ESCs and iPSCs have a powerful renewal potential and can be manipulated to differentiate into phenotypes from all three embryonic layers [34]. Human iPSCs (hiPSCs) have been used for therapeutic solutions regarding central nervous system (CNS) disorders [35], although their use for ischemic brain injury has remained limited. This cell type has been limited in translational use due to a potential risk for insertional mutagenesis or oncogenesis, production of immune-tolerable cells, and a poor integration into host circuits [34,36].

Endogenous neural stem cells (NSCs) are primarily located in the sub-ventricular zone and dentate gyrus. Exogenous NSCs are extracted from three main sources for therapeutic purposes: direct extraction from brain tissue (particularly paraventricular or olfactory bulb regions), differentiation from pluripotent stem cells, and trans-differentiation from somatic cells, such as skin fibroblasts, renal tubular cells in urine, and blood cells [37]. Initially, human NSCs held constraints: limited donor availability, low proliferation rate, and poor long-term in vitro expansion. It has recently been possible to culture them over several passages and successfully differentiate them into neuronal and glial cell types [38].

Exogenous NSCs can migrate into ischemic brain areas, where they can differentiate into neurons or astrocytes and promote endogenous NSC proliferation and differentiation [39-41]. Administration of human NSCs in a stroke rat model revealed neuroprotective effects by increasing dendritic branching, enhancing corticospinal tract projections, and inhibiting inflammation [42,43]. Additionally, an in vitro and in vivo study has suggested that improved functional outcomes in stroke rats treated with transplanted NSCs may be associated with angiogenesis stimulation and brain microvasculature formation, due to the increased expression of proangiogenic factors in NSCs [44].

A small phase I translational approach demonstrated that ipsilateral implantation of CTX0E3 human NSCs into the putamen of chronic stroke patients was safe, with no adverse events related to stem cell administration after 2 years of follow-up, and was associated with a slight improvement in the National Institutes of Health Stroke Scale (NIHSS) [45]. Notably, adverse events related to the invasive surgery required for therapy were reported.

Despite the promising utility of NSCs, some limitations exist. A very high cell dose is required for transplantation. Tissue-derived NSCs can clot in vivo, if transplanted at a high density, and may be prone to a poor survival rate [37]. Further, our understanding remains limited about the molecular mechanisms regulating endogenous NSCs for neurogenesis and methods for enhancing endogenous neurogenesis in ischemic brain injury patients [46]. More work is needed to optimize NSC application conditions prior to clinical practice.

Mesenchymal stem cells (MSCs) can be derived from a variety of sources, but are primarily harvested from the adipose tissue (AD-MSCs) or bone marrow (BM-MSCs) [47]. BM-MSC harvesting requires an invasive procedure and results in relatively low stem cell yields, whereas large quantities of AD-MSCs, which are easily accessible, abundant, and have little ethical concerns regarding their use, can be harvested using less-invasive liposuction techniques [20,48]. MSCs are the most studied stem cell in stroke treatment, providing robust information for guiding stem cell application.

Preclinical studies demonstrate that transplantation of MSCs in an early stage of stroke improves neurological functional outcomes by regulating microglia phenotypes and stimulating regenerative processes [49,50]. In cardiac arrest-induced brain injury, MSC administration, together with therapeutic hypothermia, can markedly enhance neuroprotection and reduce hippocampal neuronal death [51]. Unfortunately, MSCs are prone to accumulation and aggregation in other organs, particularly the lungs (up to 80% of cells), after intravenous (IV) administration, which may result in embolism or hemorrhage [52,53].

Human umbilical cord blood stem cells (UC-BSCs) are derived from placental tissues, following birth. They consist of hematopoietic stem cells and MSCs. These cells offer a number of key advantages, such as an ample source of cells, low donor age, and low risks to babies and mothers during harvesting, which minimizes ethical concerns. These cells can differentiate into neural progenitor cells and provide neuroprotective effects in cerebral ischemia via neurotrophic factor secretion and vascular remodeling enhancement after stroke [62,63]. UC-BSCs have protective effects against ischemic injury, resulting in brainderived neurotrophic factor expression recovery [64]. In addition, UC-BSCs can inhibit the immune response and decrease the size of the ischemic brain lesion [65]. These results indicate their potential in ischemic brain injury.

However, these cells have a key disadvantage: they do not cross the blood-brain barrier (BBB) [66]. Interestingly, it was also noted that patients who underwent UC-BSCs transplantation were more prone to CNS infections [67].

IV delivery circumvents the need to access the CNS directly, thus allowing for less invasiveness [60,68]. Intravenously injected NSCs can traverse the BBB and localize to ischemic injury areas [39]. NSCs, BM-MSCs, AD-MSCs, or UC-BSCs IV administration has shown neuroprotection in ischemic stroke [39,48,60,64,65,69].

Despite these results, stem cells delivered by the IV route are subjected to trapping in peripheral tissues (lungs, liver, spleen, and kidneys), particularly the lungs [70], limiting cell delivery to the brain. Additionally, pulmonary embolism complications were noted in a Korean family following IV AD-MSCs administration [71]. Cell size remains a major limiting factor in the intravenously delivered stem cell ability to traverse microvasculature or capillaries and avoid peripheral aggregation. Smaller cells, such as BM-MSCs (7 μm diameter), display up to a 30-fold greater passing rate through the lungs compared to larger stem cell types, such as AD-MSCs (18 μm) and NSCs (16 μm) [70,72]. Overall, it has been demonstrated that only a small number of cells reach the arterial system and target tissues, and that cells that get trapped in lungs can cause tissue damage [73]. In another study, IV BM-MSCs improved neurological function and restored BBB following damage from intracerebral hemorrhage [74]. However, BM-MSCs preferentially migrated to the spleen, being detected at a modest level at 12 hours and 11 days following treatment [75]. IV delivery of BM-MSCs in chronic stroke patients demonstrated an adequate safety profile [76]. UC-BSCs tail vein injection also reduced infarct area and restored bladder function that had been impaired by cerebral ischemia in a rat stroke model [69,77].

Intra-arterial (IA) delivery uses catheterization to guide stem cells into the carotid artery or circle of Willis, thus bypassing the initial uptake by systemic organs. IA administration has been shown to be advantageous for MSC delivery at the site of injury [78-80]. This delivery method allows for bypass of the lungs and avoids pulmonary entrapment [72]. While this method delivers a large number of cells to brain lesions, it also carries a risk of micro-embolism and cerebral ischemia due to arterial cell aggregation and clumping, resulting in reduced cerebral blood flow [81-83]. In clinical applications, injection of BM-MSCs, both IA (internal carotid artery) and IV, demonstrated safety and showed improved functional scores at 4 to 8 weeks post-therapy in chronic stroke patients [76].

Intraventricular delivery allows for direct delivery into the cerebral spinal fluid (CSF) or interstitial tissue, thereby allowing the dispersal to multiple sites of the brain and CNS through the CSF. Despite the invasiveness of this technique, complications can be reduced with the proper training and technique. Intraventricularly injected BM-MSCs were transplanted into the lateral ventricles of cuprizone mice models, thereby initiating functional remyelination, oligodendrogenesis, and endogenous NSC proliferation through paracrine effects [84]. Intrathecally-administered UC-BSCs could migrate to the ischemic area and differentiate into neurons and astrocytes, which improved motor function in stroke rats [85]. Clinical applications of this method have supported the safety and potential efficacy, with improved neurological outcomes after intraventricular transplantation of BM-MSCs into patients with chronic sequelae of hemorrhagic stroke [86,87]. In cardiac arrest-induced global brain ischemia, human NSC infusion into the lateral ventricle 3 hours following resuscitation significantly improved neurological and electrophysiological outcomes of rats [88].

Intraperitoneal (IP) administration of human MSCs (hMSCs) led to the rapid aggregation of cells in the peritoneal cavity, and only a small amount of cells migrated elsewhere [89]. When compared to hMSC IV injection, IP administration resulted in a lower accumulation in peripheral tissues, such as the lungs and liver. However, IP-delivered umbilical cord-derived mesenchymal stromal cells (UC-MSCs) showed poor localization to the ischemic frontal cortex [90], suggesting limited brain penetration.

Localized intraparenchymal delivery allows for a higher cellular volume to reach brain target lesions [91], thereby preventing offtargeting effects and improving cost-effectiveness. Intracranial transplantation of human AD-MSCs was shown to promote neuronal repair in cerebral ischemia-reperfusion injury [54]. MSC intracranial injection has been used for neonatal hypoxic-ischemic brain injury [92]. iPSC subdural injection upon infarcted brain tissue can reduce the infarct area size and improve motor function after ischemic stroke [93]. Additionally, iPSC-derived NSC intracerebral transplantation was shown to differentiate into neurons and oligodendrocytes and improve recovery in a pig stroke model [94]. UC-BSCs injected into the contralateral cortex could also migrate to the ischemic region of stroke rats [95]. Despite these accomplishments, it remains controversial whether intracranial injections are feasible for clinical therapy [54,92,96]. Direct injection is invasive, as it requires a craniotomy, and this method can cause localized trauma and BBB damage [97]. The required surgery for this delivery method can potentially disrupt local niches and result in additional traumatic injury and inflammation. Intraparenchymal delivery is also ineffective at addressing multifocal disorders; thus, its feasibility for clinical translation is still debatable [96].

Intranasal administration is a noninvasive and relatively effective drug or cell delivery route when compared to invasive intracranial transplantation. Although it has been recognized as a small agent (e.g., small molecules, proteins, viruses, bacteria, and nanoparticles) delivery method, its utility for cellular delivery has only been discovered in the last decade [98]. Intranasallydelivered cells collect beneath the nasal mucosa, close to the turbinate bones, and travel through the cribriform plate [99]. This method of delivery is non-invasive, rapid, and offers BBB bypass [100]. Cells showed improved targeting and less accumulation in peripheral organs, compared to systemic dosage [101]. Additionally, intranasal delivery allows for repeated dosages [102-105], making it more practical than other methods.

Both MSCs and NSCs have been shown to be effective agents in intranasal cell therapy [102,106]. Intranasal delivery of hypoxiapreconditioned BM-MSCs in a rat stroke model demonstrated significantly reduced cell death in peri-infarct regions [107]. Intranasal MSC treatment of neonatal hypoxic ischemic-injured mice has demonstrated long-term cognitive and sensorimotor improvement [105]. Furthermore, a single intranasal dose was sufficient for therapeutic benefits, while at least two intracranial dosages were required to achieve the same effect. Intranasallyadministered NSCs to HIE neonatal rats showed substantial migration into the brain 24 hours post-delivery and improved functional neurological outcomes [108]. In stroke, BM-MSC intranasal delivery provided neuroprotection, enhanced neurovascular regeneration, and led to improved functional recovery [63,103,109].

It is important to recognize that the clinical translation of this method remains poorly understood. Humans have a smaller olfactory bulb compared to that in the rodents utilized in these preclinical studies, and it remains unclear if this difference can hinder the translational potential of intranasal administration. Of note, a hypoxic-ischemic brain injury neonatal primate model demonstrated migration of intranasally delivered MSCs to the site of injury, suggesting a feasible translation [96].

Overall (Table 1) [42,45,51,77,78,81,82,84,90,91,100,102-105,110], IV, intracerebral, and IA routes have been the most utilized and investigated for stem cell administration in preclinical and clinical studies. IV administration facilitates widespread implementation and avoids invasive procedures, but reduces brain tissue penetration due to cell diffusion and risks accumulation in other organs such as the liver, spleen, and lungs [111]. Intracerebral injection has high therapeutic benefits, but with invasive risk. Conversely, the IA route has better CNS penetration and bypasses other organ systems, but has associated arterial micro-embolism and cell-aggregation risks [111]. Intranasal delivery, a minimally invasive novel delivery mechanism, may be an optimal route in the future, but technical aspects, such as administration technique, cell type, position of head during administration, and volume of dosage should all be considered, requiring further investigation in clinical application.

Two essential parameters that can be optimized in the use of stem cells in ischemic brain injury include the number of delivered cells and the timing of delivery. A density ranging from 5×103 to 2×10⁷ MSCs has been studied in animals, typically intravenously, immediately following ischemic stroke or over a period typically varying from 24 hours to 30 days [111]. Preclinical studies indicate that IV doses ranging from 3.6×10⁴ to 4.3×10⁷ MSCs/kg body weight provide substantial behavioral gains in rodents [117]. An additional study noted a significant lesion size reduction and sensorimotor function improvement when 0.5×10⁶ MSCs were delivered per mouse in a neonatal hypoxicischemic model. They noted that decreasing the dose to 0.25×10⁶ MSCs did not have a significant effect on sensorimotor performance or gray/white matter preservation, while increasing the dose to 1.0×10⁶ did not result in additional improvements to sensorimotor function or lesion size reduction [105]. Translationally, IV transfusion of 0.5, 1.0, and 1.5×10⁶ allogeneic BM-MSCs/kg body weight was shown to be safe in a phase I trial and suggested behavior gains in chronic stroke patients [9]. UC-BSCs (2×10⁶ cells per rat) showed safety and neuroprotection in stroke rats after IV administration 2 days following ischemia induction [77]. Additionally, IV injection of UCBSCs (1×10⁶ or 4×10⁶), 1 day after ischemic injury, has been shown to reduce neurological injury in stroke rats [118].

Regarding timing, MSCs administered 3 or 10 days postneonatal hypoxic ischemic injury showed significant improvements in motor behavior, while administration at 17 days showed neither sensorimotor function improvement nor gray/white matter loss reduction. This suggests a broader treatment window for intranasal MSC delivery, particularly when compared to the current treatment window of 6 hours for therapeutic hypothermia, which is the current standard of care [105]. In another study, hypoxic-preconditioned BM-MSCs were intranasally delivered 3, 4, 5, and 6 days after ischemic stroke in mice. These cells reached the brain 6 hours after delivery, where they increased neurogenesis, angiogenesis, and local cerebral blood flow when measured 21 days after stroke. Functional recovery was improved in these animals. This study highlighted that repeated administration of stem cells 3 to 6 days after stroke could be beneficial, thus extending the therapeutic window for stroke beyond the current 24-hour limitations [103]. BM-MSCs administration between 2 and 7 days of injury significantly increased health benefits compared to 12 to 24 hours and >7 days [119]. Delivery at 7 days post-stroke has been most commonly reported in the literature, which may be related to the specific brain environment at that time, with decreased glutamate and brain edema [120]. After stroke, the brain microenvironment is constantly changing, with increased excitotoxicity, inflammatory reaction, and trophic factor expressions [120]. Immediately following ischemic insult, ischemia-reperfusion events occur, including cellular energetic failure, mitochondrial Ca²⁺ overload, oxidative stress, BBB dysfunction, inflammation, apoptosis, and eventual neuron and glial cell necrosis [121,122], which may block or interfere with the transplanted stem cells’ neuroprotective effects. Thus, the pro-regeneration environment in the ischemic area around the 7-day time point may be considered as the optimal time point for stem cell delivery. Compared with ischemic stroke, brain injury induced secondary to cardiac arrest has different conditions related to stem cell treatment (Table 2) [10,16-18,51,123-127]. To date, BM- and AD-MSCs are the most commonly used stem cells in this pathology, and the cells are typically administered 1 to 2 hours after the return of spontaneous circulation. After 1 hour of brain ischemia, the subdural transplantation of iPSCs (1.0×10⁶) is found to improve recovery in stroke rats [93]. Recently, our lab reported that intraventricularly administered hNSCs 3 hours after 8-minute asphyxia-cardiac arrest significantly improved neurological outcomes [88]. However, an optimal stem cell delivery timing has not been fully elucidated regarding this pathology, and additional studies are needed to provide further evidence for optimal time point of stem cell delivery and timedependent pathophysiologic changes of the brain after cardiac arrest. Available preclinical studies indicate that IV doses ranging from 1×10⁶ to 5×10⁶ cells provide substantial behavioral gains in post-cardiac arrest rodents [10,16-18,51,123-125]. According to the allometric scaling approach from animals to humans recommended by the Food and Drug Administration, the equivalent MSC dose in humans is about 1.6×10⁶ MSCs/kg [9].

The hMSC donor age is another critical factor affecting functional and structural outcomes in stroke affecting functional and structural outcomes in stroke [128,129]. MSCs from older donors have a decreased proliferative capacity and trophic factor secretion [130]. Autologous transplantation has been suggested as a safe therapy with minimal immune consequences. However, stroke patients tend to be older, which limits the proliferation potential of their autologous stem cells and limits this mode of therapy in this particular pathology.

Stem cell dosage and timing remain to be optimized, both to avoid complications, such as vessel embolism and local energy deficiency, and to optimize transplanted cell survival, differentiation, and function. To achieve this, stem cell interactions with the dynamic microenvironment in the ischemic brain must be further understood. Stem cell therapy should be considered according to the delivery routes, stem cell types, and quality, progress of disease, and age of patients for optimal translation in dose, timing, and cell age.

In cardiac arrest-induced brain injury, MSC-treated rats exhibit better or equivalent functional outcomes, compared to rats treated immediately with therapeutic hypothermia (32°C to 34°C) for 2 hours after ischemia, although hypothermia was not found to markedly enhance MSC therapy benefits in this pathology [51]. However, in ischemic stroke rats, post-ischemia hypothermia induced by injection of cold saline (4°C) via carotid artery at an infusion rate of 0.6 mL/min for 5 minutes can enhance the therapeutic effects of MSC on reperfusion injury [153].

Seven days after ischemia induction, the combined utilization of UC-BSCs (injection once via the tail vein) and erythropoietin (IP exposure for 5 consecutive days) was shown to have strong neurogenic and angiogenic effects in stroke rats [154]. Fibrin glue, a biocompatible and biodegradable natural product, combined with iPSCs enhanced stoke-induced deficits recovery. The fibrin glue may also increase iPSCs utilization safety [93].

Cerium oxide nanoparticles can bind to oxygen molecules and may scavenge intracellular reactive oxygen species [155]. Hyaluronic acid-cerium oxide-labeled human umbilical cord MSCs gain combined antioxidant and anti-inflammatory effects, and significantly attenuate stroke injury [156].