The Role of Sartans in the Treatment of Stroke and Subarachnoid Hemorrhage: A Narrative Review of Preclinical and Clinical Studies

Background: Delayed cerebral vasospasm (DCVS) due to aneurysmal subarachnoid hemorrhage (aSAH) and its sequela, delayed cerebral ischemia (DCI), are associated with poor functional outcome. Endothelin-1 (ET-1) is known to play a major role in mediating cerebral vasoconstriction. Angiotensin-II-type-1-receptor antagonists such as Sartans may have a beneficial effect after aSAH by reducing DCVS due to crosstalk with the endothelin system. In this review, we discuss the role of Sartans in the treatment of stroke and their potential impact in aSAH. Methods: We conducted a literature research of the MEDLINE PubMed database in accordance with PRISMA criteria on articles published between 1980 to 2019 reviewing: “Sartans AND ischemic stroke”. Of 227 studies, 64 preclinical and 19 clinical trials fulfilled the eligibility criteria. Results: There was a positive effect of Sartans on ischemic stroke in both preclinical and clinical settings (attenuating ischemic brain damage, reducing cerebral inflammation and infarct size, increasing cerebral blood flow). In addition, Sartans reduced DCVS after aSAH in animal models by diminishing the effect of ET-1 mediated vasoconstriction (including cerebral inflammation and cerebral epileptogenic activity reduction, cerebral blood flow autoregulation restoration as well as pressure-dependent cerebral vasoconstriction). Conclusion: Thus, Sartans might play a key role in the treatment of patients with aSAH.


Preclinical Studies on Sartans in Animal Models of Ischemic Stroke
The search finally yielded 64 preclinical studies on "Sartans AND ischemic stroke", eligible for systematic review (Table 1). Of the articles included in the final analysis, a systematic review on the beneficial and non-beneficial effect in the preclinical and clinical settings was performed. Summary measures are reported as outcome measures (i.e., infarct size, neurocognition, inflammation).

Drug
Candesartan (CS), another AT 2 -1-R antagonist, reduced ischemic brain damage following MCAO occlusion [39]. CS and curcumin together significantly restored superoxide dismutase activity and blood flow compared with the untreated group [40]. Further, CS upregulated vascular endothelial growth factor (VEGF) B after induction of focal cerebral ischemia using a MCAO model. In contrast to saline-treatment after reperfusion, CS further improved neurobehavioral and motor functions and decreased infarct size [41]. VEGF-B silencing was shown to diminish CS (1 mg/kg) protective effects [42]. CS (0.3 mg/kg) was able to improve recovery from ischemic stroke in low doses by maintaining blood pressure during reperfusion [43]. CS induced early protective effects with improvement in motor function, upregulated brain-derived neutrotrophic factor (BDNF), and also reduced endoplasmatic reticulum stress markers [44]. In a MCAO BDNF, knock-out model rats received CS or saline at reperfusion for 14 days, revealing better functional outcomes, increased vascular density, and synaptogenesis in the CS (1 mg/kg) group [45,46]. In addition, CS (0.16 µM) significantly increased BDNF production [47]. Furthermore, CS (10 nM) improved cell function and viability of brain capillary endothelial cells under oxygen glucose deprivation, providing protective blood-brain-barrier (BBB) effects [48]. In other transient MCAO rat models, CS (0.1 mg/kg; 0.3 mg/kg; 1.5 or 10 mg/kg per day; 0.1, 1 and 10 mg/kg; 0.1, 0.3 or 1 mg/kg; 0.1 mg/kg twice daily; 1 mg/kg; 0.3 or 3 mg/kg per day; 0.5 mg/kg per day for 14 days; 0.1 or 0.3 mg/kg; 0.5 mg/kg per day for 3 to 14 days) showed improved neurological function with significant reduction in BBB disruption, in cerebral ischemia, and in edema [39,[49][50][51][52][53][54][55][56][57][58][59][60]. In a bilateral common carotid artery occlusion (CCAO) model in rats, pretreatment with CS (0.1 and 0.3 mg/kg) and atorvastatin significantly attenuated neurobehavioral alterations, oxidative damage, and restored mitochondrial enzyme dysfunction compared to the control group [61,62]. AT 2 -1-R administration prior to ET-1 induced MCAO provides neuroprotective effects, with CS (0.2 mg/kg per day for seven days) pretreatment attenuating infarct size and neurological deficits without altering systemic BP [63]. Pretreatment with CS for five days significantly decreased mortality, neurological deficits, and infarct size [67]. A combined inhibition of AT 2 -1-(0.05 mg/kg per day) and ET A -receptors decreased brain damage as well; additionally, an upregulation of AT 2 -1-R in ischemic middle cerebral artery smooth muscle cells (SMCs) was found [68,69]. Also, early (3 h) and delayed (24 h) effects of CS treatment (0.3 and 3 mg/kg) continued for seven days after onset of MCAO with reperfusion in normotensive rats involved a reduction of the infarct volume by low doses of CS [70]. CBF in CS (0.5 mg/kg) pretreated animals at 0.5 h after MCAO was significantly increased compared to the control group [71]. Other groups additionally showed a four-week CS-pretreatment (0.3 mg/kg per day) before MCAO clearly associated with complete reversal of a decreased lumen diameter and increased media thickness as well as decreased endothelial nitric oxide synthase (eNOS) and increased inducible nitric oxide synthase (iNOS) protein and mRNA in SR-SHR and in a normotensive control group [72].
Olmesartan (OMS), an AT 2 -1-R antagonist, has been evaluated in a bilateral CCAO model in mice, revealing improved cognitive outcome, neuroprotective effects, attenuation of oxidative hippocampal stress, and suppression of BBB disruption compared to control groups [73]. A single carotid ligation stroke model in gerbils showed that OMS (10 mg/kg per day started 36 h after stroke) was associated with an increased survival [74]. Other studies demonstrated that OMS (10 mg/kg per day for 14 days after infarct; 10 mg/kg per day for 7 days before and 14 days after infarct; 10 mg/kg per day for 7 days before infarct) treatment in a rat MCAO model showed significantly better functional scores and reduced infarct size and cell death [75]. OMS (0.01 or 0.1 µmol/kg per hour for seven days) reduced brain angiotensin II, MMP-2 and MMP-9 upregulation following brain ischemia [77].
Valsartan (VS), a selective AT 2 -1-R antagonist, reduced ischemic brain area and improved the neurological deficit after MCAO with restoration of cerebral blood flow [78]. VS significantly reduced infarct volume and improved the neurological deficit scores. VS at nonhypotensive doses significantly diminished ischemic area, neurological deficits, and reduction of cerebral blood flow as well as superoxide production [27,78,80].
Irbesartan (IS), a selective AT 2 -1-R antagonist improved motor functions, reduced infarct size and decreased the number of apoptotic cells particularly in the periinfarct area by attenuated invasion of activated microglia likewise macrophages [82][83][84].
Losartan (LS), a clinical established selective AT 2 -1-R antagonist, did not increase mortality in acute cerebral ischemia [85]. Also, LS (20 µmol/L) abolished ischemic exaggeration of cell injury [26,86]. Expression levels of pro-apoptotic genes were significant reduced by LS treatment [87]. Further LS administration initiates cerebral angiogenic response with a significantly larger vessel surface area, and administration before initiation of cerebral focal ischemia (50 mg/day for 2 weeks) markedly reduces infarct size [88].

Clinical Studies on Sartans in Ischemic Stroke
The search yielded 19 clinical studies on "Sartans AND ischemic stroke", eligible for systematic review ( Table 2). Beneficial aspects of using AT 2 -1-R antagonists before the onset of ischemic stroke have already been elucidated in a retrospective analysis of 151 patients [89].  CS has been evaluated in the Scandinavian Candesartan Acute Stroke Trial (SCAST). Within 30 h of ischemic or hemorrhagic stroke, 2029 patients either received CS-or placebo-treatment. The modified ranking Scale (mRS) was used for outcome analysis. CS showed no overall effect on vascular events in ischemic and/or hemorrhagic stroke, and the adjusted odds ratio for vascular events of patients treated within 6 h reached significance [90]. At six months, activities of daily living and level of care were assessed. In more than 1800 patients, over 1500 suffered ischemic and almost 250 hemorrhagic strokes. No statistically significant effects of CS on Barthel index or level of care could be identified [91]. Furthermore, the SCAST group evaluated whether the effect of CS treatment varies in subtypes of over 1700 ischemic strokes. Concerning functional outcomes, a trend towards a beneficial effect of CS was observed in patients with larger infarcts (total anterior circulation or partial anterior circulation) than in patients with smaller lacunar infarcts [92]. Further on, over 2000 SCAST patients were randomly allocated to placebo or CS treatment for seven days with increasing doses from 4 mg (starting day 1) to 16 mg (from day 3 to 7). After six months' follow-up, the risk of the composite vascular endpoint did not differ between the placebo and CS treatment group [93]. Also, the Acute Candesartan Cilexetil Therapy in Stroke Survivors study confirmed that administration of CS in the acute phase of stroke in 339 patients confers long-term benefits in patients who sustained acute ischemic stroke [94]. VS has been evaluated in a multicenter trial concerning efficacy and safety of modest blood pressure reduction within 48 h in more than 370 patients with acute ischemic stroke, considering the primary outcome death or dependency. The VS-treated group showed 46 of 187 patients with a 90-day mRS of 3-6, compared with 42 of 185 patients in the control group. The rate of major vascular events did not differ significantly between both groups [98]. TMS has also been evaluated concerning beneficial effects after stroke treatment. A multicenter trial, involving more than 18,500 patients with ischemic stroke, had a follow-up of 2.5 years. The primary outcome parameter was time to first recurrent stroke. Only short-term add-on TMS (80 mg/day) treatment did not mitigate this risk [100,101]. Treatment with TMS (80 mg/day) did not prevent progression of white matter lesions in patients with recent ischemic stroke [102]. Another study group enrolled 20,332 patients and analyzed 1360 patients within 72 h of ischemic stroke onset (TMS vs. placebo) concerning functional outcome after 30 days as primary outcome. Combined death or dependency did not differ between the treatment groups, showing treatment with TMS (80 mg/day) in patients with acute mild ischemic stroke and mildly elevated BP safe with no excess in adverse events [103]. Also, effects of TMS (80 mg/day) initiation early after stroke have been analyzed. From 20,332 patients with recent ischemic stroke, 10,146 patients were randomly assigned in the TMS group and 10,186 in the placebo group; 8.7% in the TMS group and 9.2% in the placebo group suffered from subsequent stroke, showing no significant reduction of recurrent stroke after early initiation [104]. LS has also been analyzed in recent clinical stroke trials. In a double-blinded multi-center trial, 196 hypertensive patients with previous ischemic stroke were randomized to cilnidipine-or LS-treatment (50-100 mg per day for four weeks) once daily for four weeks. Both treatments, however, increased global CBF despite BP lowering [105]. Additionally, the effect of long-term therapy with LS regarding cognitive function in 6206 essential hypertonic patients with additional cerebrovascular risk factors was investigated. The LS-based antihypertensive treatment increased the proportion of patients with normal cognitive function [109]. Also, the Losartan Intervention for Endpoint reduction in hypertension study group reported cardioprotective effects of a LS-based antihypertensive regimen. The incidence of any stroke, fatal stroke, and atherothrombotic stroke was significantly lower in LS-treated compared to the atenolol-treated isolated systolic hypertensive patients [106]. Other groups assessed the effect of LS treatment on mean arterial blood pressure, global, and focal CBF in 24 hypertensive patients without occlusive carotid disease 2-7 days after ischemic stroke and/or transient ischemic attack. LS (25-50 mg per day) was generally well tolerated and none of the patients suffered neurological deterioration. No changes occurred in internal carotid artery flow or cortical as well as hemispheric CBF [107].

Therapeutic Interventions After aSAH
Poor patients' outcome after aSAH is owed a multifactorial process (early brain injury, DCVS, DCI, cerebral inflammation, cortical spreading depression, loss of pressure dependent cerebral autoregulation) [4,5,7,9,[110][111][112][113]. DCVS is treated with moderate hypertensive, normovolemic, hemodilution, and in cases of therapy-refractory, DCVS with intra-arterial spasmolysis or balloon dilatation [114,115]. Research to improve poor functional outcome in patients suffering from aSAH and related DCVS is pivotal [1,5,21,116,117]. Multiple preclinical and clinical trials showed the effect of ET-1 in mediating DCVS after aSAH. CONSCIOUS-1, a randomized, double-blind, placebo-controlled study assessed the efficacy of intravenous clazosentan (ET A -R antagonist) in preventing vasospasm following aSAH. It significantly decreased angiographic DCVS with a trend for reduction in vasospasm-related morbidity/mortality [118]. CONSCIOUS-2 assigned patients with aSAH and clip ligation to clazosentan-or placebo. Thereby, clazosentan showed no significant difference in the mortality and vasospasm-related morbidity [119]. CONSCIOUS-3 assessed whether clazosentan reduced DCVS-related morbidity and mortality after aSAH and endovascular coiling. Pulmonary complications and anemia were more common in patients with clazosentan administration than in the placebo group, and mortality rates after 12 weeks were the same, respectively [120]. The REVERSE-study, infusing clazosentan intravenously in patients developing moderate to severe angiographic vasospasm after aSAH, showed a clear pharmacodynamic dilating effect on DCVS 24 h in most patients suffering aSAH, being able to reverse established angiographic vasospasm [22].
Antihypertensive agents are usually discontinued to maintain a sufficient mean arterial cerebral perfusion pressure considering the prolonged phase of DCVS between days 5 to 14 after the ictus [114]. In contrast, nimodipine, a calcium-channel antagonist, is administered for risk reduction of DCVS, yet rather its neuroprotective effects have been discussed in its beneficial role in aSAH [8,121].

Effects of Losartan Following aSAH
LS, an already well-established antihypertensive drug in daily clinical practice and well examined in preclinical and clinical settings of ischemic stroke, shows promising results by attenuating cerebral inflammation and restoring cerebral autoregulation [64,105,[122][123][124][125]. Facing preclinical aSAH research, beneficial effects of Sartans have been shown. Under already physiological conditions, LS diminished cerebral inflammation and associated DCVS [126] as well as ET-1 mediated vasoconstriction. Targeted ET B1 -and ET A -R-antagonism under LS administration revealed a direct modulatory ET B1 -R dependent effect via inducing upregulation of the NO-pathway with a significantly increased relaxation accompanied with enhanced sensitivity of the ET B1 -R [23]. After induction of aSAH, ET-1-induced vasoconstriction was likewise decreased by LS preincubation, abolished after pretreatment with an ET B1 -R antagonist. In precontracted vessels with LS and ET A -R-antagonism, ET-1 induced a higher vasorelaxation compared to the control group without, clearly demonstrating a modulatory and functional restoring effect of LS on the normally after aSAH impaired ET B1 -R function [127].
Beneficial effects of LS on ET-1-and PGF2α-mediated DCVS after aSAH in a rat model have been reported, too [23,127]. An ET-1 mediated vasoconstriction was diminished, and ET B1 -R mediated vasorelaxation under selective ET A -R blockade was restored [126,127]. In addition, PGF2α-elicited vasoconstriction of a basilar artery was markedly diminished [23,126,127]. Interestingly, several work groups could also verify positive vasomodulating effects of LS on the cerebral vessel wall, especially affecting SMCs [128,129]. Furthermore, aneurysm rupture was prevented in mice under LS treatment [129]. As already mentioned, after aSAH, increased synthesis of ET-1 triggers enhanced cerebral vasoconstriction; loss of the ET B1 -R mediated vasorelaxation contributes to this effect, too [127]. Furthermore, upregulated AT 2 -1-R and PGF2α-synthesis contribute in enhancing and maintaining cerebral vasocontraction [7,[130][131][132][133]. LS showed promising aspects in preclinical aSAH studies and therefore might have an effect in the treatment of patients with aSAH.

Discussions
This systematic review demonstrated Sartan administration after ischemic stroke clearly associated with beneficial effects on preclinical models as well regarding clinical trials. Clear evidence of which doses in preclinical and clinical settings for treatment of ischemic stroke with Sartans exactly might be useful are heterogenous and therefore not consistent yet. In a preclinical setting, Sartans significantly reduced infarct volume and edema, augmented CBF, diminished superoxide production, inflammatory processes, and disruption of the BBB. In clinical studies, clear trends towards a better functional outcome and neurocognitive function after stroke with Sartan use have been reported. Thus, the question arises whether Sartans might provide positive effects on DCVS or DCI after aSAH. In summary, LS provided in a preclinical physiological and pathophysiological setup after aSAH beneficial aspects in reducing ET-1-and PGF2α mediated cerebral vasoconstriction [126,130]. Vasoconstriction was notably reduced and the vasorelaxant properties of the ET B1 -R were restored. Furthermore, clear evidence exists, that after aSAH, AT 2 -1-R are upregulated in experimental settings [132]. Here, an additive direct antagonism on these receptors could reduce the sensitivity to an AT 2 -1-R-mediated vasocontraction to angiotensin II, too [125,134]. LS possesses beneficial aspects on cerebral epileptogenicity, which could be applied to the issue of reducing cortical spreading depression post aSAH [135][136][137][138]. Also, it is able to restore post-ischemic cerebral autoregulation after hemorrhagic stroke [134].
Considering these neuroprotective effects of LS, the ethical question arises of whether the philosophy of strictly discontinuing all antihypertensive agents after aSAH (except of new administration of nimodipine), especially of LS, should stay state of the art. Next to beneficial influences on DCVS after aSAH in rats as mentioned above, AT 2 -1-R antagonists clearly possess beneficial effects after stroke regarding cerebral inflammation, the areal of infarction, cortical spreading depression, cerebral microcirculation, and maintenance of pressure-dependent cerebral vasoconstriction [23,64,71,105,127,134,[139][140][141]. Appreciating these facts, a systemic LS administration over and above the phase of DCVS, could be a promising approach in preventing these effects; particularly because LS seems to not influence the global CBF in essential hypertonic patients, which can be set equivalent to a needed-hypertonia after aSAH [142]. Here, LS could be an interesting approach, because it increases global CBF despite lowering blood pressure [105], and is therefore capable to reduce DCI [92]. Also, considering the positive vasomodulatory influences of LS, the question arises whether after aSAH this medication should be established as secondary prophylaxis to avoid a de-novo-aneurysm genesis, ergo, if aneurysms under LS are anyway arising [143].

Translational Aspects
Both abovementioned questions after aSAH are difficult to adapt to the affected patient group, because common sense to date stays in discontinuing all antihypertensive agents after the initial bleeding event. Also, it is vague to postulate that a LS effect persists after discharging this medication on admission over the phase of DCVS for 14 days. Furthermore, the numbers of patients with LS as standard antihypertonic medication receiving follow-up angiographys are too scarce to testify a valid statement concerning case-control studies of aneurysm-growth/-development, as reviewed in our own patient series in 2009-2015. Nevertheless, LS seems to be an underrated neuroprotective drug, reducing cerebral inflammation and epileptogenicity, DCVS, and infarct size after ischemic stroke. These results of preclinical ischemic stroke and aSAH research as well as clinical ischemic stroke research could be applied in a prospective clinical setting of patients suffering aSAH. Also, the question of a de-novo-aneurysm-genesis in further cranial control imaging could be addressed.

Conclusions
There is a promising effect on LS in the treatment of ischemic stroke both in preclinical and clinical studies as well as in preclinical studies on aSAH. LS has shown to reduce ET-1-mediated vasocontraction, cerebral inflammation, and restores vasodilatory function of the ET B1 -R [26][27][28]. Thus, LS may decrease the incidence of symptomatic vasospasm and improve functional outcome in aSAH patients. Large, randomized, double-blinded clinical trials are necessary to determine its benefit in aSAH.