Clinical practice in BNCT to the brain XA0101255

Y. Nakagawa
Department of Neurosurgery,
National Kagawa Children's Hospital,
Kagawa, Japan

Abstract. Our concept of Boron Neutron Capture Therapy (BNCT) is to selectively destroy tumour cells using
the high LET particles yielded from the 10B(n,GC)7Li reactions. The effort of clinical investigators has
concentrated on how to escalate the radiation dose at the target point. BNCT in Japan combines thermal neutrons
and BSH (TS^B^HuSH). The radiation dose is determined by the neutron fluence at the target point and the
boron concentration in the rumour tissue. According to the recent analysis, the ratio of boron concentration
(BSH) in tumour tissue and blood is nearly stable at around 1.2 to 1.69. Escalation of the radiation dose was
carried out by means of improving the penetration of the thermal neutron beam. Since 1968, 175 patients with
glioblastoma (n=83), anaplastic astrocytoma (n=44), low grade astrocytoma (n=16) or other types of tumour
(n=32) were treated by BNCT at 5 reactors (HTR n=13, JRR-3 n=l, MuITR n=98, KUR n=30, JRR-2 n=33).
The retrospective analysis revealed that the important factors related to the clinical results and QOL of the
patients were minimum tumour volume radiation dose, more than 18Gy of physical dose and maximum vascular
radiation dose (less than 15Gy) in the normal cortex. We have planned several trials to escalate the target
radiation dose. One trial makes use of a cavity in the cortex following debulking surgery of the tumour tissue to
improve neutron penetration. The other trial is introduction of epithermal neutron. KUR and JRR-4 were
reconstructed and developed to be able to irradiate using epithermal neutrons. The new combination of surgical
procedure and irradiation using epithermal neutrons should remarkably improve the target volume dose
compared to the radiation dose treated by thermal neutrons.

1. I N T R O D U C T I O N

Glioblastoma is a poorly differentiated glioma and considered the most malignant

tumour of the brain. It occurs in the white matter of the cerebrum and rapidly grows and
invades the normal brain tissue from multiple directions before the time of diagnosis. Most of
the patients with such an invasive glioma, not only glioblastoma but also anaplastic
astrocytoma and low-grade astrocytoma are beyond the point of curative treatments such as
surgery, chemotherapy, and conventional radiotherapy. The proton beam therapy & heavy-ion
therapy with Bragg peak have high risk of damage to the surrounding normal brain tissue in
the same way with surgical excision. Recent trials using high dose radiation (60-70Gy)
therapy show constantly efficient results. However, radiotherapy of the whole brain produced
extensive radiation damage. From the viewpoint of the radiation effect and good quality of life
after treatment, boron neutron capture therapy (BNCT) is an ideal treatment for malignant
brain tumours. [1,2]

2. BASIC PRINCIPLE OF BNCT

We consider BNCT an intercellular internal radiation therapy. Alpha particles (^He) and
recoiling lithium-7 (^Li) nuclei yielded from the nuclear reaction between boron-10 and
thermal neutron have a high linear energy transfer (LET) and an associated high relative
biological effectiveness (RBE). Furthermore, the two particles have a short path length (5-10
mm) which is approximately equal to the diameter of the tumour cells. Selective accumulation
of 1 OB in the tumour cells and corrective irradiation with suitable thermal neutron beam can
realise cell levelled destruction of tumour tissue without significant damage to the
surrounding brain tissue. It is well known that for a successful treatment in patients with
malignant brain tumour, it is essential to secure a sufficient radiation dose (enough alpha

240
particles & recoiling lithium-7). This depends on the boron compounds that adequately
accumulate in the tumour tissue and improvement of neutron penetration in the brain. [3]

3. HISTORY OF BNCT

Clinical trials of BNCT were initiated in 1951 at Brookhaven National Laboratory by

Fair et al and developed by Sweet and Sweet et al. at Massachusetts General Hospital/
Massachusetts Institute of Technology. However, after several trials, it was discontinued in
1961 because of the discouraging clinical results. The renewal of BNCT was organised in
1968 by H. Hatanaka in Japan using a new boron compound, BSH (NaaB^HnSH) at HTR
(Hitachi training reactor) in Japan. Successively, four reactors (JRR-3; Reactor of Japan
Atomic Energy Research Institute, MuITR; Reactor of Musashi Institute of Technology, KUR;
Research Reactor Institute of Kyoto University, JRR-2; Reactor of Japan Atomic Energy
Research Institute) were authorised for medical use. Besides facilities, evolutionary
procedures and new ideal instruments were introduced into the clinical trials. One was the
diagnostic procedure such as CT scan and MRI, which made it possible to determine the size
and the depth of the tumour with greater accuracy. The other ideal instrument was prompt
gamma ray spectrometry developed by Drs. Kobayashi and Kanda- Prompt gamma ray
spectrometry has given us more accurate data on the boron concentration in rumour tissue and
blood before a decision on the radiation time is made. As various improvements progressed, a
more correct radiation plan was made and dose escalation has been tried.

4. RECENT STANDARD TECHNIQUE OF BNCT

In order to improve the neutron penetration of the brain tissue, we de-bulk the brain
tumour and make a cavity during a preliminary operation one to two weeks before BNCT.
Partial excision of the tumour also minimises the bulk of future narcotised tissue after BNCT.
The skin flap must be large enough to allow a large aperture for the neutron beam (12 cm xl2
cm). According to the MRI findings, we insert a few gold wires in the tumour or around the
tumour for measurement of neutron flux. The tip of the wire must be around the target point.
After the operation, we identify the location of the gold wires by CT and/or skull X ray. The
day before BNCT, about fifteen hours before neutron irradiation, BSH diluted in 500-ml
saline is intravenously infused for 60 minutes by drip infusion. (60-80 mg BSH/kg body
weight). The following morning the patient is taken to the reactor. Under general anaesthesia,
the patient's skin flap is reopened and the bone flap is removed. After the opening of the dura
mater, a piece of the tumour tissue is obtained for boron-10 analysis. We place an additional
two or three gold wires on the surface of the brain to measure the neutron fluence on the
irradiation field. A thin silastic rubber balloon filled with air is placed into the cavity. The
procedure maintains the size of the cavity during neutron irradiation and improves the neutron
penetration. Following the closure of the dura matter, a heat-malleable plate of a plastic
material containing ^Li-F is applied to the patient's head to protect the skin from the thermal
neutron irradiation. This "helmet" has a hole in the center to allow the neutron beam into the
tumour-harbouring area of the brain. The beam should be as free as possible from fast
neutrons and gamma rays to avoid indiscriminate radiation to the brain. The entire head is
covered with sterile plastic drapes to prevent infection. Simultaneous neutron beam
monitoring devices are attached on the surface of the brain. Gamma rays are measured by
TLD at several points of the body. The patient is moved into the irradiation chamber. Under
remote-control general anaesthesia, the head is fixed towards the neutron port and neutrons
are delivered. Blood is intermittently drawn from the patient before and after neutron
irradiation for boron-10 analysis. Boron concentration in the brain tumour and blood is

241
measured by prompt gamma ray spectrometry during the irradiation. In order to measure the
exact neutron flux at each point of interest, gold wires inserted in the tumour tissue are pulled
out at 15-30 minute intervals after the full power operation of the reactor. It is possible to
assess the exact neutron fluence at each point of interest. The plan for the remainder of the
irradiation is then based on this up-to-the-minute data regarding boron concentration and
neutron flux.

5. TIMING OF THE NEUTRON IRRADIATION

Neutron irradiation was designed according to the clinical analysis in our series. T.
Kageji et al. reported detailed pharmacokinetics and boron uptake of BSH in a recently issued
report.. Neutron irradiation was started between 10 to 20 hours after a single infusion of BSH
in consecutive trials. [4] The mean boron concentration before the neutron irradiation in the
tumour tissue was 25.8 ppm. The tumour to blood ratio (T/B) was nearly stable at around 1.2
to 1.69 in successful cases. The study showed a significant statistical correlation between
boron uptake and time interval from the infusion of BSH. Within the first 10 hours after BSH
infusion, malignant glioma tissue contained high level of boron (30-60 ppm), however; the
boron concentration in blood showed a higher level than that in the tumour tissue. Hence the
T/B ratio was below one. hi the 12-24 hours following BSH infusion, the boron concentration
in the tumour was above 20 ppm in 56% of malignant glioma patients. The T/B ratio was
more than one in 69% and two in 38% of them. These data indicated that the neutron
irradiation should be done around 15-18 hours after the BSH infusion. A positive tumour-to-
blood ratio and a uniform tumour concentration around 10^40 mg/g ^B are needed for
successful BNCT.

6. CASE REPORTS

Case 1. A 50-year-old man developed speech disturbance and right hemiparesis.

Cerebral angiography demonstrated tumour stain in the left front-parietal area. He underwent
craniotomy and the tumour was subtotally removed. Histological diagnosis was glioblastoma.
He received BNCT at MuITR in June 1972. After craniotomy under general anaesthesia, a
ping pong ball was inserted into the cavity to improve the neutron penetration. Neutron flux
was measured on the surface of the ping-pong ball and on the bottom of the cavity using gold
foils. It was 8.8E + 12n/cm2 and 5.3E + 12n/cm2 respectively. Boron concentration in the
tumour tissue was 15.3pp. and 27.3 ppm in the blood (Fig. 1-a). Retrospective analysis of the
radiation dose of boron n-alpha reaction was 7.5-16.8 Gy (physical dose). A follow up CT
scan studied 11 years after BNCT demonstrated porencephalic cyst, however, there was no
recurrence of the tumour. After 20 years the man was still active as a farmer and holds a
driving license at the age of 70 (Fig. 1-b).

242
 Lb, Follow up CT15 years after BNCT

Fig, La.Radmtkm planning

Case 2. A 60-year-old woman with glioblastoma underwent BNCT at MuITR in July

1977. A ping pong ball was inserted into the cavity and neutron flux was measured on the
surface of the ping pong ball and on the bottom of the cavity. It was 1.45E + 13n/cm2 and
7.5E + 12n/cm2 respectively. Boron concentration in the tumour tissue was 14.0ppm and
13.3ppm in the blood. According to the retrospective analysis of the radiation dose of boron n-
alpha reaction, tumour volume dose was 15.9 Gy (physical dose). Follow up MRI studied 16
years after BNCT demonstrated multi cystic lesion, however, there was no recurrence of the
tumour (Fig. 2).

Case 3. An 11-year-old girl had a huge tumour in the right frontal lobe. Histological
diagnosis was grade 3 oligo-astrocytoma. BNCT was performed at MuITR in Oct. 1981.
Neutron flux measured on the surface of the ping pong ball and on the bottom of the cavity
using gold foils was 1.46E + 13n/cm2 and 6.72E + 12n/cm2 respectively. Boron concentration
in the tumour tissue was 22.1ppm and 11.2ppm in the blood. Tumour volume radiation dose
was 23.0 Gy (physical dose). Follow up MRI studied in 1994 demonstrated porencephalic
cyst, however, there was no recurrence of the tumour (Fig. 3).

Case 4. A 41-year-old female suffered from headache epileptic seizure and right
hemiparesis. A magnetic resonance image (MRI) showed an enhanced mass in the left parietal
lobe. She underwent craniotomy and partial resection of the tumour. Histological diagnosis
was glioblastoma. BNCT was performed at KUR in Aug. 1992. Two gold wires were inserted
around the tumour. Neutron flux was measured on the surface of the brain and at the target
point. It was 1.6E + 13n/cm2 and 4.IE + 12n/cm2 respectively. Boron concentration in the
tumour tissue was 20.0ppm and 11.2ppm in the blood. Retrospective analysis of the radiation
dose of boron n-alpha reaction was 13.0 Gy (tumour volume dose). Follow up MRI
demonstrated marked decrease of the enhanced lesion (Fig 4).

243
 Fig. 2, Follow up MRA after 16 years after BNCT, Boron concentration :I4.0 ppm in tumor tissue,
133ppm on blood Radiation time : 140 min . Radiation dosed5.9 Gy (B-iO n~a)

Fig, 3, UF, Oligo-astwcytoma (G3), Planning (kft) Oct. 1981 at MulTR, Follow up MM (right) m
1994, Radiation dose at target pint was 23Gy,

244
 .: 2.31 X K

•. -y i -y v

•year-old female with gliobktstoma*, Planning (left) Aug. 1992 at KUR, Follow up MRI
{.right) 2 years after BNCT. Radiation dose at target pint "B "was IJGy.

i I.. •>..'. r:" •.- ur-old man with glloMmtoma was underwent in March, 1995 at JRR-2. Faitow up MRI
BNCT (from ieftto right; pre BNCT, Iweets, I month, 6 months after BNCT} Radk&im dose at target
pint wtis: .HGy. Abnormally enhanced lesion gradually decreased (arrow),

245
 Case 5. A 65-year-old man had glioblastoma in the bilateral frontal lobe. He underwent
craniotomy and partial resection of the tumour. Histological diagnosis was glioblastoma.
BNCT was performed at IRR-2 in March, 1995. Neutron flux was measured using gold wires
which were inserted around the tumour. It was 4.2E + 12n/cm2 at the target point. Boron
concentration in the tumour tissue was 31.0ppm and 25.0ppm in the blood. Retrospective
analysis of the radiation dose of boron n-alpha reaction was 11.0 Gy (tumour volume dose).
Follow up MRI demonstrated a gradual decreasing of the enhanced lesion (Fig 5).

7. CLINICAL OUTCOME

Since 1968, we have treated 175 patients and performed boron-neutron capture therapy
(BNCT) using 5 reactors in Japan. There were 83 patients with glioblastoma, 44 patients with
anaplastic astrocytoma and 16 patients with low grade astrocytoma (grade 1 or 2). There were
32 patients with other types of tumour. Most of the patients were followed by CT or MRI to
study the efficacy of BNCT. Retrospectively we divided the patients into two groups to
investigate the prognostic factors. One group (group 1): the patients who lived more than 3
years. The other group (group 2): the patients who lived less than 3 years. We analyzed
histology of the tumours, age of the patients, radiation time, boron concentration in the blood,
neutron fluences on the surface of the brain at the target point target depth and tumour volume
dose in each group.

Table I. Clinical outcome of the patients who lived more than 10 years
Patient Age Sex Histology T/B ratio lOBin Tumour
tumour volume dose
M.T. 50 M Glioblastoma 0.56 15.3 13.5
R.N. 60 F Glioblastoma 1.05 13.3 18.9
T.T. 30 M Chondrosarcoma 1.07 27.1 11.6
C.U. 47 F Meningioma 8.95 90.4 13.8
C.Y. 58 F Meningioma N.D N.D 12.5
K.N. 25 M An. Astrocytoma 1.15 35.2 23.1
T.M. 11 f An. Astrocytoma 1.43 11.2 14.2
Y.T. 38 F Glioblastoma 1.42 13.9 15.9
R.K. 56 M Glioblastoma 1.89 21.6 20.3
M.I. 22 M Malig. Ependymoma N.D. N.D. 14.8
K.K. 39 M An. Astrocytoma N.D. N.D. 15.2
E.M. 48 F Meningioma N.D. N.D. 9.3

CT or MRI demonstrated marked response in 3 to 6 months after BNCT in 60% of the

patients with glioma. Twelve patients (four glioblastomas and four anaplastic astrocytomas
three meningioma, one chondrosarcoma) lived more than 10 years. Seventeen patients lived
more than 5 years. There were two patients with glioblastoma, 10 patients with anaplastic
astrocytoma and one with low grade astrocytoma. Out of 143 patients with glial tumours
treated by BNCT, 27 lived or have lived longer than 3 years after BNCT. As prognostic
factors, grading of the tumour, ages of the patients and target depth were proved as important

246
factors. However, the most important factor was tumour volume radiation dose demonstrated
by boron n-alpha reaction. The tumour volume was calculated on CT or MRI findings.
Twenty-eight patients were treated before the induction of CT, therefore the patients were
excluded in this study. [5.6]. The tumour volume radiation dose in the patients with grade 2
glioma were 11.5Gy (group 1) vs. 6.7Gy (group 2), 15.6Gy (group 1) vs. 11.8 (group 2) in
grade 3 glioma and 18.2Gy (group 1) vs. 9.8Gy (group 2) in glioblastoma patients (table 2).

8. RADIATION NECROSIS

Radiation necrosis was diagnosed by CT and/or MRI; however, it is still considered

controversial to diagnose the radiation necrosis after BNCT. Radiation necrosis was
determined as follows: low intensity on MRI T-l weighted image with contrast enhancement
(+) and high intensity on T-2 weighted image. Low density area with contract enhancement
(+) on CE-CT. Radiation necrosis was found in 19 patients (19/175 10.9%). Fourteen of those
19 patients had clinical symptoms and radiographic change. Nine of the 14 had neurological
deficits such as motor weakness and speech disturbance. The patients were treated with a high
dose of steroid therapy ( Dexamethasone 32-64mg/day was tapered for one to two weeks and
changed to prednisolone 10-30mg / day per os). Of these 9, 3 patients' symptoms gradually
disappeared after using the steroid treatment. The remaining six patients had permanent mild
to slight neurological deficits. The other 5 out of 14 patients had only epileptic seizure within
1 week after BNCT. The remaining five patients had only radiographic change without
neurological deterioration. Radiation necrosis demonstrated by CT or MRI was noticed in two
months to two years after BNCT. The age of the patients with radiation necrosis is 38.5±19.0
y.o., and the age of the patients without radiation necrosis is 41.8± 18.6 y.o. The radiation
time of the patients with radiation necrosis was 254 ± 99 min., as opposed to the radiation
time of the patients without radiation necrosis, which was 218 ± 103 min. The boron
concentration in the blood of the patients with radiation necrosis was 28± 9 ppm, while the
boron concentration in the blood of the patients without radiation necrosis was 22± 10 ppm.
The maximum neutron fluence of the patients with radiation necrosis was 2.1xlO^±
Q.6xiol3 n/cm^; however, the neutron fluence of the patients without radiation necrosis was
1.7xl0l3± 0.8x1013 n/cm^. Vascular radiation dose was calculated according to the report by
Kitao and Rydin. Only one-third of the 1"B (n, a)'Li radiation occurring in vascular lumen
will be absorbed by the vascular endothelium. Lastly, the vascular radiation dose of the
patients with radiation necrosis was 21± 8.1 Gy, while the vascular dose of the patients
without radiation necrosis was 9.4 ± 5.1 Gy. These data indicated that the maximum vascular
dose should be less than 15 Gy.

Table II. Patients surviving more than 5 years after BNCT

Patient Age Sex Histology T/B ratio lOBin Tumour
tumour volume dose
P.C. 50 M Glioblastoma 0.75 15.3 13
I.M. 15 F Rhabdomiosarcoma 3.8 28 11.1
R.T.. 4 F Pontine glioma N.D N.D. 10.6
Y.A. 44 M An. Astrocytoma 1.7 13.4 16.7
Y.S. 37 M An. Astrocytoma 1.9 25.8 10.1

247
Table II. (cont.)

T.Y.. 39 F An. Astrocytoma 1.9 30.1 18.3

K.Y.. 52 F Meningioma N.D. N.D. 10.3
PJ. 40 F An. Astrocytoma 1.1 12.8 13.3
K.O.. 33 M An. Astrocytoma 1.3 18.4 17.2
Y.M. 27 M An. Astrocytoma N.D. N.D. 11.5
M.F. 42 F Meningioma N.D. N.D 10.3
N.M. 1.4 F An. Astrocytoma 1.6 28.6 8.5
R.T 41 F Glioblastoma 0.8 11.6 16.6
K.Y. 7 M An. Astrocytoma N.D. N.D 7.9
1.0. 8 F An. Astrocytoma N.D. N.D 9.7
T.S. 31 M An. Astrocytoma N.D.. N.D 13.8
H.M. 17 M An. Astrocytoma 2.6 56.1 15.2

Table III. Radiation necrosis and related factors

Necrosis (+) Necrosis (-)
Age 38.5 ±19.0 41.8 ±18.6
Radiation time 254 ± 99 218±108(min)
B-10 in blood 28.9 ± 9 22±10(ppm)
Neutron fluences 21.E±0.6 1.7E±0.8(13n/cm 2 )
Maximum
Vascular dose 21.8 ±8.1 9.4 ±5.1 (Gy)

9. NEW PROTOCOL

The tumour volume radiation dose of the last protocol was lOGy but increased up to
15Gy in the new protocol. Surgical procedures and making a cavity played an important role
not only to irradiate with sufficient neutron fluence, but also to avoid radiation side effect.
Radiation side effect or radiation necrosis was observed in 10.9 % of our series. The factors
related to radiation necrosis were maximum thermal neutron fluence on the brain surface and
vascular dose. Therefore, we decided the maximum thermal neutron fluence on the surface of
the brain should be below 2.0 + El3 n/cm2. Vascular dose in the brain tissue near the surface
of the brain or maximum vascular dose must be controlled below 15 Gy. To improve the
neutron penetration, we decided to utilize epithermal neutron beams. KUR was reconstructed
in 1997. Following the shutdown of the JRR-2, JRR-4 was renewed for medical use in 1998.
Both reactors have the capacity to yield thermal neutron beam, epithermal neutron beam and

248
mixed beams. We compared the neutron fluences at the target point and on the surface of the
cavity between a case treated by thermal neutron and one by mixed beam (Fig. 3). Thanks to
the cavity, neutron penetration was improved ca. 30% even if irradiated by thermal neutron.
The new combination of surgical procedure and irradiation using epithermal neutrons should
remarkably improve the target volume dose compared to the radiation dose treated by thermal
neutrons, seven times without cavity and 3.5 times with cavity.

9.1. Proposed protocol for malignant brain tumours in Japan in 1998

Patient selection
- Patients with glioma grade 3-4
- Less than 70 years of age
- No serious systemic disease
- Good general condition (KPS>70)
- Radiation dose
- Minimum Tumour Volume Dose: 18Gy*
- Target Volume Dose: 15Gy*
- Maximum Vascular Dose < 15Gy
*Physical dose of boron n-alpha reaction
Neutron irradiation using mixed beam of thermal neutron or epithermal neutron beam
Radiation time is decided according to the data of boron concentration in the blood or tumour
tissue and neutron flux at the target point and surface of the brain.

REFERENCES
[1] HATANAKA H.: "Boron neutron capture therapy for tumours." In: Karim ABM. &
Laws, Jr. ER.(eds.) Glioma, Springer-Verlag, 1991 p233-249
[2] NAKAGAWA Y., HATANAKA H.: "Boron Neutron Capture Therapy — Clinical Brain
Tumour Studies". Neuro Oncology (33), pi05-115 (1997)
[3] HATANAKA H., NAKAGAWA Y.: "Clinical results of long-surviving brain tumour
patients who underwent boron neutron capture therapy", hit. J. of Radiation Oncology
Biology, Physics pi061-1066 (1994)
[4] KAGEJI T., NAKAGAWA Y.: "Pharmacokinetics and boron uptake of BSH
(Na2B12HllSH) in patients with intracranial tumours". J. Neuro Oncology, (33) 117-
130(1997)
[5] Y. NAKAGAWA: "Recent study of Boron neutron capture therapy in patients with
malignant brain tumour". Cancer Neutron Capture Therapy ed. by Yutaka Mishima,
Plenum, p717-724 (1996)
[6] NAKAGAWA Y.: "What were important factors in patients treated by BNCT in Japan?"
Advances in Neutron Capture Therapy Volume 1, Medicine and Physics_Aed by
B.Larsson, Elsvier, p65-70 (1997)

249