Magnetic resonance imaging of brain angiogenesis after stroke
- First Online:
- Cite this article as:
- Seevinck, P.R., Deddens, L.H. & Dijkhuizen, R.M. Angiogenesis (2010) 13: 101. doi:10.1007/s10456-010-9174-0
Stroke is a major cause of mortality and long-term disability worldwide. The initial changes in local perfusion and tissue status underlying loss of brain function are increasingly investigated with noninvasive imaging methods. In addition, there is a growing interest in imaging of processes that contribute to post-stroke recovery. In this review, we discuss the application of magnetic resonance imaging (MRI) to assess the formation of new vessels by angiogenesis, which is hypothesized to participate in brain plasticity and functional recovery after stroke. The excellent soft tissue contrast, high spatial and temporal resolution, and versatility render MRI particularly suitable to monitor the dynamic processes involved in vascular remodeling after stroke. Here we review recent advances in the field of MR imaging that are aimed at assessment of tissue perfusion and microvascular characteristics, including cerebral blood flow and volume, vascular density, size and integrity. The potential of MRI to noninvasively monitor the evolution of post-ischemic angiogenic processes is demonstrated from a variety of in vivo studies in experimental stroke models. Finally, we discuss some pitfalls and limitations that may critically affect the accuracy and interpretation of MRI-based measures of (neo)vascularization after stroke.
KeywordsAngiogenesisMagnetic resonance imagingStrokeBrain ischemiaMicrovessel densityMicrovessel sizeCerebral blood volumeCerebral blood flowBlood–brain barrier leakageMRI
Stroke is the result of occlusion or rupture of a brain artery, leading to loss of focal blood flow and brain function. Despite significant neurological deficits, most patients experience at least some degree of spontaneous recovery , which may be augmented by therapeutic intervention . Currently, neuroprotective therapy after acute clinical stroke is restricted to the use of a single therapeutic agent, recombinant tissue plasminogen activator (rtPA), which should be given within 4.5 h after stroke to limit risk of hemorrhagic transformation . However, experimental studies suggest that restorative pharmacological and cell-based therapies have potential to improve functional outcome when initiated 24 h up to weeks after stroke . Such therapies amplify certain endogenous processes, some of which are also activated in the developing brain , that may contribute to spontaneous recovery after stroke. These events not only involve neuronal reorganization, but also include vascular remodeling through angiogenesis. Interestingly, there is an increasing amount of studies that has provided evidence for formation of new blood vessels after cerebral ischemia, and a possible significant role of angiogenesis in post-stroke recovery [5–10]. However, the exact pattern of neovascularization and its relation to restoration of function after stroke are still largely unresolved. Therefore, thorough characterization of the spatial and temporal profile of angiogenesis after ischemic stroke is of significant importance to elucidate its role in recovery and remodeling of neuronal tissue and ensuing functional outcome, which may lead to new insights for further development of neurorestorative therapies.
Magnetic resonance imaging (MRI) provides a powerful tool to noninvasively assess the evolution of various indices that can characterize cerebral tissue status after stroke [11–14]. Over the last decade, different MRI methods have also been applied to measure angiogenesis-related processes. This review describes results from such studies in experimental stroke models and discusses the principles, potentials and pitfalls of MRI methods to assess brain angiogenesis after stroke. First, the physiological process of angiogenesis after stroke will be briefly introduced. The main emphasis will be on the methodology and application of different MR imaging techniques, with and without contrast enhancement, that enable noninvasive monitoring of developing and mature microvasculature for stroke research.
Angiogenesis after stroke
Importantly, formation of new vessels after stroke may (1) contribute to recovery of tissue-at-risk by restoring metabolism in surviving neurons, (2) facilitate removal of necrotic debris, and/or (3) enhance supply of neurotrophic compounds for neuronal remodeling (e.g. synaptogenesis and neurite sprouting) [5, 6, 8–10, 19]. However, whether angiogenesis indeed gives rise to full-fledged functional vascular networks around a stroke lesion is still unclear and remains an important topic for further research.
MRI-based assessment of brain angiogenesis after stroke
In the last two decades, MRI has proven to be a valuable tool to investigate the spatiotemporal profile of ischemia-induced changes after stroke, mainly attributable to its capability to longitudinally evaluate a wide spectrum of structural and functional tissue characteristics. This versatility originates from the fact that contrast in MR images is dependent on intrinsic, biophysical tissue properties such as proton density, inter- and intramolecular magnetic interactions, oxygenation state, magnetic susceptibility, diffusion, perfusion and flow. These endogenous tissue characteristics influence the MRI signal by their effect on MR relaxation times (T1, T2 and T2*) and water proton mobility, which can be exploited to generate image contrast. Additionally, exogenous contrast agents (e.g. gadolinium chelates and iron oxide particles) can be applied to enhance endogenous contrast-generating mechanisms. This can improve anatomical and/or physiological distinction, and may allow detection of otherwise indiscernible factors. In stroke research, both endogenous and exogenous MR contrast mechanisms have been exploited to assess cerebrovascular changes associated with angiogenesis. In the following sections the various MR imaging techniques that have been utilized to assess angiogenesis in the brain will be reviewed.
MRI of vessels, blood volume and perfusion
Probably the most intuitive way to evaluate the presence of angiogenic vessels is direct visualization of the newly formed vasculature by MR angiography . However, the spatial resolution of MRI is currently limited to approximately 50 μm for small animal MR imaging, and to approximately 250 μm on clinical MRI scanners. Since angiogenic processes in brain tissue, such as endothelial sprouting and microvessel formation, take place at length scales that are at least one order of magnitude lower, direct depiction of remodeling of cerebrovascular structures is currently not feasible with MRI. Alternatively, hemodynamic parameters that are directly affected by angiogenesis, such as cerebral blood volume (CBV) and cerebral blood flow (CBF), can be measured with perfusion MR imaging techniques such as dynamic susceptibility contrast-enhanced (DSC-)MRI, steady state susceptibility contrast-enhanced (ssCE-)MRI and arterial spin labeling (ASL). DSC- and ssCE-MRI are based on detection of signal changes induced by intravascularly injected exogenous contrast agents, while ASL exploits endogeneous contrast mechanisms involving magnetically labeled arterial water. The following paragraphs provide a brief introduction to these MRI methods. For more detailed information on the principles of the different perfusion MRI techniques, we refer the reader to specific reviews [21–23].
Dynamic susceptibility contrast-enhanced MRI
DSC-MRI allows calculation of cerebral hemodynamic parameters from the time-course of signal changes induced by the first passage of a paramagnetic contrast agent after intravenous injection . The most commonly applied exogenous contrast agents are gadolinium chelates . Their high magnetic susceptibility induces local magnetic field inhomogeneities. The consequent local shortening of T2 and T2* leads to enhanced MR signal decay. By dynamically monitoring these signal changes in perfused tissue during the first passage of a bolus of injected contrast agent, hemodynamic parameters such as CBF, CBV and mean transit time (MTT) can be estimated from the linear relationship between contrast agent concentration and change in effective transverse relaxation rate (ΔR2* = Δ(1/T2*), and the central volume theory [24, 25].
Arterial spin labeling
As mentioned above, CBF can also be determined with MRI without the use of exogenous contrast agents. In ASL, radiofrequency (RF) pulses are used to magnetically alter the status of water protons in arterial blood with respect to those in stationary tissue water, thereby generating an endogenous intravascular tracer [28, 29]. Assuming that the magnetically labeled arterial blood exchanges with tissue water at the level of the capillaries, CBF can in principle be quantified from a labeling experiment and a control experiment without magnetically labeled water protons, based on the theory of diffusible tracer kinetics [28, 30, 31].
ASL has been applied in a number of experimental stroke studies to estimate CBF in relation to angiogenesis. In conjunction with an increased vascular density, which is clearly visible on the surface of the ipsilateral brain surface in Fig. 3a, Lin et al. demonstrated a significantly enhanced CBF in the perilesional cortex from day 1 to 14 after transient MCA occlusion in rats (Fig. 3b, c) . Chopp and co-workers have reported that treatment with the vasodilatator sildenafil enhanced angiogenesis and selectively increased the CBF level in the ischemic boundary in rats after embolic stroke [17, 32]. Furthermore, co-localization of the area with CBF improvement with regions with increased fractional anisotropy of white matter tissue was reported, pointing towards a link between angiogenesis and neuronal reorganization [32, 33].
Despite the significant information on tissue perfusion, MRI-based measurement of hemodynamics provides relatively low specificity for angiogenesis. For example, CBV and CBF increases may also arise in response to autoregulatory vasodilation or vessel recruitment due to arteriogenesis. Therefore, alternative MR methodologies that allow assessment of structural features of the microvasculature that may be more specifically related to angiogenesis have been recently proposed, which will be the subject of the following paragraph.
Steady state susceptibility contrast-enhanced MRI
Wu et al. have assessed the MVD in normal mouse brain based on the ratio Q, using a literature value for D = 0.664 μm2/ms . This provided an average brain MVD of 282 ± 43/mm2, which is in reasonable agreement with histologically determined values.
Recently, two studies have employed ssCE-MRI to assess vascular remodeling after experimental stroke [51, 52]. In a serial study in rats recovering from transient cerebral ischemia performed by Lin et al., an initial decrease in vascular density (based on the Q value) and an increase in vessel size (based on VSI) was observed in the reperfused cortex at day 1 and 3 . Immunohistological analysis confirmed a similar decrease in microvessel density and increase in size of vessels with a diameter larger than 30 μm. These observations were explained by a more pronounced effect of edema on compression of small capillaries as compared to large-sized vessels, leading to a shift in the calculated average vessel size. A significant increase of total CBV from day 3 to 14 in the affected hemisphere, based on ΔR2*, was speculated to be caused by improvement of collateral circulation in the relatively large microvessels. At days 14 and 21, increases in ΔR2 (microvascular CBV) and Q (microvascular density) were noticed in the reperfused cortex, which was attributed to the surge of angiogenesis. In contrast, Bosomtwi et al. observed a lowered Q-based and histology-based MVD in recovered ischemic tissue at 2 weeks after embolic stroke in rats . A possible explanation is that only a small part of the recovered region was highly angiogenic, which may have been obscured by the analysis of a relatively large region-of-interest .
MRI of blood–brain barrier permeability and tissue oxygenation
Next to assessment of hemodynamic and morphologic changes related to vascular remodeling, MRI can provide information on BBB permeability and oxygenation state of brain tissue, which both are involved in the process of angiogenesis. BBB permeability can be investigated by dynamic contrast-enhanced MRI (DCE-MRI), in which changes in MR signal, as a result of leakage of an intravascularly injected contrast agent into the interstitial space, are dynamically monitored . The oxygenation state of brain tissue can be investigated by exploiting the high magnetic susceptibility of deoxygenated hemoglobin . Since venous blood is rich of deoxygenated hemoglobin, blood oxygenation level-dependent (BOLD) MRI may detect newly developed vasculature that is rich of venous blood . Both methods will be briefly introduced in relation to their potential to evaluate angiogenesis after stroke.
Dynamic contrast-enhanced MRI
Blood oxygenation level-dependent MRI
MRI techniques, parameters, and contrast types and mechanisms to assess tissue characteristics related to vascular remodeling through angiogenesis
Δχ, contrast passage
Magnetically labeled arterial water
Δχ, pre- and post-contrast
T1, pre- and post-contrast
Discussion and future perspectives
In the search for neuroprotective and neurorestorative therapies aiming to improve functional outcome after stroke, a better understanding of the endogenous processes involved in spontaneous recovery after stroke is of major importance. A healthy tissue perfusion status, responsible for providing nutrients and oxygen and allowing clean-up of waste products, is critical for mediating functional recovery after brain injury. This process could be significantly augmented by vascular remodeling through angiogenesis. Therefore, the prospect to noninvasively assess physiological and structural characteristics of such events would improve our knowledge on brain plasticity, and may eventually help us to direct functional outcome of stroke to our benefit. This review illustrates how MRI can be applied to noninvasively monitor the evolution of various post-ischemic processes related to angiogenesis in the brain. The presented studies demonstrate that MR imaging enables assessment of a variety of (micro)vascular characteristics, such as cerebral blood flow and volume, vascular density and vessel size, BBB integrity and blood oxygenation.
Despite its potential, we need to also be aware of several pitfalls and limitations related to the application of MRI as a tool for evaluating angiogenesis after stroke. For example, an increase in Ki, i.e. contrast agent leakage, and shortening of T2*, i.e. more deoxygenated blood, which has been found in relation to angiogenesis , may also arise as a result of BBB disruption and hemorrhage in response to vascular pathology . Furthermore, MRI measurements of tissue perfusion and microvascular characteristics (CBF, CBV, Ki, MVD and VSI) rely on algorithms that are based on relatively basic biophysical and mathematical models that may be inaccurate under complex or altered conditions, as is the case in stroke pathophysiology. The methodological difficulties related to the models used in DSC-MRI and ASL to quantify perfusion, especially under abnormal flow conditions, have been extensively reviewed [21, 22, 68–70]. Furthermore, from a physiological point of view, an increase in CBF and CBV, which has been associated with post-stroke angiogenesis, may not be exclusively specific to neovascularization, as vasodilatation of existing vessels and collateral flow may also contribute. Hence, the combination of perfusion imaging with additional MR methods that provide information on microvessel density and vessel size may help elucidating the complex interdependence between CBF, CBV, MVD and VSI. This can be accomplished with ssCE-MRI, as discussed in this review. Theoretical modeling as well as Monte Carlo simulations have demonstrated the ability to quantify MVD and VSI with ssCE-MRI [34, 42, 44, 48, 49], however, in practice MVD and VSI are typically qualitatively expressed [47, 52]. Quantification of microvessel density and size based on in vivo MRI measurements have been reported based on assumptions of the diffusion coefficient D and susceptibility difference Δχ [50, 51]. Since D may considerably change under specific pathological conditions, and post-contrast Δχ will vary between subjects, it can be expected that defining D and Δχ based on literature values will affect the accuracy of quantitative MVD and VSI measurements. In addition, large variations of Q values, as well as ΔR2*/ΔR2 ratios, have been reported for rat brain [44, 48–52], which may be explained by additional biophysical assumptions and generalizations. First, the vascular network is assumed to consist of randomly oriented straight cylinders, ignoring the influence of curvature. Second, the vessel diameter within tissue-of-interest is often approximated at a single diameter. Third, isotropic diffusion is assumed. Fourth, computations rely on an intact BBB. Obviously, variations and alterations in these factors can lead to under- or overestimation of MVD and VSI calculations. Elucidation of the influence of these aspects in ssCE-MRI is critical for adequate use of this promising approach in assessing vascular characteristics in healthy and pathological brain.
A promising new MRI approach that may enable direct detection of angiogenesis is molecular imaging with contrast agents that are targeted to specific molecular markers of vascular remodeling. In oncology, MR-based molecular imaging with ανβ3-integrin-targeted paramagnetic nanoparticles has already demonstrated its potential to monitor tumor angiogenesis [71–73]. A clear example of the potential of in vivo molecular imaging of angiogenesis after stroke has recently been presented by Cai et al., who used positron emission tomography (PET) to measure upregulation of VEGF receptors with radioactively-labeled VEGF in the ischemic rat brain. For noninvasive monitoring of kinetics of expression of molecular markers, nuclear imaging techniques are valuable, however, their relatively low resolution, requirement of a radioactive tracer and lack of anatomical reference, are significant drawbacks that may be overcome by development of MR-based molecular imaging. A potential innovative way to monitor the onset of vascular remodeling in a very early stage with MRI, may be accomplished with MR reporter genes, which allow noninvasive assessment of expression of a gene of interest together with synthesis of an MR detectable by-product, such as endogenous paramagnetic ferritin [74–76].
In conclusion, in this review we have presented recent developments in the field of MR imaging that enable in vivo assessment of different physiological parameters, e.g. tissue perfusion and microvascular morphology, that can shed light on various aspects related to angiogenesis after stroke. These developments may provide important new insights into the role of vascular remodeling in brain plasticity and functional recovery after stroke, and may be valuable for monitoring of possible future therapies designed to promote (neo)vascularization in stroke patients.
The research leading to these results has received funding from the Netherlands Organization for Scientific Research (NWO) (VIDI 917.76.347), the Netherlands Ministries of Economic Affairs and Education, Culture and Science through the NL Agency project TARGET MR-IT, and the European Union’s Seventh Framework Programme (FP7/2007-2013) under grant agreements no 201024 and no 202213 (European Stroke Network). Annette van der Toorn, Ivo Tiebosch, Bas Oude Munnink and Elga de Vries are gratefully acknowledged for technical support and advice, as well as Laboratoire Guerbet (Aulnay-sous-Bois, France) for kindly providing the USPIO contrast agent P904.
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.