Targeted alpha-radionuclide therapy of functionally critically located gliomas with 213Bi-DOTA-[Thi8,Met(O2)11]-substance P: a pilot trial
- First Online:
- Cite this article as:
- Cordier, D., Forrer, F., Bruchertseifer, F. et al. Eur J Nucl Med Mol Imaging (2010) 37: 1335. doi:10.1007/s00259-010-1385-5
- 405 Views
Functionally critically located gliomas represent a challenging subgroup of intrinsic brain neoplasms. Standard therapeutic recommendations often cannot be applied, because radical treatment and preservation of neurological function are contrary goals. The successful targeting of gliomas with locally injected beta radiation-emitting 90Y-DOTAGA-substance P has been shown previously. However, in critically located tumours, the mean tissue range of 5 mm of 90Y may seriously damage adjacent brain areas. In contrast, the alpha radiation-emitting radionuclide 213Bi with a mean tissue range of 81 µm may have a more favourable toxicity profile. Therefore, we evaluated locally injected 213Bi-DOTA-substance P in patients with critically located gliomas as the primary therapeutic modality.
In a pilot study, we included five patients with critically located gliomas (WHO grades II–IV). After diagnosis by biopsy, 213Bi-DOTA-substance P was locally injected, followed by serial SPECT/CT and MR imaging and blood sampling. Besides feasibility and toxicity, the functional outcome was evaluated.
Targeted radiopeptide therapy using 213Bi-DOTA-substance P was feasible and tolerated without additional neurological deficit. No local or systemic toxicity was observed. 213Bi-DOTA-substance P showed high retention at the target site. MR imaging was suggestive of radiation-induced necrosis and demarcation of the tumours, which was validated by subsequent resection.
This study provides proof of concept that targeted local radiotherapy using 213Bi-DOTA-substance P is feasible and may represent an innovative and effective treatment for critically located gliomas. Primarily non-operable gliomas may become resectable with this treatment, thereby possibly improving the prognosis.
KeywordsGliomaSubstance PTargeted therapy213Bi
Intracranial gliomas are intrinsic brain tumours that are classified according to their histological features by the WHO grading system. WHO grade IV gliomas (glioblastoma multiforme, GBM) carry the worst prognosis with an overall survival of often less than 12 months . WHO grade II gliomas feature a better prognosis with a median survival of 8–10 years. WHO grade III gliomas (anaplastic gliomas) exhibit a prognosis closer to GBM than to the grade II gliomas; grade I gliomas are regarded as potentially curable.
The resection of GBM, followed by combined radiochemotherapy, is accepted as the current therapeutic standard . A high extent of resection was found to be positively correlated with prolonged time to tumour recurrence [1, 3, 4]. Different technical developments such as 5-aminolevulinic acid for intraoperative tumour visualization , navigational systems or intraoperative MRI [6, 7] allow a high extent of resection to be achieved. However, if the tumour is critically located in brain areas that are responsible for functions such as speech or motor function, this aggressive therapeutic approach often cannot be efficiently applied. Preservation of functions is a major goal, and so the extent of resection needs to be limited. This results in higher residual tumour volumes, which represent suboptimal prerequisites for adjuvant radiochemotherapy. In the less aggressive low-grade gliomas (WHO grade II), the situation is more complex: therapeutic recommendations range from the “wait and see” strategy over radiotherapy and/or chemotherapy to an initial radical resection [8–12]. If low-grade gliomas are functionally critically located and affect important neurological functions of the patient, there is even more therapeutic insecurity whether taking a higher risk for functional deterioration improves the patient’s overall outcome.
Besides treatment of the main tumour mass, therapy of the infiltration zone needs to be addressed as a major goal, because 95% of gliomas exhibit local recurrence .
WHO grades II–IV gliomas have been shown to consistently overexpress the transmembraneous neurokinin type-1 receptor (NK-1). NK-1 receptors have also been detected on tumour cells infiltrating the intra- and peritumoural vasculature . The physiological ligand of NK-1 receptors is substance P. The local intratumoural injection of radiolabelled substance P exploits this overexpression of the NK-1 receptor. Substance P can be radiolabelled with various radionuclides for diagnostic or therapeutic applications using the chelators DOTAGA (1,4,7,10-tetraazacyclododecane-1-glutaric acid-4,7,10-triacetic acid) or DOTA (1,4,7,10-tetraazacyclododecane-1,4,7,10-triacetic acid). Radiolabelled DOTAGA-/DOTA-substance P exhibits a preserved affinity to the NK-1 receptor in the low nanomolar range; the in vivo stability has been confirmed . Its molecular weight of only 1.8 kDa facilitates sufficient and rapid intratumoural distribution after local application. Treatment with 90Y-labelled substance P led to good functional outcome and a trend toward improved survival when evaluated as a second-line treatment for recurrent gliomas . Tumour resection following this treatment was significantly facilitated by radiation necrosis and pseudoencapsulation of the tumour. We conducted another study to evaluate this treatment as a neoadjuvant modality: 90Y-labelled substance P was locally injected after biopsy confirmation of GBM, followed by resection of the pretreated glioma. This concept was feasible without signs of decompensating intracranial pressure and relevant toxicity. Of 16 patients, 14 stabilized or improved their neurological function; the pseudocapsule-like tumour demarcation allowed a high extent of resection to be achieved in subsequent surgery (submitted).
The radionuclide used in these studies, 90Y, has several advantageous characteristics: β-particle emitter with a mean energy of 2.1 MeV, mean tissue range of 5 mm, commercially available and standardized handling. However, in functionally critical areas of the brain, the tissue range of 90Y with the resulting “cross-fire effect”  potentially damages adjacent brain areas. Alternatively, α-particle-emitting radionuclides may be powerful candidates: 213Bi as an example has a mean tissue range of only 81 µm with a high mean energy of 5.8 MeV, allowing a highly cytotoxic radiation dose to be delivered to targeted cells while sparing adjacent healthy tissue.
Targeted alpha therapy using 213Bi has been shown to exhibit a favourable toxicity profile and therapeutic potential in phase 1 trials in leukemia, non-Hodgkin’s lymphoma and melanoma [17–19]. The short-lived alpha emitter 213Bi (T1/2=46 min) is available from a radionuclide generator loaded with its longer-lived mother nuclide 225Ac (T1/2=10 days), which can be produced via radiochemical extraction from 229Th  or cyclotron-driven methods . In this pilot study, we present the results of five patients with WHO grades II–IV gliomas who have been treated with locally injected 213Bi-DOTA-substance P as the primary therapeutic modality. Besides feasibility and toxicity as primary end-points, the functional outcome was examined as a secondary end-point.
Materials and methods
Patients and characteristics
Age at Dx (years)
Diagnosis/location of tumour
Tumour volume (cm3)
Barthel Index pre-/post-therapeutic
GBM frontal L callosal
GBM frontal L (SMA precentral)
Astro WHO grade III fronto-opercular L
Astro WHO grade II frontal R (SMA)
Astro WHO grade II occipital R
The criteria for inclusion were: newly diagnosed and histopathologically confirmed unifocal glioma (WHO grades II–IV); tumour diameter max. 5 cm; no evidence for obstruction of CSF circulation or decompensating intracranial pressure; Karnofsky performance score ≥70; age 18–75 years; absence of psychological, familial, sociological or geographical conditions potentially hampering compliance with the study protocol.
DOTA-[Thi8,Met(O2)11]-substance P was synthesized as described for the corresponding DOTAGA derivative .
Production of 225Ac/213Bi, radiolabelling and stability testing
225Ac/213Bi was produced at the Institute of Transuranium Elements (Karlsruhe, Germany) by radiochemical extraction from a 229Th source [20, 23]. 213Bi was eluted from a 225Ac/213Bi-generator using 600 µl 0.1 M NaI/HCl (Merck, Darmstadt, Germany, suprapure grade) solution. Addition of 50 µl of 20% ascorbic acid (Prolabo, VWR International, Fontenay sous Bois, France, normapure grade) as radioprotectant, pH adjustment to 8.5–8.7 with 70 µl of 2 M Na2CO3 (Merck, Darmstadt, Germany, suprapure grade). Incubation of buffered 213Bi eluate (270–1174 MBq) with 10 µg (n=5, patient 1) or 30 µg (n= 22, patients 2–5) DOTA-substance P for 5 min at 95°C in a microwave oven (Biotage Initiator, Uppsala, Sweden) and addition of 30 µl 1 mM Ca-DTPA to complex-free 213Bi . Quality control by standard gradient radio-HPLC using a Chromolith Speed ROD RP-18 endcapped 50-4.6 mm column (Merck, Darmstadt, Germany) with acetonitrile-water gradient and ion thin-layer chromatography (ITLC-SG, Pall Inc., New York, NY, USA) using 0.05 M sodium citrate solution (pH 5.5) as mobile phase, followed by sterile filtration. Synthesis, quality control and sterile filtration of 213Bi-DOTA-[Thi8,Met(O2)11]-substance P was performed in less than 15 min to minimize loss of activity. The radiochemical purity of the final product was 98.0 ± 1.4% at specific activities of 20.2 ± 3.9 MBq 213Bi/µg peptide.
The stability of 213Bi-DOTA-[Thi8,Met(O2)11]-substance P was tested in human blood serum. Serum samples were prepared from blood samples taken from healthy volunteers. An aliquot of 213Bi-DOTA-[Thi8,Met(O2)11]-substance P (18.5 MBq 213Bi/µg peptide) was added to a tenfold excess of blood serum and incubated at 37°C. At various time points, an aliquot was analysed using ion thin-layer chromatography (ITLC-SA, Pall Inc., New York, NY, USA) with 0.05 M sodium citrate solution (pH 5.5) as mobile phase. Under these conditions, intact 213Bi-DOTA-[Thi8,Met(O2)11]-substance P radioconjugate remains at the bottom of the ITLC strip (Rf=0), while free 213Bi released from the radioconjugate moves with the solvent front (Rf=1). The radioconjugate was found to be very stable towards dissociation, as within 5 h, corresponding to 6.5 half-lives of 213Bi, no release of 213Bi could be observed within the uncertainty of the measurement (<2%). This is in good agreement with a previous study investigating the stability of 213Bi-DOTA-[Thi8,Met(O2)11]-substance P in the presence of a 250-fold excess of DTPA as competing ligand using radio-HPLC. In this study, a very slow release of 213Bi with a half-life of 185 h was found.
Injection of the radiopharmaceutical
Intratumoural injection of the radiopharmaceutical is performed via one to three implanted catheters [25–27]. The catheter is connected to a subcutaneous port, which is punctured for injection (Fig. 1). Application of 12 mg dexamethasone and, immediately before injection, 60 g mannitol transiently reduces intratumoural pressure. Before injection, the system is flushed with 1.5 ml human albumin 5% to coat the plastic surface. The active drug is injected in a volume of 1 ml. Finally, the system is flushed again with 1.5 ml human albumin. For each therapeutic cycle, three to five injections were performed over 2 days.
Evaluation of receptor expression and dose distribution
Each patient was injected with 1 MBq of 111In-DOTA-[Thi8,Met(O2)11]-substance P per catheter as test injection before therapy. Images were acquired immediately after injection and 4 and 24 h after injection to confirm stable and orthotopic dose distribution. The images were acquired as 3-D data set (SPECT and low-dose CT) using an integrated dual-head SPECT/CT camera (SYMBIA T2, Siemens, Malvern, PA, USA) and automatic fusion.
111In SPECT images were acquired in continuous scan mode (25 min/rotation). The windows were centred over both 111In photon peaks (245 and 172 keV, width ±20%). Therapeutic activities were administered only if the activity was found stable within the tumour, i.e. no dose deposition outside the tumour margins on SPECT/CT over 24 h which was interpreted as specific binding.
After every application of 213Bi-DOTA-[Thi8,Met(O2)11]-substance P, SPECT/CT was acquired to confirm orthotopic dose deposition. For SPECT, the window was centred at 440 keV (width ±20%). SPECT was acquired with 128 views; acquisition time was 10 s/view.
Blood sampling was performed until 4 h after injection. Aliquots of 3 ml were analysed for 213Bi activity using a NaI(Tl) bore well counter (COBRA II, D 5003 γ-system, Canberra Packard, Melbourne, Australia). The percentage of injected 213Bi in the blood pool was determined based on calculated total blood volumes.
Patients were clinically evaluated weekly within the first 8 weeks after radiopeptide application, afterwards monthly for the grade III and IV tumours and in 3-month intervals for the grade II tumours. MR imaging was performed monthly after injection of the radiopharmaceutical for the first 3 months, thereafter in 3-month intervals for the grade III and IV tumours and in 6-month intervals for the grade II tumours.
Patient characteristics and functional status
Two patients with low-grade astrocytomas (WHO grade II), one patient with anaplastic glioma (WHO grade III) and two patients with GBM (WHO grade IV) in functionally critical locations were treated with intratumourally injected 213Bi-DOTA-substance P as the primary therapeutic modality (Table 1). The preliminary diagnosis from the fresh-frozen sections was confirmed by definite histology in all cases. Four patients received 1.07–2.00 GBq 213Bi-DOTA-substance P within one therapeutic cycle, and one patient received four therapeutic cycles with a total activity of 7.36 GBq. The pre-therapeutic functional score, evaluated by the Barthel Index , was 100 out of 100 in the patients with the WHO grade II and III tumours (patients 3, 4 and 5). In these patients, this unimpaired functional status did not change during or after treatment. The two GBM patients displayed pre-therapeutic Barthel scores of 75 (patient 1) and 80 (patient 2), which means moderate restrictions in activities of daily living. Both GBM patients improved to a score of 90 after treatment, being consistent with minor impairment in daily living (Table 1).
Biodistribution and pharmacokinetics
Patient characteristics and response
Side effects and toxicity
Intratumourally injected 213Bi-DOTA-substance P was tolerated well by all patients; no relevant acute local or systemic toxicity could be found according to the National Cancer Institute (NCI) Common Toxicity Criteria (CTC) scale (version 2.0). In the long-term follow-up, no further substantial local or systemic toxicity was found. Corticosteroid therapy generally was tapered 2 weeks after radiopeptide application.
Functionally critically located malignant gliomas still represent an immense therapeutic challenge, because often the therapeutic standards for an aggressive local therapy, i.e. surgery and radiotherapy, cannot be applied. There is no doubt about the importance of sufficient local control of high- and low-grade gliomas, since virtually all of them exhibit local recurrence. In critically located gliomas, the current technical developments for a higher extent of surgical resection [5, 7, 28] are only partially useful because in this entity of tumours they carry a high risk of damaging neurological function. Awake surgery and/or intraoperative monitoring techniques represent valuable tools to avoid postoperative neurological deficits such as loss of motor function or speech, but remain without therapeutic effect on the prognostically pivotal infiltration zone. Consequently, patients with critically located gliomas are often not considered as candidates for an aggressive local therapy.
The fact that WHO grades II–IV gliomas do consistently overexpress NK-1 receptors led to the therapeutic local application of the radiolabelled NK-1 ligand DOTAGA-/DOTA-substance P. The concept, besides treatment of the main mass, is a targeted approach to tumour cells in the surrounding infiltration zone. Previous studies have shown feasibility and low toxicity of this method as a second-line therapy  as well as in a neoadjuvant setting (submitted). However, the radionuclide 90Y, which was used in these previous studies, may lead to unacceptable damage to adjacent brain areas because of its relatively high mean tissue range of 5 mm. The possibility of using 213Bi instead of 90Y may be a safer and even more efficacious alternative: The mean tissue range is only 81 µm with a high mean energy of 5.8 MeV (90Y: 2.1 MeV), corresponding to a high linear energy transfer (LET) of 72 keV/µm. Consequently, the distance between ionizing events caused by alpha particle emission of 213Bi is similar to the distance between DNA double strands, thus increasing the probability of inducing irreparable DNA double strand breaks. These prerequisites led to the design of the pilot study presented in this work: The histological diagnosis of a glioma in patients with eloquently located tumours was established by biopsy, followed by implantation of intratumoural catheters. Subsequently, 213Bi-DOTA-substance P was intratumourally injected as the primary therapeutic modality. SPECT/CT and blood sampling confirmed high retention of the radiopharmaceutical at the tumour site, in agreement with absence of toxicity to non-targeted, healthy brain tissue or other organs. MR imaging showed intratumoural changes highly suspicious for radionecrosis, which was confirmed by histopathological examination. Small tumours exhibited a complete radionecrotic appearance, whereas larger tumours seemed to be mainly necrotic in the proximity of the implanted catheters. This observation is probably due to the fact that the physical half-life of 213Bi is only 46 min, so that a significant fraction of this radionuclide is possibly decayed before a sufficient intratumoural distribution is achieved. In large tumours, possibly different radionuclides such as 225Ac with longer half-lives could be more efficacious. However, in one of two GBM patients and in the patient with the anaplastic glioma there is, despite incomplete radionecrotic demarcation of the tumours, a long time interval to further progression, so that possibly there is also a therapeutic effect in areas without radiographic evidence for radionecrosis. To the contrary, the first GBM patient, who received a relatively low dose, had only a short time interval to tumour progression. This observation may indicate a dose-effect relationship that needs to be confirmed in a larger number of patients.
The most benefit from the application of 213Bi-labelled substance P is probably seen in the patients with relatively small tumours (12.0 and 17.1 cm3), where sufficient intratumoural distribution could be achieved. Radiographically these tumours are completely necrotic demarcated; in the one patient operated on after radiopeptide therapy there was no evidence for viable tumour cells in the surrounding area. It is remarkable that the demarcated area is much larger than the radiographically evident tumour area on initial MR imaging, being indicative for the real tumour dimensions and suggestive for a therapeutic effect in the infiltration zone. Consequently, patients with small, symptomatic gliomas may be ideal candidates for this therapeutic modality: there is no significant risk for decompensation of the intracranial pressure and the intratumoural distribution of the radiopharmaceutical is sufficient to target the entire mass as well as tumour cells in the infiltration zone. For these reasons, further evaluation of this concept in this subgroup of gliomas is planned.
Targeted therapy of critically located WHO grade II–IV gliomas with locally injected 213Bi-DOTA-substance P is feasible and without relevant toxicity. Compared to therapeutic approaches using beta emitters, the treatment of gliomas using short-range alpha emitters may allow similar efficacy to be achieved with lower toxicity to healthy brain areas. Due to the relatively short half-life of 213Bi, this innovative concept probably has most of its therapeutic potential in the treatment of small, critically located gliomas and is planned for further evaluation in this group of patients.
This work was supported by the European Commission FP7, contract TARCC No.HEALTH-F2-2007-201962.