In most studies, RIT of malignant gliomas was performed via local administration of the radioimmunoconjugate (either directly into the tumor or intracavitary into tumor cysts or the surgically created resection cavity) (Reardon et al. 2007; Zalutsky 2004, 2005). Clear advantages of this locoregional approach are bypass of the blood–brain barrier, high local radiation doses but low systemic toxicity and no major interference with potential human antibodies against mouse antigen (HAMA). For this form of application radiolabeled antbodies are administered via an Ommaya- or Rickham-Reservoir, which is surgically implanted during tumor resection or during a stereotactic procedure. The reservoir consists of a small elastic silicon reservoir, which is implanted subcutaneously with a connected catheter placed into the tumor, the tumor cyst or the resection cavity. All reagents may be injected through the reservoir directly into the resection cavity (Fig. 2).

The most frequently used radionuclide for RIT of malignant glioma is ¹³¹Iodine (Bigner et al. 1998; Cokgor et al. 2000; Reardon et al. 2006, 2008; Riva et al. 1999b). Advantages of the latter are the widespread experience using this radionuclide for other cancer treatments (especially for the treatment of differentiated thyroid cancer), its cost-effective production and the ease of synthesis. Disadvantages of ¹³¹I labeling are the considerable radiation exposure for the staff caused by the additional γ-particles. Furthermore, compared to other ß-emitters the maximal energy of ¹³¹Iodine is relatively low (0.61 MeV) resulting in a minor tissue penetration (medium/maximum tissue penetration range: 0.5/2.1 mm). Some studies were performed using ⁹⁰Y-labeled mAb (Grana et al. 2002; Paganelli et al. 2001, 2006; Riva et al. 1999a). Physical (physical half life: 64.4 h) and radiobiological features of ⁹⁰Yttrium make it also suitable for RIT. The high-energy ß-particles (maximum energy: 2.26 MeV) allow a deeper particle penetration compared to ¹³¹Iodine (medium/maximum tissue penetration range: 3.6/11 mm) resulting in higher radiation doses to the target area. Staff radiation exposure is much lower and, therefore, also allows treatment of high-maintainance patients. Another interesting radionuclide for RIT is the metallic ¹⁸⁸Rhenium. Physical half-life (17 h) is much lower compared to ¹³¹Iodine (8d). Similar to ⁹⁰Yttrium the maximum energy of ß-particles is high (2.12 MeV), resulting in a similar tissue penetration (medium/maximum tissue penetration range: 3.1/10 mm). Furthermore, additional γ-emission (155 keV) is useful for dosimetry. As a generator product ¹⁸⁸Rhenium is characterized by good availability. Another promising radioisotope, which generated promising results in the treatment of other tumors (e.g., neuroendocrine tumors) but—so far—was not used for RIT of gliomas is Lutetium-177. Its lower γ-ray energy could simplify radiation protection, facilitate dosimetric,and distribution measurements and could lead to a lower radiation exposure of the normal brain with a lower risk of radiation necrosis.

Besides these ß-emitters, the Auger emitter ¹²⁵Iodine was also used for RIT (Emrich et al. 2002; Quang et al. 2004). Since the penetration range of Auger electrons is only about 100 nm, incorporated auger emitters cause very high local radiation doses without harming the surrounding tissue. Therefore, a relevant radiation effect requires the internalization of the radioimmunoconjugate into the tumor cell. Over the past decade, new alpha emitters have also emerged and started to be considered for RIT. The specific properties of α-particles (shorter path range and high linear energy transfer) offer new opportunities for this therapeutic approach. However, so far RIT with α-emitters is predominantly used in preclinical settings due to some unsolved questions like optimization of radiolabeling and availability, improvement of administration modalities, and a better understanding of treatment effects.