Community Research and Development Information Service - CORDIS


ImageLink Report Summary

Project ID: 323196
Funded under: FP7-IDEAS-ERC
Country: Germany

Mid-Term Report Summary - IMAGELINK (Multiparametric tumor imaging and beyond: Towards understanding in vivo signals)

In this project we want to gain a deeper understanding of the information that we can generate with multiparametric imaging by linking imaging to proteomics and metabolomics. This will enable us to define the molecular underlyings of imaging phenotypes.
To achieve this, we have successfully developed a milling machine that allows us to precisely excise tissue samples from small animals that were defined on multimodality multiparametric imaging. We have established a complete operating procedure including imaging experiments, rapidly freezing the animals, excising the tissue of interest and finally performing metabolomics or proteomics analysis with the tissue samples.
The used animal models were extensively tested and characterized by multiparametric imaging over a long time period in rats and mice in a subcutaneous and orthotopic setup and with different treatment regimens. Furthermore, we have developed an advanced imaging protocol and can now generate a maximum of information from each animal with high quality imaging data. The data analysis is currently in progress. We could show that during our freezing procedure the animals are frozen within 30 s, which is fast enough to ensure rapid halt of metabolism and high quality proteomics and metabolomics data. During our milling procedure, the temperature inside the animals does not exceed −27°C at any time and we could show that the accuracy in the x and y directions was 0.24±0.15 mm and 0.23±0.16 mm, respectively, while the accuracy in the z direction was 0.18±0.14 mm. The excised tissue samples were analyzed with nuclear magnetic resonance (NMR) and mass spectrometry (MS) for metabolomics and proteomics. The acquired data have proven to be of high quality and were processed with appropriate statistical methods and linking the imaging, proteomics and metabolomics data is planned.
At the same time we have developed several putative positron emission tomography (PET) tracers for the detection of senescence. We have used the HCT-116 tumor model for in vivo testing of two of the novel tracer molecules, since it is a well established model of senescence when treated with doxorubicin. One of the tracers showed significantly increased uptake in senescent tumors compared to untreated controls (control: 2.4±0.4%ID/cc; senescent: 3.8±1.1%ID/cc; p<0.0001), while the other tracer showed an increase without statistical significance (control: 1.1±0.4%ID/cc; senescent: 1.7±0.7%ID/cc). As a control we measured the already established PET tracers [F-18]FDG and [F-18]FLT which showed no significant difference in control and senescent tumors. Ex vivo analyses (beta-Galactosidase staining, immunohistology and autoradiography) supported our in vivo findings.
Furthermore, we have performed a longitudinal study with the CR-LRB-018P patient derived tumor rat model in an orthotopic and subcutaneous setting. Tumor bearing rats were measured at 3 time points starting at week 2 after tumor implantation. In the s.c. model [F-18]FDG uptake was significantly reduced at week 6 (0.9±0.3 %ID/cc)(p<0.05) compared to week 2 (1.3±0.3 %ID/cc). [F-18]FLT and [18-F]FMISO in s.c. tumors exhibited a significantly reduced uptake at week 6 (0.5±0.1 %ID/cc and 0.5±0.1 %ID/cc respectively) compared to week 2 (1.1±0.1 %ID/cc and 0.9±0.1 %ID/cc respectively; p<0.01). The o.t. tumor indicated no changes in [18-F]FMISO uptake over time and a significantly reduced uptake for [F-18]FDG at week 6 compared to week 2 (1.0±0.1 %ID/cc and 1.3±0.2 %ID/cc respectively; p<0.05) and a not significant reduction of [F-18]FLT uptake at week 6 compared to week 2 (0.5±0.1 %ID/cc and 1.0±0.5 %ID/cc respectively).
Finally, we have developed a novel machine learning framework to quantify different tumor tissue populations using longitudinal PET/MRI data and created holistic temporal profiles of growth of the necrotic and viable tumor portions. We have also devised a new algorithm to segment the tumor microenvironment using single time point imaging studies. The proposed methodology takes into account the multiparametric imaging parameters as well as a local neighborhood around each voxel and delivers a more accurate assessment of intratumor heterogeneity than the current state of the art intratumor segmentation techniques.
Overall the project is well in time and we are confident to reach all milestones within the time frame of the project.


Rebecca ROCK, (Fund manager of the Department)
Tel.: +49 7071 29 83450
Fax: +49 7071 29 4451
Record Number: 187716 / Last updated on: 2016-08-23