To the Editor: The identification of endothelial progenitor cells (EPCs) along with the description of their essential role in postnatal neovascularisation has been a major achievement for stem cell biology [1]. In particular the finding that the clonogenic potential of EPCs is reduced in patients with high cardiovascular risk [2] represents the first step towards the recognition of the therapeutic effect of EPC transplantation in cardiac and limbic ischaemia [3, 4]. However, two harmful side effects are expected to be associated with EPC transplantation as a consequence of their physiological function: malignant tumour neovascularisation and the induction of proliferative diabetic retinopathy [1]. The contribution of EPCs to tumour neovascularisation has been clearly demonstrated in animal models and confirmed, although to a lesser extent, also in humans [5]. Concerning diabetic retinopathy, recent reports have demonstrated that EPCs play a major role in the retinal neovascularisation of a mouse model of proliferative retinopathy [6], and that the concentration of stromalcell-derived factor 1, the most important chemokine mobilising EPCs, increases with the severity of diabetic retinopathy in vitreous samples of patients affected by type 2 diabetes [7]. These studies are in some way in contrast to two independent reports performed in non-complicated patients showing that type 1 diabetes per se is associated with a reduced number and function of EPCs [8], and that EPCs in type 2 diabetes are impaired in adhesion, proliferation and tubulisation [9]. Finally, a recent study has demonstrated an increased number of CD34+ mononuclear cells in patients affected by type 2 diabetes with both non-proliferative and proliferative retinopathy [10]. Although of interest, these results cannot be easily compared with the ones described above as EPCs represent only one of the different classes of circulating cells staining positively for CD34+ [1]. Taken together, and from a largely speculative point of view, these apparently discrepant results can be explained by a unifying hypothesis. Taking into account the fact that EPCs have a physiological role in the rescue and maintenance of the existing retinal capillary bed [11], it is possible to hypothesise that the reduced number and clonogenic potential of EPCs found in non-complicated diabetic patients [8, 9] might predispose these patients to the later development of diabetic retinopathy. Once the damage is widespread and specific chemokines are produced by the suffering retina, the bone marrow would respond by increasing the production of EPCs, which at this stage, possibly because of the high ambient glucose and/or the badly damaged retina, would give rise to an incompetent neoangiogenesis. To test this hypothesis and verify the possible association between the clonogenic potential of circulating EPCs and proliferative diabetic retinopathy, we isolated EPCs from whole blood obtained from 11 patients with type 1 diabetes and untreated proliferative retinopathy, from 12 patients without retinopathy despite similar age, gender distribution and duration of type 1 diabetes, and from 11 ageand sex-matched non-diabetic controls. Patients affected by type 1 diabetes took no medication other than insulin. Clinical characteristics of subjects included in the study are described in Table 1. Patients were classified with respect to the status of diabetic retinopathy by means of stereoscopic fundus examination and fluorescent angiography. Patients considered for this study either had no signs of diabetic V. Asnaghi . G. Mazzolari . M. R. Pastore . A. Maestroni . D. Ruggieri . L. Luzi . G. Zerbini (*) Renal Pathophysiology of Diabetes Unit, Division of Nutrition–Metabolism, San Raffaele Scientific Institute, Via Olgettina 60, 20132 Milan, Italy e-mail: g.zerbini@hsr.it
Read full abstract