The true incidence of PAD and CLI in diabetic patients is difficult to estimate and can be erroneous due to several reasons. A significant number of patients with PAD are asymptomatic because presence of peripheral neuropathy may mask the symptoms of IC [7]. Furthermore, the method used to diagnose PAD and CLI [symptoms of IC, palpation of peripheral pulses, ankle-brachial index (ABI) , toe pressures, and transcutaneous oxygen pressure (TcPO₂) ], as well as the criteria used for the classification of CLI (Fontaine classification or Rutherford classification [9]) has a major influence on statistics in large epidemiological studies. Moreover, the definition of CLI has changed over the last decades, with less severe ischemia of the lower limbs being classified as CLI [1]. It is also important to bear in mind that studies addressing the selected population of patients with DM and CLI are still limited and most data are from subgroup analysis of PAD studies.

Large population-based studies have shown that the prevalence of PAD is higher in diabetic patients compared with non-diabetic individuals. The Framingham study showed that there was a 3.5- and 8.6-fold excess risk among men and women, respectively, of developing PAD in patients with DM [10]. With respect to CLI, diabetic patients with PAD have a fourfold increased risk of developing CLI in comparison with non-diabetic individuals with PAD [11].

The incidence of CLI has been estimated to be approximately 500–1000 per million every year in a European or North American population [11]. The prevalence of CLI, on the other hand, in individuals 60–90 years old has been estimated to be approximately 0.5–1.2 %, but figures are inconsistent among different studies [1]. The true incidence and prevalence of CLI in diabetic patients are difficult to assess, since most CLI study populations consist of both diabetic and non-diabetic patients, while the proportion of diabetic participants varies from 35 to 80 % [1].

The natural course of CLI if left untreated is difficult to determine, since most patients with CLI receive some type of revascularization. If an exclusively conservative approach is applied in patients with CLI, irrespective of DM status, almost 50 % will end up with major amputation within 1 year [12]. However, diabetic patients with CLI appear to be at increased risk of death in comparison with non-diabetic individuals with CLI and have over 50 % mortality at 6 months if revascularization is not performed [12]. Nevertheless, although not all studies have reached similar conclusions, it seems that even when revascularization is performed in suitable patients, those with DM have higher amputation and mortality rates in comparison with non-diabetic individuals [13–16].

In a cohort of 564 diabetic patients with CLI endovascular procedure was performed in 74 % of the patients, surgical revascularization in 21, and 5 % of the patients were found unsuitable for revascularization and were treated with prostanoids [17, 18]. Major amputation rate within 30 days after intervention was only 4.1 %, while during the 6 years follow-up period 13 % of all patients underwent amputation, 8 % in the group treated with percutaneous transluminal angioplasty (PTA) , 21 % in the group treated with bypass grafting (BG) , and 59 % in the group no suitable for revascularization. Almost half of the patients died during the follow-up period [17]. Interestingly, amputated patients who had previously undergone revascularization had significantly longer survival than amputated patients who had not previously had revascularization [19]. Moreover, almost half of the participants developed CLI in the contralateral limb over a 6-year period [20]. However, the level of amputation was lower in the contralateral limb probably due to increased patient awareness and immediate management.

Most studies have shown that the 2-year mortality rates in diabetic patients with ischemic or neuro-ischemic foot ulcers or after amputation are up to 30–50 %, while the 5-year mortality rates reach up to 50–70 % [21]. It should be noted that ulcer infection is an extremely problematic issue and in combination with ischemia is associated with increased amputation risk. Presence of advanced infection and ischemia is the cause of lower limb amputations in 25–50 % of diabetic patients [21, 22].

Atherosclerosis in DM has various biologic and clinical differences from the disease in non-diabetic patients [3]. Vascular disease in DM affects both microcirculation and macrocirculation [5, 6, 22, 23].

Microvascular disease is characterized by involvement of blood vessels at the level of the arterioles and capillaries [5, 6]. Microvascular changes in DM include thickening of the basement membrane, endothelial dysfunction, and impaired vasodilatation, leading to impaired response to stress and trauma when an increased nutritional circulation is needed [5]. DM impairs endothelial function through hyperglycemia, excess circulating free fatty acids, and insulin resistance and results in inhibition of endothelial nitric oxide (ΝΟ) synthase and impaired NO-mediated dilatation [24]. In addition to reducing NO and prostacyclin, DM increases endothelin-I and angiotensin-II, which are potent vasoconstrictors [24]. Microvascular dysfunction due to diabetic neuropathy and loss of vessel auto-regulation is characterized by opening of the arterio-venous shunts, altered vascular resistance, and permeability that result in impaired capillary circulation and functional ischemia [21–23]. The abnormal distribution of blood flow at the capillary level is found in diabetic patients even in the absence of PAD, but is more pronounced in the presence of PAD [23]. Functional ischemia and skin microcirculation are further impaired by the altered hemorheology and the pro-thrombotic environment that prevails in DM [23]. Thus, although PAD is the underlying and main defect in diabetic patients with CLI, the impaired capillary circulation and the reduction in total blood flow induce a series of local responses in the microcirculation that may contribute to ischemic rest pain and foot ulceration.

Macrovascular disease is characterized by involvement of large- and medium-sized vessels. PAD in diabetic patients involves predominantly the medium-sized arteries and is due to the abnormal metabolic state that prevails in DM leading to vascular inflammation and alteration in several cell types [6]. The most important metabolic aberrations are chronic hyperglycemia, insulin resistance, dyslipidemia, and hypercoagulability, which render the arteries susceptible to atherosclerosis [6].

Hyperglycemia is a major risk factor in the pathogenesis of atherosclerosis in DM and accelerates the atherosclerotic process through several mechanisms. One of them is the non-enzymatic glycosylation of proteins and lipids that interfere with their normal function [6]. Glycosylation of low density lipoprotein (LDL) leads to functional alterations in LDL clearance and in increased susceptibility to oxidation which is strongly associated with the development of atherosclerosis. Glycosylated proteins interact with advanced glycosylated end products receptor (RAGE) on several cells such as endothelial cells, smooth muscle cells, and monocytes-derived macrophages resulting in increased oxidative stress, pro-inflammatory responses, and smooth muscle cell proliferation and migration. Elevated levels of glucose also promote protein kinase C (PKC) system that is involved in the transcription of various growth factors and in signal transduction in response to growth factors [6]. Furthermore, hyperglycemia increases intracellular concentration of glucose in platelets as its uptake is non-insulin dependent leading to decreased production of platelet-derived NO and excess production of oxygen free radicals [25]. Calcium homeostasis regulating platelet shape, secretion, aggregation, and thromboxane production also gets disturbed in DM [26]. Platelet expression of receptor proteins for von Willebrand factor and fibrin products is increased in DM, which could be the result of decreased production of the antiaggregants NO and prostacyclin, and increased production of fibrinogen and platelet activators such as thrombin and von Willebrand factor [27]. This increase in intrinsic platelet activity contributes to the state of enhanced thrombotic potential.

Moreover, patients with DM have impaired fibrinolytic activity [3]. Also, there are increased circulating levels of procoagulants such as tissue factor, factor VII, and decreased levels of anticoagulants such as antithrombin-III and Protein C, thus favoring a tendency to coagulation, impaired fibrinolysis, and persistence of thrombi [28].

DM is characterized by a chronic low- grade inflammatory state and patients with impaired glucose regulation have elevated levels of C-reactive protein (CRP) , which is strongly associated with PAD development [3, 4]. CRP has been shown to inhibit endothelial NO synthase and stimulate the production of procoagulant tissue factor, leukocyte adhesion molecules, chemotactic substances, and plasminogen activator inhibitor (PAI)-1 and thus contributing to a thrombotic environment and an abnormal regulation of the vascular tone [3, 4].

In brief, some of the cellular events in the process of atherosclerosis are as follows: DM augments the process of atheroma formation. The migration of T-lymphocytes into the intima, their activation, and secretion of cytokines is enhanced. Monocytes ingest oxidized LDL molecules on reaching the subendothelial space and become foam cells which lead to fatty streak formation, the precursors of the atheroma. The atheromatous plaque so formed is unstable as diabetic endothelial cells secrete cytokines which inhibit production of collagen by smooth muscle cells [25]. They also secrete metalloproteinases which break down the collagen in the fibrous cap of atheromas, leading to a tendency to plaque rupture and thrombus formation [29]. Endothelial cells produce increased amounts of tissue factor, a major procoagulant factor. Medial vascular smooth muscle cells migration into the intimal fatty streak lesion is enhanced. These cells then produce extracellular matrix, further aggravating atheroma formation.

CLI represents the end stage of PAD and is a state of arterial insufficiency, where lower limb perfusion is excessively reduced and nutrient supply and blood flow in the microcirculation is severely disturbed [1, 23]. The severity of CLI is related to the efficiency of collateral vessel formation as an adaptive response to ischemia. The development of new blood vessels from the pre-existing capillary network is called angiogenesis, while arteriogenesis is the formation of collateral vessels based on the growth of pre-existing arterioles and vasculogenesis is the de novo formation of new blood vessels by recruitment of precursor cell (endothelial progenitor cells, EPCs) from the bone marrow [30, 31]. Neovascularization has been described to be impaired in diabetic patients [30, 32, 33]. Thus, in patients with DM the formation of functional collateral circulation is decreased or even absent and they may present with CLI even in the absence of severe arterial obstruction of the large vessels [34].

The relative poor prognosis of CLI in patients with DM is most likely related to underlying multi-factorial pathophysiologic mechanisms, such as the association with peripheral and autonomic neuropathy and changes in microcirculation, the difficult to treat diffuse and distal distribution of arterial obstructive lesions, as well as the susceptibility to infection and the abnormal inflammation responses.

CLI is the end stage of chronic atherosclerotic disease of the lower limbs. Patients with CLI share the same traditional risk factors as patients with atherosclerosis in other vascular beds. The cardiovascular risk factors most strongly associated with CLI are DM, smoking, age > 65 years, dyslipidemia, and chronic renal disease [1, 3, 11]. Notably, diabetic patients with PAD have a fourfold increased risk in progressing to CLI in comparison with non-diabetic individuals with PAD [11].

As far as the increased predisposition to the development of PAD in diabetic patients, various risk factors have been described. Many studies have attempted to resolve this complicated issue by comparing diabetic patients with PAD to non-diabetic patients with PAD and between diabetic patients with and without PAD [10, 35–39].

Duration and degree of hyperglycemia is associated with an increased risk for PAD independently of other factors [4, 22, 35]. More specifically, every 1 % increase in HbA1c has been found to increase the risk of PAD by 28 % [4]. Increasing age also correlates strongly with PAD in patients with DM [36]. The Framingham Offspring Study reported that for each 10 years of age, the odds of PAD increased 2.6-fold [37]. Male gender has been found to correlate with increased PAD risk, whereas diabetic women have more often PAD compared to non-diabetic women of similar age [38]. The UK Prospective Diabetes Study (UKPDS) found that increased systolic blood pressure was an independent risk factor for PAD and each 10 mmHg increase in systolic blood pressure was associated with a 25 % increased risk for PAD [35].

The pattern of lower limb atherosclerotic obstructive lesions differs between diabetic and non-diabetic patients [8, 40]. Diabetic patients with CLI commonly show involvement of the arteries below the knee, especially at the tibial and peroneal arteries, unlike non-diabetic individuals who show mainly involvement of the iliac and femoral arteries [34, 41–43]. Moreover, in diabetic patients arterial obstructions are more commonly symmetrical and multi-segmental even in the collateral vessels. Non-diabetic patients with atherosclerosis usually present with single, unilateral, and proximal arterial stenosis [34, 42, 43]. Furthermore, calcification in patients with DM is diffuse, circumferential, and found in the tunica media of the arterial wall, while in patients without DM calcification is usually focal, eccentric, and found in the sub-intimal space [34]. The typical characteristics of diabetic macrovascular disease make revascularization procedures more technically challenging and demanding than in patients without DM.

Up to 75 % of diabetic patients with PAD are initially asymptomatic [22]. Eventually many patients develop symptomatic PAD that manifests as IC and progresses to CLI [7]. However, it should be emphasized that the natural history of PAD does not always follow a course of disease through standardized clinical stages. It has been reported that 5–10 % of patients with asymptomatic PAD or with IC will progress to CLI within 5 years, while 1–3 % of individuals with PAD are diagnosed with CLI at initial presentation [11]. Patients with DM are at even greater risk of PAD progression [7]. Notably, diabetic patients with decreased pain perception due to peripheral neuropathy may delay the recognition of PAD and may be diagnosed when revascularization is not an option [7, 21]. A number of studies have reported that 30–50 % of diabetic patients with foot ulcers have already gangrene at the first investigation [22].

CLI is characterized by the presence of ischemic rest pain, ischemic ulceration, and gangrene and is a limb-threatening condition. Ischemic rest pain of the lower limbs is described as burning pain mostly in the toes or the foot that most typically occurs at night when the patient is in bed and the extremities at the horizontal position; hanging the foot out of bed increases perfusion and usually relieves pain [1]. Notably, presence of rest pain depends on pain perception which is reduced or abolished in the presence of peripheral neuropathy [3]. It is important that all diabetic patients and especially those with foot ulcers are examined for the presence of peripheral neuropathy [44, 45]. The prevalence of peripheral neuropathy is 30–50 % among diabetic patients [46], while rest pain is absent in about 50 % of ischemic ulcer [34]. CLI may also present with loss of hair of the lower limbs, nail dystrophy, and loss of subcutaneous tissue leading to a shiny appearance of the skin and pallor at elevation of the feet [6].

However, every burning and aching sensation in the lower limb of diabetic patients should not be mistakenly diagnosed as CLI. Patients with DM often suffer from painful diabetic neuropathy and present with symptoms that are difficult to be distinguished from atypical ischemic rest pain. Almost 40–50 % of the patients with peripheral neuropathy have painful symptoms starting in the toes and extend proximally to involve the calves in a stocking distribution that usually worsen at night [47]. Unlike ischemic rest pain, patients with neuropathic pain do not find comfort by hanging the feet out of bed and they describe reduction in painful symptoms during walking.

Foot ulcers in diabetic patients are divided into three broad categories: neuropathic, ischemic, or neuro-ischemic [23]. Ischemic ulcers are usually seen in the toes and around the edges of the foot and have an irregular necrotic base, while neuropathic ulcers are found in the sole of the foot and are deep with surrounding callus formation [3]. It is important to bear in mind that in DM both peripheral neuropathy and PAD are major risk factors for foot ulceration. In a study by Moulik et al. in 185 diabetic patients with recent-onset full-thickness foot wounds, 45 % of the ulcers were neuropathic, 24 % ischemic, and 16 % neuro-ischemic [48]. Most studies suggest that PAD is present in up to 50 % of the diabetic patients with foot ulcer [21, 49, 50].

Furthermore, every foot ulcer should be examined for the presence of infection. Diagnosis of infection is based on local signs and symptoms of inflammation, such as warmth, redness, swelling, pain, bad odor, or purulent exudation [22]. Infected ischemic ulcers are a medical emergency and should be treated preferably within 24 h [50].

Summarizing, the diagnosis of CLI is mainly clinical based on the presence of rest pain, ulceration, and gangrene attributable to objectively proven PAD. However, patients with DM may commonly present with peripheral neuropathy and thus decreased pain perception and absence of ischemic rest pain. On the other hand, diabetic patients often present with neuropathic and infected diabetic foot ulcers even in the absence of PAD. Hence, especially in diabetic patients, the presence of ulceration and trophic changes of the lower limbs do not definitely imply that CLI is present. Nevertheless, all diabetic patients with foot ulcers should undergo vascular evaluation for the exclusion of PAD [22, 50].

The diagnosis of CLI is particularly challenging in patients with DM due to several reasons.

Palpation of peripheral pulses should be performed in all patients, although absence of pulses is unreliable for detection of PAD. The dorsalis pedis and tibialis posterior pulse may be absent congenitally in about 10 % of the population [5, 51]. Inspection of the foot may reveal findings of chronic ischemia in the periphery such as pallor on limb elevation and dependent rubor, ulceration, and distal gangrene that are all suggestive of CLI [5, 6].

Diagnosis criteria and definition of CLI have changed over the last decades. The TransAtlantic Inter-Society Consensus (TASC II) for the management of PAD in 2007 states that although the diagnosis of CLI is mainly clinical based on the presence of chronic rest pain, ulceration or gangrene attributable to proven PAD, it should be supported by objective tests [11]. The non-invasive hemodynamic tests most commonly used are measurement of ankle and toe systolic pressures and TcPO₂. For patients with ischemic rest pain in the lower limbs CLI is suggested when ankle systolic pressure is <50 mmHg and/or toe pressure <30 mmHg. For patients with ischemic ulceration or gangrene, CLI is suggested by ankle systolic pressures <70 mmHg and/or toe pressures <50 mmHg [11]. The former TASC for the management of PAD in 2000 included measurement of TcPO₂ as a diagnostic tool for the evaluation of CLI, with supine forefoot TcPO₂ values <30–50 mmHg in the presence of ischemic rest pain, ulceration, or gangrene being suggestive of CLI [52].

Diagnosis of CLI in diabetic patients based on hemodynamic tests is presented in Table 49.1.

ABI is a sensitive, specific, and reproducible screening tool for the assessment of PAD and the initial evaluation of CLI and should be measured in both lower limbs [9, 43]. ABI values <0.9 indicate presence of PAD, values <0.4 presence of CLI, while values >1.3 indicate presence of medial arterial calcification [3, 44]. ABI values <0.6 or ankle systolic pressures <50–80 mmHg are associated with poor healing in patients with foot ulcers [21, 22, 50].

The measurement of ankle pressures in patients with DM is sometimes impeded by the occlusion of tibial arteries, especially in patient with CLI, or by rigid or incompressible arteries, with pressures >250 mmHg [17, 21, 22, 51, 53, 54]. Interestingly, ankle and toe systolic pressures measurements may not be assessable in almost half of the patients with DM and CLI [17, 55, 56].

Medial arterial calcification (Mönckeberg’s sclerosis) of the lower limbs has been reported to be more frequent in patients with DM and end-stage renal disease and usually correlates with the presence of peripheral neuropathy [53, 54]. In one study up to 50 % of diabetic patients with CLI had calcified arteries of the lower limbs [18]. Presence of arterial calcification may lead to either inability to obtain the ankle systolic blood pressures or to artificially elevated values that may mask the presence of arterial insufficiency [34, 43, 53, 57]. Thus, an ankle systolic pressure ≥70 mmHg or an ABI value 0.9–1.3 should not rule out the presence of PAD and CLI in patients with DM and should be interpreted with caution, since ankle systolic pressures may be overestimated due to medial arterial calcification [22, 43, 51, 53, 55].

Notably, in one study arterial calcification, defined as ABI values >1.3 in CLI patients, was associated significantly with major amputation but not with increased mortality during follow-up [58]. Almost half of the study participants (47 %) had DM and ABI values >1.3 were associated with the presence of DM and renal insufficiency. The authors suggest that high ABI values should be considered an indicator of poor prognosis in patients with CLI.

Another study in CLI patients reported that although arterial calcification, assessed angiographically, was significantly associated with DM and hemodialysis , it was not associated with ankle systolic pressures and ABI [59]. The authors suggest that their finding could be explained by the presence of systemic calcification in patients with CLI and that the artificial elevation of ankle systolic pressure due to tibial calcification could be canceled by the artificial elevation of brachial systolic pressure due to calcified brachial arteries.

In patients where arterial calcification is present or suspected, measurement of great toe systolic blood pressures using a strain gauge sensor or a photoplethysmography is clearly recommended, since the vessels of the toes are not generally affected by media sclerosis [3, 44, 57]. However, a number of diabetic patients with CLI have toe ulcers or amputated toes that make the toe systolic pressures measurement impossible [22, 43, 55, 57]. Toe-brachial index (TBI ) values <0.7 indicate presence of PAD, while toe systolic pressures <30 mmHg are associated with poor healing in patients with foot ulcers [22, 50].

Determination of TcPO₂ may help assess CLI and local microcirculation and has a high prognostic value in predicting ulcer healing and amputations [1, 3, 21, 60] (Table 49.2). Notably, each 1 mmHg increase of TcPO₂ has been found to reduce the risk of above-the-ankle amputation by 10 % in diabetic patients with CLI [56]. TcPO₂ values <30 mmHg have been associated with poor ulcer healing, while TcPO₂ values >50 mmHg with fairly good prognosis [3, 21, 50]. TcPO₂ is a very useful and reliable test for the assessment of blood supply in tissues that can be used not only as a screening tool to select patients with CLI, but also to select the level of amputation and to evaluate the outcome of endovascular procedures [34, 61]. More specifically, after a successful revascularization procedure TcPO₂ increases and reaches a peak 4 weeks after angioplasty [62]. Presence of edema, cellulitis, hyperkeratosis, or abundant subcutaneous fat may alter TcPO₂ levels, but TcPO₂ measurement is feasible in almost all patients [43, 57].

Color duplex ultrasound is recommended as the first-line imaging modality and has good sensitivity for detecting high-grade stenosis in below-knee arteries both in diabetic and non-diabetic patients [54, 57, 63–65]. Moreover, color duplex imaging provides both hemodynamic and anatomical information of vascular lesions [51, 54, 57]. However, color duplex imaging is more difficult to perform, more time-consuming and needs extensive experience from the examiner in patients with DM [51, 64]. Newer diagnostic modalities, such as magnetic resonance angiography (MRA ) and computer tomography angiography (CTA ), are increasingly being used and should be offered to patients before considering revascularization [57, 63]. MRA is recommended to diagnose anatomical location and severity of stenosis in patients with PAD, as well as to select patients for revascularization procedures [9, 57, 63, 65]. When MRA is contraindicated, CTA can be performed. Disadvantages of CTA are the need of radiation exposure, the risk of contrast induced nephropathy, and the interference of arterial calcification with image quality [9, 57, 63, 65]. However, gadolinium use for MRA has been reported to cause nephrogenic systemic sclerosis in patients with impaired renal function [57]. It should be noted that the data about the efficiency of MRA and CTA in patients with CLI and DM are limited. Digital subtraction angiography (DSA) remains the gold standard against which all other imaging modalities should be compared and should be used when a vascular intervention is planned [3, 5, 57].

All non-invasive and invasive diagnostic modalities have their own strengths and weaknesses and none is satisfactory as a single tool for the diagnosis of CLI. Information gained from several diagnostic methods should be combined and interpreted together with the clinical presentation to make appropriate decisions for the management of each individual patient.

The primary goals of the management of diabetic patients with CLI are to relieve ischemic rest pain, heal foot ulceration, prevent amputation and cardiovascular morbidity and prolong survival [11]. Due to multi-factorial pathophysiologic mechanisms underlying diabetic CLI and the complex nature of treatment in such high-risk patients, a multidisciplinary approach is warranted and ultimately a revascularization procedure [21, 22, 50, 63, 66]. However, at present there is a lack of robust evidence to guide definite treatment strategies.

Lifestyle modification and medical treatment for cardiovascular risk factor control, as well as for pain relief are of great importance. Less well established is the role of medical management for treating symptoms of CLI and complications of lower limb hypo-perfusion. It should also be emphasized that diabetic patients with CLI usually present significant co-morbidities, such as coronary artery disease, cerebrovascular disease, carotid disease, and renal disease. Thus, screening of these patients for co-existing co-morbidities and appropriate management according to current guidelines is also important [9, 65, 67]. Appropriate foot care, sufficient footwear, debridement of foot ulcers, non-adherent dressings, and aggressive treatment of infections with antibiotics are essential for preventing amputation [3, 66, 68–70]. It is important to bear in mind that most of the following recommendations apply to patients with PAD in general and are extrapolated to the subgroup of patients with CLI.