Boron Neutron Capture Therapy

Boron Neutron Capture Therapy (BNCT) is a promising method to treat malignant brain tumors that cannot be treated by other methods. It makes use of the property of Boron-10 to split into two parts upon absorption of a thermal (low-energy) neutron. One particle, a so-called alpha particle, is released with energy of 1.47 MeV, the other particle, the Lithium-7 nucleus, with energy of 0.84 MeV. Furthermore, in 94% of the time, a gamma ray is released with energy of about 0.48 MeV, which usually escapes from the body without any interaction. The other two particles, however, the alpha particle and the Lithium-7 nucleus, have a high Linear Energy Transfer (LET), which implies that they deposit their energy within a very short range. Usually, this range is no larger than 10 m m, which corresponds with the diameter of a single cell. So, when we would be capable of injecting the tumor with a boron-containing compound, which in itself is non-toxic, and irradiate the tumor with neutrons, the tumor would be destroyed exclusively. Because it is not possible in practice to inject a deep-seated brain tumor, a tumor-seeking compound is delivered to the patient intravenously, after which the neutron irradiation can start.

The cross section for neutron absorption in Boron-10 is very high for so-called thermal neutrons, and decreases with increasing energy of the incoming neutron. This means that we should deliver preferentially thermal neutrons to the tumor to have as many neutron absorptions as possible.

To judge whether the Boron-10 absorption cross section is large or not, we should compare it with that of the other elements in the brain tissue. Compared with oxygen and carbon, the absorption cross section of Boron-10 is larger by a factor of one million at least. This means that with a boron concentration of only several parts per million, the dose rate of the boron dominates. However, compared with hydrogen and nitrogen, the boron cross section is larger by a factor of 1,000 to 10,000 "only". Because of the large atomic density of hydrogen and nitrogen in tissue, the dose rate due to neutron absorption interactions by these two nuclides can be considerable. Indeed, in practice the (n,gamma) reaction by hydrogen and the (n,proton) reaction by nitrogen can deliver a high background dose to the normal tissue.

BNCT is not a new concept. In fact, as early as 1936, Locher in the US proposed the principle to use neutron absorption reactions to radiation therapy. The first serious trials were performed in the fifties in Brookhaven using the Brookhaven Graphite Research Reactor (BGRR), while from 1959 to 1961, the Brookhaven Medical Research Reactor (a compact High Flux Reactor with a power of 5 MW) was used to this purpose. However, both reactors were used without much success. Neither successful were the irradiations at MIT in the same period using the MIT Research Reactor (MIT-RR). However, due to advancements in the tumor-affinitive boron-containing compounds, the Japanese researchers, who started irradiation in 1968, were more successful. Up till now, more than 200 patients have undergone BNCT treatment in Japan. In this intra-operative radiotherapy, the brain and tumor tissue is directly exposed to thermal neutrons. The success in Japan initiated new research in the US in the late eighties and later on also in Europe.

In the irradiations performed in the US and Europe (Petten), use is made of epithermal neutrons, which are being moderated to thermal energies in the brain tissue. As a result the maximum dose to the tumor can be delivered at greater depth up till 6-8 cm.

Current developments focus on the utilization of new neutron sources for BNCT, like a Californium source that drives a subcritical assembly, an accelerator-driven source making use of the 7Li(p,n) reaction, and neutron sources making use of the D-D and D-T reactions. All this research is driven by the fact that not many research reactors in the neighborhood of hospitals are available or suitable for BNCT. Reactors that are being used for the purpose of BNCT are upgraded by, for example, Fission Converter Plates. In this case, the thermal neutrons of the reactor are guided to an array of fuel assemblies outside the core where they initiate new fissions. The neutrons released are subsequently used for the purpose of BNCT. In other words, the fission converter plates work as a neutron booster.

In Petten, patients are irradiated with epithermal neutrons coming out of neutron beam HB11. Research is performed within the framework of the European BNCT group. The neutron beam HB11/12 in the Petten HFR has the nice feature that it has a large cross sectional area and that if faces directly one whole side of the reactor vessel. The useful diameter of the beam is 30 cm at the maximum. At present, a neutron filter is positioned in the beam leg HB11, and the epi-thermal neutrons are guided through the beam to the irradiation room. Patients can enter this shielded room through a lock, via the reactor emergency exit. Behind this, at the outside, rooms for the preparation of the patient and for further treatment after the irradiation are present. Notice the machinery needed to keep the argon in the liquid state.

Currently, the neutron filter consists of a thin layer of cadmium to capture the thermal neutrons in the beam, 15 cm of aluminium, 5 cm of sulphur, 1 cm of titanium and 150 cm of liquid argon. All these materials have scatter resonances in the high energy region to scatter down the fast neutrons to epithermal energies. There is an extra water layer that acts as a shutter. The mean (dose-weighted) neutron energy in the beam is about 10 keV. The epithermal flux is such that a therapeutic dose can be delivered in four irradiations of 30 minutes each.

When we have a look at the cross section of argon as a function of the energy of the incoming neutron, it is seen that in the keV range, around 60 keV, there exists a window in the cross section. Above this window energy, the elastic scatter cross section of argon is quite high, while for energies below the window energy, the scatter cross section of argon is quite low. As a result, fast neutrons have a large probability to scatter with the argon nuclei and to reach the epithermal energy range, while, once they are epithermal, they do not have a large probability to scatter down to the thermal energy range. As a result the use of argon seems to be a clever solution to this problem, with the only disadvantage that it is rather costly to maintain.

There exist several parameters, so-called Figures of Merit (FOM), to characterize a neutron beam for BNCT:

  1. The first is the Advantage Depth (AD), which is the depth in the tissue at which the maximum dose to the normal tissue equals the dose to the tumor. In this definition, the dose consists of the sum of the total background dose and the Boron-10 dose in the tumor. The AD indicates the depth of effective beam penetration. In other words, the AD measures the ability to penetrate tissue and to deliver a therapeutic dose at depth.
  2. The second one is the Advantage Ratio (AR). This is defined as the ratio of the dose to the tumor and the dose to the normal tissue both integrated from the body surface to a certain depth, for which usually the AD is used. The AR is a measure of the therapeutic gain, which is the ability of the beam to maximize the desired radiation dose to tumors while minimizing background dose to healthy tissue. Both the AD and the AR characterize the beam quality.
  3. The third characterization is the Advantage Depth Dose Rate (ADDR), which is the total therapeutic dose at the AD. In fact the ADDR is a measure of the ability to deliver the desired radiation dose in an acceptable treatment time.

A plot of the AD as a function of the AR for a range of ideal mono-energetic mono-directional neutron beams shows like half a donut. Thermal neutrons have very high AR values because of the relatively low dose to nomal tissue. High-energy neutrons have small AR values and small AD values because the radiation dose from the fast neutrons at the surface of the body becomes comparable to the therapeutic dose from Boron-10. Clearly, in between these two extremes, an optimum exists for epithermal neutrons with both a large AR and a large AD. Unfortunately, in the neutron beam optimization process, the in-phantom FOMs are difficult to use, because of the dosimetric calculations needed.

To overcome this problem, so-called in-air FOMs are defined that characterize a neutron beam without the need of in-phantom studies. The usual parameters are:

By means of the flux-to-dose conversion factors provided by, for example, the ICRP-51, the fast neutron dose and the gamma dose can easily be calculated. These should subsequently be converted to equivalent doses by using the so-called RBE values (the Relative Biological Effectiveness) of the different types of irradiation.

When we use these three parameters to characterize the HFR neutron beam, we find that it performs better than the neutron beam used at MITR-II, but worse than the beam used at the BMRR. Both at the BMRR and at the MITR-II reactors, the Fission Converter Based (FCB) concept is considered to boost the performance of the beam. In conclusion, an upgrade of the Petten HFR neutron beam is desirable to keep up with the other beams currently being developed.

Reactor Power (MW)
(cm-2.s-1)
(Gy.cm2/n)
(Gy.cm2/n)
HFR 45 3.3E8 8.6E-13 1.0E-12
BMRR 3 1.8E9 4.3E-13 1.3E-13
BMRR (FCB) 3 1.2E10 2.8E-13 1.0E-13
MITR-II 5 2.1E8 8.6E-13 1.3E-12
MITR-II (FCB) 5 1.7E10 1.3E-13 1.0E-13

Another parameter of interest is the directionality of the beam, which is defined as the epithermal neutron current divided by the epithermal neutron flux density. Because of the relatively long distance between the core of the Petten HFR and the irradiation position of the patient, the directionality of the Petten beam is very high.

Besides the fundamental nuclear properties of the materials that should be favorable, the materials used in a neutron filter should also meet other criteria:

Finally, we would like to make some remarks on the code system to be used. Modern Monte Carlo codes, in which the neutrons and gamma rays are tracked through the geometry from their birth to their escape or absorption, have many variance reduction methods to reduce the computation time needed. At the TU-Delft, a new method has been developed that can be very suitable to the problem of BNCT. In this method, the so-called Midway Monte Carlo method, particles are tracked from the source to the detector, while so-called adjoint particles are tracked from the detector to the source. Somewhere in between, these particles meet at an artificial surface and the detector response is calculated. Experience has shown that the computation times needed for deep-penetration methods with a relatively small source and a small detector volume may be reduced by a factor of 2-10.

Another method standard available in modern Monte Carlo codes is the so-called differential sampling method. In this method, derivatives of responses are calculated in the Monte Carlo run, which can be used afterwards to calculate the sensitivity of the detector response to certain input parameters. In this way, a first-order guess of the detector response can be made when the neutron filter arrangement is slightly modified, without having to do new Monte Carlo runs. This seems a tool very well suited to this kind of optimization problems.

LITERATURE:

For more technical information, see the contribution to the IRI annual report (2003).

The slides can be downloaded here.



For more information, please contact  j.l.kloosterman (at) tudelft.nl Home