|Home | About | Journals | Submit | Contact Us | Français|
Foot ulcers are a major complication in patients with diabetes mellitus and involve dramatic restrictions to quality of life and also lead to enormous socio-economical loss due to the high amputation rate. The poor and slow wound healing is often aggravated by the frequent comorbidity of foot ulcers with peripheral arterial disease, making the treatment of this condition even more complicated. While the local treatment of foot ulcers is mainly based on mechanical relief and prevention or treatment of infection, improving perfusion of the impaired tissue remains the major challenge in peripheral arterial disease. While focal arterial stenosis is the domain of interventional angioplasty or vascular surgery, patients with critical limb ischemia and lacking options for revascularization have a much worse prognosis, because current treatment options avoiding amputation are scarce. However, based on recent research efforts, there is rising hope for promising and more-effective therapeutic approaches for these patients. Here, we discuss the current improvements of established therapies aimed at an improvement of limb perfusion, as well as the development of novel cutting-edge therapies based on stem-cell technology. The experiences of a ‘high-volume center’ for treatment of diabetic foot syndrome with a current major amputation rate of 4% are discussed.
The diabetic foot syndrome (DF) is a major disabling complication with an estimated lifetime incidence of up to 25% in patients with diabetes mellitus [Singh et al. 2005] and peripheral arterial disease (PAD) is one of the most common comorbidities in DF. However, despite the frequency and importance of these pathological conditions, clear epidemiological data with respect to the combination of DF with critical limb ischemia (CLI) are scarce. For example, in the EURODIALE study [Prompers et al. 2007], which included 1229 patients with diabetic foot ulcers, 49% of the patients also had PAD. However, this proportion is apparently underestimated, as 32% of the subjects had an ankle-brachial index (ABI) of >1.2, indicating mediasclerosis. Therefore, we can assume a prevalence of >80% of PAD in DF. Moreover, in this study, the proportion of patients with DF and CLI with an ABI of <0.5 was 12%. In another study from Milan [Faglia et al. 2009], data were presented with respect to the long-term prognosis of patients with DF and CLI, in which 554 patients with this combination were observed over a mean of 6 years. Peripheral angioplasty was performed in 74% and vascular surgical interventions in 21% of this selected group of patients with DF and CLI. A negative option for revascularization was given in only 5% of the patients. These subjects were treated with prostanoids (60–120 µg/day alprostadil) over 5 days after arteriography. The initial major amputation (MA) rate (‘early period’, within 30 days after intervention) in this study was only 4.1%, corresponding exactly to the early MA rate of our interdisciplinary diabetic foot unit, based on 754 patients with different stages of DF [Weck et al. 2010]. During their follow up, the researchers reported MA in 13% of all subjects, 8% in the group treated with angioplasty (percutaneous transluminal angioplasty [PTA]), 21% in the vascular surgery group and 59% in the group without options for revascularization. In 40% of the subjects, they found CLI of the contralateral leg [Faglia et al. 2009]. This finding emphasizes the importance of a detailed angiological examination of both legs.
Patients with a history of diabetic foot ulcers alone already have high mortality rates, mainly due to cardiovascular events but also other causes such as cancer [Iversen et al. 2009]. However, CLI further significantly increases the mortality risk [Faglia et al. 2009]. On the other hand, treatment options for this group of patients are still very limited, demonstrating the imperative need to develop novel therapeutic strategies. Based on the current literature and our experience, we summarize the noninvasive treatment options currently available for this highly morbid group of patients with DF and CLI and draw future perspectives.
In subjects with DF, the accompanying PAD, particularly of the lower leg arteries is apparently the major risk factor for MA. Revascularization appears to be possible in the large majority of cases if these patients are referred in time to an interdisciplinary organized center offering highly skilled experience, not only in diabetology, but also interventional angiology and vascular surgery [Faglia et al. 2006, 2009; Sumpio et al. 2010]. In these specialized units, a dramatic improvement of healing rates and reduction of the frequency of MA can be accomplished.
Endovascular treatment is mostly recommended because the risk of infection is lower; subjects with DF are older and have multiple comorbidities, representing additional risk factors for vascular surgery [Zeller, 2007; Zeller et al. 2009; Alexandrescu et al. 2009]. A clinical algorithm for the diagnosis and treatment of DF with PAD is illustrated in Figure 1. In patients with DF and CLI without options for revascularization, further therapeutic approaches must be considered in order to avoid MA, although none of these noninvasive strategies have been validated in randomized clinical trials so far. This includes methods aiming at an improvement of limb perfusion, for example, prostaglandin treatment, low-dose urokinase therapy and autologous bone marrow transplantation into the affected limb.
Although the treatment of CLI with prostaglandins is not a new option, this class of drugs needs to be discussed here briefly, especially in the context of other therapeutic strategies.
The impact of prostaglandins in the treatment of PAD is still discussed in a controversial light [Lawall et al. 2009a, 2009b; Amendt, 2005; Norgren et al. 2007]. There is some evidence that prostaglandin treatment improves claudication and some evidence exists with respect to improvement of walking capacity and quality of life [Amendt, 2005]. Summarizing the literature, one can assume an improvement of wound healing, reduction of pain and a reduction of MAs in ~50% of subjects with CLI and lacking options for revascularization after local infusion of prostaglandins. Prostaglandins are administered intra-arterially or intra-venously for 3–4 or 7–28 days. Prostaglandin E1 (PGE1) is commonly applied with doses of 40 µg bid over 2–4 hours. The majority of the current guidelines recommend this procedure, but the intersociety consensus for the management of PAD (TASC II) does not [Norgren et al. 2007].
Short-term studies in subjects with CLI had no clear results. In particular, there was no ulcer healing or pain reduction in the majority of these studies. In contrast to short-term studies [Belch et al. 1983; Cronenwett et al. 1986; Schuler et al. 1984; Telles et al. 1984], the majority of long-term studies in patients with CLI demonstrated a clear reduction of pain and ulcer size, and some of these studies indicate a reduced need for MA [Norgen et al. 1990; Stiegler et al. 1992; ICAI Study Group, 1999; UK Severe Limb Ischaemia Study Group, 1991; Trübestein et al. 1987; Duthois et al. 2003; Loosemore et al. 1994; Altstaedt et al. 1993]. It should be noted that these studies are older investigations with some methodological problems and should be considered with caution.
A more recent meta-analysis of the administration of PGE1 for patients with PAD stage III or IV not eligible for arterial reconstruction shows that it not only has significant beneficial effects over placebo on ulcer healing and pain relief, but also increases the rate of patients surviving with both legs after 6-months follow up [Creutzig et al. 2004]. Despite this positive statement, the combined endpoint ‘MA and death after 6-month follow up’ seems to be very high, at 22.6%.
However, in TASC II it is concluded that the current data do not provide evidence for a significant benefit of prostaglandins in CLI with respect to amputation-free survival, whereas all other scientific associations recommend the use of prostaglandins in CLI if revascularization is not possible [Lawall et al. 2009a; Hirsch et al. 2006; Ryden et al. 2007]. In the guidelines for the diagnosis and treatment of PAD from the German Society of Angiology and Vascular Surgery, the following recommendation is given: ‘In consideration of the meta-analysis about clinical efficiency of prostanoides, the recommendation of administration of prostanoids is given with level of evidence A. This conclusion is based, (apart from the outcome of the meta-analysis), on the assessment of predominantly older studies and concordant experiences of the authors (of this guideline)’ [Lawall et al. 2009b]. Special recommendations of administration of prostaglandins on patients with DF and CLI do not exist.
Hyperfibrinogenemia with the resulting increase in plasma viscosity and erythrocyte aggregation has been demonstrated in patients with coronary heart disease and PAD [Leschke et al. 1986, 1997; Peters et al. 1999; Partsch and Jochmann, 1993; Weck et al. 2008; Rietzsch et al. 2008]. The increase in plasma viscosity, in particular, can be a flow-limiting factor and a critical determinant of oxygen supply in the poststenotic microcirculation of myocardium and diabetic foot. It has been shown that urokinase is effective in improving the microcirculation in patients with coronary heart disease [Peters et al. 1999; Leschke et al. 1996, 2003; Leschke, 2008]. Plasma fibrinogen, plasma viscosity and red blood cell aggregation were reduced significantly in these studies. In another study, low-dose urokinase was used to treat subjects with nonhealing leg ulcers. The authors reported a significant improvement of microcirculation measured by an increase of laser-Doppler indices, increase of tcpO2 and a significant decrease of plasma viscosity and plasma fibrinogen [Partsch and Jochmann, 1993].
In addition, it has been shown that urokinase is effective in improving the microcirculation in diabetic patients with PAD stages III and IV according to Fontaine (CLI) or Rutherford stages 4–6 [Leschke et al. 1997; Weck et al. 2001, 2008; Hicken et al. 1995]. These preliminary data provide some evidence that diabetic patients apparently benefit more from rheologic treatment of PAD than nondiabetic patients. However, studies on the effect of intravenous urokinase treatment in patients with CLI and diabetic foot lesions are mostly retrospective in small patient cohorts.
Therefore, we performed an open, prospective, noncontrolled, multicenter cohort study in 77 type 2 diabetic patients with CLI and diabetic foot ulceration [Weck et al. 2008]. Patients had no surgical or endovascular treatment option based on interdisciplinary consensus. Urokinase (1 IU if plasma fibrinogen ≥2.5 g/l; 0.5 IU if fibrinogen <2.5 g/l) was administered for 21 days as an intravenous infusion over 30 minutes. After 12 months, 33% of the surviving patients showed completely healed ulcers without having MA (Figure 2). The total survival rate was 85%, amputation-free survival 69% and the rate of MA was 21%. Within the course of the study, 82% of patients experienced ulcer healing at least once. It should be noted that the mortality rate and also the rate of MA in this study was surprisingly low as compared with the literature and 33% of the patients in this study even met the ultimate treatment goal of freedom from amputation and residual ulcers for at least 1 year. Although this study was not controlled, randomized or blinded, it establishes a solid basis for the use of urokinase in this special patient population. The specific reasons for the low event rates in this study can be speculated upon. For example, high levels of fibrinogen have been associated with an increased cardiovascular risk and fibrinogen is known to rise as an acute-phase protein in the acutely infected diabetic foot. Alternatively, the effect of urokinase therapy has been attributed to a decrease in fibrinogen concentration with a subsequent improvement of microcirculation [Leschke et al. 1996, 1997, 2003; Leschke, 2008]. In another observational study, the heparin-induced extracorporeal LDL-precipitation (HELP) was used as an alternative therapeutic approach in subjects with diabetic foot, CLI and sepsis [Rietzsch et al. 2008]. The risk of limb loss in these patients was extreme. HELP reduced the fibrinogen levels by 68%. Only 3 of 17 patients underwent MA, thus, the HELP approach seems to be a proof of principle for fibrinogen-lowering therapy.
Urokinase treatment in diabetic patients with chronic foot lesions and CLI appears to be feasible, safe and effective. Based on our promising clinical experience with low-dose urokinase in the treatment of DF with chronic, nonhealing ulcers and CLI without options for revascularization, we have extended the application of urokinase to the perioperative treatment of borderline amputations in some cases. With this strategy, we have seen an improvement in wound healing in some subjects with fatal distal vascularization. It should be noted that these are observational data. Nevertheless, urokinase treatment in these patients appears a reasonable option in order to avoid MA. The data and theoretical considerations on low-dose urokinase treatment in subjects with diabetic foot presented here should be treated with caution and considered as a strong endorsement for the implementation of a prospective randomized controlled study, which is now underway.
A large body of experimental evidence in mice, rats and larger animals has demonstrated the feasibility and efficacy of stem cell therapies in restoring blood flow to the critical ischemic limb. These studies demonstrated that the number of circulating endothelial precursor cells (EPCs) increases in response to ischemia [Shintani et al. 2001; Takahashi et al. 1999] and that EPCs can be found incorporated into capillaries and interstitial arteries [Shintani et al. 2001]. It is assumed that EPCs may act in a paracrine manner by secreting vascular growth factors and cytokines [Asahara et al. 1999; Kamihata et al. 2001]. In 2000, Kalka and colleagues demonstrated that intracardial application of human EPCs in nude mice improved the peripheral circulation [Kalka et al. 2000]. Recently, Turan and colleagues demonstrated that intracoronary transplantation of autologous freshly isolated bone marrow cells (BMCs) improved global ejection fraction and infarct size significantly in patients with ischemic heart disease after 3 and 12 months of transplantation [Turan et al. 2011]. These results formed the basis for the rapid growth in studies with stem cells in patients with PAD. The Therapeutic Angiogenesis using Cell Transplantation (TACT) study in 2002 was the first report on application of bone-marrow derived mononuclear cells (BMMNCs) in the treatment of patients with CLI [Tateishi-Yuyama et al. 2002]. After 24 weeks, the ABI was improved by 31% and the pain-free walking distance was increased by 80% following intramuscular injection of BMMNCs. Of course, the early TACT study has been criticized for its lack of credibility, in particular with regard to all vascular information (angiogram, ABI, walking distance). Nevertheless, these studies raised considerable interest in stem cell and/or BMMNC therapy in PAD [Fadini et al. 2010; Lawall et al. 2010]. It can be argued that the data were based on a wide variety of different experimental approaches. Furthermore, the degree of ischemia was heterogeneous in these studies and CLI was not generally present. However, despite these limitations a meta-analysis demonstrated a consistent improvement of perfusion (based on quantitation of ABI, wound healing, walking capacity, tcpO2) throughout most of the studies [Fadini et al. 2010].
In the majority of studies in humans, the intramuscular injection of BMMNCs into the gastrocnemius muscle along a symmetric grid with fixed number of injections (20–60) was the preferred mode of application [Miyamoto et al. 2004; Prochaska et al. 2009; Van Tongeren et al. 2008]. The density of preformed capillaries and collaterals is highest in close proximity to the axial arteries and collateral growth preferably occurs in these regions. Therefore, it is rational to place injections along the occluded vessels of the lower leg, as performed in the studies by Amann and colleagues [Amann et al. 2008, 2009]. A schematic display of the injections is shown in Figure 3. Formation and extension of small collateral vessels appears to be the most important physiological repair mechanism in PAD [Unthank et al. 2004], presumably due to the local production of growth factors by BMMNCs. These collaterals can form direct connections between the axial main vessels. Nevertheless, the growth capacity of these collaterals is reduced in atherosclerosis and especially in diabetic macroangiopathy [Helisch and Schaper, 2003]. In contrast to intramuscular injection, intra-arterial application guides the injected BMMNCs to the border zone of maximum ischemia [Yoshida et al. 2003]. In order to optimize the therapeutic procedure, Bartsch and colleagues conducted an ischemic preconditioning of the affected leg by means of exercise and combined this procedure with intra-arterial and intramuscular administration of BMMNCs [Bartsch et al. 2007]. Recently, Kolvenbach and colleagues applied BMMNCs as adjuvant treatment in patients with PAD and CLI during bypass surgery and/or endovascular interventions [Kolvenbach et al. 2010]. In addition, BMMNCs have also been locally applied to restore angiogenesis and promote wound healing in type 2 diabetic patients with neuro-ischemic wounds [Humpert et al. 2005].
In the majority of studies, bone marrow is the primary source of stem cell material and enrichment of mononuclear cells from the crude aspirate can be accomplished with different techniques such as density gradient centrifugation (Ficoll™) [Boyum et al. 2002] or other commercially available blood-centrifugation and plasmapheresis systems [Tateishi-Yuyama et al. 2002]. However, these are laborious and require the background of a specialized and certified hematological or immunological unit. In contrast, bedside centrifugation systems have been developed (e.g. Smart Prep®, Harvest Technologies, USA) that can be easily used on the ward. These systems yield sufficient amounts of purified BMMNCs in a short time (~1 hour) and are considerably cheaper than the other techniques mentioned [Amann et al. 2008].
Relating to the clinical outcome of BMMNC therapy in PAD, two major studies with regard to the number of treated patients and duration of follow up should be mentioned. In the TACT follow-up study, mortality and amputation-free survival were analyzed as primary endpoints [Matoba et al. 2008]. The 3-year overall mortality was only 20% and after 3 years 60% of the patients were free of amputation. The BONMOT pilot study included 51 subjects with CLI and impending risk for amputation. During 3.2 years of follow up, limb salvage was 53%. The improvement of the mean Rutherford category from 4.9 at baseline to 3.3 after 6 months appears to be clinically important [Amann et al. 2009].
Based on the combined efforts in basic together with clinical research, the therapeutic arsenal for the treatment of patients with diabetic foot ulcers and CLI is growing constantly. The current development of novel techniques and clinical protocols including stem cell-based approaches gives realistic hope for the future to substantially improve the prognosis of this multi-morbid and critically ill group of patients.
This work was supported by the BMBF (grant number FKZ01GI0924 to the Paul Langerhans Institute Dresden - DZD e.V.), DFG SFB 655 ‘from cells to tissues’ and by the Centre for Regenerative Therapy Dresden (CRTD) to SRB.
The authors declare no conflicts of interest in preparing this article.