It has been suggested that the MSCs have the capacity to release growth and trophic factors or to stimulate their release from resident neural tissue and to contribute to produce the therapeutic benefit in cerebral stroke [8]. Intravenous infusion of MSCs in experimental stroke models leads to inhibit apoptosis of cells at the lesion boundary [9] and promotes endogenous cellular activities such as proliferation [10]. Low-level basal secretion of multiple neurotrophic factors by MSCs has been observed in vitro, and it has been reported that ischemic rat brain extracts can produce neurotrophins and angiogenic growth factors in MSCs [8]. Brain-derived neurotrophic factor (BDNF) is constitutively expressed at low levels in primary human MSCs in vitro and is increased in ischemic lesions after intravenous infusion of MSC in the rat middle cerebral artery occlusion (MCAO) model [3, 11]. Transplantation of BDNF gene-modified human MSCs increased BDNF levels in ischemic lesions and stronger therapeutic effects than MSCs alone [3, 11]. Enhanced benefit was also observed with human MSC genetically modified to express GDNF (glial cell-line derived neurotrophic factor) [12]. Transplantation of BDNF-secreting MSCs into a spinal cord injury model promotes functional recovery and facilitates sprouting of corticospinal and raphespinal axons [13]. One potential advantage of a cellular therapy that delivers trophic factors to damaged neural tissue rather than systemic pharmacological approach is the reduction in potential adverse effects of systemic drug delivery.

MSCs derived from bone marrow secrete angiogenic cytokines such as vascular endothelial growth factor (VEGF) [14] and angiopoietin-1 (Ang-1) in vitro [14, 15]. VEGF has angiogenic property in nervous tissue [16] and initiates to form immature vessels by vasculogenesis/angiogenesis [17]. However, VEGF increases vascular permeability within shortly after an ischemic injury [18], which induces cerebral edema. Direct injection of VEGF into central nervous system (CNS) tissues results to open the blood-brain barrier (BBB) [19]. Ang-1 contributes maturation, stabilization, and remodeling of blood vessels [20, 21] and promotes angiogenesis in the nervous tissue [15, 22]. Ang-1 protects the vasculature from leakage [23], an action that may contribute to anti-edematic effects following cerebral ischemia. Ang-1, which is produced by pericytes [24], signals through the Tie2 family of tyrosine kinase receptors on endothelial cells to enhance blood vessel stabilization and could maintain BBB integrity and reduce “leakiness” [25, 26].

Following traumatic brain injury, pericytes migrate from the vascular wall [27], and the neurovascular unit (endothelial cell-pericyte-astrocyte-neuron) is compromised. If a similar disruption of the neurovascular unit occurs following stroke, it would be expected that MSCs might provide support for the microvasculature via Ang-1 signaling to vulnerable endothelial cells.

We have found that the infused Ang-1-MSCs genetically modified to express Ang-1 results in greater neovascularization and functional improvements than MSCs alone in MCAO rat [15]. By contrast, intravenous infusion of genetically modified MSCs that hypersecrete VEGF into MCAO model resulted in deterioration of neurological function [14], consistent with VEGF leading to increased vascular permeability. Miki et al. [28] reported that marrow stromal cells genetically modified to express VEGF, however, may have greater therapeutic effect than non-modified cells. Therefore, the level of VEGF expression may be critical in terms of potential therapeutic effects. Intravenous injection of MSCs genetically modified to express both Ang-1 and VEGF resulted in the greatest neovascularization and functional recovery [14]. Thus, an orchestrated expression of VEGF and Ang-1 may be important for appropriate neovascularization.

It has been suggested that pericytes are a source of MSCs [29, 30]. Because pericytes are disrupted after neural insult and that transplanted MSC may have therapeutic effects on microvasculature, repair of microvasculature will be an important therapeutic target for cerebral ischemia.

New neurons from progenitor cells are generated within the subventricular zone (SVZ) of the lateral ventricle and the dentate gyrus of the hippocampus in the adult mammalian brain [31]. Neural progenitor cells in the SVZ migrate through the rostral migratory stream (RMS) to the olfactory bulb where they differentiate into interneurons [32]. The number of progenitor cells within the SVZ (doublecortin-positive cells) increases after cerebral ischemic insult. It has been reported that the transplanted MSCs contribute to increase cell number [7, 10]. Shen et al. [33] demonstrated that expression of synaptophysin, a presynaptic marker, increases in MSC-treated ischemic brains, suggesting synaptogenesis.

Bang et al. [34] recruited 30 patients prospectively and randomly with cerebral infarcts in the middle cerebral artery (MCA) territory, all of whom showed severe neurological deficits in the first study to examine the feasibility, efficacy, and safety of a cell therapy approach in stroke patients using culture-expanded autologous MSCs. Of these, 5 patients received intravenous administration of 1 × 10⁸ MSCs and 25 did not. The pretreatment characteristics (clinical and radiological) were similar in both the control- and MSC-infused groups. Bone marrow collection was performed 1 week after admission and mononuclear cells were isolated. Plastic-adherent cells were cultured and expanded in fetal bovine serum. MSCs as CD34^–, CD45⁻, SH2⁺, and SH4⁺ were delivered via two infusions (5 × 10⁷ cells per infusion) at 4–5 and 7–9 weeks after stroke onset. The patients were observed over the course of a year.

The patients in the first study by Bang et al. [34] had large infarctions within the MCA territory, evaluated by diffusion-weighted MRI. The MSC-injected group showed greater functional recovery as measured by the Barthel index. The MSC group showed no deaths, stroke recurrence, or serious adverse events. This study demonstrated safety and indicated modest functional improvement, but it was emphasized that double-blinded studies with larger cohorts would be necessary to reach a definitive conclusion. A 5-year follow-up study confirmed that there were no adverse events after injection of human MSCs [35].

We reported a Phase I/II study describing a series of 12 stroke patients who received intravenous infusions of autologous bone marrow-derived MSCs [36] (Fig. 12.2). First, safety was confirmed with human MSCs in a nonhuman primate MCAO model [37]. The overall protocol of the study is outlined in Fig. 12.3. In this study, bone marrow collection was carried out within weeks after admission of the patients into our hospital. Plastic-adherent cells were cultured with patient-derived autoserum using methodologies that allowed culturing of autologous human MSCs (ahMSCs) to very high homogeneity [4, 38]. The expression pattern of cell surface antigen (CD34⁻, CD45⁻, CD73⁺, and CD105⁺) was consistent between patients. After the cells were expanded and safety and antigenic phenotype analyses were performed, the ahMSCs were cryopreserved and stored until use. On the day of infusion, cryopreserved cells were thawed and infused intravenously.

MRIs after cell infusion showed no tumor or abnormal cell growth in any of the 12 patients over 7 years. There was a trend of correlation between improvements of the National Institutes of Health Stroke Score (NIHSS) and the reduction of lesion volume within the first weeks after cell infusion, suggesting a therapeutic benefit from intravenous infusion of ahMSCs [36]. Notably, in some of these patients, the recovery rate dramatically improved within the first 2 weeks after ahMSC infusions (Fig. 12.3a ). Moreover, there was a steep reduction in lesion volume during the first 2 weeks after cell infusion (Fig. 12.3b), and the reduction in lesion volume correlated with functional improvement (Fig. 12.3c) [36]. Serial evaluations showed no severe adverse effects by cell-related, serological, or imaging-defined events.