- Email: [email protected]
Radiation injury of boron neutron capture therapy using mixed epithermal- and thermal neutron beams in patients with malignant glioma
ARTICLE IN PRESS
Applied Radiation and Isotopes 61 (2004) 1063–1067
Radiation injury of boron neutron capture therapy using mixed epithermal- and thermal neutron beams in patients with malignant glioma T. Kagejia,*, S. Nagahiroa, Y. Mizobuchia, H. Toia, Y. Nakagawab, H. Kumadac a
Department of Neurosurgery, School of Medicine, University of Tokushima, Kuramoto-cho 3-18-15, 770, Tokushima, Japan b Department of Neurosurgery, National Kagawa Children’s Hospital, Kagawa, Japan c Department of Research Reactor, Tokai Research Establishment, Japan Atomic Energy Research Institute, Tokai, Ibaraki, Japan
Abstract The purpose of this study was to clarify the radiation injury in acute or delayed stage after boron neutron capture therapy (BNCT) using mixed epithermal- and thermal neutron beams in patients with malignant glioma. Eighteen patients with malignant glioma underwent mixed epithermal- and thermal neutron beam and sodium borocaptate between 1998 and 2004. The radiation dose (i.e. physical dose of boron n-alpha reaction) in the protocol used between 1998 and 2000 (Protocol A, n ¼ 8) prescribed a maximum tumor volume dose of 15 Gy. In 2001, a new dose-escalated protocol was introduced (Protocol B, n ¼ 4); it prescribes a minimum tumor volume dose of 18 Gy or, alternatively, a minimum target volume dose of 15 Gy. Since 2002, the radiation dose was reduced to 80–90% dose of Protocol B because of acute radiation injury. A new Protocol was applied to 6 glioblastoma patients (Protocol C, n ¼ 6). The average values of the maximum vascular dose of brain surface in Protocol A, B and C were 11.474.2 Gy, 15.771.2 and 13.973.6 Gy, respectively. Acute radiation injury such as a generalized convulsion within 1 week after BNCT was recognized in three patients of Protocol B. Delayed radiation injury such as a neurological deterioration appeared 3–6 months after BNCT, and it was recognized in 1 patient in Protocol A, 5 patients in Protocol B. According to acute radiation injury, the maximum vascular dose was 15.871.3 Gy in positive and was 12.674.3 Gy in negative. There was no significant difference between them. According to the delayed radiation injury, the maximum vascular dose was 13.873.8 Gy in positive and was 13.674.9 Gy in negative. There was no significant difference between them. The dose escalation is limited because most patients in Protocol B suffered from acute radiation injury. We conclude that the maximum vascular dose does not exceed over 12 Gy to avoid the delayed radiation injury, especially, it should be limited under 10 Gy in the case that tumor exists in speech center. r 2004 Elsevier Ltd. All rights reserved. Keywords: BNCT; BSH; Epithermal neutron beam; Glioblastoma; Radiation injury
1. Introduction In 1968, Hatanaka introduced sodium borocaptate (Na2B12H11SH: BSH) as the boron carrier for BNCT in Japan. More than 170 patients with malignant intracra*Corresponding author. Tel.: +81-886-31-3111; fax: +81886-32-9464. E-mail address: [email protected] (T. Kageji).
nial tumors, especially glioblastomas, received BNCT in combination with BSH and pure thermal neutron beam between 1968 and 1998. This clinical trial produced significant improvements in clinical outcomes. Hatanaka previously reported a detailed clinical data on patients treated by BNCT (Hatanaka and Nakagawa, 1994). We found that the most important factor related to clinical outcomes was the physical radiation dose of boron n-alpha reactions. However, evaluation of related factors, i.e. maximum neutron fluence and radiation
0969-8043/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2004.05.058
ARTICLE IN PRESS 1064
T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
time, demonstrated that they were not relevant to the clinical outcome (Nakagawa and Hatanaka, 1997). Therefore, to improve the clinical results, our efforts concentrated on escalating the radiation dose of boron n-alpha reaction in the clinical target volume. The clinical outcomes were favorable in patients whose GBM were located within a 4 cm depth from the brain surface. However, they were unsatisfactory in patients whose tumors were situated in deeper regions because neutron fluence delivery into deep regions was inadequate. Therefore, the epithermal neutron beam was developed at several international institutions to improve neutron delivery. Use of the mixed neutron beam can improve thermal neutron distribution in deeper sites, which in turn elevates the therapeutic efficacy of BNCT. Another effort to improve the beam penetration was creating an air filled cavity by debulking tumor tissue. It was the most effective and significantly improved neutron penetration. We have performed clinical trials using both mixed epithermal- and thermal neutron beams and BSH since 1998. While the follow-up period is relatively short and the number of patients small (n ¼ 18), our preliminary results warrant presentation of the design of our trials and the complications we encountered.
effect of the heavy-charged particles, and to evaluate the efficacy of BNCT, we took the physical dose of the boron n-alpha reaction. We determined the irradiation field in each patient, which depended on the tumor size. A collimator of 12 cm in diameter was introduced in most patients. A thin rubber balloon filled with air was placed into the prepared tumor cavity to maintain the size of the cavity during neutron irradiation and improve neutron distribution. Real-time neutron flux is detected from the gold wires, which were pulled out 15 min after the achievement of full power of reactor. The gold wires were inserted into or around the tumor tissue. Real-time boron concentration in blood is detected from prompt gamma spectroscopy. The vascular radiation dose (vascular) is the physical dose to the endothelial cells of the vasculature in the normal cortex near the brain surface. The vascular dose rate for the 10B(n; a) 7Li reaction is calculated by dividing by three the blood boron concentration (Kitao, 1975; Rydin et al., 1976). On retrospective analysis, we can obtain the BNCT radiation dose at each point in the irradiation field using boron concentration in tumor, irradiation time and the neutron flux of gold wires. 2.3. Radiation reaction
2. Material and method 2.1. Patients and protocols In 1998, new BNCT trials were conducted at Japan Atomic Energy Research Institute (JAERI) and Kyoto University Research Reactor Institute (KUR) using mixed epithermal- and thermal neutron beams. In patients (n ¼ 8) treated according to Protocol A (1998–2000), the maximum tumor volume dose was 15 Gy. In those (n ¼ 4) treated according to doseescalation Protocol B (2000–2001), the minimum tumor volume dose was 18 Gy or, alternatively, the minimum target volume dose was 15 Gy. Since 2002, the radiation dose was reduced to 80–90% dose of Protocol B because of acute radiation injury of Protocol B. A dose-reduced Protocol C was applied to 6 patients. In all protocols, the maximum vascular radiation dose to the brain surface did not exceed 15 Gy. Of the 18 patients enrolled in these clinical trials, 16 had a histological diagnosis of grade IV GBM and one patient each had anaplastic ependymoma and PNET. 2.2. Radiation dose The equation of physical radiation for BNCT was reported previously (Nakagawa et al., 2003). We applied a new concept of BNCT to the physical radiation dose of the boron–n alpha reaction. To compare the radiation
Assessment was by MRI obtained every 2 months after BNCT. Thallium SPECT was conducted when abnormal finding was recognized on follow-up MRI. Radiation necrosis was identified on MRI as low intensity on T1 and as abnormally enhanced areas on contrast-enhanced T1-weighted images, and as highintensity on T2-weighted images.
3. Results 3.1. BNCT neutron flux on brain surface (Table 1) The neutron flux (NF; n/cm2/s) was calculated by the equation; Neutron flux (NF)irradiation time (s) = neutron fluence (n/cm2). The maximum, minimum and average values of neutron flux at brain surface were shown in Table 1. There were no significant statistical differences in them.
Table 1 Neutron flux at the brain surface (Eþ09 n=cm2 =s)
Max. surrface Min. surrface Ave. surface
Protocol A
Protocol B
Protocol C
3.10871.465 2.75971.315 2.93371.380
2.79670.172 2.30070.176 2.59270.159
2.70570.853 1.92370.530 2.30570.656
ARTICLE IN PRESS T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
3.2. BNCT radiation dose (Table 2) BNCT radiation dose at the brain surface was shown in Table 2. We estimated the maximum, minimum and average value of the radiation dose retrospectively. The maximum vascular dose in Protocol A, B and C were 11.474.2, 15.771.2 and 13.974.6 Gy, respectively. The minimum vascular dose were 10.072.9, 12.971.7 and 9.872.1 Gy, respectively. The average vascular dose were 10.773.5, 14.471.2 and 11.973.4 Gy, respectively. The maximum vascular dose in Protocol B was approximately 1.1–1.4 times higher than Protocol A and C, however, the differences were not statistically significant. The gamma dose on the brain surface in Protocol A, B and C were 7.771.8, 8.172.3 and 6.073.3 Gy, respectively. 3.3. Clinical outcomes Acute radiation injury such as a generalized convulsion within 1 week after BNCT was recognized in three patients of Protocol B. Delayed radiation injury such as Table 2 Radiation dose at the brain surface (boron-n-alpha physical dose (Gy)) Protocol
A
B
C
Max.vascular dose Min. vascular dose Average vascular dose
11.474.2 * 10:072:9 10.773.5
15.771.2 12.971.7 14.471.2
13.974.6 9.872.1 * 11:973:4
1065
a neurological deterioration appeared 3–6 months after BNCT, and it was recognized in one patient in Protocol A, five patients in Protocol B. According to acute radiation injury, the maximum vascular dose was 15.871.3 Gy in positive, it was 12.674.3 Gy in negative. The former was about 1.3 times higher than the latter, however, there was no significant difference between them. According to the delayed radiation injury, the maximum vascular dose was 13.873.8 Gy in positive, it was 13.674.9 Gy in negative. There was no significant difference between them.
3.4. Illustrative cases Case YY: Lt. Frontal GBM (Protocol C) This 61-year-old female with a history of dull headache and motor aphasia was admitted on May 2, 2002. On admission, she had motor dominant aphasia and mild right hemiparesis. Cranial MRI showed a large, ring-enhanced mass in the left frontal lobe. She underwent craniotomy and subtotal resection of the tumor on June 7, 2002. BNCT was performed at JRR-4 on August 1, 2002. On retrospective analysis, the maximum, minimum and average neutron flux at brain surface was 1.88E + 9, 1.67E + 9 and 1.76E + 9 n/cm2/s, respectively. The irradiation time was calculated as 124 min. The maximum, minimum and average vascular radiation dose of the boron n-alpha reaction was 10.1, 8.9 and 9.4 Gy, respectively. During 5 months after BNCT, the neurological status was stable with only mild aphasia, however, she had a progressive right
Fig. 1. Case YY: 61 y.o. Lt. Frontal GBM. Upper left: pre op. Right: post op. Lower 8 months after BNCT Gd-MRI (left), T2-MRI (middle) and delayed image of Tl-SPECT (right).
ARTICLE IN PRESS 1066
T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
Fig. 2. Case SS: 22 y.o. Rt. Occipital GBM Upper: pre-op Gd-MRI (right), post-op Gd-MRI (left) Middle: 10 months after BNCT Gd-MRI (left), T2-MRI (middle) and delayed image of Tl-SPECT (right). Lower: Histopathological finding after BNCT.
hemiparesis. Follow-up MRI 8 months after BNCT demonstrated an enlargement of high intensity area in the left whole frontal lobe on T2 weighted image. A GdMRI showed less enlargement of enhanced area. On the other hand, delayed image of Thallium SPECT demonstrated a high accumulation of the lesion. We diagnosed that both delayed radiation necrosis and tumor recurrence were recognized. The consciousness of the patient gradually deteriorated; she died 19 months after diagnosis (Fig. 1). Case SS: Rt. Occipital GBM (Protocol C) This 22-year-old female with a history of dull headache was admitted on October 21, 2002. On admission, she had no neurological deficits. Cranial MRI showed a large, ring-enhanced mass in the right occipital lobe. She underwent craniotomy and subtotal resection of the tumor on October 29, 2002. BNCT was performed at JRR-4 on November 20, 2002. On retrospective
analysis, the maximum, minimum and average neutron flux at brain surface was 3.79E + 9, 2.97E + 9 and 3.34E + 9 n/cm2/s, respectively. The irradiation time was calculated as 65 min. The maximum, minimum and average vascular radiation dose of the boron n-alpha reaction was 16.7, 13.1 and 14.7 Gy, respectively. She had a severe headache and vomiting for around 10 months after BNCT. Follow-up MRI demonstrated that an enlargement of high intensity area in right whole occipital lobe on T2 weighted image. A Gd-MRI showed less enlargement of enhanced area. On the other hand, delayed image of thallium SPECT demonstrated a less accumulation of the lesion. We diagnosed that only delayed radiation necrosis was recognized. She underwent a re-craniotomy to decrease intracranial pressure, and the histopathological finding was diagnosed as only a radiation necrosis. The tissue has no tumor cells (Fig. 2).
ARTICLE IN PRESS T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
4. Discussion Our previous study reported that 19 (11.9%) of 159 patients treated by BNCT between 1977 and 2001 in Japan manifested radiation necrosis. We analyzed the related factors concerning radiation necrosis. There were no significant differences between patients with and without it with respect to age, irradiation time, boron concentration in blood and maximum neutron. The maximum vascular radiation dose in patients with necrosis was 21.8 7 8.1 Gy compared to 8.9 7 3.9 Gy in those without (po0:005) (Nakagawa et al., 2003). In this study, using mixed neutron beams in BNCT, acute radiation injury was recognized in three patients of Protocol B. Delayed radiation injury appeared 3–6 months after BNCT, and it was recognized one patient in Protocol A, five patients in Protocol B. According to acute radiation injury, the maximum vascular dose was 15.871.3 Gy in positive, it was 12.674.3 Gy in negative. The former was about 1.3 times higher than the latter, however, there was no significant difference between them. According to the delayed radiation injury, the maximum vascular dose was 13.873.8 Gy in positive, it was 13.674.9 Gy in negative. There was no significant difference between them. Since 1968, we have treated the patients in the condition that the maximum vascular dose was restricted in 15 Gy, and the total amount of gamma rays remained below 10 Gy. We increased the radiation dose at the target point to prevent the tumor recurrence. As a result, the maximum vascular dose at the brain surface in the mixed neutron beam era also was increased much more compared with that in the thermal neutron era. A dose escalation protocol (Protocol B) showed both acute and delayed radiation injury in the most patients. The maximum vascular dose in Protocol B was 15.771.2 Gy, it was 1.4 times higher than that of Protocol A. With respect to gamma dose, the mean values of Protocol A and B were similar (7.7 Gy versus
1067
8.1 Gy). We concluded that radiation injury in the acute stage was led with increasing boron n-alpha physical dose rather than gamma dose. So we interrupted the Potocol B, and reduced it to 80–90%. A new protocol C showed no acute radiation injury, however, two patients suffered from delayed radiation injury. In the case YY, where the maximum vascular dose was 10.1 Gy, the delayed radiation injury led to total aphasia because the tumor was near the Broca’s area. While the dose escalation can contribute to the improvement of survival rate, it results in the radiation injury. We conclude that the maximum vascular dose should be limited below 12 Gy to prevent delayed radiation injury as much as possible. It is optimal that the maximum vascular dose should be limited below 10 Gy when the tumor exists around the speech center.
References Hatanaka, H., Nakagawa, Y., 1994. Clinical results of longsurviving brain tumor patients who underwent boron neutron capture therapy. Int. J. Radiat. Oncol. Biol. Phys. 28, 1061–1066. Kitao, K., 1975. A method for calculating the absorbed dose near interface from 10B(n, a)7 Li reaction. Radiat. Res. 61, 304–315. Nakagawa, Y., Hatanaka, H., 1997. Boron neutron capture therapy-clinical brain tumor study. J. Neuro-Oncol. 33, 105–115. Nakagawa, Y., Pooh, K., Kobayshi, T., Kageji, T., Uyama, S., Matsumura, A., Kumada, H., 2003. Clinical review of Japanese experience with boron neutron capture therapy and a proposed strategy using epithermal neutron beams. J. Neuro-Oncol. 62, 87–99. Rydin, R.A., Deutsch, O.L., Murray, B.W., 1976. The effect of geometry on capillary wall dose for boron neutron capture therapy. Phys. Med. Biol. 21, 134–138.
Applied Radiation and Isotopes 61 (2004) 1063–1067
Radiation injury of boron neutron capture therapy using mixed epithermal- and thermal neutron beams in patients with malignant glioma T. Kagejia,*, S. Nagahiroa, Y. Mizobuchia, H. Toia, Y. Nakagawab, H. Kumadac a
Department of Neurosurgery, School of Medicine, University of Tokushima, Kuramoto-cho 3-18-15, 770, Tokushima, Japan b Department of Neurosurgery, National Kagawa Children’s Hospital, Kagawa, Japan c Department of Research Reactor, Tokai Research Establishment, Japan Atomic Energy Research Institute, Tokai, Ibaraki, Japan
Abstract The purpose of this study was to clarify the radiation injury in acute or delayed stage after boron neutron capture therapy (BNCT) using mixed epithermal- and thermal neutron beams in patients with malignant glioma. Eighteen patients with malignant glioma underwent mixed epithermal- and thermal neutron beam and sodium borocaptate between 1998 and 2004. The radiation dose (i.e. physical dose of boron n-alpha reaction) in the protocol used between 1998 and 2000 (Protocol A, n ¼ 8) prescribed a maximum tumor volume dose of 15 Gy. In 2001, a new dose-escalated protocol was introduced (Protocol B, n ¼ 4); it prescribes a minimum tumor volume dose of 18 Gy or, alternatively, a minimum target volume dose of 15 Gy. Since 2002, the radiation dose was reduced to 80–90% dose of Protocol B because of acute radiation injury. A new Protocol was applied to 6 glioblastoma patients (Protocol C, n ¼ 6). The average values of the maximum vascular dose of brain surface in Protocol A, B and C were 11.474.2 Gy, 15.771.2 and 13.973.6 Gy, respectively. Acute radiation injury such as a generalized convulsion within 1 week after BNCT was recognized in three patients of Protocol B. Delayed radiation injury such as a neurological deterioration appeared 3–6 months after BNCT, and it was recognized in 1 patient in Protocol A, 5 patients in Protocol B. According to acute radiation injury, the maximum vascular dose was 15.871.3 Gy in positive and was 12.674.3 Gy in negative. There was no significant difference between them. According to the delayed radiation injury, the maximum vascular dose was 13.873.8 Gy in positive and was 13.674.9 Gy in negative. There was no significant difference between them. The dose escalation is limited because most patients in Protocol B suffered from acute radiation injury. We conclude that the maximum vascular dose does not exceed over 12 Gy to avoid the delayed radiation injury, especially, it should be limited under 10 Gy in the case that tumor exists in speech center. r 2004 Elsevier Ltd. All rights reserved. Keywords: BNCT; BSH; Epithermal neutron beam; Glioblastoma; Radiation injury
1. Introduction In 1968, Hatanaka introduced sodium borocaptate (Na2B12H11SH: BSH) as the boron carrier for BNCT in Japan. More than 170 patients with malignant intracra*Corresponding author. Tel.: +81-886-31-3111; fax: +81886-32-9464. E-mail address: [email protected] (T. Kageji).
nial tumors, especially glioblastomas, received BNCT in combination with BSH and pure thermal neutron beam between 1968 and 1998. This clinical trial produced significant improvements in clinical outcomes. Hatanaka previously reported a detailed clinical data on patients treated by BNCT (Hatanaka and Nakagawa, 1994). We found that the most important factor related to clinical outcomes was the physical radiation dose of boron n-alpha reactions. However, evaluation of related factors, i.e. maximum neutron fluence and radiation
0969-8043/$ - see front matter r 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2004.05.058
ARTICLE IN PRESS 1064
T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
time, demonstrated that they were not relevant to the clinical outcome (Nakagawa and Hatanaka, 1997). Therefore, to improve the clinical results, our efforts concentrated on escalating the radiation dose of boron n-alpha reaction in the clinical target volume. The clinical outcomes were favorable in patients whose GBM were located within a 4 cm depth from the brain surface. However, they were unsatisfactory in patients whose tumors were situated in deeper regions because neutron fluence delivery into deep regions was inadequate. Therefore, the epithermal neutron beam was developed at several international institutions to improve neutron delivery. Use of the mixed neutron beam can improve thermal neutron distribution in deeper sites, which in turn elevates the therapeutic efficacy of BNCT. Another effort to improve the beam penetration was creating an air filled cavity by debulking tumor tissue. It was the most effective and significantly improved neutron penetration. We have performed clinical trials using both mixed epithermal- and thermal neutron beams and BSH since 1998. While the follow-up period is relatively short and the number of patients small (n ¼ 18), our preliminary results warrant presentation of the design of our trials and the complications we encountered.
effect of the heavy-charged particles, and to evaluate the efficacy of BNCT, we took the physical dose of the boron n-alpha reaction. We determined the irradiation field in each patient, which depended on the tumor size. A collimator of 12 cm in diameter was introduced in most patients. A thin rubber balloon filled with air was placed into the prepared tumor cavity to maintain the size of the cavity during neutron irradiation and improve neutron distribution. Real-time neutron flux is detected from the gold wires, which were pulled out 15 min after the achievement of full power of reactor. The gold wires were inserted into or around the tumor tissue. Real-time boron concentration in blood is detected from prompt gamma spectroscopy. The vascular radiation dose (vascular) is the physical dose to the endothelial cells of the vasculature in the normal cortex near the brain surface. The vascular dose rate for the 10B(n; a) 7Li reaction is calculated by dividing by three the blood boron concentration (Kitao, 1975; Rydin et al., 1976). On retrospective analysis, we can obtain the BNCT radiation dose at each point in the irradiation field using boron concentration in tumor, irradiation time and the neutron flux of gold wires. 2.3. Radiation reaction
2. Material and method 2.1. Patients and protocols In 1998, new BNCT trials were conducted at Japan Atomic Energy Research Institute (JAERI) and Kyoto University Research Reactor Institute (KUR) using mixed epithermal- and thermal neutron beams. In patients (n ¼ 8) treated according to Protocol A (1998–2000), the maximum tumor volume dose was 15 Gy. In those (n ¼ 4) treated according to doseescalation Protocol B (2000–2001), the minimum tumor volume dose was 18 Gy or, alternatively, the minimum target volume dose was 15 Gy. Since 2002, the radiation dose was reduced to 80–90% dose of Protocol B because of acute radiation injury of Protocol B. A dose-reduced Protocol C was applied to 6 patients. In all protocols, the maximum vascular radiation dose to the brain surface did not exceed 15 Gy. Of the 18 patients enrolled in these clinical trials, 16 had a histological diagnosis of grade IV GBM and one patient each had anaplastic ependymoma and PNET. 2.2. Radiation dose The equation of physical radiation for BNCT was reported previously (Nakagawa et al., 2003). We applied a new concept of BNCT to the physical radiation dose of the boron–n alpha reaction. To compare the radiation
Assessment was by MRI obtained every 2 months after BNCT. Thallium SPECT was conducted when abnormal finding was recognized on follow-up MRI. Radiation necrosis was identified on MRI as low intensity on T1 and as abnormally enhanced areas on contrast-enhanced T1-weighted images, and as highintensity on T2-weighted images.
3. Results 3.1. BNCT neutron flux on brain surface (Table 1) The neutron flux (NF; n/cm2/s) was calculated by the equation; Neutron flux (NF)irradiation time (s) = neutron fluence (n/cm2). The maximum, minimum and average values of neutron flux at brain surface were shown in Table 1. There were no significant statistical differences in them.
Table 1 Neutron flux at the brain surface (Eþ09 n=cm2 =s)
Max. surrface Min. surrface Ave. surface
Protocol A
Protocol B
Protocol C
3.10871.465 2.75971.315 2.93371.380
2.79670.172 2.30070.176 2.59270.159
2.70570.853 1.92370.530 2.30570.656
ARTICLE IN PRESS T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
3.2. BNCT radiation dose (Table 2) BNCT radiation dose at the brain surface was shown in Table 2. We estimated the maximum, minimum and average value of the radiation dose retrospectively. The maximum vascular dose in Protocol A, B and C were 11.474.2, 15.771.2 and 13.974.6 Gy, respectively. The minimum vascular dose were 10.072.9, 12.971.7 and 9.872.1 Gy, respectively. The average vascular dose were 10.773.5, 14.471.2 and 11.973.4 Gy, respectively. The maximum vascular dose in Protocol B was approximately 1.1–1.4 times higher than Protocol A and C, however, the differences were not statistically significant. The gamma dose on the brain surface in Protocol A, B and C were 7.771.8, 8.172.3 and 6.073.3 Gy, respectively. 3.3. Clinical outcomes Acute radiation injury such as a generalized convulsion within 1 week after BNCT was recognized in three patients of Protocol B. Delayed radiation injury such as Table 2 Radiation dose at the brain surface (boron-n-alpha physical dose (Gy)) Protocol
A
B
C
Max.vascular dose Min. vascular dose Average vascular dose
11.474.2 * 10:072:9 10.773.5
15.771.2 12.971.7 14.471.2
13.974.6 9.872.1 * 11:973:4
1065
a neurological deterioration appeared 3–6 months after BNCT, and it was recognized in one patient in Protocol A, five patients in Protocol B. According to acute radiation injury, the maximum vascular dose was 15.871.3 Gy in positive, it was 12.674.3 Gy in negative. The former was about 1.3 times higher than the latter, however, there was no significant difference between them. According to the delayed radiation injury, the maximum vascular dose was 13.873.8 Gy in positive, it was 13.674.9 Gy in negative. There was no significant difference between them.
3.4. Illustrative cases Case YY: Lt. Frontal GBM (Protocol C) This 61-year-old female with a history of dull headache and motor aphasia was admitted on May 2, 2002. On admission, she had motor dominant aphasia and mild right hemiparesis. Cranial MRI showed a large, ring-enhanced mass in the left frontal lobe. She underwent craniotomy and subtotal resection of the tumor on June 7, 2002. BNCT was performed at JRR-4 on August 1, 2002. On retrospective analysis, the maximum, minimum and average neutron flux at brain surface was 1.88E + 9, 1.67E + 9 and 1.76E + 9 n/cm2/s, respectively. The irradiation time was calculated as 124 min. The maximum, minimum and average vascular radiation dose of the boron n-alpha reaction was 10.1, 8.9 and 9.4 Gy, respectively. During 5 months after BNCT, the neurological status was stable with only mild aphasia, however, she had a progressive right
Fig. 1. Case YY: 61 y.o. Lt. Frontal GBM. Upper left: pre op. Right: post op. Lower 8 months after BNCT Gd-MRI (left), T2-MRI (middle) and delayed image of Tl-SPECT (right).
ARTICLE IN PRESS 1066
T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
Fig. 2. Case SS: 22 y.o. Rt. Occipital GBM Upper: pre-op Gd-MRI (right), post-op Gd-MRI (left) Middle: 10 months after BNCT Gd-MRI (left), T2-MRI (middle) and delayed image of Tl-SPECT (right). Lower: Histopathological finding after BNCT.
hemiparesis. Follow-up MRI 8 months after BNCT demonstrated an enlargement of high intensity area in the left whole frontal lobe on T2 weighted image. A GdMRI showed less enlargement of enhanced area. On the other hand, delayed image of Thallium SPECT demonstrated a high accumulation of the lesion. We diagnosed that both delayed radiation necrosis and tumor recurrence were recognized. The consciousness of the patient gradually deteriorated; she died 19 months after diagnosis (Fig. 1). Case SS: Rt. Occipital GBM (Protocol C) This 22-year-old female with a history of dull headache was admitted on October 21, 2002. On admission, she had no neurological deficits. Cranial MRI showed a large, ring-enhanced mass in the right occipital lobe. She underwent craniotomy and subtotal resection of the tumor on October 29, 2002. BNCT was performed at JRR-4 on November 20, 2002. On retrospective
analysis, the maximum, minimum and average neutron flux at brain surface was 3.79E + 9, 2.97E + 9 and 3.34E + 9 n/cm2/s, respectively. The irradiation time was calculated as 65 min. The maximum, minimum and average vascular radiation dose of the boron n-alpha reaction was 16.7, 13.1 and 14.7 Gy, respectively. She had a severe headache and vomiting for around 10 months after BNCT. Follow-up MRI demonstrated that an enlargement of high intensity area in right whole occipital lobe on T2 weighted image. A Gd-MRI showed less enlargement of enhanced area. On the other hand, delayed image of thallium SPECT demonstrated a less accumulation of the lesion. We diagnosed that only delayed radiation necrosis was recognized. She underwent a re-craniotomy to decrease intracranial pressure, and the histopathological finding was diagnosed as only a radiation necrosis. The tissue has no tumor cells (Fig. 2).
ARTICLE IN PRESS T. Kageji et al. / Applied Radiation and Isotopes 61 (2004) 1063–1067
4. Discussion Our previous study reported that 19 (11.9%) of 159 patients treated by BNCT between 1977 and 2001 in Japan manifested radiation necrosis. We analyzed the related factors concerning radiation necrosis. There were no significant differences between patients with and without it with respect to age, irradiation time, boron concentration in blood and maximum neutron. The maximum vascular radiation dose in patients with necrosis was 21.8 7 8.1 Gy compared to 8.9 7 3.9 Gy in those without (po0:005) (Nakagawa et al., 2003). In this study, using mixed neutron beams in BNCT, acute radiation injury was recognized in three patients of Protocol B. Delayed radiation injury appeared 3–6 months after BNCT, and it was recognized one patient in Protocol A, five patients in Protocol B. According to acute radiation injury, the maximum vascular dose was 15.871.3 Gy in positive, it was 12.674.3 Gy in negative. The former was about 1.3 times higher than the latter, however, there was no significant difference between them. According to the delayed radiation injury, the maximum vascular dose was 13.873.8 Gy in positive, it was 13.674.9 Gy in negative. There was no significant difference between them. Since 1968, we have treated the patients in the condition that the maximum vascular dose was restricted in 15 Gy, and the total amount of gamma rays remained below 10 Gy. We increased the radiation dose at the target point to prevent the tumor recurrence. As a result, the maximum vascular dose at the brain surface in the mixed neutron beam era also was increased much more compared with that in the thermal neutron era. A dose escalation protocol (Protocol B) showed both acute and delayed radiation injury in the most patients. The maximum vascular dose in Protocol B was 15.771.2 Gy, it was 1.4 times higher than that of Protocol A. With respect to gamma dose, the mean values of Protocol A and B were similar (7.7 Gy versus
1067
8.1 Gy). We concluded that radiation injury in the acute stage was led with increasing boron n-alpha physical dose rather than gamma dose. So we interrupted the Potocol B, and reduced it to 80–90%. A new protocol C showed no acute radiation injury, however, two patients suffered from delayed radiation injury. In the case YY, where the maximum vascular dose was 10.1 Gy, the delayed radiation injury led to total aphasia because the tumor was near the Broca’s area. While the dose escalation can contribute to the improvement of survival rate, it results in the radiation injury. We conclude that the maximum vascular dose should be limited below 12 Gy to prevent delayed radiation injury as much as possible. It is optimal that the maximum vascular dose should be limited below 10 Gy when the tumor exists around the speech center.
References Hatanaka, H., Nakagawa, Y., 1994. Clinical results of longsurviving brain tumor patients who underwent boron neutron capture therapy. Int. J. Radiat. Oncol. Biol. Phys. 28, 1061–1066. Kitao, K., 1975. A method for calculating the absorbed dose near interface from 10B(n, a)7 Li reaction. Radiat. Res. 61, 304–315. Nakagawa, Y., Hatanaka, H., 1997. Boron neutron capture therapy-clinical brain tumor study. J. Neuro-Oncol. 33, 105–115. Nakagawa, Y., Pooh, K., Kobayshi, T., Kageji, T., Uyama, S., Matsumura, A., Kumada, H., 2003. Clinical review of Japanese experience with boron neutron capture therapy and a proposed strategy using epithermal neutron beams. J. Neuro-Oncol. 62, 87–99. Rydin, R.A., Deutsch, O.L., Murray, B.W., 1976. The effect of geometry on capillary wall dose for boron neutron capture therapy. Phys. Med. Biol. 21, 134–138.