NA-1 is a peptide that interferes with the binding of postsynaptic density protein 95 to the NMDAR and to neuronal NO synthase. This dissociates NMDAR from downstream neurotoxic cascades, which include excessive neuronal calcium influx and neurotoxic NO production. The ESCAPE-NA1 phase III trial investigated the effect of a single IV dose of NA-1 before thrombectomy in patients with ischaemic stroke, with moderate-to-good filling of collateral vessels on computed tomography (CT) angiography and no evidence of an extensive region of early infarction on CT brain. The ESCAPE-NA1 trial [8] showed that NA-1 did not improve the proportion of overall stroke patients achieving good clinical outcomes [modified Rankin score (mRS) 0–2 at 90 days] or incidence of serious adverse events compared with placebo [adjusted risk ratio (RR) 1.04, 95% confidence interval (CI) 0.96–1.14] [8]. However, in a pre-defined subgroup of patients who did not receive IV thrombolysis, NA-1 significantly increased favorable outcomes compared to placebo (59.3% vs. 49.8%). The study proposed that this could be due to enzymatic cleavage of NA-1 by plasmin, a serine protease generated from plasminogen by tissue-plasminogen activators such as alteplase, thus leading to subtherapeutic concentrations of NA-1. This hypothesis was supported by pharmacokinetic data from a subset of trial participants [80]. Therefore, further investigation is needed to identify subgroups that may potentially benefit from NA-1. Currently, there is an ongoing trial (ESCAPE-NEXT, NCT04462536) investigating the efficacy of NA-1 in patients who do not receive thrombolysis.

Memantine antagonizes overstimulated NMDAR, and preclinical studies had proposed that memantine improves outcomes through increased brain-derived neurotrophic factor signalling, reduced reactive astrogliosis, and improved vascularization [81]. In a recent human trial involving 77 participants with mild to moderate AIS (NIHSS score 16 and below), patients were given 20 mg memantine every 8 h for five days and then 20 mg daily for three months [11]. Memantine significantly improved the neurologic function of the patients, with NIHSS change after 5 days being significantly higher in the intervention group compared to the control group (−2.96 ± 0.10 in the intervention group and −1.24 ± 0.96 in the control group P < 0.0001). Memantine also significantly increased the Barthel Index (BI) scores after 3 months (6.00 ± 2.62 in the intervention group and 3.96 ± 1.76 in the control group P = 0.002) [11]. Although there is preliminary evidence that memantine may be beneficial, further larger trials are needed to better characterize its effects on AIS patients.

Novel small molecules that bind and inhibit the NMDAR and NMDA glycine site antagonists generally showed no improvement in stroke outcomes in human RCTs. This included aptiganel hydrochloride [12], AR-R15896AR [13, 14], selfotel [15], and lubeluzole [16]. The glycine site antagonists, gavestinel, GV150526, and licostinel, showed no significant differences in serious side-effects, outcome, or mortality [17–20]. Neu2000, a sulfasalazine derivative, selectively blocks NMDARs and acts as a free radical scavenger. The ongoing SONIC phase-II clinical trial has randomized AIS patients who underwent EVT into control, low-dose (2.75 g), and high-dose (5.25 g) Neu2000KWL over 5 days, and the results are yet to be published [21].

Excessive activation of AMPAR induces the postsynaptic depolarisation of glutamate and is a marker of excitotoxicity-related diseases like AIS [82]. However, in a phase II trial, AMPA antagonist ZK 200775 was found to reversibly worsen the neurological condition in patients with AIS. Furthermore, ZK 200775 exerted significant sedative effects in patients which precluded its further development as a neuroprotective agent [22].

Edaravone is an amphiphilic low-molecular-weight drug that scavenges both lipid- and water-soluble peroxyl radicals, thus preventing lipid oxidation in animal models [84]. It also inhibits oxidation, increases endothelial NO synthase (eNOS), enhances NO production, and decreases neuronal NO synthase (nNOS) and inducible NO synthase (iNOS), which are detrimental NO synthase (NOS). This may improve cerebral blood flow without peroxynitrite generation during reperfusion [85]. An RCT by Zheng and Chen [27], 2016 that recruited 60 diabetic stroke patients showed that mean NIHSS scores were lower (60%), and BI scores were 1.7-fold higher in edaravone-treated patients compared to control on day 14. Another RCT by Sharma et al. [28] 2011 involving 50 patients also found that edaravone effectively improved functional outcomes (mRS ≤ 2) in patients after AIS. This was also supported by a few retrospective observational studies [29, 30], which reported that the use of edaravone is associated with better functional independence after discharge (adjusted OR, 1.21; 95% CI 1.03–1.41) and lower hospital mortality (adjusted OR, 0.52; 95% CI 0.43–0.62). Although edaravone was the first drug to be approved for clinical treatment of AIS in Japan in 2009 [31], its benefit requires further validation in larger studies.

An RCT showed that edaravone was not inferior to ozagrel sodium (thromboxane A2 synthase inhibitor) for the treatment of acute noncardioembolic ischemic stroke in 401 patients [31], and edaravone could be more cost-effective than ozagrel [86]. Novel neurotherapeutic agents with a combination of edaravone and other active ingredients have also been developed. For instance, edaravone dexborneol, which comprises 2 active ingredients edaravone and (+)-borneol, has shown synergistic effects of antioxidant and anti-inflammatory in animal models. In a phase III RCT of 1,165 patients, the edaravone dexborneol group showed a significantly higher proportion of patients experiencing good functional outcomes on day 90 after randomization, compared with the edaravone group (mRS ≤ 1, OR 1.42, 95% CI 1.12–1.81) [32].

NXY-059 is a nitrone-based IV drug which has a free radical trapping action in animal models [87]. Although NXY-059 was consistently reported to be well-tolerated in AIS patients, NXY-059 did not improve mortality in several RCTs [33–35]. While the administration of NXY-059 within 6 h after the onset of AIS showed a reduction in terms of disability at 90 days, it did not show benefits in reducing the rates of severe and non-severe adverse events including neurologic status in a trial of 1,722 patients [35]. In the SAINT I and II RCTs, where NXY-059 was also administered within 6 h of stroke onset, an intention-to-treat analysis based on 5,028 patients from these trials confirmed an absence of benefit by comparison at each level of the mRS, as well as no difference in mortality in the treatment and control groups [33].

Citicoline may reduce free radical generation through a reversible synthetic pathway leading from choline to phosphatidylcholine, which reduces the breakdown from phosphatidylcholine into free fatty acids that are used to generate free radicals in animal models [36]. In AIS, a few RCTs reported that the administration of citicoline is associated with a significant decrease in infarct volume compared to placebo [37, 38]. In addition to infarct volume, oral administration of citicoline (500 mg/day) for 6 weeks improved NIHSS by seven or more points at week 12 [38]. In a systematic review by Dávalos et al. [39] 2002, oral citicoline within the first 24 h of onset in moderate and severe stroke could increase the chance of full recovery at 3 months (25.2% in citicoline, 20.2% in placebo, OR 1.33; 95% CI 1.10–1.62). The dose showing the largest difference with placebo was 2,000 mg, with 27.9% of patients achieving recovery. A randomised controlled study by Kuryata et al. [40] 2021 found that additional citicoline treatment could result in a decrease in the levels of glial fibrillary acidic protein (GFAP), neurofilament light subunit (NF-L) and myelin basic protein (MBP) compared to the control group. As such, citicoline could have a therapeutic effect on patients with AIS by protecting them from axonal injury, lowering levels of astrogliosis, decreasing demyelination, and protecting the function of microglia.

Natalizumab is a monoclonal immunoglobulin G4 (IgG4) antibody that binds to integrin to inhibit immune cell migration, and the main application it is approved for is the treatment of relapsing-remitting multiple sclerosis [88]. In the multicentre ACTION II trial on AIS patients with NIHSS scores of 5–23, one dose of 300 mg or 600 mg IV natalizumab was given to patients [43]. However, results showed that the composite of excellent outcome (mRS ≤ 1, BI sore ≥ 95 at 90 days) was less likely with natalizumab than with placebo (OR 0.6, 95% CI 0.39–0.93), with no significant difference in mortality or adverse events [43].

MAPC are stem cells derived from the bone marrow with immunomodulatory properties that are anti-inflammatory and has shown benefits in animal models of AIS [89]. In the phase II MASTERS RCT, 129 patients with moderately severe AIS (NIHSS 8–20) were randomised to IV MAPC or placebo between 12–48 h after symptom onset [44]. There were no differences in adverse events and global stroke recovery at 90 days, although post-hoc analysis showed efficacy in the earlier time window of 24 to 36 h. An analysis of the subgroup with stroke onset < 36 h showed the MAPC group had a higher chance of full or nearly full recovery at 1 year, and further trials evaluating the efficacy of MAPC in the earlier time window (< 36 h) are planned.

Minocycline is a tetracycline that inhibits microglial activation and causes decreased production of inflammatory cytokines including tumor necrosis factor-alpha (TNF-α) and interleukin-1 beta (IL-1β). In addition, minocycline decreases the production of inflammatory mediators like cyclooxygenase-2 (COX-2) and iNOS by inhibiting NF-κB activity. Furthermore, through decreasing T-cell proliferation, fewer inflammatory cytokines including TNF-α and interferon-γ are produced in animal studies [45]. A systematic review on minocycline reviewed 6 small, randomised trials in AIS patients, which have not demonstrated a clear neurological benefit but also were not sufficiently powered to draw definitive conclusions [45]. In another systematic review that included 7 RCTs on 426 patients, minocycline was associated with higher rates of 3-month functional independence (mRS-0-2) and BI in AIS patients [46]. The doses used ranged from 200 mg to 400 mg daily and were administered 6–24 h from the onset of symptoms. However, these trials were open-label, and larger double-blinded randomised trials are needed to confirm these results.

Imatinib is an anti-cancer drug that inhibits tyrosine kinase and specifically targets several proteins such as tyrosine kinases, Abelson (Abl)-related gene, c-Kit, platelet-derived growth factor receptor (PDGF-R), and breakpoint cluster region-Abl gene (BCR-ABL). It has been shown to have anti-inflammatory activity in reducing pro-inflammatory mediators such as TNF-α and IL-6. In a phase II RCT in 60 patients with AIS treated with thrombolysis, imatinib improved neurological outcomes by 0.6 NIHSS points per 100 mg of imatinib [47], and a larger RCT is ongoing to confirm these findings (NCT03639922).

According to Sabetghadam et al. [48] 2020, patients treated with NAC showed significantly lower mean NIHSS scores at day 90 compared to placebo. NAC treatment also reduced the blood levels of proinflammatory factors including IL-6, soluble intercellular cell adhesion molecule-1 (sICAM-1), NO, malondialdehyde (MDA), and neuron-specific enolase (NSE) and significantly increased serum levels of antioxidant biomarkers such as superoxide dismutase (SOD), glutathione peroxidase (GPx), and total thiol groups (TTG). Thus, NAC has good antioxidant and anti-inflammatory effects on patients after a stroke.

Tissue injury and stress activate endogenous mechanisms to restore homeostatic balance and prevent further damage by upregulating the synthesis of lipid signalling molecules, including palmitoylethanolamide (PEA), which has positive neuroprotective effects and reduces inflammatory events due to ischemia reperfusion injury. In an observational study, 250 stroke patients undergoing neurorehabilitation were treated for 60 days with a pharmaceutical preparation of co-ultraPEALut (Glialia), a combination of PEA, and the antioxidant flavonoid luteolin [49]. Patients showed significant improvement in neurological status, impairment of cognitive abilities, the degree of spasticity, pain, and independence in daily living activities after 30 days of treatment compared to baseline. Further RCTs are needed to confirm the neuroprotective efficacy of co-ultraPEALut in AIS.

Butylphthalide activates the Keap1-Nrf2/ARE signalling pathway which is important in maintaining the redox equilibrium state in neurons. Butylphthalide can also potentially improve the cerebrovascular reserve of patients, and enhance the establishment of collateral compensatory vessels. A retrospective study by Zhang et al. [50] showed that butylphthalide injection for 2 weeks in AIS patients improved patients’ neurological function and activities of daily living. Compared to routine treatment, the treatment group had a superior overall response rate (ORR), showed a significantly higher BI score, lower NIHSS score, higher cerebrovascular vascular reserve function, and lower pulsation index (all P < 0.05). Additionally, in the observational group, the serum level of Keap 1 was remarkably higher (P < 0.05), while the serum levels of nicotinamide adenine dinucleotide phosphate [NAD(P)H] quinone dehydrogenase 1 (NQO1), Nrf2, and ARE were evidently compared to the control (P < 0.05). As no significant adverse reactions were observed, butylphthalide may be promising as an adjuvant therapy of AIS, and there is an RCT on the efficacy of butylphthalide in AIS which is ongoing [79].

HUK is a glycoprotein that treats stroke by positive regulation of the kallikrein-kinin, which through the kinin B2 receptor acts as an antioxidant and anti-inflammatory agent in protection against stroke [90]. RESK is a phase IV study which demonstrated that the single IV dose of HUK over 50 min every day for 14–21 continuous days had a favorable safety profile, and improvements in neurological and functional outcomes were observed in 1,202 AIS patients from baseline [51]. However, as this was not placebo-controlled, its efficacy compared to conventional treatment requires further study.

Melatonin reduces oxidative stress, neuroinflammation, and apoptosis induced by an ischemic stroke. A systematic review of 6 RCTs on melatonin showed it was generally safe, but no benefits were reported regarding the quality of life despite efficacy in animal models [52]. However, the doses used in animal studies were substantially higher than those administered in the human RCTs, and this may have contributed to the lack of efficacy seen.

Uric acid is an antioxidant in animal models, and its mechanism for neuroprotection in AIS is multifaceted [91]. In an animal study, the neuroprotective property of uric acid was associated with IL-6/signal transducers and activators of transcription 3 (STAT3) signaling pathway activation, attenuation of edematogenic vascular endothelial growth factor A (VEGF-A)/MMP-9 signaling, and modulation of oxidative stress, neuroinflammation, and apoptosis [92]. In an RCT involving 411 participants by Chamorro et al. [53], it was found that the addition of uric acid to thrombolytic therapy did not increase the proportion of patients who achieved excellent outcomes (mRS 0–1) after stroke compared with placebo.

Albumin has antioxidant properties and modulates intracellular signalling of neuronal or glial cells to exert neuroprotective effects [93]. In the ALIAS RCTs I and II, treatment with 25% IV albumin did not show overall clinical benefit in AIS patients [54]. However, when albumin was initiated within 2 h, AIS patients with large cardioembolic stroke (NIHSS ≥ 15) showed a trend for benefit, suggesting that certain AIS patient subgroups most representative of preclinical settings may benefit.

Epigallocatechin gallate (EGCG) is a natural polyphenol that exerts antioxidant and anti-inflammatory effects. EGCG also induces the proliferation and differentiation of neural progenitor cells and promotes neurogenesis after AIS. In a study involving adult rats, after inducing middle cerebral artery occlusion injury, it was shown that EGCG exerted a neuroprotective effect by regulating caspase-3 and poly(adenosine diphosphate-ribose) polymerases (PARP) proteins during cerebral ischemia [94]. Another animal study confirmed that EGCG inhibited autophagy by modulating the protein kinase B (AKT)/5’adenosine monophosphate-activated protein kinase (AMPK)/mammalian target of rapamycin (mTOR) phosphorylation pathway in both in vivo and in vitro models of stroke [95].

Glibenclamide is a sulfonylurea that inhibits the sulfonylurea receptor 1-transient receptor potential melastatin 4 (Sur1-Trpm4) channel and in some cases microglial Sur1-Kir6.2 channels. Several preclinical studies have found glibenclamide to be an effective treatment in rodent models of AIS, and retrospective cohort studies suggest that being on sulfonylureas and staying on it following brain ischemia may improve outcomes [56]. Prospective RCTs are needed to test the efficacy of glibenclamide in AIS. Glyburide is another anti-diabetic drug under the sulfonylurea class. The GAMES pilot study involving a small sample size of 10 administered a 3 mg dose of RP-1127 (Glyburide) to patients with severe stroke at high risk for swelling [57]. The GAMES-RP phase II RCT on 86 participants showed no significant difference in the primary efficacy outcome [58]. An exploratory analysis based on this clinical trial showed that IV glyburide had a significant effect on the BI (P = 0.03), and treated participants had lower plasma levels of MMP-9 (189 ng/mL vs. 376 ng/mL; P < 0.001) and decreased midline shift (4.7 mm vs. 9 mm; P < 0.001) and improved survival in participants ≤ 70 years of age with large hemispheric infarction.

Statins are lipid lowering agents that upregulate endothelial NOS production, which increases cerebral blood flow, exerts anti-inflammatory effects, and may promote synaptogenesis and neurogenesis after AIS. According to the American Heart Association guidelines class IIa recommendations, among patients already taking statins at the time of AIS, a continuation of therapy during the acute period is reasonable [1]. However, the initiation of early statin therapy for its neuroprotective effect is still a topic of contention. In an RCT of 60 patients with cortical strokes, simvastatin within 3–12 h of symptom onset at a dose of 40 mg/day for one week followed by 20 mg/day until day 90 improved the neurological outcomes significantly by the third day, but there was a non-significant increase in mortality and a greater proportion of infections in the simvastatin group [59]. However, the ASSORT Trial involving 257 AIS and dyslipidemia patients did not show any superiority of early statin therapy within 24 h of admission compared with delayed statin therapy 7 days after admission on the degree of disability at 90 days after onset [60]. In this trial, possible reasons why no improvement in functional outcome was observed include the fact that patients included in the trial had relatively mild strokes with a median NIHSS of 3 and received less-intense statin regimens (atorvastatin 20 mg/day, pitavastatin 4 mg/day and rosuvastatin 5 mg/day). Future trials involving patients with more severe strokes and higher-intensity statins—atorvastatin 80 mg/day or rosuvastatin 20 mg/day or 40 mg/day can be considered. A meta-analysis of 22 observational studies on 17,554 patients showed that in-hospital statin use after IV thrombolysis was associated with a lower likelihood of symptomatic ICH, 3-month mortality, and 3-month favorable functional outcome (mRS score 0–1) but not pre-stroke statin use [61].

Cerebrolysin is a mixture of low-molecular-weight peptides and amino acids derived from the porcine brain that has potential neuroprotective properties and neurotrophic activity. A meta-analysis of 6 RCTs showed that cerebrolysin had no significant effect on 90-day functional recovery on mRS, NIHSS, and BI scales [62]. Cerebrolysin probably has little or no beneficial effect on preventing all-cause death or reducing disability in AIS, and current evidence does not support its routine use.

Epo is a hormone that acts by binding to the Epo receptor (EpoR), and reduces BBB disruption after cerebral ischemia through reducing lipid peroxidation, downregulating the vascular endothelial growth factor receptor-2 (VEGFR-2) expression in the penumbra region, and decreasing MMP-2 and MMP-9 activity in ischemic microvessels, thus reducing apoptosis. Epo also regulates cell volume and neurogenesis in areas that suffered ischemic injury and reperfusion [97]. However, in the German multicentre Epo stroke RCT on 522 patients with middle cerebral artery stroke, IV human recombinant Epo showed no beneficial effects of Epo and increased overall death rate [67].

3K3A-activated protein C is a blood protease with anticoagulant and cell-signalling properties which mechanism of action is by the activation of protease-activated receptors (PAR) 1 and 3 via non-canonical cleavage [98]. 3K3A-activated protein C has been shown to stimulate neuronal production by the human neural stem and progenitor cells in vitro [99]. The RHAPSODY trial was a phase II trial tested the use of 3K3A-activated protein C in combination with TPA or mechanical thrombectomy in AIS, which found toxicity in 7% but the possible reduction in hemorrhagic rate [70].

Other compounds that have been studied for neuroprotection include DDFPe, otaplimastat, dapsone, and theophylline. DDFPe emulsion was reported to be highly efficient at oxygen and carbon dioxide transportation in vivo. A phase Ib/II RCT involving 24 patients showed that IV DDFPe appears to be safe at all doses tested, and clinical improvements in NIHSS score and mRS were significant but compromised by a small sample size [68]. Otaplimastat inhibits MMP in animal stroke models through the upregulation of metalloproteinase-1. In the SAFE-TPA RCT, IV otaplimastat as an adjunct therapy in AIS patients receiving recombinant TPA (rTPA) was shown to be feasible and safe [69], and its functional efficacy needs to be investigated in large trials. Dapsone inhibits apoptosis, and a recent animal study suggests that dapsone represses proapoptotic proteins c-Jun N-terminal kinases (JNK), phosphatase and tensin homolog deleted on chromosome 10 (PTEN), calpain, and caspase-3 of cerebral ischemia and activates pro-survival protein brain-derived neurotrophic factor (BDNF) [100]. A pilot clinical trial conducted on 30 patients with AIS involving the middle cerebral artery territory showed that one dose of dapsone 200 mg compared to a placebo achieved improvement in NIHSS, BI, and mRS, with no adverse reactions reported. However, a larger trial is needed to study the treatment effects of dapsone [71].

Some neuroprotective agents exert neuroprotective effects by improving cerebral blood flow to ischemic tissues. These include piracetam, theophylline, and GTN, as discussed below.

Piracetam is a nootropic reported to increase cerebral blood flow in both infarcted and penumbral tissues. Although there is limited evidence regarding its neuroprotective mechanism, it has been suggested that its ability to localize polar heads in the phospholipid bilayer restores membrane fluidity and facilitates membrane-bound cell functions such as ATP production, thus preventing neuronal cell death [101]. A systematic review in 2012 on piracetam that included 3 RCTs showed that there is some suggestion (but no statistically significant result) of an unfavorable effect of piracetam on early death, but this may have been caused by baseline differences in stroke severity in the trials [63]. There is no definite evidence of a beneficial or harmful effect on death in AIS, and the routine use of piracetam in the management of AIS is not recommended.

Theophylline, derived from methylxanthine, works via smooth muscle relaxation, bronchial dilation, and CNS stimulant activities [64]. In the phase II clinical study done by Modrau et al. [64] 2020, theophylline was used as an additional thrombolytic therapy. It showed improvement in terms of NIHSS. However, major clinical improvement defined as ≥ 50% improvement in NIHSS score from baseline to 24 h follow-up, and infarct growth was not significantly different between the intervention group and the control group.

NO donors, like GTN, have vasodilatory actions and may reduce elevated blood pressure in acute stroke. A patient-level meta-analysis of 5 RCTs on 4,197 patients with acute or subacute stroke found that GTN lowered the blood pressure by 7.4/3.3 mmHg (0.99/0.44 kPa) but did not alter 90-day outcomes [65]. In a subgroup of patients who received GTN within 6 h of stroke onset, mRS improved (OR 0.52, 95% CI 0.34–0.78) and mortality was reduced (OR 0.32, 95% CI 0.14–0.78) compared to placebo [65]. However, the MR ASAP phase III RCT which tested transdermal GTN within 3 h of stroke onset vs. standard care was prematurely terminated due to safety concerns in patients with intracerebral hemorrhage, and no difference in death within 90 days or serious adverse events were observed [66].