Hyperpolarized 13C MRSI is a better predictor of survival than tumor size in treated GBM models

Presenting Author Senior Author
Name: Marina Radoul Name: Sabrina Ronen
Email: Email:
Presenting Author’s RIG/SRG: Brain Cancer MRI/MRS  
Presenting Author's Lab Location: Mission Bay   

Abstract Information
Imaging Modality: MR
Disease Application: Brain Cancer
Complete author list: Marina Radoul, Myriam M Chaumeil, Pia Eriksson, Aaron S Wang, Joanna J Phillips, Sabrina M Ronen
Abstract highlights: Our study demonstrates metabolic response of two glioblastoma models to inhibition of PI3K/mTOR/Akt signaling pathway, activated in most glioblastoma cases. Our data show that Lac/Pyr predicts significantly longer survival prior to a detectable tumor size change in both tumor models, which is extremely important in a light of clinical translation.
Glioblastoma (GBM) is one of the most aggressive brain tumor types. Current standard of care for GBM is surgical resection, radiation, and treatment with Temozolomide (TMZ). However, the heterogeneous and diffuse growth of GBM leads to resistance to traditional therapies. Complementary therapeutic approaches are therefore being investigated. The PI3K/Akt/mTOR signaling pathway is activated in ~88% of GBM [1] and is an important therapeutic target. Inhibition of the pathway leads to decreased lactate dehydrogenase A (LDH-A) expression and activity, and therefore decreased production of lactate [2, 3]. The goal of this study was to monitor metabolic response of two GBM tumor models, GS-2 and U-87, to XL765, a second-generation dual PI3K/mTOR inhibitor, alone or in combination with TMZ. This study would provide a valuable model for upcoming clinical trials of hyperpolarized 13C MRSI on GBM patients in clinical trials with PI3K/Akt/mTOR inhibitors.
Athymic nu/nu mice (6 weeks old) were injected intracranially with 3x105 GS-2 or U-87 cells [3]. Once tumors reached a diameter of ~3mm, animals were randomly divided into four treatment groups: XL765, XL765/TMZ, TMZ and controls. MRI studies were performed using a vertical wide bore Agilent 600MHz scanner. Hyperpolarized [1-13C]-pyruvic acid was injected through i.v. tail-vein catheter over 12s. 13C MRSI spectra were recorded 17s after injection using 2D-CSI (TE/TR=0.58/66ms, frequency dimension=256, phase dimension=16x16) and processed using Sivic [5]. Peak integrals were normalized to normal brain and compared between control and treated groups.
Figures A-C and D-F exhibit results obtained from GS-2 and U-87 models, respectively. Fig. A and Fig. D show the evolution of tumor size during treatment. They show that the two GBM models respond slightly differently to the same treatment. Most notably they show that XL765 does not affect the size of GS-2 tumors but leads to shrinkage of U87 tumors. However, in vivo 13C MRSI exhibited a significant decrease in average hyperpolarized [1-13C] Lac/Pyr in both models and for every treatment group when compared to control at day 9±2 for GS-2 (Fig. B) and at day 6±1 for U-87 (Fig. E), time points at which no change in tumor size could be detected. These results were associated with inhibition of PI3K/Akt/mTOR signaling and LDH-A expression (data not shown). Importantly, in every treatment group and in both models survival was longer (Fig. C and Fig. F).
Overall our data therefore show that Lac/Pyr predicts significantly longer survival prior to a detectable change in tumor size. Upon clinical translation, this approach could assist the treatment decision-making process based on individual response to drug at early time points when tumor size is still unchanged.