For a stem cell therapy to be effective in augmenting recovery after stroke, the following criteria should be satisfied [17]: therapy should be both safe and effective; applicable to a broad spectrum of stroke patients, in terms of stroke subtypes (ischemic/hemorrhagic), time (acute/chronic), and vascular territories; and it should be inexpensive and cost-effective.

At this time, rigorous reasoning is required to replicate experimental results in stroke patients. The Stroke Treatment Academic Industry Roundtable (STAIR) recommendations were developed to improve the quality of preclinical and clinical research on neuroprotective and neurorestorative drugs, and to reduce the divide between them [18]. Based on these recommendations, studies should be randomized, controlled trials. After randomization, experimental procedures may not be blinded, as bone marrow sham aspiration should not be performed in control patients. A MSC dose equivalent to that used in animal studies (1 × 10⁵–3 × 10⁶ cells/rat), and based on mean body weight, should be used. Patient selection should be based on severity and location of lesions, and time of application. Although the optimal time window for administration is undecided, subacute periods of stroke should be the first candidate time period tested, i.e., after the acute period, when patients are likely to experience rapid worsening/improvement, and before the blood–brain barrier (BBB) closes and chemokine expression (i.e., stromal cell-derived factor 1α) in the infarcted brain disappears. As all preclinical studies were conducted in animal models of middle cerebral artery occlusion, patients with middle cerebral artery territory (or anterior circulation) stroke should be selected. Patients with moderate to severe neurological disabilities could be ideal candidates. Patients suffering from very severe strokes are likely to have poor outcomes regardless of intervention, whereas patients with minor strokes are not suitable for these potentially risky experimental treatments. In addition to the clinical outcomes measured, laboratory and neuroimaging findings should be used as surrogate markers of efficacy. Lastly, patients should be followed for more than 90 days. However, long-term monitoring (> 6 months) is likely unnecessary because autologous MSCs are a clinical cell line and die within days/weeks of administration [19].

Recently, the Stem Cell Therapies as an Emerging Paradigm in Stroke (STEPS) committee suggested guidelines for both preclinical and clinical trials on stem cell therapies in stroke. The committee has suggested guidelines for bridging basic and clinical studies [20], early stage clinical trials [19], and phase II/III trials [21].

Although, to date, clinical trials of stem cell therapy have focused on the feasibility and safety of application of autologous MSCs or bone marrow cells in stroke patients, further trials to understand the therapeutic effects of stem cells are needed. Although the proportion of patients in the STARTING trial with a mRS of 0–3 significantly increased following MSC treatment, many patients in the MSC group remained significantly disabled [10].

The primary hurdles for current stem cell therapies include: the long culture period required to obtain sufficient stem cells, the relatively small number of MSCs that migrate into the injured brain, the death of stem cells within a toxic environment, limited trophic support by transplanted stem cells, and the use of xenogenic serum, with concomitant risk of transmission of prion diseases and zoonoses [22]. The following issues should be addressed in order to overcome these hurdles.