Radiotherapy and Oncology 88 (2008) 183–191
`www.thegreenjournal.com
`
`Phase II trial of BNCT
`
`Boron neutron capture therapy (BNCT) for glioblastoma
`multiforme: A phase II study evaluating a prolonged high-dose
`of boronophenylalanine (BPA)q
`
`Roger Henrikssona,*, Jacek Capalab,c, Annika Michanekd, Sten-A˚ke Lindahle,
`Leif G. Salfordf, Lars Franze´na, Erik Blomquistg,
`Jan-Erik Westlinh, A. Tommy Bergenheimi
`
`aDepartment of Radiation Sciences & Oncology, Umea˚ University, Umea˚, Sweden, bStudsvik Medical Co., Nyko¨ping, Sweden, cUnit
`for Biomedical Radiation Sciences, Rudbeck Laboratory, Uppsala University, Uppsala, Sweden, dDepartment of Oncology, Sahlgrenska
`University Hospital, Gothenburg, Sweden, eDepartment of Oncology, Karlstad Central Hospital, Sweden, fDepartment of Neurosurgery
`and Radiation Physics, Lund University Hospital, Lund, Sweden, gDepartment of Radiation Oncology, Uppsala University Hospital,
`Uppsala, Sweden, hDepartment of Oncology, Central Hospital, Eskilstuna, Sweden, iDepartment of Neurosurgery, Umea˚ University,
`Umea˚, Sweden
`
`Abstract
`
`Background and purpose: To evaluate the efficacy and safety of boron neutron capture therapy (BNCT) for
`glioblastoma multiforme (GBM) using a novel protocol for the boronophenylalanine–fructose (BPA-F) infusion.
`Patient and methods: This phase II study included 30 patients, 26–69 years old, with a good performance status of
`which 27 have undergone debulking surgery. BPA-F (900 mg BPA/kg body weight) was given i.v. over 6 h. Neutron
`irradiation started 2 h after the completion of the infusion. Follow-up reports were monitored by an independent clinical
`research institute.
`Results: The boron-blood concentration during irradiation was 15.2–33.7 lg/g. The average weighted absorbed dose
`to normal brain was 3.2–6.1 Gy (W). The minimum dose to the tumour volume ranged from 15.4 to 54.3 Gy (W). Seven
`patients suffered from seizures, 8 from skin/mucous problem, 5 patients were stricken by thromboembolism and 4 from
`abdominal disturbances in close relation to BNCT. Four patients displayed 9 episodes of grade 3–4 events (WHO). At the
`time for follow-up, minimum ten months, 23 out of the 29 evaluable patients were dead. The median time from BNCT
`treatment to tumour progression was 5.8 months and the median survival time after BNCT was 14.2 months. Following
`progression, 13 patients were given temozolomide, two patients were re-irradiated, and two were re-operated. Patients
`treated with temozolomide lived considerably longer (17.7 vs. 11.6 months). The quality of life analysis demonstrated a
`progressive deterioration after BNCT.
`Conclusion: Although, the efficacy of BNCT in the present protocol seems to be comparable with conventional
`radiotherapy and the treatment time is shorter, the observed side effects and the requirement of complex infrastructure
`and higher resources emphasize the need of further phase I and II studies, especially directed to improve the
`accumulation of 10B in tumour cells.
`
`c 2008 Elsevier Ireland Ltd. All rights reserved. Radiotherapy and Oncology 88 (2008) 183–191.
`
`Keywords: Glioblastoma; BNCT; Quality of life; Toxicity; Survival
`
`New approaches for treatment of glioblastoma multi-
`forme (GBM) are urgently needed. Boron neutron capture
`therapy (BNCT) represents an interesting modality for selec-
`tive irradiation of tumour tissue including infiltrating glioma
`cells. Following improvements in neutron beams and boron
`carriers, clinical trials of closed-skull BNCT using BPA-F
`
`(boronphenyalanine–fructose) and epithermal neutrons
`were initiated at several centers. The development of the
`BNCT concept and the present knowledge have carefully
`been reviewed by R. Barth in 2003 [2]. The BNCT therapy
`is seemingly well tolerated [1,9]. The clinical outcome of al-
`most 150 patients treated in non-randomized studies seems
`to compare with conventional and more prolonged radio-
`therapy [13]. It is noteworthy that the results obtained so
`far showed no correlation between the nominal radiation
`
`q The study was conducted as a project within the Swedish Brain
`Tumour Study Group.
`
`0167-8140/$ - see front matter c 2008 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.radonc.2006.04.015
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1074 - Page 1 of 9
`
`

`
`184
`
`BNCT and glioblastoma
`
`doses delivered to target volume and survival. One possible
`explanation of this unexpected outcome might be that the
`administration of BPA-F as a 2-h intravenous infusion at a
`dose of 250–330 mg BPA/kg body weight was effective in
`delivery of boron to only a subpopulation of tumour cells.
`This conclusion was supported by electron spectroscopy
`measurements of boron concentration in tumour cells dis-
`persed in the normal brain obtained after debulking surgery
`[22]. After a 2-h infusion of BPA, the boron concentration in
`invading tumour cells was only 37% of that in the bulk of the
`tumour, while the ratio increased to 71% and 100% when the
`infusion time was increased to 6 and 24 h, respectively. The
`advantage of longer infusion times was also supported by
`pre-clinical studies demonstrating advantages of 6-h infu-
`sion compared to 2 h [16,24]. The higher boron concentra-
`tion provides additional advantages,
`such as
`shorter
`irradiation time and lower neutron fluency to become
`needed to deliver the prescribed doses, which, in turn, re-
`duce the background radiation dose from the proton recoil
`and the secondary radiation produced by neutron capture
`in hydrogen and in nitrogen. However, the potential risk
`that a longer infusion time results in a higher accumulation
`in normal cells and greater radiation toxicity must also be
`evaluated. These observations found the rationale to in-
`crease the infusion time to 6 h and the total dose of BPA
`to 900 mg/kg body weight in the present clinical trials at
`the Studsvik facility in Sweden (for details see [8]).
`The objective of this phase II study at the Studsvik BNCT
`facility was to assess the degree of tumour control, survival,
`the safety of BNCT using a 6-h infusion of BPA-F, and the
`evaluation of the quality of life in patients with glioblas-
`toma multiforme.
`
`Patients and methods
`Patients characteristics
`Between March 2001 and February 2003, 30 patients with
`verified glioblastoma multiforme and a good performance
`status were included in this phase II protocol, designed
`according to the Geehan two-step procedure. One patient
`was not eligible since he received only 4% of the prescribed
`BNCT dose. The median age of the 29 evaluable patients (16
`male and 13 female, was 53 years (range 26–69 years). Se-
`ven patients had grade 0 of the WHO Performance Status; 19
`were of grade 1 and 3 of grade 2. Twenty-seven patients
`were subjected to debulking surgery, while three patients
`not eligible for debulking surgery had only a minimally inva-
`sive diagnostic biopsy. No other therapies in the primary
`setting were allowed. The aim was to deliver BNCT no later
`than 6 weeks following the surgery/biopsy. The study was
`approved by ethics committees at the participating hospi-
`tals. One patient was excluded from the per protocol anal-
`ysis since she was only given 4% of the planned BNCT dose
`due to bad compliance. At relapses the patients were given
`treatment at the discretion of each responsible physician.
`Tumour volume (defined as the contrast-enhancing vol-
`ume) and target (defined as the volume corresponding to
`the pre-operative tumour volume, plus oedema, plus a 2-
`cm margin) volumes were in the ranges of 14–306 cm3
`
`(median 352 cm3),
`
`(median 45 cm3) and 154–885 cm3
`respectively.
`Immediately after BPA-infusion and after neutron treat-
`ment, all patients were given high-dose betamethasone.
`As corticosteroids are known to interfere with the amino
`acid transport through the blood–brain barrier [21], the
`goal was to limit their use to a minimum before BPA-infu-
`sion. The high-dose betamethasone treatment was contin-
`ued for a few more days and thereafter gradually
`decreased and adjusted to the clinical status of each pa-
`tient by the discretion of the responsible physicians.
`
`BNCT procedure
`The treatment procedure has been described in detail
`elsewhere [8]. The standard BNCT procedure used at Studs-
`vik included two days of preparation, the day of BNCT and
`an overnight observation at the affiliated hospital. During
`the first day a CT scan was carried out with and without con-
`trast. Fiducial triangulation points, marked on the patient’s
`scalp and identified by radiographic markers, were used for
`treatment-planning and patient positioning. The treatment
`position simulation to realize the optimal irradiation geom-
`etry resulting from the treatment-planning was carried out
`at a geometrical replica of the epithermal beam port usually
`on the day before the scheduled neutron irradiation.
`Two hours after completion of the BPA-F infusion, the pa-
`tient was transferred to the Studsvik BNCT facility for neu-
`tron irradiation and put in the irradiation position as was
`previously determined during the treatment position simu-
`lation. All patients were irradiated with two fields using a
`10 · 14 cm2 collimator. The duration of irradiation was ad-
`justed to deliver the prescribed peak brain dose, which in
`turn depended upon the reactor power and the average
`blood 10B concentration during irradiation. Intercom was
`used for communication with the patient during irradiation
`and patient’s status was monitored using a closed-circuit TV
`system and a pulse-oxymeter.
`Approximately 1 h after neutron irradiation the patient
`was transported back to the hospital for an overnight
`observation.
`
`Boronophenylalanine–fructose
`Solutions of the BPA-F complex for infusion were pre-
`pared at a concentration of 30 mg BPA/ml (0.14 M) using a
`modification of previously published procedures
`(see
`[8,9]). Briefly, BPA (95% atom 10B-enriched, L-isomer, ob-
`tained from Glyconic, NY, USA) was combined in water with
`a 10% molar excess of fructose. The pH was adjusted to 9.5–
`10.0 with NaOH, the mixture was stirred until all solids dis-
`solved, and the pH was then readjusted to 7.4 with HCl. The
`concentration was attuned with water to 0.14 M. The solu-
`tion was passed through a 0.22 lm-pore sterilization filter
`(Nalge Company, Rochester, NY) and was transferred to
`sterile infusion bags. A fresh solution of BPA-F was prepared
`for each patient and was used within 48 h of preparation.
`Pyrogenicity and sterility tests were done for each batch.
`On the day of irradiation, BPA-F was infused intrave-
`nously (i.v.) over 6 h to deliver 900 mg BPA/kg body weight.
`To assess the boron concentration in the blood, blood sam-
`ples were intravenously taken just before the start of irradi-
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1074 - Page 2 of 9
`
`

`
`R. Henriksson et al. / Radiotherapy and Oncology 88 (2008) 183–191
`
`185
`
`ation, during the break between irradiations at different pa-
`tient positions, and immediately following the irradiation
`were used for calculating the average boron concentration
`during the irradiation.
`Additional samples were collected at 12, 15 and 24 h
`after the start of the infusion. Boron concentration in all
`samples was measured using induced current plasma atomic
`emission spectroscopy, for details see [5].
`
`brain cancer module (BCM 20) combined with the EORTC
`core quality of life questionnaire (QLQ-C30).
`Treatment following recurrences was at the discretion of
`each responsible physician. Thirteen of the 29 evaluable pa-
`tients were treated with temozolomide at varying time
`intervals after a progressive disease was evident. Two pa-
`tients underwent additional surgical resection at relapse,
`and two patients received additional radiotherapy.
`
`Treatment-planning
`The tumour was delineated as the contrast-enhanced
`zone based on the pre- and post-operative brain scans,
`and the clinical target volume was defined as a pre-opera-
`tive tumour volume plus oedema plus a two-centimetre
`margin. The BNCT treatment-planning software SERA [20]
`uses CT or MRI images to render the three-dimensional pa-
`tient geometry, and generate both isodose contours for
`each beam component and the total BNCT dose. The soft-
`ware requires the input of the boron concentration in the
`blood and tumour as well as the biological effectiveness fac-
`tors (RBE and C-RBE) [12], to be used. To be emphasized, in
`order to improve our evaluation of the tumour border we
`also performed a MRT/CT within 72 h after surgery to be
`able to exclude post-surgical reactions which in most stud-
`ies can hamper a correct later evaluation.
`The measured average blood–boron concentration of 10B
`during BNCT was used to calculate the dose to the normal
`brain structures. Although the results obtained from stereol-
`ogical morphometry [11] suggests that the average ratio of
`tumour to blood 10B concentrations is 3.8:1.0, it is prudent
`to assume that some parts of the tumour that were not per-
`fused as well as the average will contain somewhat less 10B.
`The 3.5:1 tumour: blood 10B ratio was used to estimate the
`radiation doses delivered to tumour cells. The RBE-weighted
`doses were calculated using the following weighting factors:
`1.3 and 3.8 for the dose from the boron neutron capture reac-
`tion in brain and tumour tissue, respectively; 3.2 for the dose
`from secondary protons and 1.0 for the gamma dose compo-
`nent (see [9]). For treatment-planning, the duration of irradi-
`ation was adjusted to limit the peak and average brain dose to
`15.0 Gy(W) and 6.0 Gy(W), respectively.
`
`Post-BNCT management and follow-up
`Following the completion of the BNCT procedure, the pa-
`tients were transferred back to the collaborating hospital
`where they remained, initially, for 48 h and then overnight
`for observation. At the completion of the in-hospital obser-
`vation the patients were discharged to the care of the des-
`ignated follow-up physicians. The post-BNCT information
`according to the study protocol was collected and con-
`trolled by an independent CRO (Clinical Research Organisa-
`tion) according to GCP (Good Clinical Practice). Clinical
`evaluation, including CT/MRT, was performed 6 weeks after
`BNCT and thereafter every third month.
`Time to progression and survival were calculated from
`the date of BNCT treatment. Survival from radiological diag-
`nosis was also calculated. All adverse effects were graded
`according to the WHO grading system of toxicity and re-
`ported in the patient’s medical records and on the case re-
`cord forms. Quality of life was evaluated using the EORTC
`
`Statistical analysis
`The Kaplan–Meier method was used for the survival anal-
`ysis and the differences between different groups were
`tested with the log rank test.
`
`Results
`Boron pharmacokinetics
`The 6-h infusion of 900 mg BPA per kg body weight re-
`sulted in an average blood–boron concentration of
`24.7 lg/g (range 15.2–33.7 lg/g) at the time of irradiation
`(approximately 2–3 h post-infusion). The detailed results
`regarding the pharmacokinetics of boron concentration in
`the blood and tissues have been discussed previously
`[6,8]. The maximum boron concentration in the blood was
`observed at the end of infusion and ranged from 23 to
`53lg/g. Mean blood–boron concentration at 6 and 18 h
`post-infusion was 18lg/g (range 12–25lg/g) and 7lg/g
`(range 3–11lg/g), respectively.
`
`Radiation dosimetry
`The radiation doses delivered to the tumour and normal
`tissue are summarized in Table 1. Radiation doses delivered
`to the tumour cells within the contrast-enhancing tumour
`volume and to the tumour cells infiltrating normal brain be-
`yond the contrast-enhancing volume were calculated
`assuming blood-to-tumour 10B concentration ratio of 3.5
`and a uniform distribution of boron within the residual tu-
`mour. Peak and average weighted absorbed doses to the
`brain were in the ranges of 7.0–15.5 Gy (W) and 3.3–
`6.1 Gy(W), respectively. The minimal weighted absorbed
`dose delivered to the tumour and target volumes ranged
`from 15.5 to 54.3 Gy and from 8.8 to 30.5 Gy, respectively.
`
`Adverse effects
`An overview of the reported adverse events from all 29
`patients is presented in Table 2. Mild and transient side ef-
`fects such as tiredness and diarrhoea occurred during the
`infusion in some of the patients. Skin reaction (erythema)
`was evident in 8 patients. In one patient itchy erythema
`of grade 3 was seen, most likely related to the anti-epileptic
`medication with carbamezapin. The patient also developed
`similar symptoms later on when he was on treatment with
`phenytoin. The symptoms disappeared completely in a cou-
`ple of weeks after the treatment was changed. One patient
`suffered from severe macroscopic haematuria during and
`following the BNCT procedure, which was completely re-
`solved within 2 weeks. This was assumed to be related to
`the infusion of a high concentration of BPA-F.
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1074 - Page 3 of 9
`
`

`
`186
`
`BNCT and glioblastoma
`
`Table 1
`Nominal radiation doses delivered to the tumour and normal tissues
`
`Peak
`Average
`Minimum
`
`Absorbed doses to the brain
`
`Absorbed doses to the tumour
`
`Weighted [Gy(W)]
`
`11.9 (8.1–15.4)
`4.6 (3.2–6.1)
`<1
`
`Physical [Gy]
`
`9.1 (6.2–12)
`3.7 (2.5–4.9)
`<1
`
`Weighted [Gy(W)]
`
`69.0 (49–99)
`58.0 (34–81)
`35.0 (15.4–54.3)
`
`Physical [Gy]
`
`21.0 (15–29)
`17.9 (11–25)
`11.1 (3.7–13.4)
`
`Mean and range are given.
`
`Table 2
`Adverse events obtained during the study period in 19 of 29
`patients
`
`Type of AE
`
`No. of events
`
`No. of patients
`
`Skin/mucosa
`Seizures (epilepsies)
`Thrombosis
`Abdominal
`Depression
`Aphasia
`CVS
`
`13
`12
`8
`5
`3
`2
`2
`
`8
`7
`6
`4
`3
`2
`2
`
`Other WHO grades 1–2 toxicities reported include transient
`lymphocytopenia, granulocytosis, skin pigmentation, patchy
`atrophy, altered taste, alteration of saliva, otitis media, exter-
`nal otitis,
`tiredness, psychosis, haematuria, alopecia, dry
`mouth, conjunctivitis, thyroid dysfunction. CVS, cerebrovascular
`insult.
`
`In seven patients, 12 epileptic seizures (grades 1–3) in
`close relationship to the delivered BNCT were reported.
`Three of the patients have had seizures also at the time
`of diagnosis, but were free from seizures after the start of
`anti-epileptic treatment.
`Abdominal disturbances, constipation and flatulence
`with pain (grades 1–2), which could not be excluded to be
`associated with BNCT, were reported by 4 patients. Venous
`thromboembolism although most likely not directly related
`to the BNCT was seen in 6 patients at different time inter-
`vals after the BNCT. In addition, a general impression ex-
`pressed by most investigators was the need for higher and
`extended treatment with corticosteroids compared to what
`is required following conventional radiotherapy.
`To summarize, the total number of adverse events re-
`ported were 61 in 19 out of the 29 evaluated patients. Nine
`grades 3–4 adverse events were reported in 4 of the 29 pa-
`tients (epileptic seizures, haematuria, thrombosis, ery-
`thema) in close proximity to the BNCT procedure.
`
`Survival
`All patients had a follow-up of at least ten months with
`regard to survival. The survival and progression-free survival
`are seen in Figs. 1 and 2. The median progression-free sur-
`vival (PFS) was 5.8 months defined as the time from the
`start of BNCT treatment and until the detection of recur-
`rence. The diagnosis of recurrences was based on CT/MRI
`and/or clinical deterioration. In 4 patients a PET-evaluation
`verified the occurrence of relapses with metabolic active
`tumours. In one patient a re-operation verified the presence
`
`T 1
`T 2
`T 3
`4
`T 5
`6
`T 7
`T 8
`T 9
`10
`11
`12
`T 13
`T 14
`15
`16
`T 17
`T 18
`19
`20
`T 21
`T 22
`23
`24
`25
`26
`27
`28
`29
`
`Timetoprogress
`Timeafter progress
`Death
`
`0
`
`6
`
`24
`18
`12
`Months after BNCT
`
`30
`
`36
`
`Fig. 1. Survival outlined for each individual patient, and status at
`the time for analysis. T, patients treated with temozolomide at
`relapse.
`
`of viable tumour cells. Neurological deterioration was pro-
`gressively seen in almost all patients during the study period
`(Fig. 2).
`Twenty-three of the 29 evaluable patients were dead at
`the time for follow-up, performed at least 10 months after
`treatment. Twenty of them died from progressive GBM,
`while one died due to pulmonary embolism 3 months after
`BNCT. In this patient an autopsy could not detect any sign
`of viable tumour. One patient died from haemorrhage in
`the tumour and pulmonary embolism, however, with pro-
`gressive brain tumour disease. One patient died after 8
`months without any verified progressive disease, but no
`other clear explanation could be detected. The overall med-
`ian survival was 14.2 months calculated from BNCT treat-
`ment and the median time from histological diagnosis to
`BNCT treatment was 40 (1–75) days. The median survival
`time from radiological diagnosis was 16 months. The patient
`excluded from the analysis, given only 4% of the prescribed
`neutron dose, died after 4 months and was not included in
`this survival analysis. At recurrence 13 of the evaluable pa-
`tients received treatment with temozolomide, at the dis-
`cretion of each responsible physician. The median survival
`of the patients that received temozolomide after tumour
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1074 - Page 4 of 9
`
`

`
`R. Henriksson et al. / Radiotherapy and Oncology 88 (2008) 183–191
`
`187
`
`progression was 17.7 months as compared to 11.5 months
`(p < 0.05) in those that had not received (Fig. 1).
`As shown in Fig. 3, no correlation between the delivered
`radiation doses and survival times, adverse effects, or qual-
`ity of life were detected.
`
`Quality of life
`EORTC QLQ 30 evaluation forms used in order to get
`information of the patients self estimated quality of life be-
`fore and after BNCT. Fig. 4A–C illustrates some of the
`parameters and show that the patients well-being progres-
`sively declined.
`
`Discussion
`The present phase II study in patients with glioblastoma
`multiforme and a good performance status suggest that a sin-
`gle session of BNCT is feasible and that the survival time
`seems to be comparable to that following multifractionated
`conventional photon irradiation. However, the effects on sur-
`vival using BNCT alone is still limited and most of the patients
`had progress of their disease already within 6 months. BNCT,
`thus, still must be considered as an experimental therapy and
`needs to be further improved before it can be adopted in the
`routine management. It is also of importance to emphasize
`recently obtained data from controlled clinical studies dem-
`onstrating that chemotherapy in addition to conventional
`irradiation displays beneficial effects [23]. Although not part
`of the present protocol, it was interesting to find that this
`observation was also evident in the present evaluation show-
`ing a clear survival advantage for those patients who were
`treated with temozolomide after recurrence.
`
`Fig. 2. Patients free from radiological and neurological progress
`after BNCT treatment
`(median progression-free survival 5.8
`months). No, number of patients. Inset: Kaplan–Meier overall
`survival curve after BNCT treatment (median survival 14.2 months).
`
`Fig. 3. Survival time versus nominal average and minimum doses delivered to tumour and target volumes.
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1074 - Page 5 of 9
`
`

`
`188
`
`BNCT and glioblastoma
`
`Fig. 4. Quality of life analysis according to the EORTC brain cancer module (BCM 20) combined with the EORTC core quality of life
`questionnaire (QLQ-C30). (A) Physical functioning. (B) Social functioning. (C) Global health status. N, the number of patients reported at each
`time point. The decreased number of patient reported are mainly due to death and a decrease in performance status.
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1074 - Page 6 of 9
`
`

`
`R. Henriksson et al. / Radiotherapy and Oncology 88 (2008) 183–191
`
`189
`
`Results of pre-clinical in vitro and in vivo studies earlier
`reported by several groups [10,16–18,24] led us to investi-
`gate a protocol with a 3-fold increase of both the time of
`infusion and the total dose of BPA-F.This was the major dif-
`ference between our studies and other clinical trials of
`BNCT using BPA-F. Although one patient displayed haemat-
`uria and the occurrence of crystalloid structures in the ur-
`ine, the applied infusion procedure seemed to be safe and
`resulted in a higher boron concentration in the blood at
`the time of neutron irradiation compared to what have been
`reported previously.
`The rationale for using a longer infusion time and a de-
`layed initiation of neutron irradiation is at least in part sup-
`ported by a recent study using microdialysis for measuring
`BPA uptake in the extracellular fluid of the glioblastoma
`and normal tissues [6]. However, it cannot be excluded that
`some glioblastoma cells might be able to pump out BPA dur-
`ing the 2- to 3-h break and thus becoming underdosed,
`explaining the disappointing clinical outcome. Neverthe-
`less, comparison presented in Table 3 suggests a possible
`benefit of the longer infusion time and/or the higher dose
`of boron on the patients’ survival. Unfortunately, the ap-
`plied procedure did not result in any significant tumour re-
`sponse as defined by the decrease in the contrast-
`enhancing volume on CT scans, and in the vast majority of
`the patients a radiological progression was seen before 6
`months post-treatment. We were aware of the limitations
`of using CT in evaluating responses and therefore the pro-
`gression was confirmed with PET scan in four patients and
`in one case with biopsy. Moreover, even if the treatment
`was generally well tolerated, the occurrence of adverse ef-
`fects and the obvious early and continuous decline in the
`quality of life must also be considered. Of course a strict
`comparison is afflicted with difficulties since the patients
`included in the various studies most likely belong to differ-
`ent prognostic groups. The patients in our study were rela-
`tively young and had a good performance status at the
`inclusion.
`It is noteworthy that most of the patients treated with
`BNCT at other centers received an aggressive treatment
`including surgery, photon irradiation, and chemotherapy
`at post-BNCT tumour recurrence, while, except in two pa-
`tients that were treated with re-irradiation and in two pa-
`tients who underwent a second surgical resection, the
`
`Table 3
`Comparison of BPA-F infusion protocols and median survival
`times (MST) from diagnosis obtained in different BNCT trials
`
`Center
`
`BPA-infusion
`
`MST (months)
`
`Time (h)
`
`Dose (mg/kg)
`
`MIT/Harvarda
`1–1.5
`250–350
`13
`BNLb
`2
`250–330
`11.9–14.8
`VTTc
`2
`290–400
`13
`Studsvikd
`6
`900
`16
`a Busse et al. 2003 [7], the occasion from which MST was cal-
`culated was not reported.
`b Diaz et al. 2003 [14].
`c Barth and Joensuu, 2007 [4].
`d The present study.
`
`only treatment offered to some of our patients (n = 13)
`was chemotherapy. Interestingly, the patients treated with
`temozolomide at recurrence in our study lived longer than
`those treated with BNCT only, 17.7 and 11.5 months,
`respectively. However, one has to be aware of that the
`treatment following recurrence was at the discretion of
`each responsible physician and did not follow a shared pro-
`tocol. Similar beneficial results using temozolomide con-
`comitantly with conventional
`radiotherapy have been
`reported in a well-conducted phase III study [23]. Although
`our results must be evaluated cautiously due to the risk of
`selections bias, it suggests that the overall survival might
`be improved if BNCT were combined with concomitant che-
`motherapy. Another interesting option could also be to find
`an optimal approach for combining conventional radiother-
`apy and BNCT [3].
`Similar to the results of the phase II part of the Brookha-
`ven study [14] no correlation between the nominal radiation
`doses delivered to the target volume and time to progres-
`sion or the total survival time was observed. One possible
`explanation could be that individual characteristics of each
`patient’s tumour, such as its aggressive behavior (number of
`clonogens), location, make it impossible to detect such cor-
`relation and the relatively small number prevents any mean-
`ingful multivariable analysis. Another could be the well-
`known heterogeneity of glioblastoma and a possible sto-
`chastic distribution of boron concentration in tumour cells.
`Thus, making our estimation of radiation doses delivered to
`the target volume of each patient by the boron neutron cap-
`ture reaction inaccurately reflects the thermal neutron flu-
`ence at, e.g. the deepest part of target volume multiplied
`by an assumed boron concentration rather than a real
`deposited high-LET radiation dose. Whichever the explana-
`tion might be, the lack of tumour control suggests that
`the doses delivered to the tumour cells were still too low.
`It is consistent with the observation from the BNCT trials
`in Japan [19], where the minimum physical (not multiplied
`by any CBE factor) radiation dose deposited by the boron
`neutron capture in the tumour volume of 15 Gy was neces-
`sary to obtain a significant improvement in survival time.
`The corresponding doses in this study were in the range
`3.7–13.4 Gy (mean 8.2 Gy). Thus, it is most plausible that
`an unaffected subpopulation of the tumour cells governs
`the patients survival.
`Although the used BNCT protocol seems to be rather well
`tolerated, it seems that neurological symptoms including
`seizures are somewhat more frequent in the first days after
`BNCT delivery than that observed in the Brookhaven trials at
`corresponding radiation doses [8], as well as following con-
`ventional photon irradiation. One reason might be that, un-
`like in the Brookhaven protocols, in the Swedish trial there
`was no limits of tumour size. Also patients with multi-focal
`tumours or tumours invading corpus callosum were in-
`cluded, which results in larger target volumes. Another rea-
`son to explain the suspected toxicity might be the longer
`BPA-infusion time and higher concentrations of boron in
`the blood and/or normal brain cells, which results in larger
`contributions of high-LET radiation to the total dose. Morris
`et al. reported [18] the CBE factor for BPA obtained from rat
`spinal cord irradiations following a 6-h infusion of BPA was
`1.5, which is 13% higher than that obtained after 2-h infu-
`
`CFAD v. Anacor, IPR2015-01776, CFAD EXHIBIT 1074 - Page 7 of 9
`
`

`
`190
`
`BNCT and glioblastoma
`
`sions [1,2] and used in the Brookhaven study. As, on aver-
`age, the physical dose resulting from neutron capture in
`boron amounts to 52% and 42% of the peak and average nor-
`mal brain dose, respectively, using the CBE factor of 1.5,
`the weighted dose limits in the Swedish protocol would cor-
`respond to 16.0 Gy(W) and 6.3 Gy(W) of the peak and aver-
`age brain dose from the Brookhaven protocols. Although
`there was no obvious relationship between side effects seen
`and estimated radiation dose, it is of importance to further
`emphasize these events and adjust the CBE factors accord-
`ingly in forthcoming trials. As reported by the patients
`themselves, there was an obvious and a continuous de-
`creased quality of life after BNCT treatment, which seems
`to be somewhat more pronounced than was seen in recent
`publications when treating patients with only photon irradi-
`ation [15,23].
`In summary, the present study investigated the possibil-
`ity to improve the outcome of BNCT for glioblastoma by
`increasing the BPA-F infusion time and dosage demonstrated
`clinical efficacy comparable to conventional photon irradia-
`tion [23]. The advantage of BNCT is that treatment takes
`only about totally 4–5 days including time for preparations,
`planning and post-treatment observation as compared to
`10–50 days required by conventional radiotherapy. On the
`other hand, the present situation with only one BNCT facil-
`ity in Sweden and few BNCT facilities worldwide requires re-
`source-demanding transportation, which together with the
`estimated higher treatment cost of BNCT as compared to
`conventional radiotherapy must also be considered when
`evaluating its applicability. Thus, it is clear that the poten-
`tial advantages of BNCT must be assessed in well-controlled
`trials considering the tumour control, survival time as well
`as the quality of life, toxicity and utilisation of health re-
`sources. Toxicity observed in this study will probably not al-
`low any further escalation of the radiation doses given to
`the normal tissues, and the presumed lack of uniformly high
`boron concentration in each of viable tumour cells could be
`a major obstacle for improvement of BNCT. Therefore, the
`search for an improved boron delivery agent and/or combi-
`nations of boron carriers must continue [2,3] Finally, since
`glioblastoma is a systemic disease within the brain, a mul-
`timodality approach using combination of various treatment
`approaches must probably be tested in order to strive for a
`significant clinical improvement.
`
`Conflict of interest
`During the period of the study several of the authors were
`supported by Studsvik