Complementary use of bioluminescence imaging and contrast-enhanced micro-computed tomography in an orthotopic brain tumor model

Development of an integrated dual-modality 3D bioluminescence tomography and ultrasound imaging system for small animal tumor imaging

Ultrasound (US) is a valuable tool for imaging soft tissue and visualizing tumor contours. Taking the benefits of US, we presented an integrated dual-modality imaging system in this paper that achieves three-dimensional (3D) bioluminescence tomography (BLT) with multi-view bioluminescence images and 3D US imaging. The purpose of this system is to perform non-invasive, long-term monitoring of tumor growth in 3D images. US images can enhance the accuracy of the 3D BLT reconstruction and the bioluminescence dose within an object. Furthermore, an integrated co-registered scanning geometry was used to capture the fused BLT and US images. We validated the system with an in vivo experiment involving tumor-bearing mice. The results demonstrated the feasibility of reconstructing 3D BLT images in the tumor region using 3D US images. We used the dice coefficient and locational error to evaluate the similarity between the reconstructed source region and the actual source region. The dice coefficient was 88.5%, and the locational error was 0.4 mm when comparing the BLT and 3D US images. The hybrid BLT/US system could provide significant benefits for reconstructing the source of tumor location and conducting quantitative analysis of tumor size.

A clinically relevant computed tomography (CT) radiomics strategy for intracranial rodent brain tumour monitoring

Here, we establish a CT-radiomics based method for application in invasive, orthotopic rodent brain tumour models. Twenty four NOD/SCID mice were implanted with U87R-Luc2 GBM cells and longitudinally imaged via contrast enhanced (CE-CT) imaging. Pyradiomics was employed to extract CT-radiomic features from the tumour-implanted hemisphere and non-tumour-implanted hemisphere of acquired CT-scans. Inter-correlated features were removed (Spearman correlation > 0.85) and remaining features underwent predictive analysis (recursive feature elimination or Boruta algorithm). An area under the curve of the receiver operating characteristic curve was implemented to evaluate radiomic features for their capacity to predict defined outcomes. Firstly, we identified a subset of radiomic features which distinguish the tumour-implanted hemisphere and non- tumour-implanted hemisphere (i.e, tumour presence from normal tissue). Secondly, we successfully translate preclinical CT-radiomic pipelines to GBM patient CT scans (n = 10), identifying similar trends in tumour-specific feature intensities (E.g. 'glszm Zone Entropy'), thereby suggesting a mouse-to-human species conservation (a conservation of radiomic features across species). Thirdly, comparison of features across timepoints identify features which support preclinical tumour detection earlier than is possible by visual assessment of CT scans. This work establishes robust, preclinical CT-radiomic pipelines and describes the application of CE-CT for in-depth orthotopic brain tumour monitoring. Overall we provide evidence for the role of pre-clinical 'discovery' radiomics in the neuro-oncology space.

Use of a Luciferase-Expressing Orthotopic Rat Brain Tumor Model to Optimize a Targeted Irradiation Strategy for Efficacy Testing with Temozolomide

Glioblastoma multiforme (GBM) is a common and aggressive malignant brain cancer with a mean survival time of approximately 15 months after initial diagnosis. Currently, the standard-of-care (SOC) treatment for this disease consists of radiotherapy (RT) with concomitant and adjuvant temozolomide (TMZ). We sought to develop an orthotopic preclinical model of GBM and to optimize a protocol for non-invasive monitoring of tumor growth, allowing for determination of the efficacy of SOC therapy using a targeted RT strategy combined with TMZ. A strong correlation (r = 0.80) was observed between contrast-enhanced (CE)-CT-based volume quantification and bioluminescent (BLI)-integrated image intensity when monitoring tumor growth, allowing for BLI imaging as a substitute for CE-CT. An optimized parallel-opposed single-angle RT beam plan delivered on average 96% of the expected RT dose (20, 30 or 60 Gy) to the tumor. Normal tissue on the ipsilateral and contralateral sides of the brain were spared 84% and 99% of the expected dose, respectively. An increase in median survival time was demonstrated for all SOC regimens compared to untreated controls (average 5.2 days, p < 0.05), but treatment was not curative, suggesting the need for novel treatment options to increase therapeutic efficacy.

Postoperative glioma segmentation in CT image using deep feature fusion model guided by multi-sequence MRIs

OBJECTIVES

Computed tomography (CT) and magnetic resonance imaging (MRI) are the most commonly selected methods for imaging gliomas. Clinically, radiotherapists always delineate the CT glioma region with reference to multi-modal MR image information. On this basis, we develop a deep feature fusion model (DFFM) guided by multi-sequence MRIs for postoperative glioma segmentation in CT images.

METHODS

DFFM is a multi-sequence MRI-guided convolutional neural network (CNN) that iteratively learns the deep features from CT images and multi-sequence MR images simultaneously by utilizing a multi-channel CNN architecture, and then combines these two deep features together to produce the segmentation result. The whole network is optimized together via a standard back-propagation. A total of 59 CT and MRI datasets (T1/T2-weighted FLAIR, T1-weighted contrast-enhanced, T2-weighted) of postoperative gliomas as tumor grade II (n = 24), grade III (n = 18), or grade IV (n = 17) were included. Dice coefficient (DSC), precision, and recall were used to measure the overlap between automated segmentation results and manual segmentation. The Wilcoxon signed-rank test was used for statistical analysis.

RESULTS

DFFM showed a significantly (p < 0.01) higher DSC of 0.836 than U-Net trained by single CT images and U-Net trained by stacking the CT and multi-sequence MR images, which yielded 0.713 DSC and 0.818 DSC, respectively. The precision values showed similar behavior as DSC. Moreover, DSC and precision values have no significant statistical difference (p > 0.01) with difference grades.

CONCLUSIONS

DFFM enables the accurate automated segmentation of CT postoperative gliomas of profit guided by multi-sequence MR images and may thus improve and facilitate radiotherapy planning.

KEY POINTS

• A fully automated deep learning method was developed to segment postoperative gliomas on CT images guided by multi-sequence MRIs. • CT and multi-sequence MR image integration allows for improvements in deep learning postoperative glioma segmentation method. • This deep feature fusion model produces reliable segmentation results and could be useful in delineating GTV in postoperative glioma radiotherapy planning.

Imaging of Mouse Brain Fixated in Ethanol in Micro-CT

Micro-CT imaging is a well-established morphological method for the visualization of animal models. We used ethanol fixation of the mouse brains to perform high-resolution micro-CT scans showing in great details brain grey and white matters. It was possible to identify more than 50 neuroanatomical structures on the 5 selected coronal sections. Among white matter structures, we identified fornix, medial lemniscus, crossed tectospinal pathway, mammillothalamic tract, and the sensory root of the trigeminal ganglion. Among grey matter structures, we identified basal nuclei, habenular complex, thalamic nuclei, amygdala, subparts of hippocampal formation, superior colliculi, Edinger-Westphal nucleus, and others. We suggest that micro-CT of the mouse brain could be used for neurohistological lesions evaluation as an alternative to classical neurohistology because it does not destroy brain tissue.

An orthotopic non-small cell lung cancer model for image-guided small animal radiotherapy platforms

METHODS:

An orthotopic non-small cell lung cancer model in NMRI-nude mice was established to investigate the complementary information acquired from 80 kVp microcone-beam CT (micro-CBCT) and bioluminescence imaging (BLI) using different angles and filter settings. Different micro-CBCT-based radiation-delivery plans were evaluated based on their dose-volume histogram metrics of tumor and organs at risk to select the optimal treatment plan.

RESULTS:

H1299 cell suspensions injected directly into the lung render exponentially growing single tumor nodules whose CBCT-based volume quantification strongly correlated with BLI-integrated intensity. Parallel-opposed single angle beam plans through a single lung are preferred for smaller tumors, whereas for larger tumors, plans that spread the radiation dose across healthy tissues are favored.

CONCLUSIONS:

Closely mimicking a clinical setting for lung cancer with highly advanced preclinical radiation treatment planning is possible in mice developing orthotopic lung tumors.

ADVANCES IN KNOWLEDGE:

BLI and CBCT imaging of orthotopic lung tumors provide complementary information in a temporal manner. The optimal radiotherapy plan is tumor volume-dependent.

A novel data management platform to improve image-guided precision preclinical biological research

OBJECTIVE:

Preclinical biological research is mandatory for developing new drugs to investigate the toxicity and efficacy of the drug. In this paper, the focus is on radiobiological research as an example of advanced preclinical biological research. In radiobiology, recent technological advances have produced novel research platforms which can precisely irradiate targets in animals and use advanced onboard image-guidance, mimicking the clinical radiotherapy environment. These platforms greatly facilitate complex research combining several agents simultaneously (in our example, radiation and non-radiation agents). Since these modern platform can produce a large amount of wide-ranging data, one of the main impediments in preclinical research platforms is a proper data management system for preclinical studies.

METHODS:

A preclinical data management system, inspired by current radiotherapy clinical data management systems was designed. The system was designed with InterSystems technology, i.e. a programmable Enterprise Service Bus solution. New DICOM animal imaging standards are used such as DICOM suppl. 187 for storing small animal acquisition context and the DICOM second generation course model.

RESULTS:

A small animal big data warehouse environment for research is designed to work with modern image-guided precision research platforms. Its modular design includes (1) a study workflow manager, (2) a data manager, and (3) a storage manager. The system provides interfaces to, e.g. preclinical treatment planning systems and data analysis plug-ins, and guides the user efficiently through the many steps involved in preclinical research. The system manages various data source locations, and arranges access to the data centrally.

CONCLUSION:

A novel preclinical data management system can be designed to improve preclinical workflow, facilitate data exchange between researchers, and support translation to clinical trials.

ADVANCES IN KNOWLEDGE:

A preclinical data management system such as the one proposed here would greatly benefit preparation, execution and analysis of biological experiments, and will eventually facilitate translation to clinical trials.

Bioluminescence tomography with structural information estimated via statistical mouse atlas registration

Due to an ill-posed and underestimated characteristic of bioluminescence tomography (BLT) reconstruction, a priori anatomical information obtained from computed tomography (CT) or magnetic resonance imaging (MRI), is usually incorporated to improve the reconstruction accuracy. The organs need to be segmented, which is time-consuming and challenging, especially for the low-contrast CT images. In this paper, we present a BLT reconstruction method based on a statistical mouse atlas to improve the efficiency of heterogeneous model generation and the accuracy of target localization. The low-contrast CT image of the mouse was first registered to the statistical mouse atlas model with the constraints of mouse surface and high-contrast organs (bone and lung). Then the other organs, such as the liver and kidney, were determined automatically through the statistical mouse atlas model. The estimated organs were then discretized into tetrahedral meshes for BLT reconstruction. The linearized Bregman method was used to solve the sparse inverse problem of BLT by minimizing the regularization function (L1 norm plus L2 norm with smooth factor). Both numerical simulations and in vivo experiments were conducted, and the results demonstrate that even though the localization of the estimated organs may not be exactly accurate, the proposed method is feasible to reconstruct the bioluminescent source effectively and accurately with the estimated organs. This method would greatly benefit the bioluminescent light source localization for hybrid BLT/CT systems.

Characterization of an Orthotopic Colorectal Cancer Mouse Model and Its Feasibility for Accurate Quantification in Positron Emission Tomography

PURPOSE

Quantification in positron emission tomography (PET) imaging of an orthotopic mouse model of colorectal cancer (CRC) is challenging due to difficult tumor delineation. We aimed to establish a reproducible delineation approach, evaluate its feasibility for reliable PET quantification and compare its added translational value with its subcutaneous counterpart.

PROCEDURES

A subcutaneous Colo205-luc2 tumor fragment harvested from a donor mouse was transplanted onto the caecum of nude mice, with (n = 10) or without (n = 10) the addition of an X-ray detectable thread. Animals underwent 2-deoxy-2-[¹⁸F]fluoro-D-glucose ([¹⁸F]FDG) PET imaging, complemented with X-ray computed tomography (CT) and magnetic resonance imaging (MRI, 7T). Animals without a thread underwent additional contrast enhanced (Exitron) CT imaging. Tumors were delineated on the MRI, μPET image or contrast enhanced μCT images and correlations between in vivo and ex vivo [¹⁸F]FDG tumor uptake as well as between image-derived and caliper-measured tumor volume were evaluated. Finally, cancer hallmarks were assessed immunohistochemically for the characterization of both models.

RESULTS

Our results showed the strongest correlation between both in vivo and ex vivo uptake (r = 0.84, p < 0.0001) and image-derived and caliper-measured tumor volume (r = 0.96, p < 0.0001) when the tumor was delineated on the MR image. Orthotopic tumors displayed an abundance of stroma, higher levels of proliferation (p = 0.0007), apoptosis (p = 0.02), and necrosis (p < 0.0001), a higher number of blood vessels (p < 0.0001); yet lower tumor hypoxia (p < 0.0001) as compared with subcutaneous tumors.

CONCLUSIONS

This orthotopic mouse model proved to be a promising tool for the investigation of CRC through preclinical imaging studies provided the availability of anatomical MR images for accurate tumor delineation. Furthermore, the tumor microenvironment of the orthotopic tumor resembled more that of human CRC, increasing its likelihood to advance translational nuclear imaging studies of CRC.

Automatic multiatlas based organ at risk segmentation in mice

OBJECTIVE:

During the treatment planning of a preclinical small animal irradiation, which has time limitations for reasons of animal wellbeing and workflow efficiency, the time consuming organ at risk (OAR) delineation is performed manually. This work aimed to develop, demonstrate, and quantitatively evaluate an automated contouring method for six OARs in a preclinical irritation treatment workflow.

METHODS:

Microcone beam CT images of nine healthy mice were contoured with an in-house developed multiatlas-based image segmentation (MABIS) algorithm for six OARs: kidneys, eyes, heart, and brain. The automatic contouring was compared with the manual delineation using three quantitative metrics: the Dice Similarity Coefficient (DSC), 95th percentile Hausdorff Distance, and the centre of mass displacement.

RESULTS:

A good agreement between manual and automatic contouring was found for OARs with sharp organ boundaries. For the brain and the heart, the median DSC was larger than 0.94, the median 95th Hausdorff Distance smaller than 0.44 mm, and the median centre of mass displacement smaller than 0.20 mm. Lower DSC values were obtained for the other OARs, but the median DSC was still larger than 0.74 for the left eye, 0.69 for the right eye, 0.89 for the left kidney and 0.80 for the right kidney.

CONCLUSION:

The MABIS algorithm was able to delineate six OARs with a relatively high accuracy. Segmenting OARs with sharp organ boundaries performed better than low contrast OARs.

ADVANCES IN KNOWLEDGE:

A MABIS algorithm is developed, evaluated, and demonstrated in a preclinical small animal irradiation research workflow.

NOTCH blockade combined with radiation therapy and temozolomide prolongs survival of orthotopic glioblastoma

Glioblastoma multiforme (GBM) is the most common malignant brain tumor in adults. The current standard of care includes surgery followed by radiotherapy (RT) and chemotherapy with temozolomide (TMZ). Treatment often fails due to the radiation resistance and intrinsic or acquired TMZ resistance of a small percentage of cells with stem cell-like behavior (CSC). The NOTCH signaling pathway is expressed and active in human glioblastoma and NOTCH inhibitors attenuate tumor growth in vivo in xenograft models. Here we show using an image guided micro-CT and precision radiotherapy platform that a combination of the clinically approved NOTCH/γ-secretase inhibitor (GSI) RO4929097 with standard of care (TMZ + RT) reduces tumor growth and prolongs survival compared to dual combinations. We show that GSI in combination with RT and TMZ attenuates proliferation, decreases 3D spheroid growth and results into a marked reduction in clonogenic survival in primary and established glioma cell lines. We found that the glioma stem cell marker CD133, SOX2 and Nestin were reduced following combination treatments and NOTCH inhibitors albeit in a different manner. These findings indicate that NOTCH inhibition combined with standard of care treatment has an anti-glioma stem cell effect which provides an improved survival benefit for GBM and encourages further translational and clinical studies.

Imaging of Orthotopic Glioblastoma Xenografts in Mice Using a Clinical CT Scanner: Comparison with Micro-CT and Histology

Purpose

There is an increasing need for small animal in vivo imaging in murine orthotopic glioma models. Because dedicated small animal scanners are not available ubiquitously, the applicability of a clinical CT scanner for visualization and measurement of intracerebrally growing glioma xenografts in living mice was validated.

Materials and Methods

2.5x10⁶ U87MG cells were orthotopically implanted in NOD/SCID/ᵞc^-/- mice (n = 9). Mice underwent contrast-enhanced (300 μl Iomeprol i.v.) imaging using a micro-CT (80 kV, 75 μAs, 360° rotation, 1,000 projections, scan time 33 s, resolution 40 x 40 x 53 μm) and a clinical CT scanner (4-row multislice detector; 120 kV, 150 mAs, slice thickness 0.5 mm, feed rotation 0.5 mm, resolution 98 x 98 x 500 μm). Mice were sacrificed and the brain was worked up histologically. In all modalities tumor volume was measured by two independent readers. Contrast-to-noise ratio (CNR) and Signal-to-noise ratio (SNR) were measured from reconstructed CT-scans (0.5 mm slice thickness; n = 18).

Results

Tumor volumes (mean±SD mm³) were similar between both CT-modalities (micro-CT: 19.8±19.0, clinical CT: 19.8±18.8; Wilcoxon signed-rank test p = 0.813). Moreover, between reader analyses for each modality showed excellent agreement as demonstrated by correlation analysis (Spearman-Rho >0.9; p<0.01 for all correlations). Histologically measured tumor volumes (11.0±11.2) were significantly smaller due to shrinkage artifacts (p<0.05). CNR and SNR were 2.1±1.0 and 1.1±0.04 for micro-CT and 23.1±24.0 and 1.9±0.7 for the clinical CTscanner, respectively.

Conclusion

Clinical CT scanners may reliably be used for in vivo imaging and volumetric analysis of brain tumor growth in mice.

An image guided small animal radiation therapy platform (SmART) to monitor glioblastoma progression and therapy response

BACKGROUND AND PURPOSE

Glioblastoma multiforme is the most common malignant brain tumor. Standard treatment including surgery, radiotherapy and chemotherapy with temozolomide is not curative. There is a great need for in vitro and in vivo models closely mimicking clinical practice to ensure better translation of novel preclinical findings.

METHODS AND MATERIALS

A 3D spheroid model was established using the U87MG cell line. The efficacy of temozolomide, RT and combinations was assessed using growth delay assays. Orthotopic glioblastoma tumors were established, different radiation doses delivered based on micro-CT based treatment planning (SmART-plan) and dose volume histograms (DVH) were determined. Tumor growth was monitored using bioluminescent imaging.

RESULTS

3D spheroid cultures showed a dose-dependent growth delay upon single and combination treatments. Precise uniform radiation was achieved in all in vivo treatment groups at all doses tested, and DVHs showed accurate dose coverage in the planning target volume which resulted in tumor growth delay.

CONCLUSION

We demonstrate that 3D spheroid technology can be reliably used for treatment efficacy evaluation and that mimicking a clinical setting is also possible in small animals. Both these in vitro and in vivo techniques can be combined for clinically relevant testing of novel drugs combined with radiation.

In vivo micro-CT imaging of untreated and irradiated orthotopic glioblastoma xenografts in mice: capabilities, limitations and a comparison with bioluminescence imaging

Small animal imaging is of increasing relevance in biomedical research. Studies systematically assessing the diagnostic accuracy of contrast-enhanced in vivo micro-CT of orthotopic glioma xenografts in mice do not exist. NOD/SCID/γc(-/-) mice (n = 27) underwent intracerebral implantation of 2.5 × 10(6) GFP-Luciferase-transduced U87MG cells. Mice underwent bioluminescence imaging (BLI) to detect tumor growth and afterwards repeated contrast-enhanced (300 µl Iomeprol i.v.) micro-CT imaging (80 kV, 75 µAs, 360° rotation, 1,000 projections, 33 s scan time, resolution 40 × 40 × 53 µm, 0.5 Gy/scan). Presence of tumors, tumor diameter and tumor volume in micro-CT were rated by two independent readers. Results were compared with histological analyses. Six mice with tumors confirmed by micro-CT received fractionated irradiation (3 × 5 Gy every other day) using the micro-CT (5 mm pencil beam geometry). Repeated micro-CT scans were tolerated well. Tumor engraftment rate was 74 % (n = 20). In micro-CT, mean tumor volume was 30 ± 33 mm(3), and the smallest detectable tumor measured 360 × 620 µm. The inter-rater agreement (n = 51 micro-CT scans) for the item tumor yes/no was excellent (Spearman-Rho = 0.862, p < 0.001). Sensitivity and specificity of micro-CT were 0.95 and 0.71, respectively (PPV = 0.91, NPV = 0.83). BLI on day 21 after tumor implantation had a sensitivity and specificity of 0.90 and 1.0, respectively (PPV = 1.0, NPV = 0.5). Maximum tumor diameter and volume in micro-CT and histology correlated excellently (tumor diameter: 0.929, p < 0.001; tumor volume: 0.969, p < 0.001, n = 17). Irradiated animals showed a large central tumor necrosis. Longitudinal contrast enhanced micro-CT imaging of brain tumor growth in live mice is feasible at high sensitivity levels and with excellent inter-rater agreement and allows visualization of radiation effects.