Early clinical experience utilizing scintillator with optical fiber (SOF) detector in clinical boron neutron capture therapy: its issues and solutions
© The Author(s). 2016
Received: 16 October 2015
Published: 9 August 2016
Real-time measurement of thermal neutrons in the tumor region is essential for proper evaluation of the absorbed dose in boron neutron capture therapy (BNCT) treatment. The gold wire activation method has been routinely used to measure the neutron flux distribution in BNCT irradiation, but a real-time measurement using gold wire is not possible. To overcome this issue, the scintillator with optical fiber (SOF) detector has been developed. The purpose of this study is to demonstrate the feasibility of the SOF detector as a real-time thermal neutron monitor in clinical BNCT treatment and also to report issues in the use of SOF detectors in clinical practice and their solutions.
Material and methods
Clinical measurements using the SOF detector were carried out in 16 BNCT clinical trial patients from December 2002 until end of 2006 at the Japanese Atomic Energy Agency (JAEA) and Kyoto University Research Reactor Institute (KURRI).
The SOF detector worked effectively as a real-time thermal neutron monitor. The neutron fluence obtained by the gold wire activation method was found to differ from that obtained by the SOF detector. The neutron fluence obtained by the SOF detector was in better agreement with the expected fluence than with gold wire activation. The estimation error for the SOF detector was small in comparison to the gold wire measurement. In addition, real-time monitoring suggested that the neutron flux distribution and intensity at the region of interest (ROI) may vary due to the reactor condition, patient motion and dislocation of the SOF detector.
Clinical measurements using the SOF detector to measure thermal neutron flux during BNCT confirmed that SOF detectors are effective as a real-time thermal neutron monitor. To minimize the estimation error due to the displacement of the SOF probe during treatment, a loop-type SOF probe was developed.
Boron neutron capture therapy (BNCT) is the combination of external irradiation (thermal neutrons or epithermal neutrons) and internal irradiation (α particle and lithium nuclei). In other words, a boron 10B compound is selectively introduced into tumor cells and is externally irradiated with thermal or epithermal neutrons. The thermal neutrons interact with 10B in the tumor cells and result in high linear energy transfer (LET) α and lithium 7Li particles through boron neutron capture reaction 10B(n,α)7Li. The very short range of α particles (~8 μm) and 7Li particles (~5 μm) helps to destroy 10B loaded tumor cells at the cellular level with minimum damage to neighboring 10B unloaded normal cells. The first clinical trial of BNCT was carried out by Farr et al. at Brookhaven National Laboratory (BNL) in 1951 but the results were not satisfactory . Later, Hatanaka et al. performed clinical trials on 13 brain tumor patients at Hitachi Training Reactor (HTR) from 1968 to 1975. The encouraging outcomes stimulated interest in BNCT . In 1987, Mishima et al. started a clinical trial of BNCT for malignant melanoma using Kyoto university reactor KUR . In a clinical study carried out by Nakawaga and Hatanaka on 149 patients treated with BNCT between August 1968 and April 1995 considered BNCT as an ideal treatment for malignant brain tumors because of the quality of life after treatment . Currently, BNCT offers the most effective treatment for primary and metastatic tumors, specifically glioblastoma multiforme and malignant melanomas for which effective therapy has not yet been developed . However, the procedure for BNCT is one of the most complex cancer treatment modalities and the effectiveness of this therapy depends on the neutron and boron distribution . In clinical BNCT, the continuous monitoring of the neutron flux distribution is necessary because the distribution and intensity may vary depending on the reactor condition or the physical feature of the patients . Their accurate and real-time assessment during irradiation is also essential for the quality assurance of the treatment.
Detectors that are used for thermal neutron monitoring include 10BF3 or 3He gas counters, ionization chambers, fission chambers and proton-recoil spectrometers . The gold wire or foil activation method is also used for the same purpose. Gas-filled detectors and activation spectrometry are considered as the primary tool for neutron beam dosimetry and monitoring in BNCT. Gas filled detectors are commonly used for phantom measurement but for in-vivo measurement they are not frequently used because of their large physical size and high sensitivity to electric noise. Gold wire is the most common method used to measure thermal neutron fluence in-vivo dosimetry. Other radiation technologies have also been developed for the same purpose, such as scintillators, thermoluminescent dosimeters (TLD), gel detectors and self-powered neutron detectors [9–12].
In most BNCT clinical cases, thermal neutron fluence has been measured by means of the gold wire activation method. However, the real-time measurement of thermal neutron flux using this method is not possible since neutron activation of the gold wire alone requires at least several minutes. In our previous work, we developed a plastic scintillator with optical fiber (SOF) for online thermal neutron measurements in BNCT [9, 13]. Details of the characteristics and properties of the SOF detector can be found in our former study. Initially, we used a boron compound as a neutron converter in the scintillator and two clinical measurements were performed using a boron loaded SOF detector. The output this detector showed good agreement with gold wire measurements but the measured value comprised of much electric noise and was latter replaced by a LiF mixed scintillator.
In the present research, measurements were carried out on a total of 16 patients using the SOF detector, until the end of 2006 at Japan Atomic Energy Agency (JAEA) / JRR4 and Kyoto University Research Reactor Institute (KURRI). In the first two clinical cases boron loaded scintillators were used, and in the remaining 14 cases LiF mixed scintillators were used. The main purpose of this study was to demonstrate the feasibility of the SOF detector as a real-time thermal neutron monitor during BNCT treatment based on results of clinical measurements. We also report in this paper about issues we experienced in the use of SOF detectors in clinical BNCT practice and their solutions.
Materials and methods
The photon signals are relayed through the optical fiber to the Photon Counting Unit and then converted into 30 ns-width TTL pulses. The pulse counts are sent to a personal computer via a universal serial bus (USB) connection. When using a plastic scintillator for counting neutrons, the signals in the detector can be produced by alpha particles, Li nuclei, recoil protons and gamma rays. It is necessary to clearly differentiate the signals produced by these particles in order to correctly estimate thermal neutron flux. The signal from gamma rays can be the main source of noise for SOF detector. Although a plastic scintillator hardly causes photoelectric effect, a signal from the Compton scatter of high-energy gamma rays contributes significantly to the total signal measured by the detector. This gamma ray contribution can be minimized if a very small detector is used for measurement. To account for the gamma ray and fast neutron signals, scintillators without 6LiF were used, which include only gamma ray and fast neutron signals. The contributions from the gamma rays and fast neutrons can be delineated and corrected based on the difference between the signals obtained from the scintillator without 6LiF and the signals from the scintillator with 6LiF.
In a few cases, a single SOF detector has been adopted to increase the number of monitoring points. A single SOF detector has almost the same composition and characteristics as the paired detector except for the gamma-ray and fast-neutron signal correction. Since the thermal neutron flux measured by the single SOF detector was similar as that of the paired detector, the single SOF was as effective as a relative neutron fluence monitor. Furthermore, because the cross-section of 6Li is almost proportional to that of 10B for neutron energy range in reactor-based BNCT, the reaction rate of 6Li is proportional to the 10B dose in the tumor.
SOF measurement in clinical use
Here, C’ + and C’ − are the measured gamma ray counts for the detectors with and without the neutron converter, respectively. The response factors R g+ and R g- were obtained from the measured counts of both detectors when only a gamma-ray field was used following Eqs. (3) and (4). These correction factors were determined from measurements using an intense pure gamma-ray source such as 137Cs.
Expected fluence and estimation error
Here, the full irradiation time was determined from constraints of skin dose or vascular dose and minimum tumor dose. The “pull-out” and “final” values were based on the SOF detector placed on the patient skin.
A summary of clinical trials in BNCT
Irrad. time [min.]
Expected fluence [n/cm2]
Estimation error (%)
1.01 × 1011
7.72 × 1011
1.12 × 1011
9.58 × 1011
8.93 × 1011
1.29 × 1011
5.02 × 1011
1.17 × 1011
4.90 × 1011
4.68 × 1011
2.18 × 1011
1.17 × 1012
2.09 × 1011
1.20 × 1012
1.11 × 1012
1.39 × 1011
6.22 × 1011
1.14 × 1011
4.82 × 1011
6.30 × 1011
1.18 × 1011
6.92 × 1011
2.03 × 1011
1.09 × 1012
1.22 × 1012
1.24 × 1011
7.25 × 1011
1.36 × 1011
7.86 × 1011
1.63 × 1012
1.41 × 1011
8.27 × 1011
3.01 × 1011
1.93 × 1012
1.80 × 1012
1.55 × 1011
6.02 × 1011
3.17 × 1011
1.21 × 1012
1.27 × 1012
7.86 × 1011
4.60 × 1012
9.27 × 1010
5.89 × 1011
5.56 × 1011
2.64 × 1011
1.03 × 1012
3.05 × 1012
1.25 × 1013
1.41 × 1013
4.19 × 1011
8.62 × 1011
3.14 × 1012
7.03 × 1012
7.13 × 1012
1.31 × 1010
2.68 × 1010
3.05 × 1012
6.49 × 1012
6.52 × 1012
6.61 × 1011
3.00 × 1012
8.30 × 1011
4.59 × 1012
4.26 × 1012
2.08 × 1011
1.17 × 1012
3.11 × 1011
1.91 × 1012
1.87 × 1012
4.43 × 1011
1.06 × 1012
2.91 × 1012
4.77 × 1012
4.85 × 1012
5.76 × 1011
1.54 × 1012
2.93 × 1012
5.51 × 1012
5.67 × 1012
A real-time monitoring by SOF detector in BNCT treatment
According to Table 1, in case 4, it was realized that the total amount of irradiation was 18.66 % lower than expected. The decline in the thermal neutron flux was observed in real-time as shown in Fig. 3d. Both the output of the monitor at the center of the ROI and the monitor near the edge of the ROI declined continuously. However, the same drastic change in the thermal flux was not observed by another SOF monitor placed in front of the collimator. This was the highest decline in the neutron flux in the entire clinical measurements. The continuous decline in the neutron flux shown by detector 1 was most probably due to the change in the position of the patient during irradiation.
Errors seem to decrease in cases 6 and 9. In case 6 (Fig. 3e), the neutron flux monitor in front of the collimator remained fairly constant. The estimation error 0.47 % in this case was most probably due to the slight change in the flux at the peripheral of the ROI. On the other hand, in case 9 (Fig. 3f) both the monitor in front of the collimator and near the periphery of interest showed decline in the thermal neutron flux. Therefore, in this case the neutron flux at the ROI was affected by the patient motion and reactor power fluctuation. In all above cases, the dislocation of SOF detector from the original position was not observed.
Figure 3g shows the location of the SOF detector probe before and after irradiation for case 14. As shown in Fig. 3g, the probe was inside a red circle before irradiation; however, after irradiation the probe was located outside of the red circle. The decline in neutron flux near the edge of the collimator was observed after 3,600 s of irradiation. Similarly, the real-time monitor at ROI showed decline in the flux after 4,800 s of irradiation. In this case, the dislocation of the probe from the original position affected the SOF measurements.
Measurement uncertainty of gold wire and SOF detector
The measurement uncertainty for the SOF detector can be deduced from Eqs. (1), (2), (6) and (7). From Eq. (7), the uncertainties for thermal neutron flux at 108 n/cm2/s and 109 n/cm2/s are 0.53 % and 0.17 %, respectively. Here, we used parameters for response ratio Rg+/Rg- = 1, gamma-ray contribution ratio C−/C+ = 0.1 and the response factor Rn+ = 2,072 n/cm2/counts from our previous work . The estimated uncertainty does not contain calibration uncertainty for absolute measurement. Similarly, measurement uncertainties for the gold wire activation method at KURRI and JRR4 were estimated as 5.82 % ± 1.73 % and 1.32 % ± 0.40 %, respectively for estimating thermal neutron flux at the beginning. The gold wire measurements at KURRI were performed on the patient skin where the average thermal neutron flux was 3.21 × 108 n/cm2/s. At JRR4, gold wire measurements were performed inside the beam port where the average thermal neutron flux was 3.83 × 109 n/cm2/s. This is 12 times higher than at KURRI. The measurement acquisition times for the activation were 60 s at KURRI and 103.8 s (30–200 s) at JRR4.
From the result section, we observed that the real-time monitoring of the thermal neutron distribution was possible with the help of the SOF detector. The gold wire method cannot be used for the same purpose because it provides only retrospective and integrated information on the neutron flux distribution. The difference in the estimated neutron fluence by the gold wire method and SOF detector was likely due to the difference in the location of the detector and the gold wire. The gold wire was placed on the patient skin at KURRI and inside the beam port at JRR4. Even at the skin surface, the position of the SOF detector did not align exactly with the position of gold wire 1 (pulled out after 15 min) or gold wire 2 (irradiated to the end of the treatment).
Generally, the gold wire is pulled out after the first 15 min of irradiation. The location of the SOF detector and the gold wire is also different (in case of JRR 4, gold wire was placed inside the port) and even on the patient skin the position of the SOF detector did not align exactly with the position of gold wire 1. This makes direct comparison between the SOF detector and gold wire measurement difficult. Similarly, TLD is only useful for gamma dosimetry. Thus, in the current research the neutron flux measured by the SOF detector was compared with the expected fluence because it was similar to that of the treatment planning system. However, the test level of the SOF detector can be further improved by comparing the results with simulation techniques which are capable of modeling complex geometries.
Issues and solutions in the clinical use of the scintillator with optical fiber (SOF) detector
We observed that the measurement uncertainty of the SOF detector was quite small and its contribution in SOF detector estimation error was not of much significance. The main source of error in SOF detector measurement at the ROI is due to the variation in the reactor condition, patient motion and dislocation of the SOF detector from the original position. Thus, the measured data should be carefully reviewed.
The displacement of the SOF detector probe from the patient skin happened during treatment in case 14 as shown in Fig. 3g. Due to the displacement of the probe, it was impossible to estimate the measured position and thus the measured value will be of less importance. A strong adhesive tape could not be used during treatment since it was harsh on the patient’s skin. The displacement of the probe also occurred as the adhesive tape was weakened by patient’s sweat during treatment. Close adhesion to the skin is especially difficult for a patient who had a craniotomy because of difficulty of shaving around the irradiated area. The neutron flux decreases as the distance from the central beam axis increases, as shown in Fig. 4. Thus, in case 14, the SOF detector displaced away from the beam axis (Fig. 3g) during irradiation and a small decline in the neutron flux was recorded by the real-time monitor (Fig. 3h) of detector 1 and 2 around 4,800 and 3,600 s, respectively. This affected the accuracy of the detected neutron flux and consequently affected the SOF measurement.
SOF detector at low neutron field
Soft error rate (SER) may be induced when a semiconductor device on a pacemaker is subjected to irradiation.. The maximum tolerable cumulative radiation dose for safe operation of a pacemaker depends highly on the pacemaker type, model and the dose rate . SER of a semiconductor device due to thermal neutron flux of 105 n/cm2/s from a nuclear reactor is about 5 times higher than environmental level thermal neutron flux . Since, the SER of the environmental thermal neutron flux is very low, the potential effect on the pacemaker by the thermal neutron field may have been negligible in this case. However, the current manuscript does not aim to determine the safe irradiation limit of pacemakers during BNCT treatment.
Comparison between boron loaded and LiF mixed SOF detector
Clinical measurements using the SOF detector to measure thermal neutron flux during BNCT treatment confirmed that SOF detectors were effective as a real-time thermal neutron monitor. The real-time monitoring of the thermal neutron by SOF detector suggests that the neutron flux distribution and intensity at the ROI vary due to the reactor condition, patient motion and dislocation of the SOF detector. To minimize the detector displacement problem, a loop-type probe was developed. The authors believe the presented work will contribute towards the field of on-line neutron monitoring in clinical BNCT irradiation.
10B, boron; 7Li, lithium; BNCT, boron neutron capture therapy; JRR4, Japan Atomic Energy Agency reactor No.4; KUR, Kyoto University Reactor; KURRI, Kyoto University Research Reactor Institute; LET, linear energy transfer; LiF, lithium fluoride; ROI, region of interest; SER, soft error rate; SOF, scintillator with optical fiber; TTL, transistor-transistor logic; USB, universal Serial Bus; α, alpha
The authors would like to thank Kenneth Sutherland for English check as native speaker.
This research was supported by a Grant-in-Aid for Young Scientists (A) (#16689022) by the Ministry of Education, Culture, Sports, Science and Technology of Japan.
Availability of data and material
Data are summarized here in detail and will not be shared.
MI developed the concept, performed the experiment, reviewed and analyzed the data, drafted the manuscript. AM, TY, JH, SM, IK, KO provided patient related data and photographs. YS, HK, provided data related to the treatment planning and the gold wire measurement. SJS provided important discussion, helped in analyzing the data and drafting the manuscript. All authors read and approved the final manuscript.
The authors declare that they have no competing interest.
Consent for publication
Ethics approval and consent to participate
This study has received ethical approval from Prof. Dr. Hiroyuki Date, Dean of Graduate School of Health Science, Hokkaido University (reference number 15–78).
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Farr LE, Sweet WH, Robertson J, Foster C, Locksley H, Sutherland D, et al. Neutron capture therapy with boron in the treatment of glioblastoma multiforme. Am J Roentgenol Radium Ther Nucl Med. 1954;71(2):279-93.Google Scholar
- Hatanaka H. A revised boron-neutron capture therapy for malignant brain tumors. J Neurol. 1975;209(2):81–94.View ArticlePubMedGoogle Scholar
- Mishima Y, Honda C, Ichihashi M, Obara H, Hiratsuka J, Fukuda H, et al. Treatment of malignant melanoma by single thermal neutron capture therapy with melanoma-seeking 10 B-compound. Lancet. 1989;334(8659):388–9.View ArticleGoogle Scholar
- Nakagawa Y, Hatanaka H. Boron neutron capture therapy: clinical brain tumor studies. J Neuro-Oncol. 1997;33(1–2):105–15.View ArticleGoogle Scholar
- Barth RF, Soloway AH, Fairchild RG, Brugger RM. Boron neutron capture therapy for cancer. Realities and prospects. Cancer. 1992;70(12):2995–3007.View ArticlePubMedGoogle Scholar
- Yamamoto T, Nakai K, Matsumura A. Boron neutron capture therapy for glioblastoma. Cancer Lett. 2008;262(2):143–52.View ArticlePubMedGoogle Scholar
- Ito Y, Katano G, Harano H, Matsumoto T, Uritani A, Kudo K, et al. Development of a tiny neutron probe with an optical fibre for BNCT. Radiat Prot Dosim. 2004;110(1–4):619–22.View ArticleGoogle Scholar
- Peurrung AJ. Recent developments in neutron detection. Nucl Instrum Methods Phys Res, Sect A. 2000;443(2):400–15.View ArticleGoogle Scholar
- Ishikawa M, Ono K, Sakurai Y, Unesaki H, Uritani A, Bengua G, et al. Development of real-time thermal neutron monitor using boron-loaded plastic scintillator with optical fiber for boron neutron capture therapy. Appl Radiat Isot. 2004;61(5):775–9.View ArticlePubMedGoogle Scholar
- Aschan C, Toivonen M, Savolainen S, Seppälä T, Auterinen I. Epithermal neutron beam dosimetry with thermoluminescence dosemeters for boron neutron capture therapy. Radiat Prot Dosim. 1999;81(1):47–56.View ArticleGoogle Scholar
- Gambarini G, Birattari C, Colombi C, Pirola L, Rosi G. Fricke gel dosimetry in boron neutron capture therapy. Radiat Prot Dosim. 2002;101(1–4):419–22.View ArticleGoogle Scholar
- Miller M, Mariani L, Gonçalves-Carralves M, Skumanic M, Thorp S. Implantable self-powered detector for on-line determination of neutron flux in patients during NCT treatment. Appl Radiat Isot. 2004;61(5):1033–7.View ArticlePubMedGoogle Scholar
- Ishikawa M, Kumada H, Yamamoto K, Kaneko J, Bengua G, Unesaki H, et al. Development of a wide-range paired scintillator with optical fiber neutron monitor for BNCT irradiation field study. Nucl Instrum Methods Phys Res, Sect A. 2005;551(2):448–57.View ArticleGoogle Scholar
- Mouton J, Haug R, Bridier A, Dodinot B, Eschwege F. Influence of high-energy photon beam irradiation on pacemaker operation. Phys Med Biol. 2002;47(16):2879.View ArticlePubMedGoogle Scholar
- Kobayashi H, Shiraishi K, Tsuchiya H, Motoyoshi M, Usuki H, Nagai Y et al., editors. Soft errors in SRAM devices induced by high energy neutrons, thermal neutrons and alpha particles. Electron Devices Meeting, 2002. IEDM’02. International; 2002:337-40. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=1175847&tag=1
- Fisher B, Abdurashitov J, Coakley K, Gavrin V, Gilliam D, Nico J, et al. Fast neutron detection with 6 Li-loaded liquid scintillator. Nucl Instrum Methods Phys Res, Sect A. 2011;646(1):126–34.View ArticleGoogle Scholar