Multi-modal MR spectroscopic imaging and PET techniques for metabolic profiling of cancer

Multi-modal imaging techniques in treatment monitoring and diagnostics of cancerThe project focuses on the development and application of noninvasive imaging methods within Magnetic Resonance (MR) modalities, including MR spectroscopic imaging (MRSI). MRSI provides information on cancer-derived metabolic abnormalities, which can be valuable for prognostication, treatment monitoring/evaluation, and diagnosis in glioma patients.In collaboration with Massachusetts General Hospital and Harvard University in Boston, USA, the potential role of MRSI technique in predicting metabolic response following anti-cancer therapy was examined. MRSI data were acquired during a combined treatment with an angiogenesis-inhibiting drug and radiochemotherapy in patients with glioblastoma tumor, the most aggressive brain tumor. Our imaging study was particularly designed around the goal of identifying early signs of treatment response. We demonstrated that metabolic biomarkers acquired by MRSI at the early stage of vascular normalization can predict early the outcome of combined cediranib/standard treatment. The results showed that including metabolic information can greatly improve the early response prediction compared to that of conventional MRI methods. The results from this study was submitted for publication in Precision Oncology. The revised version of the manuscript is under review. In addition, the results from this study was submitted to an international scientific conference (ISMRM, 2017). Two other studies were also carried out in the USA, (i) investigating the role of edited MR spectroscopic techniques and 2-dimensional MRS in clinical studies and (ii) developing a new MRS method to unambiguously detect 2-hydroxyglutarate metabolite in vivo. The first study was published in Analytical Biochemistry and the preliminary results of the second study submitted to ISMRM conference. In collaboration with Department of Radiology, Haukeland University Hospital and Department of Clinical Medicine, University of Bergen, we investigated the role of MRSI in predicting the tumor grade of endometrial cancer. We found a significant correlation between choline levels in tumor and risk features, suggesting that in vivo MRSI may potentially aid in more accurate preoperative risk stratification in endometrial cancer. A manuscript from this study is submitted for publication in European Radiology. In collaboration with Department of Neurosurgery at St Olav University Hospital, Morteza Esmaeili is now finalizing a project aiming to generate a growth pattern vector field for glioblastoma tumors. Brain tumors can interfere with global functional network organization, rather than affecting only the site of the lesion, by invading the white matter. The dominant growth-direction of brain tumors along the orientation of white matter fibers has been suspected, however, this theory may be examined using 3-dimensional MRI data. Here we aimed to investigate the hypothesis that glioma tumor cells tend to track along the white matter fibers. A manuscript is currently under preparation, and will be submitted to a peer-reviewed journal this spring. In collaboration with St Olav Hospital, we also aimed to investigate the role of multi-modal imaging in glioma characterization. The multimodal imaging technology of positron emission tomography and MR imaging (PET-MRI) has recently been realized on a clinical scale and its impact on neuro-oncology studies is starting to be investigated. Combining PET and MRSI data, a clinical study has been started on PET-MRI system in St. Olav since last year, investigating the role of multi-modal imaging modalities in the degree of improvement in tumor characterization of patients carrying gliomas. This is an ongoing study and we are recurring more glioma patients.

Multi-modal MR spectroscopic imaging and PET techniques for metabolic profiling of cancerDuring 2015, the MR spectroscopic imaging data acquired from human brains were analyzed. To facilitate 3D MRSI data analysis and visualization, in house scripts were developed and examined. The results have provided spatial metabolic profile maps from the entire brain, thus an improved tool for investigating tumor types initiated/infiltrated close to the cortical areas rich in fat content. Metabolic quantification in these areas has been a challenge using conventional MRSI protocols for glioma studies. The results was submitted and accepted for publication in Magnetic Resonance in Medicine journal. In addition, the results from the proposed MRSI protocol were presented in national and international scientific conferences in 2015. Doctoral Fellow Mingshu Shi, supervised by Morteza Esmaeili, submitted an article on 31P MRS investigation of energy metabolism alterations in heart failure tissues of rat models subjected to different exercise training intensities. Mingshu has now initiated the second 31P MRS study to investigate high phosphate energy metabolism in muscle tissues obtained from trained rat models with different VO2max capacities. The aim of this study is to investigate the impact of exercise on energy metabolism of muscles post-training at different age. In collaboration with the group at MGH/Harvard in the USA in 2015, Morteza was previously involved in a clinical trial study of forty glioblastoma patients with proposed a post-processing method. The candidate was involved in data analysis and is significantly contributing in preparation of the manuscript. Data analysis was almost finished and an article on this clinical study will be completed during 2016. The purchased 31p coil - by MR Core facility in Trondheim - could not be used as scheduled in 2015 because of some technical problems with MR system. Furthermore, the MR system will be temporarily shut downed for 3 months, from February till 1. June 2016. In order to alternatively be more efficient during this period, Morteza Esmaeili has therefore planned 2-months more stay at MGH/Harvard in Boston with intent to; (i) follow up and finalize the clinical trial study, (ii) to improve a post-processing platform for analyzing big MRSI data semi-automatically, (iii) further develop and apply the proposed method (published in 2015) on glioma patients to include 2-hydroxyglutarate quantification in metabolic map of entire brain. In 2015 and in collaboration with department of neurosurgery at St Olav hospital, Morteza Esmaeili initiated a study project for quantitative analysis of MR images obtained from glioblastoma patients. In this study the growth pattern of glioma tumors, within a certain monitoring period, will be investigated. The results from this ongoing study will be published in 2016.

Multi-modal MR spectroscopic imaging and PET techniques for metabolic profiling of cancerMagnetic resonance spectroscopic imaging (MRSI) is a non-invasive MR imaging method for metabolic mapping of biological tissue. This technique is now established in clinical MR systems and has been applied for diverse pathological studies in the human brain. Metabolism is an emerging cancer hallmark and reprogramming of metabolism is associated with many other important cancer hallmarks such as cellular proliferation, angiogenesis and mutations.The objectives of this project are to; i) establish and perform an in vivo Fast-MRSI protocol to quantify important metabolites in the human brain, ii) MRSI of human derived preclinical glioma models to study metabolic alteration during adaptive mechanisms responsible for resistance to antivascular therapy, iii) combine MRSI and positron emission tomography (PET) data of glioma and breast cancer models to comprehensively study the abnormal phospholipid metabolism in cancer cells. As the first step, the post-doctoral candidate started with developing the MRSI sequence protocol in a joint collaboration with MGH/Harvard in Boston, USA. Conventional MR sequence programs, available in clinical MR systems, do not sufficiently suppress the high intensity water and lipid signals in proton MRSI. Domination of these signals in MR spectra of the human brain hampers metabolic quantification, in particular. On the other hand lower sensitivity of MRS compared to the structural MR imaging together with the subject movements inside the scanner reduce the spectral quality and localization accuracy. Faster MRSI acquisition will significantly improve the latest limitations. We therefore aimed to establish an advanced sequence program to; (i) accelerate MRSI acquisition, (ii) Detect and correct motion artifact prospectively and real time, and (iii) develop a more sufficiently signal suppression on the lipid and water resonances. These improvements will ultimately enhance the metabolic profiling of glioma patients as well as other pathological studies of brain. The new method can be used to study metabolic alteration in glioma patients undergoing treatment to evaluate treatment response. During the research stay from July 2014-February 2015, the protocol has been developed and successfully tested in vivo in healthy volunteers. Preliminary results were submitted as an abstract to the ISMRM 2015 meeting in Toronto (1). The results will also be presented at coming national conferences. The project did not result in any published papers its first 6 months, but a publication based on the new MRSI development in Boston is planned in 2015.