface tissues. Alternatively, an "epither
mal" neutron beam (1-10,000 eV) can be produced by using
moderators or filters that slow the fast neutrons into the inter
mediate or epithermal neutron energy region. By filtering out
residual thermal neutrons with absorbers such as boron or
cadmium, a relatively pure epithermal beam can be produced
(40). This beam produces '"B-absorbing thermal neutrons,
which are the ones that interact with '"B, as it penetrates tissue
because of the moderating effects of hydrogen. Thermal neu
trons generated in tissue by such a beam actually "peak" at a
2-3 cm depth thereby circumventing problems associated with
the poor penetration of incident thermal beams. As an example,
the various beam components from the epithermal beam at the
BMRR are shown in Fig. 2. The thermal flux density generated
by the epithermal beam follows the curve for "30 ppm '"B," as
the 'Â°B(n,Â«)7Lireaction is produced by the thermal neutrons
(65). If the incident beams were a thermal beam, the falloff or
attenuation of the thermal flux would be rapid and similar to
the attenuation of the fast neutron dose (H) shown in Fig. 2.
Beam Requirements and Optimization. There is slightly in
TOTAL DOSE * 30 ppm BORON
DEPTH IN TISSUE ( cm )
Fig. 2. Dose distribution in a phantom head from the various components of
the mixed radiation field produced by the "optimized" epithermal neutron beam
at the Brookhaven medical research reactor. The beam has been "optimized" by
enhancing beam intensity, while minimi/ing fast neutron (H recoil) and y con
tamination in the incident beam. The MN(n.p)'*C and y components in Fig. 2 are
generated in tissue by the optithermal beam and cannot be reduced. Tumor would
receive a dose given by the total dose + 30 ppm boron cune (65). Time for
therapy in a single dose would be -45 min.
BORON NEUTRON CAPTURE THERAPY
creased penetration of tissues by epithermal neutrons with
increasing neutron energy so that the lowest energy fast neu
trons or the highest energy epithermal neutrons would be
optimum. For example, iron-filtered neutron beams produce
fairly pure 24-keV neutrons, but both experimental determi
nations and calculations have shown that the normal tissue
dose produced by hydrogen recoils from 24-keV neutrons is
significant and produces ~3 times the normal tissue dose than
that of an optimal epithermal neutron beam (65, 67). If, how
ever, neutrons with energies si keV are used, this harmful dose
is reduced to negligible levels (67). The acceptable level of fast
neutrons is generally believed to be ~2 x 10~" cGy per epi
thermal neutron, i.e. that dose that would be delivered by a
monoenergetic 2-keV beam (65). Current research efforts are
directed towards the production of epithermal neutron beams,
which when filtered or moderated, have a preponderance of
neutrons in the 1-1000-eV range. Since the distribution of
thermal neutrons generated at depth is only moderately affected
by the energy of the incident epithermal neutrons, it would be
best to maximize intensity by using the entire epithermal energy
region, rather than reduce intensity via a filtered monoenergetic
beam (68). When the whole reactor core is used as a source of
neutrons, suitable epithermal neutron beam intensities ^lO9 n/
cm2 sec'1 should be available with reactor powers of 1-3 MW
or more. Thus a single irradiation of 5 x IO12nlh/cnr would
take 80 min assuming that one thermal neutron was generated
per epithermal neutron (65). Reactors with this power output
include the BNL Medical Research Reactor, the Massachusetts
Institute of Technology reactor, the Georgia Institute of Tech
nology reactor, and the Power Burst Facility at the Idaho
National Engineering Laboratory.
Approximately 35 ng/g of 10B/g of tumor would be necessary
in order to raise the n,a tumor dose to levels significantly above
that delivered to normal tissues by the unavoidable n,p and n,7
reactions with nitrogen and hydrogen, respectively. With this
"optimized" epithermal neutron beam, the therapeutic gain, or
ratio of tumor dose to maximum normal tissue dose, would
approach 4 (40). It is a tenet of radiation therapy that the tumor
dose is limited by normal tissue tolerance. A therapeutic effect
could be achieved with an epithermal neutron beam delivering
5 x IO'2 (peak) n,h cm~2. The reason for this is evident from
calculated and measured dose distributions generated in a phan
tom head, from "pure" epithermal neutrons. Approximately
900 cGy (rads x RBE) would be produced by gammas and
protons from the 'H(n,7)2H and 14N(n,p)MC reactions (40, 65,
67). When this is added to 400 cGy (rads x RBE) from a
hypothetical 3.5 Â¿ig10B/g present in normal tissue assuming
one-tenth of the 35-^g IOBtumor, the normal tissue dose would
be ~1300 cGy, which approximates the normal tissue tolerance
for single dose whole brain irradiation (69).
Current efforts directed towards the modification of existing
reactors for clinical trials in the United States include those at:
(a) BNL, using the Al2O.,-moderated epithermal beam at the
BMRR; (6) MIT, using a proposed aluminum-sulfur-moderated
epithermal beam; and (c) PBF at Idaho National Engineering
Laboratory, using a proposed and yet to be installed and tested
aluminum-D2O-moderated beam. Of the three, the BMRR
beam, as shown in Fig. 2, is the only one the parameters of
which have been measured and reported. The PBF, at a power
of 20 MW, would theoretically be able to deliver therapy in a
single dose in 6 min while ~45 min would be needed for the
BMRR. Calculated parameters for the MIT reactor are prom
ising and an experimental filter is currently being installed and
tested. It is anticipated, however, that because of radiobiological
considerations such as selective repair of low LET damage in
normal tissues and redistribution of boron compounds in the
tumor, neutron irradiations will be delivered in 4-6 fractions
(70). Such fractionation would reduce the effective normal
tissue dose significantly, due to repair of the low LET compo
nent. Tissues damaged by the 'Â°B(n,Â«)7Lireaction should not
repair, due to the high LET character and the Â«and 7Li
Alternative Neutron Sources
If BNCT is shown to have therapeutic efficacy in initial
clinical trials, then alternative neutron sources become attrac
tive, due to the relative expense and paucity of suitable nuclear
reactors, as well as public concern about the siting of reactors
in metropolitan areas. Neutron sources such as Cf-252,
7Li(p,n)7Be, and spallation neutron sources are all being inves
tigated as alternatives to nuclear reactors. While ultimately
such sources may prove to be capable of providing a sufficiently
high flux of neutrons for NCT, they have not as yet been shown
to have the required intensity and purity (40, 65, 71-76).
252Cfwould be the most suitable of the various isotopie
sources that have been considered and used for neutron pro
duction (71). Neutron spectra and beam characteristics are
similar to those of reactor fission sources with a model energy
of 1 MeV. Gram amounts of 252Cf,which emits 2.34 x IO6
n/s/Mg would be necessary for beams that could be used for
therapy. The tv, of 2.65 years would provide a simple and
reliable source, which could be used in localities where no
suitable nuclear reactor is available.
Low Energy Proton Accelerators
Epithermal neutrons for BNCT also can be produced in low
energy proton accelerators (72-74) by irradiating lithium tar
gets. Proton beam currents in the range of 1-10 mA would be
required, and additional neutron moderation with materials
such as D2O would be necessary in order to obtain suitable
epithermal neutrons. Although the availability of such beams
has not been demonstrated experimentally, mA beam currents
are technically feasible and preliminary design studies have
been carried out (74).
High Energy Proton (Spallation) Sources
Neutrons of various energies are produced by irradiating
heavy elements such as copper, lead, and uranium with high
energy protons. The spectra include neutrons with energies
higher than those in fission spectra, but these can be moderated
to produce epithermal beams suitable for BNCT (75). Measure
ments carried out with 72-MeV neutrons on copper indicate
that irradiation times in the order of hours will be required
Clinical Beam Requirements
There is a consensus that the increased penetration afforded
by epithermal beams is superior to the thermal neutron beams
that currently are being used clinically in Japan. Attempts are
being made at four United States reactor facilities (BMRR,
MITR reactor, Georgia Institute of Technology research reac
tor, and PBF at Idaho National Engineering Laboratory) to
BORON NEUTRON CAPTURE THERAPY
produce beams with a preponderance of neutrons in the region
between 1 and 1000 eV using aluminum-D2O, A12O.,and alu
minum-sulfur moderators and filters. The alternative of using
scandium as a filter is being considered in the Soviet Union
(77). Clinical irradiations in Japan have been carried out with
a thermal neutron beam having an incident flux density of IO1*
ncm~2 s~'. With an incident epithermal neutron beam flux
density of IO9 ncirT2 s~' producing the same thermal neutron
flux densities throughout the tumor the epithermal beam in
stalled at the BMRR should be sufficient to produce a thera
peutic dose within ~1 h, either as a single exposure or as the
sum of a number of fractionated exposures (65, 78). While flux
levels 10 times higher (10'Â°ncm~2 s~') might be convenient for
the delivery of a single therapeutic exposure in a few minutes,
especially if a large number of patients are to be treated in a
short time, there is no known radiobiological advantage to this,
and, in fact, the reverse may be true.
Experimental Animal Studies
Soloway et al. initially reported that tumor-bearing mice
given sodium decahydrodecaborate (Na2Bi0H,â€ž)and irradiated
with 3.0-3.3 x IO'2 ncm~2 had a survival time of 62 days
compared to 23 days for reactor-irradiated and 18 days for
nonirradiated controls (79). Subsequently it was shown that a
related compound, Na2B|2HiiSH, preferentially localized in s.c.
implanted, transplantable murine ependymoblastomas (17) and
glioblastoma-like tumors (80).
Using the latter compound as a capture agent we have carried
out extensive studies with a rat brain tumor model for human
glioblastoma (81), the F98 anaplastic glioma, which has a
biological behavior similar to that of human glioblastoma mul
tiforme (82). F98 cells were implanted intracerebrally into
syngeneic CD Fischer rats and 7 to 13 days later 10B-enriched
Na2Bi2H|]SH was administered at a dose of 50 mg of com
pound/kg of body weight. At varying time intervals thereafter
ranging from 3 to 27 h, animals were irradiated at the BNL
Medical Research Reactor. The best results were seen in rats
given the capture agent 16 h prior to irradiation with 4 x IO'2
ncm~2 (429 cGy) delivered to the center of the tumor. These
animals had a mean survival time of 41 days compared to 27.7
days for unirradiated controls (81). In vitro studies with F98
cells, carried out in parallel with the in vivo experiments,
demonstrated a 99.9% reduction in the surviving fraction using
Na2Bi2HiiSH at concentrations of 50 and 100 Mg/ml and a
fluence of 10U ncm~2. The in vivo experiments reproducibly
showed an increase in life span of brain tumor-bearing rats that
were treated by means of BNCT compared to irradiated con
trols, and the in vitro results suggested that even better results
might be achieved if tumor concentrations of 10B could be
increased. Significant prolongation in survival time (60 versus
26 days for controls) has been demonstrated in another rat
model utilizing the dimer (Na4B2.,H22S2)as the capture agent
Coderre et al. (30) have treated BALB/c mice carrying s.c.
implants of the Harding-Passey melanoma by means of BNCT.
10B-enriched boronophenylalanine was administered either i.p.
or by gavage and boron concentrations, which were determined
by a track autoradiography (84), reached a maximum of 15-30
Mg'"B/g tumor 6 h following injection. Neutron irradiations
were carried out at the BNL Medical Research Reactor using
n,hfluences ranging from 2.5 to 6.7 x IO12ncm~2. At the lowest
radiation dose there was a 2-3-week delay in tumor growth
while in animals treated with the highest dose, the tumor
stopped growing and completely regressed. The X-ray dose that
gave 50% tumor control was 30 Gy, which agrees well with
estimated "effective doses" resulting from BNCT after correct
ing for the RBEs of the component radiation of the neutron
beam. These studies together with our own provide additional
experimental evidence for the therapeutic efficacy of BNCT.
Following Sweet's initial suggestion that BNCT might be
useful for the treatment of brain tumor (3), Sweet and Javid
studied the distribution of sodium tetraborate (borax) following
i.v. administration to a group of 58 patients who were undergo
ing neurosurgical biopsy or resection of their brain tumors (4).
It was observed that 3 times as much borax concentrated in
rapidly growing tumors as in surrounding normal brain tissue.
Based on these studies, a clinical trial of BNCT was initiated
in 1951 at the Brookhaven National Laboratory by Farr and
Sweet (5). A total of 10 patients with glioblastoma multiforme
were treated using borax (Na2B4O7-10H2O) as the capture
agent, followed by neutron irradiation at the BNL reactor. Five
patients received a single irradiation and 5 patients received
multiple irradiations. There was no statistically significant pro
longation of life or evidence of therapeutic efficacy. In May
1959 the BNL Medical Research Reactor became critical and
between then and May 1961 an additional group of 16 patients
were treated using either sodium tetraborate (borax) or sodium
pentaborate (Na2Bi0O|6- 10H2O) as the capture agent (85, 86).
Similarly, there was no increase in survival time or histolÃ³gica!
evidence of Â«-particle-induced radiation injury in the brains of
these patients. In the next study, carried out in 1961-1962 by
Sweet at the Massachusetts General Hospital, a series of 18
patients were treated (6). In contrast to the Brookhaven study
where there was no surgery, as much of the tumor as possible
was surgically excised, including a margin beyond grossly iden
tifiable tumor. Sixteen of the patients received an i.v. injection
of p-carboxybenzeneboronic acid and two received sodium de
cahydrodecaborate (Na2B|0Hio) via intracarotid injection. Pa
tients were irradiated at the MIT reactor following reopening
of the craniotomy wound with reflection of the scalp and dura.
Patient deaths occurred from 10 days to 11.5 months following
irradiation. Neuropathological examination at the time of au
topsy revealed extensive radiation necrosis with prominent
vascular lesions of different types in the brains of 9 of 14
patients (7). These effects were attributed to the high blood
boron concentrations at the time of neutron irradiation. These
studies led to the conclusion that it was essential to have a
boron-containing capture agent that was largely cleared from
the blood by the time neutron irradiations were carried out. A
search for more suitable boron-containing, tumor-localizing
compounds was intensified, and as described earlier in this
review, a sulfhydryl-containing polyhedral borane, Na2B,2-
H,,SH, was identified.
In 1968 Hiroshi Hatanaka, who had spent several years at
the Massachusetts General Hospital working with Sweet, ini
tiated a clinical trial in Japan utilizing Na2B,2HnSH as the
capture agent. The procedure used was similar in many respects
to that used by Sweet (17). The majority of patients had
glioblastoma multiforme, and in many instances the tumors
were recurrent. As much of the tumor as possible was surgically
removed ("debulking"), and at varying time intervals ranging
from 1 to 2 weeks following surgery the patients were given an
intracarotid infusion of '"B-enriched Na2Bi2HnSH at concen
trations ranging from 30 to 80 mg 'Â°B/kgbody weight over a
BORON NEUTRON CAPTURE THERAPY
period of 1-2 hours. This was followed approximately 12 h
later by neutron irradiation, initially carried out at the Hitachi
Training Reactor and from 1974 onwards at the Musashi
Institute of Technology reactor, using a beam of thermal neu
trons with a flux density of \0Vncm~2s~' delivered over a period
of 3-5 h. Reflection of the scalp was necessary to prevent
necrosis of the skin, which had a fairly high 10Bconcentration.
Boron concentrations in the tumor at the time of irradiation
ranged from 13 to 60 Mg/g. while that in the brain was unmea-
surable (9). Due to the rapid attenuation of thermal neutrons in
tissue, the effective depth of penetration was less than 6 cm,
thereby precluding adequate treatment of more deep seated
tumors. Hatanaka's results have been described in a number of
reports (8, 9, 86-88) and are noteworthy in several respects.
First and foremost, there was no radiation necrosis of normal
brain except in one patient who had several craniotomies and
an extraordinarily large dose of neutrons (87). This indicates
that the capture agents had been adequately cleared from the
blood. Second, of a total of approximately 77 patients who have
been treated, 38 of whom had glioblastoma multiforme, and 12
of whom had tumors located in the cerebral mantle (i.e., less
than 6 cm from the cortical surface), the reported mean survival
time was 44 months, and the median was 25.6 months. These
include several long-term survivors. One of them is a 65-year-
old man who was treated in 1972 and 17 years later is alive and
well with no evidence of tumor or neurological deficit. At least
two other patients also seem to have been cured, a 70-year-old
woman and a 13-year-old girl.
As encouraging as these results are, a number of questions
have been raised concerning Hatanaka's studies. These include
a lack of patient randomization, varying combinations of treat
ment prior to the initiation of BNCT, lack of uniformity in the
histolÃ³gica! grading of tumors, varying time intervals between
surgery and the administration of the capture agent and irra
diation, and most importantly, poor depth of penetration of the
neutron beam. What clearly is required is a controlled clinical
trial of BNCT for the treatment of glioblastoma multiforme
using currently available compounds and the best neutron
beams. At the present time plans are underway at several
institutions to carry out careful pharmacokinetic and brain
tumor localization studies in patients who are undergoing sur
gical resection of their glioblastomas. The capture agent
Na2B,2H|,SH will be administered at varying doses and time
intervals prior to surgery in order to determine the optimum
time between compound administration and neutron irradia
tion. Current interest, as described earlier in this review, focuses
on the use of an epithermal neutron beam that would have a
greater depth of penetration than a thermal beam and the
development of better tumor-localizing boron-containing com
pounds. At this point in time it is unlikely that BNCT would
be used to treat large, bulky tumors. Residual tumor that could
not be eradicated by surgery, conventional chemotherapy, or
radiotherapy could best be treated by BNCT. In order to attain
sufficient concentrations of the capture agent within the tumor,
it must have an adequate blood supply, or in the case of
micrometastases, these should be in proximity to blood vessels
through which the capture agent can diffuse and reach individ
ual tumor cells. Since the oxygen enhancement ratio of a-
particles is unity, BNCT would be highly effective against
Turning to the treatment of melanomas by means of BNCT,
Mishima and his associates have carried out pioneering work
in this area. Therapeutic efficacy initially was established in an
animal model using Duroc pigs, which develop spontaneously-
occurring melanomas (89). ["'BJBoronophenylalanine was in
jected perilesionally around the cutaneous melanoma followed
by a single neutron irradiation. As early as 2 months there was
evidence of regression and this led to a complete cure, as
evidenced by depigmentation at the melanoma site. This was
followed by a clinical study that currently is in progress. As of
November 1989 six patients with cutaneous melanoma, who
for one or another reason were not candidates for surgery, have
been treated by means of BNCT. Multiple doses of a ['"B]-
boronophenylalanine-fructose complex, which is more soluble
in water than ['Â°B]BPA,were injected perilesionally into an 80
year-old patient with a primary acrai lentiginous melanoma
occurring on the sole of the foot. After allowing for sufficient
time for the [10B]BPA to clear from surrounding normal tissues,
the patient's foot was irradiated with a dose of 1.04 x 10"
n/cm2. Within 2 weeks the melanoma showed signs of regres
sion, and this was completed by 9 weeks at which time only a
small pigmented spot remained. Two years later there was no
evidence of recurrence. An additional five patients have been
treated, and most recently the [l()B]BPA-fructose complex has
been administered i.v. Mishima et al. clearly have demonstrated
the therapeutic efficacy of BNCT for the treatment of primary
cutaneous melanoma in patients who are not candidates for
other forms of therapy. The challenge that lies ahead is to
extend this form of therapy to melanoma patients who have
disease in extracutaneous sites, which currently cannot be
treated by any available form of therapy. One such group would
be patients with cerebral mÃ©tastases,and an animal model is
being developed by us to address this problem.
The purpose of the present review was to provide an overview
of a therapeutic modality that until recently has received rela
tively little attention in the cancer literature. There are a variety
of reasons for this, not the least of which were the disappointing
results obtained in the original clinical trials that were held in
the 1950s and early 1960s. These have served as a cautionary
note against proceeding onto further clinical trials in the United
States until all of the complex questions upon which the success
of BNCT depends have been adequately addressed. These in
clude the delivery systems for 10B, the optimization of the
neutron beams to be used, careful dosimetry based on phar
macokinetic and tissue analytic studies, and the design of neu
tron sources that takes into account all of the advances that
have been made in neutron physics and nuclear engineering.
Studies in each of these areas are either planned or in progress
and it is anticipated that a carefully controlled, randomized
study could be initiated within the next few years to rigorously
assess the therapeutic efficacy of BNCT. As attractive as the
concept of BNCT is, serious problems can be encountered if all
of the various parameters are not properly optimized. For
example, intracellular distribution of boron compounds that
would be used clinically must be evaluated in order to predict
relative biological efficacy, and their pharmacokinetics must be
carefully studied in order to optimize both the absolute and
differential concentrations in tumor and normal tissues.
The true test of therapeutic efficacy for BNCT will be estab
lished only by clinical trials that bring together a most diverse
team of experts to address these complex questions. Attempts
to shortcut this process may have disastrous consequences, not
only for the patients treated but also for the future of a thera
peutic modality that otherwise might find an important place
in the armamentarium of 21st century cancer therapy.
BORON NEI IRON CAPTURE THERAPY
We thank Ada Morgan for secretarial assistance and Dr. James T.
Robertson and Dr. Frank Ellis for helpful suggestions in the preparation
of this manuscript.
1. Locher. G. L. Biological effects and therapeutic possibilities of neutrons.
Am. J. Roentgenol. Radium Ther., 36: 1-13. 1936.
2. Javid. M., Brownell, G. L., and Sweet. W. H. The possible use of neutron-
capture isotopes such as boron-10 in the treatment of neoplasms. II. Com
putation of the radiation energy and estimates of effects in normal and
neoplastic brain. J. Clin. Invest.. 31: 603-610. 1952.
3. Sweet, W. H. The use of nuclear disintegrations in the diagnosis and treat
ment of brain tumor. N. Engl. J. Med.. 245: 875-878. 1951.
4. Sweet. W. H.. and Javid. M. The possible use of neutron-capturing isotopes
such as boron-10 in the treatment of neoplasms. I. Intracranial tumors. J.
Neurosurg.. 9: 200-209. 1952.
5. Farr. L. E., Sweet, W. H.. Robertson, J. S., Foster. C. G.. Locksley. H. B.,
Sutherland. D. L., Mendelsohn, M. L., and Stickley, E. E. Neutron capture
therapy with boron in the treatment of glioblastoma multiforme. Am. J.
Roentgenol.. 71: 279-291, 1954.
6. Godwin, J. T., Farr, L. E., Sweet. \V. H.. and Robertson. J. S. Pathological
study of eight patients with glioblastoma multiforme treated by neutron-
capture therapy using boron 10. Cancer (Phila.). X:601-615. 1955.
7. Asbury, A. K., Ojeann, R. G., Nielsen. S. L.. and Sweet. W. H. Neuropath-
ologic study of fourteen cases of malignant brain tumor treated by boron-10
slow neutron capture radiation. J. Neuropathol. Exp. Neurol.. 31: 278-303.
8. Hatanaka. H. A revised boron-neutron capture therapy for malignant brain
tumors. II. Interim clinical result with the patients excluding previous treat
ments. J. Neurol., 209: 81-94, 1975.
9. Hatanaka. H.. Ramano. S.. Amano. K.. Hojo. S.. Sano. K.. Egawa, S., and
Yasukochi. H. Clinical experience of boron-neutron capture therapy for
gliomasâ€”a comparison with conventional chemo-immuno-radiotherapy. In:
H. Hatanaka (ed.). Boron Neutron Capture Therapy for Tumors. Chap. 25.
pp. 349-378. Niigata, Japan: Nishimura Co., Ltd.. 1986.
10. Mishima. Y.. Ichihashi. M.. Tsui, M., Hatta, S., Ueda. M., Honda. C., and
Susuki. T. Treatment of malignant melanoma by single thermal neutron
capture therapy with melanoma-seeking '0B-compound. Lancet. 2: 388-389.
11. Archambeau. J. O.. Alcober, V.. Calvo. W. G.. and Brenneis. H. Biological
effects of thermal neutrons and the 'Â°B(n,<r)7Li reaction. In: Biological
Effects of Neutron and Proton Irradiations, Vol. 2. pp. 55-70. Vienna:
International Atomic Energy Agency. 1964.
12. Archambeau, J. O.. Fairchild, R. G., and Brenneis. H. J. The response of the
skin of swine to increasing absorbed doses of radiation from a thermal
neutron beam, a degraded fission neutron beam, and the 'Â°B(n.<t)7Lireaction.
RadiÃ¢t.Res., 45: 145-165, 1971.
13. Pfeiffer, C. C., and Jenney, E. H. The pharmacology of boric acid and boron
compounds. Bull. Nail. Formulary Comm., 18: 57-80, 1950.
14. Locksley, H. B., and Sweet. W. H. Tissue distribution of boron compounds
in relation to neutron capture therapy of cancer. Proc. Soc. Exp. Biol. Med.,
15. Soloway, A. H. Boron compounds in cancer therapy. In: H. Stainberg and
A. L. McCloskey (eds.). Progress in Boron Chemistry. Vol. 1. Chap. 4. pp.
203-234. Elmford, NY: Pergamon Press. 1964.
16. Sweet. W. H.. Soloway, A. H.. and Brownell, G. L. Boron-slow neutron
capture therapy of gliomas. Acta Radiol. Ther. (Stockh.). /.- 114-121. 1963.
17. Soloway. A. H., Hatanaka, H., and Davis, M. A. Penetration of brain and
brain tumor. VII. Tumor-binding sulfhydryl boron compounds. J. Medicinal
Chem.. 10:114-717, 1967.
18. Alam, F., Soloway, A. H.. and Barth. R. F. Boronation of antibodies with
mercaptoundecahydro-closo-dodecaborate (2-) aniÃ³n for potential use in
boron neutron capture therapy. Int. J. Radial. Applicai. Instr. Part A, 38:
19. Slatkin, D., Micca, P., Forman, A.. Gabel. D., Wielopolski. L., and Fairchild.
R. Boron uptake in melanoma, cerebrum and blood from N.I.II, -llnsi l and
Na4B24H22S2administered to mice. Biochem. Pharmacol.. 35: 1771-1776.
20. Wellum. G. R., Tolpin, E. L, Soloway. A. H., and Kaczmarczyk. A. Synthesis
of m-disulfidobis(undecahydro-closo-dodecaborate) (4â€”)and a derived free
radical. Inorg. Chem.. 16: 2120-2122. 1977.
21. Hatanaka. H.. Amano, K.. Kanemitsu. H.. Ikcuchi. I., and Yoshizaki. T.
Boron uptake by human brain tumors and quality control of boron com
pounds. In: H. Hatanaka (ed.). Boron Neutron Capture Therapy for Tumors.
Chap. 5, pp. 77-106. Niigata. Japan: Nishimura Co.. Ltd., 1986.
22. Rofstad, E. K. Radiation biology of malignant melanoma. Acta Radiol., 25:
23. Nakanishi. T.. Ichihashi, M., Mishima, Y.. Matsuzawa. T., and Fukuda. H.
Thermal neutron capture therapy of malignant melanoma: in vitro radiobio-
logical analysis. Int. J. Radial. Biol.. 37: 573-580. 1980.
24. Mishima, Y. Neutron capture treatment of malignant melanoma using '"11
chlorpromazine compound. Pigm. Cell, /.- 215-231, 1973.
25. Fairchild. R. G., Greenbcrg. D.. Watts. K. P.. Packer. S.. Atkins. H. L., Som.
P. Hannon. S. J.. Brill. A. B.. Fand. I., and McNally, W. P. Chlorpromazine
distribution in hamsters and mice bearing transplantable melanoma. Cancer
Res.. 42: 556-561. 1982.
Nakagawa. T.. and Aono. K. Synthesis of chlorproma/ine undccahydrodo-
decaborale and nonahydrodecaborateâ€”promising agents for neutron capture
therapy of malignant melanoma. Chcm.-Pharm. Bull. (Tokyo). 24: 778-781.
Alam, F.. Bapat. B. V.. Soloway. A. H., Barlh. R. F.. Mafune. N.. and Adams.
D. M. Boronated compounds for neutron capture therapy. Sthralenther.
Onkol.. 165: 121-125. 1989.
Ichihashi. M.. Nakanishi. T.. and Mishima. Y. Specific killing effect of IOB-
/Ã¯aru-boronophenylalanine in thermal neutron capture therapy of malignant
melanoma: in vitro radiobiological evaluation. J. Invest. Dermatol.. 7Ã•Ã•:215-
Mishima. Y.. Ichihashi. M.. Tsuji, M.. Ueda. M.. Hatta. S.. Nakagawa. T.,
Tanaka. C.. and Taniyama. K. Prerequisites of first clinical trial for mela
noma selective thermal neutron capture therapy. In: H. Hatanaka (ed.).
Neutral Capture Therapy, pp. 230-236. Niigata. Japan: Nishimura Co.. Ltd..
Coderre. J. A.. Kalef-Ezra. J.. Fairchild. R. G.. Micca. P. L., Reinstein. L.
E., and Glass. J. D. Boron neutron cancer therapy of melanoma. Cancer
Res.. 48: 6313-6316. 1988.
Mishima. Y.. Ichihashi. M., Hatta. S.. Honda. C., et al. First human clinical
trial of melanoma neutron capture, diagnosis and therapy. Sthralenther.
Onkol.. 165: 251-254. 1989.
Hall, I. H., Staines, C. O., Spielvogel. B. F.. Wisian-Neilson. P.. Das. M. K..
and Wojnowich, L. Boron bctainc analogs: antitumor activity and effects on
Ehrlich ascites tumor cell metabolism. J. Pharm. Sci.. 6K:685-688. 1979.
VVhittacker, J. R. Biosynthesis of a thiouracil pheomelanin in embryonic
pigment cells exposed to thiouracil. J. Biol. Chem., 246:6217-6226. 1971.
Gabel. D. Tumor-seeking compounds for boron neutron capture therapy:
synthesis and biodistribution. In: R. G. Fairchild. V. P. Bond, and A. D.
Woodhead (eds.). Clinical Aspects of Neutron Capture Therapy, pp. 233-
241. New York: Plenum Publishing Co.. 1989.
Dougherty, T. J., Boyle. D. F., Wcishaupt. K. R.. Henderson. B. A.. Potter.
VV.R., Bellnier. D. A., and \Vityk. K. E. Photoradiation therapyâ€”clinical
and drug advances. In: D. Kessel and T. J. Dougherty (eds.). Porphyrin
Photosensitization. pp. 3-13. New York: Plenum Publishing Corp.. 1981.
Kahl. S. B., and Micca. P. Chemical and biological studies on boronated
tetraphcnyl porphyrins. In: H. Hatanaka (ed.). Neutron C'apture Therapy,
pp. 61-67. Niigata. Japan: Nishimura Co., Ltd.. 1986.
Alam. F., Bapat. B. V.. Soloway. A. H.. Barth. R. F.. Mafune. N.. and Adams.
D. M. Boronated compounds for neutron capture therapy. Strahlenther.
Onkol., 165: 121-123, 1989.
Laster, B. H., Kahl, S. B., Kalef-Ezra, J.. Popenoe, E. H., and Fairchild. R.
G. Biological efficacy of a boronalcd porphyrin as measured in cell culture.
Strahlenther. Onkol.. 165: 203-205. 1989.
Gabel. D.. Foster. S.. and Fairchild. R. G. The Monte C'arlo simulation of
the biological effect of the '"Um... i l i reaction in cells and tissue and its
implication for boron neutron capture therapy. Radial. Res.. ///: 14-25.
Fairchild. R. G.. and Bond. V. P. Current status of '"B-neutron capture
therapy: enhancement of tumor dose via beam filtration and dose rate, and
the effects of these parameters on minimum boron content: a theoretical
evaluation. Int. J. RadiÃ¢t.Oncol. Biol. Phys.. //.- 831-840, 1985.
Schinazi, R. F'., and Prusoff, W. H. Synthesis and properties of boron and
silicon substituted uracil or 2'-deoxyuridine. Tetrahedron Lett., 50: 4981-
Laster. B. H.. Schinazi. R. F., Fairchild, R. G.. Popenoe, E. A., and Sylvester,
B. Dihydroxyboryldeoxyuridine (DBDIJ): can a boronated nucleoside deceive
the DNA molecule? In: H. Hatanaka (ed.). Neutron C'apture Therapy, pp.
46-54. Niigata. Japan: Nishimura Co.. Ltd.. 1986.
Pressman. D., Day, E. D., and Blau, M. The use of paired labeling in the
determination of tumor-localizing antibodies. Cancer Res., 40: 3001-3007,
Chose. T., Cerini. M., Carter, M., and Nairn, R. C. Immunoradioactive
agent against cancer. Br. Med. J.. /: 90-93, 1967.
Ghose, T., and Nigaun, S. P. Antibody as carrier of chlorambucil. Cancer
(Phila.). 29: 1398-1400. 1972.
Hawthorne, M. F., and Wiersema, R. J., and Takasugi, M. Preparation of
tumor-specific boron compounds. 1. In vitro studies using boron labeled
antibodies and elemental boron as neutron targets. J. Medicinal Chem., Â¡5:
Sneath, R. L., Jr., Soloway, A. H.. and Dey. A. S. Protein-binding polyhedral
borancs. 1. J. Medicinal Chem.. 17: 796-799. 1974.
Sneath, R. L., Jr., Wright, J. E., Soloway, A. H., O'Keefe, S. M.. Dey, A. S.,
and Smolnycki. W. D. Protein-binding polyhedral boranes 3. J. Medicinal
Chem.. 19: 1290-1294. 1976.
Barth. R. F.. Johnson. C. W.. Wei, W-Z., Carey, W. E., Soloway, A. H.. and
McGuire. J. Neutron capture using boronated monoclonal antibody directed
against tumor-associated antigens. Cancer Detect. Prevent., 5: 315-323,
Alam. F., Soloway. A. H., McGuire, J. E., Barth, R. F., Carey, W. E., and
Adams. D. Dicesium A'-succinimidyl 3-(undecahydro-closo-dodecaboranyl-
dithio)propionate. a novel heterobifunctional boronating agent. J. Medicinal
Chem.. 28: 522-525. 1985.
Alam. F.. Soloway. A. H., Barth, R. F.. Johnson. C. W.. and Carey. W. E.
Boronation of polyclonal and monoclonal antibodies for neutron capture. In:
BORON NEUTRON CAPTURE THERAPY
First International Symposium on Neutron Capture Therapy. Cambridge.
MA. October 12-14. 1983. Brookhaven National Laboratories. Reports
BNL-51730. pp. 229-236.
52. Soloway. A. H., Barth. R. F., Alam. F., and Carey, \V. E. Boron compounds
and boronated antibodies for neutron capture therapy. In: H. Hatunaka (ed.)
Boron Neutron Capture Therapy for Tumors. Chap. 3. pp. 47-58. Niigata.
Japan: Nishimura Co.. Ltd.. 1986.
53. Tolpin, E. I., Wellum. G. R., Dohan, F. C.. Jr.. Kornblith. P. L.. and
Zamenhof, R. G. Boron neutron capture therapy of cerebral gliomas. II.
Utilization of the blood-brain barrier and tumor-specific antigens for the
selective concentration of boron in gliomas. Oncology (Basel). 32: 223-246.
54. Barth, R. F., Alam. F.. Soloway, A. H., and Adams. D. M. Delivery of boron-
10 for neutron capture therapy by means of polyclonal and monoclonal
antibodies: progress and problems. In: H. Hatanaka (ed.). Neutron Capture
Therapy, pp. 8-16. Niigata. Japan: Nishimura Co.. Ltd.. 1986.
55. Alam, F., Barth. R. F.. and Soloway A. H. Boron containing immuno-
conjugates for neutron capture therapy of cancer and for immunocytochem-
istry. Antibody. Immunoconj. Radiopharm. 2: 145-163, 1989.
56. Mizusawa, E.. Dahlman. H. L.. Bennett. S. J.. Goldenberg. D. M., and
Hawthorne. M. F. Neutron capture therapy of human cancer: in vitro results
on the preparation of boron-labeled antibodies to carcinoembryonic antigen.
Proc. Nati. Acad. Sci. USA. 79: 3011-3014. 1982.
57. Goldenberg. D. M.. Sharkey, R. M.. Primus. F. J.. Mizusawa. E.. and
Hawthorne, M. F. Neutron capture therapy of human cancer: in rivo results
on tumor localization of boron- 10-labeled antibodies to carcinoembryonic
antigen in the GW-39 tumor model system. Proc. Nati. Acad. Sci. USA, 81:
58. Alam, F., Soloway, A. H., Barth, R. F.. Mafune, N.. Adams, D. M., and
Knolh. W. H. Boron neutron capture therapy: linkage of a boronated mac-
romolecule to monoclonal antibodies directed against tumor associated an
tigens. J. Medicinal Chem.. 32: 2326-2330. 1989.
59. Barth. R. F.. Alam, F., Soloway. A. H., Adams, D. M.. and Steplewski. Z.
Boronated monoclonal antibody 17-1A for potential neutron capture therapy
of colorectal cancer. Hybridoma. 5 (Suppl. 1): 543-550. 1986.
60. Barth. R. F.. Mafune. N.. Alam. F., Adams. D. M., Soloway. A. H.. Mak-
roglou. G. E., Oredipe, O. A., Blue, T. E.. and Steplewski. Z. Conjugation,
purification and characterization of boronated monoclonal antibodies for use
in neutron capture therapy. Strahlenther. Onkol., 165: 142-145. 1989.
61. Haul.n .m. M.. Barth. R. F., Gingras, D. Londono. !.. Robinson. P. T.. Alam,
F.. Adams. D. M.. and Mattiazzi. L. Electron spectroscopic imaging for high
resolution immunocytochemistry: use of boronated protein A. J. Histochem.
Cytochem., 37: 573-580. 1989.
62. Kobayashi. T.. and Kanda, K. Analytical calculation of boron-10 dosage in
cell nucleus for neutron capture therapy. RadiÃ¢t.Res., 91: 77-94, 1982.
63. Kahl. S. B., and Callaw-ay. J. C. New tumor localizers: advances in the use
of low density lipoproteins (LDL). Strahlenther. Onkol.. 165:137-139. 1989.
64. Brugger. R. M.. Less, T. J., and Passmore. G. G. Intermediate-energy neutron
beams from reactors for NCT. In: R. G. Fairchild and V. P. Bond (eds.).
Proceedings of a Workshop on Neutron Capture Therapy, January 22-23,
1986. Brookhaven National Laboratories. Reports BNL-51995. pp. 32-41.
65. Fairchild. R. G.. Kalef-Ezra. J. A.. Saraf. S. K., Fiarman, G., Ramsey, E.,
Wielopolski. L., Laster. B., and Wheeler. F. Installation and testing of an
epithermal neutron beam at the Brookhaven Medical Research Reactor. In:
Proceedings of a Workshop on Neutron Beam Design. Development and
Performance for Neutron Capture Therapy, Massachusetts Institute of Tech
nology, March 29-31. 1989. New York: Plenum Publishing Corp.. in press,
66. Morstin, K.. and Feinendegen, L. E. Further remarks on optimization of
neutron capture therapy. In: H. Hatanaka (ed.). Neutron Capture Therapy,
pp. 75-84. Niigata. Japan: Nishimura Co.. Ltd.. 1986.
67. Fairchild, R. G., Slatkin, D. N., Gabel, D., Corderre, J.. Glass. J.. Laster. B.
H.. Borg. D. C.. Elmore. J., Foster. S.. Micca. P., and Kalef-Ezra. J. Neutron
capture therapy at Brookhaven National Laboratory. In: R. G. Fairchild and
V. P. Bond (eds.). Proceedings of a Workshop on Neutron Capture Therapy.
January 22-23, 1986, Brookhaven National Laboratories, Reports BNL-
68. Greenwood, R. C. The design of filtered epithermal neutron beams for BNCT.
In: R. G. Fairchild and V. P. Bond (eds.). Proceedings of a Workshop on
Neutron Capture Therapy. January 22-23, 1986, Brookhaven National Lab
oratories, Reports BNL-51995. pp. 123-142.
69. Ellis. F. The relationship of biological effect to dose-time-fractionation
factors in radiotherapy. Curr. Top. Radial. Res.. 4: 359-397. 1968.
70. Wambersie. A., and Feincndegen, L. E. Working group on optimization of
radiation dose delivery. In: R. G. Fairchild, V. P. Bond, and A. D. Woodhead
(eds.). Clinical Aspects of Neutron Capture Therapy, pp. 277-297. New-
York: Plenum Publishing Corp.. 1989.
71. Zamenhof. R. G.. Murray. B. W.. Brownell. G. L.. Wellum. G. R., and
Tolpin. E. I. Boron neutron capture therapy for the treatment of cerebral
gliomas. I. Theoretical evaluation of the efficacy of various neutron beams.
Med. Physics. 2: 47-60. 1975.
72. Brownell. G. Iâ€”Kirsch. J. E.. and Kehayias. J. Accelerator production of
epithermal neutrons for neutron capture therapy. In: H. Hatanaka (ed.).
Neutron Capture Therapy, pp. 127-138. Niigata. Japan: Nishimura Co.,
73. Blue. J. W.. Roberts. W. K., Blue, T. E.. Gahbauer. R. A., and Vincent, J. S.
A study of low energy proton accelerators for neutron capture therapy. In:
H. Hatanaka (ed.). Neutron Capture Therapy, pp. 147-158. Niigata. Japan:
Nishimura Co., Ltd.. 1986.
74. Wang. C-K. C.. Blue. T. E.. and Gahbauer. R. A design study of an acceler
ator-based epithermal neutron source for boron neutron capture therapy.
Strahlenther. Onkol.. 165: 75-78. 1989.
75. Conde. H.. Larsson. B.. Petterson. C.-B., Sornsuntisook, O.. Thuresson, L..
Sjostrand, N. G.. and Crawford. J. Intermediate-energy neutrons for NCT.
Status report on production and properties. In: H. Hatanaka (ed.). Neutron
Capture Therapy, pp. 139-146. Niigata, Japan: Nishimura Co.. Ltd., 1986.
76. Conde. H., Grusell, E., Larsson, B.. Rastrom. E.. Ronnquist. T., Sornsunti
sook, H.. Villa, S.. Crawford, J., Reist. H., Dahl, B., SjÃ¶strand,N. G., and
RÃ¼ssel.G. Status report on the development of a spallation neutron source
for neutron capture therapy (NCT). In: R. G. Fairchild. V. P. Bond, and A.
D. Woodhead (eds.). Clinical Aspects of Neutron Capture Therapy, pp. 319-
323. New York: Plenum Publishing Corp.. 1989.
77. Riaboukhinc. Y. Integrated approach in the planning of neutron capture
therapy. Strahlenther. Onkol.. 165: 158-162. 1989.
78. Fairchild. R. G.. Bond. V. P., and Woodhead, A. Workshop summary of
moderators' reports. In: R. G. Fairchild. V. P. Bond, and A. Woodhead
(eds.). Clinical Aspects of Neutron Capture Therapy, pp. 261-262. New
York: Plenum Publishing Corp.. 1989.
79. Entzian, W.. Soloway. A. H.. Raju. R., Sweet. W. H., and Brownell, G. L.
Effect of neutron capture irradiation upon malignant brain tumors in mice.
Acta Radiol.. 5: 95-100. 1966.
80. Hatanaka. H. Introduction. In: H. Halanaka (Ã©d.).Boron Neutron Capture
Therapy for Tumors. Chap. 1, pp. 1-28. Niigata, Japan: Nishimura Co.,
81. Clendenon. N. R.. Barth. R. F., Gordon, W. A., Goodman. J. H.. Alam. F..
Staubus, A. E.. Boesel, C. P., Yates. A. J., Moeschberger. M. L.. Fairchild.
R. G., and Kalef-Ezra. J. A. Boron neutron capture therapy of a rat glioma.
Neurosurgery. 26:47-55. 1990.
82. Clendenon, N. R., Allen, N.. Kobayashi. N., Gordon. W. A., Kartha, M.,
and Ko, L. In vivo enhancement trials of metronidazole and misonidazole in
a brain tumor model. In: L. W. Brady (ed.). Radiation Sensitizers. Cancer
Management. Vol. 5. pp. 501-506. Masson Publishing USA, Inc., 1980.
83. Slatkin, D. N.. Kalef-Ezra. J. A., Saraf. S. K., and Joel, D. D. A beam
modification assembly for experimental neutron capture therapy of brain
tumors. In: Proceedings of a Workshop on Neutron Beam Design. Devel
opment and Performance for Neutron Capture Therapy, Massachusetts
Institute of Technology. March 29-31. 1989. New York: Plenum Publishing
Corp.. in press, 1990.
84. Gabel. D., Holstein, Hâ€žLarsson, B.. Gille. L.. Ericson, G.. Sacker, D.. Som.
P.. and Fairchild, R. G. Quantitative neutron capture radiography for study
ing the biodistribution of tumor-seeking boron containing compounds. Can
cer Res.. 47: 5451-5454. 1987.
85. Farr, L. E.. Haymaker. W.. Konikowski. T.. and Lippincott, S. W. Effects of
Â«particles randomly induced in the brain in the neutron-capture treatment
of intracranial neoplasms. Int. J. Neurol., 3: 564-575, 1962.
86. Archambeau, J. D. The effect of increasing exposures of the 10B(n.a)7Li
reaction on the skin of man. Radiology, 94: 179-187. 1970.
87. Hatanaka. H.. Amano, K.. Kamano. S., and Sano. K. Clinical experience of
boron neutron capture therapy of malignant brain tumors (1968-1985). In:
H. Hatanaka (ed.). Neutron Capture Therapy, pp. 447-449. Niigata, Japan:
Nishimura Co., Ltd.. 1986.
88. Hatanaka. H., and Urano, Y'. Eighteen autopsy cases of malignant brain
tumors treated by boron-neutron capture therapy between 1968 and 1985.
In: H. Hatanaka (ed.). Boron Neutron Capture Therapy for Tumors. Chap.
26. pp. 381-416. Niigata. Japan: Nishimura Co., Ltd., 1986.
89. Mishima. Y.. Ichihashi. M., Tsui, M.. Sasumu, H., Ueda, M., Honda, C.,
and Suzuki. T. Treatment of malignant melanoma by selective thermal
neutron capture therapy using melanoma-seeking compound. J. Invest. Der-
matol.. 92 (Suppl.): 3215-3255, 1989.