Re: Angiographic Demonstration of Neoangiogenesis after Intra-Arterial Infusion of Autologous Bone Marrow Mononuclear Cells in Diabetic Patients with Critical Limb Ischemia

With interest, we read the article by Ruiz-Salmeron and colleagues in your journal (7). Although the reported results are clearly interesting and do support the earlier findings regarding bone marrow cell therapy in patients with critical limb ischemia, the article and especially the part on increased collateralization as quantified on digital substraction angiographies (DSA) using a semicomputerized method raises some questions.

We agree with Ruiz-Salmeron et al. that therapeutic neovascularization by administration of bone marrow cells is a promising new strategy for treatment of peripheral arterial disease and critical limb ischemia, in particular. Clinical benefit has been reported, however, mainly through assessment of surrogate measures for improved perfusion, such as Ankle–Brachial Index and transcutaneous oxygen measurements, techniques that have their specific limitations, especially in diabetics (10). Clinical studies, thus far, have not provided reliable and reproducible assessment of neovascularization.

Ruiz-Salmeron et al. show that intra-arterial administration of bone marrow mononuclear cells (BM-MNCs) in 20 diabetic patients with severe critical limb ischemia resulted in a notable improvement in the Rutherford-Becker classification, the University of Texas diabetic wound scale, and the Ankle–Brachial Index in the target limb after 12 months of follow-up as compared to baseline. Importantly, they also report increased vascularization after cell therapy, which was assessed at 3 months using conventional intra-arterial digital subtraction angiography (DSA) quantified by MetaMorph software.

We laud the attempt of the authors to quantify the neovessels induced with intra-arterial administration of BM-MNCs. However, the authors present a novel method to quantify vascularization without providing essential technical data and background information with regard to reliability and reproducibility. Thus far, studies on the use of DSA to evaluate neovascularization in the lower limb have almost invariably stated that the angiographic resolution is too low to identify the newly formed vessels (1,6,8,9). Although several studies have used MetaMorph software to quantify microvasculature on photographic and microscopic images (2–4), no reports seem to be available on the application of this computerized method to analyze vascularization on DSA in humans. No information is provided by the authors on inter- and intra-examiner reproducibility as well as test–retest reliability regarding the use of this method for quantification of collaterals on DSA. Furthermore, as acquisition and procedural parameters, for example, the ambient temperature, patient's heart rate, medication use (i.e., vasodilating drugs), interval between contrast injection and image acquisition, and venous overprojection potentially obscure the definition and quantification of collaterals on DSA when a comparison is made between time points (5), these factors should have been reported in more detail.

We agree with Ruiz-Salmeron et al. that quantification of vascularization is essential for optimal evaluation of the effects of cellular therapy in patients with peripheral vascular disease. Thus far, methods to quantify neovascularization on DSA have not been established. Although the method proposed in this paper is interesting, further research is necessary for its validation.