Safety and Efficacy of Transcranial Direct Current Stimulation in Acute Experimental Ischemic Stroke
Background and Purpose—Transcranial direct current stimulation is emerging as a promising tool for the treatment of several neurological conditions, including cerebral ischemia. The therapeutic role of this noninvasive treatment is, however, limited to chronic phases of stroke. We thus ought to investigate whether different stimulation protocols could also be beneficial in the acute phase of experimental brain ischemia.
Methods—The influence of both cathodal and anodal transcranial direct current stimulation in modifying brain metabolism of healthy mice was first tested by nuclear magnetic resonance spectroscopy. Then, mice undergoing transient proximal middle cerebral artery occlusion were randomized and treated acutely with anodal, cathodal, or sham transcranial direct current stimulation. Brain metabolism, functional outcomes, and ischemic lesion volume, as well as the inflammatory reaction and blood brain barrier functionality, were analyzed.
Results—Cathodal stimulation was able, if applied in the acute phase of stroke, to preserve cortical neurons from the ischemic damage, to reduce inflammation, and to promote a better clinical recovery compared with sham and anodal treatments. This finding was attributable to the significant decrease of cortical glutamate, as indicated by nuclear magnetic resonance spectroscopy. Conversely, anodal stimulation induced an increase in the postischemic lesion volume and augmented blood brain barrier derangement.
Conclusions—Our data indicate that transcranial direct current stimulation exerts a measurable neuroprotective effect in the acute phase of stroke. However, its timing and polarity should be carefully identified on the base of the pathophysiological context to avoid potential harmful side effects.
Transcranial direct current stimulation (tDCS) is currently used to modulate brain activity in many neurological disorders.1 Several tDCS approaches, based on different polarity, density of current, and timing of application, have been proposed.2 In stroke patients, the ability of tDCS to modulate cortical activity of both the lesioned and the unaffected hemisphere has been used, so far, to ameliorate chronic motor impairment, poststroke depression, and cortical deficits.3 Anodal stimulation (A-tDCS) of the affected hemisphere or cathodal stimulation (C-tDCS) of the unaffected hemisphere are currently used in stroke clinical trials with the main aim of increasing synaptic plasticity by counteracting the functional imbalance between the 2 hemispheres.4 In earlier phases of stroke, the use of tDCS is still controversial and has shown some hints of efficacy. Excitatory A-tDCS improved motor performances if applied 2 days after permanent ischemic stroke in rats,5 whereas inhibitory electric stimulation produced slight positive effects in a model of transient ischemia.6 The effects of tDCS in acute stroke are, therefore, far from being fully understood and a finetuning of the technique, based on the pathological context and timing, is still mandatory to foster bench-to-bedside translation.7 Among the proposed mechanisms of action of tDCS,8 it is known that A-tDCS is able to increase neuronal excitability and spontaneous firing, whereas C-tDCS has opposite effects.9,10 In fact, C-tDCS in healthy humans reduces neurotransmitters, such as gamma-aminobutyric acid (GABA) and glutamate, exerting an inhibitory effect.11 Interestingly, a therapeutic tDCS protocol in acute stroke could limit the massive release of excitatory neurotransmitters from damaged neurons, among which glutamate exerts a pivotal role in triggering early excitotoxic damage and cortical spreading depression waves.12 We, therefore, investigated whether the application of tDCS would exert overall inhibitory effects in the acute phase of stroke, thus limiting the ischemic damage and avoiding functional disability.
Materials and Methods
Animals and Experimental Procedures
This study was performed on a total of 137 C57BL/6 male mice purchased from Charles River Italy and aged 8 to 10 weeks (20–22 g).
A group of 36 healthy mice underwent sham middle cerebral artery occlusion (MCAO) coupled with a tDCS protocol consisting of 20′ on–20′ off–20′ on of either cathodal tDCS (C-tDCS-sham) or anodal tDCS (A-tDCS-sham), and the control group remained nonstimulated (N-tDCS-sham; Figure IA in the online-only Data Supplement). These healthy mice were euthanized for nuclear magnetic resonance (NMR) spectroscopy (C-tDCS-sham n=4; A-tDCS-sham n=4; N-tDCS-sham n=4) and Western blot analysis (C-tDCS-sham n=8; A-tDCS-sham n=8; N-tDCS-sham n=8) immediately after tDCS.
In order to test the effects of tDCS on cerebral ischemia, 101 mice were weighted and randomized into 3 treatment groups and subjected to 90-minute-long proximal MCAO.
Mice (n=75) were treated with the same tDCS protocol consisting of 20′ on–20′ off–20′ on of either cathodal tDCS (C-tDCS-MCAO n=25) or anodal tDCS (A-tDCS-MCAO n=25), and the control group remained nonstimulated (N-tDCS-MCAO n=25) starting after the first 30 minutes of MCAO (Figure IC in the online-only Data Supplement). Regional cerebral blood flow (rCBF) was calculated during MCAO and following reperfusion. A randomly selected subgroup of 9 ischemic mice was euthanized for NMR, immediately after the procedure (N-tDCS-MCAO n=3; C-tDCS-MCAO n=3; A-tDCS-MCAO n=3). Forty-three ischemic mice were weighted and euthanized for histological analysis at 24 hours (C-tDCS-MCAO n=17; A-tDCS-MCAO n=17; N-tDCS-MCAO n=18) and a subgroup of 18 animals (N-tDCS-MCAO n=6; C-tDCS-MCAO n=6; A-tDCS-MCAO n=6) was randomly selected for neurophysiological and behavioral functional assessment prior to enthanization.
Twenty-three ischemic mice were allowed to survive ≤72 hours after ischemia (C-tDCS-MCAO n=8; A-tDCS-MCAO n=8; N-tDCS-MCAO n=7) to estimate final lesion volume, pathology, and behavior.
In another experimental stimulation setup, mice (n=26) were subjected to 90-minute-long proximal MCAO and were treated with the same tDCS protocol consisting of 20′ on–20′ off–20′ on starting 4.5 hours after the induction of MCAO (C-tDCS-MCAO n=8; A-tDCS-MCAO n=8; N-tDCS-MCAO n=10; Figure II in the online-only Data Supplement). Overall mortality rate was 25% (8/32) in N-tDCS-MCAO, 6.7% (2/30) in C-tDCS-MCAO, and 20% (6/30) in A-tDCS-MCAO group.
Experimental procedures were approved by the Institutional Animal Care and Use Committee at San Raffaele Scientific Institute. All animals and results were studied in a blinded fashion for treatment. For detailed description of abovementioned techniques, see online-only Data Supplement.
tDCS Modifies Metabolite Pattern and N-Methyl-D-Aspartate Receptor Expression in Healthy Mouse Cortex
To first investigate the metabolic effect of tDCS, C57BL/6 healthy mice were subjected to either sham, cathodal, or anodal t-DCS with a protocol consisting of 20′ on–20′ off–20′ on stimulation. According to previous literature13 and pilot experiments (Figure III in the online-only Data Supplement), we identified this stimulation protocol as the best suited to induce long-term results (repeated stimulation based on priming effect) without inducing current-related tissue damage (short stimulation periods). Immediately after stimulation, we measured a wide range of metabolic cortical patterns by NMR spectroscopy (Figure 1B). C-tDCS-treated healthy mice (hereafter referred as C-tDCS-sham) showed a significant decrease in cortical glutamate compared with nonstimulated sham–treated healthy mice (N-tDCS-sham; ratio 0.75; P<0.05). A statistically significant decrease of alanine levels was also observed in the cortex of C-tDCS-sham, when compared with A-tDCS-treated healthy mice (A-tDCS-sham; ratio 0.68; P<0.05). A-tDCS-sham mice displayed instead an increase of cortical lactate compared with N-tDCS-sham mice (ratio 1.258; P<0.05; Figure 1A).
To validate whether glutamatergic synaptic transmission was effectively altered by C-tDCS, we measured by Western blot analysis the expression of 2 NMDA receptor (NMDAR) subunits: NR1 and NR2B. The former subunit is constitutively expressed, whereas the latter exerts a regulatory function during increased glutamatergic transmission.14
Although there were no differences in the expression of NR1 subunit among the 3 groups, C-tDCS-sham mice displayed a statistically significant lower expression of the NR2B subunit compared with both N-tDCS-sham (P<0.05) and A-tDCS-sham mice (P<0.05; Figure 1C). Altogether, these results indicate that C-tDCS, but not A-tDCS, is able to reduce cortical glutamate activity and metabolism.
C-tDCS Protects From Ischemic Brain Damage
In order to investigate the role of tDCS in acute stroke, we applied the identified stimulation protocol (20′ on–20′ off–20′ on) to C57BL/6 mice in the very acute phase of 90-minute-long transient proximal MCAO (Figure IC in the online-only Data Supplement).
Mean weight loss at 24 hours over baseline was comparable among the 3 treatment groups (−8.5%±1.5 SEM in N-tDCS-MCAO; −12.4%±1.2 SEM in C-tDCS-MCAO; and −10.9%±0.9 SEM in A-tDCS-MCAO), whereas modified Neurological Severity Score revealed a significant treatment effect (H=8.65; 2 d.f.; P<0.05) with an amelioration induced by C-tDCS compared with N-tDCS-MCAO and A-tDCS-MCAO mice (P<0.05) that persisted ≤3 days after ischemia (H=7.60; 2 d.f.; P<0.05; Figure 2A). We also performed a functional neurophysiological assessment of resting motor threshold (RMT) by stimulating the cortex of ischemic mice at 24 hours postischemia. RMT displayed a significant treatment effect and C-tDCS-MCAO mice had a preservation of RMT compared with A-tDCS-MCAO and N-tDCS-MCAO mice (H=7.89; 2 d.f.; P<0.05). RMT did not significantly increase over baseline in C-tDCS-MCAO mice (baseline values, 4.13±0.24 SEM; 24-hour values, 4.53±0.44 SEM; increase over baseline, 109.0%±6.73 SEM; n.s.), whereas it significantly increased in both A-tDCS-MCAO (baseline values, 4.37±0.20 SEM; 24-hour values, 6.37±0.50 SEM; increase over baseline, 159.9%±12.45 SEM; P<0.001) and N-tDCS-MCAO mice (baseline values, 4.25±0.19 SEM; 24-hour values, 6.32±0.41 SEM; increase over baseline, 150.5%±12.32 SEM; P<0.001). These findings were supportive of a neuroprotective role of C-tDCS on the ischemic sensorimotor area and corticospinal tract.
Neuropathological analysis revealed indeed that C-tDCS induced a specific preservation of the cytoarchitecture of the cerebral cortex, a feature never observed in N-tDCS-MCAO and A-tDCS-MCAO mice (Figure 2B). Stereological morphometric analysis of the ischemic lesion volume confirmed a significant treatment effect (F2,33=20.40; P<0.0001). C-tDCS-MCAO mice showed a remarkable reduced ischemic volume (34.18%±2.37 SEM) compared with both N-tDCS-MCAO (54.22%±2.30 SEM; P<0.001) and A-tDCS-MCAO mice (47.29%±2.77 SEM; P<0.001). Although the reduction of infarct volume in C-tDCS-MCAO mice persisted over time (30.46%±2.10 SEM; P<0.01 versus N-tDCS-MCAO, and P<0.001 versus A-tDCS-MCAO), at 3 days postischemia A-tDCS-MCAO mice showed an increased ischemic lesion (58.67%±2.59 SEM) compared with N-tDCS-MCAO mice (48.13%±1.25 SEM; P<0.01; Figure 2C). At the same time point, a reduction of brain edema was also observed in C-tDCS-MCAO (17.52%±2.64 SEM) compared with N-tDCS-MCAO (32.23%±2.13 SEM; P<0.01) and A-tDCS-MCAO mice (26.80%±1.13 SEM; P<0.01). Consistently, quantification of apoptotic cells in cortical and striatal areas, identified as activated Caspase3+ cells, displayed a significant treatment effect (H=18.17; 2 d.f.; P<0.001, and H=10.87; 2d.f.; P<0.01, respectively): at cortical and striatal level, C-tDCS-MCAO mice had a reduced number of Caspase3+ cells compared with both N-tDCS-MCAO (P<0.05) and A-tDCS-MCAO (P<0.01). Indeed, apoptosis of cortical neurons (Casp3+NeuN+ cells) was reduced in C-tDCS-MCAO (0.37±0.07 SEM) compared with both N-tDCS-MCAO (1.69±0.36 SEM) and A-tDCS-MCAO mice (2.4±0.33 SEM; P<0.05). On the contrary, A-tDCS-MCAO mice had a significant increase in the number of apoptotic cells compared with N-tDCS-MCAO mice (P<0.05) with an increased number of both apoptotic neurons and endothelial cells.
We further investigated whether the clinicopathological protective effect of C-tDCS in MCAO mice was preceded by an alteration of cortical metabolites. NMR metabolic profiles were assessed in all treatment groups: C-tDCS-MCAO mice showed a significant decrease of cortical glutamate (ratio 0.53; P<0.05) in comparison with N-tDCS-MCAO mice. In addition, creatine (ratio 0.69; P<0.05) and taurine (ratio 0.72; P<0.05) levels were reduced in C-tDCS MCAO mice (Figure 2D), confirming the overall inhibitory activity of C-tDCS on ischemic brain metabolism.
Inflammatory Response in MCAO Mice Treated by tDCS
In order to investigate the mechanisms underlying the effects of tDCS on MCAO mice, we evaluated the periischemic inflammatory response, namely resident ionized calcium-binding adapter molecule 1 (Iba1)+ microglia, blood-borne CD45+ central nervous system-infiltrating mononuclear cells, and myeloperoxidase (MPO)+ neutrophils, at 24 and 72 hours postischemia.
No difference in the number of Iba1+ cells was found at striatal level, whereas there was a striking difference at cortical level among the 3 treatment groups at 24 hours postischemia (H=9.62; 2 d.f.; P<0.01); a lower amount of Iba1+ cells was found in the periischemic cortex of C-tDCS-MCAO mice, compared with both N-tDCS-MCAO and A-tDCS-MCAO mice (P<0.05; Figure 3A–3C).
Quantification of periischemic CD45+ leucocytes showed also a significant treatment effect both in the periischemic cortex and striatum (H=9.673; 2 d.f.; P<0.01, and H=6.140; 2 d.f.; P<0.05). A-tDCS-MCAO mice presented increased numbers of periischemic CD45+ cells compared with both C-tDCS-MCAO (cortex, P<0.05 at 24 and 72 hours; striatum, P<0.01 at 72 hours) and N-tDCS-MCAO mice (striatum P<0.05 at 24 and 72 hours; cortex P<0.05 at 72 hours; Figure 3B and 3C). Similarly, MPO+ neutrophils in the upper ischemic cortex and striatum showed a significant treatment effect at 24 (H=9.27; 2 d.f.; P<0.01, and H=8.92; 2 d.f.; P<0.05, respectively) and 72 hours postischemia (H=5.87; 2 d.f.; P<0.05, and H=4.86; 2 d.f.; P<0.05, respectively): A-tDCS-MCAO mice showed a significant increase of infiltrating MPO+ cells with respect to both N-tDCS-MCAO (upper cortex, P<0.05; striatum, P<0.01) and C-tDCS-MCAO mice (upper cortex, P<0.01; striatum, P<0.05) 24 hours postischemia (Figure 3D). Interestingly, although MPO+ and CD45+ cells augmented in the cortex of N-tDCS-MCAO mice at 72 hours postischemia, C-tDCS-MCAO mice had a lower number of infiltrating cells compared with A-tDCS-MCAO mice (P<0.05).
A-tDCS Exacerbates Dysregulation of Postischemic Blood Brain Barrier
The increased number of inflammatory cells in A-tDCS-MCAO mice prompted us to investigate whether this stimulation protocol might have provoked a further dysregulation of the postischemic blood brain barrier (BBB). We thus evaluated the ratio of endogenous IgG extravasation in the ipsilateral ischemic over the contralateral healthy hemisphere (Figure 4A). A significant treatment effect was found (H=9.49; 2 d.f.; P<0.01, H=9.18; 2 d.f.; P<0.05, and H=7.96; 2 d.f.; P<0.05, respectively). A-tDCS-MCAO mice showed a significant increase of endogenous IgG leakage with respect to N-tDCS-MCAO (upper cortex, P<0.05) and C-tDCS-MCAO mice (upper cortex and striatum, P<0.01; lower cortex, P<0.05; Figure 4B–4D).
Consistently, structural analysis of the endothelial tight junction protein Zona Occludens-1, which is fundamental in maintaining the BBB,15 confirmed a significant treatment effect in the upper cortex (H=9.63; 2 d.f.; P<0.01). A-tDCS-MCAO mice displayed a significant lower coverage of Zona Occludens-1 over CD31+ cortical blood vessels, with respect to both N-tDCS-MCAO (P<0.01) and C-tDCS-MCAO mice (P<0.01; Figure 4C and 4D).
A-tDCS Increases Early Postischemic Reperfusion and Late Hemorrhagic Transformation
Previous evidence has shown that A-tDCS is able to increase rCBF in both healthy human and rodents.16,17 We thus assessed whether A-tDCS could influence early postischemic reperfusion (rCBF) values. No differences on rCBF were found before and during the induction of the MCAO in A-tDCS, C-tDCS, or N-tDCS-MCAO mice (F2,72=0.53; P=0.59). However, during reperfusion a significant treatment effect on rCBF in the territory of the MCA was found (F2,72=2.273; P<0.0001). Post hoc analysis revealed a significant increase of early reperfusion in A-tDCS-MCAO (P<0.001), compared with both C-tDCS-MCAO and N-tDCS-MCAO mice (Figure 5A). Considering the early postischemic reperfusion increase and the BBB structural and functional dysregulation induced by A-tDCS, we conducted a stereological analysis of cerebral bleedings. We observed a significant increase in the total number of hemorrhages (H=7.03; 2 d.f.; P<0.05) and total hemorrhagic area (H=9.55; 2 d.f.; P<0.005) in A-tDCS-MCAO mice if compared with both C-tDCS-MCAO (P<0.01) and N-tDCS-MCAO (P<0.01; Figure 5B). Cerebral bleedings were mostly found in brain areas located under the stimulating electrode, with remarkable differences among A-tDCS-MCAO, C-tDCS-MCAO, and N-tDCS-MCAO mice (Figure 5C and 5D). These data indicate that A-tDCS, in the acute phase of stroke, alters the structural and functional characteristic of the BBB, leading to hemorrhagic transformation.
Effects of tDCS When Delivered 4.5 Hours After MCAO
We finally sought to investigate whether C-tDCS had similar beneficial effects even if applied at later time points after ischemia. We applied the same 20′ on–20′ off–20′ on tDCS protocol starting 4.5 hours after the induction of 90-minute MCAO (Figure IIA in the online-only Data Supplement). We found that C-tDCS induced a functional amelioration of neurological deficits (H=7.34; 2 d.f.; P<0.05) evaluated at 24 hours postischemia, compared with N-tDCS-MCAO and A-tDCS-MCAO mice (P<0.05). The observed therapeutic effect of C-tDCS-MCAO was also paralleled by a slight reduction of the ischemic volume (C-tDCS-MCAO 42.57%±3.73 SEM; A-tDCS-MCAO 49.39%±1.89 SEM; n.s., and N-tDCS-MCAO 50.03%±2.35 SEM; n.s.) and a significant reduction of brain edema compared with A-tDCS-MCAO mice (C-tDCS-MCAO 31.86%±3.45 SEM; A-tDCS 48.33%±5.80 SEM; P<0.05, and N-tDCS-MCAO 37.68%±4.51 SEM; n.s.; Figure IIB and IIC in the online-only Data Supplement). Remarkably, we confirmed a significant increase in total hemorrhagic area (H=8.36; 2 d.f.; P<0.05) in A-tDCS-MCAO mice if compared with both C-tDCS-MCAO (P<0.01) and N-tDCS-MCAO (P<0.05; Figure IID in the online-only Data Supplement).
Our data show for the first time a protective effect of C-tDCS if applied in the very early stage of experimental ischemic stroke. As already observed in human studies,11,18 C-tDCS was able to lower cortical glutamate in our mice. This effect could be due, at least in part, to a decreased rate of glutamate synthesis subsequent to spontaneous neuronal firing modulation.19 Interestingly, we also found a lower level of alanine in C-tDCS-treated mice, which correlates with glutamate levels and may be suggestive of an increased conversion of glutamate into glutamine.20 These findings were corroborated by a striking downregulation of NR2B in C-tDCS-sham mice, suggesting that C-tDCS might be able to reduce the overall cortical glutamate activity, through the modulation of NMDAR. Previous findings showed indeed that NMDAR antagonists play a major role in attenuating most of tDCS cortical effects both in vivo21 and in vitro.22 Compared with other NMDAR subunits, NR2B preferentially contributes to pathological processes linked to overexcitation of glutamatergic pathways, representing an ideal target for neuroprotection.23,24 As a matter of fact, in mature cortical cultures and in rats subjected to brain ischemia, activation of either synaptic or extrasynaptic NR2B-containing NMDARs results in excitotoxicity and neuronal apoptosis.25
The ability of short pulses of C-tDCS to lower the excess of cortical glutamate in healthy mice prompted us to explore whether or not such stimulation protocol might have a neuroprotective role in the very early phase of ischemic stroke, which is characterized by overt glutamatergic excitotoxicity.12 We thus tested our protocol in mice undergoing 90-minute MCAO. C-tDCS was able to induce a sustained reduction of ischemic stroke volume ≤3 days postischemia when applied in the very early phase of MCAO and, although to a lesser extent, also when applied ≤4.5 hours poststroke. As a consequence, clinical deficits were significantly improved and functional preservation of the cortical and descending motor pathways (assessed with RMT) was observed. Consistently, we also found that C-tDCS was able to preserve the viability of cortical neurons, as a reduced number of Caspase3+ apoptotic neuronal cells was found in the cortex of C-tDCS-MCAO-treated mice compared with A-tDCS-MCAO and N-tDCS-MCAO mice. The role of tDCS in early stages of ischemic stroke was further explored by measuring the inflammatory reaction occurring in response to ischemia. It is known that ischemic neurons produce discrete amounts of tumor necrosis factor-α, a cytokine that can further exacerbate the ischemic damage by activating microglia.26 Additionally, glutamate per se might also exert a pivotal role in triggering microglia activation.27 Whatever the triggering factor, activated microglia produce neurotoxins, including nitric oxide and reactive oxygen species, that worsen ischemic tissue damage.28 Indeed, C-tDCS-MCAO mice showed a lower amount of cortical Iba1+ cell infiltration compared with A-tDCS-MCAO and N-tDCS-MCAO mice, and a lower amount of cortical glutamate with a proportional decrease in taurine.29 This finding might be attributed to the fact that the neuroprotective effect exerted by C-tDCS in the early phase of stroke also protects and reduces secondary injury mechanisms, such as the postischemic inflammatory reaction sustained by activated microglia.30 Nevertheless, we cannot exclude that our results might be also due, at least in part, to the ability of C-tDCS to reduce cortical-spreading depressions,31,32 a major mechanism of acute ischemic damage induced by glutamate excitotoxicity. The reduced levels of creatine after stroke in C-tDCS-MCAO mice copes with this hypothesis, because this metabolite has been associated with the electric silencing of the ischemic penumbra.33 Because the altered metabolic balance is one of the main detrimental mechanisms of acute ischemic stroke, it could be also argued that C-tDCS exerts its neuroprotective role by reducing brain metabolism in ischemic regions.
Despite previous data on the positive effects of A-tDCS in the subacute phase after cerebral ischemia (1 day and 1 week after MCAO),34 we found that A-tDCS in the very acute phase of ischemic stroke was ineffective in inducing a functional amelioration. Although this apparent discrepancy might be simply explained by the different timing and current density (2.8 versus 5.5 mA/cm2) of the 2 protocols, we favor attributing the detrimental consequences of acute A-tDCS to the alterations induced by the stimulation on vascular tone and BBB integrity. The observed facilitation of early reperfusion on filament withdrawal in A-tDCS-MCAO mice is in line with previous reports showing that A-tDCS is able to increase rCBF in cortical regions by modulating cerebral vessels.16 Because vascular and neuronal functions in the brain are closely interrelated into the brain, changes in blood perfusion during tDCS could be consistent with a primary neuronal action via neurovascular coupling.17 However, supporting evidence suggests a direct effect of tDCS on vascular response.35 Additionally, endothelial cell cultures, including models of the BBB, can be electrically stimulated and high-intensity electric stimulation was shown to increase transport across this model through a phenomenon defined as electropermeation.36 Indeed we found that A-tDCS had major effects on vasodynamic changes (increased postischemic reperfusion) and BBB structural and functional integrity in our mouse model of cerebral ischemia.
Despite increased reperfusion might be interpreted as beneficial in some cases, the observed detrimental behavioral and neuropathological outcomes in A-tDCS-MCAO mice suggests that early hyperreperfusion might represent a detrimental early hyperemia.37 In fact, although restoration of oxygen and glucose supply reinstates the oxidative phosphorylation that helps normalizing energy after reperfusion, a parallel cascade of deleterious biochemical processes can also antagonize the beneficial effect of early reperfusion.38 It has been variably reported that oxygen consumption might decrease during hyperemia despite an increased blood flow in the hyperemic areas.37 As a matter of fact, cerebral autoregulation, which usually provides dynamic protection to the brain from excessive perfusion, is often impaired after ischemic stroke due to the early release of nitric oxide (and free radicals) damaging cerebrovascular endothelium, thus inducing vasodilatation and increased cerebral vessel permeability.39,40
BBB disruption results in albumin and high-molecular-weight proteins extravasation. Accordingly, A-tDCS, but not C-tDCS, increased the extravasation of IgG from circulating blood and reduced blood vessel tight junctions. Therefore, the increased number of hemorrhages observed in A-tDCS-MCAO mice might be attributed to these latter findings and to the abovementioned increased reperfusion. Although increased blood flow induced by A-tDCS might be beneficial in subacute and chronic stroke, the present data point out that our acute stroke setting is rather detrimental as it leads to increased BBB dysfunction, increased edema, and ischemic lesion volume. In addition to its effects on BBB integrity and vascular tone, oxidative damage is also responsible for delayed tissue damage, apoptosis, and inflammation after acute ischemic stroke.41 Indeed, we found that A-tDCS applied during MCAO induced an increased number of cortical Caspase3+ cells compared with N-tDCS-MCAO mice (which did not preferentially affect neuronal or endothelial cell apoptosis), an increase of final ischemic volume, as well as a dramatic increase of blood-borne, central nervous system-infiltrating inflammatory CD45+ and MPO+ cells. Consistently, we found that A-tDCS induced an increase of brain edema and hemorrhagic transformation, also when administered 4.5 hours after MCAO.
Despite the mechanisms of tDCS in stroke are far from being fully understood, our study supports the use of cathodal tDCS in the hyperacute phase of ischemic stroke, although indicating a possible hazardous effects of A-tDCS in the very same phase. It is reasonable to think that different polarities of tDCS could have a Janus-like role in the ischemic brain, depending on the timing, site of application, and pathophysiological milieu. Nevertheless, when appropriately tuned, tDCS in acute stroke is a promising therapeutic strategy because of its low costs, fast administration, and ease of delivery to brain cortical areas, regardless of blood perfusion.
Study conception: Dr Leocani; study design: Drs Peruzzotti-Jametti, Cambiaghi, Bacigaluppi, Comi, Musco, Martino, and Leocani; data collection: Dr Peruzzotti-Jametti, Dr Cambiaghi, M. Gallizioli, E. Gaude, Dr Mari, S. Sandrone, and Dr Teneud; data analysis: Dr Peruzzotti-Jametti, Dr Cambiaghi, Dr Bacigaluppi, E. Gaude, Dr Mari, S. Sandrone, and Dr Martino; data interpretation and article writing: Drs Peruzzotti-Jametti, Cambiaghi, Bacigaluppi, Comi, Musco, Martino, and Leocani.
Sources of Funding
Joint Italian–Israeli laboratory on Brain modulation in neuroimmune, neurodegenerative, and mental disorders (Italian Ministry of Foreign Affairs), TargetBrain (EU Framework 7 project HEALTH-F2-2012–279017), and NEUROKINE (EU Framework 7 ITN project).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.113.001687/-/DC1.
- Received April 4, 2013.
- Revision received July 15, 2013.
- Accepted July 23, 2013.
- © 2013 American Heart Association, Inc.
- Baba T,
- Kameda M,
- Yasuhara T,
- Morimoto T,
- Kondo A,
- Shingo T,
- et al
- Stagg CJ,
- Best JG,
- Stephenson MC,
- O’Shea J,
- Wylezinska M,
- Kincses ZT,
- et al
- Liu Y,
- Wong TP,
- Aarts M,
- Rooyakkers A,
- Liu L,
- Lai TW,
- et al
- Rothman DL,
- Sibson NR,
- Hyder F,
- Shen J,
- Behar KL,
- Shulman RG
- Liebetanz D,
- Nitsche MA,
- Tergau F,
- Paulus W
- Zhang F,
- Guo A,
- Liu C,
- Comb M,
- Hu B
- Picconi B,
- Tortiglione A,
- Barone I,
- Centonze D,
- Gardoni F,
- Gubellini P,
- et al
- Barone FC,
- Arvin B,
- White RF,
- Miller A,
- Webb CL,
- Willette RN,
- et al
- Kaushal V,
- Schlichter LC
- Astrup J,
- Siesjö BK,
- Symon L
- Olsen TS,
- Larsen B,
- Skriver EB,
- Herning M,
- Enevoldsen E,
- Lassen NA
- Gürsoy-Ozdemir Y,
- Can A,
- Dalkara T
- Janigro D,
- West GA,
- Nguyen TS,
- Winn HR
- Eames PJ,
- Blake MJ,
- Dawson SL,
- Panerai RB,
- Potter JF