Impaired Glymphatic Perfusion After Strokes Revealed by Contrast-Enhanced MRI
A New Target for Fibrinolysis?
Background and Purpose—The aim of the present study was to investigate the impact of different stroke subtypes on the glymphatic system using MRI.
Methods—We first improved and characterized an in vivo protocol to measure the perfusion of the glymphatic system using MRI after minimally invasive injection of a gadolinium chelate within the cisterna magna. Then, the integrity of the glymphatic system was evaluated in 4 stroke models in mice including subarachnoid hemorrhage (SAH), intracerebral hemorrhage, carotid ligature, and embolic ischemic stroke.
Results—We were able to reliably evaluate the glymphatic system function using MRI. Moreover, we provided evidence that the glymphatic system was severely impaired after SAH and in the acute phase of ischemic stroke, but was not altered after carotid ligature or in case of intracerebral hemorrhage. Notably, this alteration in glymphatic perfusion reduced brain clearance rate of low-molecular-weight compounds. Interestingly, glymphatic perfusion after SAH can be improved by intracerebroventricular injection of tissue-type plasminogen activator. Moreover, spontaneous arterial recanalization was associated with restoration of the glymphatic function after embolic ischemic stroke.
Conclusions—SAH and acute ischemic stroke significantly impair the glymphatic system perfusion. In these contexts, injection of tissue-type plasminogen activator either intracerebroventricularly to clear perivascular spaces (for SAH) or intravenously to restore arterial patency (for ischemic stroke) may improve glymphatic function.
The glymphatic system allows clearance of the brain interstitial fluid through para-arterial influx and paravenular efflux of cerebrospinal fluid (CSF).1 Because the rate of metabolic byproduct generation increases dramatically after neuronal injury, waste clearance through the glymphatic system may play a central role after stroke. Here, we hypothesized that stroke affects the glymphatic system function. To test this hypothesis, we developed a minimally invasive injection method of gadolinium chelate in the cisterna magna to follow spontaneous CSF flows using MRI in mice. Subsequently, we studied the function of the glymphatic system after subarachnoid hemorrhage (SAH), intracerebral hemorrhage, common carotid artery occlusion, and embolic ischemic stroke.
For the complete details of the methods used, see the online-only Data Supplement.
The function of the glymphatic system was evaluated by T1-weighted MRI after intracisternal DOTA-Gd (Dotarem, Guerbet, France) microinjection (1 μL for 1 minute). We modified a previously reported method2 by using a pulled hematological glass micropipette (Figure I in the online-only Data Supplement). Regions of interest for subsequent signal analyses were selected as shown in Figure II in the online-only Data Supplement. Common carotid artery occlusion,3 SAH,4 and collagenase-induced intracerebral hemorrhage5 were performed as previously described. Embolic ischemic stroke was performed as described in the online-only Data Supplement. For immunohistological studies, sodium fluorescein (5 μg per mouse; Sigma-Aldrich, France) or FITC-Dextran (5 μg per mouse; Sigma-Aldrich) were injected in the cisterna magna.
MRI Allows In Vivo Assessment of the Glymphatic System Function
We first investigated the efficiency of a modified contrast-enhanced MRI protocol to image intraparenchymal CSF flow in vivo. Five minutes after intracisternal DOTA-Gd injection, the signal of the CSF cisterns became enhanced on T1-weighted images (Figure 1A). Sequential MRI revealed that the gadolinium chelate progressively entered the brain. The pathways of CSF distributions were consistent with previous reports2 (Figure 1B). At 60 minutes postinjection, the contrast material reached all the brain regions (Figure 1C). Thus, our modified contrast-enhanced MRI protocol allowed imaging of the brain glymphatic system in mice. Interestingly, other routes of CSF drainage were also revealed by this protocol, including the ethmoid sinus, peritracheal lymph nodes, optic nerves, and the cochleovestibular region (Figure 1D).
Glymphatic System Is Impaired After SAH
Then, we looked for potential effects of SAH on the glymphatic system. We induced SAH by injection of 50 μL fresh arterial blood in the prechiasmatic cistern of mice.4 Interestingly, 24 hours after SAH induction, the glymphatic system appeared severely impaired as compared with sham animals. Indeed, 60 minutes after intracisternal microinjection of DOTA-Gd, enhancement was present in the cerebellum but spared the forebrain (Figure 2A), suggesting occlusion of para-arterial influx routes. We confirmed these results by immunohistology after injection of sodium fluorescein in the cisterna magna (Figure 2B). Importantly, this glymphatic blockade persisted after bilateral craniectomy (Figure III in the online-only Data Supplement), suggesting that it is independent of SAH-induced intracranial hypertension.
Thereafter, we studied the permeability of the paravascular CSF circulation pathways 24 hours after SAH. To this aim, we injected a high-molecular-weight fluorescent dye (FITC-Dextran) in the cisterna magna. We first confirmed by histological analyses that this dye accumulates in the perivascular space after intracisternal injection (Figure IV in the online-only Data Supplement). Interestingly, in SAH animals, intracisternally injected FITC-Dextran failed to reach the perivascular spaces in the olfactory bulbs, although it accumulated in the cerebellar perivascular spaces (Figure VA in the online-only Data Supplement). Moreover, immunohistological studies revealed the presence of fibrin/fibrinogen in the perivascular spaces after SAH (Figure VB in the online-only Data Supplement). Altogether, these results suggested that perivascular spaces are occluded by fibrin clots after SAH.
To test this hypothesis, we performed intraventricular injection of a fibrinolytic agent (tissue-type plasminogen activator) 15 minutes after hemorrhage onset and evaluated the glymphatic perfusion at 24 hours. According to our hypothesis, intraventricular injection of tissue-type plasminogen activator improved glymphatic perfusion in SAH mice as compared with control saline-treated animals (Figure 2C).
Finally, to investigate the effect of glymphatic blockade on the brain clearance rate of small-molecular-weight compounds, we measured the clearance of DOTA-Gd from sham and SAH mice by serial T1-weighted MRI. As shown in Figure VI in the online-only Data Supplement, DOTA-Gd clearance from the striatum of SAH animals was slower than in control animals. This result suggested that glymphatic blockade by fibrin after SAH reduces waste clearance from the brain parenchyma. Interestingly, the clearance rate was diminished in SAH animals but was not completely abolished, suggesting the existence of other mechanisms driving DOTA-Gd clearance.
Ischemic Stroke Decreased Paravascular CSF Circulation
To study the impact of ischemic stroke on the glymphatic function, we induced embolic ischemic stroke in mice and studied the glymphatic system function by MRI and histofluorescence 3 and 24 hours after right middle cerebral artery occlusion. Interestingly, we noticed a blockade of the right glymphatic perfusion 3 hours after middle cerebral artery occlusion, as revealed by MRI (Figure 3A and 3B) and histology (Figure 3C and 3D). Again, glymphatic perfusion at 3 hours poststroke remained impaired after bilateral craniectomy (Figure VII in the online-only Data Supplement). At 24 hours after stroke onset, the glymphatic system function appeared normal. Notably, angiographies performed at the time of glymphatic imaging revealed that the middle cerebral artery were repermeabilized in all mice 24 hours after stroke onset (not shown). Thus, these results suggested that glymphatic perfusion depends on arterial patency after embolic ischemic stroke.
Glymphatic Imaging in Other Experimental Models of Cerebrovascular Diseases
In another set of experiments, we performed glymphatic imaging 24 hours after common carotid artery occlusion or intracerebral hemorrhage. Neither common carotid artery occlusion nor intracerebral hemorrhage influenced paravascular circulation of the CSF (Figure VIII in the online-only Data Supplement).
The current study demonstrates that the glymphatic system function can be evaluated in living mice using MRI coupled to a minimally invasive injection of DOTA-Gd in the cisterna magna. Our key finding is that the glymphatic system is severely impaired after SAH and in the acute phase of ischemic stroke, independently of intracranial hypertension.
Current evidence suggests that the glymphatic system plays an important role in cerebral waste clearance.6 Once impaired, the accumulation of metabolites within the parenchyma might lead to brain injury. This event could be particularly relevant during delayed cerebral ischemia occurring after SAH.7 Moreover, whether prolonged glymphatic dysfunction may promote vasospasm or microcirculatory impairment deserves further investigations.8 If true, restoring glymphatic perfusion by intraventricular fibrinolysis may be effective to improve patient outcome.
During the acute phase of ischemic stroke, the observed reduction of glymphatic perfusion may prevent adequate clearance of excitatory neurotransmitters (and other deleterious molecules) and promote neuronal death. The mechanisms responsible for this blockade remain elusive but may include: (1) a decrease of the arterial pulsation because of vessel occlusion9 or (2) the occlusion of the perivascular space because of compression by the intravascular thrombus. According to both hypotheses, the glymphatic perfusion appeared normal 24 hours after ischemic stroke, when the middle cerebral artery was repermeabilized. Consequently, beside restoration of arterial patency, intravenous thrombolysis may improve glymphatic perfusion in ischemic stroke patients. Our results support the findings from a recent study showing reduced interstitial fluid drainage after microstroke using 2-photon microscopy.10 In contrast, we failed to confirm the reduction of the ipsilateral glymphatic flow after carotid occlusion.9 Differences in the delay between arterial ligation and tracer administration (0.5 versus 24 hours) may explain this discrepancy.
In conclusion, we demonstrate that the glymphatic system is severely impaired after SAH and in the acute phase of ischemic stroke. Further studies are warranted to investigate the impact of such dysfunctions on neurological outcome. Notably, tissue-type plasminogen activator administration, either by removing perivascular fibrin after SAH or by restoring arterial patency after ischemic stroke, may improve glymphatic perfusion.
Sources of Funding
This work was supported by the Institut National de la Santé et de la Recherche Médicale.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.114.006617/-/DC1.
- Received July 23, 2014.
- Accepted August 5, 2014.
- © 2014 American Heart Association, Inc.
- Iliff JJ,
- Wang M,
- Liao Y,
- Plogg BA,
- Peng W,
- Gundersen GA,
- et al
- Gauberti M,
- Montagne A,
- Marcos-Contreras OA,
- Le Béhot A,
- Maubert E,
- Vivien D
- Sabri M,
- Ai J,
- Macdonald RL
- Gaberel T,
- Gakuba C,
- Hebert M,
- Montagne A,
- Agin V,
- Rubio M,
- et al
- Iliff JJ,
- Nedergaard M
- de Rooij NK,
- Greving JP,
- Rinkel GJ,
- Frijns CJ
- Iliff JJ,
- Wang M,
- Zeppenfeld DM,
- Venkataraman A,
- Plog BA,
- Liao Y,
- et al