This grant focuses on research to advance automated neuroimaging techniques that will enhance the potential of human neural stem cell (hNSC) implantation to treat neonatal hypoxic ischemic brain injury (HII), a common condition with serious long-term neurological and developmental disorders such as mental retardation, cerebral palsy or epilepsy. Our goal is to develop magnetic resonance imaging (MRI) and imaging processing methods in a rat pup model of neonatal hypoxic-ischemic brain injury. This model is created by tying off the middle cerebral artery and exposing the pup to hypoxia and is known as the Rice-Vannucci model (RVM). We will use complex MRI methods to automatically determine where areas of brain injury are located as well as where human neural stem cells (hNSCs) are implanted. At the present time, there are no such methods available. Developing such methods are critical if we are to use stem cells for treatment of different diseases. In the first aim (Aim1a) we will implant hNSCs into a special gel (agarose) and image the cells in a MRI scanner. We will determine the optimal number of cells to use for imaging and the strength of the MRI magnet that is most suitable. We will also look at which of four specific type of imaging modalities will give us the most useful information about hNSC activity. In Aim1b, we will compare 3 computational methods using the 10d rat pup RVM at 1, 3, 7 days after implantation. These 3 complex methods include: (i) hierarchical region splitting (HRS), (ii) graph normalized cut (N-Cut) algorithm, and (iii) Bayesian region segmentation (BRS). Testing different computational methods will help us find the one that is best at detecting and monitoring stem cells as they move through brain regions and migrate to the areas of ischemic injury. When stem cells will be used in clinical trials, it will be very important to know if they will migrate to injured tissues, when this happens, how long the cells can survive for and also the degree of increase in their volume (i.e., proliferation). In Aim 2 we will quantify the activities of the stem cells after hypoxic-ischemic injury. In Aim 2a, we will test these computational methods to see how well they describe and detail the evolution and location of hypoxic-ischemic injury as well as the implanted stem cells. In Aim 2b we will correlate the results of the information that we acquire from MRI and the 3 computational methods with examination of the brain by doing tissue staining and microscopic examination (i.e., histology and immunohistochemistry). This will allow us to check the accuracy of the imaging methods. This grant addresses many of the important technical issues that must be considered in order to improve the chance for success of NSC therapy. We strongly believe that automated and quantitative imaging analysis will be one of the most significant technological advances that will increase the likelihood for successful stem cell therapy.
Statement of Benefit to California:
This proposal will benefit children born in California who suffer acute hypoxic brain injury at birth (i.e., perinatal asphyxia or ischemic perinatal stroke) and who develop long term neurological and developmental disorders such as mental retardation, cerebral palsy or epilepsy. Treatment with stem cells should reduce the burden of illness of this very common disorder that occurs in full term newborn infants. It will also benefit the families and caretakers of affected children as it will be easier to provide care for such children as they will have less severe neurological deficits. The long-term outcome of severe perinatal brain injury is a life-time of disability and if this can be reduced, it will also reduce the costs of chronic care, hospitalizations, rehabilitation and care in long-term nursing facilities of affected children. This grant focuses on one critical and important aspect of this endeavor that is related to available automated, reproducible, rapid, and objective methods to quantify and assess hNSC activity both in animal studies but more importantly when stem cells are used in human clinical trials. By developing the imaging methods to help improve candidate selection for treatment, for monitoring stem cell implantation, and for following the effects of treatment on outcome, the studies outlined in this proposal will provide ways to use these automated imaging methods to develop ‘push-button’ automated, fast, unbiased and reproducible software tools that will be used clinically not only for treatment with stem cells but for many other acquired diseases that occur across the pediatric age spectrum.
This proposal is focused on the development of MRI technology to identify and track transplanted human neural stem cells (hNSCs) in a rat model of neonatal ischemic brain injury. The technology would allow non-invasive, in vivo imaging of transplanted hNSCs with reference to specific brain anatomy, including the ischemic lesion sites. The applicants also propose to develop software to quantify transplanted hNSC number, proliferation and potential repopulation of lesion sites. The group plans to first optimize hNSC detection in agarose gels before continuing in vivo. They plan to verify and inform their imaging based localization and quantification techniques using histology and immunohistochemistry.
In terms of impact, reviewers agreed that the proposed research has the potential to provide significant impact to the assessment of the safety, timing, and cellular properties of transplantable SCs in animal models with neural injuries. Moreover, one reviewer commented that this proposal has the potential to provide a platform for further animal model studies that include other cell-based therapies to address spinal cord and other injuries. However, despite the budding relevance of this type of imaging technology development for preclinical applications, reviewers disagreed about its clinical applicability.
Reviewers raised serious concerns about the feasibility of the proposal. They described the work as technically challenging, especially given the need for two different MRI devices to accommodate growing animals. Another concern was related to the software development proposed; namely, whether it would be applicable to other imaging systems and whether it would be open-source. One reviewer commented that the development of automated software tools is time and labor intensive and raised doubts about the attainability of the proposed milestones. Several other concerns were raised with respect to the experimental design. A reviewer noted that the applicant acknowledges the difficulty of detecting low numbers of migratory hNSCs, and so suggests larger sized transplants. However, only a subset of those cells may migrate, severely limiting their approach to monitor the location of the transplant. One reviewers’ concern was that the iron label used in the hNSCs might leak out, especially as the cells undergo apoptosis, and be taken up by neighboring microglia. Subsequently, the label would track microglia and the ordinary migration of these cells rather than any movement of stem cells. However, the reviewer noted that the use of hNSCs in primate and histological correlation with anti-human antibodies should reveal if this is a significant problem. Finally, some reviewers questioned whether the technical optimization that comprises a large part of the proposal would ever be transferable to humans.
In terms of the research team, reviewers appreciated the merits of collaboration between an MRI specialist, a physicist and a stem cell biologist. They agreed that the assembled team is well qualified to perform the MRI optimization and ischemic animal generation. However, reviewers found it unclear who in the team had the expertise in NSC transplantation. Furthermore, two reviewers noted that that the stem cell biologist is not local to the principal investigator and wondered how the 30% listed effort could be achieved from a distance. The proposal also left unclear who on the team would be responsible for developing the software tools, a significant time investment. Finally, reviewers also raised questions about the budget. They commented that the $288,000 requested for consultants and subcontracts was poorly justified and were confused by the “other expenses” category that lists $22K in imaging costs per year but produces requests of $167,000 and $158,000.
Overall, while this proposal addresses some important technical roadblocks for transplanted stem cell imaging, the reviewers raised serious questions about its clinical relevance, feasibility and budgetary justification.
During programmatic review a motion was made to move this proposal up from Tier 3, Not Recommended for Funding, to Tier 2, Recommend for Funding if Funds are Available. The maker of the motion argued that there is a real need for this type of MRI technology development and disagreed with the reviewer that questioned its clinical applicability. A member of the panel raised concern about the high number of imaging proposals already in the top funding tier but others argued that this is a critical area of research that deserves emphasis. Overall, the panel agreed with the maker of the motion that the technology and software solutions developed in this proposal could bootstrap advances in human imaging techniques. The motion to move this proposal from Tier 3 to Tier 2 carried.