Individual voxelwise dosimetry of targeted 90Y-labelled substance P radiotherapy for malignant gliomas

Abstract

Purpose: Substance P is the main ligand of neurokinin type 1 (NK-1) receptors, which are consistently overexpressed in malignant gliomas. The peptidic vector 111In/90Y-DOTAGA-substance P binds to these receptors and can be used for local treatment of brain tumours. Dosimetry for this interstitial brachytherapy has mainly been done using geometrical models; however, they often do not faithfully reproduce the in vivo biodistribution of radiopharmaceuticals, which is indispensable to correlate the deposited energy with clinical response. The aim of this study was to establish a reproducible dosimetry protocol for intratumoural radiopeptide therapy.

Methods: For test and therapeutic injections, 2 MBq of 111In-substance P and 370-3,330 MBq of 90Y-substance P, respectively, were applied in 12 patients with malignant gliomas. Over a period of 24 h, serial SPECT scans were performed on a dual-head SPECT camera. The scans were acquired in a double-energy window technique together with 99mTc-ECD in order to co-register the dose distributions with a separately acquired, contrast-enhanced CT scan. Quantitative voxelwise dose distribution maps (in Gy/GBq) were computed from these data using a mono-exponential decay approach. Pre- and post-therapeutic values were compared.

Results: Agreement between pre- and post-therapeutic dosimetry was very good and delivered absolute dose values in Gy per injected GBq. In all patients, the pretherapeutic test injection together with the CT overlay technique could predict the precise localisation of dose deposition in an anatomical context.

Conclusion: This protocol allows a precise pretherapeutic computation of the expected three-dimensional dose distribution and is clearly superior to the previously used dosimetry based on planar scintigraphic images. It has become an indispensable tool for planning intratumoural radiopeptide therapy in glioma patients.