I haven't read into the actual literature, but there certainly seem to have been a few successes due to ES. I'm a bit dubious of this claim that they ALWAYS go wrong.
ES cells always form a tumor
when implanted in a mass. They do
not form a tumor when injected in a suspension. The hypothesis is that ES cells produce autocrine/paracrine factors that stimulate differentiation. These factors need close proximity to other ES cells to reach high enough concentrations to trigger differentiation. So, the approaches with ES cells have been two:
1. Inject ES cells in a suspension. This has been tried in animals to treat animal models of Parkinson's, spinal cord injury, and cardiac repair, for example.
2. Get the ES cells to differentiate
in vitro (a culture dish) and then implant the differentiated cells. This was tried many years ago in selecting cardiac cells from differentiated ES cells and injecting them into an animal model of cardiac repair.
The problem with #2 is that, when the ES cells differentiate in culture, only a small percentage of the cells form any given tissue. ALL the tissues form in any given dish. Much research with ES cells now is to try to direct the differentiation to one tissue. However, they are trying to do this based on small molecules changing the chromatin or other structure of the DNA to induce specific differentiation.
IMO, this misses a great oppportunity. The ES cells must be producing proteins that act on neighboring cells to cause them to differentiate down a particular pathway. So, withdraw the inhibition to differentiation in the culture media (ES cells must be prevented from differentiating, in contrast to adult stem cells who must be stimulated to differentiate) and then take the conditioned media and purify the proteins in it. Then test the individual proteins on new cultures of ES cells to identify those proteins that specifically direct differentiation. Then clone the proteins. The proteins could either be used by themselves (a la BMP to induce bone) or stimulate large quantities of ES cells to differentiate
in vitro and then implant the differentiated cells.
Of course, this doesn't get around the rejection problem. How do you get an adult's ES cells? Either by cloning or by inducing adult cells to become ES cells (iPCs). The problem with iPCs is that, so far, it involves transduction of genes and some of the genes are oncogenes. So there are safety concerns.