Uptake of two 10B-compounds in liver metastases of colorectal adenocarcinoma for extracorporeal irradiation with boron neutron capture therapy (EORTC Trial 11001)
Article first published online: 5 NOV 2007
Copyright © 2007 Wiley-Liss, Inc.
International Journal of Cancer
Volume 122, Issue 5, pages 1164–1171, 1 March 2008
How to Cite
Wittig, A., Malago, M., Collette, L., Huiskamp, R., Bührmann, S., Nievaart, V., Kaiser, G. M., Jöckel, K.-H., Schmid, K. W., Ortmann, U. and Sauerwein, W. A. (2008), Uptake of two 10B-compounds in liver metastases of colorectal adenocarcinoma for extracorporeal irradiation with boron neutron capture therapy (EORTC Trial 11001). Int. J. Cancer, 122: 1164–1171. doi: 10.1002/ijc.23224
- Issue published online: 24 DEC 2007
- Article first published online: 5 NOV 2007
- Manuscript Accepted: 8 AUG 2007
- Manuscript Received: 14 MAY 2007
- European Commission. Grant Number: QLK3-CT-1999-01067
- boron neutron capture therapy;
- sodium mercaptoundecahydro-closo-dodecaborate;
Disseminated metastases of colorectal cancer in liver are incurable. The trial EORTC 11001 investigates whether autotransplantation after extracorporeal irradiation of the liver by boron neutron capture therapy (BNCT) might become a curative treatment option because of selective uptake of the compounds sodium mercaptoundecahydro-closo-dodecaborate (BSH) or L-para-boronophenylalanine (BPA). BSH (50 mg/kg bw) or BPA (100 mg/kg bw) were infused into patients who subsequently underwent resection of hepatic metastases. Blood and tissue samples were analyzed forthe 10B-concentration with prompt gamma ray spectroscopy (PGRS). Three patients received BSH and 3 received BPA. Adverse effects from the boron carriers did not occur. For BSH, the highest 10B-concentration was observed in liver (31.5 ± 2.7 μg/g) followed by blood (24.8 ± 4.7 μg/g) and tumor (23.2 ± 2.1 μg/g) with a mean 10B-concentration ratio metastasis/liver of 0.72 ± 0.07. For BPA, the highest 10B-concentration was measured in metastases (12.1 ± 2.2 μg/g) followed by liver (8.5 ± 0.5 μg/g) and blood (5.8 ± 0.8 μg/g). As BPA is transported actively into cells, viable, metabolically active cells accumulate exclusively this compound. Consequently, a model is proposed to adjust the values measured by PGRS for the proportion of viable cells to express the relevant 10B-concentration in the tumor cells, revealing a 10B-concentration ratio metastasis/liver of 6.8 ± 1.7. In conclusion, BSH is not suitable as 10B-carrier in liver metastases as the 10B-concentration in liver was higher compared to metastasis. BPA accumulates in hepatic metastases to an extent that allows for extracorporeal irradiation of the liver with BNCT. © 2007 Wiley-Liss, Inc.
The management of hepatic metastases from colorectal carcinoma represents a significant clinical problem. The best chance of cure is achieved by complete resection of the metastases, resulting in a 5-year survival rate of 25–45%.1, 2 However, only 10–15% of all patients are eligible for surgery due to the size or number of metastases.3 Moreover, local recurrence after surgery due to residual microscopic disease occurs in the majority of patients.4 Although new cytostatic drugs have a response rate of around 40%, the overall survival benefit is marginal. Up to 90% of patients with liver metastases die from the disease. Thus, attention has turned to loco-regional techniques that together with surgery may be potentially curative. Procedures like cryotherapy or radiofrequency ablation can help to treat unresectable or nontotally resectable lesions but prolongation of survival is limited.1 The use of radioactive microspheres to treat disseminated liver metastases is under investigation as a palliative modality.5
The future in cancer treatment is with dedicated targeted therapies, to selectively kill tumor cells whilst sparing surrounding healthy tissue, increasing efficacy and decreasing toxicity. One such option could be boron neutron capture therapy (BNCT) that provides through the limited spatial distribution of its effects a highly selective delivery of radiation.
BNCT is a targeted form of radiotherapy, which uses the ability of the isotope 10B to capture thermal neutrons with high probability leading to the nuclear reaction 10B(n,α,γ)7Li. This reaction produces 478 keV photons, alpha-particles and 7Li-ions, the latter 2 having a high linear energy transfer (LET) and therefore a high biological effectiveness relative to conventional irradiation. The range of these particles in tissue is 10–14 μm limiting their effects to ∼1 cell diameter. This short range offers a targeted irradiation of tumor cells, if 10B can be selectively delivered. BNCT has the potential to treat macroscopic tumors with high efficacy but also to kill tumor cell clusters embedded within normal tissue, whilst sparing surrounding healthy cells.
However, the success of BNCT ultimately depends upon the selective delivery of 10B-atoms to tumor cells. Currently, 2 experimental drugs are available for clinical investigations:
Since 1986, a research program at the University of Pavia (Italy) has been underway to investigate the possibility to cure diffuse hepatic metastases by explanting the liver and irradiating the organ with BNCT followed by an autotransplantation.18, 19, 20, 21 Within this program, 2 patients have been treated achieving tumor remission.22, 23
In principle, the delivery of BNCT to an explanted liver is especially attractive for several (radio-)biological reasons:
By explanting the liver no other organs are irradiated.
In principle, macroscopic and microscopic metastases dissiminated within healthy liver can be selectively treated.
By perfusion of the organ with Wisconsin solution (for organ preservation), 10B-containing blood is washed out, reducing the radiation dose to the vessels and decreasing the risk of radiation-induced liver disease.24
The limitations of liver transplantation in oncological patients due to the necessity of immune-suppression are not applicable in allo-transplantations. The surgical technique is similar to the technique used in living donors.
However, such an invasive procedure is only justified, if it has the potential of cure. Accepting the hypothesis that BNCT offers a curative chance for actual incurable cases, a prerequisite to any future research is the proof of a preferential delivery of a 10B-containing compound to the liver metastases. The aim of the EORTC trial 11001 was to investigate, whether specific solid tumor entities can be treated with BNCT due to a preferential uptake of BSH or BPA. This article reports the results of this biodistribution study in patients suffering from hepatic metastases of colorectal cancer.
Material and methods
The EORTC trial 11001 is a translational research/phase I trial with the goal to measure the uptake of 2 10B-compounds in tissues and blood. Prior to the planned removal of hepatic metastases, either BSH or BPA was infused intravenously. During metastasectomy, tissue and blood samples were collected. No extension of the planned surgery was allowed.
The study design was as follows: First, 3 patients were infused with BSH, followed by 3 patients infused with BPA. The samples of these 6 patients were analyzed. Furthermore, it was presumed that, because of different mechanisms of targeting a coadministration of both compounds might lead to a higher amount of 10B-atoms in single tumor cells or to a more homogeneous 10B-distribution in tumors.25, 26, 27 Therefore, an infusion of both drugs together was foreseen in another 3 patients, if BSH and BPA had both a “favorable” uptake in the metastases. “Favorable” 10B-uptake was defined as follows:
BSH: tumor-normal surrounding tissue ratio >2, tumor to blood ratio >0.6.
BPA: tumor-normal surrounding tissue ratio >2, tumor to blood ratio >1.5.
BPA and BSH: tumor-surrounding normal surrounding tissue ratio >2.
The primary endpoint was the 10B-concentration measured by prompt gamma ray spectroscopy (PGRS). The secondary endpoint was the qualitative and quantitative toxicity of the 10B-compounds as assessed according to NCI-CTC-criteria (version 2.1). As a direct benefit from participation in the trial for the individual patient was not expected, the number of included patients was limited. Descriptive statistics were applied. The Protocol Review Committee of the EORTC, the Ethics Committee of the University Duisburg-Essen and the relevant national authorities approved the trial. All patients gave written informed consent prior to their inclusion in the study.
One of the following applications of the study medication was foreseen:
BSH (50 mg/kg) infused within 1 hr. The infusion starts 12 hr prior tissue sampling.
100 mg/kg BPA infused within 1 hr. The infusion starts 2 hr prior tissue sampling.
Combination of BSH and BPA.
As each 10B-compound has a different half lives and pharmacokinetic, the schedules for blood sampling differed. In case of BSH-infusion, samples were taken prior to infusion, at the end of infusion and 3, 9, 12, 15, 24 and 48 hr after beginning of infusion. For the case of BPA, samples were taken prior to infusion, at the end of infusion and 1.5, 2, 2.5, 3, 3.5, 4, and 6 hr after start of infusion. During surgery, 1 blood sample was taken at each time point when a tissue sample was taken. On Days 1, 5, and 28 after surgery, a prospective assessment of toxicity was performed.
BSH and BPA were purchased from KATCHEM (Praha, Czech Republic). The quality of the study medication was controlled including examination of the identity of compounds by IR spectroscopy, monitoring of purity by high pressure liquid chromatography and test for pyrogenicity. The enrichment of 10B was 99% in both compounds and was tested with PGRS and inductively coupled plasma atomic emission spectroscopy.
Injection-solutions were prepared according to standard operating procedures established for the EORTC trials 11961, 11001, and 11011.28 50 mg/kg BSH were dissolved in 250 ml saline. The solution was filtered (0.22 μm) and transferred into an infusion bag using a vacuum filler assembly to prevent the oxidation of BSH. The solution can be stored for up to 6 hr. For reasons of solubility, BPA was complexed with fructose. BPA and water were stirred before 10 N NaOH was added to obtain a pH of 10.5. Fructose was added (1:1.1 molar ratio) and the pH readjusted to 7.4 with HCl. Water was added to reach a concentration of 30 mg BPA/ml. After 24 hr, the BPA-solution was filtered (0.22 μm) and transferred into an infusion bag, which could be stored for 12 days.28
Patients were eligible if a resection of MRI-diagnosed hepatic metastases of histologically proven colorectal cancer was planned. Other eligibility criteria were age ≥18 y, WHO performance status ≤2, adequate hematological values, no severe concomitant disease including phenylketonuria and absence of toxic effects of previous anticancer therapies, except alopecia. Patients having received radiation or chemotherapy within 3 months of the planned surgery were excluded.
Tissue sampling and analysis of 10B
Tissue samples were taken during surgery from tumor, liver and whenever possible from other tissues such as skin and fat. If possible, central and peripheral parts of the tumor were dissected and analyzed separately. The samples were weighed and kept in airtight containers at –20°C until analysis.
The 10B-concentration in tissue and blood was analyzed by PGRS at the High Flux Reactor Petten. PGRS quantifies the 478-keV photon emission during 10B(n,α,γ)7Li reactions. PGRS measures the integral 10B-concentration in the sample volume.29 The measured values reported here are given as mean values ± standard deviation (SD) of several samples from the same tissue. PGRS has the advantage of being an established tool for 10B-analysis in biological samples.29, 30 It is suitable for measuring a large number of samples due to the short acquisition time and the simple sample preparation. PGRS determines the 10B-concentration only, whereas chemical methods measure the sum of both naturally occurring isotopes (10B and 11B).
Pathological examination and semiquantitative analysis
All tissues removed during surgery were examined applying standard procedures. Special attention was paid to the regions where samples were taken for 10B-analysis, to histologically confirm the macroscopical appearance of samples. Additionally, a semiquantitative analysis was performed to determine the proportion of tumor cells, necrosis and stroma within the histologically proven metastases. A measuring field was superimposed on the histological sections (at a magnification of 100×; Nikon Microscope Eclipse 80i equipped with a digital camera Nikon DS-2v) shown on a computer screen. All areas of the measuring field (each area measuring 0.25 mm2, minimal count 800 areas) contained more than 50% of viable tumor cells, necrosis or stroma, respectively, were counted and subsequently the respective proportions were calculated and given as percentage of the whole area investigated.
Within 5 months, 6 patients were included in the trial, who had all completed protocol treatment and follow up investigations as foreseen. Three patients received BSH (2 female, 1 male); 3 were infused with BPA (2 female, 1 male). One patient in the BSH group suffered from the hereditary polyposis coli syndrome. This patient was young (32 y), as compared to the other patients (59–67 y).
Analysis of toxicity
Serious adverse effects did not occur. Reported adverse effects were clearly attributable to surgery, such as pain, nausea, weakness, 1 case of coagulopathy and 1 infection at the site of the central venous catheter and inflammation at the surgical site with fever. All events resolved within 4 weeks.
As surgery was given priority over the trial and as tissue sampling was influenced by the progress of the operation, samples were taken during a certain time period, which was between 10.5–11.9 hr for patients infused with BSH and 1.6–2.5 hr for patients infused with BPA. These time intervals were close enough to the planned time points of sampling to allow comparability of results. The types of tissue taken depended on surgery and differed for the individual patients. The histology of hepatic metastases of colorectal cancer is very heterogeneous, as evidenced by the semiquantitative analysis (Table I). The proportion of viable tumor cells was 1–30% of the tumor volume only (Figure 5).
|Pat ID||Histological finding||Semiquantitative analysis|
|Viable tumor cells (%)||Necrosis, mucus (%)||Stroma (%)|
|1||Partly necrotic metastasis of mucinous adenocarcinoma, G2||26||51||23|
|2||Necrotic metastasis of adenocarcinoma, G2||Metastasis 123||5||72|
|3||Metastasis of mucinous adenocarcinoma, G3||1||94||5|
|6||Metastasis of mucinous adenocarcinoma, G3||20||76||4|
|7||Partly necrotic metastasis of adenocarcinoma, G2||15||35||50|
|8||Partly necrotic metastasis of adenocarcinoma, G2||30||40||30|
10B-uptake after BSH-infusion
The 10B-concentration in the blood reached a maximum at the end of the BSH-infusion and dropped gradually thereafter. During the time period of tissue sampling, 10B-concentration in the blood was 24.8 ± 4.7 μg/g. The 10B-concentration in blood, as a function of time, is showing in Figure 1.
The mean 10B-concentration in metastases after BSH-infusion was 23.2 ± 2.1 μg/g; the metastasis/blood ratio was 1.1 ± 0.1. As opposed to Patients 1 and 2, in Patient 3 the 10B-concentration in the metastasis was slightly higher as compared to blood (10B-concentration ratio metastasis/blood 1.3 ± 0.1). In 2 patients it was possible to examine central and peripheral tumor areas. The 10B-concentration did not differ substantially (Table II). In all patients, the 10B-concentration in the liver was slightly higher as compared to blood (10B-concentration ratio liver/blood: 1.4 ± 0.1). In most samples, the 10B-concentration was higher in the normal liver as compared to the metastases (10B-concentration ratio metastasis/liver: 0.7 ± 0.1). These ratios do not fulfill the definition of the study protocol for “favorable” 10B-concentration ratios. An evaluation of the combination of both boron carriers was therefore excluded.
|Patient registration no.||1||2||3|
|Sex, age||F, 32 y||M, 59 y||F, 63 y|
|Weight, height||69 kg, 172 cm||72 kg, 176 cm||68 kg, 155 cm|
|Body surface||1.82 m2||1.85 m2||1.65 m2|
|10B-conc. in liver||28.49 ± 3.34 ppm (n = 5)||38.72 ± 3.31 ppm (n = 19)||27.35 ± 1.13 ppm (n = 15)|
|10B conc. ratio liver/blood||1.42 ± 0.09||1.37 ± 0.17||1.34 ± 0.06|
|10B-conc. in central area of metastasis||nd||1Ta: 20.34 ± 0.76 ppm (n = 3)||27.47 ± 1.14 ppm (n = 3)|
|Tb: 22.14 ± 1.95 ppm (n = 9)|
|10B conc. ratio in central metastasis/blood||nd||1Ta: 0.68 ± 0.03||1.35 ± 0.06 (n = 3)|
|Tb: 0.91 ± 0.07|
|10B-conc. in peripheral part of metastasis||2T1: 33.09 (n = 1)||1Ta: 25.93 ± 4.75 ppm (n = 7)||25.84 ± 1.85 ppm|
|T2: 20.29 ± 0.16 m (n = 2)||Tb: 21.61 ± 1.76 ppm (n = 6)|
|10B conc. ratio in peripheral metastasis/blood||2T1:1.37||Ta: 0.91 ± 0.22||1.26 ± 0.08 (n = 5)|
|T2: 1.01 ± 0.01||Tb: 0.89 ± 0.07|
|10B-conc. in skin||26.85 ± 2.55 ppm (n = 8)||nd||nd|
|10B conc ratio skin/blood||1.34 ± 0.12||nd||nd|
|10B-conc. in fat||5.67 ± 0.61 ppm (n = 3)||10.65 ± 1.25 ppm (n = 8)||4.37 ± 1.99 ppm (n = 2)|
|10B conc ratio in fat/blood||0.23 ±0.03||0.36 ± 0.04||0.17 ± 0.08|
|10B-conc. in polyp||6.69 ± 0.76 ppm (n = 5)||nd||nd|
|10B conc ratio in polyp/blood||0.50 ± 0.06||nd||nd|
|10B-conc. in muscle of intestinal wall||8.39 ± 1.61 ppm (n = 7)||nd||nd|
|10B conc muscle of intestinal wall/blood||0.63 ± 0.12||nd||nd|
In Patient 1, a metastasis of the abdominal wall was also dissected. The 10B-concentration in this metastasis was as high as 33.1 μg/g with a ratio of metastasis/blood of 1.4. The absolute 10B-concentrations in metastasis, liver and blood of the individual patients are depicted in Figure 2.
The highest 10B-concentration in other tissues was detected in skin (26.9 ± 2.6 μg/g). However, skin could be analyzed in 1 patient only. Low uptake was found in fat (6.9 ± 3.3 μg/g, 3 patients), muscle of the intestinal wall (8.4 ± 1.6 μg/g, 1 patient) and in several intestinal polyps (6.7 ± 0.8 μg/g, 1 patient), in the patient suffering from familial polyposis syndrome (Table III).
|Patient registration No.||6||7||8|
|Sex, age||F, 56 y||F, 67 y||M, 67 y|
|Weight, height||72 kg, 165 cm||65 kg, 158 cm||82 kg, 164 cm|
|Body surface||1.80 m2||1.68 m2||1.88 m2|
|10B-conc. in liver||7.31 ± 0.62 ppm (n = 21)||8.58 ± 0.32 ppm (n = 15)||9.57 ± 0.68 ppm (n = 26)|
|10B conc. ratio in liver/blood||1.42 ± 0.19||1.47 ± 0.06||1.44 ± 0.10|
|10B-conc. in central area of metastasis||nd||6.45 ± 2.78 ppm (n = 8)||14.86 ± 0.84 ppm (n = 5)|
|10B conc. ratio in central metastasis/blood||nd||1.16 ± 0.50||2.24 ± 0.13|
|10B-conc. in peripheral part of metastasis||1Ta: 11.88 ± 1.49 ppm (n = 7)||11.67 ± 1.81 ppm (n = 15)||15.76 ± 1.29 ppm (n = 16)|
|Tb: 10.74 ± 2.74 ppm (n = 10)|
|10B conc. ratio in peripheral metastasis/blood||Ta: 2.36 ± 0.10||2.29 ± 0.32||2.41 ± 0.19|
|Tb: 2.13 ± 0.54|
|10B-conc. in skin||8.50 ppm (n = 1)||nd||8.12 ± 0.52 ppm (n = 6)|
|10B conc. ratio skin/blood||1.73||nd||1.55 ± 0.10|
|10B-conc. in fat||1.87 ± 0.53 ppm (n = 5)||3.14 ± 0.08 ppm (n = 2)||1.40 ± 0.57 ppm (n = 23)|
|10B conc. ratio in fat/blood||0.38 ± 0.11||0.54 ± 0.01||0.27 ± 0.11|
10B-uptake after BPA-infusion
The pharmacokinetic data gained by analysis of the 10B-concentration in blood after infusion of BPA was modeled according to the 2-compartment model published by Kiger et al.31, which fits the data well. The 10B-concentration in blood reached a maximum at the end of the BPA-infusion and dropped gradually, but considerably faster as compared to BSH. The 10B-concentration in blood, as a function of time, is showing in Figure 3.
In contrast to BSH, in all patients who received BPA, the highest 10B-concentrations were measured in the metastases (Fig. 4). In 2 patients central and peripheral tumor parts could be distinguished. In Patient 7, the 10B-concentration was remarkably higher in peripheral areas of the metastasis as compared to central parts. In Patient 8, this difference did also exist, but to a much lower extent. The mean 10B-concentration ratio between metastasis and blood was 2.4 ± 0.3 in peripheral tumor areas and in the central tumor areas, the ratio was 1.7 ± 0.3. The mean ratios between central and peripheral metastasis and liver were 1.5 ± 0.2 and 1.2 ± 0.2 respectively. In all patients, the 10B-concentration in liver was higher as compared to blood (10B-concentration ratio liver/blood: 1.4 ± 0.1). The absolute 10B-concentrations and 10B-concentration ratios between tissue and blood of individual patients are summarized in Table III.
The 10B-concentration in skin was comparable with the 10B-concentration in liver (10B-concentration ratio skin/blood 1.6 ± 0.1, 2 patients), whereas the 10B-concentration in fat was substantially lower (10B-concentration ratio fat/blood 0.4 ± 0.1, 3 patients).
The concept of delivering dose in BNCT differs from current radiooncological techniques, which optimize the dose distribution by applying the radiation as conformal as possible to the tumor. On the contrary, BNCT irradiates an extended area where microscopic disease is expected – in the case of liver metastases even the whole organ. The selective damage to the tumor cells is not achieved by the direct action of the beam but reached by the neutron capture reactions releasing high LET-particles only where 10B-atoms are present. Also, in contrast to other anticancer drugs, a compound for BNCT does not have any therapeutic effect but is aimed exclusively to transport 10B-atoms to tumor cells. The therapeutic effect will occur only if the compound is combined with the neutron irradiation. Therefore, in the testing process of such a boron carrier, a surrogate endpoint is needed, namely the 10B-concentration and/or 10B distribution in tissues. The success of the therapy depends on favorable 10B-concentration ratios for tumor-to-blood and tumor-to-surrounding tissue. Consequently in this trial, ratios, which hypothetically lead to a therapeutic effect, have been defined as the primary endpoint.
BSH was designed for BNCT of tumors in the brain and is assumed to target brain tumors by crossing the pathologically permeable blood-brain-barrier in tumor but does not traverse the intact blood-brain-barrier protecting healthy brain. BSH is taken up in tumor lesions to a concentration which is about equal to the concentration in blood, but it is not deposited in healthy brain,32 resulting in favorable tumor/brain ratios. The low toxicity of BSH and the fact that 1 molecule transports as many as 12 10B-atoms, stimulated work to test the compound in peripheral tumors.16 In our study design, its suitability was evaluated based on the 10B-concentration ratios of metastasis/liver and metastasis/blood, with the result that BSH is not suitable to treat liver metastases from colorectal cancer. In most tumor lesions, BSH did not accumulate. Ratios above unity suggest that BSH penetrates by a mechanism other than just diffusion. The uptake mechanism of the compound is still unknown. Distribution by passive diffusion, interaction of the double negative charge of BSH with choline head groups of phosphatidylcholine34 and the incorporation into the cell by covalent mixed disulfide bonding with glutathione35 have been discussed. Trivillin et al. hypothized36 that the antitumor effect of BNCT by the chemically nonselective compound Na210B10H10, which is similar to BSH, is not caused by a direct irradiation of tumor cells but by irradiation of the endothelium of vessels. The authors suggest a differential effect between normal vasculature as opposed to tumor vessels. If such a mechanism of action would prove true, BSH might become an interesting boron carrier for peripheral tumors. However, such cannot be tested with our study design.
Toxicity was not detected after infusion of BPA as BPA-fructose complex at the dose level applied in this trial. Higher amounts suitable for treatment have also not shown significant toxicity.17, 37, 38, 39
In all 3 patients, the amount of 10B measured in the tumor by PGRS was lower than anticipated, not reaching the value defined as “favorable” in the protocol. This is in clear contradiction to all published data reporting 10B-concentration ratios of tumor to normal tissues of >2.17, 27, 37, 38, 39 It might be argued that due to the high metabolic activity of the liver, the uptake of BPA in liver is higher than in brain, skin or other relevant organs at risk leading to decreased ratios between metastases and liver. This hypothesis however is not supported by the data published by Pinelli et al., who reported a 10B-concentration ratio between metastasis/liver of 5.9 in the first patient treated in Pavia.22 Consequently, a detailed evaluation of the uptake-mechanism and of the methods of measuring and reporting the 10B-concentration in tissue is necessary.
Cellular uptake of BPA is mediated by the L-amino acid transport system for neutral amino acids.40, 41 The biochemical rationale for its preferential uptake in tumors is the increased amino acid transport across the membrane and amplified protein synthesis rate, which are both early features of malignant transformation.42, 43 A- and L-type amino acid transport has been shown to be increased in tumor cells relative to normal tissue.44, 45 As BPA is transported actively, viable, metabolically active cells exclusively take up the compound. Consequently, if the 10B-concentration is assessed with a method such as PGRS, which measures the integral concentration within a tissue volume, the measured value does not reflect the 10B-concentration in the viable cells, the only valuable information needed for BNCT. The discrepancy between integral 10B-concentration in a macroscopic volume and the intracellular 10B-concentration is especially high if the volume of the evaluated tissue contains only a small amount of viable tumor cells, as is the case in metastases of colorectal carcinoma.
Several solutions are published to correlate histological information with the 10B-concentration. Some authors reported the integral 10B-concentration, as measured with PGRS or DCP-AES without a correction but did report separately on samples, which are macroscopically different. For example, Dagrosa and coworkers.46, 47 found selective BPA-uptake in undifferentiated thyroid carcinoma in dogs. Individual samples had tumor/blood ratios between 8.36 and 0.33. Histological investigation showed a positive correlation of the amount of viable tumor cells with the 10B-concentration. An in vivo model for soft tissue sarcoma was developed in rats by Pignol et al.,48 who found a high uptake in the tumor periphery containing more viable cells (36 μg 10B/g), but a low uptake in predominantly necrotic tumor areas (2 μg10B/g).
However, some authors have reported on measured 10B-concentrations values corrected for the proportion of viable cells. Coderre et al.27 analyzed the 10B-concentration of tumor samples from glioblastoma patients and introduced a “cellularity index” by scoring nuclei in histological slices. The 10B-concentration and cellularity index showed a linear correlation. For the respective clinical trials this group used the resulting ratio to calculate and report the radiation dose in the tumor. Roveda et al. and Nano et al.18, 20 describe a method in hepatic metastases of colorectal cancer, which quantifies the 10B-concentration and its distribution in 1 slice of tissue by neutron capture radiography. The radiograph image is compared with the histology of a directly neighboring slice and the 10B-concentration in the tumor is corrected for the proportion of viable tumor cells. This method was used to quantify the 10B-concentration for the 2 patients treated in Pavia, resulting in a ratio of 5.922 in the first and 5.6 in the second patient (Zonta, personal communication).
Actually, a standardized system to measure and report 10B-concentrations in tissue does not exist, therefore published data can not be easily compared. Form the above evidence, measured data in this study should be corrected for the average proportion of viable cells within the volume of tissue analyzed, prior the comparison with published values. In a first approach, a semiquantitative evaluation of the metastases was applied. The calculated 10B-concentration ratios of metastases/liver and metastasis/blood are consequently considerably higher (Table IV) reaching a 10B-concentration ratio between metastasis and liver of 5.5–9.5 which is comparable to the data measured by the group in Pavia. However, this approach does not reflect heterogeneities of the 10B-distribution within a small piece of tissue. Moreover, the preferential accumulation in viable tumor cells cannot be directly imaged (and proven) with this method. Nonetheless, the mechanism of BPA-uptake is understood well enough to justify such mathematical correction of the values measured with PGRS.
|Pat ID||10B concentration as measured by PGRS||Percentage of viable tumor cells||Correction factor||Calculated 10B concentration (ppm)||Calculated 10B ratio between metastasis/liver||Calculated 10B ratio between metastasis/blood|
|6||Ta: 11.88 ppm||20||5||59.4||7.9||11.8|
|Tb: 10.74 ppm||53.7||7.1||10.65|
|7||Central part: 6.45 ppm||15||6.7||43.6||5.6||7.4|
|Peripheral part: 11.67 ppm||78.9||9.5||13.38|
|8||Central part: 14.86 ppm||30||3.3||49.03||5.13||7.38|
|Peripheral part: 16.03 ppm||52.9||5.5||8.0|
In conclusion, BPA accumulates in hepatic metastases to an extent that allows for extracorporeal irradiation of the liver with BNCT to treat dissiminated liver metastases.
BSH does not preferentially accumulate in hepatic metastases of colorectal cancer. BPA is taken up preferentially in liver metastases of colorectal adenocarcinoma to an extent which is high enough for therapeutic BNCT. These findings justify further investigations with the aim to develop the technique toward a treatment modality. The most important goals to be achieved within the process are:
To develop in an animal model, the surgical procedures for autotransplantation, the procedure of anesthesia to bridge the anhepatic phase and to establish the maximum tolerable radiation dose in an explanted hypoxic liver at 4°C for low and high LET irradiation.
To continue a clinical trial with some patients to further analyze the preferential accumulation of BPA in liver metastases and to try to obtain more detailed information on 10B-uptake and distribution on a cellular level. Such a trial should be used to compare the different methods for 10B-analysis. A standardized reporting of the results has to be developed.
The authors wish to thank Prof. Dr. Martin Stuschke (Department of Radiation Oncology, University Duisburg-Essen) and the Medical Faculty Essen for the continuing support throughout the project. We gratefully acknowledge Mrs. Martina Rossi and Mrs. Anja Marr (Dept. of Medical Informatics, Biometry and Epidemiology, University Duisburg-Essen) and Mrs. Alessandra Busato (Data Center, EORTC) who gave excellent support in the data management. We are indebted to Mr. Klaas Appelman and Mr. Anton Hoving (NRG, Petten) for their technical assistance with prompt gamma ray measurements.
- 10European clinical trials of boron neutron capture therapy for glioblastoma. Nuclear News 1998; 41: 54–6., , , .
- 21Evaluation of selective boron absorption in liver tumors. Strahlenther Onkol 1989; 1989: 170–2., , , , , , , , , , , .
- 22TAOrMINA: from the first idea to the application to the human liver. In: SauerweinW,MossR,WittigA, editors. Research and development in neutron capture therapy. Bologna: Munduzzi Editore, 2002., , , , , , , , , , , , , ,
- 23Optimisation of treatment planning for multi-beam Boron Neutron Capture Therapy (BNCT) using epithermal neutron beams in patients with multiple metastases to the brain from malignant melanoma. Strahlenther Onkol 2006; 182: 109., , , , .
- 47Application of boron neutron capture therapy to the treatment of anaplastic thyroid carcinoma: current status and future perspectives. Curr Opin Endocrinol Diabetes 2005; 12: 352–5., , .