Model discovery approach enables noninvasive measurement of intra-tumoral fluid transport in dynamic MRI.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is a routine method to noninvasively quantify perfusion dynamics in tissues. The standard practice for analyzing DCE-MRI data is to fit an ordinary differential equation to each voxel. Recent advances in data science provide an opportunity to move beyond existing methods to obtain more accurate measurements of fluid properties. Here, we developed a localized convolutional function regression that enables simultaneous measurement of interstitial fluid velocity, diffusion, and perfusion in 3D. We validated the method computationally and experimentally, demonstrating accurate measurement of fluid dynamics in situ and in vivo. Applying the method to human MRIs, we observed tissue-specific differences in fluid dynamics, with an increased fluid velocity in breast cancer as compared to brain cancer. Overall, our method represents an improved strategy for studying interstitial flows and interstitial transport in tumors and patients. We expect that our method will contribute to the better understanding of cancer progression and therapeutic response.

Structural and practical identifiability of contrast transport models for DCE-MRI.

Contrast transport models are widely used to quantify blood flow and transport in dynamic contrast-enhanced magnetic resonance imaging. These models analyze the time course of the contrast agent concentration, providing diagnostic and prognostic value for many biological systems. Thus, ensuring accuracy and repeatability of the model parameter estimation is a fundamental concern. In this work, we analyze the structural and practical identifiability of a class of nested compartment models pervasively used in analysis of MRI data. We combine artificial and real data to study the role of noise in model parameter estimation. We observe that although all the models are structurally identifiable, practical identifiability strongly depends on the data characteristics. We analyze the impact of increasing data noise on parameter identifiability and show how the latter can be recovered with increased data quality. To complete the analysis, we show that the results do not depend on specific tissue characteristics or the type of enhancement patterns of contrast agent signal.

Pregnancy-induced remodeling of the murine reproductive tract: a longitudinal in vivo magnetic resonance imaging study.

Mammalian pregnancy requires gradual yet extreme remodeling of the reproductive organs to support the growth of the embryos and their birth. After delivery, the reproductive organs return to their non-pregnant state. As pregnancy has traditionally been understudied, there are many unknowns pertaining to the mechanisms behind this remarkable remodeling and repair process which, when not successful, can lead to pregnancy-related complications such as maternal trauma, pre-term birth, and pelvic floor disorders. This study presents the first longitudinal imaging data that focuses on revealing anatomical alterations of the vagina, cervix, and uterine horns during pregnancy and postpartum using the mouse model. By utilizing advanced magnetic resonance imaging (MRI) technology, T1-weighted and T2-weighted images of the reproductive organs of three mice in their in vivo environment were collected at five time points: non-pregnant, mid-pregnant (gestation day: 9-10), late pregnant (gestation day: 16-17), postpartum (24-72 h after delivery) and three weeks postpartum. Measurements of the vagina, cervix, and uterine horns were taken by analyzing MRI segmentations of these organs. The cross-sectional diameter, length, and volume of the vagina increased in late pregnancy and then returned to non-pregnant values three weeks after delivery. The cross-sectional diameter of the cervix decreased at mid-pregnancy before increasing in late pregnancy. The volume of the cervix peaked at late pregnancy before shortening by 24-72 h postpartum. As expected, the uterus increased in cross-sectional diameter, length, and volume during pregnancy. The uterine horns decreased in size postpartum, ultimately returning to their average non-pregnant size three weeks postpartum. The newly developed methods for acquiring longitudinal in vivo MRI scans of the murine reproductive system can be extended to future studies that evaluate functional and morphological alterations of this system due to pathologies, interventions, and treatments.

Mitigating reactive oxygen species production and increasing gel porosity improves lymphocyte motility and fibroblast spreading in photocrosslinked gelatin-thiol hydrogels.

On-chip 3D culture systems that incorporate immune cells such as lymphocytes and stromal cells are needed to model immune organs in engineered systems such as organs-on-chip. Photocrosslinking is a useful tool for creating such immune-competent hydrogel cultures with spatial cell organization. However, loss of viability and motility in photocrosslinked gels can limit its utility, especially when working with fragile primary cells. We hypothesized that optimizing photoexposure-induced ROS production, hydrogel porosity or a combination of both factors was necessary to sustain cell viability and motility during culture in photocrosslinked gelatin-thiol (GelSH) hydrogels. Jurkat T cells, primary human CD4+ T cells and human lymphatic fibroblasts were selected as representative lymphoid immune cells to test this hypothesis. Direct exposure of these cells to 385 nm light and LAP photoinitiator dramatically increased ROS levels. Pretreatment with an antioxidant, ascorbic acid (AA), protected the cells from light + LAP-induced ROS and was non-toxic at optimized doses. Furthermore, scanning electron microscopy showed that native GelSH hydrogels had limited porosity, and that adding collagen to GelSH precursor before crosslinking markedly increased gel porosity. Next, we tested the impact of AA pretreatment and increasing gel porosity, alone or in combination, on cell viability and function in 3D GelSH hydrogel cultures. Increasing gel porosity, rather than AA pretreatment, was more critical for rescuing viability of Jurkat T cells and spreading of human lymphatic fibroblasts in GelSH-based gels, but both factors improved the motility of primary human CD4+ T cells. Increased porosity enabled formation of spatially organized co-cultures of primary human CD4+ T cells and human lymphatic fibroblasts in photo-crosslinked gels in a multi-lane microfluidic chip, towards modeling the lymphoid organ microenvironment. Some optimization is still needed to improve homogeneity between regions on the chip. These findings will enable researchers utilizing photocrosslinking methods to develop immunocompetent 3D culture models that support viability and function of sensitive lymphoid cells.

Model discovery approach enables non-invasive measurement of intra-tumoral fluid transport in dynamic MRI.

Dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) is a routine method to non-invasively quantify perfusion dynamics in tissues. The standard practice for analyzing DCE-MRI data is to fit an ordinary differential equation to each voxel. Recent advances in data science provide an opportunity to move beyond existing methods to obtain more accurate measurements of fluid properties. Here, we developed a localized convolutional function regression that enables simultaneous measurement of interstitial fluid velocity, diffusion, and perfusion in 3D. We validated the method computationally and experimentally, demonstrating accurate measurement of fluid dynamics in situ and in vivo. Applying the method to human MRIs, we observed tissue-specific differences in fluid dynamics, with an increased fluid velocity in breast cancer as compared to brain cancer. Overall, our method represents an improved strategy for studying interstitial flows and interstitial transport in tumors and patients. We expect that it will contribute to the better understanding of cancer progression and therapeutic response.

Development of a Synthetic, Injectable Hydrogel to Capture Residual Glioblastoma and Glioblastoma Stem-Like Cells with CXCL12-Mediated Chemotaxis.

Glioblastoma (GBM), characterized by high infiltrative capacity, is the most common and deadly type of primary brain tumor in adults. GBM cells, including therapy-resistant glioblastoma stem-like cells (GSCs), invade the healthy brain parenchyma to form secondary tumors even after patients undergo surgical resection and chemoradiotherapy. New techniques are therefore urgently needed to eradicate these residual tumor cells. A thiol-Michael addition injectable hydrogel for compatibility with GBM therapy is previously characterized and optimized. This study aims to develop the hydrogel further to capture GBM/GSCs through CXCL12-mediated chemotaxis. The release kinetics of hydrogel payloads are investigated, migration and invasion assays in response to chemoattractants are performed, and the GBM-hydrogel interactions in vitro are studied. With a novel dual-layer hydrogel platform, it is demonstrated that CXCL12 released from the synthetic hydrogel can induce the migration of U251 GBM cells and GSCs from the extracellular matrix microenvironment and promote invasion into the synthetic hydrogel via amoeboid migration. The survival of GBM cells entrapped deep into the synthetic hydrogel is limited, while live cells near the surface reinforce the hydrogel through fibronectin deposition. This synthetic hydrogel, therefore, demonstrates a promising method to attract and capture migratory GBM cells and GSCs responsive to CXCL12 chemotaxis.

Modeling lymphangiogenesis: Pairing in vitro and in vivo metrics.

Lymphangiogenesis is the mechanism by which the lymphatic system develops and expands new vessels facilitating fluid drainage and immune cell trafficking. Models to study lymphangiogenesis are necessary for a better understanding of the underlying mechanisms and to identify or test new therapeutic agents that target lymphangiogenesis. Across the lymphatic literature, multiple models have been developed to study lymphangiogenesis in vitro and in vivo. In vitro, lymphangiogenesis can be modeled with varying complexity, from monolayers to hydrogels to explants, with common metrics for characterizing proliferation, migration, and sprouting of lymphatic endothelial cells (LECs) and vessels. In comparison, in vivo models of lymphangiogenesis often use genetically modified zebrafish and mice, with in situ mouse models in the ear, cornea, hind leg, and tail. In vivo metrics, such as activation of LECs, number of new lymphatic vessels, and sprouting, mirror those most used in vitro, with the addition of lymphatic vessel hyperplasia and drainage. The impacts of lymphangiogenesis vary by context of tissue and pathology. Therapeutic targeting of lymphangiogenesis can have paradoxical effects depending on the pathology including lymphedema, cancer, organ transplant, and inflammation. In this review, we describe and compare lymphangiogenic outcomes and metrics between in vitro and in vivo studies, specifically reviewing only those publications in which both testing formats are used. We find that in vitro studies correlate well with in vivo in wound healing and development, but not in the reproductive tract or the complex tumor microenvironment. Considerations for improving in vitro models are to increase complexity with perfusable microfluidic devices, co-cultures with tissue-specific support cells, the inclusion of fluid flow, and pairing in vitro models of differing complexities. We believe that these changes would strengthen the correlation between in vitro and in vivo outcomes, giving more insight into lymphangiogenesis in healthy and pathological states.

Fluids and flows in brain cancer and neurological disorders.

Interstitial fluid (IF) and cerebrospinal fluid (CSF) are an integral part of the brain, serving to cushion and protect the brain parenchymal cells against damage and aid in their function. The brain IF contains various ions, nutrients, waste products, peptides, hormones, and neurotransmitters. IF moves primarily by pressure-dependent bulk flow through brain parenchyma, draining into the ventricular CSF. The brain ventricles and subarachnoid spaces are filled with CSF which circulates through the perivascular spaces. It also flows into the IF space regulated, in part, by aquaporin channels, removing waste solutes through a process of IF-CSF mixing. During disease development, the composition, flow, and volume of these fluids changes and can lead to brain cell dysfunction. With the improvement of imaging technology and the help of genomic profiling, more information has been and can be obtained from brain fluids; however, the role of CSF and IF in brain cancer and neurobiological disease is still limited. Here we outline recent advances of our knowledge of brain fluid flow in cancer and neurodegenerative disease based on our understanding of its dynamics and composition. This article is categorized under: Cancer > Biomedical Engineering Neurological Diseases > Biomedical Engineering.

Structural and practical identifiability of contrast transport models for DCE-MRI.

Compartment models are widely used to quantify blood flow and transport in dynamic contrast-enhanced magnetic resonance imaging. These models analyze the time course of the contrast agent concentration, providing diagnostic and prognostic value for many biological systems. Thus, ensuring accuracy and repeatability of the model parameter estimation is a fundamental concern. In this work, we analyze the structural and practical identifiability of a class of nested compartment models pervasively used in analysis of MRI data. We combine artificial and real data to study the role of noise in model parameter estimation. We observe that although all the models are structurally identifiable, practical identifiability strongly depends on the data characteristics. We analyze the impact of increasing data noise on parameter identifiability and show how the latter can be recovered with increased data quality. To complete the analysis, we show that the results do not depend on specific tissue characteristics or the type of enhancement patterns of contrast agent signal.

Fusobacterium nucleatum induces proliferation and migration in pancreatic cancer cells through host autocrine and paracrine signaling.

The tumor microbiome is increasingly implicated in cancer progression and resistance to chemotherapy. In pancreatic ductal adenocarcinoma (PDAC), high intratumoral loads of Fusobacterium nucleatum correlate with shorter survival in patients. Here, we investigated the potential mechanisms underlying this association. We found that F. nucleatum infection induced both normal pancreatic epithelial cells and PDAC cells to secrete increased amounts of the cytokines GM-CSF, CXCL1, IL-8, and MIP-3α. These cytokines increased proliferation, migration, and invasive cell motility in both infected and noninfected PDAC cells but not in noncancerous pancreatic epithelial cells, suggesting autocrine and paracrine signaling to PDAC cells. This phenomenon occurred in response to Fusobacterium infection regardless of the strain and in the absence of immune and other stromal cells. Blocking GM-CSF signaling markedly limited proliferative gains after infection. Thus, F. nucleatum infection in the pancreas elicits cytokine secretion from both normal and cancerous cells that promotes phenotypes in PDAC cells associated with tumor progression. The findings support the importance of exploring host-microbe interactions in pancreatic cancer to guide future therapeutic interventions.

Myeloid Derived Suppressor Cells Migrate in Response to Flow and Lymphatic Endothelial Cell Interaction in the Breast Tumor Microenvironment.

At the site of the tumor, myeloid derived suppressor cells (MDSCs) infiltrate and interact with elements of the tumor microenvironment in complex ways. Within the invading tumor, MDSCs are exposed to interstitial fluid flow (IFF) that exists within the chronic inflammatory tumor microenvironment at the tumor-lymphatic interface. As drivers of cell migration and invasion, the link between interstitial fluid flow, lymphatics, and MDSCs have not been clearly established. Here, we hypothesized that interstitial fluid flow and cells within the breast tumor microenvironment modulate migration of MDSCs. We developed a novel 3D model to mimic the breast tumor microenvironment and incorporated MDSCs harvested from 4T1-tumor bearing mice. Using live imaging, we found that sorted GR1+ splenocytes had reduced chemotactic index compared to the unsorted population, but their speed and displacement were similar. Using our adapted tissue culture insert assay, we show that interstitial fluid flow promotes MDSC invasion, regardless of absence or presence of tumor cells. Coordinating with lymphatic endothelial cells, interstitial fluid flow further enhanced invasion of MDSCs in the presence of 4T1 cells. We also show that VEGFR3 inhibition reduced both MDSC and 4T1 flow response. Together, these findings indicate a key role of interstitial fluid flow in MDSC migration as well as describe a tool to explore the immune microenvironment in breast cancer.

Towards spatially-organized organs-on-chip: Photopatterning cell-laden thiol-ene and methacryloyl hydrogels in a microfluidic device.

Micropatterning techniques for 3D cell cultures enable the recreation of tissue-level structures, but the combination of patterned hydrogels with organs-on-chip to generate organized 3D cultures under microfluidic perfusion remains challenging. To address this technological gap, we developed a user-friendly in-situ micropatterning protocol that integrates photolithography of crosslinkable, cell-laden hydrogels with a simple microfluidic housing, and tested the impact of crosslinking chemistry on stability and spatial resolution. Working with gelatin functionalized with photo-crosslinkable moieties, we found that inclusion of cells at high densities (≥ 107/mL) did not impede thiol-norbornene gelation, but decreased the storage moduli of methacryloyl hydrogels. Hydrogel composition and light dose were selected to match the storage moduli of soft tissues. To generate the desired pattern on-chip, the cell-laden precursor solution was flowed into a microfluidic chamber and exposed to 405 nm light through a photomask. The on-chip 3D cultures were self-standing and the designs were interchangeable by simply swapping out the photomask. Thiol-ene hydrogels yielded highly accurate feature sizes from 100 - 900 µm in diameter, whereas methacryloyl hydrogels yielded slightly enlarged features. Furthermore, only thiol-ene hydrogels were mechanically stable under perfusion overnight. Repeated patterning readily generated multi-region cultures, either separately or adjacent, including non-linear boundaries that are challenging to obtain on-chip. As a proof-of-principle, primary human T cells were patterned on-chip with high regional specificity. Viability remained high (> 85%) after 12-hr culture with constant perfusion. We envision that this technology will enable researchers to pattern 3D co-cultures to mimic organ-like structures that were previously difficult to obtain.

Platinum Chemotherapy Induces Lymphangiogenesis in Cancerous and Healthy Tissues That Can be Prevented With Adjuvant Anti-VEGFR3 Therapy.

Chemotherapy has been used to inhibit cancer growth for decades, but emerging evidence shows it can affect the tumor stroma, unintentionally promoting cancer malignancy. After treatment of primary tumors, remaining drugs drain via lymphatics. Though all drugs interact with the lymphatics, we know little of their impact on them. Here, we show a previously unknown effect of platinums, a widely used class of chemotherapeutics, to directly induce systemic lymphangiogenesis and activation. These changes are dose-dependent, long-lasting, and occur in healthy and cancerous tissue in multiple mouse models of breast cancer. We found similar effects in human ovarian and breast cancer patients whose treatment regimens included platinums. Carboplatin treatment of healthy mice prior to mammary tumor inoculation increased cancer metastasis as compared to no pre-treatment. These platinum-induced phenomena could be blocked by VEGFR3 inhibition. These findings have implications for cancer patients receiving platinums and may support the inclusion of anti-VEGFR3 therapy into treatment regimens or differential design of treatment regimens to alter these potential effects.

Designing Patient-Driven, Tissue-Engineered Models of Primary and Metastatic Breast Cancer.

The rising survival rate for early-stage breast cancer in the United States has created an expanding population of women in remission at risk for distant recurrence, with metastatic spread to the brain demonstrating an especially poor prognosis. The current standard of care for breast cancer brain metastases is not well defined or differentiated from the treatment of brain metastases from other primary sites. Here, we present tissue-engineered models of the primary and brain metastatic breast cancer microenvironments informed by analysis of patient tumor resections. We find that metastatic resections demonstrate distinct cellular and matrix components compared with primary resections or non-cancerous controls. Using our model systems, we find that the observed deposition of collagen I after metastasis to the brain may enhance breast cancer invasion. Future optimization of these models will present a novel platform to examine tumor-stroma interactions and screen therapeutics for the management of metastatic breast cancer.

Engineering in vitro immune-competent tissue models for testing and evaluation of therapeutics.

Advances in 3D cell culture, microscale fluidic control, and cellular analysis have enabled the development of more physiologically-relevant engineered models of human organs with precise control of the cellular microenvironment. Engineered models have been used successfully to answer fundamental biological questions and to screen therapeutics, but these often neglect key elements of the immune system. There are immune elements in every tissue that contribute to healthy and diseased states. Including immune function will be essential for effective preclinical testing of therapeutics for inflammatory and immune-modulated diseases. In this review, we first discuss the key components to consider in designing engineered immune-competent models in terms of physical, chemical, and biological cues. Next, we review recent applications of models of immunity for screening therapeutics for cancer, preclinical evaluation of engineered T cells, modeling autoimmunity, and screening vaccine efficacy. Future work is needed to further recapitulate immune responses in engineered models for the most informative therapeutic screening and evaluation.

Delivery strategies for cell-based therapies in the brain: overcoming multiple barriers.

Cell-based therapies to the brain are promising for the treatment of multiple brain disorders including neurodegeneration and cancers. In order to access the brain parenchyma, there are multiple physiological barriers that must be overcome depending on the route of delivery. Specifically, the blood-brain barrier has been a major difficulty in drug delivery for decades, and it still presents a challenge for the delivery of therapeutic cells. Other barriers, including the blood-cerebrospinal fluid barrier and lymphatic-brain barrier, are less explored, but may offer specific challenges or opportunities for therapeutic delivery. Here we discuss the barriers to the brain and the strategies currently in place to deliver cell-based therapies, including engineered T cells, dendritic cells, and stem cells, to treat diseases. With a particular focus on cancers, we also highlight the current ongoing clinical trials that use cell-based therapies to treat disease, many of which show promise at treating some of the deadliest illnesses.

Modeling Immunity In Vitro: Slices, Chips, and Engineered Tissues.

Modeling immunity in vitro has the potential to be a powerful tool for investigating fundamental biological questions, informing therapeutics and vaccines, and providing new insight into disease progression. There are two major elements to immunity that are necessary to model: primary immune tissues and peripheral tissues with immune components. Here, we systematically review progress made along three strategies to modeling immunity: ex vivo cultures, which preserve native tissue structure; microfluidic devices, which constitute a versatile approach to providing physiologically relevant fluid flow and environmental control; and engineered tissues, which provide precise control of the 3D microenvironment and biophysical cues. While many models focus on disease modeling, more primary immune tissue models are necessary to advance the field. Moving forward, we anticipate that the expansion of patient-specific models may inform why immunity varies from patient to patient and allow for the rapid comprehension and treatment of emerging diseases, such as coronavirus disease 2019.

Tear propagation in vaginal tissue under inflation.

Vaginal tearing at childbirth is extremely common yet understudied despite the long-term serious consequences on women's health. The mechanisms of vaginal tearing remain unknown, and their knowledge could lead to the development of transformative prevention and treatment techniques for maternal injury. In this study, whole rat vaginas with pre-imposed elliptical tears oriented along the axial direction of the organs were pressurized using a custom-built inflation setup, producing large tear propagation. Large deformations of tears through propagation were analyzed, and nonlinear strains around tears were calculated using the digital image correlation technique. Second harmonic generation microscopy was used to examine collagen fiber organization in mechanically untested and tested vaginal specimens. Tears became increasingly circular under pressure, propagating slowly up to the maximum pressure and then more rapidly. Hoop strains were significantly larger than axial strains and displayed a region- and orientation-dependent response with tear propagation. Imaging revealed initially disorganized collagen fibers that aligned along the axial direction with increasing pressure. Fibers in the near-regions of tear tips aligned toward the hoop direction, hampering tear propagation. Changes in tear geometry, regional strains, and fiber orientation revealed the inherent toughening mechanisms of the vaginal tissue. STATEMENT OF SIGNIFICANCE: Women's reproductive health has historically been understudied despite alarming maternal injury and mortality rates in the world. Maternal injury and disability can be reduced by advancing our limited understanding of the large deformations experienced by women's reproductive organs. This manuscript presents, for the first time, the mechanics of tear propagation in vaginal tissue and changes to the underlying collagen microstructure near to and far from the tear. A novel inflation setup capable of maintaining the in vivo tubular geometry of the vagina while propagating a pre-imposed tear was developed. Toughening mechanisms of the vagina to propagation were examined through measurements of tear geometry, strain distributions, and reorientation of collagen fibers. This research draws from current advances in the engineering science and mechanics fields with the goal of improving maternal health care.

Utilizing Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) to Analyze Interstitial Fluid Flow and Transport in Glioblastoma and the Surrounding Parenchyma in Human Patients.

BACKGROUND: Glioblastoma (GBM) is the deadliest and most common brain tumor in adults, with poor survival and response to aggressive therapy. Limited access of drugs to tumor cells is one reason for such grim clinical outcomes. A driving force for therapeutic delivery is interstitial fluid flow (IFF), both within the tumor and in the surrounding brain parenchyma. However, convective and diffusive transport mechanisms are understudied. In this study, we examined the application of a novel image analysis method to measure fluid flow and diffusion in GBM patients. METHODS: Here, we applied an imaging methodology that had been previously tested and validated in vitro, in silico, and in preclinical models of disease to archival patient data from the Ivy Glioblastoma Atlas Project (GAP) dataset. The analysis required the use of dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI), which is readily available in the database. The analysis results, which consisted of IFF flow velocity and diffusion coefficients, were then compared to patient outcomes such as survival. RESULTS: We characterized IFF and diffusion patterns in patients. We found strong correlations between flow rates measured within tumors and in the surrounding parenchymal space, where we hypothesized that velocities would be higher. Analyzing overall magnitudes indicated a significant correlation with both age and survival in this patient cohort. Additionally, we found that neither tumor size nor resection significantly altered the velocity magnitude. Lastly, we mapped the flow pathways in patient tumors and found a variability in the degree of directionality that we hypothesize may lead to information concerning treatment, invasive spread, and progression in future studies. CONCLUSIONS: An analysis of standard DCE-MRI in patients with GBM offers more information regarding IFF and transport within and around the tumor, shows that IFF is still detected post-resection, and indicates that velocity magnitudes correlate with patient prognosis.

Repeatability of tumor perfusion kinetics from dynamic contrast-enhanced MRI in glioblastoma.

BACKGROUND: Dynamic contrast-enhanced MRI (DCE-MRI) parameters have been shown to be biomarkers for treatment response in glioblastoma (GBM). However, variations in analysis and measurement methodology complicate determination of biological changes measured via DCE. The aim of this study is to quantify DCE-MRI variations attributable to analysis methodology and image quality in GBM patients. METHODS: The Extended Tofts model (eTM) and Leaky Tracer Kinetic Model (LTKM), with manually and automatically segmented vascular input functions (VIFs), were used to calculate perfusion kinetic parameters from 29 GBM patients with double-baseline DCE-MRI data. DCE-MRI images were acquired 2-5 days apart with no change in treatment. Repeatability of kinetic parameters was quantified with Bland-Altman and percent repeatability coefficient (%RC) analysis. RESULTS: The perfusion parameter with the least RC was the plasma volume fraction (v p ), with a %RC of 53%. The extra-cellular extra-vascular volume fraction (v e ) %RC was 82% and 81%, for extended Tofts-Kety Model (eTM) and LTKM respectively. The %RC of the volume transfer rate constant (K trans ) was 72% for the eTM, and 82% for the LTKM, respectively. Using an automatic VIF resulted in smaller %RCs for all model parameters, as compared to manual VIF. CONCLUSIONS: As much as 72% change in K trans (eTM, autoVIF) can be attributable to non-biological changes in the 2-5 days between double-baseline imaging. Poor K trans repeatability may result from inferior temporal resolution and short image acquisition time. This variation suggests DCE-MRI repeatability studies should be performed institutionally, using an automatic VIF method and following quantitative imaging biomarkers alliance guidelines.