Previously, we reported that histopathological studies in stroke patients have pointed out the presence of neural stem/progenitor cells in the poststroke human cerebral cortex and that the peak in endogenous neurogenesis occurs approximately 1–2 weeks after a stroke [34]. Consistent with this, analysis of the therapeutic time window in murine stroke model revealed that administration of bone marrow-derived mononuclear cells within 24 h after stroke had a mild and nonsignificant effect on brain regeneration/protection following ischemia [35], but administration of these cells between day 2 and day 14 after the ischemic event showed significant positive effects [36]. Recently, Prasad et al. reported that intravenous administration of bone marrow-derived cells for stroke patients in the chronic period is safe, but there was no beneficial effect of treatment on stroke outcome [37]. The time window after stroke onset in their study was 18.5 days (median), which seems too late to support endogenous neurogenesis after stroke. Their results may be attributed to the time lag between the onset of stroke and the peak of neurogenesis.

The intra-arterial route is well known to result in higher cell counts at the target site compared to intravenous infusion [38, 39]. Although some clinical trials of intra-arterial administration of bone marrow mononuclear cells have demonstrated feasibility and safety [38, 40], several studies reported that intra-arterial injection was not superior to intravenous injection in experimental stroke models [41, 42]. Friedrich et al. reported that there were no significant differences in neurological function with patients undergoing intra-arterial bone marrow mononuclear cells treatment, compared with the control group [39]. Moreover, generalized seizures developed in patients after intra-arterial injection of bone marrow mononuclear cells, though the relation between seizures and intra-arterial treatment is unknown [38]. Intra-arterial injection would permit a substantial lowering of the number of cells, but caution might be exercised in intra-arterial administration bone marrow mononuclear cells for stroke patients compared with intravenous administration. Results of the therapeutic and adverse effects of these treatments and different modes of cell administration are expected to emerge over the next several years. More definitive conclusions regarding differences between the intravenous and intra-arterial administration of bone marrow mononuclear cells for stroke patients await the results of an ongoing randomized and controlled multicenter trial in Spain [43].

In Table 11.1, the number of cells (from 1 × 10⁷ to 5 × 10⁸ cells) given to stroke patients in clinical trials is relatively small. A growing body of clinical and experimental evidence indicates that the number of injected cells reaching the brain parenchyma seems to be very small, i.e., preclinical studies indicate that approximately 0.02–1 % of injected cells home into the brain [42, 44, 45] and, probably, the differentiation of the stem cells into endothelial cells in the ischemic brain may not play a critical role in angiogenesis after stroke. We found that administration of a relatively small number of bone marrow-derived mononuclear cells had significantly beneficial effects on regeneration/protection of injured brain tissue in an experimental model [36].

Based on these observations, we conducted a clinical trial to enhance functional recovery through activating angiogenesis by autologous bone marrow mononuclear cells in patients with cerebral infarction. Our clinical trial was an unblended, uncontrolled phase 1/2a clinical trial. This clinical protocol was approved by the Ministry of Health, Labour and Welfare and the institutional review board of the National Cerebral and Cardiovascular Center (ClinicalTrials.gov Identifier: NCT01028794). The aim of this clinical trial was to assess the feasibility, safety, and efficacy of the intravenous transplantation of autologous bone marrow mononuclear cells into patients with stroke. The clinical trial employed a nonrandomized open-label study design for 12 stroke patients (25 ml, low-dose group, N = 6; 50 ml, high-dose group, N = 6). The outline of this protocol is shown in Fig. 11.1. Major inclusion criteria were patients with cerebral embolism, National Institute of Health Stroke Scale (NIHSS) score more than 9 on day 7 after stroke, and improvement in NIHSS in the first 7 days after onset less than 6 points. To improve the sensitivity for detecting efficacy signals, the enrollment was restricted to patients diagnosed with cerebral embolism and those expected to exhibit very poor outcomes during the chronic period at day 7 after onset.

On day 7–10 after stroke, patients had 25 ml (low-dose group) or 50 ml (high-dose group) of bone marrow cells aspirated from their iliac bone. Autologous bone marrow-derived mononuclear cells were purified by the density gradient method and administered intravenously on the day of the bone marrow aspiration. The primary endpoint was safety and improvement in the NIHSS score compared with our historical control. No side effects or safety problems were observed. A comparison of the results from patients receiving the two doses (25 and 50 ml) of bone marrow mononuclear cells showed that the higher dose was superior to the lower one in terms of showing a trend toward improved response (without statistical significance). Similarly, in comparison to historical control group, autologous bone marrow cell transplantation also showed significantly better outcomes in NIHSS score. Most of the patients showed a significant improvement in neurological function at 6 months after cell transplantation. Furthermore, analysis of cerebral blood flow and metabolism in patients after cell transplantation showed a trend favoring an increase in rCBF and rCMRO₂ in both ipsilateral and contralateral hemispheres at 6 months, compared to 1 month, after cell transplantation. Our study demonstrates that administration of autologous bone marrow mononuclear cells to patients with severe embolic stroke was feasible and safe. Furthermore, the positive trends favoring neurologic recovery in a dose-dependent manner and improvement in cerebral blood flow and metabolism in poststroke patients receiving cell therapy emphasized the potential of this approach.

For accurate assessment of the therapeutic effects of hematopoietic stem cell transplantation for stroke, improvements in clinical trial design are desirable. Fugl-Meyer Assessment is one of the most widely used quantitative measures of motor impairment after stroke [49], and assessment of motor function by Manual Muscle Testing at enrollment and follow-up with the Fugl-Meyer test would be one of the most workable designs to evaluate the effect of cell therapy. In addition, image assessment is another important component in evaluating the effect of cell therapy. Resting-state functional magnetic resonance imaging (fMRI) is one possible candidate, because it is applicable to patients with stroke who are not capable of proper performance of motor tasks [50]. Furthermore, systematic assessment of initial resting-state functional connectivity using resting-state fMRI will be able to provide prognostic prediction of later motor recovery in stroke patients. Although the presence of “responder” and “nonresponder” patients in the context of cell therapy has been reported [51], information obtained by MRI images may also provide the criteria to select only responder for enrollment in clinical trial.

The results from clinical trials, including our study, indicate that autologous hematopoietic stem cell transplantation is feasible and safe in patients with stroke and encourage the performance of randomized clinical trials to clearly prove the effect of cell therapy. In addition to hematopoietic stem cells, many kinds of stem cells have been tried in clinical trials, including autologous mesenchymal stem cells [52], allogeneic mesenchymal-derived stem cells [53], allogeneic teratocarcinoma-derived neuronal cells [6], and fetal porcine-derived neural stem cells [7]. Though the source of transplanted cells and the route for injection have varied, the major target of cell transplantation is, we believe, the modulation of the healing process after stroke (Fig. 11.2), which is similar to the wound healing process that consists of inflammatory, proliferative, and remodeling phases [54]. Optimal treatment during each phase would maximize functional recovery after stroke, and the combination of cell therapies at different phases could be one of the approaches for best recovery in the future.