Huang et al. (1), in their small seminal clinical trial of cellular therapy for critical limb ischemia in diabetic patients, did not include presence of diabetic retinopathy among their exclusion criteria. It is known that individuals with diabetes are subjected to poor blood vessel growth in ischemic hearts and limbs and increased angiogenesis in retinal complications. This so-called “diabetic paradox” has been attributed to the differential regulation of angiogenic factors in the retina versus the systemic circulation (2). While decreased levels of endothelial progenitor cells are seen in diabetic patients with peripheral arterial disease (3), a role for endothelial progenitor cells in the development of proliferative diabetic retinopathy (PDR) also has been demonstrated (4). Possible harmful side effects of progenitor cell transplantation may include pathologic neoangiogenesis favoring the development or progression of cancer or PDR (5). Moreover, although data on blood cells are not presented, the increased blood viscosity due to the leukemoid response to granulocyte colony–stimulating factor may favor retinal vessel occlusion. Therefore, it was desirable that patients were screened for PDR before and after cell transplantation, in order to identify or exclude such an undesirable effect.
Finally, the achievement of a better metabolic control in the transplant group than in the control group is not unexpected and can be explained without postulating β-cell regeneration, since wound healing per se and reduced inflammation may have improved insulin sensitivity. The authors correctly cite the hypothesis that circulating progenitor cells may have a role in rescue of pancreatic endocrine function (6), but this has not been demonstrated thus far. Transplanted cells are recruited in the pancreas of streptozotocin-induced diabetic animals, but neither sign of endocrine transdifferentiation nor improvement in blood glucose metabolism have been shown (7). Moreover, this claimed mechanism would clearly interest only those forms of diabetes due to primitive β-cell failure, while only 40% of the study patients had type 1 diabetes.