The mechanism of benefit with stem cell treatment in acute neonatal HI is uncertain, but there are several lines of thinking which are prominent, and all of the proposed mechanisms may play a role in an additive fashion.

The first possible beneficial action is the actual replacement of intrinsic cells by the transplanted stem cells. While a small percentage of stem cells engraft, and some may even take the form of neurons, their numbers do not appear to be adequate to support much enhancement of outcome [19, 23, 24].

The other prominent ideas include blood vessel regeneration, an effect on the spleen to reduce its release of inflammatory cells which might have a deleterious effect in brain, and finally, and most likely, improved survival of intrinsic brain cells promoted by various noncellular factors.

Blood vessel regeneration potentially occurs via adhesion of CXCR4-positive cells onto vascular endothelium [25], recruitment of endothelial progenitor cells [26], and formation of periendothelial vascular cells [27]. Also, crude bone marrow cells have been shown to form endothelial cells [28].

Vendrame et al. [29, 30] proposed that umbilical cord blood may reduce the brain’s inflammatory response by acting on the spleen to reduce its output of inflammatory cells. These splenic inflammatory cells apparently increase the blood–brain barrier permeability, which produces an adverse effect. Walker et al. [31] showed that intravenous injection of MAPCs impeded the splenic response to injury by lessening the spleen’s release of inflammatory cells in association with improved outcome. The work of Leonardo et al. [32] confirmed these findings.

There are numerous mechanisms by which increased intrinsic cell survival might occur. Wang et al. [12] reported that neural stem cell proliferation in the affected tissue may be promoted by the Sonic hedgehog signaling pathway. Rosenkranz et al. [33] noted that umbilical cord cell transplantation decreased the expected rise in pro-inflammatory cytokines (interleukin-1 alpha, interleukin-1beta, and tumor necrosis factor alpha) occurring after HI injury. A marker for activated microglia and macrophages was reduced. Inflammation would thereby be reduced in brain [33]. Bae et al. [34] agreed that paracrine influences were the primary influence on intrinsic cell survival. Human umbilical cord blood cells reduced upregulation of CX43 and thereby brought about attenuation of reactive gliosis [35]. Overall, inflammation was reduced. Rosenkranz et al. [36] found various increased proteins which promoted angiogenesis. Apoptosis was lessened. Vascular endothelial growth factor (VEGF) increased perhaps increasing the proliferation of blood vessels. The same group also reported that human umbilical cord blood cells increased SDF-1 expression acting in concert with the SDF-1/CXCR4 axis [2], which appears to be a major factor for cell homing. In a primate model, Li et al. [37] reported that cell transplantation caused upregulation of IL-10 expression. Neuronal apoptosis was decreased and neuronal proliferation increased. Dayer et al. [38] found that neural progenitor cells increase expression of fibroblast growth factor-2, thereby improving HI outcome. These cells are thought to generate a pool of immature neurons available for repair (Fig. 16.1).

There are no reports available as yet of controlled, double-blinded trials with stem cells in acute neonatal HI injury.

Luan et al. [39] reported a 75-day-old male who suffered severe HI injury at birth. The infant received transplantation of fetal neural stem cells into the cerebral ventricle. The infant showed improvement. There were no controls noted. In view of the relatively late transplant, it is unclear whether this child should be considered an example of acute or chronic HI injury.

Three trials in progress are listed in ClinicalTrials.gov. These are (1) Cord Blood for Neonatal Hypoxic-Ischemic Encephalopathy at Duke University, (2) Autologous Umbilical Cord Blood Transfusion for Preterm Neonates at Ain Shams University in Cairo, and (3) Autologous Stem Cells in Newborns with Oxygen Deprivation at Hospital Universitario in Monterrey, Mexico. These trials appear to be confined to single institutions.

In any single medical center, there are only relatively small numbers of term infants with acute HI encephalopathy (HIE). Consequently, considerable time will be required to acquire the necessary numbers of subjects from one hospital system. In view of the fact there is ample preclinical laboratory evidence for benefit and a robust, multileveled rationale for the use of adult stem cells in the treatment of acute HIE, it is striking that multicentered, controlled trials have not been initiated. The cost of such a trial will be significant.

Cotton et al. [40] reported the treatment of 23 infants suffering from acute HI injury with intravenous autologous umbilical cord blood. This study demonstrated that, with considerable institutional preparation and coordination, it is feasible to conduct a trial with fresh, autologous umbilical cord blood in acute HI injury in newborns. The effort used to bring about this preliminary trial was considerable.

One of the problems inhibiting such a trial is the lack of a uniform, widely available cell preparation. Although autologous umbilical cord blood could be used for this purpose, there is some variability in the quality of preparation among the samples obtained, which would undoubtedly influence the outcome of the study. A widely available commercial preparation would be advantageous.

In order to choose the subjects for such a trial the inclusion and exclusion criteria will need to be carefully constructed. The study should be confined to term or near term infants. The criteria composed by the American Congress of Obstetricians and Gynecologists [41] may prove useful. The four critical criteria are: (1) evidence of metabolic acidosis in arterial umbilical cord blood, (2) neonatal encephalopathy in infants born at 34 or more weeks of gestation, (3) CP of the spastic or dyskinetic type, and (4) exclusion of other causes. Other criteria that could be used as confirmatory would include a specific HI event, significant fetal bradycardia with abnormal decelerations, persistently low Apgar, evidence of adverse effect in other organs, and a confirmatory imaging study. Magnetic resonance imaging (MRI) is the best tool for this determination [42]. In the first 24 h, diffusion-weighted images may demonstrate increased signal intensity in central brain regions such as basal ganglia. This finding is seen in so-called severe asphyxia. T1 and T2 weighted images are more conveniently done at around 7 days of age when they diagnostically useful. Less severe or “partial” asphyxia results in a watershed pattern of injury with relative sparing of the basal ganglia.

Among the various animal models for CP perhaps the greatest difficulty is producing an injury which is sustained over a significant period of time. Rodents tend to die or recover quickly from brain injuries and consequently their motor manifestations of the injury fade rapidly. A sustained motor disability is essential to the use of the model in the experimental treatment of an injury designed to mimic long-standing brain injury.

Most of the models are variations of the Rice–Vanucci method [43] which impede delivery of oxygen to the animal brain either before or around the time of delivery. In the Rice–Vannucci method, the investigator ligates a unilateral carotid artery and then delivers a period of hypoxia. Variations on this theme have been conducted in rodents, piglets, and fetal sheep [44]. None of these variations are totally satisfactory.

Prenatal HI injury has been produced in the rabbit [45], but the injury is not sufficiently long-lasting to mimic CP in children. The induction of HI injury in prenatal rabbits by uterine ischemia results in motor disability in the animals, but once again the disability is not long lasting, perhaps less than 2 weeks. The motor disability may be assessed with a swim test [46]. The model is also amenable to study with MRI [47].

Maternal administration of lipopolysaccharide to pregnant rats late in gestation results in significant motor impairment of the offspring, but these findings tend to fade by 5 weeks [48]. A similar model has been produced in rabbits [49, 50]. This type of injury, however, does not necessarily mimic term infant HI injury and therefore would not be appropriate for a preclinical study of stem cell therapy for chronic HI injury. The model is more reminiscent of the periventricular leukomalacia of premature infants rather than term infant HI injury.

All these models suffer from the lack of similarity in structure to the human brain. The rodent models are particularly dissimilar in that the white matter is considerably less than in the mammal and differing in structure. The models overall are comparatively short-lived in terms of the motor abnormalities.

Stem cell therapy is widely proposed as a treatment for CP. Most of these therapies are offered for a fee in countries outside the USA and on a compassionate basis in at least one site in the USA. The treatment, in general, appears to be safe but the efficacy of the treatment is not clear. The problem is compounded considerably by the fact that CP is comprised by many different disorders only one of which (neonatal HI injury) is the subject of this chapter.

We are aware of several reports showing benefit. None of the reports address specifically the issue of the treatment of chronic neonatal HI injury.

Min et al. [51] reported benefit in a controlled study with allogeneic umbilical cord blood. To date, this is best study of the possible benefit of stem cell therapy in CP. In this study, erythropoietin was added to the cells. The three groups of patients were (1) the cells plus erythropoietin, (2) erythropoietin alone, and (3) the placebo group. They measured the gross motor performance measure, Bayley scales of infant development, F-fluorodeoxyglucose positron emission tomography, and diffusion tensor imaging. The first group had had significantly improved scores on the motor function measure and the Bayley scales. The diffusion tensor imaging and positron emission studies seemed to relate to the benefit seen with the cells plus erythropoietin. The subjects had a variety of diseases as a cause of their CP: periventricular leukomalacia, “diffuse encephalopathy”, focal ischemia or hemorrhage, multicystic encephalomalacia, or cerebral malformations. It was not clear how many of the children sustained HI injury as a cause of their chronic disability. Because the children were not segregated as to the cause of their CP, we cannot draw conclusions about the possible benefit of the cells in chronic HI injury.