DOI: 10.1007/s00259-016-3585-0Pages: 392-397

Intra-lesional spatial correlation of static and dynamic FET-PET parameters with MRI-based cerebral blood volume in patients with untreated glioma

1. TU München, Department of Neuroradiology, Klinikum rechts der Isar

2. TU München, TUM Neuroimaging Center (TUM-NIC), Klinikum rechts der Isar

3. TU München, Department of Nuclear Medicine, Klinikum rechts der Isar

4. TU München, Department of Neurosurgery, Klinikum rechts der Isar

5. Klinikum Bayreuth, Department of Nuclear Medicine

Correspondence to:
Jens Göttler
Tel: +49 89 4140 7631




18F-fluorethyltyrosine-(FET)-PET and MRI-based relative cerebral blood volume (rCBV) have both been used to characterize gliomas. Recently, inter-individual correlations between peak static FET-uptake and rCBV have been reported. Herein, we assess the local intra-lesional relation between FET-PET parameters and rCBV.


Thirty untreated glioma patients (27 high-grade) underwent simultaneous PET/MRI on a 3 T hybrid scanner obtaining structural and dynamic susceptibility contrast sequences. Static FET-uptake and dynamic FET-slope were correlated with rCBV within tumour hotspots across patients and intra-lesionally using a mixed-effects model to account for inter-individual variation. Furthermore, maximal congruency of tumour volumes defined by FET-uptake and rCBV was determined.


While the inter-individual relationship between peak static FET-uptake and rCBV could be confirmed, our intra-lesional, voxel-wise analysis revealed significant positive correlations (median r = 0.374, p < 0.0001). Similarly, significant inter- and intra-individual correlations were observed between FET-slope and rCBV. However, rCBV explained only 12% of the static and 5% of the dynamic FET-PET variance and maximal overlap of respective tumour volumes was 37% on average.


Our results show that the relation between peak values of MR-based rCBV and static FET-uptake can also be observed intra-individually on a voxel basis and also applies to a dynamic FET parameter, possibly determining hotspots of higher biological malignancy. However, just a small part of the FET-PET signal variance is explained by rCBV and tumour volumes determined by the two modalities showed only moderate overlap. These findings indicate that FET-PET and MR-based rCBV provide both congruent and complimentary information on glioma biology.

To access the full text, please Sign in

If you have institutional access, please click here

  • Accepted: Nov 22, 2016
  • Online: Dec 2, 2016

Article Tools